Hip

ABSTRACT

Current literature regarding complications following total joint arthroplasty have primarily focused on patients with osteoarthritis (OA), with less emphasis on the trends and in-hospital outcomes of rheumatoid arthritis (RA) patients undergoing these procedures. The purpose of this study is to analyze the outcomes and trends of RA patients undergoing total knee arthroplasty (TKA) or total hip arthroplasty (THA) compared to OA patients.

Data from the Nationwide Inpatient Sample from 2006 to 2011 was extracted using the International Classification of Diseases, Ninth Revision codes for patients that received a TKA or THA. Outcome measures included cardiovascular complications, cerebrovascular complications, pulmonary complications, wound dehiscence, and infection. Inpatient and hospital demographics including primary diagnosis, age, gender, primary payer, hospital teaching status, Charlson Comorbidity Index score, hospital bed size, location, and median household income were analyzed.

Logistic regression analysis of OA vs RA patients with patient outcomes revealed that osteoarthritic THA candidates had lower risk for cardiovascular complications, pulmonary complications, wound dehiscence, infections, and systemic complications, compared to rheumatoid patients. There was a significantly elevated risk of cerebrovascular complication in osteoarthritic THA compared to RA THA. OA patients undergoing TKA had significantly higher risk for cardiovascular and cerebrovascular complications. There were significant decreases in mechanical wounds, infection, and systemic complications in the OA TKA patients.

RA patients are at higher risk for postoperative infection, wound dehiscence, and systemic complications after TKA and THA compared to OA patients. These findings highlight the importance of preoperative medical clearance and management to optimize RA patients and improve the postoperative outcomes.

Continue to: RA is a chronic systemic inflammatory disease...

Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease that causes joint deterioration, leading to pain, disability, systemic complications, short lifespan, and decline in quality of life.^1-3 The deterioration primarily affects the synovial membranes of joints, causing arthritis and resulting in extra-articular sequelae such as cardiovascular disease,⁴ pulmonary disease,⁵ and increased infection rates.^3,6 RA is the most prevalent inflammatory arthritis worldwide and affects up to 50 cases per 100,000 in both the US and northern Europe.^2,7-9 Although the gold standard of care for these patients is medical management with immunosuppressant drugs such as disease-modifying anti-rheumatic drugs (DMARDs), total joint arthroplasty (TJA) remains an important tool in the management of joint deterioration in such patients.

Total knee arthroplasty (TKA) and total hip arthroplasty (THA) are common procedures utilized to treat disorders that cause joint pain and hindered joint mobility, including osteoarthritis (OA) and RA. Given the aging population, the amount of TKAs and THAs performed in the US has consistently increased each year, with the vast majority of this increase composed of patients with OA.¹⁰ As a result, previous studies investigated the trends and outcomes of these procedures in patients with OA, but relatively less is known about the outcomes and trends of patients with RA undergoing the same surgeries.

Given that RA is a fundamentally different condition with its own pathological characteristics, an understanding of how these differences may impact postoperative outcomes in patients with RA is important. This study aims to present a comparative analysis of the trends and postoperative outcomes between patients with RA and OA undergoing TKA and THA (Figure 1, Tables 1 and 2).

Table 1. Demographics of Total Knee Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis.

Table 2. Demographics of Total Hip Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis or Rheumatoid Arthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis

Continue to: Methods...

METHODS

Exemptions were obtained from the Institutional Review Board. Data from the Nationwide Inpatient Sample (NIS) from 2006 to 2011 were extracted using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for patients that received primary TKA or THA, as well as their comorbid conditions. No patients or populations were excluded from the sampling process. A list of all independent variables collected for analysis and provision of relevant ICD-9 codes is included in Figure 1. The NIS is the largest all-payer stratified survey of inpatient care in the US healthcare system. As of 2011, each year provides information on approximately 8 million inpatient stays from about 1000 hospitals in 46 states. All discharges from sampled hospitals are also represented in the database. All patient information is protected, and all methods were conducted in accordance with the highest ethical standards of Human and Animal Rights Research.

STATISTICAL ANALYSIS

SAS 9.2 and PROC FREQ statistics software were used to generate P values (chi square result) and analyze the trends (Cochran-Armitage). Results were weighted utilizing standard discharge weights from the NIS to ensure accurate comparison of data from different time points. P < .05 was considered statistically significant. Multivariable logistic regression analyses were performed to generate odds ratio and 95% confidence limits to assess outcomes across different demographic variables.

RESULTS

Data on 337,082 and 1,362,241 patients undergoing THA or TKA, respectively, between 2006 and 2011 were analyzed. Patients in both groups were further differentiated by a diagnosis of either OA or RA. OA was the most common diagnosis, constituting 96.8% of all arthritic THA and TKA patients. From 2006 to 2011, a 36% and 34% increase in total number of THAs and TKAs, respectively, were reported. The number of patients with OA undergoing THA and TKA steadily increased from 2006 to 2011 (Figure 2). The number of THA and TKA procedures in patients with RA followed a similar trend but at a comparatively slower rate (Figure 3). The TKA geographical trends mirrored those observed with THA. The majority of operations were performed at urban hospitals (89% THA, 87% TKA; P < .0001). Among patients with RA and OA, the majority of TKAs (47.77%; P < .0001) took place in urban non-teaching hospitals than in urban teaching hospitals (39.26%). This pattern was not the same for THA, with 44.94% being performed at urban teaching hospitals and 44.05% at urban non-teaching institutions (P < .0001). Rural hospitals accounted for a low percentage of operations for both procedures: 10.46% of THA and 12.36% of TKA (P < .0001). Large institutions (based on the number of beds) claimed the majority of cases (59% of THA and TKA).

Logistic regression analysis and odds ratios of patients with OA vs those with RA with patient outcomes adjusted for age, Charlson Comorbidity Index (CCI) score, and gender revealed that patients with OA undergoing THA had lower risk for cardiovascular (0.674; confidence interval (CI) 0.587-0.774) and pulmonary complications (0.416; CI 0.384-0.450), wound dehiscence (0.647; CI 0.561-0.747), infections (0.258; CI 0.221-0.301), and systemic complications (0.625; CI 0.562-0.695) than patients with RA. Patients with OA exhibited statistically significantly higher odds of experiencing cerebrovascular complications after THA than those with RA (1.946; CI 1.673-2.236) (Table 3). In a similar logistic regression analysis of OA vs RA in TKA, which was adjusted for age, CCI score, and gender, patients with OA had significantly higher risk for cardiovascular (1.329; CI 1.069-1.651) and cerebrovascular complications (1.635; CI 1.375-1.943) than patients with RA. Significant decreases in wound dehiscence (0.757; CI 0.639-0.896), infection (0.331; CI 0.286-0.383), and systemic complication (0.641; CI 0.565-0.729) were noted in the patients with OA and TKA (Table 4).

Table 3. Odds Ratio for In-Hospital Complications Following THA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; THA, total hip arthroplasty.

Table 4. Odds Ratio for In-Hospital Complications Following TKA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; TKA, total knee arthroplasty.

Continue to: Discussion...

DISCUSSION

Our results showed a continuous yearly increase from 2006 to 2011 in THA and TKA procedures at a rate of 36% and 34%, respectively; this result was consistent with existing literature.¹¹ Despite a substantial increase in the amount of total THA and TKA procedures, the ratio of patients with RA undergoing these operations has decreased or remained nearly the same. Similar effects were found in Japan and the US when examining patients with RA undergoing TJA procedures between 2001 and 2007 and between 1992 and 2005, respectively.^12-14 This observation may be explained by the advances and early initiation of pharmacologic treatment and the widespread use of DMARDs such as methotrexate (MTX), azathioprine, leflunomide, hydroxychloroquine, and biological response modifiers TNF-α and interleukin-1.¹⁵ These medications have drastically improved survival rates of patients with RA with impressive capabilities in symptom relief.¹⁵ With the increasing use of DMARDs and aggressive treatment early on in the disease process, patients with RA are showing markedly slow progression of joint deterioration, leading to a decreased need for orthopedic intervention compared with the general population.^13,15

When analyzing the complication rates for patients undergoing TKA and THA, we observed that patients with RA exhibited a significant increase in the rates of infections, wound dehiscence, and systemic complications prior to discharge from the hospital compared with the OA population. The increased risk of infections was reported in previous studies assessing postoperative complication rates in TJA.^16,17 A study utilizing the Norwegian Arthroplasty Registry noted an increased risk of late infection in patients with RA, leading to increased rates of revision TJA in comparison with patients with OA.¹⁶ Another study, which was based on the Canadian Institute for Health Information Discharge Abstract Database, showed that patients with RA are at an increased risk of infection only after THA and interestingly not after TKA.¹⁷ Although our study did not identify the causes of the increased infection rate, the inherent nature of the disease and the immunomodulatory drugs used to treat it may contribute to this increased infectious risk in patients with RA.^6,18 Immunosuppressive DMARDs are some of the widely used medications employed to treat RA and are prime suspects of causing increased infection rates.¹⁵ The perioperative use of MTX has not been shown to cause short-term increases in infection for patients undergoing orthopedic intervention, but leflunomide and TNF-α inhibitors have been shown to cause a significant several-fold increase in risk for surgical wound infections.^19,20

All patients with RA presented with significant increases for infection, wound dehiscence, and systemic complications, whereas only patients with RA undergoing THA showed increased risk of pulmonary and cardiovascular complications when compared with patients with OA. Surprisingly, in TKA, patients with RA were at a significantly decreased risk of cardiovascular complications. This observation was interesting due to cardiovascular disease being one of RA's most notable extra-articular features.^4,21

Patients with RA undergoing TJA also showed significantly lower cerebrovascular complications than patients with OA. The significant reduction in risk for these complications has not been previously reported in the current literature, and it was an unexpected finding as past studies have found an increased risk in cerebrovascular disease in patients with RA. RA is an inflammatory disease exhibiting the upregulation of procoagulation factors,²² so we expected patients with RA to be at an increased risk for cerebrovascular and cardiovascular complications over patients with OA. Although we are unsure why these results were observed, we postulate that pharmaceutical interventions may confer some protection to patients with RA. For example, aspirin is commonly utilized in RA for its protective anti-platelet effect²³ and may be a contributing factor to why we found low postoperative complication rates in cerebrovascular disease. However, the reason why aspirin may be protective against cerebrovascular and not cardiovascular complications remains unclear. Moreover, most guidelines suggest that aspirin be stopped prior to surgery.²⁴ Although patients with RA were younger than those with OA, age was accounted for when analyzing the data.

A major strength of the study was the large sample size and the adjustment of potential confounding variables when examining the difference in complications between RA and OA. It is also a national US study that utilizes a validated database. Given that the patient samples in the NIS are reported in a uniform and de-identified manner, the database is considered ideal and has been extensively used for retrospective large observational cohort studies.²⁵ However, the study also had some limitations due to the retrospective and administrative nature of the NIS database. Only data concerning patient complications during their inpatient stay at the hospital were available. Patients who may develop complications following discharge were not included in the data, providing a very small window of time for analysis. Another limitation with the database was its lack of ability to identify the severity of each patient's disease process or the medical treatment they received perioperatively. Finally, no patient-reported outcomes were determined, which would provide information on whether these complications affect the patients’ postoperational satisfaction in regard to their pain and disability.

CONCLUSION

As RA patients continue to utilize joint arthroplasty to repair deteriorated joints, understanding of how the disease process and its medical management may impact patient outcomes is important. This article reports significantly higher postoperational infection rates in RA than in patients with OA, which may be due to the medical management of the disease. Although new medications have been introduced and are being used to treat patients with RA, they have not altered the complication rate following TJA in this patient population. Thus, surgeons and other members of the management team should be familiar with the common medical conditions, co-morbidities, and medical treatments/side effects that are encountered in patients with RA. Future studies should delve into possible differences in long-term outcomes of patients with RA undergoing TKA and THA, as well as whether certain perioperative strategies and therapies (medical or physical) may decrease complications and improve outcomes.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

Current literature regarding complications following total joint arthroplasty have primarily focused on patients with osteoarthritis (OA), with less emphasis on the trends and in-hospital outcomes of rheumatoid arthritis (RA) patients undergoing these procedures. The purpose of this study is to analyze the outcomes and trends of RA patients undergoing total knee arthroplasty (TKA) or total hip arthroplasty (THA) compared to OA patients.

Data from the Nationwide Inpatient Sample from 2006 to 2011 was extracted using the International Classification of Diseases, Ninth Revision codes for patients that received a TKA or THA. Outcome measures included cardiovascular complications, cerebrovascular complications, pulmonary complications, wound dehiscence, and infection. Inpatient and hospital demographics including primary diagnosis, age, gender, primary payer, hospital teaching status, Charlson Comorbidity Index score, hospital bed size, location, and median household income were analyzed.

Logistic regression analysis of OA vs RA patients with patient outcomes revealed that osteoarthritic THA candidates had lower risk for cardiovascular complications, pulmonary complications, wound dehiscence, infections, and systemic complications, compared to rheumatoid patients. There was a significantly elevated risk of cerebrovascular complication in osteoarthritic THA compared to RA THA. OA patients undergoing TKA had significantly higher risk for cardiovascular and cerebrovascular complications. There were significant decreases in mechanical wounds, infection, and systemic complications in the OA TKA patients.

RA patients are at higher risk for postoperative infection, wound dehiscence, and systemic complications after TKA and THA compared to OA patients. These findings highlight the importance of preoperative medical clearance and management to optimize RA patients and improve the postoperative outcomes.

Continue to: RA is a chronic systemic inflammatory disease...

Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease that causes joint deterioration, leading to pain, disability, systemic complications, short lifespan, and decline in quality of life.^1-3 The deterioration primarily affects the synovial membranes of joints, causing arthritis and resulting in extra-articular sequelae such as cardiovascular disease,⁴ pulmonary disease,⁵ and increased infection rates.^3,6 RA is the most prevalent inflammatory arthritis worldwide and affects up to 50 cases per 100,000 in both the US and northern Europe.^2,7-9 Although the gold standard of care for these patients is medical management with immunosuppressant drugs such as disease-modifying anti-rheumatic drugs (DMARDs), total joint arthroplasty (TJA) remains an important tool in the management of joint deterioration in such patients.

Total knee arthroplasty (TKA) and total hip arthroplasty (THA) are common procedures utilized to treat disorders that cause joint pain and hindered joint mobility, including osteoarthritis (OA) and RA. Given the aging population, the amount of TKAs and THAs performed in the US has consistently increased each year, with the vast majority of this increase composed of patients with OA.¹⁰ As a result, previous studies investigated the trends and outcomes of these procedures in patients with OA, but relatively less is known about the outcomes and trends of patients with RA undergoing the same surgeries.

Given that RA is a fundamentally different condition with its own pathological characteristics, an understanding of how these differences may impact postoperative outcomes in patients with RA is important. This study aims to present a comparative analysis of the trends and postoperative outcomes between patients with RA and OA undergoing TKA and THA (Figure 1, Tables 1 and 2).

Table 1. Demographics of Total Knee Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis.

Table 2. Demographics of Total Hip Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis or Rheumatoid Arthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis

Continue to: Methods...

METHODS

Exemptions were obtained from the Institutional Review Board. Data from the Nationwide Inpatient Sample (NIS) from 2006 to 2011 were extracted using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for patients that received primary TKA or THA, as well as their comorbid conditions. No patients or populations were excluded from the sampling process. A list of all independent variables collected for analysis and provision of relevant ICD-9 codes is included in Figure 1. The NIS is the largest all-payer stratified survey of inpatient care in the US healthcare system. As of 2011, each year provides information on approximately 8 million inpatient stays from about 1000 hospitals in 46 states. All discharges from sampled hospitals are also represented in the database. All patient information is protected, and all methods were conducted in accordance with the highest ethical standards of Human and Animal Rights Research.

STATISTICAL ANALYSIS

SAS 9.2 and PROC FREQ statistics software were used to generate P values (chi square result) and analyze the trends (Cochran-Armitage). Results were weighted utilizing standard discharge weights from the NIS to ensure accurate comparison of data from different time points. P < .05 was considered statistically significant. Multivariable logistic regression analyses were performed to generate odds ratio and 95% confidence limits to assess outcomes across different demographic variables.

RESULTS

Data on 337,082 and 1,362,241 patients undergoing THA or TKA, respectively, between 2006 and 2011 were analyzed. Patients in both groups were further differentiated by a diagnosis of either OA or RA. OA was the most common diagnosis, constituting 96.8% of all arthritic THA and TKA patients. From 2006 to 2011, a 36% and 34% increase in total number of THAs and TKAs, respectively, were reported. The number of patients with OA undergoing THA and TKA steadily increased from 2006 to 2011 (Figure 2). The number of THA and TKA procedures in patients with RA followed a similar trend but at a comparatively slower rate (Figure 3). The TKA geographical trends mirrored those observed with THA. The majority of operations were performed at urban hospitals (89% THA, 87% TKA; P < .0001). Among patients with RA and OA, the majority of TKAs (47.77%; P < .0001) took place in urban non-teaching hospitals than in urban teaching hospitals (39.26%). This pattern was not the same for THA, with 44.94% being performed at urban teaching hospitals and 44.05% at urban non-teaching institutions (P < .0001). Rural hospitals accounted for a low percentage of operations for both procedures: 10.46% of THA and 12.36% of TKA (P < .0001). Large institutions (based on the number of beds) claimed the majority of cases (59% of THA and TKA).

Logistic regression analysis and odds ratios of patients with OA vs those with RA with patient outcomes adjusted for age, Charlson Comorbidity Index (CCI) score, and gender revealed that patients with OA undergoing THA had lower risk for cardiovascular (0.674; confidence interval (CI) 0.587-0.774) and pulmonary complications (0.416; CI 0.384-0.450), wound dehiscence (0.647; CI 0.561-0.747), infections (0.258; CI 0.221-0.301), and systemic complications (0.625; CI 0.562-0.695) than patients with RA. Patients with OA exhibited statistically significantly higher odds of experiencing cerebrovascular complications after THA than those with RA (1.946; CI 1.673-2.236) (Table 3). In a similar logistic regression analysis of OA vs RA in TKA, which was adjusted for age, CCI score, and gender, patients with OA had significantly higher risk for cardiovascular (1.329; CI 1.069-1.651) and cerebrovascular complications (1.635; CI 1.375-1.943) than patients with RA. Significant decreases in wound dehiscence (0.757; CI 0.639-0.896), infection (0.331; CI 0.286-0.383), and systemic complication (0.641; CI 0.565-0.729) were noted in the patients with OA and TKA (Table 4).

Table 3. Odds Ratio for In-Hospital Complications Following THA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; THA, total hip arthroplasty.

Table 4. Odds Ratio for In-Hospital Complications Following TKA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; TKA, total knee arthroplasty.

Continue to: Discussion...

DISCUSSION

Our results showed a continuous yearly increase from 2006 to 2011 in THA and TKA procedures at a rate of 36% and 34%, respectively; this result was consistent with existing literature.¹¹ Despite a substantial increase in the amount of total THA and TKA procedures, the ratio of patients with RA undergoing these operations has decreased or remained nearly the same. Similar effects were found in Japan and the US when examining patients with RA undergoing TJA procedures between 2001 and 2007 and between 1992 and 2005, respectively.^12-14 This observation may be explained by the advances and early initiation of pharmacologic treatment and the widespread use of DMARDs such as methotrexate (MTX), azathioprine, leflunomide, hydroxychloroquine, and biological response modifiers TNF-α and interleukin-1.¹⁵ These medications have drastically improved survival rates of patients with RA with impressive capabilities in symptom relief.¹⁵ With the increasing use of DMARDs and aggressive treatment early on in the disease process, patients with RA are showing markedly slow progression of joint deterioration, leading to a decreased need for orthopedic intervention compared with the general population.^13,15

When analyzing the complication rates for patients undergoing TKA and THA, we observed that patients with RA exhibited a significant increase in the rates of infections, wound dehiscence, and systemic complications prior to discharge from the hospital compared with the OA population. The increased risk of infections was reported in previous studies assessing postoperative complication rates in TJA.^16,17 A study utilizing the Norwegian Arthroplasty Registry noted an increased risk of late infection in patients with RA, leading to increased rates of revision TJA in comparison with patients with OA.¹⁶ Another study, which was based on the Canadian Institute for Health Information Discharge Abstract Database, showed that patients with RA are at an increased risk of infection only after THA and interestingly not after TKA.¹⁷ Although our study did not identify the causes of the increased infection rate, the inherent nature of the disease and the immunomodulatory drugs used to treat it may contribute to this increased infectious risk in patients with RA.^6,18 Immunosuppressive DMARDs are some of the widely used medications employed to treat RA and are prime suspects of causing increased infection rates.¹⁵ The perioperative use of MTX has not been shown to cause short-term increases in infection for patients undergoing orthopedic intervention, but leflunomide and TNF-α inhibitors have been shown to cause a significant several-fold increase in risk for surgical wound infections.^19,20

All patients with RA presented with significant increases for infection, wound dehiscence, and systemic complications, whereas only patients with RA undergoing THA showed increased risk of pulmonary and cardiovascular complications when compared with patients with OA. Surprisingly, in TKA, patients with RA were at a significantly decreased risk of cardiovascular complications. This observation was interesting due to cardiovascular disease being one of RA's most notable extra-articular features.^4,21

Patients with RA undergoing TJA also showed significantly lower cerebrovascular complications than patients with OA. The significant reduction in risk for these complications has not been previously reported in the current literature, and it was an unexpected finding as past studies have found an increased risk in cerebrovascular disease in patients with RA. RA is an inflammatory disease exhibiting the upregulation of procoagulation factors,²² so we expected patients with RA to be at an increased risk for cerebrovascular and cardiovascular complications over patients with OA. Although we are unsure why these results were observed, we postulate that pharmaceutical interventions may confer some protection to patients with RA. For example, aspirin is commonly utilized in RA for its protective anti-platelet effect²³ and may be a contributing factor to why we found low postoperative complication rates in cerebrovascular disease. However, the reason why aspirin may be protective against cerebrovascular and not cardiovascular complications remains unclear. Moreover, most guidelines suggest that aspirin be stopped prior to surgery.²⁴ Although patients with RA were younger than those with OA, age was accounted for when analyzing the data.

A major strength of the study was the large sample size and the adjustment of potential confounding variables when examining the difference in complications between RA and OA. It is also a national US study that utilizes a validated database. Given that the patient samples in the NIS are reported in a uniform and de-identified manner, the database is considered ideal and has been extensively used for retrospective large observational cohort studies.²⁵ However, the study also had some limitations due to the retrospective and administrative nature of the NIS database. Only data concerning patient complications during their inpatient stay at the hospital were available. Patients who may develop complications following discharge were not included in the data, providing a very small window of time for analysis. Another limitation with the database was its lack of ability to identify the severity of each patient's disease process or the medical treatment they received perioperatively. Finally, no patient-reported outcomes were determined, which would provide information on whether these complications affect the patients’ postoperational satisfaction in regard to their pain and disability.

CONCLUSION

As RA patients continue to utilize joint arthroplasty to repair deteriorated joints, understanding of how the disease process and its medical management may impact patient outcomes is important. This article reports significantly higher postoperational infection rates in RA than in patients with OA, which may be due to the medical management of the disease. Although new medications have been introduced and are being used to treat patients with RA, they have not altered the complication rate following TJA in this patient population. Thus, surgeons and other members of the management team should be familiar with the common medical conditions, co-morbidities, and medical treatments/side effects that are encountered in patients with RA. Future studies should delve into possible differences in long-term outcomes of patients with RA undergoing TKA and THA, as well as whether certain perioperative strategies and therapies (medical or physical) may decrease complications and improve outcomes.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

Current literature regarding complications following total joint arthroplasty have primarily focused on patients with osteoarthritis (OA), with less emphasis on the trends and in-hospital outcomes of rheumatoid arthritis (RA) patients undergoing these procedures. The purpose of this study is to analyze the outcomes and trends of RA patients undergoing total knee arthroplasty (TKA) or total hip arthroplasty (THA) compared to OA patients.

Data from the Nationwide Inpatient Sample from 2006 to 2011 was extracted using the International Classification of Diseases, Ninth Revision codes for patients that received a TKA or THA. Outcome measures included cardiovascular complications, cerebrovascular complications, pulmonary complications, wound dehiscence, and infection. Inpatient and hospital demographics including primary diagnosis, age, gender, primary payer, hospital teaching status, Charlson Comorbidity Index score, hospital bed size, location, and median household income were analyzed.

Logistic regression analysis of OA vs RA patients with patient outcomes revealed that osteoarthritic THA candidates had lower risk for cardiovascular complications, pulmonary complications, wound dehiscence, infections, and systemic complications, compared to rheumatoid patients. There was a significantly elevated risk of cerebrovascular complication in osteoarthritic THA compared to RA THA. OA patients undergoing TKA had significantly higher risk for cardiovascular and cerebrovascular complications. There were significant decreases in mechanical wounds, infection, and systemic complications in the OA TKA patients.

RA patients are at higher risk for postoperative infection, wound dehiscence, and systemic complications after TKA and THA compared to OA patients. These findings highlight the importance of preoperative medical clearance and management to optimize RA patients and improve the postoperative outcomes.

Continue to: RA is a chronic systemic inflammatory disease...

Rheumatoid arthritis (RA) is a chronic systemic inflammatory disease that causes joint deterioration, leading to pain, disability, systemic complications, short lifespan, and decline in quality of life.^1-3 The deterioration primarily affects the synovial membranes of joints, causing arthritis and resulting in extra-articular sequelae such as cardiovascular disease,⁴ pulmonary disease,⁵ and increased infection rates.^3,6 RA is the most prevalent inflammatory arthritis worldwide and affects up to 50 cases per 100,000 in both the US and northern Europe.^2,7-9 Although the gold standard of care for these patients is medical management with immunosuppressant drugs such as disease-modifying anti-rheumatic drugs (DMARDs), total joint arthroplasty (TJA) remains an important tool in the management of joint deterioration in such patients.

Total knee arthroplasty (TKA) and total hip arthroplasty (THA) are common procedures utilized to treat disorders that cause joint pain and hindered joint mobility, including osteoarthritis (OA) and RA. Given the aging population, the amount of TKAs and THAs performed in the US has consistently increased each year, with the vast majority of this increase composed of patients with OA.¹⁰ As a result, previous studies investigated the trends and outcomes of these procedures in patients with OA, but relatively less is known about the outcomes and trends of patients with RA undergoing the same surgeries.

Given that RA is a fundamentally different condition with its own pathological characteristics, an understanding of how these differences may impact postoperative outcomes in patients with RA is important. This study aims to present a comparative analysis of the trends and postoperative outcomes between patients with RA and OA undergoing TKA and THA (Figure 1, Tables 1 and 2).

Table 1. Demographics of Total Knee Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis.

Table 2. Demographics of Total Hip Arthroplasty Patients Based on Primary Diagnosis of Osteoarthritis or Rheumatoid Arthritis

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis

Continue to: Methods...

METHODS

Exemptions were obtained from the Institutional Review Board. Data from the Nationwide Inpatient Sample (NIS) from 2006 to 2011 were extracted using the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes for patients that received primary TKA or THA, as well as their comorbid conditions. No patients or populations were excluded from the sampling process. A list of all independent variables collected for analysis and provision of relevant ICD-9 codes is included in Figure 1. The NIS is the largest all-payer stratified survey of inpatient care in the US healthcare system. As of 2011, each year provides information on approximately 8 million inpatient stays from about 1000 hospitals in 46 states. All discharges from sampled hospitals are also represented in the database. All patient information is protected, and all methods were conducted in accordance with the highest ethical standards of Human and Animal Rights Research.

STATISTICAL ANALYSIS

SAS 9.2 and PROC FREQ statistics software were used to generate P values (chi square result) and analyze the trends (Cochran-Armitage). Results were weighted utilizing standard discharge weights from the NIS to ensure accurate comparison of data from different time points. P < .05 was considered statistically significant. Multivariable logistic regression analyses were performed to generate odds ratio and 95% confidence limits to assess outcomes across different demographic variables.

RESULTS

Data on 337,082 and 1,362,241 patients undergoing THA or TKA, respectively, between 2006 and 2011 were analyzed. Patients in both groups were further differentiated by a diagnosis of either OA or RA. OA was the most common diagnosis, constituting 96.8% of all arthritic THA and TKA patients. From 2006 to 2011, a 36% and 34% increase in total number of THAs and TKAs, respectively, were reported. The number of patients with OA undergoing THA and TKA steadily increased from 2006 to 2011 (Figure 2). The number of THA and TKA procedures in patients with RA followed a similar trend but at a comparatively slower rate (Figure 3). The TKA geographical trends mirrored those observed with THA. The majority of operations were performed at urban hospitals (89% THA, 87% TKA; P < .0001). Among patients with RA and OA, the majority of TKAs (47.77%; P < .0001) took place in urban non-teaching hospitals than in urban teaching hospitals (39.26%). This pattern was not the same for THA, with 44.94% being performed at urban teaching hospitals and 44.05% at urban non-teaching institutions (P < .0001). Rural hospitals accounted for a low percentage of operations for both procedures: 10.46% of THA and 12.36% of TKA (P < .0001). Large institutions (based on the number of beds) claimed the majority of cases (59% of THA and TKA).

Logistic regression analysis and odds ratios of patients with OA vs those with RA with patient outcomes adjusted for age, Charlson Comorbidity Index (CCI) score, and gender revealed that patients with OA undergoing THA had lower risk for cardiovascular (0.674; confidence interval (CI) 0.587-0.774) and pulmonary complications (0.416; CI 0.384-0.450), wound dehiscence (0.647; CI 0.561-0.747), infections (0.258; CI 0.221-0.301), and systemic complications (0.625; CI 0.562-0.695) than patients with RA. Patients with OA exhibited statistically significantly higher odds of experiencing cerebrovascular complications after THA than those with RA (1.946; CI 1.673-2.236) (Table 3). In a similar logistic regression analysis of OA vs RA in TKA, which was adjusted for age, CCI score, and gender, patients with OA had significantly higher risk for cardiovascular (1.329; CI 1.069-1.651) and cerebrovascular complications (1.635; CI 1.375-1.943) than patients with RA. Significant decreases in wound dehiscence (0.757; CI 0.639-0.896), infection (0.331; CI 0.286-0.383), and systemic complication (0.641; CI 0.565-0.729) were noted in the patients with OA and TKA (Table 4).

Table 3. Odds Ratio for In-Hospital Complications Following THA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; THA, total hip arthroplasty.

Table 4. Odds Ratio for In-Hospital Complications Following TKA for OA Patients vs RA Patients

Abbreviations: OA, osteoarthritis; RA, rheumatoid arthritis; TKA, total knee arthroplasty.

Continue to: Discussion...

DISCUSSION

Our results showed a continuous yearly increase from 2006 to 2011 in THA and TKA procedures at a rate of 36% and 34%, respectively; this result was consistent with existing literature.¹¹ Despite a substantial increase in the amount of total THA and TKA procedures, the ratio of patients with RA undergoing these operations has decreased or remained nearly the same. Similar effects were found in Japan and the US when examining patients with RA undergoing TJA procedures between 2001 and 2007 and between 1992 and 2005, respectively.^12-14 This observation may be explained by the advances and early initiation of pharmacologic treatment and the widespread use of DMARDs such as methotrexate (MTX), azathioprine, leflunomide, hydroxychloroquine, and biological response modifiers TNF-α and interleukin-1.¹⁵ These medications have drastically improved survival rates of patients with RA with impressive capabilities in symptom relief.¹⁵ With the increasing use of DMARDs and aggressive treatment early on in the disease process, patients with RA are showing markedly slow progression of joint deterioration, leading to a decreased need for orthopedic intervention compared with the general population.^13,15

When analyzing the complication rates for patients undergoing TKA and THA, we observed that patients with RA exhibited a significant increase in the rates of infections, wound dehiscence, and systemic complications prior to discharge from the hospital compared with the OA population. The increased risk of infections was reported in previous studies assessing postoperative complication rates in TJA.^16,17 A study utilizing the Norwegian Arthroplasty Registry noted an increased risk of late infection in patients with RA, leading to increased rates of revision TJA in comparison with patients with OA.¹⁶ Another study, which was based on the Canadian Institute for Health Information Discharge Abstract Database, showed that patients with RA are at an increased risk of infection only after THA and interestingly not after TKA.¹⁷ Although our study did not identify the causes of the increased infection rate, the inherent nature of the disease and the immunomodulatory drugs used to treat it may contribute to this increased infectious risk in patients with RA.^6,18 Immunosuppressive DMARDs are some of the widely used medications employed to treat RA and are prime suspects of causing increased infection rates.¹⁵ The perioperative use of MTX has not been shown to cause short-term increases in infection for patients undergoing orthopedic intervention, but leflunomide and TNF-α inhibitors have been shown to cause a significant several-fold increase in risk for surgical wound infections.^19,20

All patients with RA presented with significant increases for infection, wound dehiscence, and systemic complications, whereas only patients with RA undergoing THA showed increased risk of pulmonary and cardiovascular complications when compared with patients with OA. Surprisingly, in TKA, patients with RA were at a significantly decreased risk of cardiovascular complications. This observation was interesting due to cardiovascular disease being one of RA's most notable extra-articular features.^4,21

Patients with RA undergoing TJA also showed significantly lower cerebrovascular complications than patients with OA. The significant reduction in risk for these complications has not been previously reported in the current literature, and it was an unexpected finding as past studies have found an increased risk in cerebrovascular disease in patients with RA. RA is an inflammatory disease exhibiting the upregulation of procoagulation factors,²² so we expected patients with RA to be at an increased risk for cerebrovascular and cardiovascular complications over patients with OA. Although we are unsure why these results were observed, we postulate that pharmaceutical interventions may confer some protection to patients with RA. For example, aspirin is commonly utilized in RA for its protective anti-platelet effect²³ and may be a contributing factor to why we found low postoperative complication rates in cerebrovascular disease. However, the reason why aspirin may be protective against cerebrovascular and not cardiovascular complications remains unclear. Moreover, most guidelines suggest that aspirin be stopped prior to surgery.²⁴ Although patients with RA were younger than those with OA, age was accounted for when analyzing the data.

A major strength of the study was the large sample size and the adjustment of potential confounding variables when examining the difference in complications between RA and OA. It is also a national US study that utilizes a validated database. Given that the patient samples in the NIS are reported in a uniform and de-identified manner, the database is considered ideal and has been extensively used for retrospective large observational cohort studies.²⁵ However, the study also had some limitations due to the retrospective and administrative nature of the NIS database. Only data concerning patient complications during their inpatient stay at the hospital were available. Patients who may develop complications following discharge were not included in the data, providing a very small window of time for analysis. Another limitation with the database was its lack of ability to identify the severity of each patient's disease process or the medical treatment they received perioperatively. Finally, no patient-reported outcomes were determined, which would provide information on whether these complications affect the patients’ postoperational satisfaction in regard to their pain and disability.

CONCLUSION

As RA patients continue to utilize joint arthroplasty to repair deteriorated joints, understanding of how the disease process and its medical management may impact patient outcomes is important. This article reports significantly higher postoperational infection rates in RA than in patients with OA, which may be due to the medical management of the disease. Although new medications have been introduced and are being used to treat patients with RA, they have not altered the complication rate following TJA in this patient population. Thus, surgeons and other members of the management team should be familiar with the common medical conditions, co-morbidities, and medical treatments/side effects that are encountered in patients with RA. Future studies should delve into possible differences in long-term outcomes of patients with RA undergoing TKA and THA, as well as whether certain perioperative strategies and therapies (medical or physical) may decrease complications and improve outcomes.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

Pain control following hip arthroscopy presents a significant clinical challenge, with postoperative pain requiring considerable opioid use. Peripheral nerve blocks (PNBs) have emerged as one option to improve pain and limit the consequences of opioid use. The purpose of this study is to provide a comprehensive review of outcomes associated with PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications. A systematic review of PubMed, Medline, Scopus, and Embase databases was conducted through January 2015 for English-language articles reporting outcome data, with 2 reviewers independently reviewing studies for inclusion. When available, similar outcomes were combined to generate frequency-weighted means. Six studies met the inclusion criteria for this review, reporting on 710 patients undergoing hip arthroscopy. The mean ages were 37.0 and 37.7 years for the PNB and comparator groups, respectively, with a reported total of 281 (40.5%) male and 412 (59.5%) female patients. Postoperative post-anesthesia care unit (PACU) pain was consistently reduced in the PNB group, with the use of a lower morphine equivalent dose and lower rates of inpatient admission, compared with that in the control groups. Postoperative nausea and/or vomiting as well as PACU discharge time showed mixed results. High satisfaction and few complications were reported. In conclusion, PNB is associated with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions, though similar rates of nausea/vomiting and time to discharge were reported. Current PNB techniques are varied, and future research efforts should focus on examining which of these methods provides the optimal risk-benefit profile in hip arthroscopy.

Continue to: Hip arthroscopy has emerged...

Hip arthroscopy has emerged as a useful procedure in the diagnosis and treatment of hip pathology,^1-8 experiencing a substantial rise in popularity in recent years, with the number of procedures growing by a factor of 18 from 1999 to 2009⁹ and 25 from 2006 to 2013.¹⁰ Though hip arthroscopy is beneficial in many cases, marked postoperative pain has presented a substantial challenge, with patients requiring considerable doses of opiate-based medications in the post-anesthesia care unit (PACU).^11,12 Increased narcotic use carries increased side effects, including postoperative nausea and vomiting,¹³ and poorly managed pain leads to increased unplanned admissions.¹⁴ Furthermore, patients with chronic hip pain and long-term opioid use may experience heightened and prolonged pain following the procedure, owing to medication tolerance and reduced opioid efficacy in this setting.¹⁵

Several pain control strategies have been employed in patients undergoing hip arthroscopy. General anesthesia^16,17 and combined spinal epidural (CSE)¹⁸ are commonly used. However, such techniques rely heavily on opioids for postoperative pain control,¹¹ and epidural anesthesia commonly requires adjunctive treatments (eg, neuromuscular blockade) to ensure muscle relaxation for joint distraction.¹⁹ One technique that has been employed recently is peripheral nerve block (PNB), which has been associated with a significant decrease in postoperative opioid use and nausea and vomiting.^13,20 This method has proven successful in other fields of arthroscopy, including shoulder arthroscopy, in which it resulted in faster recovery, reduced opioid consumption,²¹ and demonstrated cost-effectiveness²² compared with general anesthesia and knee arthroscopy.^23-26 As it is a relatively new field, little is known about the use of PNB in hip arthroscopy.

The goal of this systematic review was to comprehensively review the studies reporting on PNB in hip arthroscopy. We specifically focused on outcomes, including postoperative pain; analgesic use; nausea, vomiting, and antiemetic use; discharge time; inpatient admission; and patient satisfaction, as well as the complications associated with the use of PNB. Our knowledge of outcomes associated with PNB in hip arthroscopy is based on a few individual studies that have reported on small groups of patients using a variety of outcome measures and other findings. Furthermore, each of these studies commonly reflects the experience of an individual surgeon at a single institution and, when taken alone, may not be an accurate representation of the more general outcomes associated with PNB. A comprehensive review of such studies will provide surgeons, anesthesiologists, and patients with a better understanding of the anticipated outcomes of using PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications.

MATERIALS AND METHODS

A systematic review of outcomes associated with PNB in hip arthroscopy was performed using the available English-language literature in accordance with the guidelines laid out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and included studies retrieved from the PubMed, Medline, Scopus, and Embase computerized literature databases. Searches were executed comprising all years from database inception through January 2015. Articles were retrieved by an electronic search of medical subject headings and keyword terms and their respective combinations (Table 1). The inclusion criteria for studies in this systematic review were studies that (1) were written in the English language and (2) reported explicit outcome data. The exclusion criteria were (1) review articles, meta-analyses, case reports, abstracts/conference papers, comments/letters, or technique articles without reported patient data and (2) basic research, biomechanics, or animal/cadaveric studies without reported patient data.

Table 1. Search Terms Entered to Identify English-Language Studies Through January 2015

The literature search strategy is outlined in the Figure. The initial title search yielded a subset of possible articles that were then further included or excluded on the basis of the contents of the article’s abstract, wherein articles were again selected on the basis of the aforementioned inclusion and exclusion criteria. Articles selected in both the title and abstract phases underwent full-text review, during which the full text of each qualifying article was reviewed. In addition, the reference sections from articles undergoing full-text review were scanned to identify any additional studies that had not been identified in the original literature search. Appropriate studies for final inclusion were then selected at this stage. The title, abstract, and full-text selection process were performed by 2 of the study authors (Dr. Steinhaus and Dr. Lynch), with any discrepancies being discussed and resolved by mutual agreement.

Continue to: For all 6 included studies...

For all 6 included studies,^16-18,27-29 data were collected regarding the study specifics, patients included, and outcomes measured in the study. The journal of publication, type of study, level of evidence, and type of PNB, as well as the presence of a comparator group were noted (Table 2). Patient information included the number of patients at baseline and follow-up, mean age, gender, weight, height, body mass index, American Society of Anesthesiologists (ASA) status, and the specific procedures performed. In addition, data were collected on outcomes, including postoperative pain, as well as secondary outcomes and additional findings reported by the studies (Table 3). Where possible, weighted averages were calculated across all studies to obtain aggregate data.

RESULTS

STUDY INCLUSION

Six studies, all published between 2012 and 2014, were included in this systematic review (Table 2). Three studies involved lumbar plexus block, 2 studies involved femoral nerve block, and 1 study evaluated fascia iliaca block. Two studies used a control group of patients who received only general anesthesia (compared with the treatment group who received both general anesthesia and PNB); another study compared intravenous morphine with PNB; and 1 study compared CSE alone with PNB in addition to epidural.

DEMOGRAPHIC DATA

Demographic data from the included studies are presented in Table 2. In total, 710 and 549 patients were evaluated at baseline and final follow-up, respectively, which represents a follow-up rate of 77%. The frequency-weighted mean age of patients receiving PNB was 37.0 years compared with 37.7 years in the comparison groups, and the studies reported a total of 281 (40.5%) male and 412 (59.5%) female patients. The procedures performed were heterogeneously reported; therefore, totals were not tabulated, although the reported procedures included osteochondroplasty, labral débridement, labral and/or capsular repair, gluteus minimus repair, and synovectomy.

POSTOPERATIVE PAIN

Four studies reported on postoperative pain, and these data are presented in Table 3. In a retrospective study of patients receiving femoral nerve block in addition to general anesthesia, Dold and colleagues¹⁶ noted postoperative pain at 0, 15, 30, 45, and 60 minutes following arrival in the PACU, and discovered a statistically significantly lower level of pain at 60 minutes compared with inpatients receiving general anesthesia alone. YaDeau and colleagues¹⁸ found a significantly lower level of pain at rest in the PACU for those receiving CSE and lumbar plexus blockade compared with those receiving CSE alone. This significant difference did not persist at 24 hours or 6 months after the procedure, nor did it exist for pain with movement at any time point. Similarly, Schroeder and colleagues¹⁷ examined patients receiving general anesthesia and lumbar plexus block and found a significant reduction in pain immediately postoperatively in the PACU, though these effects disappeared the day following the procedure. Krych and colleagues²⁷ also reported on postoperative pain in patients undergoing fascia iliaca blockade, although they did not include a comparator group. Outcome comparison between patients who received PNB and controls in the PACU and 1 day following the procedure are presented in Table 4.

ANALGESIC USE

Four studies reported on analgesic use after PNB, and these data are presented in Table 3. Dold and colleagues¹⁶ noted analgesic use intraoperatively, in the PACU, and in the surgical day care unit (SDCU). These authors found a significant reduction in morphine equivalent dose given in the operating room and in the PACU in the group receiving PNB, with a nonsignificant trend toward lower use of oxycodone in the SDCU. Schroeder and colleagues¹⁷ similarly reported significant reductions in morphine equivalent dose intraoperatively and in Phase I recovery for patients receiving PNB, and these differences disappeared in Phase II recovery as well as intraoperatively if the block dose was considered. In addition, these authors found a significant reduction in the use of fentanyl and hydromorphone in the operating room in the PNB group, as well as a significant reduction in the proportion of patients receiving ketorolac in the operating room or PACU. Finally, YaDeau and colleagues¹⁸ reported total analgesic usage in the PACU among PNB patients compared with those receiving CSE alone and showed a strong trend toward reduced use in the PNB group, although this difference was not significant (P = .051). PACU analgesic use is presented in Table 4.

Continue to: Postoperative nausea...

POSTOPERATIVE NAUSEA/VOMITING AND ANTIEMETIC USE

Five studies presented data on nausea, vomiting, or antiemetic use following PNB and are shown in Table 3. YaDeau and colleagues¹⁸ reported nausea among 34% of patients in the PNB group, compared with 20% in the control group, vomiting in 2% and 7%, respectively, and antiemetic use in 12% of both groups. Dold and colleagues¹⁶ identified a similar trend, with 41.1% of patients in the PNB group and 32.5% of patients in the control group experiencing postoperative nausea or vomiting, while Krych and colleagues²⁷ noted only 10% of PNB patients with mild nausea and none requiring antiemetic use. In their study of patients receiving PNB, Schroeder and colleagues¹⁷ found a significant reduction in antiemetic use among PNB patients compared with those receiving general anesthesia alone. Similarly, Ward and colleagues²⁹ noted a significant difference in postoperative nausea, with 10% of patients in the PNB group experiencing postoperative nausea compared with 75% of those in the comparator group who received intravenous morphine. The mean percentage of patients experiencing postoperative nausea and/or vomiting is shown in Table 4.

DISCHARGE TIME

Four studies presented data on discharge time from the PACU and are summarized in Table 3. Three of these studies included a comparator group. Both Dold and colleagues¹⁶ and YaDeau and colleagues¹⁸ reported an increase in the time to discharge for patients receiving PNB, although these differences were not significant. The study by Ward and colleagues,²⁹ on the other hand, noted a significant reduction in the time to discharge for the PNB group. In addition to these studies, Krych and colleagues²⁷ examined the time from skin closure to discharge for patients receiving PNB, noting a mean 199 minutes for the patients in their study. Mean times to discharge for the PNB and control groups are presented in Table 4.

INPATIENT ADMISSION

Four studies presented data on the proportion of study participants who were admitted as inpatients, and these data are shown in Table 3. Dold and colleagues¹⁶ reported no inpatient admissions in their PNB group compared with 5.0% for the control group (both cases of pain control), while YaDeau and colleagues¹⁸ found that 3 admissions occurred, with 2 in the control group (1 for oxygen desaturation and the other for intractable pain and nausea) and 1 from the PNB group (epidural spread and urinary retention). Two additional studies reported data on PNB groups alone. Krych and colleagues²⁷ observed no overnight admissions in their study, while Nye and colleagues²⁸ reported 1 readmission for bilateral leg numbness and weakness due to epidural spread, which resolved following discontinuation of the block. The mean proportion of inpatient admissions is presented in Table 4.

SATISFACTION

A total of 3 studies examined patient satisfaction, and these data are presented in Table 3. In their study, Ward and colleagues²⁹ reported a significantly greater rate of satisfaction at 1 day postoperatively among the patients in the PNB group (90%) than among patients who received intravenous morphine (25%) (P < .0001). Similarly, YaDeau and colleagues¹⁸ noted greater satisfaction among the PNB group than among the control group, with PNB patients rating their satisfaction at a mean of 8.6 and control patients at a mean of 7.9 on a 10-point scale (0-10) 24 hours postoperatively, although this difference was not significant. Finally, Krych and colleagues²⁷ found that 67% of patients were “very satisfied” and 33% were “satisfied”, based on a Likert scale.

COMPLICATIONS

Four studies presented data on complications, and these findings are summarized in Table 3. In their work, Nye and colleagues²⁸ reported most extensively on complications associated with PNB. Overall, the authors found a rate of significant complications of 3.8%. In terms of specific complications, they noted local anesthetic systemic toxicity (0.9%), epidural spread (0.5%), sensory or motor deficits (9.4%), falls (0.5%), and catheter issues. In their study of patients receiving PNB and CSE, YaDeau and colleagues¹⁸ identified 1 patient in the PNB group with epidural spread and urinary retention, while they noted 1 case of oxygen desaturation and another case of intractable pain and nausea in the group receiving CSE alone, all 3 of which required inpatient admission. They found no permanent adverse events attributable to the PNB. In another study, Dold and colleagues¹⁶ observed no complications in patients receiving PNB compared with those in 2 admissions in the control group for inadequate pain control. Similarly, Krych and colleagues²⁷ identified no complications in patients who received PNB in their study.

DISCUSSION

Hip arthroscopy has experienced a substantial gain in popularity in recent years, emerging as a beneficial technique for both the diagnosis and treatment of diverse hip pathologies in patients spanning a variety of demographics. Nevertheless, postoperative pain control, as well as medication side effects and unwanted patient admissions, present major challenges to the treating surgeon. As an adjuvant measure, peripheral nerve block represents one option to improve postoperative pain management, while at the same time addressing the adverse effects of considerable opioid use, which is commonly seen in these patients. Early experience with this method in hip arthroscopy was reported in a case series by Lee and colleagues.¹² In an attempt to reduce postoperative pain, as well as limit the adverse effects and delay in discharge associated with considerable opioid use in the PACU, the authors used preoperative paravertebral blocks of L1 and L2 in 2 patients requiring hip arthroscopy with encouraging results. Since then, a number of studies have attempted the use of PNB in hip arthroscopy.^16-18,27-29 However, we were unable to identify any prior reviews reporting on peripheral nerve blockade in hip arthroscopy, and thus this study is unique in providing a greater understanding of the outcomes associated with PNB use.

In general, we found that PNB was associated with improved outcomes. Based on the studies included in this review, there was a statistically significantly lower level of pain in the PACU for femoral nerve block (compared with general anesthesia alone)¹⁶ and lumbar plexus blockade (compared with general anesthesia¹⁷ and CSE¹⁸ alone). Nevertheless, these effects are likely short-lived, with differences disappearing the day following the procedure. In terms of analgesic use, 2 studies report significant reductions in analgesic use intraoperatively and in the PACU/Phase I recovery,^16,17 with a third reporting a strong trend toward reduced analgesic use in the PACU (P = .051).¹⁸ Finally, we report fewer admissions for the PNB group, as well as high rates of satisfaction and few complications across these studies.

Continue to: Unlike these measures...

Unlike these measures, postoperative nausea, vomiting, and antiemetic use, as well as time to discharge, showed more mixed results. With regard to nausea/vomiting, 2 studies^16,18 reported nonsignificantly increased rates in the PNB group, whereas others reported significant reductions in nausea/vomiting²⁹ and in the proportion of patients receiving antiemetics.¹⁷ Similarly, mixed results were seen in terms of patient discharge time from the PACU. Two studies^16,18 reported a nonsignificant increase in time to discharge for the PNB group, while another²⁹ noted a significant reduction for the PNB group compared with those receiving intravenous morphine. These mixed results were surprising, as we expected reductions in opioid use to result in fewer instances of nausea/vomiting and a quicker time to discharge. The reasons underlying these findings are not clear, although it has been suggested that current discharge guidelines and clinical pathways limit the ability to take advantage of the accelerated timeline offered by regional anesthesia.^16,30 As experience with PNB grows, our guidelines and pathways are likely to adapt to capitalize on these advantages, and future studies may show more reliable improvements in these measures.

While rare, the risk of bleeding requiring blood transfusion following hip arthroscopy is one of the most common complications of this procedure. Though the studies included in this review did not report on the need for transfusion, a recent study by Cvetanovich and colleagues¹⁰ used a national database and found that, of patients undergoing hip arthroscopy (n = 1338), 0.4% (n = 5) had bleeding requiring a transfusion, with 0.3% (n = 4) requiring return to the operating room, similar to an earlier study by Clarke and colleagues,³¹ who noted bleeding from the portal site in 0.4% of hip arthroscopy patients. In terms of risk factors, Cvetanovich and colleagues¹⁰ found that ASA class, older age, and prior cardiac surgery were significantly associated with minor and overall complications, whereas both regional anesthesia/monitored anesthesia care and alcohol consumption of >2 drinks a day were significantly associated with minor complications, including bleeding requiring transfusions. They noted, however, that these risk factors accounted for only 5% of the variance in complication rates, indicating that other unidentified variables better explained the variance in complication rates. These authors concluded that complications associated with hip arthroscopy are so rare that we may not be able to predict which risk factors or anesthesia types are more likely to cause them. Further characterization of bleeding following hip arthroscopy and its associated risk factors is a valuable area for future research.

LIMITATIONS

Our study contains a number of limitations. This review included studies whose level of evidence varied from I to IV; therefore, our study is limited by any bias or heterogeneity introduced in patient recruitment, selection, variability of technique, data collection, and analysis used in these studies. This heterogeneity is most apparent in the block types and comparator groups. Furthermore, several different outcome measures were reported across the 6 studies used in this review, which decreased the relevance of any one of these individual outcomes. Finally, given the limited data that currently exist for the use of PNB in hip arthroscopy, we are unable to note meaningful differences between various types of PNBs, such as differences in postoperative pain or other measures such as quadriceps weakness, which can accompany femoral nerve block.¹² While it is important to read our work with these limitations in mind, this systematic review is, to our knowledge, the only comprehensive review to date of studies reporting on PNB in hip arthroscopy, providing clinicians and patients with a greater understanding of the associated outcomes across these studies.

CONCLUSION

This systematic review shows improved outcomes and few complications with PNB use in hip arthroscopy, with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions. Although opioid use was reduced in these studies, we found similar rates of postoperative nausea/vomiting as well as similar time to discharge from the PACU, which may reflect our continued reliance on outdated discharge guidelines and clinical pathways. Current attempts to provide peripheral nerve blockade are quite varied, with studies targeting femoral nerve, fascia iliaca, L1/L2 paravertebral, and lumbar plexus blockade. Future research efforts with a large prospective trial investigating these techniques should focus on which of these PNBs presents the optimal risk-benefit profile for hip arthroscopy patients and thus appropriately address the clinical questions at hand.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

Pain control following hip arthroscopy presents a significant clinical challenge, with postoperative pain requiring considerable opioid use. Peripheral nerve blocks (PNBs) have emerged as one option to improve pain and limit the consequences of opioid use. The purpose of this study is to provide a comprehensive review of outcomes associated with PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications. A systematic review of PubMed, Medline, Scopus, and Embase databases was conducted through January 2015 for English-language articles reporting outcome data, with 2 reviewers independently reviewing studies for inclusion. When available, similar outcomes were combined to generate frequency-weighted means. Six studies met the inclusion criteria for this review, reporting on 710 patients undergoing hip arthroscopy. The mean ages were 37.0 and 37.7 years for the PNB and comparator groups, respectively, with a reported total of 281 (40.5%) male and 412 (59.5%) female patients. Postoperative post-anesthesia care unit (PACU) pain was consistently reduced in the PNB group, with the use of a lower morphine equivalent dose and lower rates of inpatient admission, compared with that in the control groups. Postoperative nausea and/or vomiting as well as PACU discharge time showed mixed results. High satisfaction and few complications were reported. In conclusion, PNB is associated with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions, though similar rates of nausea/vomiting and time to discharge were reported. Current PNB techniques are varied, and future research efforts should focus on examining which of these methods provides the optimal risk-benefit profile in hip arthroscopy.

Continue to: Hip arthroscopy has emerged...

Hip arthroscopy has emerged as a useful procedure in the diagnosis and treatment of hip pathology,^1-8 experiencing a substantial rise in popularity in recent years, with the number of procedures growing by a factor of 18 from 1999 to 2009⁹ and 25 from 2006 to 2013.¹⁰ Though hip arthroscopy is beneficial in many cases, marked postoperative pain has presented a substantial challenge, with patients requiring considerable doses of opiate-based medications in the post-anesthesia care unit (PACU).^11,12 Increased narcotic use carries increased side effects, including postoperative nausea and vomiting,¹³ and poorly managed pain leads to increased unplanned admissions.¹⁴ Furthermore, patients with chronic hip pain and long-term opioid use may experience heightened and prolonged pain following the procedure, owing to medication tolerance and reduced opioid efficacy in this setting.¹⁵

Several pain control strategies have been employed in patients undergoing hip arthroscopy. General anesthesia^16,17 and combined spinal epidural (CSE)¹⁸ are commonly used. However, such techniques rely heavily on opioids for postoperative pain control,¹¹ and epidural anesthesia commonly requires adjunctive treatments (eg, neuromuscular blockade) to ensure muscle relaxation for joint distraction.¹⁹ One technique that has been employed recently is peripheral nerve block (PNB), which has been associated with a significant decrease in postoperative opioid use and nausea and vomiting.^13,20 This method has proven successful in other fields of arthroscopy, including shoulder arthroscopy, in which it resulted in faster recovery, reduced opioid consumption,²¹ and demonstrated cost-effectiveness²² compared with general anesthesia and knee arthroscopy.^23-26 As it is a relatively new field, little is known about the use of PNB in hip arthroscopy.

The goal of this systematic review was to comprehensively review the studies reporting on PNB in hip arthroscopy. We specifically focused on outcomes, including postoperative pain; analgesic use; nausea, vomiting, and antiemetic use; discharge time; inpatient admission; and patient satisfaction, as well as the complications associated with the use of PNB. Our knowledge of outcomes associated with PNB in hip arthroscopy is based on a few individual studies that have reported on small groups of patients using a variety of outcome measures and other findings. Furthermore, each of these studies commonly reflects the experience of an individual surgeon at a single institution and, when taken alone, may not be an accurate representation of the more general outcomes associated with PNB. A comprehensive review of such studies will provide surgeons, anesthesiologists, and patients with a better understanding of the anticipated outcomes of using PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications.

MATERIALS AND METHODS

A systematic review of outcomes associated with PNB in hip arthroscopy was performed using the available English-language literature in accordance with the guidelines laid out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and included studies retrieved from the PubMed, Medline, Scopus, and Embase computerized literature databases. Searches were executed comprising all years from database inception through January 2015. Articles were retrieved by an electronic search of medical subject headings and keyword terms and their respective combinations (Table 1). The inclusion criteria for studies in this systematic review were studies that (1) were written in the English language and (2) reported explicit outcome data. The exclusion criteria were (1) review articles, meta-analyses, case reports, abstracts/conference papers, comments/letters, or technique articles without reported patient data and (2) basic research, biomechanics, or animal/cadaveric studies without reported patient data.

Table 1. Search Terms Entered to Identify English-Language Studies Through January 2015

The literature search strategy is outlined in the Figure. The initial title search yielded a subset of possible articles that were then further included or excluded on the basis of the contents of the article’s abstract, wherein articles were again selected on the basis of the aforementioned inclusion and exclusion criteria. Articles selected in both the title and abstract phases underwent full-text review, during which the full text of each qualifying article was reviewed. In addition, the reference sections from articles undergoing full-text review were scanned to identify any additional studies that had not been identified in the original literature search. Appropriate studies for final inclusion were then selected at this stage. The title, abstract, and full-text selection process were performed by 2 of the study authors (Dr. Steinhaus and Dr. Lynch), with any discrepancies being discussed and resolved by mutual agreement.

Continue to: For all 6 included studies...

For all 6 included studies,^16-18,27-29 data were collected regarding the study specifics, patients included, and outcomes measured in the study. The journal of publication, type of study, level of evidence, and type of PNB, as well as the presence of a comparator group were noted (Table 2). Patient information included the number of patients at baseline and follow-up, mean age, gender, weight, height, body mass index, American Society of Anesthesiologists (ASA) status, and the specific procedures performed. In addition, data were collected on outcomes, including postoperative pain, as well as secondary outcomes and additional findings reported by the studies (Table 3). Where possible, weighted averages were calculated across all studies to obtain aggregate data.

RESULTS

STUDY INCLUSION

Six studies, all published between 2012 and 2014, were included in this systematic review (Table 2). Three studies involved lumbar plexus block, 2 studies involved femoral nerve block, and 1 study evaluated fascia iliaca block. Two studies used a control group of patients who received only general anesthesia (compared with the treatment group who received both general anesthesia and PNB); another study compared intravenous morphine with PNB; and 1 study compared CSE alone with PNB in addition to epidural.

DEMOGRAPHIC DATA

Demographic data from the included studies are presented in Table 2. In total, 710 and 549 patients were evaluated at baseline and final follow-up, respectively, which represents a follow-up rate of 77%. The frequency-weighted mean age of patients receiving PNB was 37.0 years compared with 37.7 years in the comparison groups, and the studies reported a total of 281 (40.5%) male and 412 (59.5%) female patients. The procedures performed were heterogeneously reported; therefore, totals were not tabulated, although the reported procedures included osteochondroplasty, labral débridement, labral and/or capsular repair, gluteus minimus repair, and synovectomy.

POSTOPERATIVE PAIN

Four studies reported on postoperative pain, and these data are presented in Table 3. In a retrospective study of patients receiving femoral nerve block in addition to general anesthesia, Dold and colleagues¹⁶ noted postoperative pain at 0, 15, 30, 45, and 60 minutes following arrival in the PACU, and discovered a statistically significantly lower level of pain at 60 minutes compared with inpatients receiving general anesthesia alone. YaDeau and colleagues¹⁸ found a significantly lower level of pain at rest in the PACU for those receiving CSE and lumbar plexus blockade compared with those receiving CSE alone. This significant difference did not persist at 24 hours or 6 months after the procedure, nor did it exist for pain with movement at any time point. Similarly, Schroeder and colleagues¹⁷ examined patients receiving general anesthesia and lumbar plexus block and found a significant reduction in pain immediately postoperatively in the PACU, though these effects disappeared the day following the procedure. Krych and colleagues²⁷ also reported on postoperative pain in patients undergoing fascia iliaca blockade, although they did not include a comparator group. Outcome comparison between patients who received PNB and controls in the PACU and 1 day following the procedure are presented in Table 4.

ANALGESIC USE

Four studies reported on analgesic use after PNB, and these data are presented in Table 3. Dold and colleagues¹⁶ noted analgesic use intraoperatively, in the PACU, and in the surgical day care unit (SDCU). These authors found a significant reduction in morphine equivalent dose given in the operating room and in the PACU in the group receiving PNB, with a nonsignificant trend toward lower use of oxycodone in the SDCU. Schroeder and colleagues¹⁷ similarly reported significant reductions in morphine equivalent dose intraoperatively and in Phase I recovery for patients receiving PNB, and these differences disappeared in Phase II recovery as well as intraoperatively if the block dose was considered. In addition, these authors found a significant reduction in the use of fentanyl and hydromorphone in the operating room in the PNB group, as well as a significant reduction in the proportion of patients receiving ketorolac in the operating room or PACU. Finally, YaDeau and colleagues¹⁸ reported total analgesic usage in the PACU among PNB patients compared with those receiving CSE alone and showed a strong trend toward reduced use in the PNB group, although this difference was not significant (P = .051). PACU analgesic use is presented in Table 4.

Continue to: Postoperative nausea...

POSTOPERATIVE NAUSEA/VOMITING AND ANTIEMETIC USE

Five studies presented data on nausea, vomiting, or antiemetic use following PNB and are shown in Table 3. YaDeau and colleagues¹⁸ reported nausea among 34% of patients in the PNB group, compared with 20% in the control group, vomiting in 2% and 7%, respectively, and antiemetic use in 12% of both groups. Dold and colleagues¹⁶ identified a similar trend, with 41.1% of patients in the PNB group and 32.5% of patients in the control group experiencing postoperative nausea or vomiting, while Krych and colleagues²⁷ noted only 10% of PNB patients with mild nausea and none requiring antiemetic use. In their study of patients receiving PNB, Schroeder and colleagues¹⁷ found a significant reduction in antiemetic use among PNB patients compared with those receiving general anesthesia alone. Similarly, Ward and colleagues²⁹ noted a significant difference in postoperative nausea, with 10% of patients in the PNB group experiencing postoperative nausea compared with 75% of those in the comparator group who received intravenous morphine. The mean percentage of patients experiencing postoperative nausea and/or vomiting is shown in Table 4.

DISCHARGE TIME

Four studies presented data on discharge time from the PACU and are summarized in Table 3. Three of these studies included a comparator group. Both Dold and colleagues¹⁶ and YaDeau and colleagues¹⁸ reported an increase in the time to discharge for patients receiving PNB, although these differences were not significant. The study by Ward and colleagues,²⁹ on the other hand, noted a significant reduction in the time to discharge for the PNB group. In addition to these studies, Krych and colleagues²⁷ examined the time from skin closure to discharge for patients receiving PNB, noting a mean 199 minutes for the patients in their study. Mean times to discharge for the PNB and control groups are presented in Table 4.

INPATIENT ADMISSION

Four studies presented data on the proportion of study participants who were admitted as inpatients, and these data are shown in Table 3. Dold and colleagues¹⁶ reported no inpatient admissions in their PNB group compared with 5.0% for the control group (both cases of pain control), while YaDeau and colleagues¹⁸ found that 3 admissions occurred, with 2 in the control group (1 for oxygen desaturation and the other for intractable pain and nausea) and 1 from the PNB group (epidural spread and urinary retention). Two additional studies reported data on PNB groups alone. Krych and colleagues²⁷ observed no overnight admissions in their study, while Nye and colleagues²⁸ reported 1 readmission for bilateral leg numbness and weakness due to epidural spread, which resolved following discontinuation of the block. The mean proportion of inpatient admissions is presented in Table 4.

SATISFACTION

A total of 3 studies examined patient satisfaction, and these data are presented in Table 3. In their study, Ward and colleagues²⁹ reported a significantly greater rate of satisfaction at 1 day postoperatively among the patients in the PNB group (90%) than among patients who received intravenous morphine (25%) (P < .0001). Similarly, YaDeau and colleagues¹⁸ noted greater satisfaction among the PNB group than among the control group, with PNB patients rating their satisfaction at a mean of 8.6 and control patients at a mean of 7.9 on a 10-point scale (0-10) 24 hours postoperatively, although this difference was not significant. Finally, Krych and colleagues²⁷ found that 67% of patients were “very satisfied” and 33% were “satisfied”, based on a Likert scale.

COMPLICATIONS

Four studies presented data on complications, and these findings are summarized in Table 3. In their work, Nye and colleagues²⁸ reported most extensively on complications associated with PNB. Overall, the authors found a rate of significant complications of 3.8%. In terms of specific complications, they noted local anesthetic systemic toxicity (0.9%), epidural spread (0.5%), sensory or motor deficits (9.4%), falls (0.5%), and catheter issues. In their study of patients receiving PNB and CSE, YaDeau and colleagues¹⁸ identified 1 patient in the PNB group with epidural spread and urinary retention, while they noted 1 case of oxygen desaturation and another case of intractable pain and nausea in the group receiving CSE alone, all 3 of which required inpatient admission. They found no permanent adverse events attributable to the PNB. In another study, Dold and colleagues¹⁶ observed no complications in patients receiving PNB compared with those in 2 admissions in the control group for inadequate pain control. Similarly, Krych and colleagues²⁷ identified no complications in patients who received PNB in their study.

DISCUSSION

Hip arthroscopy has experienced a substantial gain in popularity in recent years, emerging as a beneficial technique for both the diagnosis and treatment of diverse hip pathologies in patients spanning a variety of demographics. Nevertheless, postoperative pain control, as well as medication side effects and unwanted patient admissions, present major challenges to the treating surgeon. As an adjuvant measure, peripheral nerve block represents one option to improve postoperative pain management, while at the same time addressing the adverse effects of considerable opioid use, which is commonly seen in these patients. Early experience with this method in hip arthroscopy was reported in a case series by Lee and colleagues.¹² In an attempt to reduce postoperative pain, as well as limit the adverse effects and delay in discharge associated with considerable opioid use in the PACU, the authors used preoperative paravertebral blocks of L1 and L2 in 2 patients requiring hip arthroscopy with encouraging results. Since then, a number of studies have attempted the use of PNB in hip arthroscopy.^16-18,27-29 However, we were unable to identify any prior reviews reporting on peripheral nerve blockade in hip arthroscopy, and thus this study is unique in providing a greater understanding of the outcomes associated with PNB use.

In general, we found that PNB was associated with improved outcomes. Based on the studies included in this review, there was a statistically significantly lower level of pain in the PACU for femoral nerve block (compared with general anesthesia alone)¹⁶ and lumbar plexus blockade (compared with general anesthesia¹⁷ and CSE¹⁸ alone). Nevertheless, these effects are likely short-lived, with differences disappearing the day following the procedure. In terms of analgesic use, 2 studies report significant reductions in analgesic use intraoperatively and in the PACU/Phase I recovery,^16,17 with a third reporting a strong trend toward reduced analgesic use in the PACU (P = .051).¹⁸ Finally, we report fewer admissions for the PNB group, as well as high rates of satisfaction and few complications across these studies.

Continue to: Unlike these measures...

Unlike these measures, postoperative nausea, vomiting, and antiemetic use, as well as time to discharge, showed more mixed results. With regard to nausea/vomiting, 2 studies^16,18 reported nonsignificantly increased rates in the PNB group, whereas others reported significant reductions in nausea/vomiting²⁹ and in the proportion of patients receiving antiemetics.¹⁷ Similarly, mixed results were seen in terms of patient discharge time from the PACU. Two studies^16,18 reported a nonsignificant increase in time to discharge for the PNB group, while another²⁹ noted a significant reduction for the PNB group compared with those receiving intravenous morphine. These mixed results were surprising, as we expected reductions in opioid use to result in fewer instances of nausea/vomiting and a quicker time to discharge. The reasons underlying these findings are not clear, although it has been suggested that current discharge guidelines and clinical pathways limit the ability to take advantage of the accelerated timeline offered by regional anesthesia.^16,30 As experience with PNB grows, our guidelines and pathways are likely to adapt to capitalize on these advantages, and future studies may show more reliable improvements in these measures.

While rare, the risk of bleeding requiring blood transfusion following hip arthroscopy is one of the most common complications of this procedure. Though the studies included in this review did not report on the need for transfusion, a recent study by Cvetanovich and colleagues¹⁰ used a national database and found that, of patients undergoing hip arthroscopy (n = 1338), 0.4% (n = 5) had bleeding requiring a transfusion, with 0.3% (n = 4) requiring return to the operating room, similar to an earlier study by Clarke and colleagues,³¹ who noted bleeding from the portal site in 0.4% of hip arthroscopy patients. In terms of risk factors, Cvetanovich and colleagues¹⁰ found that ASA class, older age, and prior cardiac surgery were significantly associated with minor and overall complications, whereas both regional anesthesia/monitored anesthesia care and alcohol consumption of >2 drinks a day were significantly associated with minor complications, including bleeding requiring transfusions. They noted, however, that these risk factors accounted for only 5% of the variance in complication rates, indicating that other unidentified variables better explained the variance in complication rates. These authors concluded that complications associated with hip arthroscopy are so rare that we may not be able to predict which risk factors or anesthesia types are more likely to cause them. Further characterization of bleeding following hip arthroscopy and its associated risk factors is a valuable area for future research.

LIMITATIONS

Our study contains a number of limitations. This review included studies whose level of evidence varied from I to IV; therefore, our study is limited by any bias or heterogeneity introduced in patient recruitment, selection, variability of technique, data collection, and analysis used in these studies. This heterogeneity is most apparent in the block types and comparator groups. Furthermore, several different outcome measures were reported across the 6 studies used in this review, which decreased the relevance of any one of these individual outcomes. Finally, given the limited data that currently exist for the use of PNB in hip arthroscopy, we are unable to note meaningful differences between various types of PNBs, such as differences in postoperative pain or other measures such as quadriceps weakness, which can accompany femoral nerve block.¹² While it is important to read our work with these limitations in mind, this systematic review is, to our knowledge, the only comprehensive review to date of studies reporting on PNB in hip arthroscopy, providing clinicians and patients with a greater understanding of the associated outcomes across these studies.

CONCLUSION

This systematic review shows improved outcomes and few complications with PNB use in hip arthroscopy, with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions. Although opioid use was reduced in these studies, we found similar rates of postoperative nausea/vomiting as well as similar time to discharge from the PACU, which may reflect our continued reliance on outdated discharge guidelines and clinical pathways. Current attempts to provide peripheral nerve blockade are quite varied, with studies targeting femoral nerve, fascia iliaca, L1/L2 paravertebral, and lumbar plexus blockade. Future research efforts with a large prospective trial investigating these techniques should focus on which of these PNBs presents the optimal risk-benefit profile for hip arthroscopy patients and thus appropriately address the clinical questions at hand.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

Pain control following hip arthroscopy presents a significant clinical challenge, with postoperative pain requiring considerable opioid use. Peripheral nerve blocks (PNBs) have emerged as one option to improve pain and limit the consequences of opioid use. The purpose of this study is to provide a comprehensive review of outcomes associated with PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications. A systematic review of PubMed, Medline, Scopus, and Embase databases was conducted through January 2015 for English-language articles reporting outcome data, with 2 reviewers independently reviewing studies for inclusion. When available, similar outcomes were combined to generate frequency-weighted means. Six studies met the inclusion criteria for this review, reporting on 710 patients undergoing hip arthroscopy. The mean ages were 37.0 and 37.7 years for the PNB and comparator groups, respectively, with a reported total of 281 (40.5%) male and 412 (59.5%) female patients. Postoperative post-anesthesia care unit (PACU) pain was consistently reduced in the PNB group, with the use of a lower morphine equivalent dose and lower rates of inpatient admission, compared with that in the control groups. Postoperative nausea and/or vomiting as well as PACU discharge time showed mixed results. High satisfaction and few complications were reported. In conclusion, PNB is associated with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions, though similar rates of nausea/vomiting and time to discharge were reported. Current PNB techniques are varied, and future research efforts should focus on examining which of these methods provides the optimal risk-benefit profile in hip arthroscopy.

Continue to: Hip arthroscopy has emerged...

Hip arthroscopy has emerged as a useful procedure in the diagnosis and treatment of hip pathology,^1-8 experiencing a substantial rise in popularity in recent years, with the number of procedures growing by a factor of 18 from 1999 to 2009⁹ and 25 from 2006 to 2013.¹⁰ Though hip arthroscopy is beneficial in many cases, marked postoperative pain has presented a substantial challenge, with patients requiring considerable doses of opiate-based medications in the post-anesthesia care unit (PACU).^11,12 Increased narcotic use carries increased side effects, including postoperative nausea and vomiting,¹³ and poorly managed pain leads to increased unplanned admissions.¹⁴ Furthermore, patients with chronic hip pain and long-term opioid use may experience heightened and prolonged pain following the procedure, owing to medication tolerance and reduced opioid efficacy in this setting.¹⁵

Several pain control strategies have been employed in patients undergoing hip arthroscopy. General anesthesia^16,17 and combined spinal epidural (CSE)¹⁸ are commonly used. However, such techniques rely heavily on opioids for postoperative pain control,¹¹ and epidural anesthesia commonly requires adjunctive treatments (eg, neuromuscular blockade) to ensure muscle relaxation for joint distraction.¹⁹ One technique that has been employed recently is peripheral nerve block (PNB), which has been associated with a significant decrease in postoperative opioid use and nausea and vomiting.^13,20 This method has proven successful in other fields of arthroscopy, including shoulder arthroscopy, in which it resulted in faster recovery, reduced opioid consumption,²¹ and demonstrated cost-effectiveness²² compared with general anesthesia and knee arthroscopy.^23-26 As it is a relatively new field, little is known about the use of PNB in hip arthroscopy.

The goal of this systematic review was to comprehensively review the studies reporting on PNB in hip arthroscopy. We specifically focused on outcomes, including postoperative pain; analgesic use; nausea, vomiting, and antiemetic use; discharge time; inpatient admission; and patient satisfaction, as well as the complications associated with the use of PNB. Our knowledge of outcomes associated with PNB in hip arthroscopy is based on a few individual studies that have reported on small groups of patients using a variety of outcome measures and other findings. Furthermore, each of these studies commonly reflects the experience of an individual surgeon at a single institution and, when taken alone, may not be an accurate representation of the more general outcomes associated with PNB. A comprehensive review of such studies will provide surgeons, anesthesiologists, and patients with a better understanding of the anticipated outcomes of using PNB in hip arthroscopy. We hypothesize that the use of PNB in hip arthroscopy leads to improved outcomes and is associated with few complications.

MATERIALS AND METHODS

A systematic review of outcomes associated with PNB in hip arthroscopy was performed using the available English-language literature in accordance with the guidelines laid out by the Preferred Reporting Items for Systematic Reviews and Meta-Analyses statement and included studies retrieved from the PubMed, Medline, Scopus, and Embase computerized literature databases. Searches were executed comprising all years from database inception through January 2015. Articles were retrieved by an electronic search of medical subject headings and keyword terms and their respective combinations (Table 1). The inclusion criteria for studies in this systematic review were studies that (1) were written in the English language and (2) reported explicit outcome data. The exclusion criteria were (1) review articles, meta-analyses, case reports, abstracts/conference papers, comments/letters, or technique articles without reported patient data and (2) basic research, biomechanics, or animal/cadaveric studies without reported patient data.

Table 1. Search Terms Entered to Identify English-Language Studies Through January 2015

The literature search strategy is outlined in the Figure. The initial title search yielded a subset of possible articles that were then further included or excluded on the basis of the contents of the article’s abstract, wherein articles were again selected on the basis of the aforementioned inclusion and exclusion criteria. Articles selected in both the title and abstract phases underwent full-text review, during which the full text of each qualifying article was reviewed. In addition, the reference sections from articles undergoing full-text review were scanned to identify any additional studies that had not been identified in the original literature search. Appropriate studies for final inclusion were then selected at this stage. The title, abstract, and full-text selection process were performed by 2 of the study authors (Dr. Steinhaus and Dr. Lynch), with any discrepancies being discussed and resolved by mutual agreement.

Continue to: For all 6 included studies...

For all 6 included studies,^16-18,27-29 data were collected regarding the study specifics, patients included, and outcomes measured in the study. The journal of publication, type of study, level of evidence, and type of PNB, as well as the presence of a comparator group were noted (Table 2). Patient information included the number of patients at baseline and follow-up, mean age, gender, weight, height, body mass index, American Society of Anesthesiologists (ASA) status, and the specific procedures performed. In addition, data were collected on outcomes, including postoperative pain, as well as secondary outcomes and additional findings reported by the studies (Table 3). Where possible, weighted averages were calculated across all studies to obtain aggregate data.

RESULTS

STUDY INCLUSION

Six studies, all published between 2012 and 2014, were included in this systematic review (Table 2). Three studies involved lumbar plexus block, 2 studies involved femoral nerve block, and 1 study evaluated fascia iliaca block. Two studies used a control group of patients who received only general anesthesia (compared with the treatment group who received both general anesthesia and PNB); another study compared intravenous morphine with PNB; and 1 study compared CSE alone with PNB in addition to epidural.

DEMOGRAPHIC DATA

Demographic data from the included studies are presented in Table 2. In total, 710 and 549 patients were evaluated at baseline and final follow-up, respectively, which represents a follow-up rate of 77%. The frequency-weighted mean age of patients receiving PNB was 37.0 years compared with 37.7 years in the comparison groups, and the studies reported a total of 281 (40.5%) male and 412 (59.5%) female patients. The procedures performed were heterogeneously reported; therefore, totals were not tabulated, although the reported procedures included osteochondroplasty, labral débridement, labral and/or capsular repair, gluteus minimus repair, and synovectomy.

POSTOPERATIVE PAIN

Four studies reported on postoperative pain, and these data are presented in Table 3. In a retrospective study of patients receiving femoral nerve block in addition to general anesthesia, Dold and colleagues¹⁶ noted postoperative pain at 0, 15, 30, 45, and 60 minutes following arrival in the PACU, and discovered a statistically significantly lower level of pain at 60 minutes compared with inpatients receiving general anesthesia alone. YaDeau and colleagues¹⁸ found a significantly lower level of pain at rest in the PACU for those receiving CSE and lumbar plexus blockade compared with those receiving CSE alone. This significant difference did not persist at 24 hours or 6 months after the procedure, nor did it exist for pain with movement at any time point. Similarly, Schroeder and colleagues¹⁷ examined patients receiving general anesthesia and lumbar plexus block and found a significant reduction in pain immediately postoperatively in the PACU, though these effects disappeared the day following the procedure. Krych and colleagues²⁷ also reported on postoperative pain in patients undergoing fascia iliaca blockade, although they did not include a comparator group. Outcome comparison between patients who received PNB and controls in the PACU and 1 day following the procedure are presented in Table 4.

ANALGESIC USE

Four studies reported on analgesic use after PNB, and these data are presented in Table 3. Dold and colleagues¹⁶ noted analgesic use intraoperatively, in the PACU, and in the surgical day care unit (SDCU). These authors found a significant reduction in morphine equivalent dose given in the operating room and in the PACU in the group receiving PNB, with a nonsignificant trend toward lower use of oxycodone in the SDCU. Schroeder and colleagues¹⁷ similarly reported significant reductions in morphine equivalent dose intraoperatively and in Phase I recovery for patients receiving PNB, and these differences disappeared in Phase II recovery as well as intraoperatively if the block dose was considered. In addition, these authors found a significant reduction in the use of fentanyl and hydromorphone in the operating room in the PNB group, as well as a significant reduction in the proportion of patients receiving ketorolac in the operating room or PACU. Finally, YaDeau and colleagues¹⁸ reported total analgesic usage in the PACU among PNB patients compared with those receiving CSE alone and showed a strong trend toward reduced use in the PNB group, although this difference was not significant (P = .051). PACU analgesic use is presented in Table 4.

Continue to: Postoperative nausea...

POSTOPERATIVE NAUSEA/VOMITING AND ANTIEMETIC USE

Five studies presented data on nausea, vomiting, or antiemetic use following PNB and are shown in Table 3. YaDeau and colleagues¹⁸ reported nausea among 34% of patients in the PNB group, compared with 20% in the control group, vomiting in 2% and 7%, respectively, and antiemetic use in 12% of both groups. Dold and colleagues¹⁶ identified a similar trend, with 41.1% of patients in the PNB group and 32.5% of patients in the control group experiencing postoperative nausea or vomiting, while Krych and colleagues²⁷ noted only 10% of PNB patients with mild nausea and none requiring antiemetic use. In their study of patients receiving PNB, Schroeder and colleagues¹⁷ found a significant reduction in antiemetic use among PNB patients compared with those receiving general anesthesia alone. Similarly, Ward and colleagues²⁹ noted a significant difference in postoperative nausea, with 10% of patients in the PNB group experiencing postoperative nausea compared with 75% of those in the comparator group who received intravenous morphine. The mean percentage of patients experiencing postoperative nausea and/or vomiting is shown in Table 4.

DISCHARGE TIME

Four studies presented data on discharge time from the PACU and are summarized in Table 3. Three of these studies included a comparator group. Both Dold and colleagues¹⁶ and YaDeau and colleagues¹⁸ reported an increase in the time to discharge for patients receiving PNB, although these differences were not significant. The study by Ward and colleagues,²⁹ on the other hand, noted a significant reduction in the time to discharge for the PNB group. In addition to these studies, Krych and colleagues²⁷ examined the time from skin closure to discharge for patients receiving PNB, noting a mean 199 minutes for the patients in their study. Mean times to discharge for the PNB and control groups are presented in Table 4.

INPATIENT ADMISSION

Four studies presented data on the proportion of study participants who were admitted as inpatients, and these data are shown in Table 3. Dold and colleagues¹⁶ reported no inpatient admissions in their PNB group compared with 5.0% for the control group (both cases of pain control), while YaDeau and colleagues¹⁸ found that 3 admissions occurred, with 2 in the control group (1 for oxygen desaturation and the other for intractable pain and nausea) and 1 from the PNB group (epidural spread and urinary retention). Two additional studies reported data on PNB groups alone. Krych and colleagues²⁷ observed no overnight admissions in their study, while Nye and colleagues²⁸ reported 1 readmission for bilateral leg numbness and weakness due to epidural spread, which resolved following discontinuation of the block. The mean proportion of inpatient admissions is presented in Table 4.

SATISFACTION

A total of 3 studies examined patient satisfaction, and these data are presented in Table 3. In their study, Ward and colleagues²⁹ reported a significantly greater rate of satisfaction at 1 day postoperatively among the patients in the PNB group (90%) than among patients who received intravenous morphine (25%) (P < .0001). Similarly, YaDeau and colleagues¹⁸ noted greater satisfaction among the PNB group than among the control group, with PNB patients rating their satisfaction at a mean of 8.6 and control patients at a mean of 7.9 on a 10-point scale (0-10) 24 hours postoperatively, although this difference was not significant. Finally, Krych and colleagues²⁷ found that 67% of patients were “very satisfied” and 33% were “satisfied”, based on a Likert scale.

COMPLICATIONS

Four studies presented data on complications, and these findings are summarized in Table 3. In their work, Nye and colleagues²⁸ reported most extensively on complications associated with PNB. Overall, the authors found a rate of significant complications of 3.8%. In terms of specific complications, they noted local anesthetic systemic toxicity (0.9%), epidural spread (0.5%), sensory or motor deficits (9.4%), falls (0.5%), and catheter issues. In their study of patients receiving PNB and CSE, YaDeau and colleagues¹⁸ identified 1 patient in the PNB group with epidural spread and urinary retention, while they noted 1 case of oxygen desaturation and another case of intractable pain and nausea in the group receiving CSE alone, all 3 of which required inpatient admission. They found no permanent adverse events attributable to the PNB. In another study, Dold and colleagues¹⁶ observed no complications in patients receiving PNB compared with those in 2 admissions in the control group for inadequate pain control. Similarly, Krych and colleagues²⁷ identified no complications in patients who received PNB in their study.

DISCUSSION

Hip arthroscopy has experienced a substantial gain in popularity in recent years, emerging as a beneficial technique for both the diagnosis and treatment of diverse hip pathologies in patients spanning a variety of demographics. Nevertheless, postoperative pain control, as well as medication side effects and unwanted patient admissions, present major challenges to the treating surgeon. As an adjuvant measure, peripheral nerve block represents one option to improve postoperative pain management, while at the same time addressing the adverse effects of considerable opioid use, which is commonly seen in these patients. Early experience with this method in hip arthroscopy was reported in a case series by Lee and colleagues.¹² In an attempt to reduce postoperative pain, as well as limit the adverse effects and delay in discharge associated with considerable opioid use in the PACU, the authors used preoperative paravertebral blocks of L1 and L2 in 2 patients requiring hip arthroscopy with encouraging results. Since then, a number of studies have attempted the use of PNB in hip arthroscopy.^16-18,27-29 However, we were unable to identify any prior reviews reporting on peripheral nerve blockade in hip arthroscopy, and thus this study is unique in providing a greater understanding of the outcomes associated with PNB use.

In general, we found that PNB was associated with improved outcomes. Based on the studies included in this review, there was a statistically significantly lower level of pain in the PACU for femoral nerve block (compared with general anesthesia alone)¹⁶ and lumbar plexus blockade (compared with general anesthesia¹⁷ and CSE¹⁸ alone). Nevertheless, these effects are likely short-lived, with differences disappearing the day following the procedure. In terms of analgesic use, 2 studies report significant reductions in analgesic use intraoperatively and in the PACU/Phase I recovery,^16,17 with a third reporting a strong trend toward reduced analgesic use in the PACU (P = .051).¹⁸ Finally, we report fewer admissions for the PNB group, as well as high rates of satisfaction and few complications across these studies.

Continue to: Unlike these measures...

Unlike these measures, postoperative nausea, vomiting, and antiemetic use, as well as time to discharge, showed more mixed results. With regard to nausea/vomiting, 2 studies^16,18 reported nonsignificantly increased rates in the PNB group, whereas others reported significant reductions in nausea/vomiting²⁹ and in the proportion of patients receiving antiemetics.¹⁷ Similarly, mixed results were seen in terms of patient discharge time from the PACU. Two studies^16,18 reported a nonsignificant increase in time to discharge for the PNB group, while another²⁹ noted a significant reduction for the PNB group compared with those receiving intravenous morphine. These mixed results were surprising, as we expected reductions in opioid use to result in fewer instances of nausea/vomiting and a quicker time to discharge. The reasons underlying these findings are not clear, although it has been suggested that current discharge guidelines and clinical pathways limit the ability to take advantage of the accelerated timeline offered by regional anesthesia.^16,30 As experience with PNB grows, our guidelines and pathways are likely to adapt to capitalize on these advantages, and future studies may show more reliable improvements in these measures.

While rare, the risk of bleeding requiring blood transfusion following hip arthroscopy is one of the most common complications of this procedure. Though the studies included in this review did not report on the need for transfusion, a recent study by Cvetanovich and colleagues¹⁰ used a national database and found that, of patients undergoing hip arthroscopy (n = 1338), 0.4% (n = 5) had bleeding requiring a transfusion, with 0.3% (n = 4) requiring return to the operating room, similar to an earlier study by Clarke and colleagues,³¹ who noted bleeding from the portal site in 0.4% of hip arthroscopy patients. In terms of risk factors, Cvetanovich and colleagues¹⁰ found that ASA class, older age, and prior cardiac surgery were significantly associated with minor and overall complications, whereas both regional anesthesia/monitored anesthesia care and alcohol consumption of >2 drinks a day were significantly associated with minor complications, including bleeding requiring transfusions. They noted, however, that these risk factors accounted for only 5% of the variance in complication rates, indicating that other unidentified variables better explained the variance in complication rates. These authors concluded that complications associated with hip arthroscopy are so rare that we may not be able to predict which risk factors or anesthesia types are more likely to cause them. Further characterization of bleeding following hip arthroscopy and its associated risk factors is a valuable area for future research.

LIMITATIONS

Our study contains a number of limitations. This review included studies whose level of evidence varied from I to IV; therefore, our study is limited by any bias or heterogeneity introduced in patient recruitment, selection, variability of technique, data collection, and analysis used in these studies. This heterogeneity is most apparent in the block types and comparator groups. Furthermore, several different outcome measures were reported across the 6 studies used in this review, which decreased the relevance of any one of these individual outcomes. Finally, given the limited data that currently exist for the use of PNB in hip arthroscopy, we are unable to note meaningful differences between various types of PNBs, such as differences in postoperative pain or other measures such as quadriceps weakness, which can accompany femoral nerve block.¹² While it is important to read our work with these limitations in mind, this systematic review is, to our knowledge, the only comprehensive review to date of studies reporting on PNB in hip arthroscopy, providing clinicians and patients with a greater understanding of the associated outcomes across these studies.

CONCLUSION

This systematic review shows improved outcomes and few complications with PNB use in hip arthroscopy, with reductions in postoperative pain, analgesic use, and the rate of inpatient admissions. Although opioid use was reduced in these studies, we found similar rates of postoperative nausea/vomiting as well as similar time to discharge from the PACU, which may reflect our continued reliance on outdated discharge guidelines and clinical pathways. Current attempts to provide peripheral nerve blockade are quite varied, with studies targeting femoral nerve, fascia iliaca, L1/L2 paravertebral, and lumbar plexus blockade. Future research efforts with a large prospective trial investigating these techniques should focus on which of these PNBs presents the optimal risk-benefit profile for hip arthroscopy patients and thus appropriately address the clinical questions at hand.

This paper will be judged for the Resident Writer’s Award.

ABSTRACT

The articular surface replacement (ASR) monoblock metal-on-metal acetabular component was recalled due to a higher than expected early failure rate. We evaluated the survivorship of the device and variables that may be predictive of failure at a minimum of 5-year follow-up. A single-center, single-surgeon retrospective review was conducted in patients who received the DePuy Synthes ASR™ XL Acetabular hip system from December 2005 to November 2009. Mean values and percentages were calculated and compared using the Fisher’s exact test, simple logistic regression, and Student’s t-test. The significance level was P ≤ .05. This study included 29 patients (24 males, 5 females) with 32 ASR™ XL acetabular hip systems. Mean age and body mass index (BMI) reached 55.2 years and 28.9 kg/m², respectively. Mean postoperative follow-up was 6.2 years. A total of 2 patients (6.9%) died of an unrelated cause and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom were available for follow-up. The 5-year revision rate was 34.4% (10 patients with 11 hip replacements). Mean time to revision was 3.1 years. Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with the increased rate for hip failure. Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months following the revision. This study demonstrates a high rate of failure of ASR acetabular components used in total hip arthroplasty at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is required.

Continue to: Metal-on-metal...

Metal-on-metal (MoM) articulations have been widely explored as an alternative to polyethylene bearings in total hip arthroplasty (THA), with proposed benefits including improved range of motion, lower dislocation rates, and enhanced durability.¹ Comprising cobalt and chromium, these MoM bearings gained widespread popularity in the United States, particularly in younger and more active patients looking for longer lasting devices.

The articular surface replacement (ASR) acetabular system (DePuy Synthes) was approved for sale by the US Food and Drug Administration in 2003 and implanted in an estimated 93,000 cases.² Since then, however, the early failure rate of the prosthesis has been well documented,^3-5 leading to a formal global product recall in August 2010. The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) was amongst the first to report a 6.4% rate of failure of the device at 3 years when inserted with a Corail stem.⁶ An acceptable upper rate of hip prosthesis failure is considered to reach 1% per year, with the majority of implants reporting well below this value. A 10.9% failure rate at 5 years was documented when the prosthesis was inserted for resurfacing. The National Joint Registry of England and Wales confirmed these findings and observed a 13% and 12% rate of failure at 5 years for the acetabular and resurfacing systems, respectively.² With the notable failure of the ASR system, this study reports our single-center 5-year survivorship experience and evaluates any variable that might be predictive of an early failure to aid in patient counseling.

METHODS

A single-center, single-surgeon, retrospective review of a consecutive series of patients was performed from December 2005 to November 2009. This study included all patients who underwent a primary THA with a DePuy Synthes ASR™ XL Acetabular hip system. No patients were excluded. Institutional Review Board approval was obtained. Patient demographics comprising of age, gender, and body mass index (BMI) were recorded. The primary endpoint of this study was 5-year survivorship rates. Secondary endpoints included duration to revision surgery, blood cobalt and chromium levels, time interval of blood ion tests, acetabulum size, acetabular component abduction angle, and duration to follow-up.

Candidates for the ASR™ XL Acetabular hip system included young patients and/or those considered to be physically active. In a select few, ASR devices were implanted upon patient request.

All patients underwent primary total hip replacement with a DePuy Synthes ASR™ XL uncemented acetabular component and an uncemented femoral stem (DePuy Synthes, Summit, or Tri-Lock) inserted via a standard posterior approach (Figure 1). Acetabulum sizes ranged from 52 mm to 68 mm in diameter.

All patients were followed-up yearly in the outpatient setting. Routine (yearly) metal-ion level sampling (whole blood) was started in 2010 for all patients. Laboratory tests were conducted at a single laboratory (Lab Corp.). Abduction cup inclination angles were measured by the providing surgeon using digital radiology software (GE Centricity systems).

The Student’s t-test was used to compare mean values (such as age, BMI, and metal ion levels) between the failure and no-failure groups. The 2-sided Fisher’s exact test analyzed differences in gender. Simple logistic regression analyzed variables associated with the failure group. Significance was P ≤ .05.

Continue to: Results...

RESULTS

A total of 29 patients (24 males, 5 females) with 32 ASR hip replacements were included in this study. Indications for surgery comprised osteoarthritis (28 hips, 87.5%) and avascular necrosis of the hip (4 hips, 12.5%). Mean age and BMI were 55.2 years and 28.9 kg/m², respectively. A total of 2 patients (6.9%) died of an unrelated cause (1 myocardial infarct, 1 suicide), and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom finished a 5-year minimum follow-up.

No implant failures were noted in the first year. The 5-year revision rate reached 34.4% (10 patients with 11 hip replacements). Mean time to revision for this subgroup was 3.1 years. Overall, an implant failure was observed in 37.5% of patients (11 patients with 12 hip replacements) at a mean postoperative follow-up of 6.2 years (Figure 2). Indications for implant revision were pain in 11 (92.7%) cases and infection in 1 (8.3%).

Of the 11 hips revised due to pain, 9 were performed by the original surgeon (8 were completed with primary acetabular components, 1 with a revision shell). Figure 3 shows a bilateral revision performed with primary acetabular components and retained DePuy Synthes Pinnacle femoral stems. In all these cases except 1, the ASR component was grossly loose. One case presented with pseudotumor and impingement between the femoral prosthetic neck and acetabular component after migration of a loose component. After revision, the patient returned with substantial anterior hip pain and heterotopic ossification, and failed conservative treatment, requiring another surgery with prosthesis retention, removal of heterotopic ossification, and iliopsoas lengthening. The surgery successfully relieved the symptoms. No other patients required additional surgery after their revision. In comparison to the original ASR component, the revision shell was 2 to 4 mm larger in diameter. No patient required component revision at a mean of 2.9 years after the revision surgery.

The patient with secondary revision developed a hematogenous streptococcal infection after a dental procedure performed without prophylactic antibiotics. The patient was initially lost to follow-up after the primary surgery and reported no antecedent pain prior to the revision. A substantial metal fluid collection was identified in the hip at the time of débridement and without component loosening. After débridement, the patient developed persistent metal stained wound drainage, necessitating ultimate successful treatment with a 2-stage exchange procedure.

Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with an increased rate for hip failure (Table). Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. The upper limits of blood cobalt and chromium levels reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months after the revision (Figure 4).

Table. Variables Not Associated with Early ASR Failure

Continue to: Discussion...

DISCUSSION

According to the Center for Disease Control and Prevention, 310,800 total hip replacements were performed among inpatients aged 45 years and older in the US in 2010.⁷ Specifically, in the 55- to 64-year-old age group, the number of procedures performed tripled from 2000 through 2010. As younger and more active patients opt for hip replacements, a growing need for prosthesis with enhanced durability is observed.

Despite the early proposed advantages of large head MoM bearings, our retrospective study of the DePuy Synthes ASR™ XL Acetabular hip system yielded 15.6% and 34.4% failure rates at 3 and 5 years, respectively. These higher-than-expected rates of failure are consistent with published data. The British Hip Society reported a 21% to 35% revision rate at 4 years and 49% at 6 years for the ASR XL prosthesis.⁸ In comparison, other MoM prosthesis, on average, report a 12% to 15% rate of failure at 5 years.

Considerable controversy surrounds the causes of adverse wear failure in MoM bearings.^9,10 The non-modular design of the ASR prostheses is frequently implicated as a cause of early failure. The lack of a central hole in the 1-piece component compromises the tactile feel of insertion, thereby reducing the surgeon’s ability to assess complete seating.¹¹ This condition may potentially increase the abduction angle at the time of insertion. Screw fixation of the non-modular device is not possible. The ASR XL device (148^° to 160^°) is less than a hemisphere (180^°) in size and hence features a diminished functional articular surface, further compromising implant fixation.¹¹ The functional articular surface is defined as the optimal surface area (10 mm) needed for a MoM implant.¹² Griffin and colleagues¹³ reported a 48 mm ASR XL component, when implanted at 45^° of abduction, to function similar to an implant at 59^° of abduction, leading to diminished lubrication, metallosis, and edge loading. The version of the acetabular component may similarly and adversely affect implant wear characteristics. Furthermore, the variable thickness of the implant, which is thicker at the dome and thinner at the rim, may further promote edge loading by shifting the center of rotation of the femoral head out from the center of the acetabular prosthesis.¹¹ Studies have also shown that increased wear of the MoM articulation is associated with an acetabular component inclination angle in excess of 55^°10,14 and a failure of fixation at time of implantation.¹⁵ This study, however, found no correlation between the abduction angle and risk of early implant failure for the ASR acetabular component. No correlation was also detected between the acetabulum size and revision surgery.

The AOANJRR reported loosening (44%), infection (20%), metal sensitivity (12%), fracture (9%), and dislocation of prosthesis (7%) as the indications for revision surgery for the ASR prosthesis.⁶ Furthermore, a single-center retrospective review of 70 consecutive MoM THAs with ultra-large diameter femoral head and monoblock acetabular components showed that 17.1% required revision within 3 years for loosening, pain, and squeaking.¹ Overall, 28.6% of patients reported implant dysfunction. In this study, we observed a similar rate of failure at 3 years (15.6%) for pain (11) and infection (1). The revision surgery successfully relieved all of these symptoms. One patient presented with heterotopic ossification and anterior hip pain after the original revision and required additional surgery with prosthesis retention. No patient in this series required repeat component revisions at a mean of 2.9 years after surgery. In all but 1 case, primary acetabular components were used in the revision, and in all cases except that with infection, the femoral component was retained. Replacement shells were 2 to 4 mm larger in diameter than the original ASR component.

Recently, concerns have arisen regarding the long-term effects of serum cobalt and chromium metal ions levels. Studies have shown increased serum metal ion levels,¹⁵ groin pain,¹⁶ pseudotumor formation,¹⁷ and metallosis¹⁸ after the implantation of MoM bearings. In a case study by Mao and colleagues,¹⁹ 1 patient reported headaches, anorexia, continuous metallic taste in her mouth, and weight loss. A cerebrospinal fluid analysis revealed cobalt and chromium levels at 9 and 13 nmol/L, respectively, indicating that these metal ions can cross the blood-brain barrier. Another patient reported painful muscle fatigue, night cramps, fainting spells, cognitive decline, and an inability to climb stairs. His serum cobalt level reached 258 nmol/L (reference range, 0-20 nmol/L), and chromium level totaled 88 nmol/L (reference range, 0-100 nmol/L). At 8-week follow-up after revision surgery, the symptoms of the patient had resolved, with serum cobalt levels dropping to 42 nmol/L.¹⁹ None of the patients in this study presented with any signs or symptoms of metal toxicity. The upper limits of blood cobalt and chromium levels in our study population reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. However, we noted a similar drop in post-revision blood cobalt (91% decrease) and chromium (78% decrease) levels.

In summary, our data showed a high revision rate of the DePuy Synthes ASR™ XL Acetabular hip system. Our findings are consistent with internationally published data. In the absence of reliable predictors of early failure, continued close clinical surveillance and laboratory monitoring of these patients are warranted.

CONCLUSION

This study demonstrates the high failure rate of the DePuy Synthes ASR™ XL Acetabular hip system used in THA at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is therefore required. Metal levels dropped quickly after revision, and the revision surgery can generally be performed with slightly larger primary components. Symptomatic patients with ASR hip replacements, regardless of blood metal-ion levels, were candidates for the revision surgery. Not all failed hips exhibited substantially elevated metal levels. Asymptomatic patients with high blood metal-ion levels should be closely followed-up and revision surgery should be strongly considered, consistent with recently published guidelines.²⁰

ABSTRACT

The articular surface replacement (ASR) monoblock metal-on-metal acetabular component was recalled due to a higher than expected early failure rate. We evaluated the survivorship of the device and variables that may be predictive of failure at a minimum of 5-year follow-up. A single-center, single-surgeon retrospective review was conducted in patients who received the DePuy Synthes ASR™ XL Acetabular hip system from December 2005 to November 2009. Mean values and percentages were calculated and compared using the Fisher’s exact test, simple logistic regression, and Student’s t-test. The significance level was P ≤ .05. This study included 29 patients (24 males, 5 females) with 32 ASR™ XL acetabular hip systems. Mean age and body mass index (BMI) reached 55.2 years and 28.9 kg/m², respectively. Mean postoperative follow-up was 6.2 years. A total of 2 patients (6.9%) died of an unrelated cause and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom were available for follow-up. The 5-year revision rate was 34.4% (10 patients with 11 hip replacements). Mean time to revision was 3.1 years. Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with the increased rate for hip failure. Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months following the revision. This study demonstrates a high rate of failure of ASR acetabular components used in total hip arthroplasty at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is required.

Continue to: Metal-on-metal...

Metal-on-metal (MoM) articulations have been widely explored as an alternative to polyethylene bearings in total hip arthroplasty (THA), with proposed benefits including improved range of motion, lower dislocation rates, and enhanced durability.¹ Comprising cobalt and chromium, these MoM bearings gained widespread popularity in the United States, particularly in younger and more active patients looking for longer lasting devices.

The articular surface replacement (ASR) acetabular system (DePuy Synthes) was approved for sale by the US Food and Drug Administration in 2003 and implanted in an estimated 93,000 cases.² Since then, however, the early failure rate of the prosthesis has been well documented,^3-5 leading to a formal global product recall in August 2010. The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) was amongst the first to report a 6.4% rate of failure of the device at 3 years when inserted with a Corail stem.⁶ An acceptable upper rate of hip prosthesis failure is considered to reach 1% per year, with the majority of implants reporting well below this value. A 10.9% failure rate at 5 years was documented when the prosthesis was inserted for resurfacing. The National Joint Registry of England and Wales confirmed these findings and observed a 13% and 12% rate of failure at 5 years for the acetabular and resurfacing systems, respectively.² With the notable failure of the ASR system, this study reports our single-center 5-year survivorship experience and evaluates any variable that might be predictive of an early failure to aid in patient counseling.

METHODS

A single-center, single-surgeon, retrospective review of a consecutive series of patients was performed from December 2005 to November 2009. This study included all patients who underwent a primary THA with a DePuy Synthes ASR™ XL Acetabular hip system. No patients were excluded. Institutional Review Board approval was obtained. Patient demographics comprising of age, gender, and body mass index (BMI) were recorded. The primary endpoint of this study was 5-year survivorship rates. Secondary endpoints included duration to revision surgery, blood cobalt and chromium levels, time interval of blood ion tests, acetabulum size, acetabular component abduction angle, and duration to follow-up.

Candidates for the ASR™ XL Acetabular hip system included young patients and/or those considered to be physically active. In a select few, ASR devices were implanted upon patient request.

All patients underwent primary total hip replacement with a DePuy Synthes ASR™ XL uncemented acetabular component and an uncemented femoral stem (DePuy Synthes, Summit, or Tri-Lock) inserted via a standard posterior approach (Figure 1). Acetabulum sizes ranged from 52 mm to 68 mm in diameter.

All patients were followed-up yearly in the outpatient setting. Routine (yearly) metal-ion level sampling (whole blood) was started in 2010 for all patients. Laboratory tests were conducted at a single laboratory (Lab Corp.). Abduction cup inclination angles were measured by the providing surgeon using digital radiology software (GE Centricity systems).

The Student’s t-test was used to compare mean values (such as age, BMI, and metal ion levels) between the failure and no-failure groups. The 2-sided Fisher’s exact test analyzed differences in gender. Simple logistic regression analyzed variables associated with the failure group. Significance was P ≤ .05.

Continue to: Results...

RESULTS

A total of 29 patients (24 males, 5 females) with 32 ASR hip replacements were included in this study. Indications for surgery comprised osteoarthritis (28 hips, 87.5%) and avascular necrosis of the hip (4 hips, 12.5%). Mean age and BMI were 55.2 years and 28.9 kg/m², respectively. A total of 2 patients (6.9%) died of an unrelated cause (1 myocardial infarct, 1 suicide), and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom finished a 5-year minimum follow-up.

No implant failures were noted in the first year. The 5-year revision rate reached 34.4% (10 patients with 11 hip replacements). Mean time to revision for this subgroup was 3.1 years. Overall, an implant failure was observed in 37.5% of patients (11 patients with 12 hip replacements) at a mean postoperative follow-up of 6.2 years (Figure 2). Indications for implant revision were pain in 11 (92.7%) cases and infection in 1 (8.3%).

Of the 11 hips revised due to pain, 9 were performed by the original surgeon (8 were completed with primary acetabular components, 1 with a revision shell). Figure 3 shows a bilateral revision performed with primary acetabular components and retained DePuy Synthes Pinnacle femoral stems. In all these cases except 1, the ASR component was grossly loose. One case presented with pseudotumor and impingement between the femoral prosthetic neck and acetabular component after migration of a loose component. After revision, the patient returned with substantial anterior hip pain and heterotopic ossification, and failed conservative treatment, requiring another surgery with prosthesis retention, removal of heterotopic ossification, and iliopsoas lengthening. The surgery successfully relieved the symptoms. No other patients required additional surgery after their revision. In comparison to the original ASR component, the revision shell was 2 to 4 mm larger in diameter. No patient required component revision at a mean of 2.9 years after the revision surgery.

The patient with secondary revision developed a hematogenous streptococcal infection after a dental procedure performed without prophylactic antibiotics. The patient was initially lost to follow-up after the primary surgery and reported no antecedent pain prior to the revision. A substantial metal fluid collection was identified in the hip at the time of débridement and without component loosening. After débridement, the patient developed persistent metal stained wound drainage, necessitating ultimate successful treatment with a 2-stage exchange procedure.

Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with an increased rate for hip failure (Table). Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. The upper limits of blood cobalt and chromium levels reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months after the revision (Figure 4).

Table. Variables Not Associated with Early ASR Failure

Continue to: Discussion...

DISCUSSION

According to the Center for Disease Control and Prevention, 310,800 total hip replacements were performed among inpatients aged 45 years and older in the US in 2010.⁷ Specifically, in the 55- to 64-year-old age group, the number of procedures performed tripled from 2000 through 2010. As younger and more active patients opt for hip replacements, a growing need for prosthesis with enhanced durability is observed.

Despite the early proposed advantages of large head MoM bearings, our retrospective study of the DePuy Synthes ASR™ XL Acetabular hip system yielded 15.6% and 34.4% failure rates at 3 and 5 years, respectively. These higher-than-expected rates of failure are consistent with published data. The British Hip Society reported a 21% to 35% revision rate at 4 years and 49% at 6 years for the ASR XL prosthesis.⁸ In comparison, other MoM prosthesis, on average, report a 12% to 15% rate of failure at 5 years.

Considerable controversy surrounds the causes of adverse wear failure in MoM bearings.^9,10 The non-modular design of the ASR prostheses is frequently implicated as a cause of early failure. The lack of a central hole in the 1-piece component compromises the tactile feel of insertion, thereby reducing the surgeon’s ability to assess complete seating.¹¹ This condition may potentially increase the abduction angle at the time of insertion. Screw fixation of the non-modular device is not possible. The ASR XL device (148^° to 160^°) is less than a hemisphere (180^°) in size and hence features a diminished functional articular surface, further compromising implant fixation.¹¹ The functional articular surface is defined as the optimal surface area (10 mm) needed for a MoM implant.¹² Griffin and colleagues¹³ reported a 48 mm ASR XL component, when implanted at 45^° of abduction, to function similar to an implant at 59^° of abduction, leading to diminished lubrication, metallosis, and edge loading. The version of the acetabular component may similarly and adversely affect implant wear characteristics. Furthermore, the variable thickness of the implant, which is thicker at the dome and thinner at the rim, may further promote edge loading by shifting the center of rotation of the femoral head out from the center of the acetabular prosthesis.¹¹ Studies have also shown that increased wear of the MoM articulation is associated with an acetabular component inclination angle in excess of 55^°10,14 and a failure of fixation at time of implantation.¹⁵ This study, however, found no correlation between the abduction angle and risk of early implant failure for the ASR acetabular component. No correlation was also detected between the acetabulum size and revision surgery.

The AOANJRR reported loosening (44%), infection (20%), metal sensitivity (12%), fracture (9%), and dislocation of prosthesis (7%) as the indications for revision surgery for the ASR prosthesis.⁶ Furthermore, a single-center retrospective review of 70 consecutive MoM THAs with ultra-large diameter femoral head and monoblock acetabular components showed that 17.1% required revision within 3 years for loosening, pain, and squeaking.¹ Overall, 28.6% of patients reported implant dysfunction. In this study, we observed a similar rate of failure at 3 years (15.6%) for pain (11) and infection (1). The revision surgery successfully relieved all of these symptoms. One patient presented with heterotopic ossification and anterior hip pain after the original revision and required additional surgery with prosthesis retention. No patient in this series required repeat component revisions at a mean of 2.9 years after surgery. In all but 1 case, primary acetabular components were used in the revision, and in all cases except that with infection, the femoral component was retained. Replacement shells were 2 to 4 mm larger in diameter than the original ASR component.

Recently, concerns have arisen regarding the long-term effects of serum cobalt and chromium metal ions levels. Studies have shown increased serum metal ion levels,¹⁵ groin pain,¹⁶ pseudotumor formation,¹⁷ and metallosis¹⁸ after the implantation of MoM bearings. In a case study by Mao and colleagues,¹⁹ 1 patient reported headaches, anorexia, continuous metallic taste in her mouth, and weight loss. A cerebrospinal fluid analysis revealed cobalt and chromium levels at 9 and 13 nmol/L, respectively, indicating that these metal ions can cross the blood-brain barrier. Another patient reported painful muscle fatigue, night cramps, fainting spells, cognitive decline, and an inability to climb stairs. His serum cobalt level reached 258 nmol/L (reference range, 0-20 nmol/L), and chromium level totaled 88 nmol/L (reference range, 0-100 nmol/L). At 8-week follow-up after revision surgery, the symptoms of the patient had resolved, with serum cobalt levels dropping to 42 nmol/L.¹⁹ None of the patients in this study presented with any signs or symptoms of metal toxicity. The upper limits of blood cobalt and chromium levels in our study population reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. However, we noted a similar drop in post-revision blood cobalt (91% decrease) and chromium (78% decrease) levels.

In summary, our data showed a high revision rate of the DePuy Synthes ASR™ XL Acetabular hip system. Our findings are consistent with internationally published data. In the absence of reliable predictors of early failure, continued close clinical surveillance and laboratory monitoring of these patients are warranted.

CONCLUSION

This study demonstrates the high failure rate of the DePuy Synthes ASR™ XL Acetabular hip system used in THA at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is therefore required. Metal levels dropped quickly after revision, and the revision surgery can generally be performed with slightly larger primary components. Symptomatic patients with ASR hip replacements, regardless of blood metal-ion levels, were candidates for the revision surgery. Not all failed hips exhibited substantially elevated metal levels. Asymptomatic patients with high blood metal-ion levels should be closely followed-up and revision surgery should be strongly considered, consistent with recently published guidelines.²⁰

ABSTRACT

The articular surface replacement (ASR) monoblock metal-on-metal acetabular component was recalled due to a higher than expected early failure rate. We evaluated the survivorship of the device and variables that may be predictive of failure at a minimum of 5-year follow-up. A single-center, single-surgeon retrospective review was conducted in patients who received the DePuy Synthes ASR™ XL Acetabular hip system from December 2005 to November 2009. Mean values and percentages were calculated and compared using the Fisher’s exact test, simple logistic regression, and Student’s t-test. The significance level was P ≤ .05. This study included 29 patients (24 males, 5 females) with 32 ASR™ XL acetabular hip systems. Mean age and body mass index (BMI) reached 55.2 years and 28.9 kg/m², respectively. Mean postoperative follow-up was 6.2 years. A total of 2 patients (6.9%) died of an unrelated cause and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom were available for follow-up. The 5-year revision rate was 34.4% (10 patients with 11 hip replacements). Mean time to revision was 3.1 years. Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with the increased rate for hip failure. Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months following the revision. This study demonstrates a high rate of failure of ASR acetabular components used in total hip arthroplasty at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is required.

Continue to: Metal-on-metal...

Metal-on-metal (MoM) articulations have been widely explored as an alternative to polyethylene bearings in total hip arthroplasty (THA), with proposed benefits including improved range of motion, lower dislocation rates, and enhanced durability.¹ Comprising cobalt and chromium, these MoM bearings gained widespread popularity in the United States, particularly in younger and more active patients looking for longer lasting devices.

The articular surface replacement (ASR) acetabular system (DePuy Synthes) was approved for sale by the US Food and Drug Administration in 2003 and implanted in an estimated 93,000 cases.² Since then, however, the early failure rate of the prosthesis has been well documented,^3-5 leading to a formal global product recall in August 2010. The Australian Orthopaedic Association National Joint Replacement Registry (AOANJRR) was amongst the first to report a 6.4% rate of failure of the device at 3 years when inserted with a Corail stem.⁶ An acceptable upper rate of hip prosthesis failure is considered to reach 1% per year, with the majority of implants reporting well below this value. A 10.9% failure rate at 5 years was documented when the prosthesis was inserted for resurfacing. The National Joint Registry of England and Wales confirmed these findings and observed a 13% and 12% rate of failure at 5 years for the acetabular and resurfacing systems, respectively.² With the notable failure of the ASR system, this study reports our single-center 5-year survivorship experience and evaluates any variable that might be predictive of an early failure to aid in patient counseling.

METHODS

A single-center, single-surgeon, retrospective review of a consecutive series of patients was performed from December 2005 to November 2009. This study included all patients who underwent a primary THA with a DePuy Synthes ASR™ XL Acetabular hip system. No patients were excluded. Institutional Review Board approval was obtained. Patient demographics comprising of age, gender, and body mass index (BMI) were recorded. The primary endpoint of this study was 5-year survivorship rates. Secondary endpoints included duration to revision surgery, blood cobalt and chromium levels, time interval of blood ion tests, acetabulum size, acetabular component abduction angle, and duration to follow-up.

Candidates for the ASR™ XL Acetabular hip system included young patients and/or those considered to be physically active. In a select few, ASR devices were implanted upon patient request.

All patients underwent primary total hip replacement with a DePuy Synthes ASR™ XL uncemented acetabular component and an uncemented femoral stem (DePuy Synthes, Summit, or Tri-Lock) inserted via a standard posterior approach (Figure 1). Acetabulum sizes ranged from 52 mm to 68 mm in diameter.

All patients were followed-up yearly in the outpatient setting. Routine (yearly) metal-ion level sampling (whole blood) was started in 2010 for all patients. Laboratory tests were conducted at a single laboratory (Lab Corp.). Abduction cup inclination angles were measured by the providing surgeon using digital radiology software (GE Centricity systems).

The Student’s t-test was used to compare mean values (such as age, BMI, and metal ion levels) between the failure and no-failure groups. The 2-sided Fisher’s exact test analyzed differences in gender. Simple logistic regression analyzed variables associated with the failure group. Significance was P ≤ .05.

Continue to: Results...

RESULTS

A total of 29 patients (24 males, 5 females) with 32 ASR hip replacements were included in this study. Indications for surgery comprised osteoarthritis (28 hips, 87.5%) and avascular necrosis of the hip (4 hips, 12.5%). Mean age and BMI were 55.2 years and 28.9 kg/m², respectively. A total of 2 patients (6.9%) died of an unrelated cause (1 myocardial infarct, 1 suicide), and 1 patient was lost to follow-up (3.4%), leaving 26 patients with 28 hip replacements, all of whom finished a 5-year minimum follow-up.

No implant failures were noted in the first year. The 5-year revision rate reached 34.4% (10 patients with 11 hip replacements). Mean time to revision for this subgroup was 3.1 years. Overall, an implant failure was observed in 37.5% of patients (11 patients with 12 hip replacements) at a mean postoperative follow-up of 6.2 years (Figure 2). Indications for implant revision were pain in 11 (92.7%) cases and infection in 1 (8.3%).

Of the 11 hips revised due to pain, 9 were performed by the original surgeon (8 were completed with primary acetabular components, 1 with a revision shell). Figure 3 shows a bilateral revision performed with primary acetabular components and retained DePuy Synthes Pinnacle femoral stems. In all these cases except 1, the ASR component was grossly loose. One case presented with pseudotumor and impingement between the femoral prosthetic neck and acetabular component after migration of a loose component. After revision, the patient returned with substantial anterior hip pain and heterotopic ossification, and failed conservative treatment, requiring another surgery with prosthesis retention, removal of heterotopic ossification, and iliopsoas lengthening. The surgery successfully relieved the symptoms. No other patients required additional surgery after their revision. In comparison to the original ASR component, the revision shell was 2 to 4 mm larger in diameter. No patient required component revision at a mean of 2.9 years after the revision surgery.

The patient with secondary revision developed a hematogenous streptococcal infection after a dental procedure performed without prophylactic antibiotics. The patient was initially lost to follow-up after the primary surgery and reported no antecedent pain prior to the revision. A substantial metal fluid collection was identified in the hip at the time of débridement and without component loosening. After débridement, the patient developed persistent metal stained wound drainage, necessitating ultimate successful treatment with a 2-stage exchange procedure.

Age (P = .76), gender (P = .49), BMI (P = .29), acetabular component abduction angle (P = .12), and acetabulum size (P = .59) were not associated with an increased rate for hip failure (Table). Blood cobalt (7.6 vs 6.8 µg/L, P = .58) and chromium (5.0 vs 2.2 µg/L, P = .31) levels were not significantly higher in the revised group when compared with those of the unrevised group. The upper limits of blood cobalt and chromium levels reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. In the revised group, a 91% decrease in cobalt and 78% decrease in chromium levels were observed at a mean of 6 months after the revision (Figure 4).

Table. Variables Not Associated with Early ASR Failure

Continue to: Discussion...

DISCUSSION

According to the Center for Disease Control and Prevention, 310,800 total hip replacements were performed among inpatients aged 45 years and older in the US in 2010.⁷ Specifically, in the 55- to 64-year-old age group, the number of procedures performed tripled from 2000 through 2010. As younger and more active patients opt for hip replacements, a growing need for prosthesis with enhanced durability is observed.

Despite the early proposed advantages of large head MoM bearings, our retrospective study of the DePuy Synthes ASR™ XL Acetabular hip system yielded 15.6% and 34.4% failure rates at 3 and 5 years, respectively. These higher-than-expected rates of failure are consistent with published data. The British Hip Society reported a 21% to 35% revision rate at 4 years and 49% at 6 years for the ASR XL prosthesis.⁸ In comparison, other MoM prosthesis, on average, report a 12% to 15% rate of failure at 5 years.

Considerable controversy surrounds the causes of adverse wear failure in MoM bearings.^9,10 The non-modular design of the ASR prostheses is frequently implicated as a cause of early failure. The lack of a central hole in the 1-piece component compromises the tactile feel of insertion, thereby reducing the surgeon’s ability to assess complete seating.¹¹ This condition may potentially increase the abduction angle at the time of insertion. Screw fixation of the non-modular device is not possible. The ASR XL device (148^° to 160^°) is less than a hemisphere (180^°) in size and hence features a diminished functional articular surface, further compromising implant fixation.¹¹ The functional articular surface is defined as the optimal surface area (10 mm) needed for a MoM implant.¹² Griffin and colleagues¹³ reported a 48 mm ASR XL component, when implanted at 45^° of abduction, to function similar to an implant at 59^° of abduction, leading to diminished lubrication, metallosis, and edge loading. The version of the acetabular component may similarly and adversely affect implant wear characteristics. Furthermore, the variable thickness of the implant, which is thicker at the dome and thinner at the rim, may further promote edge loading by shifting the center of rotation of the femoral head out from the center of the acetabular prosthesis.¹¹ Studies have also shown that increased wear of the MoM articulation is associated with an acetabular component inclination angle in excess of 55^°10,14 and a failure of fixation at time of implantation.¹⁵ This study, however, found no correlation between the abduction angle and risk of early implant failure for the ASR acetabular component. No correlation was also detected between the acetabulum size and revision surgery.

The AOANJRR reported loosening (44%), infection (20%), metal sensitivity (12%), fracture (9%), and dislocation of prosthesis (7%) as the indications for revision surgery for the ASR prosthesis.⁶ Furthermore, a single-center retrospective review of 70 consecutive MoM THAs with ultra-large diameter femoral head and monoblock acetabular components showed that 17.1% required revision within 3 years for loosening, pain, and squeaking.¹ Overall, 28.6% of patients reported implant dysfunction. In this study, we observed a similar rate of failure at 3 years (15.6%) for pain (11) and infection (1). The revision surgery successfully relieved all of these symptoms. One patient presented with heterotopic ossification and anterior hip pain after the original revision and required additional surgery with prosthesis retention. No patient in this series required repeat component revisions at a mean of 2.9 years after surgery. In all but 1 case, primary acetabular components were used in the revision, and in all cases except that with infection, the femoral component was retained. Replacement shells were 2 to 4 mm larger in diameter than the original ASR component.

Recently, concerns have arisen regarding the long-term effects of serum cobalt and chromium metal ions levels. Studies have shown increased serum metal ion levels,¹⁵ groin pain,¹⁶ pseudotumor formation,¹⁷ and metallosis¹⁸ after the implantation of MoM bearings. In a case study by Mao and colleagues,¹⁹ 1 patient reported headaches, anorexia, continuous metallic taste in her mouth, and weight loss. A cerebrospinal fluid analysis revealed cobalt and chromium levels at 9 and 13 nmol/L, respectively, indicating that these metal ions can cross the blood-brain barrier. Another patient reported painful muscle fatigue, night cramps, fainting spells, cognitive decline, and an inability to climb stairs. His serum cobalt level reached 258 nmol/L (reference range, 0-20 nmol/L), and chromium level totaled 88 nmol/L (reference range, 0-100 nmol/L). At 8-week follow-up after revision surgery, the symptoms of the patient had resolved, with serum cobalt levels dropping to 42 nmol/L.¹⁹ None of the patients in this study presented with any signs or symptoms of metal toxicity. The upper limits of blood cobalt and chromium levels in our study population reached 18.9 and 15.9 µg/L for the revised group and 16.8 and 5.4 µg/L for the non-revised group, respectively. However, we noted a similar drop in post-revision blood cobalt (91% decrease) and chromium (78% decrease) levels.

In summary, our data showed a high revision rate of the DePuy Synthes ASR™ XL Acetabular hip system. Our findings are consistent with internationally published data. In the absence of reliable predictors of early failure, continued close clinical surveillance and laboratory monitoring of these patients are warranted.

CONCLUSION

This study demonstrates the high failure rate of the DePuy Synthes ASR™ XL Acetabular hip system used in THA at a minimum of 5 years of follow-up. No variable that was predictive of failure could be identified in this series. Close clinical surveillance of these patients is therefore required. Metal levels dropped quickly after revision, and the revision surgery can generally be performed with slightly larger primary components. Symptomatic patients with ASR hip replacements, regardless of blood metal-ion levels, were candidates for the revision surgery. Not all failed hips exhibited substantially elevated metal levels. Asymptomatic patients with high blood metal-ion levels should be closely followed-up and revision surgery should be strongly considered, consistent with recently published guidelines.²⁰

ABSTRACT

The purpose of this study is to determine the effectiveness of tranexamic acid (TXA) alone and in conjunction with a bipolar sealer in reducing postoperative transfusions during direct anterior (DA) total hip arthroplasty (THA).

In this retrospective review, we analyzed 173 consecutive patients who underwent primary unilateral DA THA performed by 2 surgeons during a 1-year period. Subjects were divided into 3 groups based on TXA use: 63 patients received TXA alone (TXA group), 49 patients received TXA in addition to a bipolar sealer (TXA + bipolar sealer group), and 61 patients received neither TXA nor a bipolar sealer (control group). Primary end points were the transfusion rate and estimated blood loss. Secondary end points were length of stay, postoperative drop in hemoglobin, and postoperative drain output.

Two patients in the TXA group and 10 patients in the control group were transfused (P = .02). In the TXA + bipolar sealer group, 1 patient was transfused (P = .02). No significant difference in the rate of transfusion was found between the TXA group and the TXA + bipolar sealer group (P = .99). Estimated blood loss was 310.3 mL ± 182.5 mL in the TXA group (P = .004), 292.9 mL ± 130.8 mL in the TXA + bipolar sealer group (P = .003), and 404.9 mL ± 201.2 mL in the control group. 

The use of TXA, with and without the concomitant use of a bipolar sealer, decreases intraoperative blood loss and postoperative transfusion requirements. The addition of a bipolar sealer, however, does not appear to provide any additional decrease in blood loss.

Historically, patients undergoing total hip arthroplasty (THA) have significant blood loss and required blood transfusions.^1-3 Blood transfusions increase not only the risk of complications but also the cost of the procedure.^4-9 Although less invasive techniques in hip surgery may decrease blood loss,^10-12 intraoperative blood loss remains a concern. Optimization of anemia and blood conservation techniques include preoperative autologous blood donation, perioperative hemodilution, meticulous surgical hemostasis, and the use of antifibrinolytic agents.^4,5,7,13,14 Antifibrinolytics are inexpensive and have been shown to reduce blood loss during THA and total knee arthroplasty (TKA).^7,15-17

Continue to: Tranexamic acid (TXA), a synthetic analog...

Although many studies have analyzed the use of TXA and a bipolar sealer with the posterior and lateral approaches to hip arthroplasty, there is a paucity of research analyzing its use in the DA approach. This study retrospectively reviews the effectiveness of TXA alone and in conjunction with a bipolar sealer in reducing allogeneic blood transfusions in DA THA.

METHODS

This is a retrospective, comparative study evaluating the efficacy of TXA with and without a bipolar sealer in unilateral DA THA. The study included 173 patients who underwent standard DA THA performed by 2 surgeons in the period April 2013 to April 2014. Patient demographic information is summarized in Table 1.

Table 1. Demographic Data

^aResult values are expressed as mean ± standard deviation (range). ^bResult values are expressed as number of cases (percentage of column header population).

Abbreviations: BMI, body mass index; TXA, tranexamic acid.

Three cohorts were created based on intraoperative blood loss management practices at the surgeon’s discretion. The first group included 63 patients who underwent DA THA with TXA but not a bipolar sealer. The second group included 49 patients who underwent DA THA with TXA and a bipolar sealer. The third (control) group included 61 patients who underwent DA THA without TXA or a bipolar sealer. Data for the control group were collected prospectively as a part of a randomized trial, which demonstrated a reduction in transfusion requirements and blood loss with the use of a bipolar sealer in DA THA.²⁸ All patients received a surgical hemovac suction drain, which was removed at 24 hours after surgery. All patients received 40 mg of enoxaparin daily for 2 weeks for venous thromboembolism prophylaxis starting the day after surgery.

All patients in the first 2 groups received 2 g of TXA administered intravenously in 2 doses: the first dose was given preoperatively, and the second dose was given immediately postoperatively in the recovery room. The bipolar sealer was utilized as needed perioperatively according to the manufacturer’s instructions to address specific bleeding targets. The common sites and steps of a DA THA, in which bleeding typically occurs, are:

• The medial femoral circumflex artery during the approach to the capsule;
• The anterior hip capsule vessels prior to capsulotomy;
• The deep branch of the medial femoral circumflex artery and the nutrient vessels to the lesser trochanter encountered while exposing the medial neck and releasing the medial capsule;
• The posterior-superior retinacular arteries encountered after femoral neck osteotomy and removal of the femoral head along the posterior capsule; and
• The branch of the obturator artery encountered during exposure of the acetabular fovea.^29-31

At the time of this study, the transfusion criteria included hemoglobin <8 g/dL in the presence of clinical symptoms.

Continue to: Primary outcome measures...

OUTCOME MEASURES AND DATA ANALYSIS

Primary outcome measures were transfusion requirements and estimated blood loss. Secondary outcome measures were postoperative decrease in hemoglobin, length of stay, and postoperative drain output. Demographic and operative data were compared between groups to ensure that there were no statistically significant differences in blood loss and transfusion requirements. All data were recorded in a password encrypted file and subsequently transferred to the REDCap system (Research Electronic Data Capture, Vanderbilt University).

STATISTICAL ANALYSIS

A priori sample size calculation was performed on the basis of a prior study ²⁸, which evaluated surgical blood loss reduction utilizing a bipolar sealer. This study suggested a sample size of 20 per group to detect the minimal clinically important difference of 1.5 (standard deviation (SD) = 1.5, α = 0.05, β = 0.20). Additionally, a general estimate for detecting a 1-unit change on an ordinal scale of 136 (SD = 1.0, α = 0.05, β = 0.20) resulted in the same number. We conservatively chose to include at least 24 patients in each study arm in the event of greater true variance. The Wilcoxon rank-sum test was used for comparison of continuous data between groups. Differences between means were analyzed using 2-sided t tests. Comparison of categorical data was performed using Pearson’s chi-square or Fisher’s exact probability test as indicated. Ordinal ranking scores were compared using the Mantel-Haenszel test.

RESULTS

There were no statistically significant differences between groups with respect to sex, age, body mass index, or preoperative hemoglobin level (Table 1). Two patients in the TXA group and 10 patients in the control group were transfused (P = .02). In the TXA + bipolar sealer group, 1 patient was transfused (P = .02). A comparison of the transfusion rate between the TXA group and the TXA + bipolar sealer group yielded no significant difference (P = .99). The estimated blood loss was 310.3 mL ± 182.5 mL in the TXA group (P = .004), 292.9 mL ± 130.8 mL in the TXA + bipolar sealer group (P = .003), and 404.9 mL ± 201.2 mL in the control group (P = .71) (Table 2).

Table 2. Patient-Related Outcomes

^aResult values are expressed as mean ± standard deviation (range). ^bResult values are expressed as number of cases (percentage of column header population).

Abbreviation: TXA, tranexamic acid.

The total drain output was 326.3 mL ± 197.5 mL in the TXA group (P < .001 for comparison with the control group), 309.8 mL ± 196.3 mL in the TXA + bipolar sealer group (P < .001 for comparison with the control group), and 473.6 mL ± 199.7 mL in the control group (P = .58). The decrease in hemoglobin was 3.5 g/dL ± 0.8 g/dL in the TXA group (P < .001), 3.5 g/dL ± 1.1 g/dL in the TXA + bipolar sealer group (P < .001), and 4.3 g/dL ± 1.2 g/dL in the control group (Table 2). The length of stay was 2.2 ± 0.6 days for the TXA group (P = .004) and 2.2 ± 0.9 days (P = .03) for the TXA + bipolar sealer group, and 2.6 ± 0.8 days in the control group (P = .78) (Table 2).

DISCUSSION

This study shows that the use of TXA alone provides a significant decrease in transfusion rates and estimated blood loss, a benefit which was not further increased with the addition of a bipolar sealer (Table 2). Many studies have demonstrated that TXA reduces blood loss and transfusion rates in patients undergoing THA and TKA.²⁹ However, TXA’s acceptance as a more readily used hemostatic medication has been hindered by the theoretically increased risk of thromboembolism in susceptible, high-risk patients.^32-35 In a 2012 meta-analysis conducted by Yang and colleagues,³⁶ the use of TXA led to significantly less blood loss per patient and fewer transfusions without leading to an increased risk of thromboembolic events.

Continue to: Similarly, the bipolar sealer...

Although the results of our research are in line with those of previous publications, it is important to address 3 limitations within this study. First, only the control group in this study was enrolled prospectively; the remaining groups were reviewed retrospectively. Second, our adoption of TXA was recent; therefore, a confounding factor is that our surgeons had more experience in the anterior approach when using TXA. Third, the established transfusion threshold of <8 g/dl for this study led to more liberal use of transfusions. Since the conclusion of this study, we have adopted stricter transfusion criteria (hemoglobin <7.0 g/dL with clinical symptoms) which has led to even lower transfusion requirements.

CONCLUSION

In the reviewed patient population, TXA decreased blood loss and transfusion requirements following DA THA. However, the addition of a bipolar sealer did not provide an advantage. The results of this study do not support the routine use of a bipolar sealer in DA THA.

ABSTRACT

The purpose of this study is to determine the effectiveness of tranexamic acid (TXA) alone and in conjunction with a bipolar sealer in reducing postoperative transfusions during direct anterior (DA) total hip arthroplasty (THA).

In this retrospective review, we analyzed 173 consecutive patients who underwent primary unilateral DA THA performed by 2 surgeons during a 1-year period. Subjects were divided into 3 groups based on TXA use: 63 patients received TXA alone (TXA group), 49 patients received TXA in addition to a bipolar sealer (TXA + bipolar sealer group), and 61 patients received neither TXA nor a bipolar sealer (control group). Primary end points were the transfusion rate and estimated blood loss. Secondary end points were length of stay, postoperative drop in hemoglobin, and postoperative drain output.

Two patients in the TXA group and 10 patients in the control group were transfused (P = .02). In the TXA + bipolar sealer group, 1 patient was transfused (P = .02). No significant difference in the rate of transfusion was found between the TXA group and the TXA + bipolar sealer group (P = .99). Estimated blood loss was 310.3 mL ± 182.5 mL in the TXA group (P = .004), 292.9 mL ± 130.8 mL in the TXA + bipolar sealer group (P = .003), and 404.9 mL ± 201.2 mL in the control group. 

The use of TXA, with and without the concomitant use of a bipolar sealer, decreases intraoperative blood loss and postoperative transfusion requirements. The addition of a bipolar sealer, however, does not appear to provide any additional decrease in blood loss.

Historically, patients undergoing total hip arthroplasty (THA) have significant blood loss and required blood transfusions.^1-3 Blood transfusions increase not only the risk of complications but also the cost of the procedure.^4-9 Although less invasive techniques in hip surgery may decrease blood loss,^10-12 intraoperative blood loss remains a concern. Optimization of anemia and blood conservation techniques include preoperative autologous blood donation, perioperative hemodilution, meticulous surgical hemostasis, and the use of antifibrinolytic agents.^4,5,7,13,14 Antifibrinolytics are inexpensive and have been shown to reduce blood loss during THA and total knee arthroplasty (TKA).^7,15-17

Continue to: Tranexamic acid (TXA), a synthetic analog...

Although many studies have analyzed the use of TXA and a bipolar sealer with the posterior and lateral approaches to hip arthroplasty, there is a paucity of research analyzing its use in the DA approach. This study retrospectively reviews the effectiveness of TXA alone and in conjunction with a bipolar sealer in reducing allogeneic blood transfusions in DA THA.

METHODS

This is a retrospective, comparative study evaluating the efficacy of TXA with and without a bipolar sealer in unilateral DA THA. The study included 173 patients who underwent standard DA THA performed by 2 surgeons in the period April 2013 to April 2014. Patient demographic information is summarized in Table 1.

Table 1. Demographic Data

^aResult values are expressed as mean ± standard deviation (range). ^bResult values are expressed as number of cases (percentage of column header population).

Abbreviations: BMI, body mass index; TXA, tranexamic acid.

Three cohorts were created based on intraoperative blood loss management practices at the surgeon’s discretion. The first group included 63 patients who underwent DA THA with TXA but not a bipolar sealer. The second group included 49 patients who underwent DA THA with TXA and a bipolar sealer. The third (control) group included 61 patients who underwent DA THA without TXA or a bipolar sealer. Data for the control group were collected prospectively as a part of a randomized trial, which demonstrated a reduction in transfusion requirements and blood loss with the use of a bipolar sealer in DA THA.²⁸ All patients received a surgical hemovac suction drain, which was removed at 24 hours after surgery. All patients received 40 mg of enoxaparin daily for 2 weeks for venous thromboembolism prophylaxis starting the day after surgery.

All patients in the first 2 groups received 2 g of TXA administered intravenously in 2 doses: the first dose was given preoperatively, and the second dose was given immediately postoperatively in the recovery room. The bipolar sealer was utilized as needed perioperatively according to the manufacturer’s instructions to address specific bleeding targets. The common sites and steps of a DA THA, in which bleeding typically occurs, are:

• The medial femoral circumflex artery during the approach to the capsule;
• The anterior hip capsule vessels prior to capsulotomy;
• The deep branch of the medial femoral circumflex artery and the nutrient vessels to the lesser trochanter encountered while exposing the medial neck and releasing the medial capsule;
• The posterior-superior retinacular arteries encountered after femoral neck osteotomy and removal of the femoral head along the posterior capsule; and
• The branch of the obturator artery encountered during exposure of the acetabular fovea.^29-31

At the time of this study, the transfusion criteria included hemoglobin <8 g/dL in the presence of clinical symptoms.

Continue to: Primary outcome measures...

OUTCOME MEASURES AND DATA ANALYSIS

Primary outcome measures were transfusion requirements and estimated blood loss. Secondary outcome measures were postoperative decrease in hemoglobin, length of stay, and postoperative drain output. Demographic and operative data were compared between groups to ensure that there were no statistically significant differences in blood loss and transfusion requirements. All data were recorded in a password encrypted file and subsequently transferred to the REDCap system (Research Electronic Data Capture, Vanderbilt University).

STATISTICAL ANALYSIS

A priori sample size calculation was performed on the basis of a prior study ²⁸, which evaluated surgical blood loss reduction utilizing a bipolar sealer. This study suggested a sample size of 20 per group to detect the minimal clinically important difference of 1.5 (standard deviation (SD) = 1.5, α = 0.05, β = 0.20). Additionally, a general estimate for detecting a 1-unit change on an ordinal scale of 136 (SD = 1.0, α = 0.05, β = 0.20) resulted in the same number. We conservatively chose to include at least 24 patients in each study arm in the event of greater true variance. The Wilcoxon rank-sum test was used for comparison of continuous data between groups. Differences between means were analyzed using 2-sided t tests. Comparison of categorical data was performed using Pearson’s chi-square or Fisher’s exact probability test as indicated. Ordinal ranking scores were compared using the Mantel-Haenszel test.

RESULTS

There were no statistically significant differences between groups with respect to sex, age, body mass index, or preoperative hemoglobin level (Table 1). Two patients in the TXA group and 10 patients in the control group were transfused (P = .02). In the TXA + bipolar sealer group, 1 patient was transfused (P = .02). A comparison of the transfusion rate between the TXA group and the TXA + bipolar sealer group yielded no significant difference (P = .99). The estimated blood loss was 310.3 mL ± 182.5 mL in the TXA group (P = .004), 292.9 mL ± 130.8 mL in the TXA + bipolar sealer group (P = .003), and 404.9 mL ± 201.2 mL in the control group (P = .71) (Table 2).

Table 2. Patient-Related Outcomes

^aResult values are expressed as mean ± standard deviation (range). ^bResult values are expressed as number of cases (percentage of column header population).

Abbreviation: TXA, tranexamic acid.

The total drain output was 326.3 mL ± 197.5 mL in the TXA group (P < .001 for comparison with the control group), 309.8 mL ± 196.3 mL in the TXA + bipolar sealer group (P < .001 for comparison with the control group), and 473.6 mL ± 199.7 mL in the control group (P = .58). The decrease in hemoglobin was 3.5 g/dL ± 0.8 g/dL in the TXA group (P < .001), 3.5 g/dL ± 1.1 g/dL in the TXA + bipolar sealer group (P < .001), and 4.3 g/dL ± 1.2 g/dL in the control group (Table 2). The length of stay was 2.2 ± 0.6 days for the TXA group (P = .004) and 2.2 ± 0.9 days (P = .03) for the TXA + bipolar sealer group, and 2.6 ± 0.8 days in the control group (P = .78) (Table 2).

DISCUSSION

This study shows that the use of TXA alone provides a significant decrease in transfusion rates and estimated blood loss, a benefit which was not further increased with the addition of a bipolar sealer (Table 2). Many studies have demonstrated that TXA reduces blood loss and transfusion rates in patients undergoing THA and TKA.²⁹ However, TXA’s acceptance as a more readily used hemostatic medication has been hindered by the theoretically increased risk of thromboembolism in susceptible, high-risk patients.^32-35 In a 2012 meta-analysis conducted by Yang and colleagues,³⁶ the use of TXA led to significantly less blood loss per patient and fewer transfusions without leading to an increased risk of thromboembolic events.

Continue to: Similarly, the bipolar sealer...

Although the results of our research are in line with those of previous publications, it is important to address 3 limitations within this study. First, only the control group in this study was enrolled prospectively; the remaining groups were reviewed retrospectively. Second, our adoption of TXA was recent; therefore, a confounding factor is that our surgeons had more experience in the anterior approach when using TXA. Third, the established transfusion threshold of <8 g/dl for this study led to more liberal use of transfusions. Since the conclusion of this study, we have adopted stricter transfusion criteria (hemoglobin <7.0 g/dL with clinical symptoms) which has led to even lower transfusion requirements.

CONCLUSION

In the reviewed patient population, TXA decreased blood loss and transfusion requirements following DA THA. However, the addition of a bipolar sealer did not provide an advantage. The results of this study do not support the routine use of a bipolar sealer in DA THA.

ABSTRACT

The purpose of this study is to determine the effectiveness of tranexamic acid (TXA) alone and in conjunction with a bipolar sealer in reducing postoperative transfusions during direct anterior (DA) total hip arthroplasty (THA).

In this retrospective review, we analyzed 173 consecutive patients who underwent primary unilateral DA THA performed by 2 surgeons during a 1-year period. Subjects were divided into 3 groups based on TXA use: 63 patients received TXA alone (TXA group), 49 patients received TXA in addition to a bipolar sealer (TXA + bipolar sealer group), and 61 patients received neither TXA nor a bipolar sealer (control group). Primary end points were the transfusion rate and estimated blood loss. Secondary end points were length of stay, postoperative drop in hemoglobin, and postoperative drain output.

Two patients in the TXA group and 10 patients in the control group were transfused (P = .02). In the TXA + bipolar sealer group, 1 patient was transfused (P = .02). No significant difference in the rate of transfusion was found between the TXA group and the TXA + bipolar sealer group (P = .99). Estimated blood loss was 310.3 mL ± 182.5 mL in the TXA group (P = .004), 292.9 mL ± 130.8 mL in the TXA + bipolar sealer group (P = .003), and 404.9 mL ± 201.2 mL in the control group. 

The use of TXA, with and without the concomitant use of a bipolar sealer, decreases intraoperative blood loss and postoperative transfusion requirements. The addition of a bipolar sealer, however, does not appear to provide any additional decrease in blood loss.

Historically, patients undergoing total hip arthroplasty (THA) have significant blood loss and required blood transfusions.^1-3 Blood transfusions increase not only the risk of complications but also the cost of the procedure.^4-9 Although less invasive techniques in hip surgery may decrease blood loss,^10-12 intraoperative blood loss remains a concern. Optimization of anemia and blood conservation techniques include preoperative autologous blood donation, perioperative hemodilution, meticulous surgical hemostasis, and the use of antifibrinolytic agents.^4,5,7,13,14 Antifibrinolytics are inexpensive and have been shown to reduce blood loss during THA and total knee arthroplasty (TKA).^7,15-17

Continue to: Tranexamic acid (TXA), a synthetic analog...

Although many studies have analyzed the use of TXA and a bipolar sealer with the posterior and lateral approaches to hip arthroplasty, there is a paucity of research analyzing its use in the DA approach. This study retrospectively reviews the effectiveness of TXA alone and in conjunction with a bipolar sealer in reducing allogeneic blood transfusions in DA THA.

METHODS

This is a retrospective, comparative study evaluating the efficacy of TXA with and without a bipolar sealer in unilateral DA THA. The study included 173 patients who underwent standard DA THA performed by 2 surgeons in the period April 2013 to April 2014. Patient demographic information is summarized in Table 1.

Table 1. Demographic Data