Margaret C. Fang, MD, MPH

The population of patients in the US with limited English proficiency (LEP)those who speak English less than very well1is substantial and continues to grow.1, 2 Patients with LEP are at risk for lower quality health care overall than their English‐speaking counterparts.38 Professional in‐person interpreters greatly improve spoken communication and quality of care for these patients,4, 9 but their assistance is typically based on the clinical encounter. Particularly if interpreting by phone, interpreters are unlikely to be able to help with materials such as discharge instructions or information sheets meant for family members. Professional written translations of patient educational material help to bridge this gap, allowing clinicians to convey detailed written instructions to patients. However, professional translations must be prepared well in advance of any encounter and can only be used for easily anticipated problems.

The need to translate less common, patient‐specific instructions arises spontaneously in clinical practice, and formally prepared written translations are not useful in these situations. Online translation tools such as GoogleTranslate (available at http://translate.google.com/#) and Babelfish (available at http://babelfish.yahoo.com), a subset of machine translation technology, may help supplement professional in‐person interpretation and formal written translations in that they are ubiquitous, inexpensive, and increasingly well‐known and easy to use.10, 11 Machine translation has already been used in situations where in‐person interpretation is limited. For example, after the earthquake in Haiti, Creole interpreters were not widely available and a hand‐held translation application was quickly developed to meet the needs of relief workers and the population.11 However, data on the accuracy of these tools for critical clinical applications such as patient education are limited. A recent study of computer‐translated pharmacy labels suggested computer‐generated translations were frequently erratic, nonsensical, and even dangerous.12

We conducted a pilot evaluation of an online translation tool as it relates to detailed, complex patient educational material. Our primary goal was to compare the accuracy of a Spanish translation generated by the online tool to that done by a professional agency. Our secondary goals were: 1) to assess whether sentence word length or complexity mediated the accuracy of GT; and 2) to lay the foundation for a more comprehensive study of the accuracy of online translation tools, with respect to patient educational material.

Methods

Manual Evaluation: Evaluators, Domains, Scoring

We recruited three nonclinician, bilingual, nativeSpanish‐speaking research assistants as evaluators. The evaluators were all college educated with a Bachelor's degree or higher and were of Mexican, Nicaraguan, and Guatemalan ancestry. Each evaluator received a brief orientation regarding the project, as well as an explanation of the scores, and then proceeded to the blinded evaluation independently.

We asked evaluators to score sentences on Likert scales along five primary domains: fluency, adequacy, meaning, severity, and preference. Fluency and adequacy are well accepted components of machine translation evaluation,18 with fluency being an assessment of grammar and readability ranging from 5 (Perfect fluency; like reading a newspaper) to 1 (No fluency; no appreciable grammar, not understandable) and adequacy being an assessment of information preservation ranging from 5 (100% of information conveyed from the original) to 1 (0% of information conveyed from the original). Given that a sentence can be highly adequate but drastically change the connotation and intent of the sentence (eg, a sentence that contains 75% of the correct words but changes a sentence from take this medication twice a day to take this medication once every two days), we asked evaluators to assess meaning, a measure of connotation and intent maintenance, with scores ranging from 5 (Same meaning as original) to 1 (Totally different meaning from the original).19 Evaluators also assessed severity, a new measure of potential harm if a given sentence was assessed as having errors of any kind, ranging from 5 (Error, no effect on patient care) to 1 (Error, dangerous to patient) with an additional option of N/A (Sentence basically accurate). Finally, evaluators rated a blinded preference (also a new measure) for either of two translated sentences, ranging from Strongly prefer translation #1 to Strongly prefer translation #2. The order of the sentences was random (eg, sometimes the professional translation was first and sometimes the GT translation was). We subsequently converted this to preference for the professional translation, ranging from 5 (Strongly prefer the professional translation) to 1 (Strongly prefer the GT translation) in order to standardize the responses (Figures 1 and 2).

Figure 1

Figure 2

The overall flow of the study is given in Figure 3. Each evaluator initially scored 20 sentences translated by GT and 10 sentences translated professionally along the first four domains. All 30 of these sentences were randomly selected from the original, 263‐sentence pamphlet. For fluency, evaluators had access only to the translated sentence to be scored; for adequacy, meaning, and severity, they had access to both the translated sentence and the original English sentence. Ten of the 30 sentences were further selected randomly for scoring on the preference domain. For these 10 sentences, evaluators compared the GT and professional translations of the same sentence (with the original English sentence available as a reference) and indicated a preference, for any reason, for one translation or the other. Evaluators were blinded to the technique of translation (GT or professional) for all scored sentences and domains. We chose twice as many sentences from the GT preparations for the first four domains to maximize measurements for the translation technology we were evaluating, with the smaller number of professional translations serving as controls.

Figure 3

After scoring the first 30 sentences, evaluators met with one of the authors (R.R.K.) to discuss and consolidate their approach to scoring. They then scored an additional 10 GT‐translated sentences and 5 professionally translated sentences for the first four domains, and 9 of these 15 sentences for preference, to see if the meeting changed their scoring approach. These sentences were selected randomly from the original, 263‐sentence pamphlet, excluding the 30 evaluated in the previous step.

Results

Discussion

In this preliminary study comparing the accuracy of GT to professional translation for patient educational material, we found that GT was inferior to the professional translation in grammatical fluency but generally preserved the content and sense of the original text. Out of 30 GT sentences assessed, there was one substantially erroneous translation that was considered potentially dangerous. Evaluators preferred the professionally translated sentences for complex sentences, but when the English source sentence was simplecontaining a single clausethis preference disappeared.

Like Sharif and Tse,12 we found that for information not arranged in sentences, automated translation sometimes produced nonsensical sentences. In our study, these resulted from an English sentence fragment followed by a bulleted list; in their study, the nonsensical translations resulted from pharmacy labels. The difference in frequency of these errors between our studies may have resulted partly from the translation tool evaluated (GT vs programs used by pharmacies in the Bronx), but may have also been due to our use of machine translation for complete sentencesthe purpose for which it is optimally designed. The hypothesis that machine translations of clinical information are most understandable when used for simple, complete sentences concurs with the methodology used by these tools and requires further study.

GT has the potential to be very useful to clinicians, particularly for those instances when the communication required is both spontaneous and routine or noncritical. For example, in the inpatient setting, patients could communicate diet and other nonclinical requests, as well as ask or answer simple, short questions when the interpreter is not available. In such situations, the low cost and ease of using online translations and machine translation more generally may help to circumvent the tendency of clinicians to get by with inadequate language skills or to avoid communication altogether.25 If used wisely, GT and other online tools could supplement the use of standardized translations and professional interpreters in helping clinicians to overcome language barriers and linguistic inertia, though this will require further assessment.

Ours is a pilot study, and while it suggests a more promising way to use online translation tools, significant further evaluation is required regarding accuracy and applicability prior to widespread use of any machine translation tools for patient care. The document we utilized for evaluation was a professionally translated patient educational brochure provided to individuals starting a complex medication. As online translation tools would most likely not be used in this setting, but rather for spontaneous and less critical patient‐specific instructions, further testing of GT as applied to such scenarios should be considered. Second, we only evaluated GT for English translated into Spanish; its usefulness in other languages will need to be evaluated. It also remains to be seen how easily GT translations will be understood by patients, who may have variable medical understanding and educational attainment as compared to our evaluators. Finally, in this evaluation, we only assessed automated written translation, not automated spoken translation services such as those now available on cellular phones and other mobile devices.11 The latter are based upon translation software with an additional speech recognition interface. These applications may prove to be even more useful than online translation, but the speech recognition component will add an additional layer of potential error and these applications will need to be evaluated on their own merits.

The domains chosen for this study had only moderate interrater reliability as assessed by intraclass correlation and repeatability, with meaning and preference scoring particularly poorly. The latter domains in particular will require more thorough assessment before routine use in online translation assessment. The variability in all domains may have resulted partly from the choice of nonclinicians of different ancestral backgrounds as evaluators. However, this variability is likely better representative of the wide range of patient backgrounds. Because our evaluators were not professional translators, we asked a professional interpreter to grade all sentences to assess the quality of their evaluation. While the interpreter noted slightly fewer errors among the professionally translated sentences (13% vs 22%) and slightly more errors among the GT‐translated sentences (50% vs 39%), and preferred the professional translation slightly more (3.8 vs 3.2), his scores for all of the other measures were almost identical, increasing our confidence in our primary findings (Appendix A). Additionally, since statistical translation is conducted sentence by sentence, in our study evaluators only scored translations at the sentence level. The accuracy of GT for whole paragraphs or entire documents will need to be assessed separately. The correlation between METEOR and the manual evaluation scores was less than in prior studies; while inexpensive to assess, METEOR will have to be recalibrated in optimal circumstanceswith several reference translations available rather than just onebefore it can be used to supplement the assessment of new languages, new materials, other translation technologies, and improvements in a given technology over time for patient educational material.

In summary, GT scored worse in grammar but similarly in content and sense to the professional translation, committing one critical error in translating a complex, fragmented sentence as nonsense. We believe that, with further study and judicious use, GT has the potential to substantially improve clinicians' communication with patients with limited English proficiency in the area of brief spontaneous patient‐specific information, supplementing well the role that professional spoken interpretation and standardized written translations already play.

The population of patients in the US with limited English proficiency (LEP)those who speak English less than very well1is substantial and continues to grow.1, 2 Patients with LEP are at risk for lower quality health care overall than their English‐speaking counterparts.38 Professional in‐person interpreters greatly improve spoken communication and quality of care for these patients,4, 9 but their assistance is typically based on the clinical encounter. Particularly if interpreting by phone, interpreters are unlikely to be able to help with materials such as discharge instructions or information sheets meant for family members. Professional written translations of patient educational material help to bridge this gap, allowing clinicians to convey detailed written instructions to patients. However, professional translations must be prepared well in advance of any encounter and can only be used for easily anticipated problems.

The need to translate less common, patient‐specific instructions arises spontaneously in clinical practice, and formally prepared written translations are not useful in these situations. Online translation tools such as GoogleTranslate (available at http://translate.google.com/#) and Babelfish (available at http://babelfish.yahoo.com), a subset of machine translation technology, may help supplement professional in‐person interpretation and formal written translations in that they are ubiquitous, inexpensive, and increasingly well‐known and easy to use.10, 11 Machine translation has already been used in situations where in‐person interpretation is limited. For example, after the earthquake in Haiti, Creole interpreters were not widely available and a hand‐held translation application was quickly developed to meet the needs of relief workers and the population.11 However, data on the accuracy of these tools for critical clinical applications such as patient education are limited. A recent study of computer‐translated pharmacy labels suggested computer‐generated translations were frequently erratic, nonsensical, and even dangerous.12

We conducted a pilot evaluation of an online translation tool as it relates to detailed, complex patient educational material. Our primary goal was to compare the accuracy of a Spanish translation generated by the online tool to that done by a professional agency. Our secondary goals were: 1) to assess whether sentence word length or complexity mediated the accuracy of GT; and 2) to lay the foundation for a more comprehensive study of the accuracy of online translation tools, with respect to patient educational material.

Methods

Manual Evaluation: Evaluators, Domains, Scoring

We recruited three nonclinician, bilingual, nativeSpanish‐speaking research assistants as evaluators. The evaluators were all college educated with a Bachelor's degree or higher and were of Mexican, Nicaraguan, and Guatemalan ancestry. Each evaluator received a brief orientation regarding the project, as well as an explanation of the scores, and then proceeded to the blinded evaluation independently.

We asked evaluators to score sentences on Likert scales along five primary domains: fluency, adequacy, meaning, severity, and preference. Fluency and adequacy are well accepted components of machine translation evaluation,18 with fluency being an assessment of grammar and readability ranging from 5 (Perfect fluency; like reading a newspaper) to 1 (No fluency; no appreciable grammar, not understandable) and adequacy being an assessment of information preservation ranging from 5 (100% of information conveyed from the original) to 1 (0% of information conveyed from the original). Given that a sentence can be highly adequate but drastically change the connotation and intent of the sentence (eg, a sentence that contains 75% of the correct words but changes a sentence from take this medication twice a day to take this medication once every two days), we asked evaluators to assess meaning, a measure of connotation and intent maintenance, with scores ranging from 5 (Same meaning as original) to 1 (Totally different meaning from the original).19 Evaluators also assessed severity, a new measure of potential harm if a given sentence was assessed as having errors of any kind, ranging from 5 (Error, no effect on patient care) to 1 (Error, dangerous to patient) with an additional option of N/A (Sentence basically accurate). Finally, evaluators rated a blinded preference (also a new measure) for either of two translated sentences, ranging from Strongly prefer translation #1 to Strongly prefer translation #2. The order of the sentences was random (eg, sometimes the professional translation was first and sometimes the GT translation was). We subsequently converted this to preference for the professional translation, ranging from 5 (Strongly prefer the professional translation) to 1 (Strongly prefer the GT translation) in order to standardize the responses (Figures 1 and 2).

Figure 1

Figure 2

The overall flow of the study is given in Figure 3. Each evaluator initially scored 20 sentences translated by GT and 10 sentences translated professionally along the first four domains. All 30 of these sentences were randomly selected from the original, 263‐sentence pamphlet. For fluency, evaluators had access only to the translated sentence to be scored; for adequacy, meaning, and severity, they had access to both the translated sentence and the original English sentence. Ten of the 30 sentences were further selected randomly for scoring on the preference domain. For these 10 sentences, evaluators compared the GT and professional translations of the same sentence (with the original English sentence available as a reference) and indicated a preference, for any reason, for one translation or the other. Evaluators were blinded to the technique of translation (GT or professional) for all scored sentences and domains. We chose twice as many sentences from the GT preparations for the first four domains to maximize measurements for the translation technology we were evaluating, with the smaller number of professional translations serving as controls.

Figure 3

After scoring the first 30 sentences, evaluators met with one of the authors (R.R.K.) to discuss and consolidate their approach to scoring. They then scored an additional 10 GT‐translated sentences and 5 professionally translated sentences for the first four domains, and 9 of these 15 sentences for preference, to see if the meeting changed their scoring approach. These sentences were selected randomly from the original, 263‐sentence pamphlet, excluding the 30 evaluated in the previous step.

Results

Discussion

In this preliminary study comparing the accuracy of GT to professional translation for patient educational material, we found that GT was inferior to the professional translation in grammatical fluency but generally preserved the content and sense of the original text. Out of 30 GT sentences assessed, there was one substantially erroneous translation that was considered potentially dangerous. Evaluators preferred the professionally translated sentences for complex sentences, but when the English source sentence was simplecontaining a single clausethis preference disappeared.

Like Sharif and Tse,12 we found that for information not arranged in sentences, automated translation sometimes produced nonsensical sentences. In our study, these resulted from an English sentence fragment followed by a bulleted list; in their study, the nonsensical translations resulted from pharmacy labels. The difference in frequency of these errors between our studies may have resulted partly from the translation tool evaluated (GT vs programs used by pharmacies in the Bronx), but may have also been due to our use of machine translation for complete sentencesthe purpose for which it is optimally designed. The hypothesis that machine translations of clinical information are most understandable when used for simple, complete sentences concurs with the methodology used by these tools and requires further study.

GT has the potential to be very useful to clinicians, particularly for those instances when the communication required is both spontaneous and routine or noncritical. For example, in the inpatient setting, patients could communicate diet and other nonclinical requests, as well as ask or answer simple, short questions when the interpreter is not available. In such situations, the low cost and ease of using online translations and machine translation more generally may help to circumvent the tendency of clinicians to get by with inadequate language skills or to avoid communication altogether.25 If used wisely, GT and other online tools could supplement the use of standardized translations and professional interpreters in helping clinicians to overcome language barriers and linguistic inertia, though this will require further assessment.

Ours is a pilot study, and while it suggests a more promising way to use online translation tools, significant further evaluation is required regarding accuracy and applicability prior to widespread use of any machine translation tools for patient care. The document we utilized for evaluation was a professionally translated patient educational brochure provided to individuals starting a complex medication. As online translation tools would most likely not be used in this setting, but rather for spontaneous and less critical patient‐specific instructions, further testing of GT as applied to such scenarios should be considered. Second, we only evaluated GT for English translated into Spanish; its usefulness in other languages will need to be evaluated. It also remains to be seen how easily GT translations will be understood by patients, who may have variable medical understanding and educational attainment as compared to our evaluators. Finally, in this evaluation, we only assessed automated written translation, not automated spoken translation services such as those now available on cellular phones and other mobile devices.11 The latter are based upon translation software with an additional speech recognition interface. These applications may prove to be even more useful than online translation, but the speech recognition component will add an additional layer of potential error and these applications will need to be evaluated on their own merits.

The domains chosen for this study had only moderate interrater reliability as assessed by intraclass correlation and repeatability, with meaning and preference scoring particularly poorly. The latter domains in particular will require more thorough assessment before routine use in online translation assessment. The variability in all domains may have resulted partly from the choice of nonclinicians of different ancestral backgrounds as evaluators. However, this variability is likely better representative of the wide range of patient backgrounds. Because our evaluators were not professional translators, we asked a professional interpreter to grade all sentences to assess the quality of their evaluation. While the interpreter noted slightly fewer errors among the professionally translated sentences (13% vs 22%) and slightly more errors among the GT‐translated sentences (50% vs 39%), and preferred the professional translation slightly more (3.8 vs 3.2), his scores for all of the other measures were almost identical, increasing our confidence in our primary findings (Appendix A). Additionally, since statistical translation is conducted sentence by sentence, in our study evaluators only scored translations at the sentence level. The accuracy of GT for whole paragraphs or entire documents will need to be assessed separately. The correlation between METEOR and the manual evaluation scores was less than in prior studies; while inexpensive to assess, METEOR will have to be recalibrated in optimal circumstanceswith several reference translations available rather than just onebefore it can be used to supplement the assessment of new languages, new materials, other translation technologies, and improvements in a given technology over time for patient educational material.

In summary, GT scored worse in grammar but similarly in content and sense to the professional translation, committing one critical error in translating a complex, fragmented sentence as nonsense. We believe that, with further study and judicious use, GT has the potential to substantially improve clinicians' communication with patients with limited English proficiency in the area of brief spontaneous patient‐specific information, supplementing well the role that professional spoken interpretation and standardized written translations already play.

The population of patients in the US with limited English proficiency (LEP)those who speak English less than very well1is substantial and continues to grow.1, 2 Patients with LEP are at risk for lower quality health care overall than their English‐speaking counterparts.38 Professional in‐person interpreters greatly improve spoken communication and quality of care for these patients,4, 9 but their assistance is typically based on the clinical encounter. Particularly if interpreting by phone, interpreters are unlikely to be able to help with materials such as discharge instructions or information sheets meant for family members. Professional written translations of patient educational material help to bridge this gap, allowing clinicians to convey detailed written instructions to patients. However, professional translations must be prepared well in advance of any encounter and can only be used for easily anticipated problems.

The need to translate less common, patient‐specific instructions arises spontaneously in clinical practice, and formally prepared written translations are not useful in these situations. Online translation tools such as GoogleTranslate (available at http://translate.google.com/#) and Babelfish (available at http://babelfish.yahoo.com), a subset of machine translation technology, may help supplement professional in‐person interpretation and formal written translations in that they are ubiquitous, inexpensive, and increasingly well‐known and easy to use.10, 11 Machine translation has already been used in situations where in‐person interpretation is limited. For example, after the earthquake in Haiti, Creole interpreters were not widely available and a hand‐held translation application was quickly developed to meet the needs of relief workers and the population.11 However, data on the accuracy of these tools for critical clinical applications such as patient education are limited. A recent study of computer‐translated pharmacy labels suggested computer‐generated translations were frequently erratic, nonsensical, and even dangerous.12

We conducted a pilot evaluation of an online translation tool as it relates to detailed, complex patient educational material. Our primary goal was to compare the accuracy of a Spanish translation generated by the online tool to that done by a professional agency. Our secondary goals were: 1) to assess whether sentence word length or complexity mediated the accuracy of GT; and 2) to lay the foundation for a more comprehensive study of the accuracy of online translation tools, with respect to patient educational material.

Methods

Manual Evaluation: Evaluators, Domains, Scoring

We recruited three nonclinician, bilingual, nativeSpanish‐speaking research assistants as evaluators. The evaluators were all college educated with a Bachelor's degree or higher and were of Mexican, Nicaraguan, and Guatemalan ancestry. Each evaluator received a brief orientation regarding the project, as well as an explanation of the scores, and then proceeded to the blinded evaluation independently.

We asked evaluators to score sentences on Likert scales along five primary domains: fluency, adequacy, meaning, severity, and preference. Fluency and adequacy are well accepted components of machine translation evaluation,18 with fluency being an assessment of grammar and readability ranging from 5 (Perfect fluency; like reading a newspaper) to 1 (No fluency; no appreciable grammar, not understandable) and adequacy being an assessment of information preservation ranging from 5 (100% of information conveyed from the original) to 1 (0% of information conveyed from the original). Given that a sentence can be highly adequate but drastically change the connotation and intent of the sentence (eg, a sentence that contains 75% of the correct words but changes a sentence from take this medication twice a day to take this medication once every two days), we asked evaluators to assess meaning, a measure of connotation and intent maintenance, with scores ranging from 5 (Same meaning as original) to 1 (Totally different meaning from the original).19 Evaluators also assessed severity, a new measure of potential harm if a given sentence was assessed as having errors of any kind, ranging from 5 (Error, no effect on patient care) to 1 (Error, dangerous to patient) with an additional option of N/A (Sentence basically accurate). Finally, evaluators rated a blinded preference (also a new measure) for either of two translated sentences, ranging from Strongly prefer translation #1 to Strongly prefer translation #2. The order of the sentences was random (eg, sometimes the professional translation was first and sometimes the GT translation was). We subsequently converted this to preference for the professional translation, ranging from 5 (Strongly prefer the professional translation) to 1 (Strongly prefer the GT translation) in order to standardize the responses (Figures 1 and 2).

Figure 1

Figure 2

The overall flow of the study is given in Figure 3. Each evaluator initially scored 20 sentences translated by GT and 10 sentences translated professionally along the first four domains. All 30 of these sentences were randomly selected from the original, 263‐sentence pamphlet. For fluency, evaluators had access only to the translated sentence to be scored; for adequacy, meaning, and severity, they had access to both the translated sentence and the original English sentence. Ten of the 30 sentences were further selected randomly for scoring on the preference domain. For these 10 sentences, evaluators compared the GT and professional translations of the same sentence (with the original English sentence available as a reference) and indicated a preference, for any reason, for one translation or the other. Evaluators were blinded to the technique of translation (GT or professional) for all scored sentences and domains. We chose twice as many sentences from the GT preparations for the first four domains to maximize measurements for the translation technology we were evaluating, with the smaller number of professional translations serving as controls.

Figure 3

After scoring the first 30 sentences, evaluators met with one of the authors (R.R.K.) to discuss and consolidate their approach to scoring. They then scored an additional 10 GT‐translated sentences and 5 professionally translated sentences for the first four domains, and 9 of these 15 sentences for preference, to see if the meeting changed their scoring approach. These sentences were selected randomly from the original, 263‐sentence pamphlet, excluding the 30 evaluated in the previous step.

Results

Discussion

In this preliminary study comparing the accuracy of GT to professional translation for patient educational material, we found that GT was inferior to the professional translation in grammatical fluency but generally preserved the content and sense of the original text. Out of 30 GT sentences assessed, there was one substantially erroneous translation that was considered potentially dangerous. Evaluators preferred the professionally translated sentences for complex sentences, but when the English source sentence was simplecontaining a single clausethis preference disappeared.

Like Sharif and Tse,12 we found that for information not arranged in sentences, automated translation sometimes produced nonsensical sentences. In our study, these resulted from an English sentence fragment followed by a bulleted list; in their study, the nonsensical translations resulted from pharmacy labels. The difference in frequency of these errors between our studies may have resulted partly from the translation tool evaluated (GT vs programs used by pharmacies in the Bronx), but may have also been due to our use of machine translation for complete sentencesthe purpose for which it is optimally designed. The hypothesis that machine translations of clinical information are most understandable when used for simple, complete sentences concurs with the methodology used by these tools and requires further study.

GT has the potential to be very useful to clinicians, particularly for those instances when the communication required is both spontaneous and routine or noncritical. For example, in the inpatient setting, patients could communicate diet and other nonclinical requests, as well as ask or answer simple, short questions when the interpreter is not available. In such situations, the low cost and ease of using online translations and machine translation more generally may help to circumvent the tendency of clinicians to get by with inadequate language skills or to avoid communication altogether.25 If used wisely, GT and other online tools could supplement the use of standardized translations and professional interpreters in helping clinicians to overcome language barriers and linguistic inertia, though this will require further assessment.

Ours is a pilot study, and while it suggests a more promising way to use online translation tools, significant further evaluation is required regarding accuracy and applicability prior to widespread use of any machine translation tools for patient care. The document we utilized for evaluation was a professionally translated patient educational brochure provided to individuals starting a complex medication. As online translation tools would most likely not be used in this setting, but rather for spontaneous and less critical patient‐specific instructions, further testing of GT as applied to such scenarios should be considered. Second, we only evaluated GT for English translated into Spanish; its usefulness in other languages will need to be evaluated. It also remains to be seen how easily GT translations will be understood by patients, who may have variable medical understanding and educational attainment as compared to our evaluators. Finally, in this evaluation, we only assessed automated written translation, not automated spoken translation services such as those now available on cellular phones and other mobile devices.11 The latter are based upon translation software with an additional speech recognition interface. These applications may prove to be even more useful than online translation, but the speech recognition component will add an additional layer of potential error and these applications will need to be evaluated on their own merits.

The domains chosen for this study had only moderate interrater reliability as assessed by intraclass correlation and repeatability, with meaning and preference scoring particularly poorly. The latter domains in particular will require more thorough assessment before routine use in online translation assessment. The variability in all domains may have resulted partly from the choice of nonclinicians of different ancestral backgrounds as evaluators. However, this variability is likely better representative of the wide range of patient backgrounds. Because our evaluators were not professional translators, we asked a professional interpreter to grade all sentences to assess the quality of their evaluation. While the interpreter noted slightly fewer errors among the professionally translated sentences (13% vs 22%) and slightly more errors among the GT‐translated sentences (50% vs 39%), and preferred the professional translation slightly more (3.8 vs 3.2), his scores for all of the other measures were almost identical, increasing our confidence in our primary findings (Appendix A). Additionally, since statistical translation is conducted sentence by sentence, in our study evaluators only scored translations at the sentence level. The accuracy of GT for whole paragraphs or entire documents will need to be assessed separately. The correlation between METEOR and the manual evaluation scores was less than in prior studies; while inexpensive to assess, METEOR will have to be recalibrated in optimal circumstanceswith several reference translations available rather than just onebefore it can be used to supplement the assessment of new languages, new materials, other translation technologies, and improvements in a given technology over time for patient educational material.

In summary, GT scored worse in grammar but similarly in content and sense to the professional translation, committing one critical error in translating a complex, fragmented sentence as nonsense. We believe that, with further study and judicious use, GT has the potential to substantially improve clinicians' communication with patients with limited English proficiency in the area of brief spontaneous patient‐specific information, supplementing well the role that professional spoken interpretation and standardized written translations already play.

Recombinant tissue plasminogen activator (tPA), approved for use in the United States for the treatment for acute ischemic stroke since 1996, improves overall outcomes from ischemic stroke when administered to selected patients.14 Several prominent guidelines, including the Brain Attack Coalition and the American Stroke Association, have recommended increasing the use of tPA for acute ischemic stroke.57 In addition, in 2003 the Joint Commission on Accreditation of Healthcare Organizations developed a disease‐specific certification program to designate certain institutions Primary Stroke Centers, with one of the performance measures being the availability of thrombolysis.8

Despite guidelines and regulatory agencies promoting the use of thrombolysis for ischemic stroke, previous studies have shown disappointingly low rates of use.912 The goals of this study were to assess whether national trends in the use of thrombolysis for acute ischemic stroke have increased in light of increased regulatory activity as well as to identify patient characteristics associated with thrombolytic administration.

Materials and Methods

Data for this study were obtained from the 2001 through 2006 National Hospital Discharge Survey (NHDS), a nationally representative sample of inpatient hospitalizations conducted annually by the National Center for Health Statistics.13 The NHDS collects data on approximately 300,000 patients from about 500 short‐stay nonfederal hospitals in the United States and uses a 3‐stage sampling strategy that allows for extrapolation to national level estimates. Response rates typically exceed 90% from participating hospitals. The survey collects demographic data, including age, sex, race, hospital geographic region, hospital bedsize and patient insurance status. In addition, up to 7 diagnostic and 4 procedural codes from the hospitalization are available, as is hospitalization length of stay and patient discharge disposition. No information on timing of symptoms, degree of neurologic compromise, or results of imaging tests were available in the NHDS.

We searched for all patients age 18 years or older with a primary diagnostic code of ischemic stroke using the International Classification of Diseases, 9th Edition, Clinical Modification (ICD‐9‐CM) codes 433, 434, 436, 437.0, and 437.1, excluding codes with a fifth digit of 0 (which indicated arterial occlusion without mention of cerebral infarction). We then searched for the presence of an ICD‐9‐CM procedure code for injection or infusion of thrombolytic agent (code 99.10). Specific comorbid conditions associated with ischemic stroke were identified by searching for specific ICD‐9‐CM codes, including for heart failure, coronary artery disease, hypertension, diabetes mellitus, and atrial fibrillation. To provide a general assessment of the severity of illness of the patients, we calculated an adapted Charlson comorbidity score for each patient using available secondary discharge diagnosis codes.14 We also searched for codes corresponding to intracranial hemorrhage, a complication associated with tPA administration.

Results

From years 2001 through 2006, we identified 22,842 patients with a primary diagnosis of ischemic stroke. Of these, 313 (1.37%, 95% confidence interval [CI], 1.22‐1.53%) had a procedure code for injection or infusion of thrombolysis. Using NHDS sample weights, these numbers corresponded to an estimated 2.55 million hospitalizations for ischemic stroke in the United States during the time period and to 35,082 patients receiving intravenous thrombolytics. Although the overall rate of thrombolysis administration was quite low overall, the administration rate increased over time, from 0.87% [95% CI, 0.61‐1.22%] of stroke patients in year 2001 to 2.40% [95% CI, 1.95‐2.93%] in year 2006 and with a particular increase especially noted after year 2003 (P <0.001 for trend, Figure 1).

Figure 1

On bivariate analysis, a lower proportion of African‐American patients received tPA compared to white patients (0.8% vs. 1.5%, P = 0.003), while a higher proportion of patients with atrial fibrillation received tPA (2.3% vs. 1.2%, P < 0.001). Older patients were less likely than younger patients to receive tPA (Table 1). The rate of intracranial hemorrhage was significantly higher in patients who received tPA (5.4% vs. 0.6%, P < 0.001) and the overall inpatient mortality in patients who received tPA was 9.0%. Mortality in patients receiving tPA continued to be higher than in patients who did not receive tPA even when patients with intracranial hemorrhage were excluded (8.1% vs. 5.3%, P < 0.001). Larger hospitals were more likely to administer tPA to patients with ischemic stroke, with a 1.79% administration rate in hospitals with 300 beds compared to 0.90% in hospitals with 100 to 199 beds and 0.52% in hospitals with 6 to 99 beds (P < 0.001).

After adjusting for patient and hospital characteristics, the absolute rate of thrombolysis administration increased by an average of 0.19% per year (95% CI, 0.12‐0.26%). Factors that were significantly associated with administration of thrombolytics included being hospitalized in a larger hospital, having a history of atrial fibrillation, and a higher Charlson comorbidity index (Table 2). Patients aged 80 years or older, African American patients, and those with diabetes mellitus were significantly less likely to receive thrombolysis.

Discussion

Despite strong recommendations from guidelines and regulatory agencies, national rates of intravenous thrombolysis for ischemic stroke continue to be quite low overall. However, tPA administration appears to have increased from previous years and particularly increased in years after the Joint Commission began to accredit institutions as Primary Stroke Centers.11 The oldest patients and African Americans were less likely to receive thrombolytics, while patients with atrial fibrillation were more likely to receive thrombolysis, potentially related to atrial fibrillation causing more severe strokes.15 A total of 5.4% of patients who received tPA were diagnosed with intracranial hemorrhage, and the inpatient mortality rate of patients with tPA was 9.0%.

The exact optimal rate of thrombolysis administration for the patients in our study is unknown, as the NHDS database lacked detailed information on factors that would preclude tPA administration such as late timing of presentation and mild stroke symptoms.3 Studies conducted in stroke registries and regional settings have found that only approximately 15% to 32% of patients presenting with ischemic stroke arrive within 3 hours of symptom onset, and of these, only about 40% to 50% are eligible for tPA clinically.9, 10, 1619 However, even among presumed eligible patients, tPA administration rates only range between 25% and 43%,17, 19, 20 and the ideal rate is likely to be higher than the very low rates we observed in our study. Newer evidence that extending the time window where tPA may be given safely may increase the number of eligible patients.21

Patients who received thrombolysis had higher mortality rates than patients who did not. Although we were unable to determine a causal association, prior observational studies of tPA administration for acute stroke have found that patients with more severe neurologic deficits were more likely to receive thrombolysis.17, 18 The 9.0% inpatient case‐fatality rate observed in our study compares favorably to the 13.4% mortality rate after tPA reported in a post‐approval meta‐analysis of safety outcomes22 and the rate of intracranial hemorrhage in our analysis was similar to those observed in other settings.9, 2225 We were unable to determine whether intracranial hemorrhages in our study were as a result of tPA administration or whether patients who received tPA were more likely to have intracranial hemorrhages detected, such as may be due to increased frequency of head imaging.

Larger hospitals were more likely to administer tPA. This may reflect regionalization of stroke care, particularly in those designated as stroke centers of excellence. As well, there is some evidence that there is a learning curve with thrombolysis administration, where guideline‐recommended practice and use of tPA increases with additional experience with the drug.9, 26 Promoting systems that allow for rapid triage and diagnosis of acute stroke should be encouraged and hospital leaders should develop strategies that allow for early recognition of potential tPA candidates.

There are several limitations to our analysis. The NHDS does not collect detailed data on clinical or presenting features of stroke, and so we lacked information on stroke severity and eligibility for administration of thrombolysis. Our study may have underestimated the overall rates of thrombolysis, as it was dependent on diagnostic codes. A previous study of 34 patients who received tPA found that although the 99.10 code was 100% specific, the code identified only 17 patients who actually received tPA (sensitivity of 50%).20 Another study comparing Medicare administrative claims data to actual pharmacy billing charges for tPA found that administrative data underestimated the rate of tPA administration by approximately 25% to 30%.12 If a diagnostic code sensitivity of 50% was assumed, rates of tPA administration in our study may have been as high as 4.8% (95% CI, 4.1‐5.5%) by year 2006.

Conclusion

In conclusion, the use of intravenous thrombolysis in patients admitted with acute ischemic stroke in the United States has risen over time, but overall use remains very low. Further efforts to improve appropriate administration rates should be encouraged, particularly as the acceptable time‐window for using tPA widens.

Recombinant tissue plasminogen activator (tPA), approved for use in the United States for the treatment for acute ischemic stroke since 1996, improves overall outcomes from ischemic stroke when administered to selected patients.14 Several prominent guidelines, including the Brain Attack Coalition and the American Stroke Association, have recommended increasing the use of tPA for acute ischemic stroke.57 In addition, in 2003 the Joint Commission on Accreditation of Healthcare Organizations developed a disease‐specific certification program to designate certain institutions Primary Stroke Centers, with one of the performance measures being the availability of thrombolysis.8

Despite guidelines and regulatory agencies promoting the use of thrombolysis for ischemic stroke, previous studies have shown disappointingly low rates of use.912 The goals of this study were to assess whether national trends in the use of thrombolysis for acute ischemic stroke have increased in light of increased regulatory activity as well as to identify patient characteristics associated with thrombolytic administration.

Materials and Methods

Data for this study were obtained from the 2001 through 2006 National Hospital Discharge Survey (NHDS), a nationally representative sample of inpatient hospitalizations conducted annually by the National Center for Health Statistics.13 The NHDS collects data on approximately 300,000 patients from about 500 short‐stay nonfederal hospitals in the United States and uses a 3‐stage sampling strategy that allows for extrapolation to national level estimates. Response rates typically exceed 90% from participating hospitals. The survey collects demographic data, including age, sex, race, hospital geographic region, hospital bedsize and patient insurance status. In addition, up to 7 diagnostic and 4 procedural codes from the hospitalization are available, as is hospitalization length of stay and patient discharge disposition. No information on timing of symptoms, degree of neurologic compromise, or results of imaging tests were available in the NHDS.

We searched for all patients age 18 years or older with a primary diagnostic code of ischemic stroke using the International Classification of Diseases, 9th Edition, Clinical Modification (ICD‐9‐CM) codes 433, 434, 436, 437.0, and 437.1, excluding codes with a fifth digit of 0 (which indicated arterial occlusion without mention of cerebral infarction). We then searched for the presence of an ICD‐9‐CM procedure code for injection or infusion of thrombolytic agent (code 99.10). Specific comorbid conditions associated with ischemic stroke were identified by searching for specific ICD‐9‐CM codes, including for heart failure, coronary artery disease, hypertension, diabetes mellitus, and atrial fibrillation. To provide a general assessment of the severity of illness of the patients, we calculated an adapted Charlson comorbidity score for each patient using available secondary discharge diagnosis codes.14 We also searched for codes corresponding to intracranial hemorrhage, a complication associated with tPA administration.

Results

From years 2001 through 2006, we identified 22,842 patients with a primary diagnosis of ischemic stroke. Of these, 313 (1.37%, 95% confidence interval [CI], 1.22‐1.53%) had a procedure code for injection or infusion of thrombolysis. Using NHDS sample weights, these numbers corresponded to an estimated 2.55 million hospitalizations for ischemic stroke in the United States during the time period and to 35,082 patients receiving intravenous thrombolytics. Although the overall rate of thrombolysis administration was quite low overall, the administration rate increased over time, from 0.87% [95% CI, 0.61‐1.22%] of stroke patients in year 2001 to 2.40% [95% CI, 1.95‐2.93%] in year 2006 and with a particular increase especially noted after year 2003 (P <0.001 for trend, Figure 1).

Figure 1

On bivariate analysis, a lower proportion of African‐American patients received tPA compared to white patients (0.8% vs. 1.5%, P = 0.003), while a higher proportion of patients with atrial fibrillation received tPA (2.3% vs. 1.2%, P < 0.001). Older patients were less likely than younger patients to receive tPA (Table 1). The rate of intracranial hemorrhage was significantly higher in patients who received tPA (5.4% vs. 0.6%, P < 0.001) and the overall inpatient mortality in patients who received tPA was 9.0%. Mortality in patients receiving tPA continued to be higher than in patients who did not receive tPA even when patients with intracranial hemorrhage were excluded (8.1% vs. 5.3%, P < 0.001). Larger hospitals were more likely to administer tPA to patients with ischemic stroke, with a 1.79% administration rate in hospitals with 300 beds compared to 0.90% in hospitals with 100 to 199 beds and 0.52% in hospitals with 6 to 99 beds (P < 0.001).

After adjusting for patient and hospital characteristics, the absolute rate of thrombolysis administration increased by an average of 0.19% per year (95% CI, 0.12‐0.26%). Factors that were significantly associated with administration of thrombolytics included being hospitalized in a larger hospital, having a history of atrial fibrillation, and a higher Charlson comorbidity index (Table 2). Patients aged 80 years or older, African American patients, and those with diabetes mellitus were significantly less likely to receive thrombolysis.

Discussion

Despite strong recommendations from guidelines and regulatory agencies, national rates of intravenous thrombolysis for ischemic stroke continue to be quite low overall. However, tPA administration appears to have increased from previous years and particularly increased in years after the Joint Commission began to accredit institutions as Primary Stroke Centers.11 The oldest patients and African Americans were less likely to receive thrombolytics, while patients with atrial fibrillation were more likely to receive thrombolysis, potentially related to atrial fibrillation causing more severe strokes.15 A total of 5.4% of patients who received tPA were diagnosed with intracranial hemorrhage, and the inpatient mortality rate of patients with tPA was 9.0%.

The exact optimal rate of thrombolysis administration for the patients in our study is unknown, as the NHDS database lacked detailed information on factors that would preclude tPA administration such as late timing of presentation and mild stroke symptoms.3 Studies conducted in stroke registries and regional settings have found that only approximately 15% to 32% of patients presenting with ischemic stroke arrive within 3 hours of symptom onset, and of these, only about 40% to 50% are eligible for tPA clinically.9, 10, 1619 However, even among presumed eligible patients, tPA administration rates only range between 25% and 43%,17, 19, 20 and the ideal rate is likely to be higher than the very low rates we observed in our study. Newer evidence that extending the time window where tPA may be given safely may increase the number of eligible patients.21

Patients who received thrombolysis had higher mortality rates than patients who did not. Although we were unable to determine a causal association, prior observational studies of tPA administration for acute stroke have found that patients with more severe neurologic deficits were more likely to receive thrombolysis.17, 18 The 9.0% inpatient case‐fatality rate observed in our study compares favorably to the 13.4% mortality rate after tPA reported in a post‐approval meta‐analysis of safety outcomes22 and the rate of intracranial hemorrhage in our analysis was similar to those observed in other settings.9, 2225 We were unable to determine whether intracranial hemorrhages in our study were as a result of tPA administration or whether patients who received tPA were more likely to have intracranial hemorrhages detected, such as may be due to increased frequency of head imaging.

Larger hospitals were more likely to administer tPA. This may reflect regionalization of stroke care, particularly in those designated as stroke centers of excellence. As well, there is some evidence that there is a learning curve with thrombolysis administration, where guideline‐recommended practice and use of tPA increases with additional experience with the drug.9, 26 Promoting systems that allow for rapid triage and diagnosis of acute stroke should be encouraged and hospital leaders should develop strategies that allow for early recognition of potential tPA candidates.

There are several limitations to our analysis. The NHDS does not collect detailed data on clinical or presenting features of stroke, and so we lacked information on stroke severity and eligibility for administration of thrombolysis. Our study may have underestimated the overall rates of thrombolysis, as it was dependent on diagnostic codes. A previous study of 34 patients who received tPA found that although the 99.10 code was 100% specific, the code identified only 17 patients who actually received tPA (sensitivity of 50%).20 Another study comparing Medicare administrative claims data to actual pharmacy billing charges for tPA found that administrative data underestimated the rate of tPA administration by approximately 25% to 30%.12 If a diagnostic code sensitivity of 50% was assumed, rates of tPA administration in our study may have been as high as 4.8% (95% CI, 4.1‐5.5%) by year 2006.

Conclusion

In conclusion, the use of intravenous thrombolysis in patients admitted with acute ischemic stroke in the United States has risen over time, but overall use remains very low. Further efforts to improve appropriate administration rates should be encouraged, particularly as the acceptable time‐window for using tPA widens.

Recombinant tissue plasminogen activator (tPA), approved for use in the United States for the treatment for acute ischemic stroke since 1996, improves overall outcomes from ischemic stroke when administered to selected patients.14 Several prominent guidelines, including the Brain Attack Coalition and the American Stroke Association, have recommended increasing the use of tPA for acute ischemic stroke.57 In addition, in 2003 the Joint Commission on Accreditation of Healthcare Organizations developed a disease‐specific certification program to designate certain institutions Primary Stroke Centers, with one of the performance measures being the availability of thrombolysis.8

Despite guidelines and regulatory agencies promoting the use of thrombolysis for ischemic stroke, previous studies have shown disappointingly low rates of use.912 The goals of this study were to assess whether national trends in the use of thrombolysis for acute ischemic stroke have increased in light of increased regulatory activity as well as to identify patient characteristics associated with thrombolytic administration.

Materials and Methods

Data for this study were obtained from the 2001 through 2006 National Hospital Discharge Survey (NHDS), a nationally representative sample of inpatient hospitalizations conducted annually by the National Center for Health Statistics.13 The NHDS collects data on approximately 300,000 patients from about 500 short‐stay nonfederal hospitals in the United States and uses a 3‐stage sampling strategy that allows for extrapolation to national level estimates. Response rates typically exceed 90% from participating hospitals. The survey collects demographic data, including age, sex, race, hospital geographic region, hospital bedsize and patient insurance status. In addition, up to 7 diagnostic and 4 procedural codes from the hospitalization are available, as is hospitalization length of stay and patient discharge disposition. No information on timing of symptoms, degree of neurologic compromise, or results of imaging tests were available in the NHDS.

We searched for all patients age 18 years or older with a primary diagnostic code of ischemic stroke using the International Classification of Diseases, 9th Edition, Clinical Modification (ICD‐9‐CM) codes 433, 434, 436, 437.0, and 437.1, excluding codes with a fifth digit of 0 (which indicated arterial occlusion without mention of cerebral infarction). We then searched for the presence of an ICD‐9‐CM procedure code for injection or infusion of thrombolytic agent (code 99.10). Specific comorbid conditions associated with ischemic stroke were identified by searching for specific ICD‐9‐CM codes, including for heart failure, coronary artery disease, hypertension, diabetes mellitus, and atrial fibrillation. To provide a general assessment of the severity of illness of the patients, we calculated an adapted Charlson comorbidity score for each patient using available secondary discharge diagnosis codes.14 We also searched for codes corresponding to intracranial hemorrhage, a complication associated with tPA administration.

Results

From years 2001 through 2006, we identified 22,842 patients with a primary diagnosis of ischemic stroke. Of these, 313 (1.37%, 95% confidence interval [CI], 1.22‐1.53%) had a procedure code for injection or infusion of thrombolysis. Using NHDS sample weights, these numbers corresponded to an estimated 2.55 million hospitalizations for ischemic stroke in the United States during the time period and to 35,082 patients receiving intravenous thrombolytics. Although the overall rate of thrombolysis administration was quite low overall, the administration rate increased over time, from 0.87% [95% CI, 0.61‐1.22%] of stroke patients in year 2001 to 2.40% [95% CI, 1.95‐2.93%] in year 2006 and with a particular increase especially noted after year 2003 (P <0.001 for trend, Figure 1).

Figure 1

On bivariate analysis, a lower proportion of African‐American patients received tPA compared to white patients (0.8% vs. 1.5%, P = 0.003), while a higher proportion of patients with atrial fibrillation received tPA (2.3% vs. 1.2%, P < 0.001). Older patients were less likely than younger patients to receive tPA (Table 1). The rate of intracranial hemorrhage was significantly higher in patients who received tPA (5.4% vs. 0.6%, P < 0.001) and the overall inpatient mortality in patients who received tPA was 9.0%. Mortality in patients receiving tPA continued to be higher than in patients who did not receive tPA even when patients with intracranial hemorrhage were excluded (8.1% vs. 5.3%, P < 0.001). Larger hospitals were more likely to administer tPA to patients with ischemic stroke, with a 1.79% administration rate in hospitals with 300 beds compared to 0.90% in hospitals with 100 to 199 beds and 0.52% in hospitals with 6 to 99 beds (P < 0.001).

After adjusting for patient and hospital characteristics, the absolute rate of thrombolysis administration increased by an average of 0.19% per year (95% CI, 0.12‐0.26%). Factors that were significantly associated with administration of thrombolytics included being hospitalized in a larger hospital, having a history of atrial fibrillation, and a higher Charlson comorbidity index (Table 2). Patients aged 80 years or older, African American patients, and those with diabetes mellitus were significantly less likely to receive thrombolysis.

Discussion

Despite strong recommendations from guidelines and regulatory agencies, national rates of intravenous thrombolysis for ischemic stroke continue to be quite low overall. However, tPA administration appears to have increased from previous years and particularly increased in years after the Joint Commission began to accredit institutions as Primary Stroke Centers.11 The oldest patients and African Americans were less likely to receive thrombolytics, while patients with atrial fibrillation were more likely to receive thrombolysis, potentially related to atrial fibrillation causing more severe strokes.15 A total of 5.4% of patients who received tPA were diagnosed with intracranial hemorrhage, and the inpatient mortality rate of patients with tPA was 9.0%.

The exact optimal rate of thrombolysis administration for the patients in our study is unknown, as the NHDS database lacked detailed information on factors that would preclude tPA administration such as late timing of presentation and mild stroke symptoms.3 Studies conducted in stroke registries and regional settings have found that only approximately 15% to 32% of patients presenting with ischemic stroke arrive within 3 hours of symptom onset, and of these, only about 40% to 50% are eligible for tPA clinically.9, 10, 1619 However, even among presumed eligible patients, tPA administration rates only range between 25% and 43%,17, 19, 20 and the ideal rate is likely to be higher than the very low rates we observed in our study. Newer evidence that extending the time window where tPA may be given safely may increase the number of eligible patients.21

Patients who received thrombolysis had higher mortality rates than patients who did not. Although we were unable to determine a causal association, prior observational studies of tPA administration for acute stroke have found that patients with more severe neurologic deficits were more likely to receive thrombolysis.17, 18 The 9.0% inpatient case‐fatality rate observed in our study compares favorably to the 13.4% mortality rate after tPA reported in a post‐approval meta‐analysis of safety outcomes22 and the rate of intracranial hemorrhage in our analysis was similar to those observed in other settings.9, 2225 We were unable to determine whether intracranial hemorrhages in our study were as a result of tPA administration or whether patients who received tPA were more likely to have intracranial hemorrhages detected, such as may be due to increased frequency of head imaging.

Larger hospitals were more likely to administer tPA. This may reflect regionalization of stroke care, particularly in those designated as stroke centers of excellence. As well, there is some evidence that there is a learning curve with thrombolysis administration, where guideline‐recommended practice and use of tPA increases with additional experience with the drug.9, 26 Promoting systems that allow for rapid triage and diagnosis of acute stroke should be encouraged and hospital leaders should develop strategies that allow for early recognition of potential tPA candidates.

There are several limitations to our analysis. The NHDS does not collect detailed data on clinical or presenting features of stroke, and so we lacked information on stroke severity and eligibility for administration of thrombolysis. Our study may have underestimated the overall rates of thrombolysis, as it was dependent on diagnostic codes. A previous study of 34 patients who received tPA found that although the 99.10 code was 100% specific, the code identified only 17 patients who actually received tPA (sensitivity of 50%).20 Another study comparing Medicare administrative claims data to actual pharmacy billing charges for tPA found that administrative data underestimated the rate of tPA administration by approximately 25% to 30%.12 If a diagnostic code sensitivity of 50% was assumed, rates of tPA administration in our study may have been as high as 4.8% (95% CI, 4.1‐5.5%) by year 2006.

Conclusion

In conclusion, the use of intravenous thrombolysis in patients admitted with acute ischemic stroke in the United States has risen over time, but overall use remains very low. Further efforts to improve appropriate administration rates should be encouraged, particularly as the acceptable time‐window for using tPA widens.

Atrial fibrillation (AF), the most common clinically significant cardiac arrhythmia, affects over 2.3 million people in the United States.1 AF is associated with an increased risk of stroke and heart failure and independently increases the risk of all cause mortality.26 As such, AF confers a staggering healthcare cost burden.7, 8 Pharmacologic treatments to restore sinus rhythm in patients with AF are associated with a considerable relapse rate911 and the development of nonpharmacologic treatments for AF, such as catheter ablation procedures,1214 may be significantly more successful in restoring and maintaining sinus rhythm.15, 16 Despite relatively poor results from early catheter ablation techniques, the practice has evolved and boasts short‐term success rates as high as 73% to 91% depending on the specific type of procedure.17

In light of the success of ablative therapy, this approach, which was once used primarily in younger patients with structurally intact hearts, has been expanded to include more medically complex patients, including elderly patients, those with cardiomyopathy, and those with implanted devices.16, 18 At the same time, catheter ablation is not without complications, with major complications observed in up to 6% of cases,19 and significant costs.20 Moreover, while the most optimistic randomized control data demonstrate the ability of catheter ablation to prevent the recurrence of AF at 1 year,12, 21, 22 long‐term outcome data are lacking, particularly in patients older than 65 years or those with heart failure.17, 23

The encouraging results supporting catheter ablation continue to stimulate the utilization of catheter ablation practices and spur innovations in ablation techniques.24 The American College of Cardiology/American Heart Association/European Society of Cardiology consensus guidelines recommend consideration of ablative therapy in many instances of AF.17 AF is primarily a disease of older adults25 and although most studies have focused on younger individuals,26 it is possible that increasing numbers of older patients are receiving ablation therapy.16 Although single center studies are available,16 there are few data about the characteristics of patients undergoing ablative therapy on a national level. In order to better understand the current use of catheter ablation treatment for AF, we analyzed data from the National Hospital Discharge Survey (NHDS) to explore trends in patient characteristics and rates of ablation procedures in hospitalized patients with AF from the years 1990 to 2005.

Methods

The NHDS is a nationally representative study of hospitalized patients conducted annually by the National Center for Health Statistics,27 which collects data from approximately 270,000 inpatient records using a representative sample of about 500 short‐stay nonfederal hospitals in the United States. Data for each patient are obtained for age, sex, hospital geographic region (Northeast, Midwest, South, West), and hospital bed size, as well as up to 7 diagnostic codes and 4 procedural codes using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD‐9‐CM). Of note, data on race/ethnicity were not consistently coded in the NHDS and are therefore not included in this analysis.

We searched for all patients age 18 years or older who had an ICD‐9‐CM diagnosis of AF (427.31). Of these patients, we then identified those who had a procedure code for nonsurgical ablation of lesions or tissues of the heart via peripherally‐inserted catheter or an endovascular approach (37.34). We also searched for specific ICD‐9‐CM‐coded diagnoses corresponding to higher stroke risk according to the (CHADS₂) risk index,28 where 1 point is assigned for congestive heart failure, hypertension, age >75 years, or diabetes mellitus, and 2 points for prior stroke or transient ischemic attack. We calculated a CHADS₂ score for each patient.

Results

From 1990 to 2005, we identified 269,471 hospitalizations in the NHDS with a diagnosis of AF, of which 1,144 (0.42%) had a procedure code for catheter ablation. When extrapolated to national estimates, this corresponds to 32 million hospitalizations of patients with AF in the United States during the time period, of which 133,003 underwent ablation. The proportion of patients with AF who had ablation increased significantly over time, from 0.06% in 1990 to 0.79% in 2005 (P < 0.001 for trend; Figure 1).

Figure 1

On univariate analysis, people with AF undergoing ablation were on average younger and more likely to be male than those who did not have ablation (Table 1). The rate of catheter ablation was higher in patients younger than 50 years (1.75%) compared to 0.55% in patients aged 50 to 79 years, and 0.16% in patients aged 80 years or older. However, ablation rates increased significantly in all age groups over time, with no one age group increasing at a significantly faster rate than the others (P value for interaction between age categories and hospitalization year = 0.7; Figure 2). People undergoing ablation tended to have lower CHADS₂ stroke risk scores and fewer risk factors for stroke, including heart failure, coronary artery disease, and diabetes mellitus (Table 1).

Figure 2

People who underwent ablation were more likely to have private insurance as their primary source of payment and less likely to have Medicare (Table 1). Ablation rates were higher among patients with AF hospitalized in the Western and Southern regions of the United States (0.52% and 0.53%, respectively), compared to rates in the Midwest (0.30%) and Northeast (0.40%). Hospital bed‐size was significantly related to the frequency of ablation, with the overall rate of ablation in patients with AF being 0.04% in hospitals with 6 to 99 beds compared to 1.37% in hospitals with at least 500 beds (P < 0.001). Length of stay was shorter in patients with ablations compared to patients without ablation therapy, and patients with ablation were more likely to be discharged home (Table 1). The inpatient mortality rate in patients undergoing ablation was quite low (0.96%).

In multivariate analysis, the likelihood of ablation therapy in a hospitalized patient with AF increased by 15% per year (95% confidence interval [CI], 13%‐16%) over the time period, adjusted for clinical and hospital characteristics. The likelihood of ablation decreased with older age (adjusted odds ratio [aOR], 0.7 [95% CI, 0.6‐0.7] for each decade of age over 50 years) and for each 1‐point increase in CHADS₂ score (aOR, 0.7 [95% CI, 0.7‐0.8]). Ablation was significantly more likely to be performed in hospitals with larger bed‐sizes (aOR, 27.4 [95% CI, 16.1‐46.6] comparing bed‐size of 500+ to bed‐size of 6 to 99) and in patients with private insurance (aOR, 1.4 [95% CI, 1.2‐1.6]; Table 2). The goodness‐of‐fit of the model was appropriate, with a nonsignificant Hosmer‐Lemeshow test P value of 0.13.

To account for the possibility that the ablation procedure was not specifically for AF, we performed a subgroup analysis that excluded all patients who also had diagnostic codes for supraventricular or ventricular tachycardias (427.0, 427.1, 427.2, and 427.4), or atrial flutter (427.32). Of the 269,471 hospitalizations with AF, 23,069 (8.6%) had a code for an arrhythmia in addition to AF. When we excluded patients with other arrhythmias, we identified 691 patients who underwent ablation and who only had a diagnosis of AF. An analysis of this subset yielded results similar to the full analysis (Table 2). The likelihood of ablation therapy in this subset of patients with only AF increased by 14% per year (95% CI, 11%‐16%), adjusting for patient age, sex, insurance status, CHADS₂ score, hospital region, and hospital bed‐size.

Discussion

The proportion of hospitalized patients with AF who undergo ablation therapy in the United States has been increasing by approximately 15% per year over the last 15 years. Patients receiving ablation therapy are more likely to be younger, have private insurance, and have fewer stroke risk factors. These demographics likely reflect the fact that these ablations are elective procedures that are preferentially performed in healthier, lower‐risk patients. Despite these preferences, the rate of ablation therapy has been increasing significantly across all age groups, even in the oldest patients.

Though limited by relatively short follow‐up data, published studies of ablation therapies for AF show promising results,17, 26 and initial cost analyses suggest possible fiscal benefits of ablation for AF.20 Despite a paucity of randomized clinical trials comparing ablation to pharmacologic rhythm and rate control, studies suggest that quality of life may be significantly improved with ablation as compared to antiarrhythmic drugs.21 This may be because ablation may reduce AF‐related symptoms.12 As ablation becomes more widespread and recommended, physicians, including hospitalists, may be increasingly likely to refer their patients for ablation, even for patient subgroups who were not well‐represented in clinical trial settings.

The inpatient mortality rate in patients undergoing ablation therapy was quite low in our study, although ablation is not without some risk of procedure‐related stroke and other complications.19 An analysis of the compiled studies on ablation for AF estimates that major complication such as cardiac tamponade or thromboembolism occur in as many as 7% of patients.26 Patients are at highest risk for embolic events, such as transient ischemic attacks or ischemic strokes, in the immediate hours to weeks after ablation. An estimated 5% to 25% of patients will develop a new arrhythmia at some point in the postablation period and other complications, including esophageal injury, phrenic nerve injury, groin hematoma, and retroperitoneal bleed, have been observed.26 Increasing comanagement of postablation patients will necessitate that hospitalists understand the potential complications of ablation as well as current strategies for bridging anticoagulation therapy.

Few data are available about the safety and efficacy of catheter ablation for patients over the age of 65 years. In fact, the mean age of patients enrolled in most clinical trials of catheter ablation was younger than 60 years.26, 29 There are also limited data about the long‐term efficacy of ablation therapy in patients with structural heart disease30; despite this, our study shows that a quarter of patients with AF undergoing ablation therapy in the United States have diagnosed heart failure. As always, the optimistic introduction of new technologies to unstudied patient populations carries the risk of unintended harm. Hospitalists are well situated to collect and analyze outcome data for older patients with multiple comorbidities and to provide real‐time monitoring of potential complications.

Few studies have focused on the demographic and comorbid characteristics of patients undergoing ablation for AF on a national level. One study examined characteristics of patients referred to a single academic center for AF ablation from 1999 to 2005 and found that referred patients have, over time, been older (mean age 47 years in 1999 versus 56 years in 2005), have more persistent AF, larger atria, and were more likely to have had a history of cardiomyopathy (0% in 1999 versus 16% in 2006).16 This study also reported that men were consistently more likely to be referred for ablation than women. These results are generally consistent with our findings.

Our study has several limitations. The exact indication and specific type of ablation were not available in the NHDS, and it is possible that the ablation procedure was for an arrhythmia other than AF. However, our analysis of the subset of patients who only had AF as a diagnosis yielded results similar to the full analysis. We were unable to assess specific efficacy or complication data, but mortality was low and patients tended to have short hospital stays. Because the NHDS samples random hospitalizations, it is possible that some patients were overrepresented in the database if they were repeatedly hospitalized in a single year. This could potentially bias our results toward an overestimate of the number of patients who receive ablation.

It remains unclear what proportion of AF ablation procedures occur in the outpatient versus inpatient setting. Inpatient versus outpatient status is not specified in the few single‐center ablation experiences reported in the literature,16 and the few trials reported are not reliable for determining practice in a nonstudy setting. The most recent (2006) Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert Consensus Statement on Catheter and Surgical Ablation of AF recommends aggressive anticoagulation in the periprocedure period with either heparin or low‐molecular‐weight heparins, followed by a bridge to warfarin.17 It makes intuitive sense that patients undergoing ablation for AF would be admitted at least overnight to bridge anticoagulation therapy and monitor for complications, but widespread use of low‐molecular‐weight heparin may make hospitalization less necessary. The observation that patients undergoing ablation had shorter hospital stays does not necessarily imply that ablation procedures shorten hospital stays. Rather, the data almost certainly reflect the fact that ablations are mostly elective procedures performed in the setting of planned short‐term admissions.

Our study provides important epidemiologic data about national trends in the use of ablation therapy in hospitalized patients with AF. We find that the rate of catheter ablation in patients with AF has been increasing significantly over time and across all age groups, including the oldest patients. As the proportion of patients with AF who receive ablation therapy continues to increase over time, comprehensive long‐term outcome data and cost‐effectiveness analyses will be important.

Atrial fibrillation (AF), the most common clinically significant cardiac arrhythmia, affects over 2.3 million people in the United States.1 AF is associated with an increased risk of stroke and heart failure and independently increases the risk of all cause mortality.26 As such, AF confers a staggering healthcare cost burden.7, 8 Pharmacologic treatments to restore sinus rhythm in patients with AF are associated with a considerable relapse rate911 and the development of nonpharmacologic treatments for AF, such as catheter ablation procedures,1214 may be significantly more successful in restoring and maintaining sinus rhythm.15, 16 Despite relatively poor results from early catheter ablation techniques, the practice has evolved and boasts short‐term success rates as high as 73% to 91% depending on the specific type of procedure.17

In light of the success of ablative therapy, this approach, which was once used primarily in younger patients with structurally intact hearts, has been expanded to include more medically complex patients, including elderly patients, those with cardiomyopathy, and those with implanted devices.16, 18 At the same time, catheter ablation is not without complications, with major complications observed in up to 6% of cases,19 and significant costs.20 Moreover, while the most optimistic randomized control data demonstrate the ability of catheter ablation to prevent the recurrence of AF at 1 year,12, 21, 22 long‐term outcome data are lacking, particularly in patients older than 65 years or those with heart failure.17, 23

The encouraging results supporting catheter ablation continue to stimulate the utilization of catheter ablation practices and spur innovations in ablation techniques.24 The American College of Cardiology/American Heart Association/European Society of Cardiology consensus guidelines recommend consideration of ablative therapy in many instances of AF.17 AF is primarily a disease of older adults25 and although most studies have focused on younger individuals,26 it is possible that increasing numbers of older patients are receiving ablation therapy.16 Although single center studies are available,16 there are few data about the characteristics of patients undergoing ablative therapy on a national level. In order to better understand the current use of catheter ablation treatment for AF, we analyzed data from the National Hospital Discharge Survey (NHDS) to explore trends in patient characteristics and rates of ablation procedures in hospitalized patients with AF from the years 1990 to 2005.

Methods

The NHDS is a nationally representative study of hospitalized patients conducted annually by the National Center for Health Statistics,27 which collects data from approximately 270,000 inpatient records using a representative sample of about 500 short‐stay nonfederal hospitals in the United States. Data for each patient are obtained for age, sex, hospital geographic region (Northeast, Midwest, South, West), and hospital bed size, as well as up to 7 diagnostic codes and 4 procedural codes using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD‐9‐CM). Of note, data on race/ethnicity were not consistently coded in the NHDS and are therefore not included in this analysis.

We searched for all patients age 18 years or older who had an ICD‐9‐CM diagnosis of AF (427.31). Of these patients, we then identified those who had a procedure code for nonsurgical ablation of lesions or tissues of the heart via peripherally‐inserted catheter or an endovascular approach (37.34). We also searched for specific ICD‐9‐CM‐coded diagnoses corresponding to higher stroke risk according to the (CHADS₂) risk index,28 where 1 point is assigned for congestive heart failure, hypertension, age >75 years, or diabetes mellitus, and 2 points for prior stroke or transient ischemic attack. We calculated a CHADS₂ score for each patient.

Results

From 1990 to 2005, we identified 269,471 hospitalizations in the NHDS with a diagnosis of AF, of which 1,144 (0.42%) had a procedure code for catheter ablation. When extrapolated to national estimates, this corresponds to 32 million hospitalizations of patients with AF in the United States during the time period, of which 133,003 underwent ablation. The proportion of patients with AF who had ablation increased significantly over time, from 0.06% in 1990 to 0.79% in 2005 (P < 0.001 for trend; Figure 1).

Figure 1

On univariate analysis, people with AF undergoing ablation were on average younger and more likely to be male than those who did not have ablation (Table 1). The rate of catheter ablation was higher in patients younger than 50 years (1.75%) compared to 0.55% in patients aged 50 to 79 years, and 0.16% in patients aged 80 years or older. However, ablation rates increased significantly in all age groups over time, with no one age group increasing at a significantly faster rate than the others (P value for interaction between age categories and hospitalization year = 0.7; Figure 2). People undergoing ablation tended to have lower CHADS₂ stroke risk scores and fewer risk factors for stroke, including heart failure, coronary artery disease, and diabetes mellitus (Table 1).

Figure 2

People who underwent ablation were more likely to have private insurance as their primary source of payment and less likely to have Medicare (Table 1). Ablation rates were higher among patients with AF hospitalized in the Western and Southern regions of the United States (0.52% and 0.53%, respectively), compared to rates in the Midwest (0.30%) and Northeast (0.40%). Hospital bed‐size was significantly related to the frequency of ablation, with the overall rate of ablation in patients with AF being 0.04% in hospitals with 6 to 99 beds compared to 1.37% in hospitals with at least 500 beds (P < 0.001). Length of stay was shorter in patients with ablations compared to patients without ablation therapy, and patients with ablation were more likely to be discharged home (Table 1). The inpatient mortality rate in patients undergoing ablation was quite low (0.96%).

In multivariate analysis, the likelihood of ablation therapy in a hospitalized patient with AF increased by 15% per year (95% confidence interval [CI], 13%‐16%) over the time period, adjusted for clinical and hospital characteristics. The likelihood of ablation decreased with older age (adjusted odds ratio [aOR], 0.7 [95% CI, 0.6‐0.7] for each decade of age over 50 years) and for each 1‐point increase in CHADS₂ score (aOR, 0.7 [95% CI, 0.7‐0.8]). Ablation was significantly more likely to be performed in hospitals with larger bed‐sizes (aOR, 27.4 [95% CI, 16.1‐46.6] comparing bed‐size of 500+ to bed‐size of 6 to 99) and in patients with private insurance (aOR, 1.4 [95% CI, 1.2‐1.6]; Table 2). The goodness‐of‐fit of the model was appropriate, with a nonsignificant Hosmer‐Lemeshow test P value of 0.13.

To account for the possibility that the ablation procedure was not specifically for AF, we performed a subgroup analysis that excluded all patients who also had diagnostic codes for supraventricular or ventricular tachycardias (427.0, 427.1, 427.2, and 427.4), or atrial flutter (427.32). Of the 269,471 hospitalizations with AF, 23,069 (8.6%) had a code for an arrhythmia in addition to AF. When we excluded patients with other arrhythmias, we identified 691 patients who underwent ablation and who only had a diagnosis of AF. An analysis of this subset yielded results similar to the full analysis (Table 2). The likelihood of ablation therapy in this subset of patients with only AF increased by 14% per year (95% CI, 11%‐16%), adjusting for patient age, sex, insurance status, CHADS₂ score, hospital region, and hospital bed‐size.

Discussion

The proportion of hospitalized patients with AF who undergo ablation therapy in the United States has been increasing by approximately 15% per year over the last 15 years. Patients receiving ablation therapy are more likely to be younger, have private insurance, and have fewer stroke risk factors. These demographics likely reflect the fact that these ablations are elective procedures that are preferentially performed in healthier, lower‐risk patients. Despite these preferences, the rate of ablation therapy has been increasing significantly across all age groups, even in the oldest patients.

Though limited by relatively short follow‐up data, published studies of ablation therapies for AF show promising results,17, 26 and initial cost analyses suggest possible fiscal benefits of ablation for AF.20 Despite a paucity of randomized clinical trials comparing ablation to pharmacologic rhythm and rate control, studies suggest that quality of life may be significantly improved with ablation as compared to antiarrhythmic drugs.21 This may be because ablation may reduce AF‐related symptoms.12 As ablation becomes more widespread and recommended, physicians, including hospitalists, may be increasingly likely to refer their patients for ablation, even for patient subgroups who were not well‐represented in clinical trial settings.

The inpatient mortality rate in patients undergoing ablation therapy was quite low in our study, although ablation is not without some risk of procedure‐related stroke and other complications.19 An analysis of the compiled studies on ablation for AF estimates that major complication such as cardiac tamponade or thromboembolism occur in as many as 7% of patients.26 Patients are at highest risk for embolic events, such as transient ischemic attacks or ischemic strokes, in the immediate hours to weeks after ablation. An estimated 5% to 25% of patients will develop a new arrhythmia at some point in the postablation period and other complications, including esophageal injury, phrenic nerve injury, groin hematoma, and retroperitoneal bleed, have been observed.26 Increasing comanagement of postablation patients will necessitate that hospitalists understand the potential complications of ablation as well as current strategies for bridging anticoagulation therapy.

Few data are available about the safety and efficacy of catheter ablation for patients over the age of 65 years. In fact, the mean age of patients enrolled in most clinical trials of catheter ablation was younger than 60 years.26, 29 There are also limited data about the long‐term efficacy of ablation therapy in patients with structural heart disease30; despite this, our study shows that a quarter of patients with AF undergoing ablation therapy in the United States have diagnosed heart failure. As always, the optimistic introduction of new technologies to unstudied patient populations carries the risk of unintended harm. Hospitalists are well situated to collect and analyze outcome data for older patients with multiple comorbidities and to provide real‐time monitoring of potential complications.

Few studies have focused on the demographic and comorbid characteristics of patients undergoing ablation for AF on a national level. One study examined characteristics of patients referred to a single academic center for AF ablation from 1999 to 2005 and found that referred patients have, over time, been older (mean age 47 years in 1999 versus 56 years in 2005), have more persistent AF, larger atria, and were more likely to have had a history of cardiomyopathy (0% in 1999 versus 16% in 2006).16 This study also reported that men were consistently more likely to be referred for ablation than women. These results are generally consistent with our findings.

Our study has several limitations. The exact indication and specific type of ablation were not available in the NHDS, and it is possible that the ablation procedure was for an arrhythmia other than AF. However, our analysis of the subset of patients who only had AF as a diagnosis yielded results similar to the full analysis. We were unable to assess specific efficacy or complication data, but mortality was low and patients tended to have short hospital stays. Because the NHDS samples random hospitalizations, it is possible that some patients were overrepresented in the database if they were repeatedly hospitalized in a single year. This could potentially bias our results toward an overestimate of the number of patients who receive ablation.

It remains unclear what proportion of AF ablation procedures occur in the outpatient versus inpatient setting. Inpatient versus outpatient status is not specified in the few single‐center ablation experiences reported in the literature,16 and the few trials reported are not reliable for determining practice in a nonstudy setting. The most recent (2006) Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert Consensus Statement on Catheter and Surgical Ablation of AF recommends aggressive anticoagulation in the periprocedure period with either heparin or low‐molecular‐weight heparins, followed by a bridge to warfarin.17 It makes intuitive sense that patients undergoing ablation for AF would be admitted at least overnight to bridge anticoagulation therapy and monitor for complications, but widespread use of low‐molecular‐weight heparin may make hospitalization less necessary. The observation that patients undergoing ablation had shorter hospital stays does not necessarily imply that ablation procedures shorten hospital stays. Rather, the data almost certainly reflect the fact that ablations are mostly elective procedures performed in the setting of planned short‐term admissions.

Our study provides important epidemiologic data about national trends in the use of ablation therapy in hospitalized patients with AF. We find that the rate of catheter ablation in patients with AF has been increasing significantly over time and across all age groups, including the oldest patients. As the proportion of patients with AF who receive ablation therapy continues to increase over time, comprehensive long‐term outcome data and cost‐effectiveness analyses will be important.

Atrial fibrillation (AF), the most common clinically significant cardiac arrhythmia, affects over 2.3 million people in the United States.1 AF is associated with an increased risk of stroke and heart failure and independently increases the risk of all cause mortality.26 As such, AF confers a staggering healthcare cost burden.7, 8 Pharmacologic treatments to restore sinus rhythm in patients with AF are associated with a considerable relapse rate911 and the development of nonpharmacologic treatments for AF, such as catheter ablation procedures,1214 may be significantly more successful in restoring and maintaining sinus rhythm.15, 16 Despite relatively poor results from early catheter ablation techniques, the practice has evolved and boasts short‐term success rates as high as 73% to 91% depending on the specific type of procedure.17

In light of the success of ablative therapy, this approach, which was once used primarily in younger patients with structurally intact hearts, has been expanded to include more medically complex patients, including elderly patients, those with cardiomyopathy, and those with implanted devices.16, 18 At the same time, catheter ablation is not without complications, with major complications observed in up to 6% of cases,19 and significant costs.20 Moreover, while the most optimistic randomized control data demonstrate the ability of catheter ablation to prevent the recurrence of AF at 1 year,12, 21, 22 long‐term outcome data are lacking, particularly in patients older than 65 years or those with heart failure.17, 23

The encouraging results supporting catheter ablation continue to stimulate the utilization of catheter ablation practices and spur innovations in ablation techniques.24 The American College of Cardiology/American Heart Association/European Society of Cardiology consensus guidelines recommend consideration of ablative therapy in many instances of AF.17 AF is primarily a disease of older adults25 and although most studies have focused on younger individuals,26 it is possible that increasing numbers of older patients are receiving ablation therapy.16 Although single center studies are available,16 there are few data about the characteristics of patients undergoing ablative therapy on a national level. In order to better understand the current use of catheter ablation treatment for AF, we analyzed data from the National Hospital Discharge Survey (NHDS) to explore trends in patient characteristics and rates of ablation procedures in hospitalized patients with AF from the years 1990 to 2005.

Methods

The NHDS is a nationally representative study of hospitalized patients conducted annually by the National Center for Health Statistics,27 which collects data from approximately 270,000 inpatient records using a representative sample of about 500 short‐stay nonfederal hospitals in the United States. Data for each patient are obtained for age, sex, hospital geographic region (Northeast, Midwest, South, West), and hospital bed size, as well as up to 7 diagnostic codes and 4 procedural codes using the International Classification of Diseases, 9th Revision, Clinical Modification (ICD‐9‐CM). Of note, data on race/ethnicity were not consistently coded in the NHDS and are therefore not included in this analysis.

We searched for all patients age 18 years or older who had an ICD‐9‐CM diagnosis of AF (427.31). Of these patients, we then identified those who had a procedure code for nonsurgical ablation of lesions or tissues of the heart via peripherally‐inserted catheter or an endovascular approach (37.34). We also searched for specific ICD‐9‐CM‐coded diagnoses corresponding to higher stroke risk according to the (CHADS₂) risk index,28 where 1 point is assigned for congestive heart failure, hypertension, age >75 years, or diabetes mellitus, and 2 points for prior stroke or transient ischemic attack. We calculated a CHADS₂ score for each patient.

Results

From 1990 to 2005, we identified 269,471 hospitalizations in the NHDS with a diagnosis of AF, of which 1,144 (0.42%) had a procedure code for catheter ablation. When extrapolated to national estimates, this corresponds to 32 million hospitalizations of patients with AF in the United States during the time period, of which 133,003 underwent ablation. The proportion of patients with AF who had ablation increased significantly over time, from 0.06% in 1990 to 0.79% in 2005 (P < 0.001 for trend; Figure 1).

Figure 1

On univariate analysis, people with AF undergoing ablation were on average younger and more likely to be male than those who did not have ablation (Table 1). The rate of catheter ablation was higher in patients younger than 50 years (1.75%) compared to 0.55% in patients aged 50 to 79 years, and 0.16% in patients aged 80 years or older. However, ablation rates increased significantly in all age groups over time, with no one age group increasing at a significantly faster rate than the others (P value for interaction between age categories and hospitalization year = 0.7; Figure 2). People undergoing ablation tended to have lower CHADS₂ stroke risk scores and fewer risk factors for stroke, including heart failure, coronary artery disease, and diabetes mellitus (Table 1).

Figure 2

People who underwent ablation were more likely to have private insurance as their primary source of payment and less likely to have Medicare (Table 1). Ablation rates were higher among patients with AF hospitalized in the Western and Southern regions of the United States (0.52% and 0.53%, respectively), compared to rates in the Midwest (0.30%) and Northeast (0.40%). Hospital bed‐size was significantly related to the frequency of ablation, with the overall rate of ablation in patients with AF being 0.04% in hospitals with 6 to 99 beds compared to 1.37% in hospitals with at least 500 beds (P < 0.001). Length of stay was shorter in patients with ablations compared to patients without ablation therapy, and patients with ablation were more likely to be discharged home (Table 1). The inpatient mortality rate in patients undergoing ablation was quite low (0.96%).

In multivariate analysis, the likelihood of ablation therapy in a hospitalized patient with AF increased by 15% per year (95% confidence interval [CI], 13%‐16%) over the time period, adjusted for clinical and hospital characteristics. The likelihood of ablation decreased with older age (adjusted odds ratio [aOR], 0.7 [95% CI, 0.6‐0.7] for each decade of age over 50 years) and for each 1‐point increase in CHADS₂ score (aOR, 0.7 [95% CI, 0.7‐0.8]). Ablation was significantly more likely to be performed in hospitals with larger bed‐sizes (aOR, 27.4 [95% CI, 16.1‐46.6] comparing bed‐size of 500+ to bed‐size of 6 to 99) and in patients with private insurance (aOR, 1.4 [95% CI, 1.2‐1.6]; Table 2). The goodness‐of‐fit of the model was appropriate, with a nonsignificant Hosmer‐Lemeshow test P value of 0.13.

To account for the possibility that the ablation procedure was not specifically for AF, we performed a subgroup analysis that excluded all patients who also had diagnostic codes for supraventricular or ventricular tachycardias (427.0, 427.1, 427.2, and 427.4), or atrial flutter (427.32). Of the 269,471 hospitalizations with AF, 23,069 (8.6%) had a code for an arrhythmia in addition to AF. When we excluded patients with other arrhythmias, we identified 691 patients who underwent ablation and who only had a diagnosis of AF. An analysis of this subset yielded results similar to the full analysis (Table 2). The likelihood of ablation therapy in this subset of patients with only AF increased by 14% per year (95% CI, 11%‐16%), adjusting for patient age, sex, insurance status, CHADS₂ score, hospital region, and hospital bed‐size.

Discussion

The proportion of hospitalized patients with AF who undergo ablation therapy in the United States has been increasing by approximately 15% per year over the last 15 years. Patients receiving ablation therapy are more likely to be younger, have private insurance, and have fewer stroke risk factors. These demographics likely reflect the fact that these ablations are elective procedures that are preferentially performed in healthier, lower‐risk patients. Despite these preferences, the rate of ablation therapy has been increasing significantly across all age groups, even in the oldest patients.

Though limited by relatively short follow‐up data, published studies of ablation therapies for AF show promising results,17, 26 and initial cost analyses suggest possible fiscal benefits of ablation for AF.20 Despite a paucity of randomized clinical trials comparing ablation to pharmacologic rhythm and rate control, studies suggest that quality of life may be significantly improved with ablation as compared to antiarrhythmic drugs.21 This may be because ablation may reduce AF‐related symptoms.12 As ablation becomes more widespread and recommended, physicians, including hospitalists, may be increasingly likely to refer their patients for ablation, even for patient subgroups who were not well‐represented in clinical trial settings.

The inpatient mortality rate in patients undergoing ablation therapy was quite low in our study, although ablation is not without some risk of procedure‐related stroke and other complications.19 An analysis of the compiled studies on ablation for AF estimates that major complication such as cardiac tamponade or thromboembolism occur in as many as 7% of patients.26 Patients are at highest risk for embolic events, such as transient ischemic attacks or ischemic strokes, in the immediate hours to weeks after ablation. An estimated 5% to 25% of patients will develop a new arrhythmia at some point in the postablation period and other complications, including esophageal injury, phrenic nerve injury, groin hematoma, and retroperitoneal bleed, have been observed.26 Increasing comanagement of postablation patients will necessitate that hospitalists understand the potential complications of ablation as well as current strategies for bridging anticoagulation therapy.

Few data are available about the safety and efficacy of catheter ablation for patients over the age of 65 years. In fact, the mean age of patients enrolled in most clinical trials of catheter ablation was younger than 60 years.26, 29 There are also limited data about the long‐term efficacy of ablation therapy in patients with structural heart disease30; despite this, our study shows that a quarter of patients with AF undergoing ablation therapy in the United States have diagnosed heart failure. As always, the optimistic introduction of new technologies to unstudied patient populations carries the risk of unintended harm. Hospitalists are well situated to collect and analyze outcome data for older patients with multiple comorbidities and to provide real‐time monitoring of potential complications.

Few studies have focused on the demographic and comorbid characteristics of patients undergoing ablation for AF on a national level. One study examined characteristics of patients referred to a single academic center for AF ablation from 1999 to 2005 and found that referred patients have, over time, been older (mean age 47 years in 1999 versus 56 years in 2005), have more persistent AF, larger atria, and were more likely to have had a history of cardiomyopathy (0% in 1999 versus 16% in 2006).16 This study also reported that men were consistently more likely to be referred for ablation than women. These results are generally consistent with our findings.

Our study has several limitations. The exact indication and specific type of ablation were not available in the NHDS, and it is possible that the ablation procedure was for an arrhythmia other than AF. However, our analysis of the subset of patients who only had AF as a diagnosis yielded results similar to the full analysis. We were unable to assess specific efficacy or complication data, but mortality was low and patients tended to have short hospital stays. Because the NHDS samples random hospitalizations, it is possible that some patients were overrepresented in the database if they were repeatedly hospitalized in a single year. This could potentially bias our results toward an overestimate of the number of patients who receive ablation.

It remains unclear what proportion of AF ablation procedures occur in the outpatient versus inpatient setting. Inpatient versus outpatient status is not specified in the few single‐center ablation experiences reported in the literature,16 and the few trials reported are not reliable for determining practice in a nonstudy setting. The most recent (2006) Heart Rhythm Society/European Heart Rhythm Association/European Cardiac Arrhythmia Society Expert Consensus Statement on Catheter and Surgical Ablation of AF recommends aggressive anticoagulation in the periprocedure period with either heparin or low‐molecular‐weight heparins, followed by a bridge to warfarin.17 It makes intuitive sense that patients undergoing ablation for AF would be admitted at least overnight to bridge anticoagulation therapy and monitor for complications, but widespread use of low‐molecular‐weight heparin may make hospitalization less necessary. The observation that patients undergoing ablation had shorter hospital stays does not necessarily imply that ablation procedures shorten hospital stays. Rather, the data almost certainly reflect the fact that ablations are mostly elective procedures performed in the setting of planned short‐term admissions.

Our study provides important epidemiologic data about national trends in the use of ablation therapy in hospitalized patients with AF. We find that the rate of catheter ablation in patients with AF has been increasing significantly over time and across all age groups, including the oldest patients. As the proportion of patients with AF who receive ablation therapy continues to increase over time, comprehensive long‐term outcome data and cost‐effectiveness analyses will be important.

There is a growing demand for hospitalists in the United States. In academic settings, hospitalists are called on to perform a variety of duties, from leading quality improvement initiatives to serving on hospital committees to helping to offset restrictions on work hours of the house staff.1 Although hospitalists may be well positioned to take on these roles, obtaining adequate protected time and recognition for such contributions remains a challenge. The existing promotion and tenure processes at academic institutions may not give adequate consideration to such responsibilities. Hospitalists who do not meet the traditional benchmarks of teaching and research may suffer in their career advancement and, ultimately, in their desire to remain in academics. Developing a sustainable and long‐term career in hospital medicine is important not only from a professional developmental standpoint, but also because it may lead to better patient care; evidence from a large multicenter hospitalist study suggests that physician experience is linked to improved patient care and outcomes.2 Thus, it behooves academic medical centers that employ hospitalists to create rewarding hospitalist career paths.

Goldman described academic hospital medicine as comprising periods of systole, during which hospitalists provide clinical care, and periods of diastole, the portion of a hospitalist's time spent in nonclinical activities.3 Far from being a period of relaxation, diastole is an active component of a hospitalist's work, the time devoted to the pursuit of complementary interests, career advancement, and job diversity. A well‐thought‐out plan for the diastolic phase of a hospitalist job description can lead to significant improvement in quality, education, research, and outcomes for an academic medical center.4 A good balance of systole and diastole allows for focus on career development and advancement and has the potential to be very helpful in preventing burnout. This is of particular concern to academic hospitalists, who report working longer hours, feeling more stress, and worrying more about burnout than their nonhospitalist colleagues.5 This suggests the diastolic phase is an important part of creating a sustainable hospitalist job and should be funded as part of an academic hospitalist position.

Although the optimal balance of systole and diastole to prevent burnout is not known, outlining clear expectations is an important strategy for preparing physicians for a sustainable academic hospitalist career. This is an important issue, given the increasing number of residency graduates who are choosing careers in hospital medicine.6 Based on the reported career plans of residents taking internal medicine in‐training exams from 2002 through 2006, the number of residents going into hospital medicine has more than doubled, from 3% (in 2002) to 6.5% (in 2006). The goal of this article is to compare and contrast several career paths that balance systole and diastole in academic hospital medicine. Specifically, we review training opportunities for becoming a successful hospitalist‐educator, hospitalistquality expert, hospitalist‐investigator, and hospitalist‐administrator.

EDUCATION (THE HOSPITALIST‐EDUCATOR)

Hospitalists in academic centers often play central roles as teachers and leaders in medical education. This is not surprising given that most teaching of medical trainees occurs in the inpatient setting.7 Furthermore, several studies have consistently demonstrated that trainee satisfaction with teaching by hospitalists is high, and hospitalists are rated as more effective teachers than traditional subspecialist ward attendings.810

A typical hospitalist‐educator position is 80%‐90% clinical time, with 10%‐20% set aside for teaching. However, academic hospitalists are often expected to teach medical trainees concurrently with their clinical care activities, rather than during a separate, protected time.11 Thus, most hospitalist‐educator responsibilities do not occur during diastole, as may be conceived, but instead are add to the systole. Small amounts of protected diastolic time for a hospitalist‐educator can be used for related administrative activities, such as writing letters of recommendation, mentoring students and residents, doing creative thinking and curriculum development, and conducting educational research, such as evaluating a new educational program or curriculum. Some hospitalist‐educator positions, such as director of the residency program or internal medicine clerkship, are exceptions in that they generally include a greater amount of protected time, which may be earmarked for administrative activities and hands‐on teaching.

CLINICAL QUALITY AND OPERATIONS IMPROVEMENT (THE HOSPITALISTQUALITY EXPERT)

Hospitalists are increasingly being called on to lead clinical quality and operations improvement at academic teaching hospitals. Benefits to the institution include the consistent presence of a committed physician who is able to plan and execute change in the context of clinical care. This is in contrast to the transient nature of residents and nonhospitalist attending physicians, whose ability to participate in such initiatives is impaired by the scheduling of their rotations. Hospitalists, however, are often able to cultivate long‐standing relationships with nurses, case managers, and hospital administrators, thereby building the institutional clout to lead such initiatives while considering views from all the necessary stakeholders.22 Thus, they are in a good position to serve as physician champions and expedite the adoption of new innovations within hospitalist groups and among other physician groups and clinical staff.23, 24

RESEARCH (THE HOSPITALIST‐INVESTIGATOR)

Few hospitalists devote most of their time to clinical research. Having a strong research base is essential for the field of hospital medicine to gain credibility as a distinct specialty.4 Although the initial research in hospital medicine sought to prove the value of the field itself, hospitalists have now begun to focus on quality improvement and outcomes research.3032 Because of their unique position in clinical care, hospitalists are well situated to oversee inpatient data collection and perform research on a variety of conditions ranging from acute coronary syndromes to venous thromboembolism. Another potential area of research for hospitalists is participation in clinical trials focused on the inpatient setting. Although the proportion of time spent in research can vary widely, to become an independently successful clinical researcher typically requires a substantial amount of time be devoted to research. In general, at least 50% protected time, greater if possible, is recommended.

ADMINISTRATION (THE HOSPITALIST‐ADMINISTRATOR)

Physician leaders in hospital administration are not new. Many hospitals already include physicians in senior management positions, such as chief medical officer.44 Naturally, a career in hospital administration is another potential path for diastole in academic medical centers.

CONCLUSIONS

As hospital medicine continues to grow and evolve, designing sustainable and rewarding academic careers will be crucial to the success of the field. Being able to balance clinical systole time with obtaining the skills to support nonclinical diastole time is important to ensuring a successful career as an academic hospitalist. We have described several possible career paths in teaching, research, quality improvement, and administration. By preparing future hospitalists with the knowledge and skills required to assume a variety of roles during their diastolic time, we hope to encourage the growth of hospitalist leaders with well‐developed skill sets. If hospitalists adequately prepare themselves, academic hospital medicine will likely remain sustainable and rewarding, and future generations of trainees will be inspired and prepared to enter the field.

There is a growing demand for hospitalists in the United States. In academic settings, hospitalists are called on to perform a variety of duties, from leading quality improvement initiatives to serving on hospital committees to helping to offset restrictions on work hours of the house staff.1 Although hospitalists may be well positioned to take on these roles, obtaining adequate protected time and recognition for such contributions remains a challenge. The existing promotion and tenure processes at academic institutions may not give adequate consideration to such responsibilities. Hospitalists who do not meet the traditional benchmarks of teaching and research may suffer in their career advancement and, ultimately, in their desire to remain in academics. Developing a sustainable and long‐term career in hospital medicine is important not only from a professional developmental standpoint, but also because it may lead to better patient care; evidence from a large multicenter hospitalist study suggests that physician experience is linked to improved patient care and outcomes.2 Thus, it behooves academic medical centers that employ hospitalists to create rewarding hospitalist career paths.

Goldman described academic hospital medicine as comprising periods of systole, during which hospitalists provide clinical care, and periods of diastole, the portion of a hospitalist's time spent in nonclinical activities.3 Far from being a period of relaxation, diastole is an active component of a hospitalist's work, the time devoted to the pursuit of complementary interests, career advancement, and job diversity. A well‐thought‐out plan for the diastolic phase of a hospitalist job description can lead to significant improvement in quality, education, research, and outcomes for an academic medical center.4 A good balance of systole and diastole allows for focus on career development and advancement and has the potential to be very helpful in preventing burnout. This is of particular concern to academic hospitalists, who report working longer hours, feeling more stress, and worrying more about burnout than their nonhospitalist colleagues.5 This suggests the diastolic phase is an important part of creating a sustainable hospitalist job and should be funded as part of an academic hospitalist position.

Although the optimal balance of systole and diastole to prevent burnout is not known, outlining clear expectations is an important strategy for preparing physicians for a sustainable academic hospitalist career. This is an important issue, given the increasing number of residency graduates who are choosing careers in hospital medicine.6 Based on the reported career plans of residents taking internal medicine in‐training exams from 2002 through 2006, the number of residents going into hospital medicine has more than doubled, from 3% (in 2002) to 6.5% (in 2006). The goal of this article is to compare and contrast several career paths that balance systole and diastole in academic hospital medicine. Specifically, we review training opportunities for becoming a successful hospitalist‐educator, hospitalistquality expert, hospitalist‐investigator, and hospitalist‐administrator.

EDUCATION (THE HOSPITALIST‐EDUCATOR)

Hospitalists in academic centers often play central roles as teachers and leaders in medical education. This is not surprising given that most teaching of medical trainees occurs in the inpatient setting.7 Furthermore, several studies have consistently demonstrated that trainee satisfaction with teaching by hospitalists is high, and hospitalists are rated as more effective teachers than traditional subspecialist ward attendings.810

A typical hospitalist‐educator position is 80%‐90% clinical time, with 10%‐20% set aside for teaching. However, academic hospitalists are often expected to teach medical trainees concurrently with their clinical care activities, rather than during a separate, protected time.11 Thus, most hospitalist‐educator responsibilities do not occur during diastole, as may be conceived, but instead are add to the systole. Small amounts of protected diastolic time for a hospitalist‐educator can be used for related administrative activities, such as writing letters of recommendation, mentoring students and residents, doing creative thinking and curriculum development, and conducting educational research, such as evaluating a new educational program or curriculum. Some hospitalist‐educator positions, such as director of the residency program or internal medicine clerkship, are exceptions in that they generally include a greater amount of protected time, which may be earmarked for administrative activities and hands‐on teaching.

CLINICAL QUALITY AND OPERATIONS IMPROVEMENT (THE HOSPITALISTQUALITY EXPERT)

Hospitalists are increasingly being called on to lead clinical quality and operations improvement at academic teaching hospitals. Benefits to the institution include the consistent presence of a committed physician who is able to plan and execute change in the context of clinical care. This is in contrast to the transient nature of residents and nonhospitalist attending physicians, whose ability to participate in such initiatives is impaired by the scheduling of their rotations. Hospitalists, however, are often able to cultivate long‐standing relationships with nurses, case managers, and hospital administrators, thereby building the institutional clout to lead such initiatives while considering views from all the necessary stakeholders.22 Thus, they are in a good position to serve as physician champions and expedite the adoption of new innovations within hospitalist groups and among other physician groups and clinical staff.23, 24

RESEARCH (THE HOSPITALIST‐INVESTIGATOR)

Few hospitalists devote most of their time to clinical research. Having a strong research base is essential for the field of hospital medicine to gain credibility as a distinct specialty.4 Although the initial research in hospital medicine sought to prove the value of the field itself, hospitalists have now begun to focus on quality improvement and outcomes research.3032 Because of their unique position in clinical care, hospitalists are well situated to oversee inpatient data collection and perform research on a variety of conditions ranging from acute coronary syndromes to venous thromboembolism. Another potential area of research for hospitalists is participation in clinical trials focused on the inpatient setting. Although the proportion of time spent in research can vary widely, to become an independently successful clinical researcher typically requires a substantial amount of time be devoted to research. In general, at least 50% protected time, greater if possible, is recommended.

ADMINISTRATION (THE HOSPITALIST‐ADMINISTRATOR)

Physician leaders in hospital administration are not new. Many hospitals already include physicians in senior management positions, such as chief medical officer.44 Naturally, a career in hospital administration is another potential path for diastole in academic medical centers.

CONCLUSIONS

As hospital medicine continues to grow and evolve, designing sustainable and rewarding academic careers will be crucial to the success of the field. Being able to balance clinical systole time with obtaining the skills to support nonclinical diastole time is important to ensuring a successful career as an academic hospitalist. We have described several possible career paths in teaching, research, quality improvement, and administration. By preparing future hospitalists with the knowledge and skills required to assume a variety of roles during their diastolic time, we hope to encourage the growth of hospitalist leaders with well‐developed skill sets. If hospitalists adequately prepare themselves, academic hospital medicine will likely remain sustainable and rewarding, and future generations of trainees will be inspired and prepared to enter the field.

There is a growing demand for hospitalists in the United States. In academic settings, hospitalists are called on to perform a variety of duties, from leading quality improvement initiatives to serving on hospital committees to helping to offset restrictions on work hours of the house staff.1 Although hospitalists may be well positioned to take on these roles, obtaining adequate protected time and recognition for such contributions remains a challenge. The existing promotion and tenure processes at academic institutions may not give adequate consideration to such responsibilities. Hospitalists who do not meet the traditional benchmarks of teaching and research may suffer in their career advancement and, ultimately, in their desire to remain in academics. Developing a sustainable and long‐term career in hospital medicine is important not only from a professional developmental standpoint, but also because it may lead to better patient care; evidence from a large multicenter hospitalist study suggests that physician experience is linked to improved patient care and outcomes.2 Thus, it behooves academic medical centers that employ hospitalists to create rewarding hospitalist career paths.

Goldman described academic hospital medicine as comprising periods of systole, during which hospitalists provide clinical care, and periods of diastole, the portion of a hospitalist's time spent in nonclinical activities.3 Far from being a period of relaxation, diastole is an active component of a hospitalist's work, the time devoted to the pursuit of complementary interests, career advancement, and job diversity. A well‐thought‐out plan for the diastolic phase of a hospitalist job description can lead to significant improvement in quality, education, research, and outcomes for an academic medical center.4 A good balance of systole and diastole allows for focus on career development and advancement and has the potential to be very helpful in preventing burnout. This is of particular concern to academic hospitalists, who report working longer hours, feeling more stress, and worrying more about burnout than their nonhospitalist colleagues.5 This suggests the diastolic phase is an important part of creating a sustainable hospitalist job and should be funded as part of an academic hospitalist position.

Although the optimal balance of systole and diastole to prevent burnout is not known, outlining clear expectations is an important strategy for preparing physicians for a sustainable academic hospitalist career. This is an important issue, given the increasing number of residency graduates who are choosing careers in hospital medicine.6 Based on the reported career plans of residents taking internal medicine in‐training exams from 2002 through 2006, the number of residents going into hospital medicine has more than doubled, from 3% (in 2002) to 6.5% (in 2006). The goal of this article is to compare and contrast several career paths that balance systole and diastole in academic hospital medicine. Specifically, we review training opportunities for becoming a successful hospitalist‐educator, hospitalistquality expert, hospitalist‐investigator, and hospitalist‐administrator.

EDUCATION (THE HOSPITALIST‐EDUCATOR)

Hospitalists in academic centers often play central roles as teachers and leaders in medical education. This is not surprising given that most teaching of medical trainees occurs in the inpatient setting.7 Furthermore, several studies have consistently demonstrated that trainee satisfaction with teaching by hospitalists is high, and hospitalists are rated as more effective teachers than traditional subspecialist ward attendings.810

A typical hospitalist‐educator position is 80%‐90% clinical time, with 10%‐20% set aside for teaching. However, academic hospitalists are often expected to teach medical trainees concurrently with their clinical care activities, rather than during a separate, protected time.11 Thus, most hospitalist‐educator responsibilities do not occur during diastole, as may be conceived, but instead are add to the systole. Small amounts of protected diastolic time for a hospitalist‐educator can be used for related administrative activities, such as writing letters of recommendation, mentoring students and residents, doing creative thinking and curriculum development, and conducting educational research, such as evaluating a new educational program or curriculum. Some hospitalist‐educator positions, such as director of the residency program or internal medicine clerkship, are exceptions in that they generally include a greater amount of protected time, which may be earmarked for administrative activities and hands‐on teaching.

CLINICAL QUALITY AND OPERATIONS IMPROVEMENT (THE HOSPITALISTQUALITY EXPERT)

Hospitalists are increasingly being called on to lead clinical quality and operations improvement at academic teaching hospitals. Benefits to the institution include the consistent presence of a committed physician who is able to plan and execute change in the context of clinical care. This is in contrast to the transient nature of residents and nonhospitalist attending physicians, whose ability to participate in such initiatives is impaired by the scheduling of their rotations. Hospitalists, however, are often able to cultivate long‐standing relationships with nurses, case managers, and hospital administrators, thereby building the institutional clout to lead such initiatives while considering views from all the necessary stakeholders.22 Thus, they are in a good position to serve as physician champions and expedite the adoption of new innovations within hospitalist groups and among other physician groups and clinical staff.23, 24

RESEARCH (THE HOSPITALIST‐INVESTIGATOR)

Few hospitalists devote most of their time to clinical research. Having a strong research base is essential for the field of hospital medicine to gain credibility as a distinct specialty.4 Although the initial research in hospital medicine sought to prove the value of the field itself, hospitalists have now begun to focus on quality improvement and outcomes research.3032 Because of their unique position in clinical care, hospitalists are well situated to oversee inpatient data collection and perform research on a variety of conditions ranging from acute coronary syndromes to venous thromboembolism. Another potential area of research for hospitalists is participation in clinical trials focused on the inpatient setting. Although the proportion of time spent in research can vary widely, to become an independently successful clinical researcher typically requires a substantial amount of time be devoted to research. In general, at least 50% protected time, greater if possible, is recommended.

ADMINISTRATION (THE HOSPITALIST‐ADMINISTRATOR)

Physician leaders in hospital administration are not new. Many hospitals already include physicians in senior management positions, such as chief medical officer.44 Naturally, a career in hospital administration is another potential path for diastole in academic medical centers.

CONCLUSIONS

As hospital medicine continues to grow and evolve, designing sustainable and rewarding academic careers will be crucial to the success of the field. Being able to balance clinical systole time with obtaining the skills to support nonclinical diastole time is important to ensuring a successful career as an academic hospitalist. We have described several possible career paths in teaching, research, quality improvement, and administration. By preparing future hospitalists with the knowledge and skills required to assume a variety of roles during their diastolic time, we hope to encourage the growth of hospitalist leaders with well‐developed skill sets. If hospitalists adequately prepare themselves, academic hospital medicine will likely remain sustainable and rewarding, and future generations of trainees will be inspired and prepared to enter the field.