Extra ED Physician Shift Fails to Cut LOS

CHICAGO – Adding an additional physician shift in the ambulatory pod of a typical, high-volume academic center with severe emergency department overcrowding did not reduce overall length of stay in the randomized, controlled PICO-D trial.

"Despite the fact you could say this is a negative trial, the physician group almost unanimously voted to keep this intervention because they were seeing fewer patients, but their quality of life was improved," Dr. Brian H. Rowe said at the annual meeting of the Society for Academic Emergency Medicine. "There was more overlap, more discussion, and less staying 3, 4 hours after the end of their shift."

Patrice Wendling/IMNG Medical Media

The unblinded, parallel-group trial compared the traditional schedule of three emergency physician shifts per day (9 a.m. to 5 p.m., 2 p.m. to 10 p.m., and 7 p.m. to 3 a.m.) with four physician shifts per day (9 a.m. to 5 p.m., 1 p.m. to 9 p.m., 5 p.m. to 1 a.m., and 9 p.m. to 5 a.m.) in the ambulatory pod at the University of Alberta Hospital in Edmonton. The schedules were computer generated in 2-week blocks for a total of 12 weeks during mid-2011.

The tertiary care center ED has about 60,000 adult and 30,000 pediatric visits per year, static 8-hour shifts that are similar on weekdays and weekends, and no on-call system. It is the last fee-for-service ED in Canada and has serious overcrowding issues, said Dr. Rowe, a clinical emergency physician at the hospital and professor and research director of emergency medicine at the University of Alberta.

In the 3 months prior to and during the intervention, patient volumes were similar at 15,135 vs. 14,005, as was median patient age (both 46 years) and CTAS (Canadian Triage and Acuity Scale) status (23.8% vs. 22.5% levels 1, 2 [highest acuity]).

Physician initial assessment times decreased significantly from a median of 76 minutes on control days to 69 minutes on intervention days (P less than .001), Dr. Rowe said.

The median length of stay was not significantly decreased by the intervention for admitted patients (10.5 hours to 10.2 hours; P = .27) or for discharged patients (4.1 hours to 3.9 hours; P = .06).

In multiple linear regression modeling that adjusted for confounders such as age, sex, CTAS status, and consultations, the intervention did provide a statistically significant influence on overall length of stay (P = .003), he said.

In addition, the left-without-being-seen rate dropped significantly, from 5.1% on control days to 3.7% on intervention days (P less than .001). The proportion of patients leaving against medical advice was similar during both periods (0.7% vs. 0.5%; P = .08).

Although physicians saw five fewer patients per shift in the ambulatory pod on the intervention days (decreasing from 27 to 22 patients), the ED physicians rated their satisfaction higher, Dr. Rowe observed.

He said that this is the third randomized, controlled trial to be conducted in his institution’s ED, and that the physician group was particularly resistant to changes in the staffing model. Although there has been considerable research on throughput interventions to reduce ED overcrowding, volume-based staffing has been infrequently described in the literature.

Session moderator and emergency physician Dr. Robert A. Lowe of Oregon Health and Science University in Portland asked whether the lack of substantial change in length of stay is an argument that the problem of ED overcrowding is really due to a shortage of inpatient beds.

Dr. Rowe agreed, and said the physician group was able to demonstrate to the administration that despite their efforts, the problem remained.

"So it’s a good argument to administrations to stop blaming us," Dr. Lowe added.

Dr. Rowe reported no relevant conflicts of interest.

CHICAGO – Adding an additional physician shift in the ambulatory pod of a typical, high-volume academic center with severe emergency department overcrowding did not reduce overall length of stay in the randomized, controlled PICO-D trial.

"Despite the fact you could say this is a negative trial, the physician group almost unanimously voted to keep this intervention because they were seeing fewer patients, but their quality of life was improved," Dr. Brian H. Rowe said at the annual meeting of the Society for Academic Emergency Medicine. "There was more overlap, more discussion, and less staying 3, 4 hours after the end of their shift."

Patrice Wendling/IMNG Medical Media

The unblinded, parallel-group trial compared the traditional schedule of three emergency physician shifts per day (9 a.m. to 5 p.m., 2 p.m. to 10 p.m., and 7 p.m. to 3 a.m.) with four physician shifts per day (9 a.m. to 5 p.m., 1 p.m. to 9 p.m., 5 p.m. to 1 a.m., and 9 p.m. to 5 a.m.) in the ambulatory pod at the University of Alberta Hospital in Edmonton. The schedules were computer generated in 2-week blocks for a total of 12 weeks during mid-2011.

The tertiary care center ED has about 60,000 adult and 30,000 pediatric visits per year, static 8-hour shifts that are similar on weekdays and weekends, and no on-call system. It is the last fee-for-service ED in Canada and has serious overcrowding issues, said Dr. Rowe, a clinical emergency physician at the hospital and professor and research director of emergency medicine at the University of Alberta.

In the 3 months prior to and during the intervention, patient volumes were similar at 15,135 vs. 14,005, as was median patient age (both 46 years) and CTAS (Canadian Triage and Acuity Scale) status (23.8% vs. 22.5% levels 1, 2 [highest acuity]).

Physician initial assessment times decreased significantly from a median of 76 minutes on control days to 69 minutes on intervention days (P less than .001), Dr. Rowe said.

The median length of stay was not significantly decreased by the intervention for admitted patients (10.5 hours to 10.2 hours; P = .27) or for discharged patients (4.1 hours to 3.9 hours; P = .06).

In multiple linear regression modeling that adjusted for confounders such as age, sex, CTAS status, and consultations, the intervention did provide a statistically significant influence on overall length of stay (P = .003), he said.

In addition, the left-without-being-seen rate dropped significantly, from 5.1% on control days to 3.7% on intervention days (P less than .001). The proportion of patients leaving against medical advice was similar during both periods (0.7% vs. 0.5%; P = .08).

Although physicians saw five fewer patients per shift in the ambulatory pod on the intervention days (decreasing from 27 to 22 patients), the ED physicians rated their satisfaction higher, Dr. Rowe observed.

He said that this is the third randomized, controlled trial to be conducted in his institution’s ED, and that the physician group was particularly resistant to changes in the staffing model. Although there has been considerable research on throughput interventions to reduce ED overcrowding, volume-based staffing has been infrequently described in the literature.

Session moderator and emergency physician Dr. Robert A. Lowe of Oregon Health and Science University in Portland asked whether the lack of substantial change in length of stay is an argument that the problem of ED overcrowding is really due to a shortage of inpatient beds.

Dr. Rowe agreed, and said the physician group was able to demonstrate to the administration that despite their efforts, the problem remained.

"So it’s a good argument to administrations to stop blaming us," Dr. Lowe added.

Dr. Rowe reported no relevant conflicts of interest.

CHICAGO – Adding an additional physician shift in the ambulatory pod of a typical, high-volume academic center with severe emergency department overcrowding did not reduce overall length of stay in the randomized, controlled PICO-D trial.

"Despite the fact you could say this is a negative trial, the physician group almost unanimously voted to keep this intervention because they were seeing fewer patients, but their quality of life was improved," Dr. Brian H. Rowe said at the annual meeting of the Society for Academic Emergency Medicine. "There was more overlap, more discussion, and less staying 3, 4 hours after the end of their shift."

Patrice Wendling/IMNG Medical Media

The unblinded, parallel-group trial compared the traditional schedule of three emergency physician shifts per day (9 a.m. to 5 p.m., 2 p.m. to 10 p.m., and 7 p.m. to 3 a.m.) with four physician shifts per day (9 a.m. to 5 p.m., 1 p.m. to 9 p.m., 5 p.m. to 1 a.m., and 9 p.m. to 5 a.m.) in the ambulatory pod at the University of Alberta Hospital in Edmonton. The schedules were computer generated in 2-week blocks for a total of 12 weeks during mid-2011.

The tertiary care center ED has about 60,000 adult and 30,000 pediatric visits per year, static 8-hour shifts that are similar on weekdays and weekends, and no on-call system. It is the last fee-for-service ED in Canada and has serious overcrowding issues, said Dr. Rowe, a clinical emergency physician at the hospital and professor and research director of emergency medicine at the University of Alberta.

In the 3 months prior to and during the intervention, patient volumes were similar at 15,135 vs. 14,005, as was median patient age (both 46 years) and CTAS (Canadian Triage and Acuity Scale) status (23.8% vs. 22.5% levels 1, 2 [highest acuity]).

Physician initial assessment times decreased significantly from a median of 76 minutes on control days to 69 minutes on intervention days (P less than .001), Dr. Rowe said.

The median length of stay was not significantly decreased by the intervention for admitted patients (10.5 hours to 10.2 hours; P = .27) or for discharged patients (4.1 hours to 3.9 hours; P = .06).

In multiple linear regression modeling that adjusted for confounders such as age, sex, CTAS status, and consultations, the intervention did provide a statistically significant influence on overall length of stay (P = .003), he said.

In addition, the left-without-being-seen rate dropped significantly, from 5.1% on control days to 3.7% on intervention days (P less than .001). The proportion of patients leaving against medical advice was similar during both periods (0.7% vs. 0.5%; P = .08).

Although physicians saw five fewer patients per shift in the ambulatory pod on the intervention days (decreasing from 27 to 22 patients), the ED physicians rated their satisfaction higher, Dr. Rowe observed.

He said that this is the third randomized, controlled trial to be conducted in his institution’s ED, and that the physician group was particularly resistant to changes in the staffing model. Although there has been considerable research on throughput interventions to reduce ED overcrowding, volume-based staffing has been infrequently described in the literature.

Session moderator and emergency physician Dr. Robert A. Lowe of Oregon Health and Science University in Portland asked whether the lack of substantial change in length of stay is an argument that the problem of ED overcrowding is really due to a shortage of inpatient beds.

Dr. Rowe agreed, and said the physician group was able to demonstrate to the administration that despite their efforts, the problem remained.

"So it’s a good argument to administrations to stop blaming us," Dr. Lowe added.

Dr. Rowe reported no relevant conflicts of interest.