Username:
Abstract
Study Sites
- University of Alberta
Additional Resources
A Structured Approach to Shift Scheduling in the Emergency Department: Implementation of Simple Statistical Methods to Evaluate Physician Workload
Jeffrey Michael Franc
Manuela Verde
Darren Nichols
Francesco Della Corte
Objectives
The University of Alberta hospital emergency department provides care for approximately 150 adult patients each day. Emergency physician consists of seven shift of eight hours each. Informal observation suggests that the patient volume appears to be unequally distributed among the shifts, with certain shifts being routinely over-worked while others and under-worked. A statistical analysis of the patient volume seen during each shift may allow a more rational approach to shift scheduling.
MethodsPatient volume for forty-nine consecutive shifts was obtained by direct observation from the computer tracking system. .
Hypothesis
Test Statistic
Statistical Methods
H0: Patient volume is equal for all shifts
H1: Patient volume is not equal for all shifts
F-test
Differences in patient volume between each of the seven shifts was compared using Analysis of Means (ANOM). Possible cofactors including daily patient volume, average triage score and operator, were assessed using ANOM and Analysis of Variance(ANOVA).
Results
Mean patient volume per shift was 18.8 (SD=7.5). ANOM revealed a significant difference between shifts with a large range between the highest and lowest volume shift (10.9 – 29.4). Three shifts were consistently below mean patient volume, and two consistently above. Patient daily volume and average triage score were not shown to be significant using ANOVA. Unfortunately, it was difficult to separate operator effect from the shift effect due to the small sample size and large number of operators. Simple statistical functions for analysis of patient volume are presented.
Analysis of Variance
Response: total
Df Sum Sq Mean Sq F value Pr(>F)
date 6 47.55 7.925 0.5161 0.7867303
operator 21 1500.84 71.468 4.6543 0.0024360 **
shift 6 903.24 150.540 9.8037 0.0002419 ***
avg.ctas 1 6.09 6.092 0.3967 0.5389394
Residuals 14 214.98 15.355
---
Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
 |
 |
|
Conclusions
The null hypothesis of equivalent patient volumes between shifts was rejected, as some shifts clearly performed above the mean and others below. A suggestion for an alternative shift pattern was described, which should be implemented on a trial basis and the study methods repeated. Further studies are likely to be indicated to further characterize the effect of average triage score and operator. Future replication of the study methods at other sites is suggested.
|
About Us |
|
|
|
|
|