Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

Health Services Research & Development

Veterans Crisis Line Badge
Go to the ORD website
Go to the QUERI website

PPO 09-259 – HSR&D Study

New | Current | Completed | DRA | DRE | Portfolios/Projects | Centers | Career Development Projects

PPO 09-259
Incorporating Patient Complexity into Physician Workload Assessment
Kathlyn E Fletcher MD MA
Clement J. Zablocki VA Medical Center, Milwaukee, WI
Milwaukee, WI
Funding Period: April 2010 - March 2011

BACKGROUND/RATIONALE:
More than 500,000 hospitalizations occur yearly in the VAMC system. Inpatient physician workload is a potential contributing factor to patient adverse events. The current method for approximating inpatient physician workload is census. This is thought to be a poor estimate of workload because it does not take into account any patient factors, such as acuity and comorbidities. It is common nursing practice to generate an acuity score for individual patients in order to allocate work. Drawing on nursing practice and human factors engineering, we propose a pilot study to establish a method for collecting data on physician workload as it relates to individual patients.

OBJECTIVE(S):
1)To establish a method for measuring the workload generated by individual patients. This will be accomplished by directly observing physicians as they admit patients, and then recording the amount of time that physicians spend on individual patients; and
2)To explore the ability of several factors to predict the amount of physician time spent on individual patients during the first 12 hours of admission.

METHODS:
We recruited 25 physicians at the Milwaukee VAMC. Data collection occurred when the physicians were admitting patients. Direct observation was used to perform detailed task analysis. This resulted in a record of how interns spent time overall and also how (and how much) time interns spent on individual patients. Retrospective chart review occurred for up to two patients admitted by each participating physician, and these patients were consented. The chart review included demographics and other data needed to complete standardized instruments assessing acuity. We also assessed subjective workload during the observation by asking the interns to rate their workload at random intervals and by having the observer rate the intern's workload at random intervals, both using a 6-20 scale (20=most workload). We approached the analysis so that we could describe how interns spent their time overall (i.e. on all patients and all tasks during the admission period) and also how they spent their time on individual patients. Analysis included calculating descriptive statistics for the amount of time spent on individual patients in the first 4, 6, and 8 hours after admission. We analyzed data collected from patient chart review and the correlations between patient characteristic with the total amount of time spent by the physician on the individual patients. This work resulted in the establishment of a method to collect the data needed for task analysis of physicians with respect to individual patients. It also provides preliminary data for a larger multi-center study to identify a prospective method for predicting the likely workload associated with individual patients.

FINDINGS/RESULTS:
Intern participation was 69% (25 of 36). Mean intern age was 28.6, and 56% were women. Patient participation was 60% (26 out of 43). Mean patient age was 62.5 and 96% were men. We first looked at the overall distribution of intern time on all patients and tasks. The largest proportion (40%) of intern time was spent in documentation (writing orders, reviewing charts and writing notes). Thirty percent of time was spent on non-patient communication (e.g. talking with nurses and other doctors). Only 12% of intern time was at the bedside. Downtime activities, transit and teaching/learning accounted for 11%, 5%, 2%, respectively. With respect to time spent on individual patients, we found that interns spent a mean of 69 (SD 36) minutes with each new admission in the first 4 hours after admission, 87 (SD 48) minutes after 6 hours and 106 (SD 69) minutes at 8 hours. Few patients were actually observed for 12 hours after admission. The mean workload subjective workload score as reported by interns was 12.17 (3.76) while the mean observer-related workload score was 11.7 (SD 3.13). In a mutivariable model, patient demographic factors (age, gender, co-morbidity score) did not predict the amount of time spent on patients in the first 4 hours of admission, although these were preliminary analyses that were not powered to detect a difference.

IMPACT:
The immediate impact of this work is two-fold. First, it provides information about how interns spend their time on call. We have highlighted how little formal and semi-formal teaching and learning occur during these periods. We have also shown that only 12% of time is spent at the bedsides of patients. These are important findings because shift lengths for interns during this study were up to 30 hours long (although we only observed approximately 14 hours per intern-from 1PM to 5AM), and beginning July 1, 2011 intern shifts will be no longer than 16 hours. This may result in even less time for teaching and learning and patient interactions. The second aspect of this study that is impactful is that this is the first internal medicine project to document how much time is spent on admitting individual patients. The next steps are to consider how to make that time more efficient, safe and patient-centered.

PUBLICATIONS:

Journal Articles

  1. Fletcher KE, Visotcky AM, Slagle JM, Tarima S, Whittle JC, Weinger MB, Schapira MM. Self-Reported Subjective Workload of On-Call Interns. Journal of graduate medical education. 2013 Jun 26; doi: http://dx.doi.org/10.4300/JGME-D-12-00241.1.
  2. Fletcher KE, Visotcky AM, Slagle JM, Tarima S, Weinger MB, Schapira MM. The composition of intern work while on call. Journal of general internal medicine. 2012 Nov 1; 27(11):1432-7.


DRA: Health Systems
DRE: Research Infrastructure
Keywords: Organizational Planning, Practice patterns, Predictive Modeling, Qualitative Methods, Safety, Utilization patterns
MeSH Terms: none

Questions about the HSR&D website? Email the Web Team.

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.