Issues in Determining Appropriate Levels of Hospitalist Staffing
- Staffing model: shift vs. call
- Scheduling approach: number of days on/off
- Non-patient care responsibilities: teaching, research, committees, etc.
- Staffing philosophy: part-time vs. full-time preference
Benchmark information is extremely helpful in determining physician capacity for a hospital medicine program. These benchmarks provide a point of comparison for hospitalist leaders developing staffing models. Medians for inpatient, non-patient, and on-call hours from the 2004 SHM Productivity and Compensation Survey are documented in Figure 5 (page 52).
The simplified example in Box 6, based on Kaiser time estimates, illustrates how demand, work, and physician capacity can be used to determine the number of hospitalists required to support a program.
As an alternative methodology or for comparative purposes, RVUs can be used rather than time. Box 7 uses RVUs from Figure 4 (initial hospital care: 1.28 RVUs; subsequent hospital care: .64 RVUs; hospital discharge < 30 minutes: 1.28 RVUs). The lowest level RVU values are used because they are consistent with the Kaiser example. Also, the median RVUs per year from Figure 5 are used (2961 for a hospital-based program).
Understand Your Work Environment
When a hospitalist program leader begins the process of developing a staffing model, it is important that he or she understands how the unique goals and characteristics of the program affects staffing. For example:
- Hospitalist-only groups are often driven by revenue. It is likely that these programs will expect hospitalists to do more billable work (i.e., see more patients)
- Academic programs typically have a broad range of other, non-patient care responsibilities, including teaching, research, and committee work. The hospitalists in these programs may see fewer patients.
The data from the 2004 SHM Productivity and Compensation Survey (Figure 5) confirms these differences. For inpatient hours worked, the national medians for these two different employment models differ by 23% (1700 vs. 2210). For RVUs worked, the national medians for the two different employment models differ by 17% (3000 vs. 3600).
Summary
Determining the right level of hospitalist staffing is important because it can positively or negatively affect the hospital medicine program and the hospital. Understaffing can lead to physician burn-out and adversely affect physician performance and hospital utilization. Overstaffing can affect the program’s financial performance and undercut the credibility of the program. The right staffing models and formulas, however, can help create a successful hospitalist work environment.
Summary of Recommendations
- There is no industry standard for a hospitalist staffing model. The analysis can be time-based or RVU-based, census driven, or based on any combination of output measures.
- Inpatient utilization drives the requirements for hospitalist staffing. A thorough analysis of historical inpatient utilization data is critical to developing a staffing model.
- In addition to understanding past utilization, projecting future inpatient demand is also important. Critical change factors include trends in: 1) the age and severity of patients; 2) population growth or decline; 3) payer sources; and 4) referral patterns.
- The services (work) performed by the hospital medicine program should be clearly identified and factored into the staffing formula. Brainstorming with the hospitalist group can be an effective technique for ensuring that the analysis is credible.
- Stakeholders should be involved early and often in developing a staffing model and in making staffing decisions.
- In developing a staffing model, particularly in the beginning stages, focus on the process and the methodology and not on the outcome (i.e., “my program needs 6 physicians”).
- Understand how the unique goals and characteristics of your hospital medicine program affect your staffing model.
