We read with interest the article by Iannuzzi et al,1 which, in staffing a large academic health center internal medicine service, presents a financial comparison of direct patient care costs and length of stay between the hospitalist-resident teams and the hospitalist-physician assistant (PA) teams. In this letter, we would like to make 3 points.
First, the comparison between hospitalist-resident and hospitalist-PA teams may not be fair or rational. Residents are trainees in a formal program of graduate medical education, and there is an educational component to their role. Conversely, a PA employed by an academic health center is a permanent employee of the institution, is not in training, uses fewer resources, and their compensation is at least one-third more than a resident's stipend. As trainees, physician residents are subject to supervision from faculty, and they tend to spend time on the wards for a defined period of time. Hospitalist-PAs are more likely to be permanent workers on wards, bringing greater continuity of care. PAs' supervision entails collaboration with physicians as colleagues. In our own observations, PAs also provide assistance to trainees and maintain continuity of care during periods of trainee turnover. All these variables are not accounted for in the article's analysis and could affect the results. We therefore suggest that these differences make the comparison discussed in the article methodologically inappropriate.
Second, the analysis by Iannuzzi and colleagues1 is at odds with findings of similar studies in health services research publications. For example, several recent comparisons have shown not only beneficial effects from the utilization of PA/nurse practitioner (NP) providers as hospitalists but also have demonstrated cost savings.2–4 One study demonstrated that collaborative physician-NP multidisciplinary care management of hospitalized medical patients reduced length of stay and improved profit without affecting readmissions or mortality.2 Another study found that a service staffed by hospitalists and a PA can provide a safe alternative to a resident service for general medicine inpatients.3 A third study found that hospitalist-PA team-based general medicine inpatient care was associated with a higher length of stay, and had similar charges, readmission rates, and inpatient mortality as compared to traditional resident-based teams.4 Other examples are available if general medicine is to be the standard for comparative studies. We do not think that the 25-year-old data in the Knickman et al5 reference to be a contemporary one, and the citation of neonatal intensive care unit providers to be appropriate for discussion references of an adult medicine ward situation.
Third, the authors use the term “midlevel provider” when referring to PAs. This term is considered demeaning to PAs and NPs alike, and suggests that users do not have an understanding of their roles. Does the term presume physicians are at the superior level and nurses at the inferior level of some team arrangement? This type of diminutive collective term is considered pejorative by both health professional associations.6
