National health and health care surveys provide essential information for health care delivery system analysis and planning. However, because many are oriented toward traditional physician care, they may not accurately reflect the growing impact of nonphysician clinicians such as physician assistants (PAs) and nurse practitioners (NPs).
In this study, we assess several surveys for use in research on PAs and NPs and identify a need for new survey methodologies that can better support incorporation of PAs/NPs into research for health care delivery system analysis and planning purposes. Owing to predicted physician shortages, economists and workforce planners are pursuing strategies to increase the U.S. physician workforce (Cooper et al. 2002; COGME 2005; Association of American Medical Colleges 2006). Because PAs and NPs provide services that often overlap substantially with those of physicians (Scheffler, Waitzman, and Hillman 1996), and because the ratio of approximately one PA or NP for every five practicing physicians is rapidly increasing (Hooker and Berlin 2002; Winget al. 2002; Bureau of Labor Statistics 2006), PAs and NPs could buffer the predicted physician shortage and reduce the number of additional physicians needed to meet the nation’s needs.
Inpatient care is a growing arena for PAs and NPs, but their services are similarly difficult to extract from hospitalization records. Neither the Medical Expenditure Panel Survey (MEPS) Hospital Inpatient files (Agency for Healthcare Research and Quality 2004) nor the Nationwide Inpatient Sample (Healthcare Cost and Utilization Project [HCUP] 2006), both of which collect representative hospitalization data, include data on PAs and NPs. Finally, national health and health care surveys provide data that are used to assess health care needs, quantify care provided, and examine patterns and effectiveness of care. They provide the most detailed information on what providers actually do during patient encounters, so may be the best source for addressing how PA and NP activities compare with and affect physician activities.
We recognize that the surveys were designed for purposes other than PA/NP research, at a time when PAs and NPs comprised a much smaller portion of the health care workforce, so their utility for PA/NP related research might vary. In this investigation, we assess several of the widely used surveys to determine their applicability for research on team care and PA/NP care.
With a predicted physician shortage, an aging and increasingly diverse population, limited resources, and rising health care costs, accurate information about the health care workforce is fundamental for developing strategies to meet future health care delivery needs. Our analysis of existing national health care surveys shows that research based on several widely used surveys may inadequately account for the contributions of PAs and NPs and create a distorted picture of the actual health care provided.
Without modifications to existing surveys or new surveys designed to fully reflect the activities of PAs and NPs, policy makers may underestimate their current and potential contributions and their influence on numbers of physician needed. National health care surveys should illuminate the roles of PAs and NPs in patient care at both the individual provider and health care team levels. Toward this end, we advocate modifying existing national surveys rather than creating additional profession-specific surveys.
Surveys that include only the activities of one profession cannot provide a full understanding of the interplay among the different health professions. Modification of existing surveys will allow comparison and correlation of PA/NP data with existing data and could support analyses of the effects of PAs and NPs on the health care delivery system in general and on the activities of physicians in particular. These team care analyses are critical if we are to address pressing questions about the numbers and ideal mix of providers needed to meet future U.S. needs. Download entire article (PDF)
HSR: Health Services Research 42:5 (October 2007)