Abstract
Introduction: US primary care physicians and their office staff have experienced large increases in time-consuming requirements for prior authorization (PA) of tests, medications, and other clinical services in recent years. This report presents results of 2 similar studies in which physicians and office staff self observed and reported the amount of time spent on PA activities.
Methods: Physicians and office staff from 12 primary care offices in northeastern United States recorded request type, reporter role, and time spent for each PA event at the time of the PA activity. Costs were estimated using salary data from the US Bureau of Labor Statistics (study 1) and from Salary.com (study 2). Time and costs were estimated for the practices in each study.
Results: The mean annual projected cost per full-time equivalent physician for PA activities ranged from $2,161 (study 1) to $3,430 (study 2). Using self-reporting at the time of the event, we found that preauthorization is a measurable burden on physician and staff time.
Conclusions: Further studies that include cost-benefit analyses, estimates of opportunity costs and costs of delayed testing and treatment, as well as the “hassle factor” for patients and physicians, are warranted.
- Electronic Health Records
- Health Insurance
- Practice-based Research
- Practice Management
- Prior Authorization
Physicians across the United States are burdened with an increasing requirement by insurers for “prior authorization” (PA) of services for patients.1 However, there are limited data on the scope and cost of PA for primary care offices. National surveys conducted in recent years support the belief that the PA burden is substantial, with cost estimates to the US health care system of $23 to $31 billion each year,2 or $82,9753 to $85,2764 per full-time equivalent (FTE) physician. In response, the American Medical Association5 has called for standardization of PA costs in response to the administrative and financial burdens that seem to be growing annually.
The studies cited above all relied on the recall of survey respondents. Our 2 research groups in the northeastern United States recently studied the burden of PA using techniques of real-time self-observation, using similar but different methods to capture costs. We arrived at similar conclusions and present the findings of both studies in this report.
Methods
Each study used a “card study” design, with respondents recording the amount of time spent on each individual instance of request for PA of a procedure, medication, or diagnostic test. Each group excluded PA requests for specialist referrals because those have been a part of practice-payer interactions for decades. Data about purpose, amount of time spent, and occupation of each individual who worked on each PA request (physician, mid-level provider, registered nurse, licensed practical nurse, or clerical/administrative staff) were recorded on event forms and collated electronically. A total of 9 practices enrolled in study 1 (from Upstate New York as part of the SALT-Net practice-based research network) and 3 enrolled in study 2 (from Southeastern Pennsylvania).
Costs were extrapolated using 2 slightly different methods. For study 1, data from the US Bureau of Labor Statistics (US average) for 2010 were used to calculate cost per week for primary care practices. Costs per event were estimated by multiplying the US Bureau of Labor Statistics 2010 average hourly wage for each staff type involved by the time recorded in each request. For study 2, average medical wages employed in Pennsylvania were obtained from Salary.com, and these figures were confirmed by the participating physicians and managers as reasonable estimates. For each study, wages were adjusted by a factor of 1.25 to account for 25% fringe costs.
Results
For study 1, 442 cards were submitted and 435 were analyzed (7 included insufficient data). Mean projected annual cost per FTE was $2,161.75 (range, $926.13–$4,604.11) for the 9 practices. For study 2, mean projected annual cost was $3,430.35 per FTE (range, $1,797.31–$6,066.67) for the 3 practices. Results are summarized in Table 1. The largest costs were due to clerical and physician time spent on PA activities. Physician time is limited and seems to be shifted to clerical workers. Cost and time comparisons between staff type in the 2 studies are shown in Figure 1.
Cost of Prior Authorization Requests Over a 4-Week Period in Study 1* and Over a 3-Week Period in Study 2†
Percentage of time and cost recorded for each staff type during each study. PA, physician assistant; RN, registered nurse; LPN, licensed practical nurse.
Conclusions
The estimates of PA costs presented here are dramatically different from those previously published.2⇓–4 Previous estimates of costs for PA exceeding $80,000 per year, and current estimates of $2,120 to $3,420 per year, on average, should be viewed as bookend estimates, with the true costs lying somewhere in between (especially when accounting for opportunity and efficiency costs). The current studies did not capture opportunity costs, such as lost productivity, or costs related to operational efficiency. In addition, we have demonstrated that it is possible for practices to measure their own PA activities and can estimate costs derived from their true expenditures. Given that true costs in individual practices and systems are measurable and that the estimates presented here are certainly low, policy makers, insurers, or large-practice organizations would be ill-advised to act on these results without confirmation in their own settings.
It is important to note that the results presented here do not capture the health-related consequences of delayed or denied patient treatment or the frustration experienced by both physicians (and staff) and patients because of PA requirements, which may lead to burnout and dissatisfaction. On the other side, the studies presented here did not weigh the observed time and associated costs against savings for any health system.
Nevertheless, these studies serve to demonstrate that, however measured, PA requirements have measureable costs associated with them. Further research on opportunity and efficiency costs, with larger samples over longer periods of time, are warranted.
Acknowledgments
Kathleen Barzee, MPH, CPH, assisted with the collection of data and recruitment of practices and in drafting some sections of prior versions the manuscript (study 1 only). Natalie Jones, BS, and Morgan Pratte provided editorial assistance with the manuscript. Brian A. Hannah, MD, MS; Scott Heisman, MS, PMP; Andrew Cymerman, MS; Gregory T. Soltner, DO; Hal Tragash, PhD; and Roger McGruder, MA, assisted in conducting study 2.
Notes
-
This article was externally peer reviewed.
-
Funding: Study 1 was funded by the New York State Academy of Family Physicians and by Health Resources and Services Administration (HRSA) grants D54HP05462/D5AHP19904 and D54HP23297. Life Laboratory (www.lifelaboratory.org) secured funding and provided research design and oversight for study 2.
-
Conflict of interest: none declared.
- Received for publication March 1, 2012.
- Revision received August 14, 2012.
- Accepted for publication August 22, 2012.
References
- 1.
- 2.
- 3.
- 4.
- 5.