HIV/AIDS News & Information

In managed care, financial incentives and utilization review create conflicts of interest for physicians. We sought to determine whether these incentives would lead physicians to deny indicated services. We surveyed internists practicing in areas with at least 30% penetration of managed care. Our questionnaire included four scenarios in which a test or referral is indicated according to clearly established practice guidelines. We randomly assigned physicians to receive one of five versions of the questionnaire, which differed only in the type of reimbursement incentive and utilization review that applied to the scenarios. We received responses from 710 (70%) of 1,009 internists. Although physicians underutilized services regardless of incentives in all scenarios, physicians whose questionnaires depicted full capitation said that they would order fewer services than physicians whose questionnaires depicted fee-for-service. In the scenario in which an x-ray of the lumbosacral spine is indicated for a patient with low back pain, 86% of physicians randomized to the full capitation version said that they would order the test compared to 94% in the fee-for-service version. Similarly, physicians randomized to scenarios requiring utilization review said that they would order fewer services than those randomized to scenarios requiring completion of an insurance form. Scenarios depicting managed care incentives caused consistent, modest underutilization compared to fee-for-service scenarioes, although physicians underutilized services under all financial incentives and utilization review. In response, physicians must develop better methods for detecting underutilization and devise programs to increase the provision of indicated services.

Managed care has grown in response to rising health care expenditures under fee-for-service reimbursement. In managed care, financial incentives and utilization review encourage physicians to decrease services. These incentives and utilization review affect physician behavior and have resulted in lower utilization of health care services in managed care. One concern is that such incentives and utilization review may not differentiate between services that are unindicated and those that are clearly indicated. Incentives that decrease unindicated services benefit patients and promote quality care. Incentives and utilization review under managed care that lead physicians to withhold clearly indicated services create serious ethical conflicts, however, and result in poorer quality of care. Specifically, these incentives create conflicts of interest for physicians and raise concerns that in response to these incentives, physicians may with hold indicated services. Furthermore, such conflicts of interest may erode patient trust in physicians.

Comparisons of the quality of care between managed care and fee-for-service settings find equivocal and inconsistent results. Little rigorous, empirical evidence exists, however, to evaluate whether capitation and utilization review lead physicians to withhold clearly indicated medical services. We sought to determine whether financial incentives and utilization review in managed care lead physicians to withhold services that are indicated according to well-established practice guidelines. We examined scenarios in which services were clearly indicated because these situations raise the strongest ethical concerns regarding harm to patients. To define indicated services, we used evidence-based guidelines published in the medical literature, and we asked about clinical scenarios that fell clearly within these guidelines.

We conducted a randomized questionnaire study of physicians to answer the following research question: “Do financial incentives and utilization review in managed care lead physicians to state that they would deny patients clearly indicated tests and referrals?”

Methods
Subjects
We surveyed internists in the American College of Physicians (ACP) who 1) self-identified as general internists in group or solo practice primarily involved in patient care and 2) practice in metropolitan statistical areas with a population greater than 250,000 and at least 30% penetration of managed care. We excluded physicians who primarily conduct research or practice in a group-model health maintenance organization (HMO) or Veterans Affairs, military, or public institution. cheap levitra online

Questionnaire
Our questionnaire included four case scenarios in which a test or referral was indicated according to evidence-based guidelines published in the peer reviewed literature (Table 1). For each scenario, physicians were asked to indicate whether they would order the test or referral recommended in the guideline and, if not, to write which test or referral they would order, if any. We pretested the scenarios with academic general internists.

TABLE 1 — Cose Scenarios and Indicated Tests or Referrals Included in Each Questionnaire

Physicians were randomized to receive one of five versions of the questionnaire (Figure 1): These versions differed only in the type of financial incentive and utilization review that applied to all four scenarios. We randomized three types of financial incentives: 1) fee-for-service, 2) capitation with a bonus which stated, “assume that each patient has an insurance plan for which you receive a monthly capitation payment with a potential bonus at the end of the year depending on the cost of your referrals and tests”, and 3) full capitation which stated, “assume that each patient has an insurance plan for which you receive a monthly capitation payment from which you pay for all referrals and tests.”

Figure 1.—Randomization to financial incentive and utilization review.

We further randomized according to the utilization review that applied to the scenarios. All fee-for-service versions of the questionnaire required completion of an insurance form as the utilization review. The capitation versions were further randomized to two types of utilization review: 1) review by a committee of colleagues or 2) preauthorization by telephone. Thus, physicians were assigned one of five versions of the questionnaire, with each version including the identical case scenario. flomax canada

Each questionnaire also asked about demographics and practice characteristics.
Our protocol was approved by the University of California, San Francisco Committee on Human Research.

Data Analysis
We recorded the percentage of physicians who stated that they would order the indicated test or referral or who stated that they would order a different test or referral that would achieve the same goal. For the scenario in which the patient with low back pain needs an x-ray of the lumbosacral spine in order to rule out cancer or infection, a referral to an orthopedic surgeon was counted as an appropriate measure because the referral would likely lead to an x-ray of the lumbosacral spine. A computed tomographic (CT) scan or magnetic resonance imaging (MRI) was also counted as an appropriate measure, although referral to physical therapy was not. A treadmill test with thallium or a stress echocardiogram was also scored correct for the patient with atypical angina for whom a treadmill test is indicated. For the scenario in which a referral to a gastroenterolo-gist for upper endoscopy is indicated, we considered any referral to a gastroenterologist or for upper endoscopy correct. For the scenario depicting a man who remains depressed after medical therapy, we considered a referral to any mental health professional correct.

For each scenario, we performed a chi-square test to assess trends in the proportion of physicians who indicated that they would order the test or referral according to the type of financial incentive or utilization review to which they were randomized. We hypothesized that among financial incentives, full capitation would be the most restrictive and fee-for-service the least restrictive. Similarly, we hypothesized that among types of utilization review, preauthorization by telephone would lead physicians to state that they would order the fewest number of indicated tests while completion of an insurance form would lead physicians to indicate that they would order the most. We analyzed the data to examine this predicted trend. We performed all statistical analyses using STATISTICA 4.1 for the Macintosh (StatSoft; Tulsa, OK).