Public Expectations and Attitudes for Annual Physical Examinations and Testing

Context Recent preventive health guidelines recommend against comprehensive annual examination of healthy adults. Yet, many Americans are accustomed to receiving such examinations. Contribution This population-based survey of adults in three U.S. cities suggests that many people expect annual physical examinations that include many tests (for example, complete blood counts) that prevention guidelines do not recommend. Conversely, adults do not feel strongly about receiving tests and counseling that have proven benefit. Desires for tests decrease as out-of-pocket costs increase. Implications Public education about preventive health interventions is needed. The Editors The model of the comprehensive annual physical examination advocated by the American Medical Association in the 1920s (1) was the standard of care until the 1970s, when principles of evidence were first applied to the components of the periodic health evaluation (2-5). In recent years, several expert panels have examined the content of and appropriate mechanisms for providing preventive services to asymptomatic adults. In 1979, the Canadian Task Force on the Periodic Health Examination first suggested that the few preventive health care interventions that are well supported by data could be done during visits for short-term and long-term care and did not require scheduled annual physical examinations (6). Since then, the American College of Physicians, the American Medical Association, the U.S. Preventive Services Task Force (USPSTF), and the U.S. Public Health Service have all agreed that routine annual checkups for healthy adults should be abandoned in favor of a more selective approach to preventing and detecting health problems (7-12). Little is known about public acceptance of this change in emphasis. A 1984 study (13) showed that patients in a university-based family practice expected a comprehensive annual physical examination with a battery of routine tests. A more recent survey (14) showed that British general practice patients favored general health screening; however, that study did not evaluate expectations for specific tests. Patient satisfaction with medical care has been linked to expectation for services (15, 16). If the public is unaware that an annual physical examination and accompanying laboratory and miscellaneous testing are no longer considered valuable, the physicians who follow contemporary recommendations risk having dissatisfied patients. Public knowledge and acceptance of these changes in recommendations regarding an annual physical examination are unknown. We designed and implemented a survey instrument to ascertain the public's perception of the need for and content of an annual physical examination. Because medical services usually involve a charge or cost to the recipient, we also ascertained the effect of an imposed dollar cost on public desire for an annual examination. Methods We performed a two-phase study. In phase I, we evaluated public expectations and attitudes of Denver, Colorado, residents about an annual physical examination and preventive health care. In phase II, we assessed public attitudes about an annual physical examination in two other metropolitan areas (Boston, Massachusetts, and San Diego, California) and ascertained the effect of a financial barrier on desire for annual evaluation. Phase I For phase I, we developed a telephone questionnaire that was administered to Denver-area adults Appendix. The questionnaire assessed demographic information (sex, age, ethnicity/race, annual income, education level), usual source of medical care (health maintenance organization [HMO], private-sector clinic, hospital-based clinic, community clinic, or none), frequency of visits to a physician in the past year, the presence of any of five chronic medical conditions (hypertension, heart disease, lung disease, diabetes mellitus, or cancer), and smoking status. These factors were selected to determine whether demographic variables, system of medical care, pattern of physician use, and the presence of one or more chronic diseases are variables that could affect respondents' attitudes about an annual examination. We asked respondents to agree or disagree with the following statement: In addition to seeing my regular doctor when I am sick or for chronic medical problems, I need an annual physical exam. We then asked which items, from a fixed list of history, physical examination, or blood or other tests, should be included in an annual physical examination. The instrument was developed after we reviewed similar published instruments (17). It was reviewed by several physicians knowledgeable in general internal medicine and preventive health care and by an expert with substantial training and experience in survey design and administration; the survey was also pretested on a sample of 20 persons. On the basis of this process, the survey was modified slightly. A professional telephone surveyor administered the survey to Denver-area adults 18 years of age or older, who were selected by random-digit dialing in fall 1997. We estimated that a sample size of 600 respondents would detect 20% differences among respondents and provide sufficient power to detect differences when the data were analyzed according to age, sex, or usual source of medical care. Supplement. Appendix: Data Dictionary: Annual Physical Patient Questionnaire Phase II Phase II was conducted in a new sample, primarily to ascertain whether the relatively high public expectation for annual physical examination found in Denver would be replicated in two other diverse metropolitan areas located in different regions of the country. As we had done for phase I, we determined demographic characteristics, source of medical care, frequency of physician visits, presence of five chronic medical conditions, smoking status, and response to the following statement: In addition to seeing my regular doctor when I am sick or for chronic medical problems, I need an annual physical exam. The remainder of the questionnaire was modified for phase II (Appendix). For some persons, obtaining an annual physical examination involves payment of a fee. To ascertain the potential effect of a financial barrier, respondents affirming the need for an annual physical examination were next asked whether they would still want a physical examination if they had to pay a $150 charge. Subsequently, respondents were asked if they felt they needed eight selected tests (urinalysis; stool tests for blood; chest radiography; mammography and Papanicolaou [Pap] smear [for women only]; and tests for prostate-specific antigen [PSA] [for men only], cholesterol, and blood glucose levels) every 1 to 3 years. Those who affirmed the need for any of these tests were immediately asked whether they would still want the test if payment of a specified amount were required ($10 for urinalysis, $20 for cholesterol test, $20 for blood glucose test, $20 for fecal occult blood testing, $50 for PSA test, $125 for chest radiography, $150 for Pap smear, and $160 for mammography). We determined the charges after surveying several Denver-area hospitals, clinics, and commercial laboratories and determining, on the basis of these sources, mean dollar amounts for all care-related charges, such as facility and laboratory fees and physician charges. The same professional telephone surveyor used in phase I administered the phase II questionnaire in spring 1998 to Boston, San Diego, and Denver residents at least 18 years of age. We selected the respondents by random-digit dialing. For both study phases, 75% of telephone calls were made after 6:00 p.m. The telephone surveyor spoke only English; thus, potential nonEnglish-speaking respondents were excluded. Statistical Analysis We performed statistical comparisons of categorical responses between groups by using the chi-square test for unpaired categorical data and the McNemar chi-square test for paired categorical data (that is, for the effect of charge on test expectations). We compared between-group data for continuous variables using the Student t-test. We used multivariable logistic regression analysis to measure independent associations between expectations for annual physical examination and respondent age, sex, ethnicity/race, education level, annual income level, smoking status, HMO enrollment, regularity of physician visits, frequency of physician visits, and city. We combined phase I and II data for multivariable analysis. We treated age (<65 vs. 65 years), ethnicity/race (white vs. nonwhite), education level (high school vs. >high school), annual income level (<$30 000 vs. $30 000), and frequency of physician visits in past year (0 vs. >0) as dichotomous variables. We did this mainly to provide more stable parameter estimates in the logistic regression model, given the limited sample size, and to yield measures of association that are easily interpreted. Medicare eligibility at 65 years of age was another reason to categorize patients according to age as a dichotomous variable, because propensity to seek care and to expect an annual examination could vary by insurance coverage. Variables were included in the model if the bivariate association with expectations for annual physical examination had a P value less than 0.20. Because only age did not fulfill this criterion, all variables were included in the final model. All analyses were conducted by using SAS software, version 8.0 (SAS Institute, Inc., Cary, North Carolina). Finally, because reporting odds ratios for common outcomes (>10%) can overestimate the magnitude of the association, we converted odds ratios to relative risk (RR) ratios; this was done according to the Flanders and Rhodes method, by using marginal standardization (18). Results are reported as adjusted RRs with 95% CIs, which were computed by using bootstrap resampling. Role of the Funding Source The funding source from divisional funds had no role in t