The DECISIONS Study: A Nationwide Survey of United States Adults Regarding 9 Common Medical Decisions

摘要

Patient involvement is required before patients' preferences can be reflected in the medical care they receive. Furthermore, patients are a vital link between physicians' assessments of patients' needs and actual implementation of appropriate care. Yet no study has specifically examined how and when a representative sample of patients considered, discussed, and made medical decisions. Objective. To identify decision prevalence and decision-making processes regarding 1) initiation of prescription medications for hypertension, hypercholesterolemia, or depression; 2) screening tests for colorectal, breast, or prostate cancer; and 3) surgeries for knee or hip replacement, cataracts, or lower back pain. Design. Computer-assisted telephone interview survey. Setting. Nationally representative sample of US adults in households with telephones. Participants. 3010 English-speaking adults age 40 and older identified using a stratified random sample of telephone numbers. Measurements. Estimated prevalence of medical decisions, defined as the patient having initiated medications, been screened, or had surgery within the past 2 years or having discussed these actions with a health care provider during the same interval, as well as decision-specific data regarding patient knowledge, attitudes and patient-provider interactions. Results. 82.2% of the target population reported making at least 1 medical decision in the preceding 2 years. The proportion of decisions resulting in patient action varied dramatically both across decision type (medications [61 %] v. screening [83%] v. surgery [44%]; P < 0.001), and within each category (e.g., blood pressure medications [76%] v. cholesterol medications [55%] vs. depression medications [48%]; P < 0.001). Respondents reported making more decisions if they had a primary care provider or poorer health status and fewer decisions if they had lower education, were male, or were under age 50. Limitations. Retrospective self-reports may incor