PREVALENCE AND PROCESS OF MEDICAL DECISION MAKING IN THE US: REPORTS FROM THE DECISIONS SURVEY

Purpose: To gather nationally representative data regarding the frequency and process of patient decision making regarding nine common medical decisions.

Methods: The National Survey of Medical Decisions (DECISIONS) is a nationwide, random digit dial telephone interview survey of English-speaking adults age 40 and older who reported initiating prescription medications (for hypertension, hypercholesterolemia, or depression), undergoing cancer screening tests (for colorectal, breast or prostate cancer), having elective surgery (for knee/hip replacement, cataracts, or lower back pain), or discussing these actions with a health care provider within the past two years. Survey questions allowed estimation of national decision prevalence and gathered extensive data regarding patients' perceptions of the decision-making process, patient-provider communications, and other related variables such as information sources and insurance status.

Results: 3,010 respondents completed interviews, corresponding to an 86.5% cooperation rate and an AAPOR RR4 response rate of 51.6%.  71.9% (95% CI: 69.8%, 73.8%) of respondents had discussed cancer screening tests, 42.9% (95% CI: 40.8%, 45.1%) had considered initiating medications, and 16.3% (95% CI: 14.8%, 17.8%) had discussed surgical interventions in the past two years. The proportion of decisions that resulted in an action within two years varied widely across decision types (range: 28%-88%; F=174.66, p<0.001). Patients described significant variations across decisions in who first initiated discussion of the topic (F=14.68, p<0.001), with the highest rate of patient initiation for knee and hip replacement surgery (41%). Across all decisions, many patients reported that health care providers did not ask them about their preferences (range: 20%-66%) or provide them with information about reasons not to take action (range: 20%-80%). Patients' reported role in decision making also varied across decisions. For example, among the cancer screening decisions, prostate cancer screening decisions were least likely to be patient-driven while breast cancer screening decisions were most likely to be so (32% vs. 46%; overall F=5.45, p<0.001).

Conclusions: Medical decisions that have significant life-saving, quality of life, and cost implications are a pervasive part of life for most middle-aged and older Americans but do not always result in immediate action. Patients report frequently making medical decisions without being asked their opinion or receiving information about reasons not to take medical action, two aspects of the decision making process which may directly affect patients' ability to make informed medical decisions.