A qualitative study to identify reasons for discharges against medical advice in the cardiovascular setting

BACKGROUND Cardiovascular disease (CVD) is responsible for the largest number of discharges against medical advice (AMA). However, there is limited information regarding the reasons for discharges AMA in the CVD setting. OBJECTIVE To identify reasons for discharges AMA among patients with CVD. DESIGN Qualitative study using focus group interviews (FGIs). PARTICIPANTS A convenience sample of patients with a CVD-related discharge diagnosis who left AMA and providers (physicians, nurses and social workers) whose patients have left AMA. PRIMARY AND SECONDARY OUTCOMES To identify patients' reasons for discharges AMA as identified by patients and providers. To identify strategies to reduce discharges AMA. APPROACH FGIs were grouped according to patients, physicians and nurses/social workers. A content analysis was performed independently by three coauthors to identify the nature and range of the participants' viewpoints on the reasons for discharges AMA. The content analysis involved specific categories of reasons as motivated by the Health Belief Model as well as reasons (ie, themes) that emerged from the interview data. RESULTS 9 patients, 10 physicians and 23 nurses/social workers were recruited for the FGIs. Patients and providers reported the same three reasons for discharges AMA: (1) patient's preference for their own doctor, (2) long wait time and (3) factors outside the hospital. Patients identified an unmet expectation to be involved in setting the treatment plan as a reason to leave AMA. Participants identified improved communication as a solution for reducing discharges AMA. CONCLUSIONS Patients wanted more involvement in their care, exhibited a strong preference for their own primary physician, felt that they spent a long time waiting in the hospital and were motivated to leave AMA by factors outside the hospital. Providers identified similar reasons except the patients' desire for involvement. Additional research is needed to determine the applicability of results in broader patient and provider populations.