Background There is limited information in the literature about reasons for discharges against medical advice (DAMA) as supplied by patients and providers. Information about the reasons for DAMA is necessary for identifying workable strategies to reduce the likelihood and health consequences of DAMA. The objective of this study is to identify the reasons for DAMA based on patient and multicategory provider focus-group interviews (FGIs).
Methods Patients who discharged against medical advice between 2006 and 2008 from a large, academic medical centre along with hospital providers reporting contact with patients who left against medical advice were recruited. Three patient-only groups, one physician-only group and one nurse/social worker group were held. Focus-group interviews were transcribed, and a thematic analysis was performed to identify themes within and across groups. Participants discussed the reasons for patient DAMA and identified potential solutions.
Results Eighteen patients, five physicians, six nurses and four social workers participated in the FGIs. Seven themes emerged across the separate patient, doctor, nurse/social worker FGIs of reasons why patients leave against medical advice: (1) drug addiction, (2) pain management, (3) external obligations, (4) wait time, (5) doctor's bedside manner, (6) teaching hospital setting and (7) communication. Solutions to tackle DAMA identified by participants revolved mainly around enhanced communication and provider education.
Conclusions In a large, academic medical centre, the authors find some differences and many similarities across patients and providers in identifying the causes of and solutions to DAMA, many of which relate to communication.
- against medical advice
- qualitative research
- qualitative research
- quality of care
Statistics from Altmetric.com
- against medical advice
- qualitative research
- qualitative research
- quality of care
Approximately 1% of inpatient discharges are against medical advice.1–3 Discharges against medical advice (DAMA) ie, describing a situation in which the patient leaves the hospital without physician authorisation, are of interest because they are associated with higher patient morbidity and risk for hospital readmission.3–12 Discharges against medical advice are also of interest because they can present ethical challenges for the physician.13 Information from both providers and patients regarding the reasons for DAMA are necessary for identifying workable strategies to reduce DAMA. However, there has been no study to date comparing the reasons for DAMA from multiple perspectives.
Reasons for DAMA among patients diagnosed as having psychiatric conditions or with a history of alcohol abuse have been identified in studies using a variety of methods including the retrospective review of patient medical records, interviews with patients or staff, and follow-up via telephone or mail.14 15 Reasons identified for DAMA from a detoxification/dual diagnosis unit at the Philadelphia Veterans Administration include family emergency obligations, financial obligations and legal issues.15 However, DAMA also occur in other disease settings. Awareness of this is reflected in the increasing number of studies of DAMA among patients admitted to a general (ie, Medicine) service. A few studies have examined reasons for DAMA either from an inpatient setting5 or from an emergency department setting.16–18 In the study by Baptist et al, the reasons for DAMA were identified following a review of the medical records of asthma patients who left against medical advice and included: dissatisfaction with care; patient was feeling better; and family obligations.5 Patient medical records provide a reliable source of information regarding the clinical condition of the patient and may contain limited notes reflecting communications between patients and their providers from the medical staff's point of view; however, these notes likely do not provide the complete picture from patients' perspectives. Direct patient interviews may provide additional information that would be missing from a medical record, due either to the provider's failure to enquire or to the patient's failure to provide complete information.16 To our knowledge, reasons for DAMA based on direct patient and non-physician provider interviews in a general Medicine service have not been reported.
Given these, the current paper sought to explore the topic of DAMA using direct patient and multicategory provider interviews. The objective of this study was to identify the reasons for DAMA among patients admitted to the Medicine service at a large urban teaching hospital from a variety of perspectives. A secondary objective was to compare and contrast stated reasons for DAMA and the suggested solutions, as offered by patients, physicians, nurses and social workers.
Healthcare providers and patients were recruited to participate in focus-group interviews (FGIs) to discuss why patients admitted to the Medicine service at a large teaching hospital leave against medical advice. A total of five FGIs were held: three patient-only groups, one physician-only group and one nurse/social worker group. All FGIs took place in October 2008 in Maryland. Patient and provider groups were held separately in order to minimise incentives to withhold information about the reasons for DAMA. The study was marked as exempt by the University of Maryland Baltimore Institutional Review Board (H30288).
Potential participants for the patient-only FGIs were identified from hospital discharge data. DAMA constitute a standard category of discharge disposition as noted in the hospital discharge record. Hospital discharge data available for analysis were limited to patients aged 18 years or older who left against medical advice between 1 July 2006 and 30 June 2008. To reduce the likelihood that patients required detoxification or psychiatric services, patients with a primary admitting diagnosis of alcohol abuse, drug abuse or psychoses were excluded (see appendix A for ICD-9 codes). Participant recruitment to the patient FGI was based on an initial invitation letter sent via mail. A follow-up telephone call was made after 1 week of non-response to the initial invitation letter. General announcement posters were placed in common areas throughout the hospital inviting providers to participate in the FGI. A $50 honorarium for each participant was set using the wage-payment model.19
Conducting the focus groups
The methodological framework to develop a topic guide was based on the cognitive constructs (perceived susceptibility to health consequences due to DAMA, perceived severity of health consequences due to DAMA, benefits and costs of DAMA) of the Health Belief Model (HBM).20 This topic guide was reviewed by clinicians (ES and MRW), a hospital administrator and a health services researcher trained in qualitative analysis (FGP), and was modified as needed to direct the conversation.
Each FGI lasted approximately 1 h. The provider groups were held in a convenient hospital location, and the patient FGIs were held in a university facility to minimise patient discomfort, given the interview topic. The same moderator (EO) guided all five FGIs. Two research assistants attended the FGIs as observers. All participants were informed that the discussion would be audio-recorded and that the transcriptions would be verbatim, anonymous and confidential. Each participant verbally agreed to these conditions.
The recordings were manually transcribed by EO and MZ. Each transcription was subject to an additional review for accuracy by someone other than the original transcriber, and then the associated audiotapes were destroyed. The researchers independently analysed each transcript and identified themes. A thematic analysis was used to identify themes within the text of all five FGI transcripts.21 Themes were identified separately for the patient group, the physician group and the combined nurses and social workers group, for a total of three groups.
A total of 276 patients meeting the inclusion criteria were contacted by letter. Sixty-two envelopes were returned by the post office, labelled with ‘wrong address/undeliverable.’ One hundred and seventy-one patients did not respond to the letter. There was a 20% ((43/(276–62))×100) response rate for the patient FGIs: 20 used the response card method, and 23 were contacted via telephone call. Ten of these 43 patients did not respond to a follow-up telephone call and, therefore, were not scheduled for an FGI. Thirty-three respondents were placed in a scheduled FGI, with a final participation count of 18 patients due to no-shows. Patient demographic information was collected from discharge data; however minimal demographic information for providers was collected, consistent with the IRB-approved protocol. The average age of patient participants was 45 (range=25 to 63), 61% were male, and 73% were African–American. Five physicians (two males) participated in the physician-only FGI, and 10 females (60% nurses) participated in the nurse/social worker group interview.
Reasons for discharges against medical advice
Seven themes emerged across the patient, doctor and nurse/social worker FGIs of reasons for DAMA: (1) drug seeking patients, (2) pain management, (3) other obligations, (4) wait time, (5) doctor's bedside manner, (6) teaching-hospital setting and (7) communication. Patients' and providers' comments illustrate these themes (see table 1). First, drug-seeking patients are likely to leave the hospital against medical advice to acquire their drug of choice (ie, alcohol, cocaine, heroin). Second, while some patients may leave against medical advice to seek their drug of choice, other patients indicated that they are prejudged by providers who assume that they are drug-seeking when rather they truly are in pain. Some providers echoed this sentiment, reporting that sometimes they assume patients are drug-seeking, when in fact they truly are in pain.
Third, often patients have other obligations to attend to outside the hospital such as childcare, paying rent, attending a court hearing or attending work. Fourth, patients, doctors, nurses and social workers reported that patients left against medical advice because they did not expect an extended hospital stay, the test results were not coming back soon enough, or the doctor was taking too long to conference with the patient. Fifth, patients reported that doctors do not always have proper bedside manner, and both groups of providers admitted that doctors can be rude to patients at times, leading patients to leave against medical advice. Sixth, the teaching hospital setting was described as confusing and frustrating to some patients, causing them to leave against medical advice.
Finally, two types of communication problems were identified as responsible for DAMA. The first communication difficulty is between providers. For example, patients and providers report that patients become frustrated by inconsistent messages from doctors and nurses. Doctors report that they sometimes have to ask the patient about the treatment plan that a previous doctor had established. Nurses and social workers are also aware that there is poor communication between providers and that the poor communication frustrates the patients. The second communication difficulty is between providers and patients. For example, providers report that patients do not understand the severity of their illness when making the decision to leave against medical advice, while patients report that providers use words that are hard to understand. Reasons for DAMA mentioned during the patient FGIs that did not emerge as a theme include: (1) patients leaving because they want a second opinion from another doctor; (2) the patient feels better; (3) lack of health insurance coverage; and (4) the hospital room was dirty.
First, patients would like providers to communicate what they know about the consequences of leaving against medical advice. Second, patients feel that the providers should spend more time convincing patients not to leave.
Nurses' and social workers' solutions
Nurses suggest improved communication between doctors and nurses regarding patient discharge orders and lab tests ordered. Nurses and social workers suggest improved communication between nurses and patients in what concerns sharing information about the general hospital setting, the functioning of a teaching hospital (eg, schedule for morning rounds) and the discharge process as distinct from the discharge event. In addition, nurses and social workers suggest better management of patients' expectations as described in figure 1.
Alterations to the hospital system suggested by nurses to prevent DAMA include nurses attending patient–doctor consultations as a way to stay abreast of patient management. Nurses suggest implementing a system that rewards professional behaviour and outlines consequences for providers who fall short. Also, nurses and social workers identified a need to modify the perception that everyone who asks for pain medication is a drug misuser.
Similar to the nurses' and social workers' solution, doctors indicate that improved communication between provider team members and with patients would help reduce DAMA. However, physicians feel that nurses need to be more responsive to patient phone calls and provide more current information to patients. Other solutions offered include: (1) communicating with the patient's primary care provider; (2) physician education (see figure 2); (3) patient advocate as a resource for patients.
There is no information available in the published literature that compares and contrasts the predictors of a discharge against medical advice in a general inpatient population from patients' and providers' viewpoints. We used the focus-group methodology to identify reasons for DAMA from these perspectives. The use of focus groups is advantageous when collecting the desired information on a large scale is cost-prohibitive, the overall topic of interest is a sensitive one, or the intent is to identify underlying reasons for observed behaviour.22 23 Both patients and providers identified similar reasons for DAMA from a medicine service as identified in table 1. Thus, the current study provides new evidence to support the limited literature24 on patient-reported reasons for DAMA in any disease setting.
The degree of consonance between patients and providers suggests that there is sufficient common ground to support a coordinated intervention effort to reduce the frequency of avoidable DAMA. The need for an intervention effort as the next step towards reducing the number of DAMA has been identified previously.16 A refinement suggested by the combination of patient and provider FGI data from the present study relates to the idea of a coordinated intervention effort. A coordinated intervention effort recognises the joint role of patients and providers in the series of events leading up to a discharge against medical advice and actively seeks both patient and provider input in designing the intervention, defining the scope of action and identifying success metrics. A coordinated effort also builds on the themes and lessons from patient stories regarding both actual and averted DAMA, identifies and addresses disagreements within provider groups regarding which group of providers is most responsible for effectuating change and leverages the common desire for better outcomes shared by patients and providers.
Despite the consonance in reasons, there were subtle indications of dissonance within providers regarding the party responsible for effectuating change. There was a feeling among some doctors interviewed that nurses are responsible for the breakdown in communication with patients and thus bear a larger share of the burden of improving communication with the patient; the opposite viewpoint was expressed in the nurse/social worker groups. For example, a few doctors noted that by the time they are called to speak with the patient, the situation already has progressed to a point beyond which they are able to exert a positive influence. On the other hand, participants in the nurse/social worker group would note that the long time interval between the patient's admission and consultation with the attending physician often contributes to the patient deciding to leave against medical advice. The correlation between perceptions and reality justifies further examination into the extent to which the stated provider viewpoints are generalisable and/or may be strong enough to torpedo interventions designed to reduce DAMA.
Lastly, we examined whether the consonance in identifying potential causes translated into consonance in identifying potential solutions. Improved communication was a common solution identified across the focus groups and in previous literature,24 but this consonance masks differences in conceptualisation and implementation that have not been highlighted to date. Even when participants stated a common definition of ‘improved communication,’ each group placed the responsibility for improving communication with others: patients looked to doctors; doctors looked to patients and to nurses; nurses looked to doctors (see figure 1). For example, patients and providers expressed a need for more honesty in communication but did not agree on the responsible group.
Actionable solutions included: communicating with the patient in an empathetic tone and using lay English terms, nurses attending the patient–doctor consultation, educating physicians about the results from research on DAMA, providing physician education about strategies for communicating with patients likely to leave against medical advice, physician education about the consequences of leaving against medical advice and informing the patient about the patient advocate service. The solutions are directly relevant to discussions in the literature about ensuring fully informed decision-making24 in the case of a patient who seeks to leave against medical advice. The need for communicating with the patient about the consequences of DAMA has been noted previously,16 24 and the current study provides new information suggesting that, in some situations, lack of knowledge of DAMA consequences on the part of the physicians may mean that the patient is less able to make a fully informed decision about whether or not to leave against medical advice.
Some limitations to the qualitative study apply. The FGIs were conducted in a large teaching hospital setting, and the results may not generalise to community or smaller hospitals. Individuals in the patient groups may have withheld relevant information concerning the reasons for DAMA. The moderator involved all participants using direct questions; however, their involvement does not guarantee that all concerns related to DAMA held by the participant were voiced. Non-response bias is a concern for this study due to the low patient participation rate. Despite the promises of anonymity in responses, the nature of the subject matter may have discouraged patients who feel uncomfortable about their unauthorised discharge from responding to the request to participate in the study.
Knowledge of the reasons for discharges against medical advice in a large, academic medical centre can demystify the process of identifying workable strategies to reduce the likelihood of discharges against medical advice. In this study, we identified several reasons, largely related to communication, for discharges against medical advice. Areas of consonance and dissonance related to the definition and implementation of improved communication between patients and (among) providers. The general consonance across patient and provider focus-group interview groups, coupled with the actionable solutions identified by the participants, suggests that effective strategies to reduce discharges against medical advice are possible and would be welcomed.
The authors are grateful for the logistics support provided by I Connerney. The authors also are grateful for the research assistance provided by E Loh. This work was supported by the University of Maryland General Clinical Research Center Grant M01 RR016500, General Clinical Research Centers Program, National Center for Research Resources (NCRR) and NIH. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of the NIH.
Appendix A ICD-9 Codes used for exclusion criteria
Patients with a primary admitting diagnosis of alcohol abuse
ICD9: 265.2, 291.1–291.3, 291.5–291.9, 303.0, 303.9, 357.5, 425.5, 535.3, 571.0–571.1, 980.x, V11.3
Patients with primary admitting diagnosis of drug abuse
ICD9: 292.x, 304.x, 305.2–305.9, V65.42
Patients with primary admitting diagnosis of psychoses
ICD9: 293.8, 295.x, 296.04, 296.44, 296.54, 297.x, 298.x
Funding EO is supported by a National Institutes of Health K12 Career Development Award 1K12RR023250-01.
Competing interests None.
Ethics approval Ethics approval was provided by the University of Maryland Baltimore Institutional Review Board.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.