|Home | About | Journals | Submit | Contact Us | Français|
Patients who leave hospitals against medical advice (AMA) may be at risk for adverse health outcomes. Their decision to leave may not be clearly understood by providers. This study explored providers’ experiences with and attitudes toward patients who leave the hospital AMA.
To explore providers’ experiences with and attitudes toward patients who leave the hospital AMA.
We conducted interviews with university-based internal medicine residents and practicing internal medicine clinicians caring for patients at a community hospital from July 2006 to August 2007. We approached 34 providers within 3 days of discharging a patient AMA. The semi-structured instrument elicited perceptions of care, emotions, and challenges faced when caring for patients who leave AMA. Using an editing analysis style, investigators independently coded transcripts, agreeing on the coding template and its application.
All 34 providers (100%) participated. Providers averaged 32.6 years of age, 22 (61%) were men, 20 (59%) were housestaff from three residency programs, 13 (38%) were faculty, hospitalist physicians, or chief residents serving as ward attendings, and one (3%) was a physician assistant.
Four themes emerged: 1) providers’ beliefs that patients lack insight into their medical conditions; 2) suboptimal communication, mistrust, and conflict; 3) providers’ attempts to empathize with patients’ concerns; and 4) providers’ professional roles and obligations toward patients who leave AMA.
Our study revealed that patients who leave AMA influence providers’ perceptions of their patients’ insight, and their own patient–provider communication, empathy for patients, and professional roles and obligations. Future research should investigate educational interventions to optimize patient-centered communication and support providers in their decisional conflicts when these challenging patient–provider discussions occur.
Patients go to hospitals to receive care for illnesses they cannot manage at home. Appropriate inpatient care requires that providers understand and treat each patient’s presenting complaints and address comorbidities, health behaviors, and psychosocial risk factors that may contribute to poor outcomes. In their report Crossing the Quality Chasm, the Institute of Medicine described optimal inpatient care as patient-centered: “providing care that is respectful of and responsive to individual patient preferences, needs, and values, and ensuring that patient values guide all clinical decisions.”1
One in every 70 hospital discharges occurs against medical advice (AMA).2 Patients who leave AMA may be at risk for adverse health outcomes leading to more frequent hospital readmissions and increased morbidity and mortality.3–6 Prior retrospective chart review studies have found that the likelihood of AMA discharges is associated with hospital factors such as size and location2 and patient factors such as mental illness,4,7 substance abuse,4,5,7–9 gender,2,5,7–9 income,2,5 lack of medical insurance,8,9 whether the individual is a Medicaid recipient,2,9 race,2,8–10 and young age2,5,7,9,11. One chart review found that providers document patients’ reasons for leaving AMA as personal issues, financial problems, and legal matters.12
A patient’s choice to leave AMA may represent a breakdown in communication, which in turn relates to a variety of patient, provider, and relationship factors.13–15 In addition, this decision may create conflict for providers in balancing their ethical obligations: respecting patient autonomy and self-determination, and protecting the patient from harm.16 While some information is known about patient factors associated with leaving AMA, no studies have assessed providers’ perspectives on their interactions with these patients. Since providers may have insight into the events leading to premature departure, our research focused on providers’: 1) actions when they found out a patient wished to leave prematurely; 2) feelings regarding the experience; and 3) lessons learned.
We conducted one-on-one interviews to investigate providers’ experiences with and perspectives regarding patients who leave the hospital AMA. We chose a qualitative methodology, which allows greater in-depth exploration and emergence of themes unanticipated by the investigators than is possible using closed-ended questionnaires.17 We also attempted to telephone patients within 7 days of an AMA discharge to learn their reasons for leaving and perspectives on care. Although multiple attempts were made, we reached only four patients, which we deemed an insufficient number to analyze.
We targeted all internal medicine providers (physicians-in-training, attending physicians, or physician assistants) providing inpatient care for a patient who left AMA from an urban hospital in Connecticut. This 405-bed hospital located in Waterbury, Connecticut, hosts inpatient rotations for three internal medicine residencies from the Yale University School of Medicine, has a 13-provider hospitalist service and a cadre of community physicians who admit patients with and without resident coverage. Waterbury has a population of 107,000 and a per capita average income $4,000 less than the national average.18 See Table 1 for other regional and national demographic comparisons.
From July 2006 to August 2007, we examined daily hospital discharge logs for patients who left AMA. We contacted each AMA patient’s medical providers within three days of discharge in order to obtain the most accurate account of experiences and emotions. One provider had two AMA patients but only the interview about the first patient was included in our analysis. Sampling was discontinued when thematic saturation was achieved.17
We obtained written consent from each participant. The Waterbury Hospital Institutional Review Board and the Yale University Human Investigation Committee approved the study protocol.
One investigator (DMW) conducted audiotaped, semi-structured interviews with consenting respondents in a private setting inside the hospital. Each interview began with the participant describing the medical circumstances surrounding the patient’s admission to the hospital. The interview instrument was in part derived from the Picker Patient Experience Questionnaire,19 a survey instrument with high construct validity and internal consistency designed to examine patients’ hospital experiences. We included the 10 items deemed most pertinent to communication with patients who choose to leave AMA (Appendix). The interviewer used reflective probes to encourage respondents to clarify and expand on their statements. We obtained patient demographic characteristics by chart review.
We analyzed transcripts using an “editing analysis style,” wherein the analysts seek meaningful segments of data to create codes, deriving the coding template from the data itself.17 The two study investigators independently analyzed six randomly selected transcripts to generate a list of codes summarizing the subjects’ statements. Each code could occur more than once in a transcript, and each quotation could be associated with more than one code. Investigators compared codes and negotiated discrepancies to create a consensus coding template, discussing and agreeing on the application of final coding categories. We independently applied this template to all remaining transcripts. Coder agreement was high (κ=0.97). Themes presented were chosen by consensus based on the most frequent codes. We reviewed all available quotations for associated themes and chose representative quotations with the aim of presenting viewpoints from all provider types. To help establish validity of our themes, a random sample of 50% of provider respondents reviewed our data and agreed with our summary.
We managed and analyzed data using Atlas.ti 5.0 software (Atlas.ti GmbH, Berlin, Germany, 2005), a qualitative analysis software developed from a type of editing analysis style called the grounded theory approach.
Patients who left AMA were more likely to be older, male, unemployed, and uninsured compared to regional and national demographics (Table 1). Many patients smoked (75%), and some were involved with illicit drugs (38%).
All 34 clinicians agreed to be interviewed (100% response rate). Providers included resident trainees, attending physicians, and one physician assistant. More trainees had patients who left AMA compared to private or hospitalist physicians. Provider characteristics are described in Table 2. Interviews lasted 8 to 25 minutes.
Providers reported spending an average of 44 minutes with patients discussing their decisions to leave AMA. A majority of providers stated that their patients explicitly told them why they were leaving AMA. However, most providers speculated that patients’ reasons may have been multi-factorial and only partially disclosed to them. Actions taken by providers at the time the patient wished to leave AMA and their assessment of the patient are presented in Table 3.
Four themes emerged from the interviews: 1) providers’ beliefs that patients lack insight into their medical conditions; 2) suboptimal communication, mistrust, and conflict; 3) providers’ attempts to empathize with patients’ concerns; and 4) providers’ professional roles and obligations towards patients who leave AMA. Table 4 shows code counts for these themes.
Theme 1: Providers’ beliefs that patients lack insight into their medical conditions Regardless of patients’ stated motivations, providers felt that patients’ decisions to leave AMA demonstrated lack of comprehension about the dangers of their current medical conditions. This left providers attempting to provide their patients with medical information in hopes of convincing them to stay.
Lack of Insight Concern about patients’ lack of insight — expressed by a majority of providers — was explained by one attending physician this way:
He did not seem to make the connection that dealing with his active inpatient issues was an important part of his recovery from his illness.
One resident felt that his patient did not comprehend what hospitalization and hospital care meant:
I don’t think that the patient ever really wanted to be in the hospital. It is almost like he thought that the hospital was like fast food. You are hungry and you go and get a meal really quickly, and then you get out. That is pretty much it.
Feeling Fine Many providers attributed lack of insight to patients “feeling fine,” either having recovered from their symptoms or having an underlying asymptomatic condition. A resident stated:
I do not think that he had any real understanding of how sick he was. I think that, in his mind, what he was saying was true. He felt better. He was eating. His strength was improved, and there was no reason for him to stay there.
Providing Medical Information Providing medical information to patients was a common strategy for trying to convince patients to stay in the hospital. Providers gave patients detailed medical information regarding the dangers of leaving, including health risks of deferring treatment, need for further diagnostic testing, and potential for serious illness or death.
One intern detailed:
I talked about the risks of him going home. Not being monitored by medical personnel. We talked about the risk of him being home with two young children and no adults. The risk of him going into a seizure and not having anyone to help him, especially when his medications are not therapeutic.
Theme 2: Suboptimal Communication, Mistrust, and Conflict Overall, providers felt that the quality of provider–patient communication was suboptimal, either due to resistance on the patient’s part or miscommunication by providers.
Withholding Information: “I Just Want to Go Home.” One type of suboptimal communication was demonstrated by providers’ perceptions that patients did not truly articulate a reason for leaving AMA. Providers most often said that patients simply stated they did not want to be in the hospital or wanted to go home. As noted by an attending:
All she said was that she wanted to go home and sleep in her own bed.
Similarly, a resident discussed the reasons for his patient’s decision:
He mentioned that he was just tired of being in the hospital. He did not give any particular irritating or inciting problems for those feelings. He said just very nonspecifically a phrase like, “I have things to do.” There was no particular thing that seemed to be bothering him.
Suspicion about True Motives The lack of explicit rationale led some providers to speculate that “wanting to go home” masked patients’ true reasons for leaving AMA, including intentions to engage in less desirable activities. Many providers cited alcohol or other substance abuse as potential motivators, as a resident described:
My thought is that he may not have been interested in doing the rehabilitation, as well as his need to do more substance abuse, which he cannot do in a hospital. When the social worker talked to him, he had no interest in quitting his cocaine or heroin....
This intern explained how suspicion of substance abuse surfaces in her conversations with certain patients who choose to leave AMA:
I think that is a question that we as housestaff ask ourselves every time: ‘Why did this patient want to leave?’ Was he withdrawing from alcohol? He knows that if he goes outside that he will either drink again or can potentially get into serious problems with withdrawal.
Conflict Providers also noted past conflicts with patients. One attending felt that he could never see eye to eye with his patient regarding medical diagnoses:
[The patient] actively disagreed when you suggested that anything else might be wrong.
Another attending relayed a similar story of conflict:
People were telling him that he had colitis, but he said, ‘Well, no, I do not have any diarrhea. They are diagnosing me wrong’.... [I replied,] ‘We are not saying that you have diarrhea, it is just that is what it is termed and so we need to investigate that.’
Another attending described conflict over concerns for patient deception:
I felt just frustrated and taken advantage of. From the beginning when I heard her story it just seemed like this was a woman who was drug-seeking and then, after I met her and examined her, she seemed so fine. I really felt like she was trying to manipulate the system, and I resented it.
Theme 3: Providers’ Attempts to Empathize with Patients’ Concerns Some providers expressed empathy for their patients’ feelings and decisions.
Competing Responsibilities Providers acknowledged that some patients may have competing concerns that they prioritized over their health.
An intern described his patient’s family obligations and commitments:
He stayed with his sister with two young children.... He felt that it was his responsibility to go back home and take care of those kids so that his sister could go back to work.
An attending cited her patient’s responsibilities as a parent and caregiver as motivating factors to leave:
She said that she has some personal issues to take care of, and there was an issue with one of her children and court, and then her grandchildren that she takes care of, and she really needed to get home.
Fear Other providers felt that patients chose to leave because of emotional distress with their medical conditions.
An attending felt that the patient’s fear about his illness was truly behind his desire to leave:
He definitely had some fears that he discussed even on the last day. He had multiple friends who had died from HIV-related complications.
Upset About Care Some providers expressed understanding of why patients might have felt anger or frustration with the care they received in the hospital.
An intern explained how the patient felt uninvolved in his care:
I think that he got frustrated with sitting in a hospital bed and not much was changing, and waiting for something to happen, and not knowing what decisions, advice, or suggestions were being debated among his caregivers.
Communicating Empathy and Respect to Patients The capacity to understand patients’ competing concerns and emotional reactions to their health allowed some providers to try different approaches to discussing patients’ decisions to leave AMA.
The most common communication approach by providers was to spend time with the patient. One intern reflected on the importance of time:
Foremost, I spent a great deal of time with the person. I spent an hour with the patient. I took the time to inquire why the patient wanted to leave the hospital.
Showing caring for patients was also a common communication strategy, as noted by this intern:
If you put in enough effort and communicate in an open, caring fashion, they tend to listen and respect your opinion.
Providers also described the importance of eliciting patients’ concerns. In the words of one attending:
Most of the time, it requires sitting down with someone and identifying their concerns.
This intern recounted a similar strategy with his patient:
Mainly, I started with an opening question to see what was on his mind.... What is the reasoning around his wanting to leave?
Providers also discussed negotiating with patients to meet other concerns and needs in a way that would allow them to stay in the hospital:
I try to sit down and talk to them and try to understand what their frustrations are. I try to get to the core of why they want to leave AMA. I offer them support and try to get social work involved. If necessary, I try to do some give and take. Sometimes the patients may want something as simple as a pass to just leave the floor to see the light of day. Sometimes you can get someone to assist them to go down. Just simple things like that.
Theme 4: Providers’ Professional Roles and Obligations toward Patients Who Leave AMA Providers reflected how their experiences with patients leaving AMA made them reconsider their professional roles and obligations toward challenging patient relationships.
Self-Doubt and Uncertainty about Missed Opportunities Providers most often discussed self-doubt about their inability to secure patients’ diagnoses by the time they left the hospital. One resident noted:
Because he had something going on but we just could not figure out what it was....
Providers also discussed uncertainty regarding other resources that could have been offered to the patient. One attending recounted:
I don’t know what else, what other resources we have to talk to the patient about. It is too complicated, patient care is, and the time we spend.... I am not an expert in social services, but maybe they can spend some time.
Many providers experienced a missed opportunity to interact with their patients prior to their leaving, as explained by one attending:
I just wish that we were able to know that he was leaving. I wish that somehow we could have identified that he was leaving the hospital and tried to intervene.
Missed opportunities often occurred because providers were busy with other patients or activities, and thus unavailable to the patient. One resident reflected:
I did not feel very good about the way that he left. I think that it was mainly because we were not there and there was a lot going on that day.
The Dilemma of Respecting Patient Autonomy In achieving empathy for patients’ non-medical concerns, providers still struggled with their patients’ right to prioritize those concerns over their health. An intern described this dilemma as follows:
Especially in these situations, it is hard to understand a patient’s rationales because, from a care providers’ perspective, health is the most important thing, because that is what we try to address. A lot of time, patients have other circumstances in life that are more important to them than health.
The importance of allowing patients’ their autonomy was acknowledged by one attending who asserted:
I think that it was a difficult situation, a troubling situation because I acknowledge and believe the importance of giving people determination over their own decision making.
Another attending felt that yielding control and autonomy in making decisions would help convince patients to stay:
Sometimes I feel that if you say to the patient, ‘You know you are a grown up and you are free to leave, but these are the reasons why I think that you should stay and this is what is going to happen to you if you leave’ then they often calm down and make the decision to stay.
Maintaining Responsibility versus Accepting Futility Many providers maintained their sense of responsibility to patients who wished to leave AMA, with a majority participants citing ways they would improve future interactions. One intern discussed improving her communication with patients by providing more information regarding their condition and treatment:
I would make more of an effort to help patients understand their condition, the implications of their condition, and the prognosis, and then allow them to make a decision on whether or not to leave the hospital.
An attending took responsibility for eliciting the patient’s concerns:
I would identify the reason that they are in the hospital from their perspectives and try to talk about it at length so the patients do not get disillusioned.
At the same time, an overlapping group of participants expressed the view that they did all they could do to help their patients. A few providers felt that trying to change approaches would be futile. One intern reflected:
I feel like I covered all of the bases with my approach to him, and unfortunately, regardless of our efforts, we can not make a patient do what we want them to do all of the time.
The sense of futility was shared by another intern who said:
Despite all of our efforts, there is no way to ensure 100% of the people are going to stay in the hospital. I think that the lesson to this is that, no matter what we do sometimes you can’t keep everyone in-house. It is unfortunate.
Based on the codes and themes from our study, we have created a pictorial representation from the provider perspective of what providers and patients think, say, and hear when communicating about decisions to leave AMA. (See Fig. 1)
Patients who leave the hospital prior to a desired endpoint present a challenge to many providers. These challenges often evoke strong feelings, including concern for a patient’s health and safety, frustration, and ineptitude.
Four themes emerged from our study. The first involved the providers’ assessment that patients lack insight into their medical condition. To address this issue, providers often spent time discussing the patient’s medical condition and treatment options. Despite attempts at patient education, providers’ and patients’ views toward health, disease, and illness may differ fundamentally. Arthur Kleinman distinguishes between the illness — the patient’s “lived experience” — and the disease — the provider’s biomedical explanation for what is occurring.20 A patient’s meaning of illness may be shaped by their symptoms, how those symptoms translate into knowledge about themselves and the culture of their wider psychosocial environment.20 Thus, the common reaction to leaving AMA — offering medical information to raise the patient’s awareness about their disease — may be insufficient by itself.
In our second theme, we found that suboptimal communication, mistrust, and conflict emerged as potential reasons patients left AMA. A study evaluating views of hospitalization by patients and providers found little agreement between the two groups with respect to justification of hospital stay and discharge planning.21 This discrepancy was felt to be due in part to poor patient-physician communication.21 Another study to determine what patients’ value most during hospitalization established that confidence and trust in providers, treatment with respect and dignity, and adequacy of involvement in care were the most important factors.22 These studies support the importance of effective communication and relationship building skills. Governing and accreditation bodies23–25 and the Institute of Medicine1 also recognize these skills as paramount; they have also been linked to important outcomes including improved diagnostic and clinical proficiency,26,27 reduced patient emotional distress,26 and increased patient and provider satisfaction.28–30
Many providers attempted to empathize with their patients in the third theme that emerged. In doing so, they made an effort to elicit concerns and communicate in an open, caring fashion. Some providers recognized a power struggle between themselves and their patients regarding individual needs. These providers identified the importance of negotiation and working together to find common ground as keys to persuading patients to stay in the hospital. Educators recommend that providers in challenging interactions take an approach characterized by collaboration through improved partnership, and appropriate use of power, negotiation, and empathy.31,32 Opening the lines of communication earlier and often during patient interactions may have prevented many of the misconceptions providers described in our study from occurring (Fig. 1).
Our fourth theme addressed professional roles and obligations toward patients who leave AMA. The code of medical ethics,33 which emphasizes respecting the rights of patients and their right to self-determination, appears to conflict with what providers learn regarding the legality of patients who leave AMA. Literature about AMA patients often focuses on the provider’s duty to determine patient capacity and qualifications for involuntary commitment.34,35 Scholarship in this domain discusses risk management strategies, noting that improving patient-physician communication can lead to decreased patient complaints and fewer malpractice claims.36,37 Articles also articulate the importance of careful and thorough documentation of the actions taken by the patient and the provider in the sequence of events that lead up to and include leaving the hospital AMA.34 These legal and procedural themes echo loudly in providers’ thoughts as they try to balance patient autonomy, potential harm, and beneficence. This leaves some providers conflicted about their roles and obligations.
What strategies could help providers struggling with the four themes described above? Interpersonal conflicts may be remedied if providers embrace different responsibilities in their care. Carrese discusses the exploration of treatment refusal in order to better understand the patient’s beliefs, expectations, fears, and personal needs.16 He recommends exploring “religious beliefs, cultural background, various psychosocial factors, previous interactions with the health care system, influential personal experiences, or the preferences of family members or friends.”16 Beginning in the first interview, this process may reveal that patients don’t actually lack insight, but are informed about their health in different ways. If, despite this effort, a patient still desires to leave before treatment is deemed completed, Swota recommends maximizing patient autonomy through mandatory post-hospital follow-up contact.38 These two methods use a patient-centered approach to care, and move beyond conflict, and allow both parties to effectively carry out their individual roles and obligations.
Providers struggling to come to terms with how these experiences shape their professional identity may benefit from reflecting on their own assumptions, feelings, and attitudes when a patient wants to leave AMA. Based on the themes in this study, we offer eight questions that providers may ask themselves — rather than their patients — to assess their own capacity and insight in these situations (Table 5).
Our study has several limitations. First, as a qualitative study limited to clinicians at a single institution, our findings may not apply to health care providers in other hospitals or settings. At the same time, our study design was strengthened by our high response rate and our inclusion of providers from a variety of levels. Second, our study did not include the perspectives of the patients who left AMA, their significant others, or their other care providers in the hospital. Although soliciting patients’ perspectives was an additional aim of this study, the investigators were unable to recruit sufficient numbers of participants for theoretical saturation. Given the fact that many of these challenges are interpersonal in nature, it is especially important that future research gather data from all stakeholders in the decision to leave AMA.
In conclusion, our study revealed that patients who leave AMA raise questions for providers about their patients’ level of insight, quality of communication, need for empathy and professional roles and obligations. Future research should investigate educational interventions to optimize patient-centered communication and support providers in their decisional conflicts when these challenging patient–provider discussions occur.
None. No financial disclosures.
Conflicts of Interest None disclosed.