Mutual Mistrust in the Medical Care of Drug Users

Abstract

OBJECTIVE

Caring for patients who are active drug users is challenging. To better understand the often difficult relationships between illicit drug–using patients and their physicians, we sought to identify major issues that emerge during their interactions in a teaching hospital.

DESIGN

Exploratory qualitative analysis of data from direct observation of patient care interactions and interviews with drug-using patients and their physicians.

SETTING

The inpatient internal medicine service of an urban public teaching hospital.

PARTICIPANTS

Nineteen patients with recent active drug use, primarily opiate use, and their 8 physician teams.

RESULTS

Four major themes emerged. First, physicians feared being deceived by drug-using patients. In particular, they questioned whether patients' requests for opiates to treat pain or withdrawal might result from addictive behavior rather than from “medically indicated” need. Second, they lacked a standard approach to commonly encountered clinical issues, especially the assessment and treatment of pain and opiate withdrawal. Because patients' subjective report of symptoms is suspect, physicians struggled to find criteria for appropriate opiate prescription. Third, physicians avoided engaging patients regarding key complaints, and expressed discomfort and uncertainty in their approach to these patients. Fourth, drug-using patients were sensitive to the possibility of poor medical care, often interpreting physician inconsistency or hospital inefficiency as signs of intentional mistreatment.

CONCLUSION

Physicians and drug-using patients in the teaching hospital setting display mutual mistrust, especially concerning opiate prescription. Physicians' fear of deception, inconsistency and avoidance interacts with patients' concern that they are mistreated and stigmatized. Medical education should focus greater attention on addiction medicine and pain management.

There is a great deal of antagonism set up, because the doctors are the ones with the keys to the “narc” cabinet… and the patients are the ones who need and want the narcotics, both for real and objectifiable and unobjectifiable reasons, and that puts all the doctors in a difficult position.—Senior Medical Resident

When patients who are actively using illicit drugs are admitted to the hospital, physicians in teaching hospitals confront some of their most challenging work.¹ Addressing the dangers of an acute illness while giving appropriate consideration to the addiction that has often led to hospitalization can be complex. Patients may not be ready for addiction treatment, and even if they are, access to such treatment is limited.^2,3 Acute and chronic psychosocial issues can manifest as behavioral problems¹ that, coupled with the stresses of medical training, may be frustrating for physicians and staff.

Historically, physicians have been excluded from a major role in the treatment of opiate addiction.^4–6 The Harrison Act of 1914 and the subsequent prosecution of physicians who maintained opiate addicts in a medical setting provided stark incentives to avoid treating addiction problems.^5–8 Under current law, physicians may treat opiate withdrawal symptoms in addicted patients who are hospitalized for medical conditions other than addiction. Further addiction treatment, including methadone maintenance, is strictly regulated, requiring special state and federal registration.⁹ This isolation of addiction treatment has contributed to the well-recognized lack of physician skills in the screening, assessment, treatment, and referral of patients with substance abuse problems.^10–12

Interest in expanding physician involvement in assessing and treating substance use disorders has been sparked by the recognition that these disorders are associated with a major proportion of preventable deaths.^13,14 Patient outcomes have been improved after brief physician interventions to help patients quit smoking¹⁵ and decrease problem drinking.^16–18 The integration of opiate addiction treatment with methadone into medical practice has been the focus of several successful trials and is becoming recognized as a legitimate treatment for successfully stabilized patients.^19–23 New addiction treatment medications are likely to give further momentum to this integration in the future.^2,24,25

In spite of these advances, little empirical data has informed the physician's role in opiate addiction management in the teaching hospital. To develop preliminary data and describe existing patterns of care and tension in this demanding arena, we investigated the experiences of drug-using patients and their physician teams as they interacted over the course of inpatient treatment in a public teaching hospital.

METHODS

Design

We chose focused ethnography^26,27 as our research method, a qualitative approach especially useful for generating initial themes and hypotheses concerning a relatively unexplored social setting. Researchers use empirical means such as direct observation or in-depth interviews in order to arrive at an understanding of the beliefs and behaviors of participants in a particular context, and allow questions to emerge from these broad inquiries rather than specifying hypotheses in advance. Because predetermined categories of events are not counted, and participants are not surveyed regarding prespecified beliefs, this method does not generate the denominator data needed to report meaningful rates or proportions. Instead, it supports an iterative process of observation and thematic development that strives to incorporate multiple perspectives, and aims to identify critical issues and hypotheses for further study.

In the medical setting, many focused ethnographies rely on the qualitative analysis of in-depth interviews.^28,29 In this study, we supplemented interview data with direct observation of interactions between hospitalized drug-using patients and their physician teams in order both to capture qualities of their relationship that might not be easily articulated³⁰ and to corroborate participants' stated beliefs with observed behaviors.

The study was approved by the Human Subjects Review Committee of the University of Washington in Seattle. Because of the potentially sensitive data collected, a Certificate of Confidentiality was obtained from the National Institute on Drug Abuse. Written informed consent was obtained from all patients and physicians who were tape-recorded, and verbal consent obtained from indirectly involved staff.

Research Setting

The study was conducted over approximately 20 weeks between June and December 1997 on the inpatient internal medicine service of a public urban teaching hospital. At the time of the study, no addiction consultation services were available, and waiting periods for methadone maintenance programs were at least 1 to 2 months.

Participants

Eight inpatient physician teams participated in the study. Patients actively engaged in illicit injection drug or crack cocaine use were recruited during the course of thirty-one 24-hour periods. The admitting resident identified potential patients using routine clinical information available from emergency department records and staff. Patients were approached for consent after the decision to admit had been made, but before the inpatient team had performed their initial assessment.

Data Sources

Patients and their physician teams were followed throughout the patients' hospitalization. Initial interactions between patients and their physicians were observed and tape-recorded when feasible. Morning work rounds were observed each day, as were approximately half of teaching rounds. While the focus was on the patients and their primary medical teams, some interactions between patients and the nursing staff, social workers, and consultants were also observed. These data, along with informal observation and discussion of the course of care with patients, physicians, and staff were documented using ethnographic field notes. These notes included handwritten recordings of conversations, comments, sequences of events, and researcher reflections on developing themes.²⁷

Close to the time of patient discharge, semistructured tape-recorded interviews were conducted with each patient and with at least one of their physicians. Interviews began with open-ended questions about the events of the particular hospitalization (e.g., “How did things go with this patient?”). Areas that patients or physicians identified as troubling or associated with conflict were probed further. Patients and physicians were also asked to reflect on previous noteworthy experiences related to the medical care of drug users. A single researcher (JOM) collected all data after being introduced to both patients and physicians as a physician-researcher uninvolved in the patients' medical care.

Analysis

Tape recordings were transcribed when technically feasible and transcriptions reviewed for accuracy. Field notes, tape recordings, and transcripts were reviewed multiple times throughout the data collection process and coded to identify major themes. Tentative themes were explored in greater depth with subsequent subjects and then modified; this process was repeated throughout the study. Frequently encountered or emotionally charged themes were grouped and studied for patterns and connections, and data discordant with these major themes was particularly noted. Trustworthiness,³¹ the qualitative research analog of reliability and validity, was enhanced through the iterative process of reviewing emergent themes with multiple patient and physician participants. In addition, the data were reviewed with a medical anthropologist and an addictive behaviors expert, and presented for review by colleagues in a variety of settings.

RESULTS

Participants

Twelve percent of the patients admitted to the teams under study were known to be active users of injection drugs or crack cocaine at the time of admission, and were therefore eligible for the study. Nineteen of 27 eligible patients were enrolled, with 7 eligible patients not enrolled due to the logistics of consent during busy call nights. One patient refused to participate, after suffering an iatrogenic pneumothorax.

Patients' median age was 45 (range 32 to 70); 12 were male and 7 female. Eleven were white, 7 African American and 1 Latino. Five had not completed high school, 2 had completed high school and 12 had attended some college. Six were homeless and 2 were employed. Fourteen were current daily drug injectors, 3 had stopped daily use less than 1 month prior to admission (2 in jail and 1 through inpatient detoxification), 1 had been injecting daily until given pain medications for a severe arm infection, and 1 currently smoked crack cocaine and injected heroin occasionally. The main substance of choice was heroin for 18 subjects and crack for 1, though all study patients used more than 1 substance, and 9 regularly injected a combination of heroin and cocaine. The mean duration of use was 13 years (range 1.5 to 35 years). Ten patients had been in methadone maintenance treatment at some time, and all but one had been in some form of drug treatment. Sixteen had a history of drug-related incarceration. Seventeen had previously suffered medical complications of needle use, including 16 with soft-tissue infections, 3 with endocarditis, and 2 with HIV. Hepatitis and overdose were not explicitly assessed. The most common admitting diagnoses were soft tissue (47%) or pulmonary (21%) infections.

Physician subjects included 11 junior residents (“interns”), 1 fourth-year medical student acting as a junior resident, 8 senior residents (post-graduate year 2 or 3) and 8 attending physicians. Twenty-one were male and 8 female; 24 were white, 3 Asian American, 1 African American and 1 Latino. No physician subjects refused to participate.

Major Themes

Four major themes describe the interactions between these primarily opiate-addicted patients and their physician teams. First, physicians were fearful of being deceived by patients with opiate addiction. Second, they lacked a standard approach to assessment or treatment of clinical issues commonly encountered in this setting, especially the management of pain and opiate withdrawal. Third, physicians avoided engaging patients regarding key patient complaints, and expressed discomfort and uncertainty in their approach to these patients. Fourth, patients were sensitive to the possibility of poor medical care, often interpreting physician inconsistency as a sign of intentional mistreatment.

While these themes were manifest in a wide variety of issues, they were particularly crystallized in the context of opiate prescription. All patients in our sample were prescribed opiates for the treatment of pain or withdrawal. Opiate prescription was a common subject during the interactions between physicians and patients during work and teaching rounds and in interviews with both patients and physicians when they were asked to describe previous difficult encounters. While the prescription of opiates was not universally problematic and was only occasionally the subject of direct conflict, it was an area in which both patients and physicians expressed ambivalence and discomfort.

Fear of Deception

Physicians consistently described their apprehension about being deceived by the patients under study, and opiate prescription was a focal point for such fear. They wondered whether requests for opiate treatment might result from the patients' addictive behavior rather than from what physicians might perceive as medically indicated treatment, and feared being manipulated into inappropriate prescribing. While physicians did not express mistrust of every patient, the “legitimacy” of patient requests was an active concern, and they mentioned previous negative experiences with drug-using patients as powerful influences. As described by two medical residents:

All of us go through a little bit of a hitch every time we are requested to prescribe narcotics for our patients…. Are they trying to get more out of me than they really should have? The last thing I want to do is over-dose them or reinforce this behavior (of) trying to coax more drugs out of you.—Senior Medical Resident

When the patient is always seeking, there is a sort of a tone, always complaining and always trying to get more. It's that seeking behavior that puts you off, regardless of what's going on, it just puts you off.—Junior Medical Resident

Even when drug-using patients and their physicians were able to develop good rapport during a hospitalization, physicians commonly viewed this as atypical.

Many patients recognized physicians' fear of being deceived or manipulated, and often wondered whether this influenced their treatment. One 34-year-old man who developed an excellent relationship with his medical team described a previous hospital experience:

Maybe they thought I was coming in to get drugs or something, to get high. I didn't care what they gave. Just a local would have been OK. It's painful to cut into someone's arm like that. I would have thought they would realize that.

Physicians were not alone in expressing negative expectations of drug-using patients. Some of the most emotional comments condemning manipulative, “drug-seeking” patients came from other drug-using patients, who strongly resented the resulting difficulty in obtaining legitimate pain management.

No Standard Approach

Assessments of patients' substance use history, current drug use patterns, and related symptoms were inconsistent. In particular, the evaluation and treatment of pain and withdrawal were extremely variable, with no common approach or clearly articulated standards. Physicians were often aware of their own inconsistency:

I don't know if it's arbitrary, if someone gives me the right feeling, or dupes me enough that they kind of talk me into it…. I give it to some people and am a lot more stingy with other people. I don't know why.—Senior Medical Resident

Everybody has an idea of how to do it, and they are all different.—Junior Medical Resident

Attending physicians rarely gave guidance regarding treatment of pain or withdrawal in spite of house staff uncertainty.

Patients' subjective statements concerning pain or incipient withdrawal were sometimes accepted. At other times, skepticism of patients' motives led to attempts to judge the appropriateness of opiate prescription on the basis of objective evidence such as vital signs.

… since there is this manipulative interaction, almost antagonistic interaction, most doctors take the tack of being cautious, and if in error under-treating, wait for vital sign abnormalities or objective findings, and in the meantime the patients are uncomfortable. We just treat them differently.—Senior Medical Resident

In searching for criteria to determine when opiates might be withheld, some physicians went beyond overt patient signs or symptoms and focused on very subtle clues.

I can tell they are playing games by their intonation, their voice, their body language. They are saying, “I will talk the way you want to get the drugs I need.” It's all veiled in a whole body language to get the drug. Being ill is secondary.—Junior Medical Resident

While all patients eventually received methadone or other opiates to assuage pain or withdrawal, physicians struggled with the rationale for withdrawal treatment. Citing the long waiting lists for methadone maintenance treatment, one resident described the hospital use of methadone as “a bridge to nowhere.” While some physicians saw methadone as a tool to assist patients in adhering to needed medical treatment, others suggested limiting its use to those with more severe medical conditions.

Patients who had had multiple encounters with the medical system noticed this variability most. A 32-year-old musician who had just undergone an incision and drainage of a hand abscess commented:

The last time, they took me to the operating room, put me to sleep, gave me pain meds, and I was in and out in two days…. This crew was hard! It's like the Civil War. “He's a trooper, get out the saw….”

Patients offered various interpretations of physician variability, attributing it to lack of interest, poor clinical skills, or physician bias against drug users.

Avoidance

Physicians focused primarily on familiar acute medical problems, and avoided the more uncertain areas of assessing or intervening in the underlying addiction problem. Even the acute medical issues of pain and withdrawal were evaded by physicians during encounters with patients.

Resident: “Good morning.”

Patient: “I'm in terrible pain.”

Resident: “This is Dr. Attending and Dr. Intern, who will be taking care of you.”

Patient: “I'm in terrible pain.”

Attending: “We're going to look at your foot.”

Patient: “I'm in terrible pain….”

Resident: “Did his dressing get changed?”

Patient: “Please don't hurt me.”

While this dialog shows a profound avoidance of a patient's pain, most other interactions were not so extreme. Nevertheless, patients initiated discussions of pain and withdrawal far more often than physicians, and treatment plans were seldom mentioned, especially during initial interactions. The risk and benefits of methadone treatment were infrequently disclosed. Similarly, while physicians were concerned with the possibility of in-hospital illicit drug use, they discussed it with patients only in reaction to suspicious events. The possibility that such use might result from inadequate treatment of pain or withdrawal was never mentioned.

Patient Fear of Mistreatment

These opiate-addicted patients interpreted physician inconsistency and avoidance as signs of bias. Patients were fearful that they would be punished for their drug use by poor medical care. They were concerned that even delays easily attributable to hospital inefficiency actually represented intentional mistreatment. Even subtle clues to physicians' condescending or hostile attitudes became magnified for patients.

I mentioned that I would need methadone, and I heard one of them chuckle…in a negative, condescending way. You're very sensitive because you expect problems getting adequate pain management because you have a history of drug abuse…. He showed me that he was actually in the opposite corner, across the ring from me.

While nearly all patients expressed some fear of mistreatment, patients who described prior negative encounters with the medical system were most vigilant and least likely to give physicians the benefit of the doubt. They, in turn, were perceived by physicians to be the least cooperative and truthful, and their hospitalizations were characterized by more negative interactions and frustration on the part of both patients and physicians.

DISCUSSION

This study employed a unique ethnographic approach to shed light on the nature of interactions between physicians and opiate-addicted patients in a teaching hospital. Drug-using patients and their physicians were mutually suspicious and uncertain about how to approach each other, and opiate prescription for the treatment of pain or withdrawal was a common focal point of their distrust. Physicians' approaches to these clinical issues were extremely variable. This inconsistency and the avoidance of key addiction and pain issues frequently interacted with patients' fear of mistreatment, resulting in poor communication and frustration. Prior experiences of the exceptionally difficult drug user or of the seemingly abusive and stigmatizing physician powerfully influenced subsequent interactions.

These data are consistent with previous assertions that physicians are hesitant to treat pain in patients with substance abuse problems.^5,32–35 The management of pain is a well-documented area of poor physician performance,^36–38 and when patients are opiate tolerant, the technical difficulty increases.³⁹ Yet in the hospital setting, pain treatment is generally safe, has minimal addictive potential,^39–41 and enjoys legal protections.^5,8 Similarly, the treatment of opiate withdrawal symptoms in the hospital is permitted outside federally licensed addiction programs.^9,42 Such treatment can minimize diagnostic confusion caused by these symptoms⁴³ and allow opiate-addicted patients access to hospital care without mandatory detoxification. Delay in treating pain or withdrawal, whether due to the fear of deception or inexperience, may lead to in-hospital illicit drug use or reduce patients' willingness to remain in the hospital.⁴³ It may also hinder the establishment of a therapeutic relationship that might more effectively address a patient's primary addiction problem.

The broad and multifaceted ethnographic approach used in this study aims to develop specific themes that are inherently exploratory, and that require confirmation and extension through a variety of other methodologies. Should these themes be supported, more general hypotheses could be broached that have implications for medical education in both pain management and addiction medicine.

This research raises fundamental questions about attitudes toward drug users in the teaching hospital. Physicians did not clearly identify patients' primary addiction as a medical disorder requiring careful evaluation and treatment. Thus, problematic behaviors tended to be interpreted in terms of patient deception or manipulation rather than as manifestations of a medical disorder. Physicians felt obliged to guard the keys to the “narc” cabinet, distinguishing the “drug seeker” from the patient with “legitimate” medical issues. Yet this role is in stark contrast to a patient-centered approach to medical care that values the empathetic elicitation of patients' symptoms and autonomous preferences for treatment. Thus, physicians found themselves in a gray area between patient advocacy and police oversight.

The conflicting roles experienced by physicians on the medical wards are arguably not the result of clinical or regulatory demands, since pain and withdrawal symptom management is both clinically prudent and legally sanctioned. Rather, it might be hypothesized that they reflect the influence of negative societal attitudes toward opiate-addicted persons. Previous reports have documented negative physician attitudes toward patients with addictive disorders.^4,44 The pressures to combat addiction through coercive means may spill into the practice of medicine at the expense of optimal patient care practices and the development of empathetic care providers. Thus, another possible hypothesis is that the tension of these conflicting roles threatens the development of core physician values and professional identity at a formative time in physician education. Educational efforts to improve care in this setting might therefore need to address the influence of societal attitudes toward addiction on the development of physicians' professional identity.

Yet physicians did not directly articulate the tension between these dual expectations. They expressed a vague sense of ambivalence or defensiveness but, unlike many patients, did not attribute avoidance or inconsistency to punishment or retaliation for drug addiction. These physician behaviors that appear associated with polarized interactions between physicians and drug-using patients may result in part from insufficient clinical tools to address pain and addiction management. This hypothesis is supported by a well-documented under-emphasis on both pain management and addiction issues in medical education. The additional complexity encountered when both pain and addiction complicate clinical care suggests that teaching these simultaneously might be of benefit. Recent efforts to forge ties between pain and addiction specialists^45–47 are important first steps that could lead to improved training for physicians. The implementation of blended addiction and pain management education may benefit from the involvement of hospital addiction medicine and pain consultation services.

In addition to providing physicians with the tools to assess and treat pain and withdrawal, educational efforts in this arena could also promote more effective counseling approaches to drug users in the hospital setting. A proactive rather than reactive discussion of opiate prescription in the hospital could provide an opportunity to reduce miscommunication and relieve patients' fears. The development of constructive relationships might also be facilitated by motivational interviewing,⁴⁸ a nonconfrontational approach to addressing addictive behavior matched to a patient's stage of motivation. This approach encourages the active exploration of a patient's ambivalence regarding addiction in order to both reduce resistance to behavior change and promote the development of a rapport that gives credence to directive recommendations about the importance of addiction treatment.⁴⁹

The educational deficiencies in both pain management and addiction medicine can be at least partially attributed to the historical separation of addiction treatment from medical practice. Physicians have been discouraged from using opiates to treat pain by the fear of causing addiction or incurring medical board sanctions. Future initiatives enabling physician prescription of medications to treat addiction would give physicians medical tools to feel more effective in the face of addiction and add further relevance to addiction education.

As these data were collected, certain limitations became apparent. Patients with opiate addiction, in contrast to patients with addictions to other drugs, were likely over-sampled, since potential medical treatment of withdrawal gives opiate users a strong incentive to disclose their addiction. Thus, the study findings pertain only to opiate-addicted patients. Patient interactions with nursing staff, the literal holders of “the keys to the ‘narc cabinet’” were not sampled consistently because nurses' unscheduled visits are difficult to anticipate. Similarly, the attending physician perspective is under-represented, because most observations of attending physician encounters with patients were in the context of team rounds the day after admission. The focus of this study was on the central role of medical house officers, highlighting the arena of internists' formative experiences with the care of these patients.

All participants knew they were under observation, and therefore may have avoided more overt statements of bias or acts of hostility. Consequently, this data might underestimate the extent of such behavior for both physicians and patients; it is unlikely that the act of observation made things worse.

This ethnographic analysis relies on data sampled from the complex social environment of the hospital, including interactions between multiple participants in more- and less-formal medical settings (rounds, patient visits, interviews, informal discussions, etc.), rather than a more systematic sampling of a narrower range of data. Thus the observer's ability to accurately and fairly select and record events determines the scope and focus of any data that others might independently assess. While physician and patient subjects were given the opportunity to reflect on the major themes as they were being developed, they did not participate in subsequent analysis and might not have agreed with the final interpretations.

The medical care of opiate-addicted patients in a teaching hospital requires physicians to simultaneously treat acute medical problems, manage pain and withdrawal, and attend to an addiction that has often caused physical and psychosocial devastation. Physicians in this setting often lack important clinical skills and experience in addiction medicine and pain management. Moreover, physicians and patients enter their relationship with an uncertainty and trepidation that may complicate clinical management and threaten the development of professional identity and empathy. Efforts to provide physicians with further addiction and pain management tools, and to integrate addiction interventions into medical settings may be first steps toward altering the dynamic of mutual mistrust that can come between physicians and drug-using patients.