How doctors discuss major interventions with high risk patients: an observational study

Introduction

Doctors are increasingly confronted by major, life threatening illness in patients who have serious comorbidities. Decision making in these circumstances poses major difficulties, and doctorsperceive that they are poorly prepared for this task.1 The task is made harder by the need to reach decisions about an acute illness or crisis quickly, often within minutes, and by the probability that the doctor who must make the decision has never previously met the patient.

The choices are usually stark. Withholding treatment almost always results in the patient's death (although not always quickly), while intervention is likely to result in considerable complications and a protracted course, often culminating in death. Neither option represents a good or obvious choice. Withholding treatment without the agreement of a patient or relatives is liable togenerate complaint, but mentioning a treatment that the doctor may feel to be inappropriate may encourage patients and families to demand it.

Despite recognition that this is a common problem, for which doctors are poorly prepared, there has been little research of doctors' discussions when confronted by difficult choices for a patient close to the end of life. We therefore conducted this study of junior doctors' behaviour in such a situation.

Participants and methods

Clinical scenario—We designed a “no win” scenario in which a decision not to operate would result in death (leaking abdominal aortic aneurysm), while comorbidities (end stagecardiac, respiratory, and renal disease) would make it highly likely that intervention would result in serious complications and prolonged stay in intensive care, probably culminating in death. An actor was briefed to take the role of the patient with very limited functional status, who wasstruggling to live alone after the death of his wife and who had been considering death before his aorta ruptured. For the interviews, the actor lay on a trolley in a hospital gown, wearing an oxygen mask and displaying obvious discomfort on movement. The actor was instructed to specifically ask for advice in the event that the interviewing doctor did not offer it.

Participants—Using the same actor and scenario on a weekly basis over four weeks, we invited 30 junior doctors in training (3-5 years after graduation) to speak to the “patient” about his condition and treatment options. The doctors were aware that they were speaking to an actor and were told that the aim of the study was to see what they discussed when faced with a potential major intervention in a high risk patient.

Recording of interviews—Only the actor and doctor were present for each interview. No time limit was set—we instructed the doctors to take their leave when they were satisfied that the interview had reached a natural conclusion. We recorded the interviews on audiotape, and subsequently transcribed the recording. To maintain anonymity, the doctors were not identified on the audiotape or on the transcript. After transcription the tapes were erased on the instruction of the ethics committee.

Scoring of transcripts—Four specialists (three intensive care consultants and one nephrologist) independently reviewed the transcripts (that is, without discussion between them). They had agreed on the aspects of the interviews to be scored before the study. These included explanation of the medical condition and the interventional or non-interventional options and exploration of the patient's prior functional status, values, and fears. In addition, the response to the request for advice was to be noted. Each of the aspects was scored on a four point scale (from 0 (not discussed) to 3 (good discussion)) by each reviewer.

Statistical analysis—We determined a median score for each aspect reviewed by each of the four specialists and combined these scores to provide an overall median (interquartile range) score for each aspect. We compared the combined scores for the different aspects by applying the sign test to the paired differences. We determined inter-rater reliability for the specialists using Kendall's coefficient of concordance.

Results

The four reviewers showed fair inter-rater reliability (Kendall's coefficient of concordance for explanation of condition 0.54, of functional status 0.82, of values 0.65, of fears 0.49, of interventional option 0.56, and of non-interventional option 0.58; all P < 0.0001). The reviewers considered the medical condition to be well described generally (median score 2.7 (interquartile range 2.1-3.0), whereas the patient's functional status, values and wishes, and fears were generally not addressed or were only poorly covered (median scores 0.5 (0.0-1.0), 0.5 (0.0-1.0), and 0.0 (0.0-1.5), respectively; fig 1).

Fig 1

Quality of 30 junior doctors' discussions of treatment options for a patient with acute, life threatening illness and serious comorbidities. (Figure shows median, interquartile range, and range of scores)

• Download figure
• Open in new tab
• Download powerpoint

The interventional (surgical) option was not significantly better described than the non-interventional option (median scores 2.0 (1.25-2.75) and 2.0 (1.25-2.0) respectively; fig 2). The explanation of the interventional option often did not describe the likely postoperative course but remained focused on the immediate surgical risk. Explanation of the consequences of non-intervention was often limited to the probability of death.

Fig 2

Quality of 30 junior doctors' explanations of treatment options for a patient with acute, life threatening illness and serious comorbidities. (Figure shows median, interquartile range, and range of scores)

• Download figure
• Open in new tab
• Download powerpoint

All but one doctor indicated that they wished to involve the family in discussions, but none identified a preferred surrogate decision maker. Six doctors (20% (95% confidence interval 6% to 34%)) suggested that the patient should speak with his family alone to help them to arrive at a decision (without the doctor being present to explain the medical situation). The doctors were reluctant to give advice, despite it being requested: two (6% (0 to 14%)) stated that a doctor could not give advice (stated reasons were “My job as a doctor is to give you the options,” “I can't answer that, because I am not you,” “I have given you the options, you'll have to decide”), while 17 (57% (39% to 75%)) simply restated the medical risks without offering advice. Of the 11 (36% (18% to 53%)) who did give advice, eight (72%) advocated intervention.

Discussion

This study showed that the junior doctors explained the medical situation reasonably well but generally did not consider the extent to which disease was affecting the life of the patient. Nordid they consider the treatment options in the context of the patient's fears or hopes. They focused on the disease rather than approaching the problem as a composite of the disease and the person.

Patients consider it to be important for doctors to give attention to patient values rather than just to the disease process.2 This is not a new issue.The eminent physician William Osler (1849-1919) wrote: “It's more important to know what sort of person this disease has, than what sort of disease this person has.” This view is probably even more pertinent today, with the advance of medical specialisation, the ageing population, and the vastly increased range of possible treatments that can be offered.

Proper informed consent requires a balanced presentation of the available treatment options. Our analysis of the junior doctors' presentation of the interventional and non-interventional options indicates that there is plenty of room for improvement.