There has been an increasing emphasis in recent years on the need to involve patients more fully in decisions about their own care. This reflects the importance of respecting patient autonomy in contemporary bioethics as well as the belief that better informed patients will have better outcomes,1^–4 will comply better with treatment5 and will make choices that reflect their own values rather than those of professionals.

However, previous population-based and clinical studies suggest that a substantial proportion of patients do not wish to be active participants in making decisions about treatment,6^–17 although there is considerable variation depending on treatment settings and conditions studied. Also, there may be a discrepancy between patients’ wish for information and their desire to participate in actual decision-making; thus, patients who want to receive information about their condition are not always willing to use that information to make the final decision about their treatment.8^–11

Patient preferences for involvement in decision-making may differ according to age, gender, socio-economic status, severity of illness and the importance of the decision to be made. A recent systematic review of the role of demographic variables in determining preferences regarding participation in decision-making found that younger, better-educated patients and women were consistently more likely to prefer a more active role in decision-making.13 In several studies, age was the most important predictor of patient preference for participation in decision-making, with older people, defined as those aged 70 or more, less likely to want an active role in decision-making.610^–12

Most studies to date have been conducted in stable outpatient populations6911 or in patients with a single well-defined condition1310 or facing a particular decision.13^–17 The preferences of patients in these settings and the problems they face may differ from those of patients who are acutely ill. We examined patient preferences for information and for participation in decision-making, and the determinants of these preferences in patients recently admitted to an acute hospital.

METHODS

Participants

Consecutive acute medical admissions to Merlin Park Hospital, a 300-bed acute teaching hospital, were considered for inclusion in the study. We excluded those who were considered by their doctors to be too ill to participate, those with major communication or cognitive difficulties and those who refused to participate.

The study was approved by the local research Ethics Committee.

Assessment

Each patient underwent a standardised assessment conducted by a study doctor who had not previously been involved in his or her care. The same doctor also interviewed the clinician responsible for the patient’s care. Demographic information, including age, sex, occupation and socio-economic group (5-group classification18) and details of the patient’s illness were recorded.

Abbreviated mental test

This 10 item test provides a brief assessment of cognitive function.19 A cut-off of 7/8 has satisfactory sensitivity and specificity for identification of cognitive impairment. In this study, we automatically excluded those with a score less than 3. Other patients were excluded only if significant difficulties with understanding and responding emerged during the interview process.

Severity of illness

A subjective rating of overall illness severity as mild (1), moderate (2) or severe (3) was made on initial assessment by the study physician based on the approach of Charlson and colleagues.20

Degner control of preferences scale21^–23

Degner Control of Preferences Scale allows the patient to choose one of five levels of involvement in decision-making ranging from (1) “I prefer to make the final decision about what treatment I will receive” through (3) “I prefer that my doctor and I share responsibility for deciding which treatment is best for me” to (5) “I prefer to leave all decisions regarding my treatment to my doctor.” We followed Degner in classifying the choice of cards 1 or 2 as preferring an active role, the choice of card 4 a collaborative role and cards 4 or 5 a passive role in decision-making.22

The five Degner cards were placed in front of the patient in a random arrangement. The patient was then encouraged to read the cards and, with their own condition in mind, to pick out the one which they most agreed with. The same procedure was followed with the physician looking after the patient. This clinician, who was unaware of the patient’s choice, was asked to predict the patient’s choice of card.

Desire for information

Patients’ desire for information about their illness was assessed by asking them to respond to the question “How much information do you want to receive about your condition?” by marking one of the responses, ranging from (1) “very extensive” to (5) “none,” on a 5-point Likert scale. Again, the same scale was administered to the patient’s physician, who was asked to predict the patient’s response.

Analyses

Results of the Degner scale and the desire for information scale were regarded as the primary outcome measures. It was expected that these data would not have a normal distribution, and Spearman’s rho or chi square tests were used to examine the relationships between these outcomes and potential predictors such as age, sex, socio-economic group, severity of illness and cognitive status. The kappa statistic was calculated as a measure of agreement between patient preferences and clinician prediction of those preferences; this has a value of 1.0 when agreement is perfect, a value of −1.0 when disagreement is perfect and a value of 0 when agreement is no better than chance.24

We wished to have 80% power, with a 0.05 type I error probability, to detect a 15% difference in the proportion of patients older and younger than 70 years opting for an active role in medical decision-making. The sample size requirement for this task was 152.

RESULTS

Of 189 patients admitted during the study period, 152 (80%) satisfied the inclusion criteria. Thirty-seven patients were excluded because of one or more cognitive/communication problems (26), being too ill to participate (16) or refusal (1). Demographic and clinical details of participating patients are shown in table 1.

Of the 152 patients, 93 (61%) favoured a passive approach to decision-making (either “leave all decisions to the doctor” or “doctor makes final decision but seriously considers my opinion” (fig 1). In contrast, 101 (66%) patients sought “very extensive” or “a lot” of information about their condition. There was no relationship between patient preferences for involvement in decision-making and for information about their condition (Spearman rho correlation 0.02, p = 0.8).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Patient preferences for involvement in decision-making and for information.

Of the 152 patients, 88 (58%) were aged 70 years or more. There was no difference in the proportion of older and younger patients seeking an active role in decision-making (14/88 (16%) vs 10/64 (16%), Pearson χ² 0.002, p = 0.96) or seeking “a lot” or “very extensive” information (54/88 (61%) vs 47/64 (73%), Pearson chi square 2.4, p = 0.12).

Table 2 shows the relationships between the specified outcome measures and predictor variables. There was no clinically or statistically significant correlation between any of the predictors and the Degner scale. There was a weak (0.19) significantly positive association between increased cognitive function and increased desire for information.

There was no agreement between patient preferences on the Degner scale and clinical prediction of those preferences (kappa = 0.04, p = 0.7); there was agreement in 34 pairs, clinician overestimation of patient preference for involvement in decision-making in 74 pairs and clinician underestimation of that preference in 44 pairs. There was a weak (kappa = 0.17) but statistically significant agreement (p = 0.006) between patient preferences on the desire for information scale and clinical prediction; there was agreement in 55 pairs, clinician overestimation of patient desire for information in 29 pairs and clinician underestimation of that desire in 68 pairs.

DISCUSSION

There has been a major shift in medical culture in recent years from an attitude of medical paternalism, where physicians may make decisions with little reference to patients’ preferences, to one in which respect for patient autonomy and informed participation of patients in medical decisions are predominant values. Shared decision-making is often seen as the ideal model for the doctor–patient relationship, since it respects both the superior knowledge of the former and the rights of patients to have their own values and preferences inform decisions.1217 However, it does not enhance patient autonomy, and might even be harmful, to force participation in decision-making on patients who do not want it.25

This study indicates that many acute hospital patients prefer a relatively passive role in decision-making. This is in accordance with the findings of previous population-based and clinical studies.6^–13 Our results also support the distinction noted in some previous studies between patients’ desire for information about their condition and their desire to be actively involved in making the ultimate decision.8^–10 Like Strull and colleagues,6 we noted that clinicians underestimated patients’ desire for information and overestimated their wish to participate in medical decision-making.

Many studies have examined the relationship between demographic and clinical variables and patient preferences.8^–15 However, even if accurate predictors of preferences can be identified at a population level, they will not help determine the preferences of individual patients. In any event, in the current study of acute medical inpatients, there were no clinically significant and few statistically significant associations between clinical or demographic details and decision-making or information-seeking preferences. In particular, age was not a determinant of patient preferences. These results reinforce the necessity of discussion with the individual if shared decision-making is to be achieved.

The major strength of this study is that it was conducted in an unselected and representative group of acutely ill medical admissions. Preferences have not been assessed in this population before. The study included a high proportion of older patients, and patients with mild cognitive impairment were included if possible. Patients made their choices with their own current condition in mind, rather than relying on hypothetical vignettes or future decisions. Thus, the results are more applicable to day-to-day hospital practice than those of previous studies that have focused on stable out-patient populations.

Our results suggest that many patients do not subscribe to the ideal of a shared decision-making process. However, a shortcoming of the present study is that it did not examine the actual decisions faced by individual patients. A study of preferences for involvement in specific decisions might yield different results. There is some evidence, for example, that patients prefer to have more input into major than into minor decisions and that preferences may change at different stages of their illness.1326

The relationship between patients and their doctors may also affect patients’ information and decision-making preferences. Many patients, and indeed clinicians, may be poorly informed of the potential benefits from active participation in decision-making.27 Education of patients about these benefits may improve their willingness to take a more active role, especially in those circumstances where knowledge of the patient’s values and circumstances is essential to making the best decision.12 For example, in a recent study of patients with prostate cancer, willingness to participate in decision-making increased significantly after the use of a decision aid.28 The authors’ suggestion that preferences should only be elicited after adequate education has occurred warrants further assessment.

Identifying factors that influence patients’ preference for involvement in decision-making may help make clinicians more sensitive to the communication needs of their patients. However, there is also a risk that it might lead to attempts to predict patient preferences based on stereotypes or assumptions, for example about older people. The most important finding of this study is that patients’ preferences for information and for involvement in decision-making are highly variable. They cannot be predicted from demographic or clinical data or by clinician judgement and can only be determined by asking the individual patient. Identifying patient preferences in this manner will allow doctors to provide better patient-centred care and communication.