Acceptance of shared decision making with reference to an electronic library of decision aids (arriba-lib) and its association to decision making in patients: an evaluation study

In shared decision making (SDM), patients are empowered in a way that they actively ask questions and participate in decisions about their health care on the basis of their preferences and values [1, 2]. In clinical practice, patients are not regularly asked about their preferences [3]. Scheibler et al. [4] present results where about one-half of patients want to be involved in decision making, but a far smaller percentage of them are actually involved. The willingness of physicians to involve their patients in decision making is considerably lower than the preference of their patients. Physicians need to assess their patients' preferences before starting to discuss the reason for consultation [5].

Decision aids based on SDM are designed to help patients make informed choices among diagnostic or treatment options by delivering evidence-based information on options and outcomes. They are supposed to supplement the counselling process and can be delivered in different formats before, during, or after the consultation [6]. Decision aids are reported to increase knowledge, reduce decisional conflict, cause greater satisfaction with decision making, support more realistic expectations, achieve a greater likelihood of being able to make a decision, result in an increased association between patient values and decisions, support patient participation, and enhance communication between physicians, patients, and their relatives [7]. Decision aids should not substitute personal counselling because uncertain patients would then be abandoned [8]. Several authors argue for the need to develop evidence-based decision aids for a wide range of clinical applications that should display this evidence on a basic level to be understandable for the patient. They should be interactive so that individual risk data can be entered, and the effects of certain treatments can be seen immediately. Pros and cons can be discussed by using weighted scales [1].

A patient decision aid can be regarded as a complex intervention because it consists of several presumably relevant components. Therefore, there is a need to model components of a complex intervention and to perform exploratory trials to pre-test preliminary versions of an intervention. Outcomes of potential relevance like patient characteristics might then represent endpoints in later controlled trials [9]. Consequently, decision aids require rigorous field testing to assess patients' and physicians' attitudes towards them; recently, several studies attempted to carry out such field tests [10‐12]. Therefore, it is strongly recommended to evaluate decision support systems in a real world setting with multi-perspective, multi-method studies ahead of dissemination for routine use. Such studies should contain a variety of aspects, use multiple methods, apply flexible study designs with longitudinal measures, and do formative and summative evaluations. Most studies in this area concentrate on physicians, not on patients or other users [1, 11, 13‐15].

We undertook a mixed method evaluation study using quantitative and qualitative methods (patient interviews, focus groups). The importance of mixed method research in complex interventions like decision aids is underlined by the study of Protheroe et al. [9]. In their pragmatic randomized controlled trial of a decision aid for women who attended their primary care physician because of menorrhagia, they found that women with less formal education reported greater benefits in qualitative measures. In contrast, the quantitative analysis revealed that women with more formal education benefited most from the intervention. This inter-method discrepancy emphasizes the need for a multi method approach when examining such a complex intervention.

The aim of our study was to evaluate the acceptance of SDM with reference to an interactive, transactional, and evidence-based library of decision aids by patients and physicians in the primary care context.

We performed a mixed method evaluation study. According to the taxonomy of mixed methods as designed by Palinkas et al. [16], we sequentially collected quantitative and qualitative data. Our intention was to use the qualitative method to answer questions raised by quantitative data (function: expansion). We intended to build the qualitative data upon our quantitative data set (process: connect). Here, we present quantitative data of patients on the acceptance of SDM with reference to our electronic library of decision aids (arriba-lib) and its association with decision making. The analyses of our comprehensive qualitative data and the integration of quantitative and qualitative data will be presented in different publications.

The average age of the 29 participating primary care physicians was 52.2 years (sd 5.1 years; range: 43 to 64 years). Eighteen were male (62%). The average time practicing was 14 years (sd 7.5 years; range: 0 to 27 years).

The module for cardiovascular prevention was selected in 128 patients (67%), the diabetes modules in 43 patients (22%), coronary heart disease in 8 patients (4%), atrial fibrillation in 8 patients (4%), and depression in 3 patients (2%).

The 45 patients who had been excluded from the study did not differ significantly from participants regarding gender (χ² = 1.69, p = 0.19) and age (t = 1.69, p = 0.09; d = 0.28). The medium age of the 192 participating patients was 62.4 years (sd 11.8 years; range: 23 to 83 years). There was an equal distribution regarding gender, with 97 males (50.5%) and 95 females (49.5%). A majority of 70.3% had a formal education of eight years or less, 14.6% had a formal education of up to 10 years, and 15.1% of more than 10 years. In our sample, 122 patients (63.5%) preferred SDM with their physicians, and exactly the same proportion mentioned that SDM actually had taken place.

In 46 patients (24.0%) the reason for consultation was a check-up, 34 patients (17.7%) attended their physicians for a monitoring visit, and 20 patients (10.4%) were seen in the context of a disease management programme. The remaining patients came with acute complaints or to discuss results of laboratory examinations.

Two months after the consultation 133 patients (69.3% of the original sample) took part in a short telephone follow-up interview.

We conducted a study to evaluate the acceptance of SDM with reference to an interactive, transactional, and evidence-based library of decision aids and its associations with decision making in patients in primary care practice. The majority of patients perceived that different aspects of the SDM process had actually taken place. Patients were highly satisfied with the encounter. In a brief telephone interview two months after the consultation, two-thirds of the patients stated that they would like to be counselled again with arriba-lib.

There was a high match between preferred and perceived decision making in patients. More than two-thirds of patients said that a decision could be made after the consultation. This was not associated with gender or education, but elderly patients were more likely to say that no decision could be made. More than three-quarters of those reached in the telephone interview implemented the decision within an interval of two months after the consultation. Elderly patients were more likely to say that they had implemented the decision.

Our study has several limitations. It is possible that physicians did not necessarily perform consecutive patient recruitment, and instead treated some patients as usual. This may have led to a positive selection of patients who were already favourably inclined to SDM. This positive selection bias concerning SDM may also be true of the participating physicians because just 32% of the invited physicians took part in our study. Results of statistical analyses within a small evaluation study should always be treated with caution and should be regarded as preliminary [28]. We had no control group in our study, so we cannot compare our results to the situation of usual care.

There is a lack of an accepted primary outcome regarding the use of decision aids. Possible categories to classify measures of effectiveness are knowledge, decision process (e.g., satisfaction and participation preference), decision outcomes (e.g., has a treatment decision been made, adherence), health status, and economic measures. In our sample, patients were highly satisfied with arriba-lib and the process of SDM. Whether this can be solely attributed to the programme is debatable. In a different study, we found that patients were highly satisfied with their physicians regardless of SDM being applied or not [25]. This challenges patient satisfaction to be an adequate measure in evaluating decision aids. It further has to be mentioned that the questionnaires used may have primarily measured satisfaction with the SDM process or just patient satisfaction with their physician in general. Our telephone interviews have the same limitations than other telephone surveys, e.g., social desirability. Validity checks were not possible because we were not allowed to view patient records. It was not recorded what kind of decisions had been made. We were primarily interested in the acceptance of an SDM approach in connection with our electronic library. The implementation of a decision also depends on what kind of decision was made.

Congruence between preferred and perceived decision making predicted adherence to medical decisions, while age and gender did not have any explanatory power [31]. Our results partially support these conclusions of Jahng et al. because we also found a high congruence between preferred and perceived decision making and a high adherence to decisions.

Our findings suggest that our electronic library of decision aids has a positive association with decision making in patients and that future preference for it is high, regardless of patient characteristics like gender, education, or age. Sepucha et al. [32] report that people with lower education, lower income, and high trust in their doctor overestimate their state of being informed about medical issues. Therefore, it is important that physicians check the level of understanding in patients during consultation with a decision aid.

In their updated systematic review, Légaré et al. found time constraints, patient characteristics, and the clinical situation to be the most often reported barriers for the implementation of SDM [33, 34]. On the basis of our results, we agree with the authors that physicians should not assume that patients with certain sociodemographic characteristics are not fit for SDM. Instead, the encounter should be adapted to the individual patient.

Consequently, there is a need for constant evaluation of measures used in the area of SDM. In the near future, we will conduct a randomised controlled trial that will attempt to find active ingredients in the risk-presenting part of the programme. Different methods of risk presentation will be applied. Physicians and patients will be asked which presentation they find most suitable. In another study, we will check the validity of our cardiovascular prevention module by longitudinal epidemiological data.

SDM with reference to our comprehensive electronic library of decision aids (arriba-lib) was accepted by a high number of patients. It has positive associations with general aspects of decision making in patients and can be used for patient groups with a wide range of individual characteristics.