Patient opinions and desires concerning medical services (e.g. medication, therapy modalities) are often not sufficiently considered. However, taking into account concepts like “patient involvement” and “shared decision making”, an understanding of patient priorities concerning treatment decision making is required. Furthermore, it is important to evaluate whether the attending physicians sufficiently know the preferences of their patients to consider them adequately in (mutual) treatment decisions.
In a previous study (N = 218), we analyzed patient preferences in the treatment of AIDS/HIV using direct assessments and a Discrete Choice Experiment . The study presented here used the same assessments for a survey among physicians (N = 131), however, asking not about their own preferences, but about their view of patient preferences. Subsequently, a Discrete Choice Experiment was used to determine the degree to which each therapy characteristic influenced the decisions.
The survey was offered paper-based and as online questionnaire. The intense usage of the paper version was probably due to the recruitment procedure, which included sending off the paper-based questionnaire. As the recruitment concentrated on physicians known for large numbers of AIDS/HIV patients, the study results can be seen as being representative only for this group of physicians. However, it was supposed that physicians experienced with treating a large proportion of HIV/AIDS-patient are also knowledgeable to patient needs and priorities. The best, albeit most demanding, approach would be to ask physicians and their (real) patients at the same time. However, this was not possible within the scope of this study. In order to achieve highest possible comparability we recruited highly experienced physicians that treat a large number of HIV/AIDS patients. Moreover, HIV/AIDS patients are highly active and well organized in terms of self-help groups. They are highly informed and often have long-term relationships with their physicians, which enables shared-decision making as well as a good judgment by physicians.
For the direct assessment of preferences physicians had to assess 26 treatment characteristics which had been found to be relevant in the previous study  based on literature research and a qualitative pre-study. As expected, a ceiling effect was found with ratings in the upper range of the scales for many of the characteristics. This might be due to the fact that, resulting from the qualitative pre-study, mainly important characteristics were presented. Within the rating, the top of the priority list is marked by preferences such as “drug does not affect or does not much affect appearance”, “self-application is possible”, but also by avoidance of side effects.
In most of the treatment characteristics the assessments of physicians seen in this study and patients found in the previous study  are congruent. In some characteristics, however, there are obvious differences: thus, compared to patient assessments, physicians tend to underestimate the relevance of the characteristics “Drug allows further therapy options in the future”, “Drug does not generate resistance”, “Drug can be used also in case of comorbidities” and “Dosing of drug may vary to current state”. Effects of socio-demographic baseline characteristics of the physicians showed that professionally more experienced physicians assessed “High efficacy”, “Not generating resistance” and “Unlikely long term side-effects” as being more important, while physicians with large numbers of patients for example assessed “Unlikely long term side-effects”, “Improvement of physical state” and of “Emotional and mental state” as more important. Whereas these assessments of more experienced physicians seem to be closer to the patient preferences, the assessments of less experienced physicians might truly reflect a different patient population. However, these findings can also be a misinterpretation of patient preferences due to less patient contacts.
In a second step, preferences were measured using a Discrete Choice Experiment in which the physicians had to choose eight times between two treatment options, which were described by combinations of 6 quality aspects. The influence of each treatment characteristic on the treatment decision was evaluated.
The clearly highest relevance for the treatment selection of physicians as well as patients was found for emotional quality of life indicated by the fact that the disease was not obvious for other persons. From the patient point of view the avoidance of physical impairments such as diarrhea and nausea followed next, and with some distance the possibility of participating in social life – physicians assessed both of these aspects at the same level.
As well as in the previous study on patient assessments of treatment preferences, the DCE results of the current study correspond partly to the direct measurement of needs where emotional, physical and social quality of life were also in the upper range of preferences. Again the characteristics “Maximum increase of life expectancy” and “Avoidance of the risk of long term side-effects” seem to be far less important in DCE than in the direct assessment. This might be due to the fact that the direct assessment using ratings has certain ceiling effects, resulting in all attributes being important. On the contrary the DCE focusses on trade-offs between the attributes. This might result in lower assessments.
As the DCE results show particularly those aspects are weighted high that emphasize HIV/AIDS more as a chronic disease than as directly life threatening condition. As a consequence, therapy adherence of the patients may be increased by better focusing on quality of life arguments than efficacy and safety aspects. However, these arguments differ from results shown in other studies like pill count, dosing frequency and adverse events as shown by Stone et al.  or resistance, regimen convenience and sleep disturbance shown by Beusterien et al. .
Compared to patients’ preferences concerning the therapy of multiple myeloma , the results of the present study demonstrate preference ratings, which may be more affected by the chronic aspect of the disease, whereas preferences of multiple myeloma patients seem to be more affected by the progressive character of their disease leading to increase of life expectancy and possibility of further treatment options as most important factors. Interestingly, in the present study physician assessments of patient preferences were somewhat closer to patient assessments than in multiple myeloma patients. Maybe the preferred recruitment of more experienced physicians in the present study has had an influence and possibly indicates the importance and the need to communicate patient preferences at least to less experienced physicians.
The current study has shown that the preferences of patients and physicians assessing their impression on patients’ preferences were mostly concordant. However, as various studies and a recent review on the concordance of patients and physicians have shown patient preferences and expert judgments can differ  and that the results should always be interpreted in the light of actual circumstances given in a study .
Some limitations of the current study are to be discussed. Because the DCE is only manageable using a limited number of characteristics and pairs to be compared, decisions were made during the composition of characteristics reflecting the factor structure found in the direct assessment and the construction of comparisons . Doing this, however, some problems remained: dosing aspects were presented as maximum 3 tablets per day versus more, whereas the direct preference measurement revealed more facets of application such as dosing according to current health status, treatment free periods or self-application. As these and some more aspects of application affected at least two factors in the factor analysis, the characteristic “dosing flexibility” may have been defined too simple to reflect the complexity of dosing characteristics. Furthermore, the definition of poles of the characteristics may have influenced the preference decisions: the item “increase of life expectancy” was presented as “increase” versus “maximum increase”. Maybe the perceived difference between these two poles was too small to prefer this characteristic against others. Thus, the somewhat surprising low importance of this efficacy parameter may be the consequence of too close definitions of the poles of this characteristic.
Another limitation arises from the two different samples assed for the comparison. The patient recruitment was performed by the German Competence Network for HIV/AIDS, while the recruitment of physicians was performed using a nationwide database from Janssen-Cilag GmbH. Hence the patients in the previous study are maybe dissimilar from those treated by the doctors in this study. Moreover physicians may have based their assessments on thousands of their patients. These patients might not be similar to the 218 patients in the patient sample.