Background
There is an increasing body of empirical research on patient preferences towards health care and healthcare outcomes [
1]. However, the use of this research in healthcare policy decision making seems limited [
2]-[
5]. Much of the literature on the importance of incorporating the patient perspective in healthcare policy decisions focuses on direct, active participation (e.g. membership of committees) [
6]. However, it is questioned whether active participation is the best way of incorporating the patient perspective into healthcare policy decisions [
7]. Using research on patient preferences is another way for incorporating the patient perspective in healthcare policy decisions, which may contribute to the evidence-base of active patient participation [
7],[
8]. This article contributes to the literature on incorporating the patient perspective in healthcare policy decisions by focusing on the use of research on patient preferences. In this paper we focus on two types of decisions: pharmaceutical coverage decisions (i.e. uptake of removal of a pharmaceutical in the benefit package) and clinical practice guideline (CPG) development. Integration of research evidence on patient preferences in pharmaceutical coverage decisions and CPG is important for several reasons. A first argument is that research evidence on patient preferences adds to other knowledge bases and is therefore important for decision making. Both pharmaceutical coverage decisions and CPG are – at least in part – evidence-based and informed by research on safety, effectiveness and/or cost-effectiveness. Likewise, using the available research on patient preferences would make healthcare policy-making more evidence-based with regard to the patient perspective. In addition, studies on patient preferences may report on the relative value of outcomes and experiences, which is not covered by most patient reported outcomes and experience measures used in clinical studies. Further, incorporation of research on patient preferences may foster more patient-centeredness of health care in general [
6], and even so on the individual level (shared decision making). Finally, when constructing quality adjusted life years (QALYs), the general public’s valuation is used for patients’ health states. As patient preferences may differ from preferences of the general public, the patient perspective gets short shrift. Second argument is that research evidence on patient preferences may serve as a source of information for patient representatives in healthcare decision making (empowerment) and provide what representatives bring to the table with more scientific foundation. A final argument is that patient preferences, amongst others, are known to influence the cost-effectiveness of healthcare services [
9]. Furthermore, currently CPG do not always reflect patients’ preferences [
10] which causes low adherence to and low acceptability of guidelines [
11]. Thus, consideration of research evidence patient preferences in healthcare decision making may improve the quality of decisions and yield return on investment on research on patient preferences.
This paper describes the first step of the Patient-VIP study (Patient Values In Policy making) [
12], which is an explorative study on the integration of research on patient preferences in pharmaceutical coverage decisions and CPG. In three sub studies the Patient-VIP study aims to develop a decision framework for the integration of evidence on patient preferences in pharmaceutical coverage decisions and CPG. The objective of the current paper is to explore whether and how such research is incorporated in coverage decision procedures and CPG development. We applied a two-step analysis for this study. First we studied the written guidance for pharmaceutical coverage decisions and CPG development for their content on using research on patient preferences (theoretical perspective). Second we studied specific cases of pharmaceutical coverage decisions and CPGs in order to explore if results of research on patient preferences were considered (practical perspective). We performed both steps for five European countries, namely the Netherlands, England & Wales, Scotland, Germany and France. We choose the Netherlands because it is the origin of the Patient-VIP study. England & Wales, Scotland (as representatives of the United Kingdom), Germany and France were chosen because they are neighbour countries of the Netherlands and the three largest economies in Europe. We also choose these countries because of relevant differences with respect to the funding system of the healthcare system (tax-funded in England & Wales and Scotland versus social insurance-based in France, the Netherlands and Germany). In addition, the chosen countries have leading organisations in guideline development, which will ascertain the availability of literature. Studying the integration of evidence on patient preferences in different countries will reveal similarities and/or differences between countries, which provide valuable information when working towards systematic integration of evidence on patient preferences in pharmaceutical coverage decisions and CPG.
Discussion
The results of this explorative study confirm our expectation that systematic consideration of research on patient preferences is limited in both pharmaceutical coverage decisions and CPG (clinical practice guideline) development, or if it is considered, this is not visible. This result is more strongly present in pharmaceutical coverage decisions than in CPG development. The coverage case studies showed that research on patient preferences were mainly found in the documents text as research findings on quality of life. The CPG guidance documents indicate that the consideration of research on patient preferences is optional (the Netherlands), not mandatory but encouraged (England & Wales) or completely absent (France and Germany). Only the Scottish procedure explicitly mentions a “search for patient evidence”, which is executed before the start of the CPG development and is informative for the development of review questions. Several CPG cases reported research questions relating to patient preferences, for which a literature search was performed, in a way comparable to other research questions. The case studies on CPG showed several references with terms related to preferences. The terms most frequently used were “quality of life”, “needs”, “perspective” and “well-being”.
The absence of instructions for the consideration of results of research on patient preferences in pharmaceutical coverage decisions confirms the limited focus on this type of research in health technology assessment [
3]. In CPG development the importance of incorporating research on patient preferences is more acknowledged, as could be derived from the results of the case studies. The GRADE classification system for the strength of recommendations in guidelines, holds that patient preferences, among others, can affect the strength of recommendations [
47]. Our results on the consideration of research on patient preferences for CPGs compare well with a recent Dutch study, which found that in 13 out of 62 guidelines an attempt was made to use research on patient preferences [
6].
The limited attention for the consideration of research on patient preferences, especially in pharmaceutical coverage decisions, is in contrast with the strong movement towards patient involvement in healthcare decision making. This suggest that there may be barriers for the consideration of research on patient preferences in pharmaceutical coverage decisions and CPG development. Literature mentions several barriers that have normative origin, conceptual origins, methodological origins, procedural origin and practical origin. Normative barrier is that HTA and health economics community, the predominant measure in economic evaluation is the Quality Adjusted Life Year (QALY), which is determined by using preferences of a general population for health states of patients. Brazier et al. performed a review mentioning all positive, normative and methodological arguments both in favour and against the use of patient preferences [
48]. Arguments relate for example to the allocation of collective resources, the foundations of welfare economics, as well as systematic differences between public and patient values. Second normative barrier is that the QALY is derived from patient outcomes for quality of life related to health. However, patients may derive utility from aspects beyond health, e.g. process or organisation of care, which is purposely excluded from the QALY [
49]. It should be determined whether society is willing to pay for the preferences beyond health outcomes. If we decide to incorporate these preferences, the methodological issues of the limitations of the QALY should be addressed. Procedural barrier is that the place of evidence on patient preferences in pharmaceutical coverage decisions and CPG development is not well determined. There is debate on the potential benefits and downsides of the integration of evidence on patient preferences. In the field of CPG development some consider using research on patient preferences as an unsubstantiated add on to already lengthy and complex guidelines that may further decrease adherence, usability and possibilities for transfer of knowledge into translate into practice [
50]. Some view patient preferences as fundamentally a property of the individual and question the usefulness of aggregate preferences for individual decision making [
3],[
50]. Indeed, patient preferences can be very heterogeneous. However, current decision making is also informed by aggregate data on effectiveness and cost-effectiveness. Aggregate data on patient preferences are however not a substitute for elicitation of individual preferences when clinicians discuss treatment choices with their patients. Contrary, aggregate evidence on patient preferences may indicate that decisions are preference-sensitive [
3],[
5] which may stimulate clinicians to help patients construct their individuel preferences. In addition, research in general supports individual decision making, as more information may reduce uncertainty. Other procedural barrier is that the importance attached to evidence on patient preferences is, purposely, not equal to that placed on evidence on effectiveness and cost-effectiveness [
3],[
51]. It remains unclear whether this type of evidence should get some weight in pharmaceutical coverage decisions and guideline development, and how this relates to other evidence.
Barrier of conceptual origin is that previous studies showed that there is much variability in terminology used for preferences [
14]. This is confirmed in our study in the procedures for guideline development and demonstrated in the CPG case studies. We retrieved several discussion papers on the subject of integration of patient preferences in healthcare decision making, using a wide spectrum of terms. In addition to the ambiguity in terminology, there is heterogeneity in the type of study design and measurement techniques that constitute preference related research [
3],[
13]. Furthermore, it is stated that evidence on patient preferences is often of low quality and shows variability in results [
5]. It is argued that the variability in terminology and different measuring methods result in difficulties in the search for these studies, the synthesis of evidence and the judgement whether the results are valid for the purpose of pharmaceutical coverage decisions and CPG. Some authors argue that there is a lack on research on patient preferences, which is considered a practical barrier [
50],[
51]. Indeed, at the time of pharmaceutical coverage decisions are made research on patient preferences is often not available. However, for CPG it is not clear whether there is indeed an absolute lack on this type of evidence or that the available research is not recognised as such. Finally, a methodological barrier is that it is argued that there are no clear existing methods for incorporating evidence on patient preferences in the decision- or development process [
51].
Before systematic integration of research on patient preferences in the procedures of pharmaceutical coverage decisions and CPG development will succeed, the previously described issues need to be addressed. The lack of conceptual and methodological clarity in preference research can be addressed by the development of a conceptual map on preferences, incorporating all contributions from all relevant disciplines and a taxonomy to systematically categorise the different types of research (and measurement techniques), as suggested by Brooker et al. [
13] and Chong et al. [
3]. The procedural issues need to be addressed by reaching consensus on the place of patient preferences in decision making and addressing possible barriers for integration. To address both issues we currently undertake a qualitative study as was prescribed in a previously published research protocol [
12]. This qualitative study constitutes interviews with relevant stakeholders in which the opinions and ideas of relevant stakeholders with regard to the facilitators and barriers for integrating research on patient preferences in pharmaceutical coverage decisions and CPG, as well as the taxonomy for research on patient preferences values, will be addressed. The results of the current study will be used as input for the qualitative study and the subsequent development of a taxonomy for research on patient preferences.
Our study has several limitations. We only studied written documents for answering our research question in this explorative study. Written guidance documents provide guidance on how procedures should be executed and the written pharmaceutical coverage decisions and CPG are the result of the procedure that was followed. We therefore considered them a reliable source of information. We acknowledge however that there may be discrepancies between what is written down and the practical execution that was not reported in the coverage decision or CPG, as decision making can be seen as a dynamic process which is sensitive to contextual factors. Exploring this possible discrepancy requires other research methodologies and was beyond the scope of this paper.
Unfortunately, we did not receive a member check response for both pharmaceutical coverage decisions and CPG for all countries. As a result, we may have missed valuable additional information. However, the responses we did receive did not change our conclusions. We therefore do not expect that additional responses would have changed our conclusions.
Despite the member check, and although we have thoroughly reviewed the national procedures of pharmaceutical coverage decisions and CPG development on the integration of research values, it is possible that we missed documents that may have included more information on this subject. The number of case studies we performed is not sufficient to draw firm conclusions. Our results are at best indicative for current practice., The focus of our study was the use of results of research on patient preferences (passive patient participation) and not on other forms of (active) patient participation (e.g. membership of CPG development group). In the member check the confusion on the difference between active participation and the consideration of research on patient preferences became clear. Indeed we do realise that patient participation is a broad concept which has many forms. Much is already known on active participation in pharmaceutical coverage decisions and CPG development and we intentionally focussed on the consideration of research as little is known on this topic. However, there is a grey area when defining research. Research includes existing, secondary research, but it can be debated whether it also includes results of primary research (e.g. focus groups) performed for the purpose of the specific coverage decision or guideline under development. Another limitation may be that although we included a broad range of terms that referred to patient preferences, this range may have been incomplete, and we may have missed relevant terms. Nonetheless, this would mean the variability in terms used to refer to patient preferences is even larger than presumed, it emphasises the need for more uniformity and transparency in terminology.
Competing interests
The authors declare that they have no competing interests.