OBJECTIVE: The objective of the research described in this report was to develop a framework to evaluate the economics associated with the use of patient decision aids (PDA) used within a shared decision making (SDM) process.
METHODS: A systematic review of existing economic evaluations of PDAs, and a literature review of systematic reviews of PDAs were undertaken. Studies identified were summarised, outcomes extracted and tabulated, and a thematic analysis was conducted to identify main patterns and themes that emerged from the data extracted from the reviews. Input and opinions of specialist experts in the field of PDAs and SDM were obtained during an interactive workshop. The results generated from these three pieces of work were used to inform and develop a conceptual framework for economic evaluations of PDAs used in a SDM process.
RESULTS: Literature reviews: Just five existing economic evaluations of PDAs were identified. The PDAs evaluated were used in a variety of conditions covering either primary or secondary care, with 4 evaluations set in the UK and 1 in Finland. The main limitations of the existing evaluations were the short time horizons (maximum 2 years), the outcomes reported (only one presented a formal incremental cost per quality adjusted life year (QALY)), and the restricted focus within the evaluation (i.e. the effects of patient satisfaction or preferences on health related quality of life were not incorporated). A review (2014) including 115 studies of RCTs of PDAs (compared to usual care and/or alternative interventions) was used as the basis for the second review. The main outcomes assessed included the attributes of choices made and the attributes of the decision-making process. Secondary outcomes included behavioural, health outcomes, and health-system effects. The RCTs covered decisions ranging from screening through treatment and surgery, and predominantly related to prostate cancer screening (n=15), colon cancer screening (n=10), or hormone replacement therapy (n=10). In summary, comparing the use of PDAs to usual care, PDAs improved people’s knowledge of the options available, reduced decisional conflict relating to feeling uninformed, and reduced the proportion of people who were unclear about their personal values. PDAs stimulated people to take a more active role in decision making and improved congruence between patient’s values and the option chosen. PDAs had a more variable effect on the consultation time and the choice of intervention, and did not appear to have any adverse effect on either health outcomes or satisfaction. However, there was insufficient evidence to determine the effects of PDAs on patient-practitioner communication, adherence with the chosen option or the costs and resource use.
WORKSHOP: The main messages emerging from the workshop suggested the following concepts were worthy of consideration: the ‘quality’ of the treatment decision; both health and non-health benefits of PDAs; the potential inability of a single generic measure capturing all benefits; process outcomes and non-tangible effects (increase in dignity, or increase in anxiety); the similarity of SDM and a basic standard of care; conflicts with the QALY maximisation model in terms of individual’s preferences; and finally, the possible reduction in efficiency and potential trading of non-health benefits and QALYs to utilise the current standard framework.
CONCEPTUAL FRAMEWORK: PDAs may impact on processes, outcomes and costs. While the reviews provided clear evidence on improvements of patients’ knowledge of the outcomes of alternative interventions, and providers’ understanding of the preferences and values of patients, the evidence was more mixed for patient satisfaction, health outcomes, resource use and cost impact. The literature suggested little or no health benefits from PDAs with the main benefits likely to arise from non-health effects such as reduced decisional conflict and satisfaction with the decision making process. These require quantifying in terms of equivalent lost benefits from displaced activities in the NHS due to any additional costs imposed by PDAs. In addition to non-health effects, the current QALY model makes assumptions about people’s preferences for health over time and uncertainty, and health states are usually valued using general population valuation preferences rather than patients. Any deviations in patient preferences from these assumptions may result in patients making choices that are not considered cost-effective under the QALY framework.
CONCLUSION/SUMMATION: The implications for economic evaluations of PDAs within SDM is that the framework needs to be extended beyond health to better incorporate what matters to patients, but this raises important normative concerns and conflicts with the current aim of cost-effectiveness analysis to maximise health measured through the QALY. We have provided a framework for extending economic evaluation and the types of data to be collected, but further research is required in order develop methods for putting it into practice.
