Hidden Treasures and Secret Pitfalls: Application of the Capability Approach to ParkinsonNet

Abstract

Abstract

In the Netherlands, the largest health technology assessment (HTA) program funds mainly (cost-)effectiveness studies and implementation research. The cost-effectiveness studies are usually controlled clinical trials which simultaneously collect cost data. The success of a clinical trial typically depends on the effect size for the primary outcome, such as health gains or mortality rates. A drawback is that in case of a negative primary outcome, relevant other (and perhaps more implicit) benefits might be missed. Conversely, positive trials can contain adverse outcomes that may also remain hidden. The capability approach (developed by Nobel Prize winner and philosopher Sen) is an instrument that may reveal such “hidden treasures and secret pitfalls” that lie embedded within clinical trials, beyond the more traditional outcomes. Here, we exemplify the possible merits of the capability approach using a large clinical trial (funded by the HTA program in the Netherlands) that aimed to evaluate the ParkinsonNet concept, an innovative network approach for Parkinson patients. This trial showed no effects for the primary outcome, but the ParkinsonNet concept tested in this study was nevertheless met with great enthusiasm and was rapidly implemented throughout an entire country, and meanwhile also internationally. We applied the capability approach to the ParkinsonNet concept, and this analysis yielded additional benefits within several capability domains. These findings seems to substantiate the claim that richer policy debates may ensue by applying the capability approach to clinical trial data, in addition to traditional outcomes.

INTRODUCTION

The Health Care Efficiency Program (Programma DoelmatigheidsOnderzoek) of the Netherlands Organization for Health Research and Development (ZonMw) is the largest health technology assessment (HTA) program in the Netherlands. The HTA program currently has an annual budget of about 10 million euros, which is used to finance (cost-)effectiveness studies and implementation research [1]. The cost-effectiveness studies are usually controlled clinical trials, which simultaneously collect cost data.

The success of clinical trials currently depends on how the trial scores on its primary outcomes. These outcomes typically focus on health gains, e.g. improvements in the UPDRS motor score in new treatment trials for patients with Parkinson’s disease. Whilst such health gains are obviously important, additional benefits and costs – some with potentially great societal relevance – may remain hidden, in particular for complex interventions that aim to enhance quality of life. For example, many trials do not address more elusive benefits such as feelings of self-containment or empowerment among trial participants. Indeed, it has been suggested that complex interventions and interlinked packages of care challenge the application of current methods of economic evaluation that focus on health gains [2]. Typically, missing ‘hidden treasures’ can produce type II mistakes (inappropriately denying funding for a seemingly ineffective treatment). Consequently, public funds may be misallocated and opportunities to create societal gains are missed. Such misallocation is not just hypothetical.In many countries, access to new medicines or technology is often decided based on strict health criteria, such as costs per Quality-Adjusted Life Year (QALY). In some countries (such as France) it is even explicitly forbidden to allocate health budgets based on anything other than health outcomes. The allocation of scarce public resources is more important today than ever before, in light of the economic crisis, shrinking government budgets and rising healthcarecosts.

CAPABILITY APPROACH

Payne has suggested to define a sub-set of services that one can evaluate with a modified framework, using different evaluation metrics [2]. Which criteria should be met for such a metric? The methodology should be rooted in the literature, be non-controversial (because it needs to be widely adopted) and the metric needs to be able to pick up the benefits that clinical trials hide. Here, we want to discuss a metric that meets those criteria. Specifically, one option to boost the yield of clinical trials is the so-called ‘capability approach’ which was developed by Nobel prize winner Amartya Sen [3]. The capability approach emphasizes the importance of ‘functional capabilities’. Capabilities are defined in terms of the substantive freedoms people have reason to value, instead of utility or access to resources (Panel 1). The emphasis is not only on how human beings actually function, but on their capability to function. This approach to human well-being stresses the importance of freedom of choice, individual heterogeneity and multi-dimensional nature of welfare or wellbeing. In practical terms, nine functional capabilities can be distinguished (Table 1) [4].

PANEL 1. DEFINITION OF CAPABILITIES

Capabilities are possibilities to do things; they lend you the power to reach a wanted situation, to use acquired commodities in ways you like to do because they are meaningful to you, or to become the person you want to be. For example, an effective network allows one to communicate with colleagues or friends. Functionings are capabilities which a person actually uses or in which he or she participates, such as working, being educated or healthy. The distinction between functionings and capabilities is between what is realized and what would could possibly be to realized in the future. In this paper we use a simplified version of the original theory in order to facilitate the application.

The capability approach offers several added valuesto clinical trials. First, by providing a complete list of wellbeing capabilities, it functions as a sanity check to researchers who might otherwise overlook certain benefits or costs of interventions, but with fewer methodological aggregation problems that foil quality of life indicators. Sen himself wanted to solve the aggregation problem by democratic processes, but this proved unpractical. This has been ‘solved’ by showing all capabilities that yield benefits in monetary terms, and by showing non-monetary benefits, such that they could contribute to decision making [4]. The capability approach potentially provides a richer evaluative space, enabling better evaluation of health interventions.

Second, the capability approach not only focuses on outcomes realized in the present time, but also incorporates capabilities that contain a value in itself and that may translate into potential future benefits. The capability approach thus provides a platform for weighing quantifiable benefits (and costs) with other benefits (and costs), creating possibilities to incorporate important gains that health(care) interventions bring to society that would otherwise be missed.

Third, the capability approach forces researchers to refect more carefully about ethical intricacies, theoretical and empirical inconsistencies and the implications of value judgments that are taken for granted. This helps to develop a more coherent discipline, firmly grounded in societies’ values [5].

The capability approach has been used scarcely in health economics, let alone in clinical medicine [6, 7]. But there are many examples of healthcare interventions where important benefits (some of them non-quantifiable) are at stake. Blood donation, organ donation, preventive programs for obesity: examples of interventions where benefits are characterized by enhanced freedom and capabilities of patients and professionals (but not necessarily by measurable health gains). Such interventions often affect certain parts of society in a disproportionate fashion, and there are often ethical considerations at stake. It is crucial that these features are displayed in a structured way, so that benefits can be properly weighed and political decision making is facilitated.

We illustrate the merits of the capability approach using a large clinical trial, which evaluated the merits of ParkinsonNet, a new healthcare approach for Parkinson patients that was funded by the HTA program in the Netherlands between 2005–2008 [8]. We show that this traditionally designed trial likely failed to take several genuine benefits into account.

THE PARKINSONNET MODEL

ParkinsonNet consists of professional regional networks which key components are: (a) commitment to work according to mono- and multidisciplinary guidelines; (b) inclusion of a limited number of motivated healthcare providers; (c) preferred referrals by physicians and patients towards ParkinsonNet experts; (d) educational programs for professionals, with a baseline course, continuous learning on the job and online- plus offline interaction among professionals within the regional networks; (e) integrated care delivery to patients by stimulating collaboration between therapists, neurologists, and patients; (f) transparency about quality of services and health outcomes; (g) patient-centered approach, including e.g. online communities, routinely collection and feedback of patient experiences; and (h) adoption of supportive information technology platforms, e.g. a healthcare search engine (www.parkinsonZorgzoeker.nl), online communities for both patients and professionals (www.ParkinsonConnect.nl), an innovation technology platform with tools to support self-management (www.ParkinsonNeXt.nl), and a monthly online television show to offer education to patients (www.ParkinsonTV.nl) [9].

CLASSICAL EVALUATIONS OF PARKINSONNET

To illustrate the need for the capability approach, it is important to choose an initiative that both scored negatively in a classical trial, but is nevertheless unambiguously successful in everyday practice. ParkinsonNet does not need ‘data drenching’ type ex-post justifications in order for it to receive credits. Two early evaluations of ParkinsonNet showed a substantial impact on the caseload of physiotherapists (specialized ParkinsonNet therapist treated three times as many patients compared to regular therapists), and also demonstrated improvements in disease knowledge and better adherence to clinical guidelines [10, 11]. However, the first main evaluation study that followed, a large RCT which included 699 patients, showed that ParkinsonNet did not improve disability levels, functional mobility and mobility-related quality of life [8]. The trial did show that ParkinsonNet was associated with substantial reductions in healthcare costs. A subsequent additional analysis of a national healthcare claims database showed that in those regions that adopted the ParkinsonNet early, a shift had occurred from institutionalized care to expert treatment within the community, that disease complications were less common (including a 50% reduction in hip fractures) and, again, that healthcare costs were much lower [9]. A third evaluation consisted of a non-randomized controlled trial that compared usual care to the combination of a regional ParkinsonNet model plus a specialized Parkinson center of excellence in that same region. The primary outcome measures, i.e. quality of life and outcomes of daily living, showed only small benefits of the integrated care approach (presumably because the expert center was not adequately connected to the surrounding ParkinsonNet) [12]. Other studies did not evaluate ParkinsonNet per se, but rather used the specialized ParkinsonNet infrastructure as “test bed” to evaluate the merits of specific allied health interventions [13]. Specifically, one large-scale RCT showed that specifically trained ParkinsonNet occupational therapists are highly effective in improving performance in daily activities, when providing individualized occupational therapy according to a practice-based guideline [14]. Another large RTC showed that personal guidance of Parkinson’s patients by ParkinsonNet physiotherapists – aiming to increase their levels of physical activity –led to improvements in outdoor activities and to greater physical fitness [15–17]. Based on this converging evidence, the Dutch health insurance companies are now offering a structural financial reimbursement that ascertains ParkinsonNet care throughout the Netherlands.

CAPABILITY APPROACH FOR PARKINSONNET

We will now highlight several examples of potentially hidden treasures within ParkinsonNet by applying the capability approach. In the first formal RCT that evaluated ParkinsonNet, there were no differences between ParkinsonNet and usual care for the primary outcomes (quality of life and activities of daily living). However, patients in ParkinsonNet clusters may have felt better equipped to achieve self-management of their disease, e.g. the ability to self-select their health team. Thus, ParkinsonNet possibly yielded subjective health benefits to participants that extended beyond mere health-related quality of life. However, this was never formally tested within the trial.

Feelings of safety were not measured in the ParkinsonNet trial, but it is clearly a genuine possible benefit. Within ParkinsonNet, patients likely felt much safer in the hands of experts who treated many other patients with the same disorder. Also, health professionals may have felt safer within ParkinsonNet clusters: because of its complexity and diffuse treatment options, Parkinson’s disease can be a scary condition for insufficiently trained professionals. Presumably, training changed the perceptions of Parkinson’s disease for ParkinsonNet professionals, making management of this disease much more gratifying, but this was never measured.

The ParkinsonNet trial only measured the direct effects of education, such as formal knowledge of evidence-based guidelines (this was significantly improved in ParkinsonNet clusters) [10, 11]. However, education likely offered extra benefits that were not measured, including the ability for future use of the received education, e.g. more freedom to change jobs. Another element could be the intrinsic effect of being higher educated. Such aspects can be quantified, allowing for a richer set of potential benefits.

The ability to be intellectually challenged is not traditionally assessed in clinical trials. However, it could be an important reason why ParkinsonNet therapists enjoy being part of a professional network, because they receive more complex patients, and because they have ample opportunity to discuss complex cases with peers within the network.

ParkinsonNet facilitates self-care and self-management, allowing for a shift from formal to informal care, e.g. by letting patients decide when and to whom they are being referred. Such shifts are much needed, because patients can actively participate in their own disease management, and also as solution to the shrinking volume in health professionals in our ageing society. The ParkinsonNet trial did not measure the merits of civil participation, but positive perceptions about its potential value – in particular for the future – could well have played a role in the rapid dissemination of the ParkinsonNet concept in The Netherlands.

Cultural elements are rarely assessed in clinical trials. However, ParkinsonNet has changed the ‘culture of care’ for Parkinson patients, from one of mono-disciplinary soloists into one of multidisciplinary, collaborative care. Such aspects could be incorporated in future trials.

The quality and impact of social networks is another missed domain. With the advent of social media, participation in networks is becoming increasingly important. ParkinsonNet offers participants various opportunities for improved social interaction, both ‘offline’ (conferences) and ‘online’ (via web-based communities), for both professionals and patients. Traditional trials do not assess the quality of social networks, but it could be important to assess this in future trials. Indeed, new tools are now becoming available to assess the quality of social networks [18, 19].

The example of ParkinsonNet shows that using the capability approach leads to richer set of possible outcomes. Additional merits were identified for both patients and professionals. A number of the potential benefits are subjective in nature or simply impossible to quantify. Yet, given the almost automatic bias inpolitical decision making and science against softer values, it would be appropriate to have a platform available in which those values find a place so they can enter the decision process. Redefining value in healthcare is also noticed in other discourses, where the capability approach may benefit from: health in itself should no longer be the ultimate goal of care, but patients should be supported and stimulated to create the ability to adapt and participate, despite limitations due to a disease [20]. Moreover, Porter emphasizes that healthcare organizations need to redesign their strategy and define what value they want to create for theircustomers [21].

The example also shows that doctors, practitioners and policy makers should be more aware that revealing trial benefits may require different kind of data and other forms of data gathering. A number of discussed hidden benefits of ParkinsonNet remain speculative in nature, in the sense that appropriate data and surveys were lacking to allow for scientific evidence of these benefits. Future trials should move beyond anecdotal evidence and properly assess this. The goal of this paper is not an ex post justification for ParkinsonNet’s success, but a plea for applying the capability approach as an alternative evaluation metric. Next, research is needed to develop methodologically sound approaches to measure these capabilities.

CONCLUSIONS

We conclude that clinical trials can contain benefits or pitfalls with potentially great societal relevance that may remain hidden when the focus is merely on traditional outcomes. Therefore, we propose to routinely include the capability approach within future randomized clinical trials, along with traditional outcomes, and to use its outcome as an additional guide when making strategic decisions about implementation of new treatments into everyday clinical practice. The capability approach emphasizes the capability to function and the freedoms people have reason to value, instead of how people actually function. As such, the capability approach has the potential to reveal many hidden benefits or pitfalls of clinical trials. Equally, there could be merit in looking back at clinical trials with negative outcomes on organizational and other health interventions where one could suspect similar types of lost benefits that showed up here for the ParkinsonNet healthcare concept.

CONFLICT OF INTEREST

MM currently is director of ParkinsonNet and BRB was, together with MM, founder of ParkinsonNet. The development of ParkinsonNet was financially supported by the Radboud university medical center, ZonMw, the National Parkinson Foundation, Stichting Robuust, Stichting Nuts Ohra, VGZ, CZ, Achmea, Menzis, and Zorgverzekeraars Nederland. BRB was supported by a research grant of the National Parkinson Foundation. MC, MJF, WA and MJN have no conflict of interest to report.