Seeking the state of the art in standardized measurement of health care resource use and costs in juvenile idiopathic arthritis: a scoping review

Juvenile idiopathic arthritis (JIA) is the most common chronic rheumatologic disease in childhood [1]. Children with JIA suffer from joint inflammation, stiffness, contractures, and pain, which can lead to fatigue, growth abnormalities and functional impairment [2‐5]. Due to its chronic nature, children diagnosed with JIA are at higher risk of developing behaviour and/or psychiatric disorders (e.g. social isolation, depression, anxiety) [6, 7], they miss more school or work days due to symptoms or medical appointments [8, 9], and have lower Health-Related Quality of Life (HRQoL) [6] than their peers. In addition, there is an associated risk of developing extra-articular diseases such as uveitis, which may lead to blindness if left undiagnosed or untreated [2]. Treatment of JIA consists of a combination of pharmacological, physical and occupational therapy, and psychological support [3]. In patients with active disease or with inadequate response to conventional pharmacological treatment (e.g. MTX, NSAIDs and DMARDs), the use of a (much more expensive) biologic DMARD is recommended [4, 10, 11].

Besides the direct negative health impact of JIA, it may result in lifelong functional limitations [12], lower educational attainment [13], higher unemployment rates [13] and lower HRQoL [14, 15]. In addition, JIA does not only affect the patient, but it also leads to significant out-of-pocket costs and to productivity losses for parents or caregivers (i.e. absenteeism and presenteeism). Consequently, JIA is associated with considerable financial burden to society.

In order to analyse the cost impact of JIA, it is necessary to identify, measure and value resource use associated with the treatment or management strategies. These costs can be measured in a full economic evaluation, in which both costs and outcomes of alternative treatment or management strategies are analysed. Costs can also be measured in partial economic evaluations such as a cost description or a cost-outcome description [16]. Identifying the perspective of the analysis (ranging from health care system to full societal), determines what categories of costs are included in these analyses. As illustrated above, the impact of JIA (in terms of costs and health outcomes) cannot be fully captured by only considering the health care related costs of JIA, such as physician visits, the use of medication, or hospitalizations. Instead, it should also capture the costs of JIA borne outside of the health care sector, including productivity losses (not only for patients, but also for parents, caregivers or siblings), as well as out-of-pockets costs, such as costs for home and/or car alterations, extra school costs, transportation costs, etc. Therefore, when performing an economic evaluation of JIA, the use of a societal perspective is recommended [17, 18]. Although general guidance on conducting and reporting economic evaluations is available [19‐21], these do not include a standardized list of resource use items or any standardized resource use data collection instrument to include in such evaluations. There have been efforts to standardize these resource use items but to date this is limited to adult conditions in the UK from a payer perspective [22]. In addition, two recent systematic reviews did not identify any validated standardized instruments for collecting health care resource use in children [23, 24]. Although a number of (non-validated) instruments for different childhood diseases (although not including JIA) can be found in an online database of such instruments, there is considerable variation in the resource use items they include [25]. We therefore concluded there is no standardized guidance regarding what types of resource use items to include in an economic evaluation of interventions for childhood diseases generally, nor for JIA specifically.

This study therefore aims to describe current practice on identification and measurement of health care resource use in the health economic literature on JIA. It is hypothesized that the lack of available guidance will be reflected in literature. A second objective is to identify commonly used sources to collect health care resource use and unit costs.¹ The findings from this study will be taken into account in a multicenter, international collaborative project into management strategies for JIA, conducted in the Netherlands and Canada, named UCAN CAN-DU.² The results of the current study will, ultimately, facilitate the identification of a standard set of core resource use items to improve the quality and comparability of health economic evaluations of JIA interventions.

A scoping review was conducted to identify resource use items and unit costs included in economic evaluations or costing/resource utilization studies in JIA [26, 27]. The search was performed in PubMed and Embase in July 2018. JIA specific disease terms were combined with the search filters for economic evaluations as recommended by the Canadian Agency for Drugs and Technologies in Health (CADTH) [28] (Additional file 1). Studies were included if they met the following criteria: peer reviewed articles presenting primary research; the study type included a full or partial economic evaluation which included costs, or was an analysis of resource utilization; the study population focused on JIA patients and/or the potential consequences of JIA into adulthood; English language. As the aim of this scoping review was to capture current practice including all economic evaluation or costing/resource utilization studies that have been performed in JIA, a broadly defined search strategy was used (e.g. also including the abbreviations ‘JA’, ‘JRA’ and ‘JCA’). Unlike with a systematic review, a quality assessment was not part of the inclusion criteria for this study. Duplicates were removed electronically, and then titles and abstracts were screened by two reviewers (MMAK and GC). Related published literature reviews were excluded, but the reference lists were manually searched for relevant studies [11, 29‐32]. Full text review was conducted (by MMAK and GC), and disagreements resolved by consensus. A third reviewer (LGL) was available if consensus could not be reached.

All included full-text articles were analysed (by MMAK) to identify which resource use items were measured, or which items were mentioned as (potentially) relevant (e.g. in the discussion section of the manuscript) but not measured. Subsequently, for each study, the data sources used to obtain resource use and unit costs were identified.

Medication use, outpatient and inpatient hospital visits, medical professional, other medical visits, and laboratory tests were found to be the most commonly reported health care related resource use items. Besides these, when considering resource use items related to costs of lost productivity and costs borne by families, productivity losses among parents and/or caregivers were most frequently reported. However, the extent and way these are incorporated differs considerably between studies. Other resource use items to quantify impacts outside the health care system, including social care services, school costs, school days lost and productivity losses among patients, as well as transportation costs and (other) out-of-pocket costs, were less frequently reported.

The most commonly reported resource use items (i.e. medication, medical professional and other medical visits, outpatient and inpatient hospital visits) mirror the core items identified for an adult generic health care resource use instrument [51], with the addition of laboratory testing. The addition of laboratory testing in the JIA population aligns with clinical practice, given that the medications prescribed require regular lab monitoring (both in adults and children). In addition, the relatively common inclusion of productivity losses of caregivers is reflective of this unique consideration in the evaluation of childhood diseases.

When considering the 20 articles included, there is strong variation in the perspectives (or combination of perspectives) that were applied. In addition, the great majority of these studies used a one-year time horizon, although this is insufficient to capture the full economic implications of childhood diseases like JIA. The choice for these different perspectives and time horizons may however be partly explained by the different aims of the articles, ranging from (specifically) investigating out-of-pocket costs and the use of complementary or alternative health care [35, 37, 46], to investigating a specific intervention or type of medication [38, 44, 49]. Therefore, the high heterogeneity between the articles in this scoping review limits their comparability. In addition, as studies are performed in different countries, with different health care systems, the resource use and cost items covered may differ between these health care systems, thereby further limiting the comparability between the studies in this scoping review. Consequently, the full health economic impact of JIA cannot be reliably quantified.

Although questionnaires or inquiry with parent(s) or the child were found to be the most frequently used data source to obtain resource use, the impact of JIA on societal costs (including lost productivity) is often not incorporated. More specifically, the impact of JIA in terms of missed school days was only incorporated in 3 articles (although none of them quantified this impact in terms of costs), making it likely that many health economic evaluations underestimated the full potential impact of JIA.

Although medical records will likely represent the most reliable source of evidence for collecting medical resource use (in contrast to questionnaires or inquiry with parents), results indicate that this is not common practice. As results from questionnaires are prone to recall bias, the accuracy of the outcomes of these studies may be limited. When considering unit costs, the results indicate that these were most frequently obtained from medical charges or reimbursement fees, or from published or available data(bases), literature, or costing guidelines. Although medical charges and reimbursement fees may not provide the actual costs incurred by the health service provided, they are however often considered the best available source of evidence.

This is the first study to provide a comprehensive overview of all resource use items included in studies evaluating JIA-related resource use and/or costs. Although some resource use items identified (e.g. splints and devices) may be specific to JIA, most items will also apply to other (chronic) childhood diseases. However, as mentioned previously, no validated health care resource data collection instrument for children could be found. Therefore, the results from the current study, in terms of the importance of improved guidance regarding how to quantify resource use and unit costs, are likely also valuable to other (chronic) childhood diseases.

The results of this scoping review are limited by the quality of reporting on the included resource use items. One limitation concerns the level of detail reported for each item. For example, although medical professional visits were included in most of the articles, the type of medical professional that was consulted was not always stated explicitly. This lack of detail limits the transparency and reproducibility of the study’s results.

Another limitation concerns the accuracy in reporting data sources for quantifying resource use and obtaining unit costs. This specification is often general, for example stating that all resource use was obtained from physician inquiry and/or available literature, or stating that all medical costs were obtained from medical charges and/or published literature, without exactly specifying which data source was used.

In health economic evaluations, the use of a wider and/or societal perspective is increasingly recommended as the preferred perspective, and is required when it is likely that this wider and/or societal impact will substantially impact the results [17, 18, 52]. Thus, as JIA leads to productivity losses among both patients and parents and/or caregivers [9, 34, 36] and comes with significant costs borne by families, it can be concluded that this impact is often overlooked as only 7 out of 20 articles report the use of a societal (or social) perspective. More specifically, when considering the resource use and costs that were actually included in these 7 articles, this varies from only incorporating health care costs and productivity losses (although use of the term societal perspective should not be justifiable purely based on the inclusion of productivity losses [53]), to also incorporating out-of-pocket costs, lost school days, social costs, and/or costs of education. Thus, there is great variation in the way the societal perspective is conceptualized and interpreted in economic evaluations, which has also been reported in literature [53, 54].

Another aspect that may be overlooked in economic evaluations of JIA concerns the use of complementary and alternative health care. As the results of the current study show that this type of health care is used by a substantial proportion of JIA patients, incorporating the accompanying costs is likely relevant to capture out-of-pocket costs for families. Consequently, current economic evaluations of JIA (as well as other (chronic) childhood diseases) may underestimate the real-life impact of JIA.

In addition, the terminology used to state the perspective of the evaluation was not always clear. To illustrate this, 2 articles that reported either the use of a ‘direct medical cost perspective’ [36] or a ‘direct and indirect cost perspective’ [39] both reported (besides medical costs) working time lost from parents or caregivers and/or family borne costs. Consequently, the quality and comparability of studies evaluating resource use and costs in JIA would benefit from improved guidance regarding: 1) which perspective can and should be applied, 2) which cost and resource use items should be included, and 3) how these items can be quantified. Such improved guidance will likely increase the quality and comparability between these studies [51], both for JIA as well as for other chronic childhood diseases.

Although there is some consistency in terms of items that are included in studies examining resource use and unit costs of JIA from the health care system perspective, there is heterogeneity in what is included within perspectives going beyond that. However, as the impact of productivity losses and family borne costs among JIA patients (and among other chronic childhood diseases) pose a substantial burden to society, these studies should go beyond the health care system perspective to capture this full societal impact. Therefore, improved guidance for conducting and reporting (health) economic evaluations of JIA is required. This can be achieved by the development of a standardized list of items for collecting resource use and unit costs in chronic childhood diseases. In particular it should be emphasized how to quantify this societal impact of these childhood diseases over a prolonged time-period. Incorporating these recommendations will likely increase the quality and comparability of health economic evaluations of both JIA and other chronic childhood diseases.