Using Proportional and Absolute Shortfall in Reimbursement Decision-Making in the Netherlands: Implications for Oncology Drugs for Older Patients Authors: Vivian Reckers-Droog, Steef Redeker, Jan Busschbach, Sylvia Vijgen, Angèl Link, Lonneke Timmers Published: 2026 ABSTRACT AbstractIn the Netherlands, proportional shortfall is the primary metric for threshold-setting in cost-effectiveness analyses, while absolute shortfall is also reported and considered during appraisal. Although conceptually and mathematically related, the two metrics can yield discordant results. Their theoretical differences are well recognised; nevertheless, debate continues regarding which metric should be used for operationalising disease severity in health technology assessment (HTA). This study describes the rationale behind the combined use of proportional and absolute shortfall by the National Health Care Institute (ZIN) and provides empirical evidence on how the two severity metrics relate in real-world reimbursement practice. We analysed proportional and absolute shortfall levels reported in 56 of all 57 appraised reimbursement dossiers published by ZIN between May 2018 and April 2026. While the two metrics are systematically related, we observed a distinct cluster of 28 dossiers showed high proportional shortfall (mean 0.88) but relatively low absolute shortfall (mean 14.28 QALYs). Of these, 27 concerned oncology drugs for older patients, which were significantly more likely to fall into this cluster than non-oncology drugs. These findings demonstrate that differences between proportional and absolute shortfall, previously discussed largely at a theoretical level, translate into observable patterns in reimbursement practice. By making explicit how the two metrics profile technologies differently, this study contributes empirical evidence to ongoing policy discussions about the operationalisation of disease severity in HTA. More specifically, the analysis clarifies how ZIN’s current practice of reporting both metrics may support more transparent, better-informed and consistent use of disease severity in reimbursement decision-making. Key Points for Decision Makers Proportional and absolute shortfall quantify disease severity differently and can prioritise different patient groups in reimbursement decisions. Analysis of Dutch reimbursement dossiers shows that oncology drugs for conditions affecting older patients often have high proportional shortfall but relatively low absolute shortfall.Reporting both metrics can help decision makers understand these differences and improve transparency in how disease severity informs reimbursement decisions. Introduction Disease severity plays an important role in determining whether a new health technology is eligible for reimbursement in a growing number of countries. In the Netherlands, reimbursement decisions are informed, among other criteria, by the cost-effectiveness of a technology. Disease severity is incorporated into the decision-making framework through the application of higher cost-effectiveness thresholds (referred to as maximum reference values in the Netherlands), which are used to evaluate the incremental cost-effectiveness ratio (ICER) of technologies indicated for more severely ill patients [1]. Table 1 presents an overview of the severity categories and corresponding maximum reference values per quality-adjusted life year (QALY) gained. The upper bound of €80,000 per QALY is informed by earlier policy advice on acceptable cost-effectiveness thresholds [2], while the specific category cut-offs were selected pragmatically to distinguish between broad severity levels for the purpose of linking them to differentiated cost-effectiveness thresholds in a way that is both policy-relevant and operational in practice [1, 3, 4]. Table 1Maximum reference values (in €) per QALY gained [1, 3]Proportional shortfallMaximum reference value per QALY gained (in €)0.10–0.4020,0000.41–0.7050,0000.71–1.0080,000No maximum reference value per QALY gained is specified for proportional shortfall 0.00–0.09. Health technologies indicated for conditions within this range are, in principle, not reimbursed, and the category is therefore not reported in policy documents [1, 2]. Severity-related considerations may still play a role in individual cases (e.g. episodic conditions such as depression) [3] Maximum reference values (in €) per QALY gained [1, 3] No maximum reference value per QALY gained is specified for proportional shortfall 0.00–0.09. Health technologies indicated for conditions within this range are, in principle, not reimbursed, and the category is therefore not reported in policy documents [1, 2]. Severity-related considerations may still play a role in individual cases (e.g. episodic conditions such as depression) [3] While the conceptual model behind this approach was introduced as early as 2001, it was not until 2018 that the operationalisation of disease severity in terms of proportional shortfall for threshold-setting in the assessment stage of health technology assessment (HTA), together with the consideration of absolute shortfall in the appraisal stage, was formalised by the National Health Care Institute (Zorginstituut Nederland, ZIN) [3]. In parallel with developments in the Netherlands, several other countries have adopted similar approaches. Norway uses absolute shortfall to adjust the cost-effectiveness threshold, applying greater weights to technologies targeting more severely ill patients [5]. In Lithuania, proportional shortfall is similarly used to adjust the threshold [6, 7]. In England and Wales, both severity metrics can be used to modify the valuation of QALY gains in ICER estimations [8]. There are important differences between countries, such as the size of the weight or modifier assigned to severity, the number of severity categories used and whether shortfall is discounted. Although the theoretical differences between proportional and absolute shortfall are well recognised [3, 5, 8], key questions remain regarding which metric should be used to operationalise disease severity and what the implications are for reimbursement decision-making [3–5, 8–10]. While previous analyses focused on how severity weights based on proportional and absolute shortfall are implemented and distributed across severity categories in decision-making [8], less attention has been paid to the relationship between these metrics at the level of individual cases and to how potential discordance between them may affect decision outcomes. To address this gap, this paper describes the rationale behind ZIN’s joint reporting of proportional and absolute shortfall in HTA and examines how these two metrics relate to each other in real-world reimbursement practice [8]. The analysis highlights how proportional and absolute shortfall capture different aspects of disease severity, each of potential relevance for equity-oriented decision-making, and identifies areas where a more balanced or integrated use of disease severity could provide a fuller understanding of disease severity in reimbursement decisions. Proportional and Absolute Shortfall Both proportional and absolute shortfall rely on estimates of prospective health outcomes, typically expressed in terms of remaining quality-adjusted life-years (QALYs). In what follows, we introduce proportional shortfall first to reflect its primary role in the assessment stage of HTA in the Netherlands, followed by absolute shortfall. Proportional shortfall is defined as the average loss of remaining QALYs due to a disease (i.e. their absolute shortfall), expressed as a proportion of the average total remaining QALYs patients would be expected to have in the absence of the disease [6]. In decision-making practice, it is used to normalize patients’ QALY loss by comparing it with the remaining QALY expectancy of healthy individuals in the population (typically stratified by age and, depending on the jurisdiction, also by sex) [2]: Absolute shortfall is defined as the difference between patients’ expected remaining QALYs in the absence of the disease and their expected remaining QALYs with the disease [6]. In decision-making practice, this is operationalised by comparing the average expected QALYs of healthy individuals in the population (of the same age and/or sex) with the average expected QALYs of patients with the disease [2]: Currently, in the Netherlands, proportional and absolute shortfall are calculated accounting for patients’ remaining QALYs while having access to standard care [3]. As such, in decision-making practice, these metrics reflect how much of patients’ remaining QALY expectancy is lost owing to the disease if the new technology is not reimbursed from public funding (i.e. if only standard care is provided). Rationale for Using Proportional Shortfall for Threshold-Setting One of the main arguments put forward by ZIN for using proportional rather than absolute shortfall for threshold-setting was to reduce the risk of age-related bias in reimbursement decisions informed by cost-effectiveness analysis (CEA) results [3]. More specifically, the aim was to avoid a “double disadvantage” for older patients, as they tend to gain fewer QALY from treatment and, therefore, have less favourable ICERs in CEAs [3]. Age is generally not considered a morally relevant patient characteristic for priority setting in healthcare and, in principle, the use of proportional shortfall aims to treat proportional health losses similarly across age groups [9, 11]. By expressing QALY loss relative to what is still possible given patients’ age, proportional shortfall results in smaller differences across age groups than absolute shortfall. This helps to avoid that older patients are systematically deprioritised in reimbursement decisions, as might be the case using absolute shortfall [9]. In fact, older patients may even be implicitly prioritised under proportional shortfall, since they are, ceteris paribus, more likely than younger patients to lose a larger proportion of their remaining QALYs [4]. Note that, in this context, prioritisation refers to the level of the applicable cost-effectiveness threshold against which a technology is evaluated, influencing the conditions under which it may be considered acceptable for reimbursement (e.g. required price reductions), rather than directly determining patient access. This feature of proportional shortfall may also be seen as a disadvantage. Although it was introduced to minimise age-related bias, it may in turn disadvantage younger patients. In addition, while empirical evidence suggests that members of the public may generally prefer prioritising more over less severely ill patients, it also suggests that they may prefer prioritising younger over older patients, even when the latter are more severely ill [12, 13]. This implies that absolute shortfall may align more closely with societal preferences than proportional shortfall [7]. This interpretation should, however, be considered in light of how evidence on cost-effectiveness and disease severity are used in practice. A counterargument in favour of proportional shortfall is that it may counterbalance the fact that ICERs tend to be more favourable for younger patients, who have a greater potential to gain QALYs. In this sense, proportional shortfall may help balance age-based prioritisation in CEA-based decision-making. While absolute shortfall may align more closely with public preferences for prioritising younger patients, such preferences are seldom elicited in the practical context of CEAs, where technologies are evaluated in isolation rather than through explicit comparisons across patient groups. It is therefore not evident that absolute shortfall would better reflect societal preferences in actual decision-making. Moreover, applying absolute shortfall consistently in both severity weighting and cost-effectiveness estimations could reinforce the advantage younger patients already have in terms of QALY gains. This raises concerns about fairness and distributive justice, particularly regarding access to new health technologies for severely ill older patients. In recognition of the limitations of proportional shortfall and societal preferences for prioritising younger patients, ZIN requests information on both proportional and absolute shortfall, including all underlying components of the calculations, since 2018 [3]. While absolute shortfall is not used to determine the maximum reference value against which the ICER of a health technology is evaluated during the assessment stage of HTA, it can still be taken into account by the Insured Package Advisory Committee (Adviescommissie Pakket, ACP) during appraisal [8]. Relationship between Proportional and Absolute Shortfall in Reimbursement Practice To understand the potential implications of using proportional rather than absolute shortfall for threshold-setting, we analysed all 57 reimbursement dossiers appraised by the ACP between May 2018 and April 2026. The dossiers all concerned requests for reimbursement of pharmaceuticals, including 37 oncology drugs (64.9%) and 20 non-oncology drugs (35.1%). A total of 17 pharmaceuticals had an orphan designation, 8 of which were indicated for oncological conditions and 9 for non-oncological conditions. Table 2 provides an overview of the dossiers, including oncological indication (yes/no), orphan designation (yes/no), proportional and absolute shortfall levels, the reimbursement recommendation and whether absolute shortfall was explicitly considered by the ACP in its recommendation as reported in the minutes of their public meetings. Table 2Overview of reimbursement dossiers (N = 57)IDSubstance nameOncological conditionOrphan designationaProportional shortfallAbsolute shortfallDiscordantbReimbursement recommendationcAbsolute shortfall influentialdACP 74-5Osimertinib (case of activating EGFR-mutations)YesNo0.8815.95YesConditionalNoACP 75-5Axicabtagene ciloleucelYesNo0.83NSNAConditionalNoACP 76-5DurvalumabYesNo0.8314.69YesConditionalNoACP 77-4Venetoclax, in combination with rituximabYesNo0.769.40YesConditionalNoACP 78-4Ipilimumab in combination with nivolumabYesYes0.7611.79YesYesNoACP 79-4Dabrafenib in combination with trametinibYesNo0.4810.46YesConditionalNoACP 81-5Vortigern neparvovecNoYes0.6941.00NoConditionalNoACP 82-4OlaparibYesNo0.6516.53NoYesNoACP 83-4Dabrafenib in combination with trametinibYesNo0.499.94NoConditionalNoACP 83-5Avelumab in combination with axitinibYesNo0.8716.65YesConditionalNoACP 84-4IbrutinibYesYes0.5612.00NoConditionalNoACP 85-4Tezacaftor/ivacaftor in combination with ivacaftorNoNo0.7531.07NoConditionalNoACP 86-6Trastuzumab-emtansineYesNo0.3410.01NoConditionalNoACP 87-4Venetoclax in combination with obinutuzumabYesNo0.416.02NoConditionalNoACP 88-4Polatuzumab vedotionYesYes0.9111.82YesConditionalNoACP 90-4GivosiranNoYes0.6021.31YesConditionalNoACP 91-4Elexacaftor/tezacaftor/ivacaftor in combination with ivacaftor monopreparaatNoNo0.6934.47NoNoNoACP 91-4Elexacaftor/tezacaftor/ivacaftor in combination with ivacaftor monopreparaatNoNo0.8140.06NoNoNoACP 93-4NiraparibYesNo0.7815.18YesConditionalNoACP 94-4Betibeglogene autotemcelNoYes0.5327.8NoConditionalNoACP 94-5AvelumabYesNo0.8511.91YesConditionalNoACP 95-4TafamidisNoNo0.777.70YesConditionalNoACP 97-4OsimertinibYesNo0.549.79NoConditionalNoACP 97-5VenetoclaxYesNo0.918.39YesConditionalNoACP 98-6Ipilimumab in combination with nivolumabYesNo0.9513.29YesConditionalNoACP 99-4TucatinibYesNo0.9524.4NoConditionalNoACP 102-4RisdiplamNoYes0.9767.51NoConditionalNoACP 102-5BempedoïnezuurNoNo0.375.84NoConditionalNoACP 102-6CannabidiolNoNo0.9064.09NoConditionalNoACP 103-4Sacituzumab govitecanYesNo0.9825.07YesConditionalNoACP 104-4PegcetacoplanNoYes0.195.54NoConditionalNoACP 104-5Icosapent ethylNoNo0.386.57NoConditionalNoACP 104-6Atidarsagen autotemcelNoYes0.9970.00NoConditionalYesACP 105-4Trastuzumab-deruxtecanYesNo0.8719.6YesConditionalNoACP 107-4PCSK9-remmers evolocumabNoNo0.163.47NoConditionalNoACP 107-5Brexucabtagene autoleucelYesNo0.8815.46YesConditionalNoACP 107-6AvacopanNoNo0.335.88NoNoNoACP 108-4AvacopanNoNo0.335.88NoConditionalNoACP 109-4LenvatinibYesNo0.9414.18YesConditionalNoACP 114-4SelumetinibYesYes0.5435.15NoConditionalNoACP 116-4Axicabtagene ciloleucelYesNo0.528.71NoConditionalNoACP 117-4Olipudase alfaNoYes0.7440.27NoConditionalNoACP 118-4TebentafuspYesYes0.9317.46YesConditionalNoACP 120-4Ciltacabtagene autoleucelYesYes0.9317.44YesConditionalNoACP 121-6AvapritinibYesYes0.9012.14YesConditionalNoACP 121-7Lutetium (177Lu)-vipivotide tetraxetanYesNo0.9210.40YesConditionalNoACP 121-8TalazoparibYesNo0.725.87YesConditionalNoACP 122-4TeclistamabYesNo0.9215.91YesConditionalNoACP 123-4Trastuzumab–deruxtecanYesNo0.9214.89YesConditionalNoACP 124-4OmaveloxolonNoYes0.7541.70NoConditionalNoACP 124-5Exagamglogene autotemcel (exa-cel)NoYes0.6631.69NoConditionalNoACP 125-4Enfortumab vedotin in combination with pembrolizumabYesNo0.8511.55YesConditionalNoACP 126-4RipretinibYesYes0.9920.00YesConditionalNoACP 127-4Glofitamab in combination with gemcitabine and oxaliplatineYesNo0.8711.34YesConditionalNoACP 128-4Exagamglogene autotemcelNoNo0.8039.73YesConditionalNoACP 130-4Serplulimab in combination with chemotherapyYesNo0.9113.00YesConditionalNoACP 130-5Belantamab mafodotin in combination with bortezomib and dexamethasonYesNo0.7112.71YesConditionalNoACP Adviescommissie Pakket (Insured Package Advisory Committee)aOrphan designation in accordance with the criteria laid down by the European Medicines Agency (yes/no)bHealth technology is indicated as discordant (yes) in case proportional shortfall is 0.71–1.0 and absolute shortfall is ≤ 26 QALYs, with cut-off values determined from Fig. 1cConditional recommendations refer to recommendations to not reimburse a health technology unless price negotiations are successful in bringing an incremental cost-effectiveness ratio (ICER) below the relevant maximum reference value; dabsolute shortfall explicitly considered and of influence on the ACP’s reimbursement recommendation (yes/no) Overview of reimbursement dossiers (N = 57) ACP Adviescommissie Pakket (Insured Package Advisory Committee) aOrphan designation in accordance with the criteria laid down by the European Medicines Agency (yes/no) bHealth technology is indicated as discordant (yes) in case proportional shortfall is 0.71–1.0 and absolute shortfall is ≤ 26 QALYs, with cut-off values determined from Fig. 1 cConditional recommendations refer to recommendations to not reimburse a health technology unless price negotiations are successful in bringing an incremental cost-effectiveness ratio (ICER) below the relevant maximum reference value; dabsolute shortfall explicitly considered and of influence on the ACP’s reimbursement recommendation (yes/no) Absolute shortfall was reported alongside proportional shortfall in 56 of the 57 dossiers. In one dossier (ID 104-6), absolute shortfall was explicitly considered by the ACP and of influence on its reimbursement recommendation. Although the recommendation was to reimburse the pharmaceutical conditional on successful price negotiations, the ACP deemed an ICER above the maximum reference value of €80,000 (given that the proportional shortfall was 0.99) acceptable owing to the patients’ high absolute shortfall of 70 QALYs. The ACP did not specify the extent to which the ICER should be reduced for the price of the pharmaceutical to be considered societally acceptable. Figure 1 shows the relationship between patients’ proportional and absolute shortfall based on the 56 reimbursement dossiers, where each point represents a single dossier. As illustrated, there is a generally positive association between the two metrics. Relationship between absolute and proportional shortfall in 56 reimbursement dossiers. Dossiers with high proportional shortfall and relatively low absolute shortfall (n = 28) are indicated by solid black circles, and other dossiers are indicated by open circles. Vertical dashed lines indicate the cut-off values for proportional shortfall categories used in the Netherlands As defined in Sect. 2, absolute shortfall equals proportional shortfall multiplied by expected remaining QALYs in the healthy population. Because expected remaining QALYs vary across dossiers, primarily reflecting differences in patients’ age and remaining life expectancy, the relationship does not follow a single linear pattern with a constant slope. Despite the systematic relationship, proportional and absolute shortfall do not always align in how they characterise disease severity in practice. Among dossiers with a proportional shortfall between 0.5 and 1.0, absolute shortfall varies considerably, ranging from around 10 to more than 50 QALYs. This shows that the choice of metric can have practical consequences for how technologies are positioned within the severity framework and, ultimately, for how they are assessed in reimbursement decision-making. A cluster of 28 dossiers (indicated by black solid circles in Fig. 1) highlights discordance between the two metrics. This is illustrated by the dossiers all reporting a high proportional shortfall, with a mean (standard deviation, SD) of 0.88 (0.07), placing them in the highest severity category under the decision-making framework in the Netherlands, but a relatively low absolute shortfall, with a mean (SD) of 14.28 (4.57) QALYs. This pattern reflects cases in which patients have relatively limited remaining QALY expectancy, resulting in high proportional shortfall but relatively low absolute shortfall. Of the 28 dossiers, 1 concerned the treatment of a cardiovascular condition, and 27 concerned oncology treatments (73.0% of all appraised oncology dossiers), indicating that oncology drugs were more likely to fall into the cluster of high proportional shortfall and relatively low absolute shortfall than non-oncology drugs. Implications of Discordance between Proportional and Absolute Shortfall The observed discordance between proportional and absolute shortfall in several reimbursement dossiers highlights that these two metrics operationalise disease severity in different ways. Proportional shortfall emphasizes the proportion of remaining QALYs lost owing to illness, aligning more closely with notions of fairness based on prospective health [10]. In contrast, absolute shortfall reflects the total number of QALYs lost and therefore tends to give greater weight to younger patients with more life to lose. As such, it aligns more closely—though not entirely—with notions of fairness based on the fair innings principle [10, 14]. This distinction implies that proportional shortfall may, in several cases, lead to more favourable assessments for health technologies targeting older populations who lose a high proportion but not a high absolute number of QALYs. This is consistent with the key reasons ZIN uses proportional shortfall for threshold-setting, i.e. to recognize the relative impact of illness on remaining life expectancy and counterbalance the potential prioritisation of younger patients in CEAs [2]. However, the discordance between the two metrics has important implications for prioritising patients of different ages in reimbursement decision-making. This raises broader concerns about equity, alignment with societal values and the potential for implicit, yet systematic, (age-based) prioritisation depending on the severity metric used and the position in the equity debate chosen. These concerns lie at the heart of the ongoing debate about which severity metric should be used for threshold-setting or for modifying the valuation of QALY gains. Absolute shortfall is not used to define severity categories in the Netherlands. Nonetheless, in countries where it is used for this purpose, a minimum of 18 discounted QALYs is required to qualify for the highest severity category in England and Wales [8] and 20 or more QALYs in Norway [5]. On the basis of these cut-off levels and the observed relationship between proportional and absolute shortfall, shaped by differences in expected remaining QALYs, it may be reasonable to assume that the 28 technologies would also not qualify for the highest severity category in the Netherlands if absolute shortfall were used for threshold-setting. Although speculative, this could influence reimbursement recommendations. For example, technologies that currently qualify for the highest severity category under proportional shortfall might fall into a lower severity category if quantified using absolute shortfall, potentially resulting in a lower applicable threshold and the recommendation of more stringent conditions for reimbursement. Indeed, using absolute shortfall could in several cases alter prioritisation across disease areas and patient populations, with significant implications for fairness and resource allocation. If absolute instead of proportional shortfall were used for threshold-setting in the Netherlands, prioritisation would likely shift towards health technologies targeting younger populations and conditions associated with large lifetime health losses. While this shift could increase the average health of the population in terms of lifetime QALYs, it could result in the deprioritisation of many oncology drugs. The extent of this shift would depend on how absolute shortfall is categorised and weighted within the decision-making framework. Such changes could affect not only reimbursement decisions and related conditions but also perceptions of legitimacy and fairness in healthcare resource allocation. While the scenario above is hypothetical (and a similar line of reasoning may also apply to cases with moderate proportional shortfall but relatively high absolute shortfall), the finding that nearly all technologies in the cluster with high proportional shortfall but relatively low absolute shortfall concern oncology drugs for older patients has direct implications for the use of disease severity in appraisals. Under the combined use of proportional and absolute shortfall in appraisal, these technologies yield discordant indications of disease severity: appearing highly severe when quantified in terms of proportional shortfall but less so when quantified in terms of absolute shortfall. This discordance highlights the need to recognise the relative severity of the disease in older patients as well as the absolute health loss across the lifespan. An explicit appraisal and weighing of the two metrics (particularly for cases like in this cluster) is, therefore, warranted. This could help ensure that combined use of both metrics achieves its intended balance: acknowledging the moral relevance of remaining health loss while avoiding systematic disadvantage for older patients with life-threatening conditions such as cancer. Conclusions In the Netherlands, proportional shortfall is the primary metric for threshold-setting in CEAs, while absolute shortfall is also reported and considered during appraisal. Our findings show that although the two metrics are systematically related, they can yield notable discordance. In particular, reimbursement dossiers for oncology drugs for older patients often report high proportional shortfall but relatively low absolute shortfall. These findings demonstrate that differences between proportional and absolute shortfall, previously discussed largely at a theoretical level, translate into observable patterns in reimbursement practice. The joint reporting of both metrics in such cases may contribute to a more transparent, better-informed and consistent use of disease severity in reimbursement decision-making. In doing so, the moral relevance of remaining health loss can be acknowledged while remaining attentive to the risk of systematic disadvantage for older patients with life-threatening conditions such as cancer. Funding No funds, grants, or other support were received for conducting this study. Data availability The data underlying this study were derived from publicly available reimbursement dossiers published by the National Health Care Institute (Zorginstituut Nederland, ZIN) and available at www.zorginstituutnederland.nl. The extracted and analysed data are included in the manuscript. Declarations Conflict of Interest Dr. Reckers-Droog reports no conflict of interest. Dr. Redeker reports being employed by the National Health Care Institute (Zorginstituut Nederland, ZIN). Dr. Busschbach reports being member of the Insured Package Advisory Committee (Adviescommissie Pakket, ACP). Dr. Vijgen reports being employed by the National Health Care Institute (Zorginstituut Nederland, ZIN). Ms. Link reports being employed by the National Health Care Institute (Zorginstituut Nederland, ZIN). Dr. Timmers reports being employed by the National Health Care Institute (Zorginstituut Nederland, ZIN). No other conflicts of interest are reported. Ethical Approval Ethical approval was not required for this study according to the Medical Research Involving Human Subjects Act in the Netherlands. AI Use and Disclosure Statement ChatGPT (GPT-5) by OpenAI was used in a limited capacity to support this work, specifically for language refinement and grammar checking. All content, analyses, interpretations, and conclusions were conceived, written, and verified by the authors. Author Contributions Conceptualization: Vivian Reckers-Droog, Steef Redeker, Jan Busschbach, Sylvia Vijgen, Angèl Link, Lonneke Timmers; methodology: Vivian Reckers-Droog, Steef Redeker; Software: Vivian Reckers-Droog, Steef Redeker; validation: Vivian Reckers-Droog, Steef Redeker; Formal analysis: Vivian Reckers-Droog, Steef Redeker; investigation: Vivian Reckers-Droog, Steef Redeker; Resources: Steef Redeker, Angèl Link; writing – original draft preparation: Vivian Reckers-Droog; writing – review and editing: Vivian Reckers-Droog, Steef Redeker, Jan Busschbach, Sylvia Vijgen, Angèl Link, Lonneke Timmers; Visualization: Vivian Reckers-Droog; Steef Redeker; supervision: Jan Busschbach, Lonneke Timmers; project administration: Vivian Reckers-Droog; Steef Redeker. References Associated Data Data Availability Statement The data underlying this study were derived from publicly available reimbursement dossiers published by the National Health Care Institute (Zorginstituut Nederland, ZIN) and available at www.zorginstituutnederland.nl. The extracted and analysed data are included in the manuscript.