Introduction

An estimated 537 million people, of which 44.7% undiagnosed, are nowadays living with diabetes mellitus worldwide. Type 2 Diabetes Mellitus (T2DM), formerly known as non-insulin-dependent or adult-onset diabetes, accounts for 90% of the population with diabetes1. Additional estimates indicate 541 million people having impaired glucose tolerance and 319 million having impaired fasting glucose levels, otherwise defined as prediabetes1, and are as such at increased risk to progress to T2DM2.

T2DM often occur together with hypertension3, another major risk factor for cardiovascular disease. Regarding its prevalence, hypertension affects around 1.28 billion or one-third of all adults between 30 and 79 globally, with 46% of them being unaware of the condition (https://www.who.int/news-room/fact-sheets/detail/hypertension). Prevalence is slightly decreasing in high-income countries but it is still on the rise in low- and middle-income countries4. An additional quarter to half of the global adult population is presumed to have pre-hypertension, defined as high normal office systolic and diastolic BP5.

Both diabetes and hypertension impose a substantial burden on healthcare budgets with 11.5 and 10% of global health expenditures spent on diabetes1 and high BP4, respectively. It is encouraging that preventive measures targeting modifiable lifestyle risk factors could result in substantial health and economic gains1,6. Lifestyle interventions focusing on diet modifications and increased physical activity have been proven effective in reducing HbA1c-levels and BP values7, and despite some discrepant results and varying study quality, these lifestyle interventions were found to be cost-effective as well8,9,10. Lifestyle interventions are thus valid strategies but the cost-effectiveness of various programmes and their drivers still need to be better documented. Staff labour cost is such an important driver, accounting for the larger part of lifestyle intervention costs and could thus be a potential target for improving cost-effectiveness11, for instance with the use of digital health interventions.

There is increasing evidence of digital health interventions as a practical, low labour and low cost delivery mode12 that can foster clinical-effectiveness and cost-effectiveness of such lifestyle measures. The potential is great since 92% of the global population uses a mobile phone13. The clinical effectiveness of digital health in diabetes and hypertension management has been confirmed14,15,16,17,18,19,20 but limited cost-effectiveness data have been referred to as one of the major barriers for widespread implementation21. Yet, Iribarren et al.22 reported that 75% of full economic studies (i.e., a comparison of both costs and health consequences of two or more alternatives) in a broad range of conditions found cost-effective or cost-saving results for mobile health solutions. Specifically, for T2DM, previous systematic reviews focused mainly on partial economic evaluations (i.e., single programme description of both costs and health consequences, or cost description/analysis of one or more alternatives)23,24. Nevertheless, the few full economic evaluations showed favourable results for digital health intervention modes such as phone/video calls, Short Message Service (SMS), and telemonitoring23.

The current study aims to systematically review full health economic evaluations of digital health interventions targeting the prevention and treatment of T2DM and/or hypertension in adults with (pre)diabetes and/or (pre)hypertension: smartphone applications, text-messaging, and websites are the subject of investigation19,20. This systematic review extends previously published systematic reviews that were mainly based on partial economic evaluations. Although such evaluations can provide some trends in this field, it is critically important in rapidly evolving research fields, such as digital health, to synthesize and report on more comprehensive economic evaluations25. The current systematic review focuses on three specific digital technology modes and discusses the health economic evidence for each mode separately. Finally, this review puts emphasis on the methodological quality appraisal, with a focus on making informed recommendations to improve methodological quality.

Results

Study selection

The process of study identification, screening, and inclusion is displayed in the PRISMA flow diagram in Fig. 1. From 3056 studies (2503 after duplicate removal) identified through database searches, 14 studies evaluated the value for money of either website26,27,28,29,30, text-messaging27,31,32,33,34,35,36, or smartphone application interventions29,37,38,39. An overview of study characteristics, intervention details, and health economic outcomes can be found in Table 1. Supplementary note 1 shows excluded studies at full text screening with reasons. Augustovski et al.31 and Zhang et al.36 reported on the same trial but the former was a trial-based analysis while the latter was model-based to extrapolate costs and effects on the long-term.

Fig. 1
figure 1

PRISMA flow diagram.

Full size image
Table 1 Evidence table.
Full size table

Study characteristics

Included studies reflected a broad geographic distribution with one study conducted in North America26, one in Central America32, two in South America31,36, four in East Asia34,37,38,39, one in South Asia33, two in the Middle East30,35, and three in Europe27,28,29. Eight studies included people with T2DM27,29,30,32,33,35,37,39, two included people with prediabetes26,34, and four studies focused on people with hypertension28,31,36,38.

Five studies were within-trial analyses with a time horizon between 6 and 18 months and a public healthcare system perspective27,28,31,33,39, while one was a retrospective matched cohort study applying similar analytics35. The within-trial analysis of Derakshandeh-Rishehri et al.30 applied a patient perspective but this is disputable. One study used a decision tree-based model with a time horizon of 6 months and a patient perspective38. Five studies used a Markov model to estimate long-term (i.e., 10 years to lifetime) costs and effects based on clinical trial inputs, and whereby three applied a public healthcare system perspective29,32,34,36 and one a healthcare payer perspective26. Finally, there was one Markov-model study which did not directly stem from one particular implementation study (i.e., all input parameters were literature driven) and which applied a 20-year horizon37. All studies with a time horizon of more than 1 year applied discount rates for both future costs and health outcomes between 3 and 5%26,31,32,34,36,37.

Interventions

Four studies evaluated the use of smartphone applications, one in people with hypertension38 and three in people with T2DM29,37,39. Smartphone applications were used for monitoring, treatment adaptation, and communication between patients and healthcare professionals (in Tsuji et al.37, also for communication with family). The smartphone applications in Li et al.39 and Cunningham et al.29 were also used for patient education. The smartphone application in Zhang et al.38 included a health agenda (i.e., reminders for follow-up). None of the four studies on smartphone applications included non-digital intervention features.

Seven text-messaging27,31,32,33,34,35,36 and five website-based studies26,27,28,29,30 were included. One intervention combined text-messaging and websites27, while five other interventions also comprised (non-)digital health modalities such as the implementation of a case manager or teleconsultation26,28,31,32,35,36. Text-messaging was used to encourage the adoption of healthier lifestyle behaviours by participants. The length of the intervention ranged from 16 weeks to 2 years, and the frequency of text messages could be as high as daily but it was not always reported. The website-based intervention component consisted of educational web pages and social network support groups, often in addition to teleconsultation, face-to-face follow-up, and/or telemonitoring.

Interventions were compared to care as usual26,29,30,31,33,35,36,37,38 or an enhanced version of care as usual (comprising self-management training, education, and/or physician training)27,28,32,38,39.

Health outcomes

Ten studies reported on the cost per quality adjusted life year (QALY) as the primary health economic outcome27,28,29,31,32,33,34,36,37,38. Some studies reported clinical outcomes such as systolic blood pressure reduction28,31, HbAc1 reduction30,33,35, proportion of population reaching hypertension31 or glycemic control39, life years gained34, and points gained on the problem areas in diabetes control (PAID) scale27. The cost-minimisation study of Chen et al.26 reported on the return on investment.

Quality appraisal

Table 2 shows the critical appraisal of selected studies for the evaluation of their quality. More than half of the included studies did not provide sufficient detail on the comparative alternatives (i.e., what does care as usual actually mean). Nine studies did not describe important costing aspects such as how the costs were measured or the sources of cost valuation26,28,29,30,33,35,37,38,39. A rather short time horizon was applied in more than half of the studies27,28,30,31,33,35,38,39 despite a long-time horizon being recommended in evaluating cost-effectiveness of chronic diseases to capture all relevant costs and effects. Moreover, all but one study27 did not provide sufficient argumentation for choosing another perspective to the societal one. Finally, only six studies reported both probabilistic sensitivity results plus another kind of sensitivity analysis such as threshold analysis or one-way sensitivity analysis on top of the point estimate results27,28,29,31,32,36.

Table 2 Quality appraisal with the CHEC-list.
Full size table

Data synthesis

Among the studies expressing results in QALYs, the ICURs varied between dominant (i.e., less costly and better health outcomes) and €75,233/QALY, with a median of €3840/QALY (interquartile range €16,179). One study did not find a QALY difference (Fig. 2). None of the three digital health intervention modes was associated with substantially better cost-effectiveness results than the others. Four out of fourteen studies (one on text messaging, two on mainly smartphone applications, and one on website-based education) reported cost-saving results26,29,34,39.

Fig. 2: Incremental cost-utility results (ICUR) estimates of included studies.
figure 2

Note that McManus et al.28 did not calculate an ICUR as QALY difference was insignificant. ICUR estimates in Cunningham et al.29 and Wong et al.34 were dominant. CG control group. *: asterisk denotes studies targeting populations with hypertension; studies without an asterisk include people with (pre)diabetes.

Full size image

Smartphone applications were appraised by the studies’ authors as cost-effective37,38 or dominant29,39 compared to usual care. However, the cost-effective results in Tsuji et al.37 were associated with considerable uncertainty and should be confirmed by future trial data, as effectiveness data were simulated and the prediction model had been built on major assumptions. Li et al.39 did not report uncertainty analyses. Furthermore, the smartphone application in Zhang et al.38 was reported as not cost-effective compared to a self-management intervention: QALY gain was higher but at a considerable cost: a self-management strategy appeared to be the preferred strategy from a health economic perspective (Fig. 2).

Text-messaging alone, or in combination with other intervention aspects (such as teleconsultation, telemonitoring, case management), was found to be cost-effective27,31,32,33,35,36 or even cost-saving34. Although QALY gains were limited (ranging from a 0.01 increment per target person after 6 months in Islam et al.33 to a 0.22 increment per target person taking a lifetime horizon in Gilmer et al.32), the ICUR appeared to be robust in probabilistic sensitivity analysis27,31,32,36. This can be related to the low intervention costs since Islam et al.33 demonstrated that programme costs could at least be doubled while remaining cost-effective. Wong et al.34 even calculated that programme costs could be 50 times greater before the break-even point would be reached. Moreover, Li et al.27 argued that the health economic results of text-messaging can be even further improved by upscaling so that the cost per person decreases. Importantly, the ICUR in Gilmer et al.32 turned cost-effective only after 10–20 years, which was inconsistent with other studies that demonstrated cost-effectiveness in the short term27,31,33.

Website-based interventions appeared to be cost-effective27,28,30, dominant29, or cost-saving26, even though only a natural effect (i.e., a reduction in systolic blood pressure; the incremental number of QALYs was not significant) was found in the study by McManus et al.28. Yet scenario analysis, in which the intervention effect partly faded away, and probabilistic sensitivity analyses showed the results to be robust at given thresholds26,27,28,29.

Sensitivity and subgroup analyses were limited in most studies, which restricts the identification of cost-effectiveness drivers. First, Augustovski et al.31 reported on patient baseline characteristics: the intervention appeared to have greater value for money in populations of younger age, subjects with higher cardiovascular risk, higher body mass index, and women. The gender difference has been reported by Cunningham et al.29 as well.

Secondly, intervention aspects influenced the ICER/ICUR as well. A less intensive so less costly intervention following lower treatment adherence was reported by Augustovski et al.31, thus being indicative of better cost-effectiveness although the observed differences were not statistically significant. Meanwhile drop-out rates did not impact the ICER/ICUR in Wong et al.34. Costs were important drivers of cost-effectiveness in other studies as well33,38.

Third, modelling assumptions was the third and most investigated pillar of what drives cost-effectiveness results. The value for money improved with longer time horizons26,32, and the impact of transition probabilities, utility values, and discount rate on the ICUR were mixed34,37,38.

Whether digital health interventions targeting (pre)T2DM versus hypertension populations resulted in different cost-effectiveness outcomes, is difficult to assess because only three studies targeted populations with hypertension. However, it seems that digital health interventions targeting (pre)T2DM populations showed consistently positive cost-effective results26,27,29,30,32,33,34,35,37,39, while cost-effectiveness results in hypertension populations were more mixed28,31,36,38.

Whereas six studies evaluated one particular digital health mode, there were two studies that combined two of the digital health modes under investigation27,29, two studies where the digital health mode was part of a broader digital intervention including telemonitoring26,35, and four studies (three interventions) where the digital health mode was part of a broader health system intervention including digital and non-digital components28,31,32,36. Website-interventions, text messaging, and smartphone applications were complemented by, or were seen as a complement of, other intervention components in four out of five, four out of six, and one out of four times, respectively. Gilmer et al.32 and Zhang et al.36 evaluated two of the broader health system interventions and found relatively higher health effects (0.22 and 0.13 QALYs, respectively) compared to stand-alone interventions. Note that these two studies applied a long-term perspective, contrary to McManus et al.28 who evaluated a broad health system intervention and who found only a systolic blood pressure reduction on the short-term but no QALY improvement.

Discussion

This review aimed to synthesize the available health economic evidence of digital health interventions in populations with or at risk of T2DM and/or hypertension. Digital health interventions were restricted to smartphone applications, text-messaging, and website-based interventions. The three digital health intervention modes were found to be cost-effective or cost-saving compared to care as usual and, most of the time, to enhanced care as usual too. Median ICUR of cost-utility studies was low with €3840/QALY.

Recent meta-analyses from our team have shown the three digital health interventions to be equally effective in reducing BP in adults with hypertension, while text-messaging and smartphone application interventions were associated with increased improvements in glycaemic control compared to website-based interventions in adults with T2DM19,20. However, increased effects did not always offset additional costs: when comparing the three digital intervention modes with (enhanced) care as usual, our analysis did not show a strong preference in terms of cost-effectiveness for one particular mode.

Digital health interventions seem to be consistently cost-effective in populations with (pre)T2DM but not in populations with hypertension. One possible explanation could be that the cost-effectiveness of implementing a digital health mode depends on the perceived severity of a condition and hence the urge to act upon. Hypertension is so widespread that some might perceive it merely as a risk factor instead of a disease40, so patients and professionals could be less motivated to do something about it. For example, New Zealand does not have hypertension guidelines but bases its care recommendations on a cardiovascular risk score41. Moreover, a global consensus definition of hypertension is lacking (see for example the definitions of different leading organisations: https://tinyurl.com/whohyp, https://tinyurl.com/cdchyp, https://tinyurl.com/mayhyp, https://tinyurl.com/nhshypdef). Smartphone applications, websites and text-messaging may have a significant clinical impact on BP, but there are possibly other approaches or other health objectives that better justify the money invested. This remains to be tested as the health economic evidence of smartphone apps, text-messaging, and website-based interventions in populations with hypertension remains very limited.

Among other process evaluation constructs, adherence and reach are two important ones with a major impact on digital health interventions’ cost-effectiveness42. Patients who adhere with a smartphone application showed, for instance, better medication adherence43. However, high drop-out rates of 40% (95% CI 16–63%) in RCT’s testing smartphone applications have been demonstrated as well44. It has been suggested that attrition could, for instance, be reduced by using user feedback to enhance user experience, by enabling the possibility for users to contact health professionals (a so-called hybrid model), by focusing on self-management skills, by increasing health literacy, and by combining smartphone applications with internet or telehealth solutions44.

Differentiating between primarily digital health interventions and primarily health system interventions with a digital component is warranted. Our results suggest that health system interventions might have the potential to gain more health effects on the long-term compared to a stand-alone digital health mode intervention, although current evidence is limited and mixed. However, Augustovski et al.31 suggested better cost-effectiveness when the intervention was less intense. Although their statement should be interpreted cautiously because of overlapping confidence intervals between the different intervention intensities, these observations might be in line with the results of a meta-analysis on drop-out rates of exercise interventions that demonstrated a higher likelihood of drop-out in more intensive interventions45. Therefore, future interventions should carefully consider which features needs to be combined, knowing that more intervention features could improve effectiveness but a too intense intervention may also increase complexity of use thus having a detrimental impact on both drop-out and cost-effectiveness. Participant input via co-design may be of a help from the evidence gathering to the real-world testing stage46.

Given that there are hundreds of millions of people with or at risk of T2DM and hypertension, it is important to keep an eye on the scalability and budget impact of a new programme42. Whereas clinical effectiveness on the individual level can be optimised by adding possible intervention components to tailor care, less elaborated programmes may have a higher reach resulting in more population benefit within a closed budget. This could be of particular importance to digitally less-developed countries where digital interventions might be relatively more expensive. Scrutinizing the optimal intervention dose in different health systems including digitally less-developed countries is therefore paramount. Our results indicate for instance that text-messaging is appraised as cost-effective across studies, either in combination with other intervention features or not. Self-monitoring can also be a very powerful strategy to improve cost-effectiveness as well. It might therefore be an option to integrate such functionalities in smartphone applications.

Our quality appraisal demonstrated important methodological shortcomings. Based on these, our four key lessons for future health economic evaluations of digital health interventions are:

  1. 1.

    Health economic results can only be appraised correctly with an elaborate research question and sufficient context. The competing alternatives under investigation – care as usual in particular – should be detailed. The study’s perspective should be justified and the applied time horizon should capture relevant long-term costs and effects of preventive measures9,11. In this regard, it is important to stress the cost-effectiveness results of digital therapeutics despite the sometimes quite short time horizons applied.

  2. 2.

    Transparency is pivotal when reporting applied costs: which costs have been included exactly (which refers to the perspective) and how these were measured and valued should be stated.

  3. 3.

    Health economic evaluations of digital health interventions often come with data uncertainty and assumptions. One-way and probabilistic sensitivity analyses are at least needed to address these uncertainties, preferably in different subgroups. Following key lesson 1 on applying an appropriate long time horizon, it is pivotal to scrutinize the impact of the intervention effect’s sustainability on the health economic outcome, especially given the high attrition and dropout rates in for instance app-based interventions44.

  4. 4.

    An ICUR does not have an intrinsic value and should always be evaluated in light of a willingness-to-pay threshold. Most included studies applied a threshold value of one to three times the gross domestic product (GDP) per capita, as recommended by the World Health Organisation47,48, but critics argued that a more conservative threshold of ±50% the GDP per capita would better capture opportunity costs48,49. Note that, given such a conservative threshold, most cost-effectiveness estimates of digital health interventions remain cost-effective. Furthermore, some studies applied natural units (e.g., cost per percentage HbA1c reduction). For instance, Derakhshandeh-Rishehr et al.30, Faleh Al-Mutairi et al.35 and McManus et al.28 reported an increase in health effect at an increased cost and stated the result was cost-effective although no willingness-to-pay threshold or valid argumentation for the applied willingness-to-pay threshold was reported, respectively.

These key lessons should be considered in future research. Such studies should also strive to address evidence gaps in the field. Head-to-head studies are definitely needed to determine the digital health mode with the best value for money in different subgroups operating within a particular health system. The uncertainty associated with long-term health economic evaluations can be reduced by designing trials with longer clinical follow-up periods so the sustainability of the intervention effect can be modelled more precisely. Moreover, budget impact estimates are truly relevant for policy makers given the high prevalence of T2DM and hypertension, while uptake and attrition rates should also be taken into consideration as they can also have a significant effect on the costs.

The most important strengths of this review are the complementarity with previously published meta-analyses19,20 scrutinizing the effectiveness of the three digital health intervention modes, and the thorough quality appraisal resulting in several key lessons for health economic research.

However, this systematic review also has limitations. First, the adult filter is not consistent between the five searched databases. In Medline, the adult filter is >19 years of age, whereas for EMBASE and PsycINFO it is >18. However, the proportion of the population with T2DM or hypertension at that age is small50,51. Second, studies on people with or at risk of T2DM or hypertension were included but the small number of studies impeded appropriate subgroup analyses. What may work in one population may not work in another. Third, only English articles were included and this may limit our conclusions, especially since T2DM and hypertension prevalence are high in large non-English-speaking countries such as India52 and China53. However, included studies from the Americas, Europe, Asia, and the Middle-East reflected a geographically and demographically diverse population. Fourth, digital health solutions in five website-based or text-messaging studies have been augmented with other intervention features such as healthcare professional education, telemonitoring and/or (tele)consultations. It is therefore not clear whether the intervention effect arises due to these additional intervention features or due to the digital intervention component. Fifth, the number of full health economic papers remain scarce, especially compared to the accumulating amount of clinical effectiveness evidence, and the results of Tsuji et al.37 are based on disputable assumptions. Because of the low number of included papers, additional analyses of the impact of study quality on results were not conducted. Sixth, no head-to-head health economic studies of the three digital intervention modes were found. Seventh, health economic studies might be subject to multiple sources of publication bias including a publication bias in first health outcome publications and next economic publications. Funnel plotting to investigate possible publication bias was not an option in this study but Moschonis et al.19 and Siopis et al.20 demonstrated respectively a small and non-existing publication bias in our health outcome reviews. It is of course still possible that a publication bias favoring cost-effective or cost-saving results remains in economic publications, especially given the suboptimal reporting of sensitivity analyses54.

In conclusion, health economic evidence suggests that smartphone application, text-messaging, and website-based interventions are cost-effective and, in some cases, even cost-saving. It shows how challenging, but at the same time how possible, it can be to improve the health of the population while saving money.

While previous research demonstrated that the three digital health intervention modes were equally clinically-effective in adults with hypertension, and that text-messaging and smartphone application interventions worked significantly better than website-based interventions in adults with T2DM, no cost-effectiveness evidence was found supporting one particular digital health intervention mode over another. Moreover, text-messaging, smartphone application, and website-based interventions appeared to be consistently cost-effective in populations with (pre)T2DM, but not in populations with hypertension.

Based on the available evidence, policy makers and clinicians should make decisions on the most appropriate digital health interventions based on available budgets and well-defined health objectives. The high penetration rate of digital applications in diverse populations is a strength but it is pivotal to keep process evaluation constructs in mind. Key lessons for future health economic studies on how to design studies and report on the results are given. It is important to pay special focus on the context, report the costs included and how these were measured and valued, conduct sufficient sensitivity analyses, and appraise the cost-effectiveness result more critically in light of a reasoned willingness-to-pay threshold. Head-to-head studies are missing while this would enhance understanding and practice substantially. It is strongly recommended to consistently include a cost-effectiveness work package alongside clinical trials55.

Methods

Literature search

The protocol (PROSPERO CRD42021247845) and reporting of this systematic review were consistent with the 2020 PRISMA guidelines56. Five electronic databases (Medline via Ovid, Embase via embase.org, CENTRAL via cochranelibrary.com, CINAHL via EBSCO, and APA PsycInfo via Proquest) were systematically searched for scientific publications on September 2, 2022. The applied search strategy consisted of population-related and intervention-related keywords, developed by Moschonis et al.19 and Siopis et al.20, combined with a search string to detect economic evaluations, developed by Werbrouck et al.57. The latter was originally based on previously published search strings. References58,59, but was broadened to maximize sensitivity57. The search strategy is further completely consistent with Moschonis et al.19 and Siopis et al.20 the literature search was restricted by age (adults only), publication date (1 January 2009 onwards to include contemporary evidence only), and language (English), if the search engine allowed to do so. The search strategies for CENTRAL and CINAHL were further restricted to trials only and peer-reviewed manuscripts, respectively. The final search string can be found in Supplementary Methods 1. The search terms and inclusion criteria targeted a broad spectrum of studies with digital components to maximize detection rate. However, only studies with at minimum a smartphone application, text-messaging, or website-based intervention were eventually withheld. Backward and forward citation tracking were performed to identify any studies missed by the search strategy.

As this study is a systematic review, ethical approval was not applicable.

Study selection and data extraction

Titles and abstracts were screened with Rayyan60 by two independent reviewers (RW and NV) based on a priori developed eligibility criteria (Table 3). Importantly, not only head-to-head studies directly comparing the three digital health modes were included, but studies comparing the intervention including a digital health mode to usual care were included as well. Discrepancies were discussed between the two reviewers until consensus was reached. A third reviewer (LA) was available but did not have to step in as there were no discrepancies left.

Table 3 Eligibility criteria.
Full size table

Eligible full texts were screened by the first author (RW) and one-third of these full texts were screened by a second author (NV). Reference lists from articles that fitted the inclusion criteria were checked for missed articles. The following predetermined data were extracted from all included articles:

  • General study characteristics: publication year, country, participant characteristics, intervention alternatives;

  • Methods: study perspective (i.e., point of view), economic evaluation type, analytic approach, time horizon (i.e., period of analysis), discount rate (i.e., to convert a value received in the future to a value today), reference year of costs, willingness-to-pay threshold (i.e., what is society prepared to pay for health), intervention costs, health resource use and data sources, information regarding the base case and sensitivity analyses;

  • Results and conclusion: (incremental) costs and effects, results from sensitivity analyses, author’s conclusions.

Quality appraisal

As recommended by van Mastrigt et al.61, study quality has been appraised with the Consensus on Health Economic Criteria (CHEC) list62, since this checklist enables the assessment of both trial- and model-based economic evaluations25,61. The two independent reviewers (RW and NV) followed Werbrouck et al.57, who suggested small adaptions to the checklist (e.g., ‘not applicable’ was a valid answer option next to yes or no: for instance, whether or not discounting (item 14) was applied, was only considered applicable if a study’s time horizon was >1 year. Such adaptions resulted in a more valid appraisal of individual studies’ quality)57. Discrepancies were discussed by the two reviewers until consensus was reached by specifying assessment criteria.

Evidence table and analysis

The evidence table summarises study characteristics, treatment alternatives, and results from the incremental base case analyses. Sensitivity analyses are addressed in text. The following methodology applies:

  • The treatment in the comparator group has been dichotomised into care as usual (CAU) and enhanced care as usual (CAU + ). In the case of the latter, further description was provided.

  • Perspective could either be (i) the public healthcare system perspective (i.e., the third-party payer perspective), (ii) the healthcare payer perspective (i.e., including patient costs next to third-party payer costs), (iii) the societal perspective (i.e. including the payer perspective and costs from productivity losses), (iv) the patient perspective, or (v) the organisational perspective. In the case that the perspective was not explicitly stated, the authors made a judgement. A perspective could also be called ‘limited’: a limited societal perspective may for instance account for non-medical costs (i.e., costs such as transport costs to the hospital, which are costs outside the healthcare sector, but directly relatable to the disease) but not for indirect non-medical costs (i.e., productivity losses due to absenteeism or presenteeism).

  • In order to improve the comparability between studies from different countries and different reference years61, costs and incremental cost-effectiveness ratios (ICERs)/incremental cost-utility ratios (ICURs) were converted via an online calculator (https://eppi.ioe.ac.uk/costconversion/) to 2022 Euro currency values with Belgium as the reference country, to account for purchasing power parities.

Results were analysed together and per delivery mode, disease, and outcome measure. Moreover, possible cost-effectiveness drivers were explored.

Reporting summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.