A Scoping Review Mapping Economic Evaluations of Midwifery Service Provision and the Midwifery Workforce

Abstract

Background

Midwives are essential in achieving universal health coverage targets and the health targets of the Sustainable Development Goals, yet a significant global shortfall exists in the midwifery workforce. Economic evaluations of midwifery are scarce but can assist in supporting evidence-informed decision-making for sustainable and equitable health care for women and girls.

Objectives

This review aimed to systematically identify, map and report on available literature regarding economic evaluations conducted on midwifery service provision and the midwifery workforce in all settings.

Methods

A scoping review was conducted following the Joanna Briggs Institute methodology. A comprehensive search strategy was developed and run in six health databases. Peer-reviewed studies and unpublished research theses conducting economic evaluations on midwifery service provision or midwifery workforce strategies were included. Sources were limited to English-language literature published in the past 20 years. Identified sources were screened and reviewed, and data from included sources were extracted, reviewed, mapped and synthesised to report findings. Quality appraisal was conducted on all included sources using the Joanna Briggs Institute Critical Appraisal Checklist for Economic Evaluations.

Results

A total of 32 studies were included in the review. Most were from high-income countries (26/32), and very few were from low- and middle-income countries (6/32). The quality of included studies varied greatly. Under half of the studies conducted full economic evaluations (15/32), and the remainder were partial economic evaluations (17/32). Most studies evaluated midwifery service provision (29/32) through either midwife-led models of care (15/29) or by place of birth (13/29), mostly for low-risk women (23/29) from the perspective of healthcare funders. Evaluation of midwifery education programs was less common, and these were all conducted in low- and middle-income countries (3/32). Most studies concluded that midwifery service provision was cost-saving, cost-effective or cost-beneficial.

Conclusions

Our review identified a significant gap in economic evaluation of midwifery from low- and middle-income countries. However, there is ongoing need for robust, quality economic evaluations on midwifery service provision and workforce strategies in all global regions. Such studies would further support health policymakers and governments to make evidence-informed decisions to address midwifery workforce shortages and provision of evidence-based and respectful care that meets the healthcare needs of women and girls.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40258-025-00962-z.

Key Points for Decision Makers

Information on the costs of sexual, reproductive, maternal, newborn and adolescent health and midwifery care is essential to inform investment decision-making.

Economic data relating to midwifery is scarce, and many economic evaluations do not adequately reflect the complex and multifaceted health interventions provided by midwives within the full scope of midwifery practice.

There is a clear need for ongoing, robust investment into quality economic evaluations in all global regions to enable informed decisions for policy formulation and resource allocation.

Introduction

Evidence supports midwives as key providers of sexual, reproductive, maternal, newborn and adolescent health (SRMNAH) care. It is estimated that when midwives are educated to global standards, deployed appropriately, regulated and enabled to practice in enabling environments, they can provide up to 90% of SRMNAH care [1]. Midwifery interventions are essential to achieve universal health coverage targets and meet the health targets of the Sustainable Development Goals, which apply to all signatory countries [2–4]. Midwifery care saves the lives of women and newborns, reduces unnecessary interventions, contributes to positive childbirth experiences and uses resources more efficiently than other models of maternal–newborn care [5–11].

There is a critical shortfall of midwives globally. Workforce estimates report a shortfall of 1.1 million in the global SRMNAH health workforce, 900,000 of whom are midwives [1]. While the shortfall is most significant in low- and middle-income countries (LMICs), where the burden of maternal and newborn morbidity and mortality and stillbirth is highest [1], high-income countries also report midwifery shortages, particularly in rural and remote areas [12, 13]. This workforce shortfall exists owing to scarce investments in the midwifery workforce, lack of prioritisation of SRMNAH care needs of women and girls, poor recognition of the role of midwives in providing critical health services and the perception that the health workforce is a cost rather than an investment [1, 2, 14, 15]. The chronic underinvestment in the midwifery workforce continues despite high-quality evidence supporting midwifery care and despite ongoing calls for investment from peak global organisations [1–4, 16, 17].

Economic evaluations are critical to guiding decision-making in health resource allocation towards evidence-informed, sustainable and equitable investment in the health economy [18, 19]. This includes analyses conducted to identify the costs, cost-effectiveness, cost–benefits, cost-utility and return on investment for SRMNAH interventions, including for the continuum of midwifery service provision and the midwifery workforce. Such analyses build on the body of evidence that supports investment in the midwifery workforce and service provision which would lead to a higher coverage of interventions and a stronger maternal health workforce [20, 21]. While information on the costs of SRMNAH and midwifery care is essential to inform investment decision-making, this data relating to midwifery is often scarce, thereby limiting the capacity for health care policy makers to make informed decisions on resource allocation [22].

A total of four reviews related to the economic evaluation of midwifery service provision have been previously published [9, 23–25], all of which included sources of evidence from only high-income countries (HICs). One systematic review synthesised economic evaluations of midwifery service provision in the United Kingdom (UK) [24]; another investigated cost-effectiveness studies of maternity models of care (inclusive of midwives, doulas and physicians) in the United States of America (USA) and Canada, limited only to studies that used decision-tree or Markov models [25] and a third systematic review on midwifery continuity models of care included costs as a secondary outcome, highlighting a limited body of available evidence [9]. A fourth review focussed on evidence regarding midwifery service provision for women with complex pregnancies in HICs and similarly found limited and methodologically varied economic literature [23]. Each of these reviews focussed exclusively on economic evaluation of midwifery service provision, were relatively narrow in scope, and were not able to include sources of evidence regarding midwifery in LMICs.

The aim of this scoping review was to systematically identify, map and report on available literature regarding economic evaluations and costing studies conducted on midwifery service provision and the midwifery workforce in all settings in the last 20 years and to identify research gaps to inform future research priorities. Our review aimed to answer the question: “What economic evaluations and costing studies have been undertaken on midwifery service provision, education and workforce, and in which contexts have these evaluations taken place?” Secondary research objectives included:

• To identify the types of economic evaluations and costing studies conducted on midwifery service provision, education and workforce and the perspectives considered;

• To describe the costs, resource use, outcomes and benefits that have been measured and valued relating to midwifery service provision, education and workforce and

• To synthesise and describe key findings, knowledge gaps, challenges and opportunities regarding economic evaluation and costing of midwifery service provision, education and workforce.

Methods

A scoping review was identified as the most appropriate method to investigate our research question, to systematically identify the features, nature and volume of the available evidence using a broader approach to evidence synthesis [26, 27], as we anticipated a wide variety of economic evaluations and varied quality of identified literature. We conducted the scoping review in line with the Joanna Briggs Institute (JBI) methodology for scoping reviews [28] and our findings are presented as per the Preferred Reporting Items for Systematic Reviews and Meta-Analyses – Scoping Review Extension (PRISMA-ScR) [29]. We developed an a priori protocol for this review, which was registered and is accessible on the Open Science Framework; DOI: 10.17605/OSF.IO/GNFYM.

We followed JBI’s Population, Concept, Context (PCC) guide to determine the review question, objectives and sources of evidence for inclusion. Professional midwives were the principal population of focus for the review as defined by the International Confederation of Midwives [30]; key definitions relating to midwives and midwifery care are detailed in Box 1. Economic evaluation was the concept under investigation, defined in this review as any study conducted to evaluate or analyse the costs or cost-effectiveness of midwifery using any of the following methods: cost-effectiveness analysis, cost-utility analysis, cost–benefit analysis, costing analysis or return on investment. These methods are further described in Box 2. The context was global, and therefore, all countries and regions were included.

Box 1 Midwifery definitions informing this review.

• ICM definition of the midwife – “A midwife is a person who has successfully completed a midwifery education programme based on the ICM Essential Competencies for Midwifery Practice and the framework of the ICM Global Standards for Midwifery Education, recognised in the country where it is located; who has acquired the requisite qualifications to be registered and/or legally licensed to practice midwifery and use the title ‘midwife,’ and who demonstrates competency in the practice of midwifery” [30 p1].

• Midwifery scope of practice – Midwives provide skilled supportive and preventive care for all women and newborns, promote normal reproductive processes, first-line management of complications, skilled emergency care and referral to appropriate care providers. Midwives provide respectful care tailored to individual needs during pregnancy, labour and the postpartum/newborn period and provide integrated care across facility and community settings [7, 30].

• Midwifery-led models of care – “Midwifery models of care are models of care in which the main care providers for women and newborns, starting from pre-pregnancy and continuing all the way through the postnatal period, are educated, licensed, regulated midwives who autonomously provide and coordinate respectful high-quality care across their full scope of practice, using an approach that is aligned with the midwifery philosophy of care” [31 px].

Box 2 Definitions of economic evaluations for inclusion.

• Cost-effectiveness analyses (CEA) – Evaluations in which the intervention cost is related to a single clinical or natural measure of effectiveness, e.g. deaths and cases. CEAs predominantly consider health sector costs.

• Cost-utility analyses (CUA) – Evaluations in which the intervention cost is related to a multidimensional measure of effectiveness which considers not only the outcomes but the valuation of benefits, i.e. a measure of utility such as quality-adjusted life-years (QALYs) or disability-adjusted life-years (DALYs). CUAs predominantly consider health sector costs.

• Cost–benefit analyses (CBA) – Evaluations in which the intervention cost is related to a valuation of the benefits using a common or equal unit of measure, typically monetary. CBAs require the collection of all costs and benefits encompassing the perspectives of all stakeholders, e.g. health care providers, recipients of care, third party financiers or broader society.

• Costing analyses – Costing comparisons of alternative interventions or cost descriptions of a single intervention. This includes cost-minimisation analyses, cost-comparison analyses or cost descriptions of interventions or programs.

• Return on investment analyses (ROI) – A measurement of the amount of return on an intervention relative to the intervention cost. Intervention benefits are expressed in monetary terms to enable comparison against costs and reporting of the ROI as a metric.

  Adapted from: [32, 33]

Eligibility Criteria

Sources of evidence were considered for inclusion if they met the following criteria: peer-reviewed publication; focussed on the midwifery workforce including pre-service and in-service midwifery education, deployment strategies, workforce retention, or midwifery service provision; constituted a full economic evaluation or costing analysis; published between 2003 and 2024 in the English language and full-text accessible through library access or via authors.

In line with the review objectives and PCC criteria, we considered a broad range of studies investigating the costs or cost-effectiveness of the midwifery workforce or continuum of midwifery service provision. We excluded studies that evaluated specific perinatal interventions or outcomes where midwives or midwifery was not considered the population or intervention in itself. Studies were excluded if the population of interest was not clearly identified as midwives or nurse-midwives, such as skilled birth attendants, maternal health workers, obstetric nurses or lay and traditional birth attendants. Studies were also excluded if the intervention or service provided was a comprehensive service that included midwives as only one component of a package of care.

Evidence syntheses, such as systematic, scoping and literature reviews, were excluded; however, reference lists of relevant reviews were hand-searched for eligible peer-reviewed studies. Grey literature was excluded as we sought economic evaluations that had undergone a peer-review quality control process, which cannot be guaranteed in grey literature. The exception was for unpublished higher-degree research theses, which were considered for inclusion. We could not consider sources of evidence in languages other than English owing to resource limitations. Detailed inclusion and exclusion criteria are outlined in Table 1.

Table 1 .

Inclusion and exclusion criteria for sources of evidence

Inclusion criteria	Exclusion criteria
Population
Midwifery service provision – care or services provided by midwives or nurse-midwives as defined the International Confederation of Midwives (see Box 1 above) [30]	Care or service provision, education, or deployment or retention strategies of other health professionals or workers such as nurses, doctors, community health workers, traditional birth attendants or maternal health workers
Midwifery or nurse-midwifery education – pre-service or in-service	Single perinatal interventions which may be delivered by midwives or nurse-midwives
Deployment, retention or other strategies relating directly to the midwifery workforce	Comprehensive, multidisciplinary services where midwives are not the primary population/ intervention or cannot be identified as primary care providers
Concept
Economic evaluations that consider comparative analysis of costs and outcomes including cost-effectiveness analysis, cost–benefit analysis, cost-utility analysis or return on investment analysis	Financial evaluations of private businesses or insurance groups
Economic evaluations that consider only costs including cost-minimisation analysis, cost-comparison analysis and cost-description analysis
Studies that include economic evaluation or costing analysis as primary or secondary study outcomes
Context
All countries, regions and sub-regions globally
Study design
A peer-reviewed research study or unpublished research thesis	Systematic, scoping or literature reviews, other evidence syntheses, conference abstracts or grey literature
Randomised control trial, cohort study, prospective or retrospective design, case study, case series, economic or mathematical modelling or before–after comparison	Policy or research models or guides
Year of publication was 2003 or later	Year of publication was before 2003
Published in English	Published in languages other than English

Search Strategy

We undertook a three-phase approach to searching the literature in accordance with the JBI methodology [28]. The first stage was an initial limited search of two databases, MEDLINE and Web of Science, to identify key text words contained in the titles and abstracts of relevant studies. Key text words and subject headings identified from this stage were reviewed by the research team and an experienced health research librarian [28]. We then refined and further developed the search terms and subject headings in collaboration with the health research librarian, resulting in a logic grid (displayed in Table 2) which was developed into a complete search strategy that we then adapted and ran in six databases – MEDLINE, Web of Science, Embase, CINAHL, PsychINFO and NHS EED (complete search strategy in Supplementary Table S1). The NHS EED is an archive database; therefore, results were limited to studies published between 2003 and 2014. In the final search phase, we hand-searched reference lists of all sources of evidence included in the review and the reference lists of relevant existing reviews.

Table 2.

Logic grid used to inform search terms in database searching

Keywords
“midwife” or “midwives” or “midwifery” or “nurse-midwife” or “student midwife” or “student nurse-midwife” or “graduate midwife” or “graduate nurse-midwife” or “midwifery education” or “midwifery workforce”	AND	“cost” or “cost effectiveness” or “cost assessment” or “cost evaluation” or “cost analysis” or “cost model” or “cost benefit” or “cost outcome” or “cost reduction” or “cost utility” or “economic evaluation” or “economic assessment” or “economic analysis” or “economic model” or “economic benefit” or “health economics” or “return on investment” or “investment case” or “financing”

MeSH terms and Subject Headings
Midwifery/ or Nurse Midwives/ or midwife/ or nurse midwife/ or midwifery education/ or midwifery student/ or nurse midwifery education/	AND	“costs and cost analysis”/ or economic evaluation/ or economics, hospital/ or economics, medical/ or economics, nursing/ or “health care cost”/ or Health Care Rationing/ or Investments/ or “Resource Allocation”/ or “Budgets”/ or “Cost Containment”/

Source of Evidence Screening and Selection

The search was conducted in May 2024 across all six databases and the Open Access Theses and Dissertations website (https://www.oatd.org/). Search results were imported into EndNote referencing software [34], and initial duplicates were removed. Results were then uploaded to Covidence review software [35], where further duplicates were removed prior to screening, review and extraction. Three reviewers independently conducted title and abstract screening of sources (B.C., A.F. and C.H.). Two reviewers then screened and discussed the full-text sources to determine eligibility for inclusion or exclusion in the review (B.C. and V.S.); any discrepancies were discussed and resolved with a third reviewer (C.H.). Reference lists of all included sources were hand-searched by one reviewer (B.C.) to identify any further sources for consideration.

Data Extraction and Charting

Two reviewers (B.C. and V.S.) developed a data extraction form that was pilot tested on a small sample of sources to determine if all relevant data were captured; several adjustments were made before conducting data extraction for all sources. Data items extracted included: author and publication details, study location and income setting, study aim(s) and workforce focus, research methodology, type of economic evaluation, perspective and time horizon, comparator(s), participant number and risk classification (if applicable), cost source, health outcomes (if applicable), key findings and reported costs, study strengths and limitations. Data were extracted by the lead author (B.C.) and recorded in the form, which was iteratively checked and discussed with a second author (V.S.) to ensure validity and consistency. Quality appraisal was conducted on all included sources using the JBI Quality Appraisal Checklist for Economic Evaluations [36]. The JBI appraisal tool was chosen as it can be applied to any form of economic evaluation [36]; the tool was applied to studies that considered both costs and outcomes and those that only considered costs, with appraisal questions applying only to the costing component of the latter group of studies. Quality appraisal on all sources was conducted independently by two authors (B.C. and V.S.); appraisal results were then compared and discussed to reach consensus, and a third author (C.H.) was consulted to resolve any conflicts. Sources were not excluded on the basis of quality as the scope of the review was broad and intended to capture all sources regardless of quality; however, we intended to report the quality of the economic component of each source.

Results

Search Results

The final search results and the selection and screening process is outlined in Figure 1. A total of 3229 sources of evidence were identified in the search, of which 1229 were duplicates. A total of 2005 sources underwent title and abstract screening, and 1920 did not meet the inclusion criteria, which left 85 studies for full-text review. In total, 53 of these studies were excluded as they did not investigate professional midwives or nurse-midwives according to our definition (n = 22), were not economic evaluations or costing analyses (n = 12), were not peer-reviewed research or an unpublished thesis (n = 15) or were not available in English (n = 5); all abstracts that proceeded to full-text review were accessible at full-text (reasons for these exclusions are detailed in Supplementary Table S2). This left a total of 32 studies included in the final review.

Fig. 1.

PRISMA flow diagram of study selection process

Quality Appraisal of Included Studies

Results of the quality appraisal are presented in Table 3. Partial economic evaluations met fewer criteria than full economic evaluations overall; only 5 of 17 partial economic evaluations met 70% of the criteria, compared with just over half of full economic evaluations meeting 70% of the criteria for reporting on economic evaluations. This is likely due to the significant differences in research design and varying degrees of costing analyses within the studies. Higher-rated studies were more likely to be full economic evaluations using CEA, CUA or CMA methods to analyse comparative costs and outcomes of midwifery service provision. In addition, studies that collected primary data for resource use and costs and evaluations that were presented alongside RCTs or prospective study designs rated higher. Poorer quality studies tended to analyse costs using retrospective datasets and secondary data sources, with some exceptions, such as modelling studies that employed rigorous methods [37–39].

Table 3.

Quality appraisal of included studies

Reference	Q1	Q2	Q3	Q4	Q5	Q6	Q7	Q8	Q9	Q10	Q11*	% met
Full economic evaluations
Attanasio et al. 2020 [58]	Y	N	N	N	N	Y	N/A	U	Y	N	N/A	30%
Bernitz et al. 2012 [62]	Y	Y	N	Y	N	Y	N/A	Y	Y	Y	N/A	70%
Boukhalfa et al. 2024 [39]	Y	Y	N	Y	N	Y	Y	Y	N	N	N/A	60%
Callander et al. 2020 [41]	Y	N	Y	Y	U	Y	Y	Y	Y	U	N/A	70%
Callander et al. 2021 [46]	Y	Y	U	Y	U	Y	N/A	Y	Y	Y	N/A	70%
Callander et al. 2024a [68]	Y	Y	Y	Y	U	Y	N/A	Y	Y	N	N/A	70%
Fawsitt et al. 2017 [52]	Y	Y	U	Y	N	U	N/A	N	Y	U	N/A	40%
Hitzert et al. 2017 [55]	Y	Y	Y	Y	Y	U	N/A	N	Y	Y	N/A	70%
Isaline et al. 2019 [60]	Y	Y	N	Y	U	U	N/A	Y	Y	N	N/A	50%
Kenny et al. 2015 [53]	Y	Y	Y	U	Y	Y	N/A	Y	Y	U	N/A	70%
Koto et al. 2019 [48]	Y	N	U	Y	U	Y	N/A	Y	Y	U	N/A	50%
Manasyan et al. 2011 [65]	Y	N	N	Y	N	N	N/A	N	N	N	N/A	20%
Petrou et al. 2004 [63]	Y	Y	Y	N	Y	Y	N/A	N	Y	Y	N/A	70%
Schroeder et al. 2012 [56]	Y	N	Y	Y	Y	Y	N/A	Y	Y	N	N/A	70%
Walters et al. 2015 [50]	Y	U	Y	U	U	Y	N	Y	Y	N	N/A	50%
Partial economic evaluations
Altman et al. 2017 [59]	Y	Y	Y	U	Y	Y	N/A	N	N	Y	N/A	60%
Bartlett et al. 2014 [64]	Y	N	Y	N	Y	Y	N	N	N	U	N/A	40%
Callander et al. 2024b [40]	Y	Y	Y	Y	N	Y	N	N	N	Y	N/A	60%
Cicero et al. 2022 [61]	Y	Y	Y	N	N	P	N/A	N	N	N	N/A	35%
Gao et al. 2014 [42]	Y	Y	Y	Y	N	Y	N/A	N	N	U	N/A	50%
Hendrix et al. 2009 [54]	Y	Y	Y	U	Y	Y	N/A	N	Y	Y	N/A	70%
Hu et al. 2024a [37]	Y	Y	Y	Y	Y	Y	N/A	N	N	U	N/A	60%
Hu et al. 2024b [38]	Y	Y	Y	Y	Y	Y	N	N	N	U	N/A	60%
Janssen et al. 2015 [51]	Y	Y	Y	N	Y	Y	N/A	N	N	Y	N/A	60%
Muula et al. 2006 [66]	Y	N	N	N	Y	Y	P	N	Y	N	N/A	45%
O'Brien et al. 2010 [49]	Y	Y	N	N	U	Y	N	N	N	N	N/A	30%
Scarf et al. 2020 [47]	Y	Y	Y	Y	Y	Y	N/A	N	N	N	N/A	60%
Schroeder et al. 2017 [57]	Y	Y	Y	Y	Y	Y	P	N	N	Y	N/A	75%
Toohill et al. 2012 [43]	Y	Y	Y	Y	Y	Y	N/A	N	N	Y	N/A	70%
Tracy et al. 2013 [44]	Y	Y	Y	Y	Y	Y	N/A	N	N	Y	N/A	70%
Tracy et al. 2014 [45]	Y	Y	Y	Y	Y	Y	N/A	N	N	Y	N/A	70%
Zainullah et al. 2014 [67]	N	N/A	U	Y	Y	Y	N	N	N	U	N/A	30%

*Q11 marked as not applicable owing to the global scope of this review and the vastly different contexts of each study setting

JBI Critical Appraisal Checklist for Economic Evaluations [36]

Q1 – Is there a well-defined question/objective?

Q2 – Is there a comprehensive description of alternatives

Q3 – Are all important and relevant costs and outcomes for each alternative identified?

Q4 – Has clinical effectiveness been established?

Q5 – Are costs and outcomes measured accurately?

Q6 – Are costs and outcomes valued credibly?

Q7 – Are costs and outcomes adjusted for differential timing?

Q8 – Is there any incremental analysis of costs and consequences?

Q9 – Were sensitivity analyses conducted to investigate uncertainty in estimates of costs or outcomes?

Q10 – Do study results include all issues of concern to users?

Q11 – Are the results generalizable to the setting of interest in the review?

Y, yes; N, no; U, unclear; P, partially; N/A, not applicable

Where Midwifery Has Been Costed

The overwhelming majority of midwifery-related economic evaluations have been undertaken in high-income countries (HICs), predominantly in Australia, North America and Europe (n = 26). All economic evaluations of midwifery from HICs evaluated midwifery service provision. No evaluations of midwifery education programs were identified from HICs. The highest number of studies from a single country was Australia (n = 10), where most studies analysed the costs of midwife-led care compared with standard hospital care [38, 40–46], and two investigated costs relating to place of birth [37, 47]. A total of three studies from Canada similarly focussed on midwife-led care in hospitals [48–50], and one investigated the costs of midwife-attended home births [51]. Two studies each from Ireland [52, 53], the Netherlands [54, 55], and the UK [56, 57] all evaluated midwifery services by place of birth compared with either standard hospital care or obstetric-led care units, while the two studies from the USA evaluated midwife-led care in hospital settings [58, 59]. In total, one study was found from each of the following HICs: Belgium [60], Italy [61], and Norway [62], focussing on place of birth coordinated by midwives; and Switzerland [63], focussing on place of postpartum care coordinated by midwives.

Of all sources of evidence included in this review, only six conducted economic evaluations on midwifery in LMICs [39, 64–68]. Pre-service midwifery education was costed in Afghanistan [67] and Malawi [66], one in-service education program for nurse-midwives was evaluated in Zambia [65], multiple midwife-led birth centres were evaluated across Bangladesh, Pakistan and Uganda [68] and two studies modelled the cost impact of scaling-up the midwifery workforce. One of these modelling studies included the costs of pre-service and in-service midwifery education in addition to service provision costs in Morocco [39]; the other modelling study considered only service provision costs across 58 LMICs [64]. Only one study was identified that evaluated midwifery service provision at the facility-level in LMICs [68]. A full list of countries and income levels represented in this review are outlined in Table 4.

Table 4.

Included sources of evidence by country and income level

Reference	Country/countries	Income level*
Altman et al. 2017 [59]	USA	High income
Attanasio et al. 2019 [58]	USA	High income
Bartlett et al. 2014 [64]	58 LMICs	Low income Lower-middle income Upper-middle income
Bernitz et al. 2012 [62]	Norway	High income
Boukhalfa et al. 2024 [39]	Morocco	Lower middle income
Callander et al. 2020 [41]	Australia	High income
Callander et al. 2021 [46]	Australia	High income
Callander et al. 2024a [68]	Bangladesh; Pakistan; Uganda	Low income Lower-middle income
Callander et al. 2024b [40]	Australia	High income
Cicero et al. 2022 [61]	Italy	High income
Fawsitt et al. 2017 [52]	Ireland	High income
Gao et al. 2014 [42]	Australia	High income
Hendrix et al. 2009 [54]	Netherlands	High income
Hitzert et al. 2017 [55]	Netherlands	High income
Hu et al. 2024a [37]	Australia	High income
Hu et al. 2024b [38]	Australia	High income
Isaline et al. 2019 [60]	Belgium	High income
Janssen et al. 2015 [51]	Canada	High income
Kenny et al. 2015 [53]	Ireland	High income
Koto et al. 2019 [48]	Canada	High income
Manasyan et al. 2011 [65]	Zambia	Lower-middle income
Muula et al. 2006 [66]	Malawi	Low income
O'Brien et al. 2010 [49]	Canada	High income
Petrou et al. 2004 [63]	Switzerland	High income
Scarf et al. 2020 [47]	Australia	High income
Schroeder et al. 2012 [56]	UK	High income
Schroeder et al. 2017 [57]	UK	High income
Toohill et al. 2012 [43]	Australia	High income
Tracy et al. 2013 [44]	Australia	High income
Tracy et al. 2014 [45]	Australia	High income
Walters et al. 2015 [50]	Canada	High income
Zainullah et al. 2014 [67]	Afghanistan	Low income

*As defined by the World Bank Group at time of publication [69]

How Has Midwifery Been Costed

Economic evaluations that considered both the costs and the consequences of midwifery (full economic evaluations) and those that considered only the costs (partial economic evaluations) were identified in this review. Synthesised results from data extraction are presented in Table 5. Over half (n = 9/15) of the full economic evaluations were cost-effectiveness analyses measuring the incremental costs and incremental health outcomes through averted morbidity or averted medical interventions [48, 50, 55, 56, 58, 60, 62, 65, 68]. In total, two were cost–benefit analyses valuing health effects in monetary terms; one converted women’s willingness to pay into monetary benefits [52], and the other assigned monetary value to lives saved [39]. A total of two studies were cost-utility analyses measuring health outcomes in quality-adjusted life-years [41, 46]; and two cost-minimisation analyses compared costs where interventions and comparators were found to have comparable health outcomes [53, 63]. The remaining 17 studies were partial economic evaluations, predominantly cost comparisons (n = 15) [37, 38, 40, 42–45, 47, 49, 51, 54, 57, 59, 61, 64] or cost analyses with no comparator (n = 2) [66, 67].

Table 5.

Economic evaluations of midwifery by income level

	High income countries (n = 26)	Low- and middle-income countries (n = 6)	Total (n = 32)
Economic evaluation design
Full economic evaluation	12	3	15
Cost-benefit analysis	1	1
Cost-effectiveness analysis	7	2
Cost-minimisation analysis	2	–
Cost-utility analysis	2	–	2
Partial economic evaluation	14	3	17
Cost-comparison analysis	14	1	15
Cost analysis		2	2
Research design
Case study approach		2	2
Cross-sectional study	1	–	1
Longitudinal cohort study	1	–	1
Matched control design	1	–	1
Micro-costing study	2	–	2
Prospective cohort study	4	–	4
Randomised control trial	5	–	5
Retrospective and prospective cohort study	1	1	2
Retrospective cohort study	4	–	4
Willingness-to-pay survey	1	–	1
Cost analysis design only	2	1	3
Modelling design only	4	2	6
Modelling used
No	19	3	22
Yes	7	3	10
Budget impact modelling	1		1
Decision tree modelling	3	1	4
Markov microsimulation model	1		1
Population-based simulation modelling	2	2	4
Perspective
Education funder		1	1
Healthcare funder/provider	24	1	25
Societal	2	2	4
Not stated/unclear		2	2
Study focus
Education		3	3
Midwifery pre-service education		2	2
Midwifery in-service education		1	1
Service provision and education		1	1
Service provision – midwife-led care	14	1	15
Service provision – place of birth	12	1	13
Midwife-led birth unit	5	1	6
Midwife-led birth unit and home birth	4		4
Midwife-led home birth	3		3
Comparator
Standard care (varied descriptions)	14	2	16
Obstetrician-led care	8	1	9
Family physician-led care	1		1
Family physician-led and obstetrician-led care	1		1
Standard care and obstetrician-led care	1		1
Short-stay hospital birth	1		1
No comparator		3	3
Recipients of care
All women	4	2	6
Low-risk nulliparous women	2		2
Low-risk women	20	1	21
Low-risk newborns		1	1
N/A (education only)		2	2
Time horizon for primary economic evaluation
< 1 year	21	1	22
≥ 1 year	4	1	5
2 years	1		1
6 years (education)		1	1
Unclear		3	3
Antenatal care costs considered
Antenatal care considered	18	2	20
16 weeks onwards	1		1
36 weeks onwards	1		1
37 weeks onwards	1		1
39 weeks onwards	1		1
Not stated	14	2	16
Not considered/costed	8	1	9
N/A (Education only)		3	3
Intrapartum and birth care costs considered
Intrapartum care considered	26	3	29
Inpatient birth	1		1
Inpatient labour and birth	25	3	28
Not considered/costed
N/A (Education only)		3	3
Postpartum care costs considered
Postpartum care considered	19	2	21
7 days	2		2
2 weeks	1		1
4 weeks	3		3
6 weeks	6		6
2 months	1		1
3 months	1		1
1 year	2		2
2 years	1		1
Not stated	2	2	4
Newborn only		1	1
7 days		1	1
Not considered/costed	7	1	8
N/A (education only)		2	2

The economic evaluations included in this review were conducted within a wide variety of research designs. Only five evaluations were conducted alongside, or using data from, randomised control trials [40, 44, 53, 62, 63], and six evaluations were conducted within either prospective cohort studies [43, 54–56] or prospective and retrospective (before–after) cohort studies [42, 65]. In total, ten studies used modelling: four cost-effectiveness analyses used decision-tree models [41, 48, 50, 58], four studies used population-based simulation or micro-simulation modelling [37–39, 64], one cost-utility analysis employed Markov microsimulation modelling [41] and one cost-comparison analysis modelled budget impact [40]. A small number of studies conducted costing analyses without otherwise specifying research design [51, 61, 66]. Other research designs and other findings are outlined in Table 5.

What Has Been Costed Relating to Midwifery

Midwifery Service Provision

Detailed data relating to time horizons, resource use and cost sources, costs and health outcomes for studies that evaluated midwifery service provision are presented in Supplementary Table S3. Data detailing the comparators, outcome measures and economic outcomes of all full economic evaluations are presented in Supplementary Table S5. Of the 29 studies investigating midwifery service provision, 15 evaluated midwife-led care, 13 evaluated place of birth and 1 study considered costs of both midwife-led care and education. Midwifery service provision was primarily evaluated for a population of low-risk women (n = 20), or sub-populations of low-risk primiparous women (n = 2) [45, 54], and low-risk women having an uncomplicated vaginal birth [47]. Varied criteria were used to determine which women fell into the defined low-risk category. Low-risk women were the recipients of care in all studies evaluating place of birth (n = 13) and in most (n = 10/16) that investigated midwife-led care. Midwife-led care for all women, regardless of the presence of risk factors, was evaluated by fewer studies (n = 6); two studies evaluated midwife-led care at the facility level [44, 46], and one evaluated midwife-led care for a priority population of Indigenous women [42]. A total of three studies considered midwife-led care for all women at a population level: two at the national level [39, 64] and one at a sub-national (state) level [38].

All studies investigating midwifery service provision focussed on costs associated with labour and birth, as this period of perinatal care is most likely to incur the greatest use of resources in hospital settings and, therefore, a large proportion of the costs. In addition, 20 of these studies measured resource use and costs in the antenatal period, although many (n = 14) did not specify an exact or consistent gestational period, likely reflecting the varied lengths of antenatal care according to gestation at booking. Those that did specify timeframe by gestation varied from 16 weeks onwards [54] to 39 weeks onwards [41]. Most studies (n = 21) accounted for resource use and costs in the postpartum period, with the timeframe of postpartum resource use varying considerably amongst these studies. The shortest postpartum period beyond inpatient hospital stay was 7 days [52, 55], and the longest was 2 years [41].

Resources were measured and valued using many different sources. Most studies used secondary data sources or a combination of primary and secondary data. The use of activity-based funding records was a popular means to identify and value resource use in settings where these methods were available, predominantly in Australia and Canada [37, 38, 40–44, 46, 48, 50, 51, 61, 62]. Hospital or birth centre accounts data and medical record resource use were measured in many studies that investigated facility-specific midwifery service provision [40, 44–47, 53, 55–57, 59, 62, 63]. Other secondary data sources were commonly used to identify resource use not directly associated with hospital inpatient costs, such as physician, obstetrician or general practitioner fees; ambulance or transport services; pharmaceutical services; provincial or national government accounts and outpatient service use. Most studies relied solely on secondary data, with only 11 studies incorporating primary data collection to measure resource use [40, 47, 52–57, 61, 63, 68]. All studies measured resource use from multiple sources, with the exception of one study that relied solely on per-woman cost charges from records of one health insurer [60]. One study considered use of resources and costs incurred by women or their support network measured through self-reported questionnaires, in line with the societal perspective taken for this study [63].

Midwife-Led Care

In total, 16 studies investigated midwife-led care, which was described variously according to the context and health system structure of the study location. This included evaluations of midwife-led and nurse-midwife-led care in Canada and the USA, respectively [48–50, 58, 59], caseload midwifery in Australia [38, 40–46], a general description of midwifery care in LMICs [39, 64] and postpartum midwifery care in Switzerland [63]. Comparators for these studies also reflected the dominant care models of each context. Comparators in North American studies were predominantly family physician-led and obstetrician-led care [48, 50, 58, 59], and one comparator was standard care otherwise not described [49]. Comparators in Australian studies were predominantly described as standard care in public hospitals with mixed providers working within a medical model [38, 40–44, 46]. One study considered private obstetric-led care as a second comparator in addition to standard care [45]. Standard inpatient postpartum care was the comparator to one study evaluating home-based midwifery postpartum care [63]. The costs of scaling-up obstetricians or a combination of midwives and obstetricians were used as comparators in one modelling study of multiple LMICs [64].

A total of 6 of the 16 studies on midwife-led care were full economic evaluations, comparing costs and outcomes, and one study was a cost-minimisation analysis of an RCT that presented no differences in health outcomes measured [63]. The primary health outcomes compared with costs in the full economic evaluations included maternal and newborn mortality and stillbirth [39], neonatal morbidity averted as measured by Neonatal Intensive Care Unit (NICU) admission [48, 50], maternal QALYs [41, 46] and a dual outcome of preterm birth and episiotomy [58]. The remaining nine partial economic evaluations either presented a measure of resource use and costs alone or presented costs alongside health outcomes but did not conduct a comparative analysis using economic evaluation methods.

Place of Birth

In total, 13 studies conducted economic evaluations on birthplace within care models provided by midwives. Places of birth evaluated included birth at home [37, 47, 54–56, 60, 61], in alongside midwifery units [56, 62], in freestanding midwifery units [52, 53, 56, 57, 61] or in midwife-led birth centres (MLBCs) [47, 55, 68] primarily compared with birth in a hospital or standard obstetric unit. Costs of birth in MLBCs were also compared with costs of standard care in three different countries, which included birth in a hospital or birth at home with various or no professional care providers [68]. Costs of birth at home were compared with birth in a short-stay hospital wherein care was provided by either midwives or general practitioners [54]. Of the seven studies that conducted a full economic evaluation on place of birth, the most common type of evaluations were cost-effectiveness analyses (n = 5), with one cost–benefit analysis and one cost-minimisation analysis identified. Primary outcomes for the cost-effectiveness analyses centred around the avoidance of clinical procedures during labour and birth [55, 60, 62] and averted morbidity and mortality [55, 56, 68]. The one cost–benefit analysis conducted a contingent valuation method to determine women’s willingness to pay as the monetised value of the benefits of care [52].

Midwifery Education

Few studies identified in this review economically evaluated midwifery education, and of the four identified, all were conducted in LMICs. Supplementary Table S4 presents detailed data on the time horizons, resource use, costs and conclusions for these studies. In total, three studies presented cost analyses of pre-service midwifery education [66, 67], one of which included pre-service and in-service education costs in a workforce modelling cost–benefit analysis [39]. The study from Afghanistan was a costing analysis of two midwifery pre-service education programs and presented program running total costs according to number of students, graduates and deployed midwives [67]. Costs were from the perspective of the education system funder, which was an international aid donor. In contrast, the study from Malawi took a societal perspective to analyse costs of nurse-midwife education in addition to primary and secondary schooling costs, and to present the potential economic losses to society resulting from outwards migration of nurse-midwives [66]. The study from Morocco similarly considered a societal perspective in keeping with cost–benefit analysis methodology and presented cost estimates of pre-service midwifery education, costs of simulation training and other in-service education activities alongside midwifery workforce and service provision costs [39].

Economic evaluation of an in-service midwifery education program was conducted in one study which used cost-effectiveness analysis methods to evaluate the implementation of an essential newborn care education program for a cohort of clinical nurse-midwives [65]. The precise perspective taken was not clearly documented; however, resource use and costs appear to be from the perspective of the education funder, in this case a research implementation project. This study compared health outcomes of the in-service education program with standard care (before the intervention), presenting the costs as cost per newborn life saved resulting from the impact of the education intervention. Comparators were not considered in the other three education studies.

Reported Findings of Economic Evaluations

Of the 32 sources included in this review, 30 reported conclusions regarding the costs, cost-effectiveness or cost–benefit of midwifery. Overwhelmingly, it was concluded that midwifery offers cost-savings, is cost-effective or is cost-beneficial, as presented in Table 6. Precise costs and outcomes reported for all studies are presented in Supplementary Tables S3, S4 and S5. Midwifery service provision was found to be marginally more costly than the comparator in one study, yet the authors concluded that midwife-led care remained cost-effective owing to improved health outcomes [48]; and one study concluded that midwifery care combined with obstetric care would present cost-savings when scaling-up maternal health across 58 LMICs, as opposed to either profession alone [64]. Midwifery care was reported to have comparable costs when caring for low-risk women having uncomplicated vaginal births at home, in a birth centre or in hospital [47] or when caring for low-risk women giving birth in a birth centre or in hospital [55, 60].

Table 6.

Reported findings of economic evaluations on midwifery

	High income countries (n = 26)	Low- and middle-income countries (n = 6)	Total (n = 32)
Midwifery costs reported
Less than comparator	20	1	21
Similar to or less than comparator*	2	1	3
Similar to comparator	3		3
More than comparator	1	1	2
N/A (no comparator)		3	3
Study conclusion regarding costs or cost-effectiveness
Midwifery care and obstetric care combined is cost-saving		1	1
Midwifery care has comparable costs	2		2
Midwifery care is comparable to cost savings^	3		3
Midwifery care is cost-beneficial	1	1	2
Midwifery care is cost-effective	2		2
Midwifery care offers cost savings	17	1	19
Midwifery care costs more but is cost-effective	1		1
In-service education is cost-effective		1	1
N/A (education cost description only)		2	2

*Studies that presented costs from different study sites, intervention arms or perspectives, with some costing results presented as being comparable to the comparator and some costing results as less than the comparator [40, 55, 68]

^Studies that presented differing results from two perspectives [40] or two intervention arms [55]

Considering the 11 studies that rated highest in quality appraisal (those that met 70% or more of the reporting criteria), all concluded that the provision of midwifery services is cost-saving. Midwife-led care, through caseload midwifery models in Australia, was determined to cost less than standard care models and private obstetric care and offers cost-savings to healthcare funders [41, 43–46]. Intrapartum and birth care in midwife-led units in Ireland and the UK were also concluded to cost the healthcare funder less than standard obstetric-led units, and therefore, they also presented cost-savings [53, 57]. MLBCs evaluated across three LMICs were reported to have varied and site-specific costs that tended to be comparable to or less than standard maternity care; however, overall, it was concluded that MLBCs could be cost-effective [68]. Midwifery care for women birthing at home in the Netherlands was reported as cost-effective, while birth in a birth centre had comparable costs to hospital birth [55]. In the UK, midwifery care for women at home or in a midwife-led unit was cost-saving compared with birth in standard obstetric units [56]. Finally, home-based midwifery postpartum care was reported as cost-saving to society compared with standard inpatient postpartum for women in Switzerland [63].

Discussion

Our review identified 32 economic evaluations of midwifery service provision and midwifery education conducted in a range of global contexts. While the majority of identified literature focuses on midwifery service provision in high-income countries, we identified a small number of studies focussing on broader components of the midwifery workforce, namely pre-service and in-service education. We did not identify any economic evaluations conducted on other components of the midwifery workforce, such as deployment or retention strategies. Our review identified very few studies from LMICs, in keeping with findings from Friedman et al. [70] which suggest that there has been little progress in this field of research over the past 10 years. This finding is significant given that the burden of maternal and newborn morbidity and mortality, stillbirths and midwifery workforce shortages continues to lie with LMICs [71–73].

Most economic evaluations of midwifery investigated the provision of midwifery services around the immediate intrapartum and birth period; very few studies evaluated service provision reflecting the full scope of midwifery practice, pre-service midwifery education programs or components of education, and fewer still conducted economic evaluations on in-service education programs for midwives. We did not identify any economic evaluations of midwifery recruitment strategies, deployment strategies for midwives, continued professional development strategies for midwives or any other component of the midwifery workforce. Return-on-investment analysis was limited to one study that was presented as a national-level cost–benefit analysis; this was the only study identified that accounted for a combination of in-service and pre-service education and service provision costs and benefits [39]. This finding is consistent with a recent review that mapped midwifery-related research across the Asia-Pacific region, which identified that research on the returns on investment for midwifery education, regulation, deployment and retention was lacking [74].

Of the sources of evidence included, a range of economic evaluation methods were used to measure and value costs and to measure and value outcomes within full economic evaluations. These economic studies were conducted within or alongside a diversity of research designs. Few economic evaluations were conducted alongside randomised control trials or prospective studies, and very few conducted time-motion bottom-up micro-costing methods, which are viewed as the most accurate means to measure resource use in hospitals [75, 76]. Many studies instead relied on available top-down, retrospective cost data repositories to estimate costs. The use of retrospective cost data in economic evaluation is often less detailed than prospectively collected cost data and relies on the availability, accessibility and accuracy of the original data source; however, it is more efficient to use in terms of time and cost [75]. Accessible top-down healthcare cost databases, such as activity-based funding or health insurance databases, are less commonly available in LMICs [77, 78], which is likely a contributing factor to fewer economic evaluations on midwifery service provision emerging from LMICs. Other studies have suggested that the paucity of health-related costing studies in LMICs stems from limited technical capacity [79], lack of interest from researchers and policymakers [22] and minimal experimental research to generate quality data [70].

In terms of the results reported by the economic evaluations, our finding that nearly all studies reported that provision of midwifery care presents cost savings or is cost-effective, predominantly to the healthcare funder, is similar to other reviews reporting cost savings of midwife continuity of care models [9, 25], cost savings of midwife-led care in the UK [24], and cost savings of midwifery-led care for women with complex pregnancies [80]. These reviews all report a scarcity of high-quality economic evaluations on midwifery service provision which presents significant limitations when synthesising evidence to inform and support policymakers to implement and strengthen midwifery models of care. Similarly, in our review, we identify the critical lack of economic evaluations of the midwifery workforce in LMICs, a finding that restricts the capacity to advocate for midwifery as a cost-effective strategy to address maternal and newborn health outcomes and, in doing so, support the strengthening of health systems [81, 82]. However, the evidence that is available points towards positive economic outcomes for midwifery service provision. This should be considered by policymakers, alongside recent recommendations by the WHO, to transition to midwifery models of care when designing or redesigning maternal health services [31].

The perspective taken in economic evaluations can significantly alter findings, as the perspective dictates the scope of measurement and valuation of resource use and costs [83]. While the perspective of the healthcare funder generally considers only direct health costs incurred by funders, this perspective does not account for individual out-of-pocket costs, productivity losses or flow-on costs beyond the health sector; these costs can only be considered through a societal perspective [83]. Our review revealed that most studies consider the perspective of the healthcare funder at different levels; very few studies consider a societal perspective, and we did not identify any studies that consider the perspective of individuals. This finding is in keeping with a large review of economic evaluations of maternal health interventions which found only 2.2% of studies consider the perspective of individuals; however, this review did find that 20% of studies took a societal perspective, unlike our findings [84]. Taking a societal perspective can identify if healthcare costs have transferred from the provider and onto individuals, which is particularly important when care is provided outside the hospital system, as can be the case in the provision of midwifery services or in settings where individuals incur large out-of-pocket expenses for healthcare services, as is the case in many LMICs [83]. However, taking a societal perspective will likely require greater time and resources to measure broader resource use and costs.

Midwifery care is a complex and multifaceted health intervention that combines primary health care with acute, emergency and curative health care and person-centred, respectful maternity care for women, newborns, partners and communities [7, 85]. The continuum of midwifery care spans the perinatal period from pre-conception care to antenatal, intrapartum, birth, postpartum and newborn care to family planning and abortion services [30]. Midwifery, therefore, presents a considerable difference from traditional subjects of health economic evaluation that tend to be singular, specific health interventions to address or treat a particular health outcome. Similar challenges in conducting economic evaluations have been described for other complex interventions, such as integrated care or strategies in health systems strengthening. In these interventions, health outcomes and measurement of resource use and costs are often narrowed down to singular components of care with shorter time horizons which do not reflect the comprehensive and multidimensional health outcomes or non-health outcomes such as satisfaction with care provided by such services [81, 86, 87]. This may be a contributing factor to the limited availability and quality of economic evaluations of midwifery globally.

Future economic evaluations of midwifery should take into consideration the full continuum of midwifery service provision and the midwifery workforce as an intervention that contributes to the strengthening of health systems, progress towards universal health coverage and achievement of the health-related Sustainable Development Goals. Owing to the complex and multifaceted interventions offered by midwifery care and the long-term health sequalae for women and newborns, future economic evaluations should consider broad resource use and costs to account for the continuum of care, consider longer time horizons to encompass longer-term health outcomes and take a societal perspective to account for costs incurred by women, families and societies. Strategies to support the midwifery workforce to enhance quality care should also be considered in future economic evaluations. Furthermore, economic research in this field should be supported by funders and research capacity strengthened for midwifery researchers and researchers based in LMICs.

Strengths and Limitations

To our knowledge, our review is the first to systematically identify, map and report on economic evaluations and costing studies on the midwifery workforce and service provision broadly, and with global scope. Our broad criteria for inclusion enabled a greater number of studies to be identified, resulting in a comprehensive synthesis of evidence relating to the midwifery workforce. There were some limitations to our review. Owing to English language restrictions and the omission of grey literature, some sources of evidence may not have been identified in our search strategy; this is especially true of sources from LMICs. As the scope of our review included all types of economic evaluation, with the intention of identifying, mapping and describing the available literature, we were unable to systematically synthesise the findings of included sources. Moreover, we acknowledge that there are challenges in synthesising literature from vastly different contexts and health systems with significant differences in resourcing and financing. Our mapping of the literature across contexts highlights the disparity in research between HICs and LMICs and the variation in study quality across all contexts.

Conclusions

This scoping review identified 32 economic evaluations relating to midwifery which were conducted in a number of countries over the past 20 years, covering primarily midwifery service provision and, to a lesser degree, midwifery education programs. While a small number of studies conducted evaluations in LMICs, the vast majority were conducted in higher-income countries and focussed on the provision of midwifery care for women with low-risk pregnancies. Most evaluations do not reflect the complex and multifaceted health interventions provided by midwifery care within the full scope of midwifery practice. There is ongoing need for robust, quality economic evaluations in midwifery in all global regions. Such studies would further support health policymakers and governments to make evidence-informed decisions to address midwifery workforce shortages and to ensure the provision of evidence-based and respectful care that meets the SRMNAH needs of women and girls. This aligns with universal health coverage targets and the Sustainable Development Goals.

Supplementary Information

Below is the link to the electronic supplementary material.

Declarations

Conflict of interest

All authors declare that no competing interests exist.

Funding

Open Access funding enabled and organized by CAUL and its Member Institutions. Boe Calvert is supported by an Australian Government Research Training Program Scholarship. Caroline S.E. Homer and Sarah Bar-Zeev receive funding from an Australian National Health and Medical Research Council Investigator Grant (no. 2016379).

Data Availability

All data supporting the findings of this study are available within the paper and its Supplementary Information.

Ethics Approval

Not applicable.

Consent to Participate

Not applicable.

Consent for Publication (from Patient/Participants)

Not applicable.

Code Availability

Not applicable.

Author Contributions (CRediT Statement)

Boe Calvert carried out conceptualization, investigation, data curation, validation, formal analysis, visualization, writing – original draft and writing – review and editing. Caroline SE Homer carried out conceptualization, supervision, validation, formal analysis, visualization and writing – review and editing. Sarah Bar-Zeev carried out supervision, visualization, validation and writing – review and editing. Alicia Ferguson carried out data curation, validation and writing – review and editing. Vanessa Scarf carried out conceptualization, supervision, investigation, validation, formal analysis, visualization and writing – review and editing.