Impact factor (WEB OF SCIENCE - Clarivate)

2 year: 7.2 | 5 year: 6.6


A systematic review of the effect of performance-based financing interventions on out-of-pocket expenses to improve access to, and the utilization of, maternal health services across health sectors in sub-Saharan Africa

Miriam Nkangu1,2, Julian Little1, Olumuyiwa Omonaiye3,4, Roland Pongou5, Raywat Deonandan6, Robert Geneau7, Sanni Yaya8,9

1 School of Epidemiology and Public Health, University of Ottawa, Ottawa, Canada
2 Health Promotion Alliance Cameroon (HPAC), Yaounde, Cameroon
3 Centre for Quality and Patient Safety Research, Institute for Health Transformation, Deakin University, Melbourne Burwood Campus, Australia
4 Centre for Quality and Patient Safety Research -Eastern Health Partnership, Box Hill, Victoria, Australia
5 Department of Economics, University of Ottawa, Ottawa, Canada
6 Interdisciplinary School of Health Sciences, University of Ottawa, Ottawa, Canada
7 Faculty of Health Science, University of Cape Town, South Africa
8 School of International Development and Global Studies, University of Ottawa, Ottawa, Canada
9 The George Institute for Global Health, Imperial College London, London, UK

DOI: 10.7189/jogh.13.04035




Performance-based financing (PBF) assumes that subsidizing user fees for maternal health services to reduce out-of-pocket expenses will expand coverage and reduce inequities in access to maternal health services. It is usually associated with process changes, and the idea that increasing a facility’s resources from PBF interventions can improve the availability of equipment, drugs, and medical supplies at the facility, has an indirect effect on out-of-pocket expenses. Assessment of complex interventions such as PBF requires consideration of specific underlying assumption or theories of change. Such assessment will allow a better and broader understanding of the system’s strengths and weaknesses, where the gaps lie, whether the theory of change is sound, and will inform policy design and implementation.


Following Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) checklist, we performed a systematic review and a critical appraisal of selected studies using the risk-of-bias criteria developed by the Cochrane Effective Practice and Organisation of Care. We used the Grading of Recommendation and Evaluation, Development and Assessment framework for assessing the overall strength of the evidence.


After the abstract screening (n = 9873), we deemed 302 as relevant for full-text screening and assessed 85 studies for review eligibility. Finally, we included 17 studies in the review. We could not conduct a meta-analysis, so we report a narrative synthesis. As an add-on to an existing payment mechanism, PBF may facilitate the removal of operational barriers to enhance utilization of certain maternal health services in some contexts, especially in public facilities.


PBF strategies may potentially decrease out-of-pocket expenses for specific maternal health services, especially in settings that have already instituted some form of user fee exemption policies on maternal health services. The implementation of PBF can be considered a potential access instrument in reducing out-of-pocket expenses to stimulate demand for maternal services. However, the implementation approaches employed will determine utilization, taking into consideration existing equitable and inequitable access characteristics which vary by context.


PROSPERO CRD42020222893

Print Friendly, PDF & Email

Following the 1987 Bamako Initiative, whose aim was to address severe concerns in financing care in the fields of primary health care and maternal health, most sub-Saharan African countries adopted user fees as a means of financing health care, which resulted in an increase in out-of-pocket (OOP) expenses to consumers [16]. This policy was part of a structural adjustment program promoted by the World Bank [1,2]. However, the introduction of user fees was later found to be a major barrier in accessing essential maternal and child health services, especially among poor individuals, with potential adverse consequences of impoverishing households, eg, catastrophic payments [16]. Recently, there has been an increase in the momentum and international discussions on the need to introduce reforms to subsidize or abolish OOP expenses as a strategy to increase access to, and utilization of, maternal health services, especially in sub-Saharan Africa [4].

Generally, OOP expenditures are defined as the direct payments made by individuals to health care providers at the time of service utilization and are categorized into formal and informal fees [7]. These payments include fees for medications, contraceptives, laboratory tests, provider services, facility fees, and transportation expenses and exclude any prepaid payments, such as insurance fees [79]. Financial aspects have also been reported to be a major barrier to accessing antenatal care (ANC) and skilled birth delivery in most sub-Saharan African countries [16], where these financial barriers, among other factors, have contributed to high rates of maternal mortality, reaching up to 600 per 100 000 live births [1013]. Many studies performed in Ghana, Burkina Faso, Sierra Leone, Malawi, Senegal, Kenya, and Burundi, among other countries, have shown that abolishing user fees has a substantial positive effect on the number of institutional deliveries [1,4,1016] · In Burkina Faso and Ghana, user fee reforms were found to result in up to a 25 percentage-point increase in institutional deliveries [4]. Other studies have found inequities in outcomes, benefiting rich individuals more than poor individuals [4].

To address some of these maternal health problems, performance-based financing (PBF) is being implemented in many sub-Sahara African countries, largely focused on maternal health services [7]. PBF is a supply-side strategy that pays providers incentives based on pre-defined quality and quantity criteria [8,9,1719]. It aims at paying providers incentives to change their behaviour (whereby they are motivated with incentives through policy changes) and stimulate demand for services by a change in cost and increase the ability of individuals to access services based on affordability [1719]. One of the pathways through which PBF can influence the use of maternal services is the effect on OOP expenses [1719]. For example, PBF strategies may reduce OOP expenses to enhance utilization and receive financial rewards [1719] and, in some cases, reduce informal or unofficial fees, which are payments made in addition to official fees, also called “under-the-table” payments [8,9]. PBF assumes that “subsidizing user fees and incentivizing providers to improve financial protection by reducing OOP expenditures will increase the utilization of maternal services and reduce inequities” [8]. Within this assumption, PBF intersects with universal health coverage (UHC), defined as “expanding the coverage of health services for the general population, especially for the poorest” [8,9].

PBF is usually associated with process changes. For example, increasing a facility’s resources due to PBF interventions can improve the availability of equipment, drugs, and medical supplies at the facility, thus indirectly affecting OOP expenses regarding cost for drugs purchase within the facility compared to private pharmacy off the health facility, which may directly reduce OOP expenses related to ANC, contraceptives, and skilled birth delivery [1720]. In some PBF designs, demand-side incentives to offset OOP expenses are incorporated as best practices to enhance effectiveness [8,17,18,21,22], and the exemption of user fees for the most vulnerable group intended to stimulate demand from this group [17,18]. However, it is also possible that PBF could increase the level of inequality in access if providers only select services or target services that are easier to reach [17,23].

Previous reviews have reported PBF interventions on OOP expenses as a secondary outcome with a focus only on consultation fees in low- and middle-income countries [20,2427]. However, a consultation fee is only one of the items constituting OOP expenses. Previous reviews did not specify what aspects of consultation fees may include other services not related to maternal health. Through this review, we investigated the effect of PBF on each OOP expenses for maternal services as defined above, including both demand-side (implemented to enhance PBF) and supply-side PBF strategies. We build on the above PBF assumption on subsidizing user fees to reduce OOP expenses and apply Andersen et al.’s contextual and enabling characteristics of the behavioural model on health care utilization to discuss the outcomes, using their six dimensions of access (Table S1 in the Online Supplementary Document) [28]. As per the protocol [29], the rationale for applying this model [28] was to understand the interplay of the contextual and individual characteristics and how PBF, as a mechanism to stimulate behaviour change, interacts with this to increase access and utilization of maternal services and improve equity.

In most sub-Saharan African countries, health care is largely delivered in both public and private sectors, and PBF interventions are mostly implemented across all health sectors. Additionally, OOP expenses for maternal health services in the public sector differ from those in the private sector, where higher OOP expenses in accessing maternal services have been reported [16,19]. Given the high OOP expenses that impede access to maternal services in sub-Saharan Africa, evidence is lacking on how PBF mediates OOP expenses to improve access to ANC, skilled birth delivery, and family planning across the health sectors.

Assessment of complex interventions such as PBF requires a review of the approaches aimed at assessing specific underlying assumptions or theories of change. This will allow a better and broader understanding of the system’s strengths and weaknesses, where the gaps lie, and whether the assumption or theory of change is sound, in order to inform policy design and implementation [1928,30,31]. It also helps with identifying missing links or gaps in the process to further inform programs and intervention designs [1928,30,31]. Such understanding can inform country-level policies for PBF on ANC, skilled birth delivery, and family planning. This review aligns with the growing call for UHC and for addressing equity in access to maternal health services, as well as the global action for health promotion, which calls for sustainable financing [32]. Thus, we sought to assess the effect of PBF interventions on OOP expenses as a mediating factor toward improving access to, and the utilization of ANC, skilled birth delivery, and family planning. This review was guided by the following questions: 1) What are the PBF supply-side and/or demand-side interventions implemented or identified toward reducing OOP expenses for ANC, skilled birth delivery, and family planning in sub-Saharan Africa? 2) What is the effect of PBF interventions on OOP expenses in improving access to and the utilization of ANC, skilled birth delivery, and family planning and across health sectors in sub-Saharan Africa? We used the population/intervention/comparison/outcome (PICO) framework in formulating the research questions [29].


Protocol and registration

We registered the protocol (PROSPERO CRD42020222893) and published it with BMC Systematic Reviews [29] and reported this review using the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA 2020) reporting guidelines [33,34]. We performed a critical appraisal of selected studies using the risk-of-bias criteria developed by the Cochrane Effective Practice and Organization of Care (EPOC) Group for randomized and non-randomized studies [35]. We used the Grading of Recommendation and Evaluation, Development and Assessment (GRADE) tool to assess the overall strength of the evidence with respect to effect estimates [3638]. Additionally, two reviewers (MN, OO) independently assessed the quality of systematic reviews using the Assessing the Methodological Quality of Systematic Reviews (AMSTAR) tool [39].

Search strategy, selection criteria, and data extraction

We designed our search terms in consultation with an information specialist, after which we combined the search terms for PBF and sub-Saharan Africa. We first developed the search criteria in Ovid (MEDLINE) and then adapted the search strategy to the following databases: EMBASE, CINAHL, Cochrane Library, and CABI. We also searched the World Bank e-library (for PBF impact evaluations), Department for International Development research output database, and ELDIS development database. We did not set a publication year limit, and the last search was conducted on 30 December 2022. We screened the reference lists of all the included papers for additional studies. Additionally, we searched clinical trial registries to identify completed and/or in-progress studies and hand-searched relevant study references. The full electronic search strategy for all databases is available in Table S2 in the Online Supplementary Document.

To be included in the review, each study had to meet the following criteria: 1) the intervention must be a PBF strategy and may include a demand-side intervention implemented as part of a PBF program, 2) the intervention must include consideration of aspects of OOP expenses aimed at either directly or indirectly enhancing access and utilization, 3) the study results must quantitatively report on the impact on access to and the utilization of ANC, skilled birth delivery, and family planning, irrespective of the direction of the effect. We excluded papers that reported the outcomes, but did not assess or report on any pathway of PBF intervention in reducing OOP expenses. We included randomized controlled trials (RCTs) and quasi-experimental studies conducted in sub-Saharan Africa and published in English or French. We also assessed the effect of PBF across the different health sectors. For the outcome, this included any reported PBF interventions on OOP expenses for ANC, skilled birth delivery, and family planning or any reported (or observed) changes in ANC, skilled birth delivery, and family planning attributed directly or indirectly to PBF effect on OOP changes and across the health sector. Finally, we excluded commentaries, perspectives, expert opinions, conference proceedings, and editorials.

Study selection and data extraction process

Two reviewers (MN, OO) screened the studies and assessed them for eligibility. We exported the search results to Endnote for duplication removal and referencing and to Covidence [40] for additional duplicate removal and independent screening. MN and OO conducted the initial title and abstract screening, developing the inclusion and exclusion criteria through calibration on a small number of studies to establish clarity. We resolved any disagreements through consensus, consulting additional reviewers (SY and JL) if consensus was not achieved. We retained all studies reporting an intervention using PBF and our outcomes. When we could not retrieve full-text articles could, we contacted the authors. We excluded studies that only reported PBF interventions toward ANC, skilled birth delivery, and family planning without reporting any intervention effect or policy change on OOP expenses. We also excluded all studies that were not performed within sub-Saharan Africa.

Two researchers (MN, OO) independently extracted the data from the included studies using a data extraction form adapted from EPOC [41]. We resolved any disagreements through discussion, consulting other reviewers (JL and SY) when consensus was not achieved. We contacted the authors of the included studies (twice if necessary) when the extraction was difficult or unclear or when data were missing. We extracted data on the year of PBF implementation, health sectors, and health system financing mechanisms in place before PBF, with a specific focus on the health system policies that were implemented on maternal health before PBF. We then extracted methodological and population-level data on the study design and objectives, population of interest, study outcomes, and any equity considerations. Additionally, we extracted data on access measures adapting the six dimensions of access as described by Andersen et al. [28]. For the intervention, we extracted data on any reported OOP expenses related to medication costs (specifically for those during ANC), contraception fees, skilled birth delivery, laboratory tests, and ANC fees.

Assessment of risk of bias in individual studies

We used the EPOC risk-of-bias tool to assess the quality of the methodology used for RCTs, controlled before-and-after studies (CBA), and interrupted time series analyses (ITS) [34]. We applied this tool according to its criteria for the three study designs, based on which they are categorized as “high”, “low”, and “unclear” [35]. Two reviewers (MN and OO) independently assessed risk of bias, resolving any disagreements through consensus.

Summary measures of effects

We used the GRADE tool for the summary of findings [36]. Only studies that met the following criteria were included in the appraisal: RCTs, CBAs, and ITS with a predefined shape that clearly states the pre-intervention period and clearly defines the point of analysis as the point of intervention [35]. We only narratively described the studies that did not meet these criteria. We extracted post-intervention effect estimates for RCTs, CBAs, and recorded changes in the level and slope for ITS. Difference-in-differences analysis was the most used analysis across all study designs, adjusting for covariates and potential confounders. Most of the studies generated relative effects reported as regression betas. Some of them did not report confidence intervals, or in some cases, a standard deviation. We recalculated the regression betas for eight studies (two RCTs and six CBAs) using the approach described in the Cochrane review by Diaconu et al. [25]. Thus, we divided the effect estimate of the beta by the baseline control group mean and multiplied it by 100 to obtain the relative percentage change for the outcome attributable to the intervention [25]. We did not re-analyse the data for the two ITS and did not extract data for results presented graphically. We then appraised the analysis for each study included in the summary of findings table to ensure that the studies are adjusted for clustering and controlled for potential confounders.

Poor reporting of data prevented adequate assessment of statistical heterogeneity and made it impossible to perform a robust meta-analysis. We also noted this as a gap in the way measures of precision were reported which limited the possibility of pooling the data for meta-analysis. Therefore, we reported the results as a narrative synthesis using the synthesis without meta-analysis guidelines (SWiM) [42] and excluded less robust studies in the summary of findings table.

To be eligible for the GRADE synthesis, each study must have measured OOP expenses in one of the following ways and reported the effect estimates: the probability of paying for OOP expenses during skilled birth delivery (or the percentage of women who paid for delivery), the probability of paying for any ANC visits, or the probability of paying for family planning. Alternatively, each study must have reported any change in user fees or presented results that compare any additional demand-side subsidies to offset OOP expenses vs a control group or reported results or effects of any policy change because of PBF on OOP expenses for ANC, skilled birth delivery and family planning.

We carried out a subgroup assessment for each outcome if the authors reported and disaggregated results that compared the additional incentives or strategies to offset OOP expenses on the outcome and by wealth quintile. This was achieved by extracting data for each outcome as reported in the study according to specific intervention groups based on the additional incentives and/or wealth categories. We did a sensitivity analysis based on the study design by reporting the results for each outcome assessed using an RCT [25]. Thereafter, we assessed the certainty of the evidence using the EPOC group worksheet criteria (high, moderate, low, or very low, as well as the GRADE consideration for the risk of bias, inconsistency of the results, imprecision, indirectness, and publication bias) [3538]. In the absence of meta-analyses, we used the EPOC table recommended for presenting findings without a meta-analysis, assessed the level of certainty to present the overall direction of effect of the intervention, and provided justifications for the assessed outcome [3538]. GRADE assessments were conducted by two independent reviewers (MN and OO), and disagreements were resolved through discussions and consultation with SY and JL.


We screened a total of 9873 abstracts, 302 of which were relevant for full-text screening. Seventeen studies could not be retrieved. A further 200 were excluded, and 85 were PBF studies eligible for additional screening considering the initial objective of updating some of the outcomes of an earlier Cochrane review published in 2012. However, the Cochrane review was updated in the meantime [25], so we included 17 studies which were within the focus of this review (see PRISMA flow diagram in Figure 1 and list of excluded studies in Table S3 in the Online Supplementary Document.

Figure 1.  PRISMA flow diagram, adapted from [34].

Most studies included a control group, some used matched controls, some were evaluated at the subnational and national level, and most were evaluated at the district and facility level. Nine of the studies were CBA studies [17,19,21,4348], two ITS [49,50], four RCTs [22,5153], one a cluster-RCT [54], and another a cross-sectional study [55] (Table 1). The studies were conducted in Burkina Faso, Burundi, Cameroon, the Democratic Republic of Congo, Gambia, Malawi, Nigeria, and Tanzania. PBF was implemented between 2006 and 2012, and most were either pilot studies or used data from the pilot studies. Most countries deliver care in both the public and private sectors (Table 2), and OOP expenses varied by sector and were reported to be higher in the private than the public sector, but the results were generally not reported in a disaggregated form. Four studies (from Malawi, Gambia, and Nigeria) combined demand-side strategies (conditional cash transfers) with PBF to offset OOP expenses [21,22,46,47]. Some incentivized traditional birth attendants for referrals and outreach programs and the use of home visits to enhance access [21,22,55]. The study durations ranged from six to 57 months. Some studies adjusted for clustering at the level of the facility (even when the allocation was not done at the facility level). According to the risk of bias assessment (Figure 2 and in Table S4 in the Online Supplementary Document), most CBA were of low quality mostly due to lack of randomization and allocation concealment. Some of the RCTs had high risk of bias, mostly in allocation concealment, and it was unclear how incomplete outcome data were addressed.

Table 1.  General methodological characteristics of PBF-included studies.

WordPress Data Table

ANC – antenatal care, HMIS – health management information system, DHIS – district health information system, DHS – demographic and health survey, ANOVA – analysis of variance, CCT – conditional cash transfer, PBF – performance-based financing, MCH – maternal child health, CBA – controlled before-and-after study, CTL – control, INT – intervention, P4P – pay-for-performance

*Outcomes column is only for the outcomes of interest to this review and not necessarily for all the outcomes studied by each paper.

Table 2.  Health system characteristics of PBF-included studies including health financing mechanism before PBF, pathways of OOP expenses and overall findings

WordPress Data Table

RMNCH – reproductive maternal newborn child health, SBD – skilled birth delivery, P4P – pay-for-performance, OOP – out-of-pocket, CCT – conditional cash transfer

*Findings column is only for the outcomes of interest to the review and not necessarily all the outcomes studied by each paper.

Figure 2.  Risk of bias assessment for the 16 included studies. Panel A. Interrupted time series. Panel B. Randomized controlled trials. Panel C. Controlled before-and-after studies. *Cross-sectional study.

Characteristics of PBF intervention strategies, including setting and health policies in place

Most of the PBF interventions were compared against standard care; in some cases, PBF schemes were compared with nationwide free services. Of the eight countries included in the review (Table 2), Tanzania reported an unenforced exemption policy in public facilities before PBF [17,19,48]. Following PBF, providers re-enforced the user fee exemption policy to reduce OOP expenses [19,45]. The two CBA studies reported a significant reduction in OOP expenses and increased utilization of institutional deliveries [19,45]. One study further examined this mediating effect of OOP expenses on skilled birth delivery in Tanzania and across sectors and reported an increase of 48% in institutional deliveries in private and 78% in public facilities [19]. Burundi reported a user fee exemption policy for deliveries and caesarean sections before PBF, as well as some potential constraints on transportation costs for poor individuals [43,44]. The government replaced the income loss from abolishing user fees (six months before PBF) by incorporating payments of maternal and child services to the PBF program [43,44]. Before PBF, Burkina Faso had a user fee exemption policy for ANC, as well as 80% exemption of user fees for delivery care and full exemption for very poor individuals [49,54]. After its implementation, Burkina Faso introduced a nationwide user fee policy for pregnant and lactating women [49,54].

Demand-side intervention such as conditional cash transfer was reported in Nigeria, Malawi, and Gambia [21,22,46,47,53]. Women were provided cash incentives for either ANC and/or skilled birth delivery and, in some cases, transportation costs to offset some elements of OOP expenses to enhance the utilization of ANC and skilled birth delivery. In Malawi, before PBF, all primary health care services included in the essential health package were intended to be free at the point of care in both the public and private sectors [50]. However, evidence indicates that such services are not effectively available, subjecting women to substantial OOP expenses [50]. PBF did not result in any substantial effect on household costs, but an observed substantial reduction in indirect costs for households benefitting from conditional cash transfers was reported, which suggests PBF can potentially reduce overall burdens on households [46].

Notably, Cameroon, the Democratic Republic of Congo, Nigeria, and Gambia did not report any user fee exemption policies before or after PBF [21,22,51,52]. However, in Cameroon, the government introduced subsidies for delivery kits simultaneously during the introduction of PBF [51], and OOP expenses for some family planning contraceptives and informal costs were reduced because of PBF [51,55]. In the Democratic Republic of Congo, providers considerably reduced OOP expenses for specific services [52].

In the pilot PBF study in Cameroon, PBF led to a significant reduction in household informal OOP expenses with a difference of US$3.64, as well as a significant reduction in laboratory and x-ray fees by US$2.38 and official provider fees for antenatal care by US$1.68 [51]. A cross-sectional study in Cameroon reported the use of PBF home visits to improve the utilization of modern family planning contraceptives by 59% vs 46% in the control group [55]. This was reported in an area where most of the population are farmers, with limited time to listen to family planning health promotion messages from the radio [55]. However, the authors did not report if there were any implications on cost, but rather reported that condoms were free at public health facilities, in addition to a subsidy for other modern family planning contraceptives in public facilities at an average cost of US$2.63; but they did not mention if this was due to PBF or not [55].

One RCT from the Democratic Republic of Congo assessed the percentage of paying for consultations and sending gifts to providers, but this was not disaggregated and may have included other forms of consultation not related to ANC, family planning, and/or skilled birth delivery [52]. However, the study reported a significant cost reduction for drugs [52] and noted that most facilities with user fees reduced due to PBF were those with initially high user fees [52]. Generally, user fees have been reported to be higher in private than in public facilities [19,48,51,52]. In Tanzania, the improved availability of drugs in turn minimized the need to pay for drugs in private pharmacies, which are usually expensive [19,45]. One RCT study from Cameroon reported on the availability of family planning contraceptives but did not report any effect associated with this on cost [51].

Impact on outcomes: Direction of relative effect as rated in GRADE

We used the GRADE tool to assess the overall strength of the evidence with respect to effect estimates and reported results [36,37]. We present and briefly summarize the outcome details in the summary of findings table (Table 3) using EPOC’s recommended plain language summary approach [36] and the relative effect. Generally, the evidence is of low certainty. We noted that, PBF interventions with a standard comparator or standard care yielded positive results compared to when PBF was compared against other additional demand side enhancing interventions. A total of 10 studies met the eligibility criteria as described above for the GRADE assessment.

Table 3.  Effect of PBF on OOP expenses to improve access and use of ANC, skilled birth delivery and family planning – GRADE summary of findings*

WordPress Data Table

GRADE – Grading of Recommendations Assessment, Development and Evaluation, ANC – antenatal care, RCT – randomized controlled trial, OOP – out-of-pocket expense, CBA – controlled before-and-after, PBF – performance-based financing, N/A – not applicable, EmOC – emergency obstetric care

*Most studies had limitations on one or more aspects of risk of bias and indirectness (as most did not directly meet the patient/population, intervention, comparison, and outcome (PICO) assessment in terms of reporting the mediating effect or disaggregating data correctly) and inconsistency.

‡Definitions from EPOC Group below on the grading of the evidence. High – this research provides a very good indication of the likely effect. The likelihood that the effect will be substantially different is low. Moderate – this research provides a good indication of the likely effect. The likelihood that the effect will be substantially different is moderate. Low – this research provides some indication of the likely effect. However, the likelihood that it will be substantially different is high. Very low – this research does not provide a reliable indication of the likely effect. The likelihood that the effect will be substantially different is very high. Substantially different – a large enough difference that it might affect a decision [36].

‡The RCT study [54] that reported on subgroup for almost all indicators had concerns for risk of bias and reported contextual limitations and timing of data collection. Subgroup for additional demand-side incentives implies either conditional cash transfer support, community-based program exempting poor and/or user fee exemption policy for the poor or specific policy on user fee exemption following PBF implementation.

Effect on skilled birth delivery

All 10 studies (from five countries) reported on skilled birth delivery. There was moderate evidence from two CBA studies from Tanzania that measured the probability of paying OOP expenses for skilled birth delivery, and relative effect indicates that PBF may decrease OOP expenses for skilled birth delivery by 31% to 42% (Table 3). The two studies from Tanzania indicate PBF’s relative effect on OOP expenses may probably improve skilled birth delivery by 9.4% in private and 8.3% in public facilites [19,45]. Low certainty evidence from three studies (two RCTs and one CBA) that measured the effect of PBF on skilled birth delivery without reporting the mediating role on OOP expenses indicate PBF may improve skilled birth delivery from as low as 0.05% to 5.6% [49,51,54]. Three CBA studies reported on PBF effect on skilled birth delivery in public facilities [19,44,45], while four other studies (two RCTs and one ITS) reported no significant change in institutional deliveries [4952]. However, one ITS study from Malawi reported on a higher baseline utilization of skilled birth delivery of 90% [50]. A pilot CBA study in Burundi reported an increased percentage of institutional deliveries which was limited to selected provinces, but found no impact on this indicator at the national level [43,44]. With PBF additional equity-enhancing components as reported in one RCT study, it may slightly decrease utilization of skilled birth delivery as low as 0.009% [54].

Effect on ANC

Almost all studies reported high baseline utilization rates for the first ANC visit, ranging between 78% and 96% with low certainty of evidence. Although PBF had little or no impact on this indicator, an observed increase in the quality of ANC provided was reported, represented by increased consumption of anti-tetanus vaccines [43,44,54], antimalarial drugs [19,45,50], and ion-based medications [45,54], which are all components related to ANC care. Four studies reported on the use of ANC visits; PBF may lead to little or no difference in ANC visits ranging from 1.14% to 3.3% [44,45,50,51]. Moreover, three studies reported on ANC visits in the 1st trimester, with little or no difference indicating a relative change of 0.08% or a slight reduction by 0.1% [44,50,54]. In one cluster randomized controlled trial (C-RCT) study, PBF with an additional enhancing financing mechanism slightly reduced the utilization of ANC in the 1st trimester by 0.1% [54], however, the study reported power limitations [54]. Three studies reported on the use of four or more ANC visits; they found that PBF may improve utilization by 5.5%, while possibly slightly reducing it in some contexts by 0.01% [45,50,54]. In one RCT study, PBF with additional enhancing financing mechanisms slightly reduced four or more ANC visits by 0.06% [54]. For the interrupted time series, the average monthly provision of ANC increased in Burkina Faso with the implementation of the nationwide free program [49]. No study reported on ANC visits in the second and/or third trimester. Rather, they focused more on the use of more than one ANC or any ANC, which does not help capture appropriate trends in utilization to monitor and understand the breakpoints and relevant gaps in access and utilization. One CBA study reported a greater effect on the probability of paying OOP expenses during ANC but a positive effect on the probability of giving gifts during ANC [45].

Effect on family planning

Seven studies reported PBF effect on family planning with low certainty of evidence [43,45,47,4951,54]. The relative effect of two studies (CBA and a C-RCT) indicated that PBF may improve utilization of family planning by 0.18%, while another study found that it may decrease utilization by 1.8% [45,54]. Another C-RCT study found that PBF with additional equity elements may have the relative effect of decreasing utilization of family planning by 0.19% [54] yet it also reported power limitations. Three other studies where the relative effect could not be calculated reported increased use of modern family planning [43,50,51] and one study reported no impact on family planning with limitations on power and the timing of data collection [50]. One study reported on family planning counselling [50]. A pilot study in Burundi reported an increase in the use of family planning which was limited to selected provinces, but found no impact on these indicators when the study was extended to the national level [43,44]. In Burkina Faso, the average monthly provision of family planning increased with the implementation of the nationwide free program [49].

Impact on equity

Five [17,44,46,49,54] of the 10 assessed studies reported on the distributional effect of PBF by wealth group. All study countries (ie, Tanzania, Burkina Faso, Burundi, and Malawi) implemented some form of a user fee exemption intervention either as a national policy exemption, demand-side financing, or additional community-based insurance targeting poor individuals. All studies used household characteristics and asset ownership to categorize wealth groups into various quintiles. One CBA study further used the predisposing and enabling characteristics suggested by Andersen et al. to define the population groups [17]. However, none of the studies provided a definition of poor in the context of PBF vis-à-vis the context of the study in relation to the scale being used. One C-RCT study noted this as a limitation that the wealth grouping in their study did not match the characteristics of those targeted as poor individuals in the PBF community-based program [54].

A Tanzanian study reported increased utilization of institutional deliveries among women from the intermediate wealth quintile and uninsured households likely due to the increased adherence to user fee exemption in public facilities and the improved availability of drugs [17,19]. An indication of a pro-poor effect on the rate of institutional deliveries in public facilities was reported in Tanzania, with a change in health care cost amongst the poor and intermediate households and women in rural districts but not among the least poor [17,45]. Some limitations were reported in the statistical power and timing of data collection in one of the studies [17]. In Burundi, an increase in the proportion of institutional deliveries, especially among better-off individuals, was reported, but there was no effect on poor individuals and no changes in the place of residence were found [44].

In Burkina Faso, the relative effect observed in one C-RCT study on utilization of ANC in the first trimester had a slight improvement (0.10%) for the lowest wealth quintile and a slight decrease (0.14%) for the highest quintile [54]. There was a similar pattern for four or more ANC visits and a slight improvement (0.03%) in utilization for the lowest quintile and a decreased (0.10%) for the highest quintile [54]. With additional financing incentives, the study reported a slight decrease across all income groups in utilization for family planning with a relative effect of 0.19% for the lowest quintile and 0.33% for the highest quintile [54]. Following the abolishment of user fees in Burkina Faso, the Ministry of Health removed the additional payments for equity interventions that were targeted for certain groups of population and services [54]. The equity interventions did not generate additional benefits compared to PBF alone for women and those from the poorest category [49,54]. Some potential contextual limitations were reported in relation to inadequate human resources and structural constraints, delays in payments, and previous experience with poor implementation approaches [49,54]. For skilled birth delivery, there was a slight increase in utilization among the highest quintile (0.0007%) and a slight decrease for the lowest quintile (0.0001) [54]. The study reported some power and contextual limitations as the study data was collected about one year after the launch of the national free health care policy targeting women, which also influenced PBF equity measures as some additional compensations were abolished which may have also demotivated health workers from exploring innovative approaches to stimulate demand [49,54]. Similarly, in the Democratic Republic of Congo, abolishing additional compensation was reported to have demotivated workers and increased the rate of staff absenteeism [52].

In Burundi, there was no indication of the heterogeneity of effects for ANC and family planning by poverty status [43], no improvement in institutional deliveries, and no difference between public and private facilities [44]. However, an increase in institutional deliveries was reported among nonpoor individuals (wealthier group), but no effect was found among poor individuals [44]. The Burundi study also reported that although the user costs have been waived, the transportation costs may have likely affected the ability of poor women to deliver in a health facility [44].


Generally, the evidence we found in our review is of low certainty. We identified that PBF as a stand-alone supply side strategy can motivate providers to reduce OOP expenses for certain maternal services, such as in Cameroon, and the Democratic Republic of Congo. However, the reduction may not stimulate demand to generate impact on utilization due to other contextual and individual demand-side barriers. In settings in which policies regarding free health care at the point of use have been instituted, such as Tanzania, or in which services have been subsidized, such as Burkina Faso, PBF likely facilitates or enables the utilization of skilled birth delivery. While the effective implementation of free health care policy or subsidized services may be hindered by operational barriers at the health facility or hidden costs, PBF may facilitate the removal of such operational barriers to enhance utilization of services especially in public facilities.

Impact on access and utilization

Based on the earlier stated PBF assumption, PBF can be considered a behavioural change mechanism whereby providers are motivated with incentives through policy changes (contextual enabling characteristics) to stimulate demand by a change in cost (as one of the possible pathways), thereby increasing the ability of individuals to access services based on affordability (individual enabling characteristics) and quality of care [19,28]. According to the Andersen et al.’s framework, these contextual and individual enabling characteristics are a function of potential access, which may be either equitable or inequitable [28]. Thus, any PBF policy on OOP expenses is an instrument of potential access. However, these policy strategies are also approached differently by health facilities and health sectors to enhance the individual enabling characteristics towards improving access. These enabling characteristics thus translates to realized access, which is the actual utilization of services [28]. Most studies reported an improvement in the availability of drugs, which is an indicator of potential access in terms of an indirect potential reduction in OOP expenses for women who otherwise need to purchase drugs from the private sector, which is expensive [19]. In some countries, PBF favoured individuals from the middle-income and sometimes high-income quintiles, and in limited cases those from the low-income quintile. One question, however, is how this translates to effective access, which is a function of both potential and realized access.

In Tanzania, PBF was found to be effective in re-enforcing the user fee exemption policy and led to an immediate increase in the utilization of skilled birth delivery, with an even greater increase observed in public facilities [19,45]. In Malawi, PBF was found to be effective in improving access to ANC because of “resilience against interrupted supply chains and improvements in attendance rates” [50]. PBF was also found to be effective in the timing of ANC during the first trimester in some countries [,50,54], and for even more than four ANC visits in some contexts [44,50], but not among the poor quintile for more than four ANC visits in Burkina Faso [54]. In Burkina Faso and Burundi, an inequitable increase in utilization was observed among nonpoor individuals [44,49,54]. In the Democratic Republic of Congo, despite an increased in efforts by providers to reduce prices, PBF did not increase utilization, and the study suggests poor strategies employed by providers to stimulate demand given an existing poor level of satisfaction toward the quality of services, especially when most of the population are poorly educated and informed [52]. In Burkina Faso, poor implementation approaches and delay in payments were reported [49]. In Cameroon, the increased efforts made by providers to reduce user fees, including reductions in the prices of laboratory services and significant reductions in informal fees, among other efforts, did not lead to any significant change in the level of utilization for antenatal care and skilled birth delivery, suggesting existing high OOP expenses [51] ·

Policy implications

The evidence on the effect of PBF as a way of reducing financial barriers by decreasing OOP expenses to improve access is low and of poor quality, especially when isolating the effect of PBF as a supply side strategy from other demand side strategies like conditional cash transfers and/or policy like removal of user fees.

PBF designs may appear similar, but implementation varies by context and contextual characteristics seem to play a major role in determining the outcome of the PBF. Based on the findings, PBF effect on equity was not significant both in settings with existing high OOP expenses and those with free or subsidized services, and/or combined implementation of CCT and supply. In some settings, there was no effect on equity. Therefore, it is important to carefully design PBF interventions to align with country specific problems and contexts and or assess other strategies that may generate the same results with significant effect on equity. OOP is an indicator for assessing equity, so all its elements should be incorporated to understand various equitable and inequitable process changes; if a policy is not impacting the poor or addressing equity concerns, it should be reviewed and re-strategized. Since PBF is an addition to existing payment mechanisms (with the mixed results of PBF effect from studies and possible administrative cost implications in the implementation), it is worth assessing its cost/benefit as a policy mechanism within specific context to enhance access to maternal health services. Most studies did not conduct such assessments, indicating an important research gap.

The PBF definition on equity aspects has not been well studied [56]; as reported in one of the studies, the classification of “poor” was not in line with the PBF community approach [54].There is a need to clearly define the “poor” within each context and align with the PBF definition of who constitute the poor, to ensure appropriate classification and reporting to inform policy intervention strategies towards achieving universal health coverage.

Methodological considerations and study limitations

We observed some level of bias in all studies included in this review. There are substantial differences in the reporting of effect estimates, intervention designs, outcomes, and study groups, which precluded statistical pooling of the results. The effect estimates reported for PBF on OOP expenses for the public and private sectors were often absent; some studies rather controlled for it and did not use the opportunity to assess potential differences. Most importantly, PBF has been implemented in most countries over a reasonable period. In countries where some policy changes and adaption are ongoing [57], research must focus on designing and collecting new data to assess the policy over time and not focus only on data from the pilot studies which may have their own limitations. Moreover, most pilot studies are designed and conducted in an ideal situation, therefore, the findings may not reflect the actual implementation processes and challenges over time. This implies that there is limited evidence to inform relevant areas of the policy that may require revision or adoption to support advocacy and sustainability.

We assessed the quality of the review online using the AMSTAR tool and any disagreements were discussed to obtain consensus. The results from the two independent reviewers indicates moderate quality rating, which implies more than one non-critical weakness, but no critical flaws [39]. This means that the review may provide an accurate summary of the results of the included studies [39]. One weakness of this review, as per the AMSTAR tool, is the lack of a meta-analysis to enable an assessment of publication bias; furthermore, we did not assess funding sources from the individual sources [39].

One strength of this systematic review is its focus on sub-Saharan Africa where there is an interest in assessing the PBF-induced effect on OOP expenses to improve access and utilization of maternal services [30,31]. We initially aimed to update some outcomes of a published Cochrane review [58], but it had been updated and published during the review process [25]. However, some studies included here were not considered in the summary of findings of the Cochrane review, especially those studies published after 2018. The protocol was slightly modified at full text screening to exclude some of the papers that did not provide detailed information on OOP expenses but still met the PICO criteria; at this stage, we also excluded some of the papers that were included in the updated Cochrane review. Finally, the dimension of access should be interpreted with caution; it is focused specifically on discussing the PBF assumption on OOP expenses as used in this review and based on what was reported in the studies, which may not comprehensively reflect what is being implemented in the country.


The results of our review are mixed, with important contextual differences and implementation approaches. Given that PBF is an addendum to an existing payment mechanism to stimulate behaviour change, the implementation of PBF can be considered a potential access instrument in reducing OOP expenses to stimulate demand for maternal services. However, the implementation approaches employed will determine utilization (realized access), taking into consideration existing equitable and inequitable access characteristics which vary by context.

Additional material

Online Supplementary Document


We would like to thank the Librarian Lindsay Sikora at the university of Ottawa who provided support, advise and peer review in developing the search strategy. We are also grateful to Nigussie Assefa for providing feedback on the statistical approach.

Data availability: All data relevant for the study are included in the review and/or attached as supplementary material.

[1] Funding: No funding was provided for this review.

[2] Authorship contributions: MN conceived the idea for the systematic review, conducted screening, data collection, analysis, investigation, methodology, and wrote the original draft of the manuscript. SY and JL supervised the methodology, analysis and critically revised the manuscript. OO assisted in the screening and data collection. RD, RP, RG and OO critically revised the manuscript and provided feedback. All authors reviewed and agreed on the final version of the manuscript.

[3] Disclosure of interest: The authors completed the ICMJE Disclosure of Interest Form (available upon request from the corresponding author (and disclose no relevant interests).


[1] S Dzakpasu, T Powell-Jackson, and OMR Campbell. Impact of user fees on maternal health service utilization and related health outcomes: a systematic review. Health Policy Plan. 2014;29:137-150. DOI: 10.1093/heapol/czs142. [PMID:23372035]

[2] Akin JS, Birdsall N, De Ferranti DM, Bank W. Financing Health Services in Developing Countries: An Agenda for Reform. Washington, D.C.: World Bank; 1987.

[3] B McKinnon, S Harper, JS Kaufman, and Y Bergevin. Removing user fees for facility-based delivery services: a difference-in-differences evaluation from ten sub-Saharan African countries. Health Policy Plan. 2015;30:432-41. DOI: 10.1093/heapol/czu027. [PMID:24816570]

[4] T Leone, V Cetorelli, S Neal, and Z Matthews. Financial accessibility and user fee reforms for maternal healthcare in five sub-Saharan countries: a quasi-experimental analysis. BMJ Open. 2016;6:e009692. DOI: 10.1136/bmjopen-2015-009692. [PMID:26823178]

[5] Y Beaugé, V Ridde, E Bonnet, S Souleymane, N Kuunibe, and M De Allegri. Factors related to excessive out-of-pocket expenditures among the ultra-poor after discontinuity of PBF: a cross-sectional study in Burkina Faso. Health Econ Rev. 2020;10:36 DOI: 10.1186/s13561-020-00293-w. [PMID:33188618]

[6] IB Meda, A Baguiya, V Ridde, HG Ouédraogo, A Dumont, and S Kouanda. Out-of-pocket payments in the context of a free maternal health care policy in Burkina Faso: a national cross-sectional survey. Health Econ Rev. 2019;9:11 DOI: 10.1186/s13561-019-0228-8. [PMID:30919219]

[7] United States Agency for International Development (USAID). Out-of-pocket Health Expenditures Module: Questionnaire and Interviewer’s Manual, Demographic and Health Surveys Methodology 2008. Available: Accessed: 10 October 2021.

[8] Fritsche G, Soeters R, Meessen B. Performance-based financing toolkit. Washingdon, D.C: World Bank; 2014.

[9] SINA Health. Performance-based financing in action. 2018. Available: Accessed: 7 November 2021.

[10] JL Alvarez, R Gil, V Hernández, and A Gil. Factors associated with maternal mortality in Sub-Saharan Africa: an ecological study. BMC Public Health. 2009;9:462 DOI: 10.1186/1471-2458-9-462. [PMID:20003411]

[11] J Borghi, T Ensor, A Somanathan, C Lissner, and A Mills. Mobilising financial resources for maternal health. Lancet. 2006;368:1457-65. DOI: 10.1016/S0140-6736(06)69383-5. [PMID:17055948]

[12] MO hIarlaithe, N Grede, S de Pee, and M Bloem. Economic and social factors are some of the most common barriers preventing women from accessing Maternal and Newborn Child Health (MNCH) and Prevention of Mother-to- Child Transmission (PMTCT) services: A Literature Review. AIDS Behav. 2014;18:S516-30. DOI: 10.1007/s10461-014-0756-5. [PMID:24691921]

[13] G Manthalu, D Yi, S Farrar, and D Nkhoma. The effect of user fee exemption on the utilization of maternal health care at mission health facilities in Malawi. Health Policy Plan. 2016;31:1184-92. DOI: 10.1093/heapol/czw050. [PMID:27175033]

[14] M Lagarde and N Palmer. The impact of user fees on health service utilization in low- and middle-income countries: how strong is the evidence? Bull World Health Organ. 2008;86:839-48. DOI: 10.2471/BLT.07.049197. [PMID:19030689]

[15] B McKinnon, S Harper, and JS Kaufman. Who benefits from removing user fees for facility-based delivery services? Evidence on socioeconomic differences from Ghana, Senegal and Sierra Leone. Soc Sci Med. 2015;135:117-23. DOI: 10.1016/j.socscimed.2015.05.003. [PMID:25965892]

[16] R Maini, R Van den Bergh, J van Griensven, K Tayler-Smith, J Ousley, and D Carter. Picking up the bill – improving health-care utilisation in the Democratic Republic of Congo through user fee subsidisation: a before and after study. BMC Health Serv Res. 2014;14:504 DOI: 10.1186/s12913-014-0504-6. [PMID:25370385]

[17] P Binyaruka, B Robberstad, G Torsvik, and J Borghi. Who benefits from increased service utilisation? Examining the distributional effects of payment for performance in Tanzania. Int J Equity Health. 2018;17:14 DOI: 10.1186/s12939-018-0728-x. [PMID:29378658]

[18] M Steenland, PJ Robyn, P Compaore, M Kabore, B Tapsoba, and A Zongo. Performance-based financing to increase utilization of maternal health services: Evidence from Burkina Faso. SSM Popul Health. 2017;3:179-84. DOI: 10.1016/j.ssmph.2017.01.001. [PMID:29349214]

[19] L Anselmi, P Binyaruka, and J Borghi. Understanding causal pathways within health systems policy evaluation through mediation analysis: an application to payment for performance (P4P) in Tanzania. Implement Sci. 2017;12:10 DOI: 10.1186/s13012-016-0540-1. [PMID:28148305]

[20] A Das, SS Gopalan, and D Chandramohan. Effect of pay for performance to improve quality of maternal and childcare in low- and middle-income countries: a systematic review. BMC Public Health. 2016;16:321 DOI: 10.1186/s12889-016-2982-4. [PMID:27074711]

[21] GM Ashir, HV Doctor, and GY Afenyadu. Performance based financing and uptake of maternal and child health services in yobe sate, northern Nigeria. Glob J Health Sci. 2013;5:34-41. DOI: 10.5539/gjhs.v5n3p34. [PMID:23618473]

[22] L Ferguson, R Hasan, C Boudreaux, H Thomas, M Jallow, and G Fink. Project Implementation Committee (PIC). Results-based financing to increase uptake of skilled delivery services in The Gambia: using the ‘three delays’ model to interpret midline evaluation findings. BMC Pregnancy Childbirth. 2020;20:712 DOI: 10.1186/s12884-020-03387-9. [PMID:33228543]

[23] E Paul, L Albert, BN Bisala, O Bodson, E Bonnet, and P Bossyns. Performance-based financing in low-income and middle-income countries: isn’t it time for a rethink? BMJ Glob Health. 2018;3:e000664. DOI: 10.1136/bmjgh-2017-000664. [PMID:29564163]

[24] N James, K Lawson, and Y Acharya. Evidence on result-based financing in maternal and child health in low- and middle-income countries: a systematic review. Glob Health Res Policy. 2020;5:31 DOI: 10.1186/s41256-020-00158-z. [PMID:32626825]

[25] K Diaconu, J Falconer, A Verbel, A Fretheim, and S Witter. Paying for performance to improve the delivery of health interventions in low- and middle-income countries. Cochrane Database Syst Rev. 2021;5:CD007899. [PMID:33951190]

[26] C Blacklock, E MacPepple, S Kunutsor, and S Witter. Paying for Performance to Improve the Delivery and Uptake of Family Planning in Low- and Middle-Income Countries: A Systematic Review. Stud Fam Plann. 2016;47:309-24. DOI: 10.1111/sifp.12001. [PMID:27859313]

[27] NS Singh, RJ Kovacs, R Cassidy, SR Kristensen, J Borghi, and GW Brown. A realist review to assess for whom, under what conditions and how pay for performance programmes work in low- and middle-income countries. Soc Sci Med. 2021;270:113624. DOI: 10.1016/j.socscimed.2020.113624. [PMID:33373774]

[28] Andersen R, Davidson P, Baumeister S. Changing the U.S. health care system: Key issues in health services policy and management. Chapter: Improving access to care in America. In: Kominski F, editor. New Jersey: John Wiley and Sons; 2013.

[29] M Nkangu, J Little, O Omonaiye, and S Yaya. The effect of performance-based financing interventions on out-of-pocket expenses intended to improve access to and utilization of maternal health services in sub-Saharan Africa: protocol for a systematic review and meta-analysis. Syst Rev. 2022;11:133 DOI: 10.1186/s13643-022-01990-9. [PMID:35773732]

[30] White H. Using causal chain analysis in systematic reviews. 2021. Available: Accessed: 14 April 2023.

[31] DP MacKinnon, AJ Fairchild, and MS Fritz. Mediation analysis. Annu Rev Psychol. 2007;58:593 DOI: 10.1146/annurev.psych.58.110405.085542. [PMID:16968208]

[32] World Health Organization Global Action Plan for healthy lives and wellbeing Available: Accessed: 1 March 2020.

[33] A Liberati, DG Altman, J Tetzlaff, C Mulrow, PC Gøtzsche, and JPA Loannidis. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Med. 2009;6:e1000100. DOI: 10.1371/journal.pmed.1000100. [PMID:19621070]

[34] MJ Page, JE McKenzie, PM Bossuyt, I Boutron, TC Hoffmann, and CD Mulrow. The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ. 2021;372:n71 DOI: 10.1136/bmj.n71. [PMID:33782057]

[35] Effective Practice and Organisation of Care (EPOC). Suggested risk of bias criteria for EPOC reviews: EPOC resources for review authors. 2017. Available: Accessed: 1 July 2021.

[36] Cochrane Effective Practice and Organisation of Care (EPOC). EPOC worksheets for preparing a Summary of Findings (SoF) table using GRADE: EPOC Resources for review authors. 2017. Available: Accessed: 1 July 2021.

[37] M Hultcrantz, D Rind, EA Akl, S Treweek, RA Mustafa, and A Iorio. The GRADE Working Group clarifies the construct of certainty of evidence. J Clin Epidemiol. 2017;87:4-13. DOI: 10.1016/j.jclinepi.2017.05.006. [PMID:28529184]

[38] Deeks JJ, Higgins JP, Altman DG. Chapter 10: Analysing data and undertaking meta-analyses. In: Higgins JP, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, et al, editors. Cochrane Handbook for Systematic Reviews of Interventions version 6.0 (updated July 2019). London: Cochrane; 2019.

[39] BJ Shea, BC Reeves, GA Wells, M Thuku, C Hamel, and J Moran. AMSTAR 2: a critical appraisal tool for systematic reviews that include randomized or non-randomized studies of healthcare interventions, or both BMJ. 2017;358:j4008 DOI: 10.1136/bmj.j4008. [PMID:28935701]

[40] Covidence. Covidence Systematic review management software. Available: Accessed: 1 November 2020.

[41] Cochrane Effective Practice and Organisation of Care (EPOC). EPOC Resources for review authors. 2017. Available: Accessed: 1 February 2020.

[42] M Campbell, JE McKenzie, A Sowden, SV Katikireddi, SE Brennan, and S Ellis. Synthesis without meta-analysis (SWiM) in systematic reviews: reporting guideline. BMJ. 2020;368:l6890 DOI: 10.1136/bmj.l6890. [PMID:31948937]

[43] I Bonfrer, R Soeters, E Van de Poel, O Basenya, G Longin, and F van de Looij. Introduction of performance-based financing in Burundi was associated with improvements in care and quality. Health Aff (Millwood). 2014;33:2179-87. DOI: 10.1377/hlthaff.2014.0081. [PMID:25489036]

[44] I Bonfrer, E Van de Poel, and E Van Doorslaer. The effects of performance incentives on the utilization and quality of maternal and childcare in Burundi. Soc Sci Med. 2014;123:96-104. DOI: 10.1016/j.socscimed.2014.11.004. [PMID:25462610]

[45] P Binyaruka, E Patouillard, T Powell-Jackson, G Greco, O Maestad, and J Borghi. Effect of Paying for Performance on Utilisation, Quality, and User Costs of Health Services in Tanzania: A Controlled Before and After Study. PLoS One. 2015;10:e0135013. DOI: 10.1371/journal.pone.0135013. [PMID:26317510]

[46] J Chinkhumba, M De Allegri, J Mazalale, S Brenner, D Mathanga, and AS Muula. Household costs and time to seek care for pregnancy related complications: The role of results-based financing. PLoS One. 2017;12:e0182326. DOI: 10.1371/journal.pone.0182326. [PMID:28934320]

[47] S Brenner, J Mazalale, D Wilhelm, RC Nesbitt, TJ Lohela, and J Chinkhumba. Impact of results-based financing on effective obstetric care coverage: evidence from a quasi-experimental study in Malawi. BMC Health Serv Res. 2018;18:791 DOI: 10.1186/s12913-018-3589-5. [PMID:30340491]

[48] Canavan A, Swai G. Payment for Performance (P4P) Evaluation: Tanzania Country Report for Cordaid Godfrey Swai, National Consultant Width 1. Amsterdam: KIT Development Policy & Practice; 2008.

[49] N Kuunibe, J Lohmann, M Hillebrecht, HT Nguyen, G Tougri, and M De Allegri. What happens when performance-based financing meets free healthcare? Evidence from an interrupted time-series analysis. Health Policy Plan. 2020;35:906-17. DOI: 10.1093/heapol/czaa062. [PMID:32601671]

[50] S Brenner, RP Chase, SA McMahon, J Lohmann, CJ Makwero, and AS Muula. Effect Heterogeneity in Responding to Performance-Based Incentives: A Quasi-Experimental Comparison of Impacts on Health Service Indicators Between Hospitals and Health Centers in Malawi. Health Syst Reform. 2020;6:e1745580. DOI: 10.1080/23288604.2020.1745580. [PMID:32521206]

[51] D de Walque, PJ Robyn, H Saidou, G Sorgho, and M Steenland. Looking into the performance-based financing black box: evidence from an impact evaluation in the health sector in Cameroon. Health Policy Plan. 2021;36:835-47. DOI: 10.1093/heapol/czab002. [PMID:33963406]

[52] Huillery E, Seban J. Performance-based financing, motivation, and final output in the health sector: experimental evidence from the Democratic Republic of Congo. 2014. Available: Accessed: 14 April 2023.

[53] L Ferguson, C Boudreaux, M Cheyassin Phall, B Jallow, MN Fofana, and L Njie. Facility and Community Results-Based Financing to Improve Maternal and Child Nutrition and Health in The Gambia. Health Syst Reform. 2022;8:2117320. DOI: 10.1080/23288604.2022.2117320. [PMID:36084280]

[54] T Mwase, J Lohmann, S Hamadou, S Brenner, SMA Somda, and H Hien. Can Combining Performance-Based Financing with Equity Measures Result in Greater Equity in Utilization of Maternal Care Services? Evidence From Burkina Faso. Int J Health Policy Manag. 2022;11:308-22. [PMID:32729286]

[55] TO Egbe, J Atashili, E Talla, and MBS Atanga. Effect of performance-based financing home visiting on the use of modern methods of contraception in the kumbo east health district, Cameroon. Contracept Reprod Med. 2016;1:19 DOI: 10.1186/s40834-016-0030-5. [PMID:29201408]

[56] V Ridde, L Gautier, AM Turcotte-Tremblay, I Sieleunou, and E Paul. Performance-based Financing in Africa: Time to Test Measures for Equity. Int J Health Serv. 2018;48:549-61. DOI: 10.1177/0020731418779508. [PMID:29932352]

[57] S Abimbola. Evidence as Cliché: Using Trials to Tamper with Governance. Bull CODESRIA. 2020;4:9-15.

[58] S Witter, A Fretheim, FL Kessy, and AK Lindahl. Paying for performance to improve the delivery of health interventions in low- and middle-income countries. Cochrane Database Syst Rev. 2012;CD007899. DOI: 10.1002/14651858.CD007899.pub2. [PMID:22336833]

Correspondence to:
Miriam N Nkangu
School of Epidemiology and Public Health, University of Ottawa
600 Peter Morand Crescent, Ottawa, ON K1G 5Z3
Ottawa, Canada
[email protected]