Does pay-for-performance design matter? Evidence from Brazil.

Pay-for-performance (P4P) schemes have been shown to have mixed effects on health care outcomes. A challenge in interpreting this evidence is that P4P is often considered a homogenous intervention, when in practice schemes vary widely in their design. Our study contributes to this literature by providing a detailed depiction of incentive design across municipalities within a national P4P scheme in Brazil (PMAQ) and exploring the association of alternative design typologies with the performance of primary health /care providers. We carried out a nation-wide survey of municipal health managers to characterize the scheme design, based on the size of the bonus, the providers incentivized, and the frequency of payment. Using OLS regressions and controlling for municipality characteristics, we examined whether each design feature was associated with better family health team performance. To capture potential interactions between design features, we used cluster analysis to group municipalities into five design typologies and then examined associations with quality of care. A majority of the municipalities included in our study used some of the PMAQ funds to provide bonuses to family health team workers, while the remaining municipalities spent the funds in the traditional way using input-based budgets. Frequent bonus payments (monthly) and higher size bonus allocations (share of 20-80%) were strongly associated with better team performance, while who within a team was eligible to receive bonuses did not in isolation appear to influence performance. The cluster analysis showed what combinations of design features were associated with better performance. The PMAQ score in the 'large bonus/many workers/high-frequency' cluster was 8.44 points higher than the 'no bonus' cluster, equivalent to a difference of 21.7% in the mean PMAQ score. Evidence from our study shows how design features can potentially influence health provider performance, informing the design of more effective P4P schemes.

Health facility assessment of small and sick newborn care in low- and middle-income countries: systematic tool development and operationalisation with NEST360 and UNICEF.

BACKGROUND: Each year an estimated 2.3 million newborns die in the first 28 days of life. Most of these deaths are preventable, and high-quality neonatal care is fundamental for surviving and thriving. Service readiness is used to assess the capacity of hospitals to provide care, but current health facility assessment (HFA) tools do not fully evaluate inpatient small and sick newborn care (SSNC). METHODS: Health systems ingredients for SSNC were identified from international guidelines, notably World Health Organization (WHO), and other standards for SSNC. Existing global and national service readiness tools were identified and mapped against this ingredients list. A novel HFA tool was co-designed according to a priori considerations determined by policymakers from four African governments, including that the HFA be completed in one day and assess readiness across the health system. The tool was reviewed by > 150 global experts, and refined and operationalised in 64 hospitals in Kenya, Malawi, Nigeria, and Tanzania between September 2019 and March 2021. RESULTS: Eight hundred and sixty-six key health systems ingredients for service readiness for inpatient SSNC were identified and mapped against four global and eight national tools measuring SSNC service readiness. Tools revealed major content gaps particularly for devices and consumables, care guidelines, and facility infrastructure, with a mean of 13.2% (n = 866, range 2.2-34.4%) of ingredients included. Two tools covered 32.7% and 34.4% (n = 866) of ingredients and were used as inputs for the new HFA tool, which included ten modules organised by adapted WHO health system building blocks, including: infrastructure, pharmacy and laboratory, medical devices and supplies, biomedical technician workshop, human resources, information systems, leadership and governance, family-centred care, and infection prevention and control. This HFA tool can be conducted at a hospital by seven assessors in one day and has been used in 64 hospitals in Kenya, Malawi, Nigeria, and Tanzania. CONCLUSION: This HFA tool is available open-access to adapt for use to comprehensively measure service readiness for level-2 SSNC, including respiratory support. The resulting facility-level data enable comparable tracking for Every Newborn Action Plan coverage target four within and between countries, identifying facility and national-level health systems gaps for action.

Measuring healthcare payor management practices in England.

Good management practice in healthcare payors and providers is considered central to ensuring health systems respond to population needs, contain costs, and improve both quality and outcomes. However, the evidence to support this assertion is sparce. While a quantitative link between better management practice and improved patient outcomes has been demonstrated for healthcare providers, no such link has been identified for healthcare payors. The lack of a robust tool to assess the management practices of healthcare payors has impeded such quantitative assessments. We report upon a novel tool developed to measure and assess 11 management practices in all 152 healthcare payors within England's National Health Service in 2010. We have tested the acceptability, reliability and validity of this tool using rigorous analytic methods and present four key findings. First, performance of the tool is strong and comparable to management practice scorecards used in other settings. Second, exploratory factor analysis indicates the tool measures two distinct latent factors of healthcare payor management practice with high internal consistency and reliability. Third, there is evidence of assessment and score validity. Fourth, payor management practice variations are associated with the degree of supervisory oversight. While deploying such a tool is challenging, these results suggest that healthcare payor management practices can be measured and assessed robustly. This could enable governments, and others, to identify how payor management practices influence health system performance and to estimate what health system performance improvements they should expect from interventions designed to improve the management practices of their local healthcare payors.

Management Practices and Quality of Care: Evidence from the Private Health Care Sector in Tanzania.

We measure the adoption of management practices in over 220 private for-profit and non-profit health facilities in 64 districts across Tanzania and link these data to process quality-of-care metrics, assessed using undercover standardised patients and clinical observations. We find that better managed health facilities are more likely to provide correct treatment in accordance with national treatment guidelines, adhere to a checklist of essential questions and examinations, and comply with infection prevention and control practices. Moving from the 10th to the 90th percentile in the management practice score is associated with a 48% increase in correct treatment. We then leverage a large-scale field experiment of an internationally recognised management support intervention in which health facilities are assessed against comprehensive standards, given an individually tailored quality improvement plan and supported through training and mentoring visits. We find zero to small effects on management scores, suggesting that improving management practices in this setting may be challenging.

Trading HIV for sheep: Risky sexual behavior and the response of female sex workers to Tabaski in Senegal.

We use a cohort of female sex workers (FSWs) in Senegal to show how large anticipated economic shocks lead to increased risky sexual behavior. Exploiting the exogenous timing of interviews, we study the effect of Tabaski, the most important Islamic festival celebrated in Senegal, in which most households purchase an expensive animal for sacrifice. Condom use, measured robustly via the list experiment, falls by between 27.3 percentage points (pp) (65.5%) and 43.1 pp (22.7%) in the 9 days before Tabaski, or a maximum of 49.5 pp (76%) in the 7 day period preceding Tabaski. The evidence suggests the economic pressures from Tabaski are key to driving the behavior change observed through the price premium for condomless sex. Those most exposed to the economic pressure from Tabaski were unlikely to be using condoms at all in the week before the festival. Our findings show that Tabaski leads to increased risky behaviors for FSWs, a key population at high risk of HIV infection, for at least 1 week every year and has implications for FSWs in all countries celebrating Tabaski or similar festivals. Because of the scale, frequency, and size of the behavioral response to shocks of this type, policy should be carefully designed to protect vulnerable women against anticipated shocks.

Investment case for small and sick newborn care in Tanzania: systematic analyses.

BACKGROUND: Small and sick newborn care (SSNC) is critical for national neonatal mortality reduction targets by 2030. Investment cases could inform implementation planning and enable coordinated resource mobilisation. We outline development of an investment case for Tanzania to estimate additional financing for scaling up SSNC to 80% of districts as part of health sector strategies to meet the country's targets. METHODS: We followed five steps: (1) reviewed national targets, policies and guidelines; (2) modelled potential health benefits by increased coverage of SSNC using the Lives Saved Tool; (3) estimated setup and running costs using the Neonatal Device Planning and Costing Tool, applying two scenarios: (A) all new neonatal units and devices with optimal staffing, and (B) half new and half modifying, upgrading, or adding resources to existing neonatal units; (4) calculated budget impact and return on investment (ROI) and (5) identified potential financing opportunities. RESULTS: Neonatal mortality rate was forecast to fall from 20 to 13 per 1000 live births with scale-up of SSNC, superseding the government 2025 target of 15, and close to the 2030 Sustainable Development Goal 3.2 target of <12. At 85% endline coverage, estimated cumulative lives saved were 36,600 by 2025 and 80,000 by 2030. Total incremental costs were estimated at US$166 million for scenario A (US$112 million set up and US$54 million for running costs) and US$90 million for scenario B (US$65 million setup and US$25 million for running costs). Setup costs were driven by infrastructure (83%) and running costs by human resources (60%). Cost per capita was US$0.93 and the ROI is estimated to be between US$8-12 for every dollar invested. CONCLUSIONS: ROI for SSNC is higher compared to other health investments, noting many deaths averted followed by full lifespan. This is conservative since disability averted is not included. Budget impact analysis estimated a required 2.3% increase in total government health expenditure per capita from US$40.62 in 2020, which is considered affordable, and the government has already allocated additional funding. Our proposed five-step SSNC investment case has potential for other countries wanting to accelerate progress.

Devices and furniture for small and sick newborn care: systematic development of a planning and costing tool.

BACKGROUND: High-quality neonatal care requires sufficient functional medical devices, furniture, fixtures, and use by trained healthcare workers, however there is lack of publicly available tools for quantification and costing. This paper describes development and use of a planning and costing tool regarding furniture, fixtures and devices to support scale-up of WHO level-2 neonatal care, for national and global newborn survival targets. METHODS: We followed a systematic process. First, we reviewed planning and costing tools of relevance. Second, we co-designed a new tool to estimate furniture and device set-up costs for a default 40-bed level-2 neonatal unit, incorporating input from multi-disciplinary experts and newborn care guidelines. Furniture and device lists were based off WHO guidelines/norms, UNICEF and national manuals/guides. Due to lack of evidence-based quantification, ratios were based on operational manuals, multi-country facility assessment data, and expert opinion. Default unit costs were from government procurement agency costs in Kenya, Nigeria, and Tanzania. Third, we refined the tool by national use in Tanzania. RESULTS: The tool adapts activity-based costing (ABC) to estimate quantities and costs to equip a level-2 neonatal unit based on three components: (1) furniture/fixtures (18 default but editable items); (2) neonatal medical devices (16 product categories with minimum specifications for use in low-resource settings); (3) user training at device installation. The tool was used in Tanzania to generate procurement lists and cost estimates for level-2 scale-up in 171 hospitals (146 District and 25 Regional Referral). Total incremental cost of all new furniture and equipment acquisition, installation, and user training were US$93,000 per District hospital (level-2 care) and US$346,000 per Regional Referral hospital. Estimated cost per capita for whole-country district coverage was US$0.23, representing 0.57% increase in government health expenditure per capita and additional 0.35% for all Regional Referral hospitals. CONCLUSION: Given 2.3 million neonatal deaths and potential impact of level-2 newborn care, rational and efficient planning of devices linked to systems change is foundational. In future iterations, we aim to include consumables, spare parts, and maintenance cost options. More rigorous implementation research data are crucial to formulating evidence-based ratios for devices numbers per baby. Use of this tool could help overcome gaps in devices numbers, advance efficiency and quality of neonatal care.

Determinants of government spending on primary healthcare: a global data analysis.

In 2018 global leaders renewed their political commitment to primary healthcare (PHC) ratifying the Declaration of Astana emphasising the importance of building a sustainable PHC system based on accessible and affordable delivery models strengthened by community empowerment. Yet, PHC often remains underfunded, of poor quality, unreliable and not accountable to users which further deprives PHC of funding. This paper analyses the determinants of PHC expenditure in 102 countries, and quantitatively tests the influence of a set of economic, social and political determinants of government expenditure on PHC. The analysis is focused on the determinants of PHC funding from government sources as the government is in a position to make decisions in relation to this expenditure as opposed to out-of-pocket spending which is not in their direct control. Multivariate regression analysis was done to determine statistically significant predictors.Our analysis found that some economic factors-namely Gross Domestic Product (GDP) per capita, government commitment to health and tax revenue raising capacity-were strongly associated with per capita government spending on PHC. We also found that control of corruption was strongly associated with the level of total spending on PHC, while voice and accountability were positively associated with greater government commitment to PHC as measured by government spending on PHC as a share of total government health spending.Our analysis takes a step towards understanding of the drivers of PHC expenditure beyond the level of national income. Some of these drivers may be beyond the remit of health policy decision makers and relate to broader governance arrangements and political forces in societies. Thus, efforts to prioritise PHC in the health agenda and increase PHC expenditure should recognise the constraints within the political landscapes and engage with a wide range of actors who influence decisions affecting the health sector.

Understanding why EmpaTeach did not reduce teachers' use of violence in Nyarugusu Refugee Camp: A quantitative process evaluation of a school-based violence prevention intervention.

EmpaTeach was the first intervention to address teacher violence to be tested in a humanitarian setting and the first to focus on reducing impulsive use of violence, but a cluster randomised trial found no evidence that the intervention was effective in reducing physical and emotional violence from teachers. We aimed to understand why. We conducted a quantitative process evaluation to describe the intervention implementation process (what was implemented and how); examine teachers' adoption of positive teaching practices (was the content of the intervention taken up by participants), and test mechanisms of impact underlying the program theory (how the intervention was supposed to produce change). Despite participation in the intervention activities and adoption of intervention-recommended strategies (classroom management and positive disciplinary methods), we show that teachers who used more positive discipline did not appear to use less violence; and teachers in intervention schools did not experience gains in intermediate outcomes such as empathy, growth mindset, self-efficacy or social support. Our findings suggest that the intervention did not work due to the failure of some key hypothesised mechanisms, rather than because of implementation challenges.

Does increased provider effort improve quality of care? Evidence from a standardised patient study on correct and unnecessary treatment.

BACKGROUND: Poor quality of care, including overprovision (unnecessary care) is a global health concern. Greater provider effort has been shown to increase the likelihood of correct treatment, but its relationship with overprovision is less clear. Providers who make more effort may give more treatment overall, both correct and unnecessary, or may have lower rates of overprovision; we test which is true in the Tanzanian private health sector. METHODS: Standardised patients visited 227 private-for-profit and faith-based facilities in Tanzania, presenting with symptoms of asthma and TB. They recorded history questions asked and physical examinations carried out by the provider, as well as laboratory tests ordered, treatments prescribed, and fees paid. A measure of provider effort was constructed on the basis of a checklist of recommended history taking questions and physical exams. RESULTS: 15% of SPs received the correct care for their condition and 74% received unnecessary care. Increased provider effort was associated with increased likelihood of correct care, and decreased likelihood of giving unnecessary care. Providers who made more effort charged higher fees, through the mechanism of higher consultation fees, rather than increased fees for lab tests and drugs. CONCLUSION: Providers who made more effort were more likely to treat patients correctly. A novel finding of this study is that they were also less likely to provide unnecessary care, suggesting it is not simply a case of some providers doing "more of everything", but that those who do more in the consultation give more targeted care.

Pay-for-performance for primary health care in Brazil: A comparison with England's Quality Outcomes Framework and lessons for the future.

Pay-for-performance (P4P) has been widely applied in OECD countries to improve the quality of both primary and secondary care, and is increasingly being implemented in low- and middle-income countries. In 2011, Brazil introduced one of the largest P4P schemes in the world, the National Programme for Improving Primary Care Access and Quality (PMAQ). We critically assess the design of PMAQ, drawing on a comparison with England's quality and outcome framework which, like PMAQ, was implemented at scale relatively rapidly within a nationalised health system. A key feature of PMAQ was that payment was based on the performance of primary care teams but rewards were given to municipalities, who had autonomy in how the funds could be used. This meant the incentives felt by family health teams were contingent on municipality decisions on whether to pass the funds on as bonuses and the basis upon which they allocated the funds between and within teams. Compared with England's P4P scheme, performance measurement under PMAQ focused more on structural rather than process quality of care, relied on many more indicators, and was less regular. While PMAQ represented an important new funding stream for primary health care, our review suggests that theoretical incentives generated were unclear and could have been better structured to direct health providers towards improvements in quality of care.

Supply, then demand? Health expenditure, political leanings, cost obstacles to care, and vaccine hesitancy predict state-level COVID-19 vaccination rates.

OBJECTIVES: To examine predictors of state-level COVID-19 vaccination rates during the first nine months of 2021. METHODS: Using publicly available data, we employ a robust, iteratively re-weighted least squares multivariable regression with state characteristics as the independent variables and vaccinations per capita as the outcome. We run this regression for each day between February 1 and September 21, the last day before vaccine booster rollout. RESULTS: We identify associations between vaccination rates and several state characteristics, including health expenditure, vaccine hesitancy, cost obstacles to care, Democratic voting, and elderly population share. We show that the determinants of vaccination rates have evolved: while supply-side factors were most clearly associated with early vaccination uptake, demand-side factors have become increasingly salient over time. We find that our results are generally robust to a range of alternative specifications. CONCLUSIONS: Both supply and demand-side factors relate to vaccination coverage and the determinants of success have changed over time. POLICY IMPLICATIONS: Investing in health capacity may improve early vaccine distribution and administration, while overcoming vaccine hesitancy and cost obstacles to care may be crucial for later immunisation campaign stages.

School violence, depression symptoms, and school climate: a cross-sectional study of Congolese and Burundian refugee children.

Forcibly displaced children are at increased risk of violence and mental health disorders. In refugee contexts, schools are generally perceived as protective environments where children can build a sense of belonging and recover from trauma. Evidence shows that positive school climates can support student skills development and socio-emotional wellbeing and protect them against a host of adverse outcomes. However, schools are also places where children may experience violence, from both teachers and peers. Prevalence estimates of violence against children in humanitarian settings are scarce and evidence on the relationship between school climate and student outcomes in these contexts is non-existent. The aim of the study is to estimate the prevalence of school-based violence against children and to explore the association between perceptions of school climate and students' experiences and use of violence and their depression symptoms. We relied on data from a cross-sectional survey of students and teachers in all primary and secondary schools in Nyarugusu Refugee Camp in Tanzania, conducted as part of a cluster randomised controlled trial, to compute prevalence estimates and used mixed logistic regression analysis to assess the association between school climate and students' outcomes. We found that students in Nyarugusu experienced high levels of violence from both peers and teachers in both primary and secondary schools in the camp, with little difference between boys and girls. Nearly one in ten students screened positive for symptoms of depression. We found that opportunities for students and teachers to be involved in decision-making were associated with higher odds of violent discipline and teachers' self-efficacy was a significant protective factor against student depression symptoms. However, generally, school-level perceptions of school climate were not associated with student outcomes after adjusting for potential confounders. Our findings suggest that interventions to prevent and respond to teacher and peer violence in schools and to support students' mental health are urgently needed. Our results challenge the assumption that education environments are inherently protective for children and call for further investigation of norms around violence among students and teachers to better understand the role of school climate in refugee settings.

Performance bonuses and the quality of primary health care delivered by family health teams in Brazil: A difference-in-differences analysis.

BACKGROUND: Pay-for-performance (P4P) programmes to incentivise health providers to improve quality of care have been widely implemented globally. Despite intuitive appeal, evidence on the effectiveness of P4P is mixed, potentially due to differences in how schemes are designed. We exploited municipality variation in the design features of Brazil's National Programme for Improving Primary Care Access and Quality (PMAQ) to examine whether performance bonuses given to family health team workers were associated with changes in the quality of care and whether the size of bonus mattered. METHODS AND FINDINGS: For this quasi-experimental study, we used a difference-in-differences approach combined with matching. We compared changes over time in the quality of care delivered by family health teams between (bonus) municipalities that chose to use some or all of the PMAQ money to provide performance-related bonuses to team workers with (nonbonus) municipalities that invested the funds using traditional input-based budgets. The primary outcome was the PMAQ score, a quality of care index on a scale of 0 to 100, based on several hundred indicators (ranging from 598 to 660) of health care delivery. We did one-to-one matching of bonus municipalities to nonbonus municipalities based on baseline demographic and economic characteristics. On the matched sample, we used ordinary least squares regression to estimate the association of any bonus and size of bonus with the prepost change over time (between November 2011 and October 2015) in the PMAQ score. We performed subgroup analyses with respect to the local area income of the family health team. The matched analytical sample comprised 2,346 municipalities (1,173 nonbonus municipalities; 1,173 bonus municipalities), containing 10,275 family health teams that participated in PMAQ from the outset. Bonus municipalities were associated with a 4.6 (95% CI: 2.7 to 6.4; p < 0.001) percentage point increase in the PMAQ score compared with nonbonus municipalities. The association with quality of care increased with the size of bonus: the largest bonus group saw an improvement of 8.2 percentage points (95% CI: 6.2 to 10.2; p < 0.001) compared with the control. The subgroup analysis showed that the observed improvement in performance was most pronounced in the poorest two-fifths of localities. The limitations of the study include the potential for bias from unmeasured time-varying confounding and the fact that the PMAQ score has not been validated as a measure of quality of care. CONCLUSIONS: Performance bonuses to family health team workers compared with traditional input-based budgets were associated with an improvement in the quality of care.

Pushy Patients Or Pushy Providers? Effect Of Patient Knowledge On Antibiotic Prescribing In Tanzania.

Antimicrobial resistance is one of the most serious threats to global health, but little progress has been made in reversing its spread. Inappropriate use of antibiotics in humans is a major driver of antimicrobial resistance, and rates are high and growing in lower- and middle-income countries. Antibiotics are thought to be subject to supplier-induced demand, whereby providers prescribe them to patients who do not know they are unnecessary. We conducted a randomized field experiment in 227 private health facilities in Tanzania, with standardized patients presenting uncomplicated upper respiratory tract infection symptoms. Standardized patients were randomly assigned to express knowledge (informed) or not (uninformed) that antibiotics were not required to treat them. There was a very high rate of inappropriate antibiotic prescription, with 86.0 percent of informed standardized patients and 94.8 percent of uninformed standardized patients prescribed an antibiotic, for an adjusted difference of 7.8 percentage points between the groups. This small effect suggests that broader health systems factors are at play and that interventions should be aimed at systems, health facilities, and providers.

Poor quality for the poor? A study of inequalities in service readiness and provider knowledge in Indonesian primary health care facilities.

BACKGROUND: For many low and middle-income countries poor quality health care is now responsible for a greater number of deaths than insufficient access to care. This has in turn raised concerns around the distribution of quality of care in LMICs: do the poor have access to lower quality health care compared to the rich? The aim of this study is to investigate the extent of inequalities in the availability of quality health services across the Indonesian health system with a particular focus on differences between care delivered in the public and private sectors. METHODS: Using the Indonesian Family Life Survey (wave 5, 2015), 15,877 households in 312 communities were linked with a representative sample of both public and private health facilities available in the same communities. Quality of health facilities was assessed using both a facility service readiness score and a knowledge score constructed using clinical vignettes. Ordinary least squares regression models were used to investigate the determinants of quality in public and private health facilities. RESULTS: In both sectors, inequalities in both quality scores existed between major islands. In public facilities, inequalities in readiness scores persisted between rural and urban areas, and to a lesser extent between rich and poor communities. CONCLUSION: In order to reach the ambitious stated goal of reaching Universal Health Coverage in Indonesia, priority should be given to redressing current inequalities in the quality of care.

Effect of a multifaceted intervention to improve clinical quality of care through stepwise certification (SafeCare) in health-care facilities in Tanzania: a cluster-randomised controlled trial.

BACKGROUND: Quality of care is consistently shown to be inadequate in health-care settings in many low-income and middle-income countries, including in private facilities, which are rapidly growing in number but often do not have effective quality stewardship mechanisms. The SafeCare programme aims to address this gap in quality of care, using a standards-based approach adapted to low-resource settings, involving assessments, mentoring, training, and access to loans, to improve clinical quality and facility business performance. We assessed the effect of the SafeCare programme on quality of patient care in faith-based and private for-profit facilities in Tanzania. METHODS: In this cluster-randomised controlled trial, health facilities were eligible if they were dispensaries, health centres, or hospitals in the faith-based or private for-profit sectors in Tanzania. We randomly assigned facilities (1:1) using computer-generated stratified randomisation to receive the full SafeCare package (intervention) or an assessment only (control). Implementing staff and participants were masked to outcome measurement and the primary outcomes were measured by fieldworkers who had no knowledge of the study group allocation. The primary outcomes were health worker compliance with infection prevention and control (IPC) practices as measured by observation of provider-patient interactions, and correct case management of undercover standardised patients at endline (after a minimum of 18 months). Analyses were by modified intention to treat. The trial is registered with ISRCTN, ISRCTN93644888. FINDINGS: Between March 7 and Nov 30, 2016, we enrolled and randomly assigned 237 health facilities to the intervention (n=118) or control (n=119). Nine facilities (seven intervention facilities and two control facilities) closed during the trial and were not included in the analysis. We observed 29 608 IPC indications in 5425 provider-patient interactions between Feb 7 and April 5, 2018. Health facilities received visits from 909 standardised patients between May 3 and June 12, 2018. Intervention facilities had a 4·4 percentage point (95% CI 0·9-7·7; p=0.015) higher mean SafeCare standards assessment score at endline than control facilities. However, there was no evidence of a difference in clinical quality between intervention and control groups at endline. Compliance with IPC practices was observed in 8181 (56·9%) of 14 366 indications in intervention facilities and 8336 (54·7%) of 15 242 indications in control facilities (absolute difference 2·2 percentage points, 95% CI -0·2 to -4·7; p=0·071). Correct management occurred in 120 (27·0%) of 444 standardised patients in the intervention group and in 136 (29·2%) of 465 in the control group (absolute difference -2·8 percentage points, 95% CI -8·6 to -3·1; p=0·36). INTERPRETATION: SafeCare did not improve clinical quality as assessed by compliance with IPC practices and correct case management. The absence of effect on clinical quality could reflect a combination of insufficient intervention intensity, insufficient links between structural quality and care processes, scarcity of resources for quality improvement, and inadequate financial and regulatory incentives for improvement. FUNDING: UK Health Systems Research Initiative (Medical Research Council, Economic and Social Research Council, UK Department for International Development, Global Challenges Research Fund, and Wellcome Trust).