Data fusion for predicting long-term program impacts.

Policymakers often require information on programs' long-term impacts that is not available when decisions are made. For example, while rigorous evidence from the Oregon Health Insurance Experiment (OHIE) shows that having health insurance influences short-term health and financial measures, the impact on long-term outcomes, such as mortality, will not be known for many years following the program's implementation. We demonstrate how data fusion methods may be used address the problem of missing final outcomes and predict long-run impacts of interventions before the requisite data are available. We implement this method by concatenating data on an intervention (such as the OHIE) with auxiliary long-term data and then imputing missing long-term outcomes using short-term surrogate outcomes while approximating uncertainty with replication methods. We use simulations to examine the performance of the methodology and apply the method in a case study. Specifically, we fuse data on the OHIE with data from the National Longitudinal Mortality Study and estimate that being eligible to apply for subsidized health insurance will lead to a statistically significant improvement in long-term mortality.

How personnel diversity and affective bonds affect performance-based financing: a moderator analysis of a difference-in-difference estimator.

To spur improvement in health-care service quality and quantity, performance-based financing (PBF) is an increasingly common policy tool, especially in low- and middle-income countries. This study examines how personnel diversity and affective bonds in primary care clinics affect their ability to improve care quality in PBF arrangements. Leveraging data from a large-scale matched PBF intervention in Tajikistan including 208 primary care clinics, we examined how measures of personnel diversity (position and tenure variety) and affective bonds (mutual support and group pride) were associated with changes in the level and variability of clinical knowledge (diagnostic accuracy of 878 clinical vignettes) and care processes (completion of checklist items in 2485 instances of direct observations). We interacted the explanatory variables with exposure to PBF in cluster-robust, linear regressions to assess how these explanatory variables moderated the PBF treatment's association with clinical knowledge and care process improvements. Providers and facilities with higher group pride exhibited higher care process improvement (greater checklist item completion and lower variability of items completed). Personnel diversity and mutual support showed little significant associations with the outcomes. Organizational features of clinics exposed to PBF may help explain variation in outcomes and warrant further research and intervention in practice to identify and test opportunities to leverage them. Group pride may strengthen clinics' ability to improve care quality in PBF arrangements. Improving health-care facilities' pride may be an affordable and effective way to enhance health-care organization adaptation.

Tracking health system performance in times of crisis using routine health data: lessons learned from a multicountry consortium.

COVID-19 has prompted the use of readily available administrative data to track health system performance in times of crisis and to monitor disruptions in essential healthcare services. In this commentary we describe our experience working with these data and lessons learned across countries. Since April 2020, the Quality Evidence for Health System Transformation (QuEST) network has used administrative data and routine health information systems (RHIS) to assess health system performance during COVID-19 in Chile, Ethiopia, Ghana, Haiti, Lao People's Democratic Republic, Mexico, Nepal, South Africa, Republic of Korea and Thailand. We compiled a large set of indicators related to common health conditions for the purpose of multicountry comparisons. The study compiled 73 indicators. A total of 43% of the indicators compiled pertained to reproductive, maternal, newborn and child health (RMNCH). Only 12% of the indicators were related to hypertension, diabetes or cancer care. We also found few indicators related to mental health services and outcomes within these data systems. Moreover, 72% of the indicators compiled were related to volume of services delivered, 18% to health outcomes and only 10% to the quality of processes of care. While several datasets were complete or near-complete censuses of all health facilities in the country, others excluded some facility types or population groups. In some countries, RHIS did not capture services delivered through non-visit or nonconventional care during COVID-19, such as telemedicine. We propose the following recommendations to improve the analysis of administrative and RHIS data to track health system performance in times of crisis: ensure the scope of health conditions covered is aligned with the burden of disease, increase the number of indicators related to quality of care and health outcomes; incorporate data on nonconventional care such as telehealth; continue improving data quality and expand reporting from private sector facilities; move towards collecting patient-level data through electronic health records to facilitate quality-of-care assessment and equity analyses; implement more resilient and standardized health information technologies; reduce delays and loosen restrictions for researchers to access the data; complement routine data with patient-reported data; and employ mixed methods to better understand the underlying causes of service disruptions.

Texts don't nudge: An adaptive trial to prevent the spread of COVID-19 in India.

We conduct an adaptive randomized controlled trial to evaluate the impact of a SMS-based information campaign on the adoption of social distancing and handwashing in rural Bihar, India, six months into the COVID-19 pandemic. We test 10 arms that vary in delivery timing and message framing, changing content to highlight gains or losses for either one's own family or community. We identify the optimal treatment separately for each targeted behavior by adaptively allocating shares across arms over 10 experimental rounds using exploration sampling. Based on phone surveys with nearly 4,000 households and using several elicitation methods, we do not find evidence of impact on knowledge or adoption of preventive health behavior, and our confidence intervals cannot rule out positive effects as large as 5.5 percentage points, or 16%. Our results suggest that SMS-based information campaigns may have limited efficacy after the initial phase of a pandemic.

Knowledge as a Predictor of Insurance Coverage Under the Affordable Care Act.

BACKGROUND: The Affordable Care Act established policy mechanisms to increase health insurance coverage in the United States. While insurance coverage has increased, 10%-15% of the US population remains uninsured. OBJECTIVES: To assess whether health insurance literacy and financial literacy predict being uninsured, covered by Medicaid, or covered by Marketplace insurance, holding demographic characteristics, attitudes toward risk, and political affiliation constant. RESEARCH DESIGN: Analysis of longitudinal data from fall 2013 and spring 2015 including financial and health insurance literacy and key covariates collected in 2013. SUBJECTS: A total of 2742 US residents ages 18-64, 525 uninsured in fall 2013, participating in the RAND American Life Panel, a nationally representative internet panel. MEASURES: Self-reported health insurance status and type as of spring 2015. RESULTS: Among the uninsured in 2013, higher financial and health insurance literacy were associated with greater probability of being insured in 2015. For a typical uninsured individual in 2013, the probability of being insured in 2015 was 8.3 percentage points higher with high compared with low financial literacy, and 9.2 percentage points higher with high compared with low health insurance literacy. For the general population, those with high financial and health insurance literacy were more likely to obtain insurance through Medicaid or the Marketplaces compared with being uninsured. The magnitude of coefficients for these predictors was similar to that of commonly used demographic covariates. CONCLUSIONS: A lack of understanding about health insurance concepts and financial illiteracy predict who remains uninsured. Outreach and consumer-education programs should consider these characteristics.

Preparedness of Americans for the Affordable Care Act.

This paper investigates whether individuals are sufficiently informed to make reasonable choices in the health insurance exchanges established by the Affordable Care Act (ACA). We document knowledge of health reform, health insurance literacy, and expected changes in healthcare using a nationally representative survey of the US population in the 5 wk before the introduction of the exchanges, with special attention to subgroups most likely to be affected by the ACA. Results suggest that a substantial share of the population is unprepared to navigate the new exchanges. One-half of the respondents did not know about the exchanges, and 42% could not correctly describe a deductible. Those earning 100-250% of federal poverty level (FPL) correctly answered, on average, 4 out of 11 questions about health reform and 4.6 out of 7 questions about health insurance. This compares with 6.1 and 5.9 correct answers, respectively, for those in the top income category (400% of FPL or more). Even after controlling for potential confounders, a low-income person is 31% less likely to score above the median on ACA knowledge questions, and 54% less likely to score above the median on health insurance knowledge than a person in the top income category. Uninsured respondents scored lower on health insurance knowledge, but their knowledge of ACA is similar to the overall population. We propose that simplified options, decision aids, and health insurance product design to address the limited understanding of health insurance contracts will be crucial for ACA's success.

Perceptions of e-cigarettes: a comparison of adult smokers and non-smokers in a Mechanical Turk sample.

BACKGROUND: Given plans to extend its regulatory authority to e-cigarettes, the Food and Drug Administration (FDA) urgently needs to understand how e-cigarettes are perceived by the public. OBJECTIVES: To examine how smoking status impacts adult perceptions and expectations of e-cigarettes. METHODS: We used Mechanical Turk (MTurk), a "crowdsourcing" platform, to rapidly survey a large (n = 796; female = 381; male = 415), diverse sample of adult ever (44%) and never smokers (56%), including ever (28%) and never (72%) users of e-cigarettes. RESULTS: Smokers and non-smokers learned about e-cigarettes primarily through the internet and conversations with others. Ever smokers were more likely than never smokers, and female current smokers were more likely than female former smokers, to have learned about e-cigarettes from point of sale advertising (p's < 0.05) and to believe that e-cigarettes help smokers quit (ps < 0.05). Among never users of e-cigarettes, current smokers were more likely than never smokers and former smokers to report that they would try e-cigarettes in the future (ps < 0.01). Current smokers' top reason for wanting to try e-cigarettes was to quit or reduce smoking (56%), while never and former smokers listed curiosity. In contrast, female current smokers' top reason for not trying e-cigarettes was health and safety concerns (44%) while males were deterred by expense (44%). CONCLUSIONS: Adult smokers and non-smokers have different perceptions and expectations of e-cigarettes. Public health messages regarding e-cigarettes may need to be tailored separately for persons with and without a history of using conventional cigarettes. Tailoring messages by gender within smoker groups may also improve their impact.

Policies related to opioid agonist therapy for opioid use disorders: The evolution of state policies from 2004 to 2013.

BACKGROUND: State Medicaid policies play an important role in Medicaid enrollees' access to and use of opioid agonists, such as methadone and buprenorphine, in the treatment of opioid use disorders. Little information is available, however, regarding the evolution of state policies facilitating or hindering access to opioid agonists among Medicaid enrollees. METHODS: During 2013-2014, we surveyed state Medicaid officials and other designated state substance abuse treatment specialists about their state's recent history of Medicaid coverage and policies pertaining to methadone and buprenorphine. We describe the evolution of such coverage and policies and present an overview of the Medicaid policy environment with respect to opioid agonist therapy from 2004 to 2013. RESULTS: Among our sample of 45 states with information on buprenorphine and methadone coverage, we found a gradual trend toward adoption of coverage for opioid agonist therapies in state Medicaid agencies. In 2013, only 11% of states in our sample (n = 5) had Medicaid policies that excluded coverage for methadone and buprenorphine, whereas 71% (n = 32) had adopted or maintained policies to cover both buprenorphine and methadone among Medicaid enrollees. We also noted an increase in policies over the time period that may have hindered access to buprenorphine and/or methadone. CONCLUSIONS: There appears to be a trend for states to enact policies increasing Medicaid coverage of opioid agonist therapies, while in recent years also enacting policies, such as prior authorization requirements, that potentially serve as barriers to opioid agonist therapy utilization. Greater empirical information about the potential benefits and potential unintended consequences of such policies can provide policymakers and others with a more informed understanding of their policy decisions.

Where Is Buprenorphine Dispensed to Treat Opioid Use Disorders? The Role of Private Offices, Opioid Treatment Programs, and Substance Abuse Treatment Facilities in Urban and Rural Counties.

POLICY POINTS: Buprenorphine is an effective opioid dependence treatment that has expanded access to care since its 2002 approval, but it can only be prescribed by physicians waivered to treat a limited number of individuals. We examined the impact of 2006 legislation that increased waivered physician patient limits from 30 to 100 on buprenorphine use, and found that 100-patient-waivered physicians were significantly associated with growth in buprenorphine use, with no such relationship for 30-patient-waivered physicians. Policies relaxing patient limits may be more effective in increasing buprenorphine use than alternatives such as opening new substance abuse treatment facilities or increasing the overall number of waivered physicians. CONTEXT: Opioid use disorders are a significant public health problem. In 2002, the FDA approved buprenorphine as an opioid use disorder treatment when prescribed by waivered physicians who were limited to treating 30 patients at a time. In 2006, federal legislation raised this number to 100 patients. Although federal legislators are considering increasing these limits further and expanding prescribing privileges to nonphysicians, little information is available regarding the impact of such changes on buprenorphine use. We therefore examined the impact of the 2006 legislation-as well as the association between urban and rural waivered physicians, opioid treatment programs, and substance abuse treatment facilities-on buprenorphine distributed per capita over the past decade. METHODS: Using 2004-2011 state-level data on buprenorphine dispensed and county-level data on the number of buprenorphine-waivered physicians and substance abuse treatment facilities using buprenorphine, we estimated a multivariate ordinary least squares regression model with state fixed effects of a state's annual total buprenorphine dispensed per capita as a function of the state's number of buprenorphine providers. FINDINGS: The amount of buprenorphine dispensed has been increasing at a greater rate than the number of buprenorphine providers. The number of physicians waivered to treat 100 patients with buprenorphine in both rural and urban settings was significantly associated with increased amounts of buprenorphine dispensed per capita. There was no significant association in the growth of buprenorphine distributed and the number of physicians with 30-patient waivers. CONCLUSIONS: The greater amounts of buprenorphine dispensed are consistent with the potentially greater use of opioid agonists for opioid use disorder treatment, though they also make their misuse more likely. The changes after the 2006 legislation suggest that policies focused on increasing the number of patients that a single waivered physician could safely and effectively treat could be more effective in increasing buprenorphine use than would alternatives such as opening new substance abuse treatment facilities or raising the overall number of waivered physicians.

Effect of Chiranjeevi Yojana on institutional deliveries and neonatal and maternal outcomes in Gujarat, India: a difference-in-differences analysis.

OBJECTIVE: To evaluate the effect of the Chiranjeevi Yojana programme, a public-private partnership to improve maternal and neonatal health in Gujarat, India. METHODS: A household survey (n = 5597 households) was conducted in Gujarat to collect retrospective data on births within the preceding 5 years. In an observational study using a difference-in-differences design, the relationship between the Chiranjeevi Yojana programme and the probability of delivery in health-care institutions, the probability of obstetric complications and mean household expenditure for deliveries was subsequently examined. In multivariate regressions, individual and household characteristics as well as district and year fixed effects were controlled for. Data from the most recent District Level Household and Facility Survey (DLHS-3) wave conducted in Gujarat (n = 6484 households) were used in parallel analyses. FINDINGS: Between 2005 and 2010, the Chiranjeevi Yojana programme was not associated with a statistically significant change in the probability of institutional delivery (2.42 percentage points; 95% confidence interval, CI: -5.90 to 10.74) or of birth-related complications (6.16 percentage points; 95% CI: -2.63 to 14.95). Estimates using DLHS-3 data were similar. Analyses of household expenditures indicated that mean household expenditure for private-sector deliveries had either not fallen or had fallen very little under the Chiranjeevi Yojana programme. CONCLUSION: The Chiranjeevi Yojana programme appears to have had no significant impact on institutional delivery rates or maternal health outcomes. The absence of estimated reductions in household spending for private-sector deliveries deserves further study.

Impact of the influx of Syrian refugees on domestic violence against Jordanian women: Evidence from the 2017-18 Jordan Population and Family Health Survey.

The 2011 Syrian crisis led to a large influx of refugees into neighboring countries, including Jordan. The resulting stress on local host communities could heighten the risk of domestic violence against Jordanian women. We utilized multilevel propensity score weighting and data from the 2017-18 Jordan Population and Family Health Survey to empirically test for differences in outcomes related to domestic violence, marital control, and justification of wife-beating between Jordanian communities with varying density levels of Syrian women. We did not find systematic differences in these outcomes across communities. However, we cannot rule out effects that may not be statistically detectable with our sample but could still be substantively meaningful.

Performance of predictive algorithms in estimating the risk of being a zero-dose child in India, Mali and Nigeria.

INTRODUCTION: Many children in low-income and middle-income countries fail to receive any routine vaccinations. There is little evidence on how to effectively and efficiently identify and target such 'zero-dose' (ZD) children. METHODS: We examined how well predictive algorithms can characterise a child's risk of being ZD based on predictor variables that are available in routine administrative data. We applied supervised learning algorithms with three increasingly rich sets of predictors and multiple years of data from India, Mali and Nigeria. We assessed performance based on specificity, sensitivity and the F1 Score and investigated feature importance. We also examined how performance decays when the model is trained on older data. For data from India in 2015, we further compared the inclusion and exclusion errors of the algorithmic approach with a simple geographical targeting approach based on district full-immunisation coverage. RESULTS: Cost-sensitive Ridge classification correctly classifies most ZD children as being at high risk in most country-years (high specificity). Performance did not meaningfully increase when predictors were added beyond an initial sparse set of seven variables. Region and measures of contact with the health system (antenatal care and birth in a facility) had the highest feature importance. Model performance decreased in the time between the data on which the model was trained and the data to which it was applied (test data). The exclusion error of the algorithmic approach was about 9.1% lower than the exclusion error of the geographical approach. Furthermore, the algorithmic approach was able to detect ZD children across 176 more areas as compared with the geographical rule, for the same number of children targeted. INTERPRETATION: Predictive algorithms applied to existing data can effectively identify ZD children and could be deployed at low cost to target interventions to reduce ZD prevalence and inequities in vaccination coverage.

Disruption Of Non-COVID-19 Health Care In Latin America During The Pandemic: Effects On Health, Lessons For Policy.

COVID-19 had severe direct and indirect effects on health and well-being in Latin America. To understand the extent to which disruptions among non-COVID-19-related health services affected population health, we used administrative data from the period 2015-21 to examine public hospital discharges and mortality for conditions amenable to health care in four Latin American countries: Brazil, Ecuador, Mexico, and Peru. Between March 2020 and December 2021, hospitalization rates for these conditions declined by 28 percent and mortality rates increased by 15 percent relative to prepandemic years. Noncommunicable diseases accounted for 89 percent of this rise in mortality. The poorest states in each country experienced relatively larger increases in mortality. Our results, which focus on the health effects of service disruption, suggest that maintaining health care services in this region during the pandemic could have avoided at least 96,000 deaths. Policies should focus on maintaining essential health care services during emergencies, particularly for patients with noncommunicable diseases, and on minimizing negative consequences by ensuring coordinated and continuous care; leveraging alternative modalities of care, such as telemedicine; broadening the role of nonphysician health care workers; and expanding options for medication delivery.

Lives Versus Livelihoods: The Epidemiological, Social, And Economic Impact Of COVID-19 In Latin America And The Caribbean.

During the COVID-19 pandemic, Latin American and Caribbean countries implemented stringent public health and social measures that disrupted economic and social activities. This study used an integrated model to evaluate the epidemiological, economic, and social trade-offs in Argentina, Brazil, Jamaica, and Mexico throughout 2021. Argentina and Mexico displayed a higher gross domestic product (GDP) loss and lower deaths per million compared with Brazil. The magnitude of the trade-offs differed across countries. Reducing GDP loss at the margin by 1 percent would have increased daily deaths by 0.5 per million in Argentina but only 0.3 per million in Brazil. We observed an increase in poverty rates related to the stringency of public health and social measures but no significant income-loss differences by sex. Our results indicate that the economic impact of COVID-19 was uneven across countries as a result of different pandemic trajectories, public health and social measures, and vaccination uptake, as well as socioeconomic differences and fiscal responses. Policy makers need to be informed about the trade-offs to make strategic decisions to save lives and livelihoods.

Role of User Benefit Awareness in Health Coverage Utilization among the Poor in Cambodia.

The objective of this study was to understand the steps to health coverage benefit utilization in Cambodia toward improving access to health care and financial risk protection for the poor. We particularly examine the role of user awareness in the pathway to care seeking and benefit utilization with respect to the Health Equity Funds (HEF). Using 2016 survey data that were nationally representative of households with children under two years of age, we used a series of logistic regression models to evaluate associations between respondents' awareness of benefits, public health care seeking behaviors, coverage benefit claims, and out-of-pocket expenditures. Beneficiaries were generally aware of their entitlements, although their awareness of specific benefits, such as transport reimbursement, was relatively lower. Awareness of free services at public health centers was associated with twice the odds of having ever visited a public provider for outpatient care, while awareness of free services at public hospitals was associated with higher odds of always seeking inpatient care in the public sector. Study findings point to the decision of where to seek care as the critical point in the pathway to HEF utilization. If the decision had already been made to go to a public provider, it was likely that HEF benefits were claimed. Interventions that prompt appropriate care seeking in the public sector may do the most to improve HEF utilization and subsequently improve access to care through sufficient financial risk protection.

The Quality of Primary Care in Cambodia: An Assessment of Knowledge and Effort of Public Sector Maternal and Child Care Providers.

Improving the quality of primary care is essential for achieving universal health coverage in low- and middle-income countries. This study examined the level and variation in primary care provider knowledge and effort in Cambodia, using cross-sectional data collected in 2014-2015 from public sector health centers in nine provinces. The data included clinical vignettes and direct observations of processes of antenatal care, postnatal care, and well-child visits and covered between 290-495 health centers and 370-847 individual providers for each service and type of data. The results indicate that provider knowledge and observed effort were generally low and varied across health centers and across individual providers. In addition, providers' effort scores were generally lower than their knowledge scores, indicating the presence of a "know-do gap." Although higher provider knowledge was correlated with higher levels of effort during patient encounters, knowledge only explained a limited fraction of the provider-level variation in effort. Due to low baseline performance and the know-do gap, improving provider adherence to clinical guidelines through training and practice standardization alone may have limited impact. Overall, the findings suggest that raising the low quality of care provided by Cambodia's public sector will require multidimensional interventions that involve training, strategies that increase provider motivation, and improved health center management. The authors reported there is no funding associated with the work presented in this article.

Effect of the suspension of the J&J COVID-19 vaccine on vaccine hesitancy in the United States.

On April 13, 2021, U.S. authorities announced an investigation into potential adverse events associated with the Johnson & Johnson (Janssen, J&J) COVID-19 vaccine and recommended "a pause in the use of this vaccine out of an abundance of caution." We examined whether public attitudes toward COVID-19 vaccination shifted after this recommended suspension using an interrupted time series with data from the Census Bureau's Household Pulse Survey, which was fielded bi-weekly between January 6 and April 26, 2021. We found no significant changes in trends of the proportion of the U.S. adult population hesitant about getting a COVID-19 vaccine, but a significant increase in concerns about safety and efficacy of COVID-19 vaccines among the already hesitant population.

The role of teams in shaping quality of obstetrical care: a cross-sectional study in Dire Dawa, Ethiopia.

OBJECTIVES: To examine how characteristics of clinical colleagues influence quality of care. DESIGN: We conducted a cross-sectional observational study examining the associations between quality of care and a provider's coworkers, controlling for individual provider's characteristics and contextual factors. SETTING: Nine health facilities in Dire Dawa Administration, Ethiopia, from December 2020 to February 2021. PARTICIPANTS: 824 clients and 95 unique providers were observed across the 9 health facilities. OUTCOME MEASURES: We examine the quality of processes of intrapartum and immediate postpartum care during five phases of the delivery (first examination, first stage of labour, third stage of labour, immediate newborn care and immediate maternal postpartum care). RESULTS: For the average client, 50% of the recommended routine clinical actions were completed during the delivery overall, with immediate maternal postpartum care being the least well performed (17% of recommended actions). Multiple healthcare providers were involved in 55% of deliveries. The number of providers contributing to a delivery was unassociated with the quality of care, but a one standard deviation increase in the coworker's performance was associated with a 2% point increase in quality of care (p<0.01); this association was largest among providers in the middle quartiles of performance. CONCLUSIONS: A provider's typical performance had a modest positive association with quality of delivery care given by their coworker. As delivery care is often provided by multiple healthcare providers, examining the dynamics of how they influence one another can provide important insights for quality improvement.

Effect of lifting COVID-19 restrictions on utilisation of primary care services in Nepal: a difference-in-differences analysis.

INTRODUCTION: An increasing number of studies have reported disruptions in health service utilisation due to the COVID-19 pandemic and its associated restrictions. However, little is known about the effect of lifting COVID-19 restrictions on health service utilisation. The objective of this study was to estimate the effect of lifting COVID-19 restrictions on primary care service utilisation in Nepal. METHODS: Data on utilisation of 10 primary care services were extracted from the Health Management Information System across all health facilities in Nepal. We used a difference-in-differences design and linear fixed effects regressions to estimate the effect of lifting COVID-19 restrictions. The treatment group included palikas that had lifted restrictions in place from 17 August 2020 to 16 September 2020 (Bhadra 2077) and the control group included palikas that had maintained restrictions during that period. The pre-period included the 4 months of national lockdown from 24 March 2020 to 22 July 2020 (Chaitra 2076 to Ashar 2077). Models included month and palika fixed effects and controlled for COVID-19 incidence. RESULTS: We found that lifting COVID-19 restrictions was associated with an average increase per palika of 57.5 contraceptive users (95% CI 14.6 to 100.5), 15.6 antenatal care visits (95% CI 5.3 to 25.9) and 1.6 child pneumonia visits (95% CI 0.2 to 2.9). This corresponded to a 9.4% increase in contraceptive users, 34.2% increase in antenatal care visits and 15.6% increase in child pneumonia visits. Utilisation of most other primary care services also increased after lifting restrictions, but coefficients were not statistically significant. CONCLUSIONS: Despite the ongoing pandemic, lifting restrictions can lead to an increase in some primary care services. Our results point to a causal link between restrictions and health service utilisation and call for policy makers in low- and middle-income countries to carefully consider the trade-offs of strict lockdowns during future COVID-19 waves or future pandemics.

COVID-19 and resilience of healthcare systems in ten countries.

Declines in health service use during the Coronavirus Disease 2019 (COVID-19) pandemic could have important effects on population health. In this study, we used an interrupted time series design to assess the immediate effect of the pandemic on 31 health services in two low-income (Ethiopia and Haiti), six middle-income (Ghana, Lao People's Democratic Republic, Mexico, Nepal, South Africa and Thailand) and high-income (Chile and South Korea) countries. Despite efforts to maintain health services, disruptions of varying magnitude and duration were found in every country, with no clear patterns by country income group or pandemic intensity. Disruptions in health services often preceded COVID-19 waves. Cancer screenings, TB screening and detection and HIV testing were most affected (26-96% declines). Total outpatient visits declined by 9-40% at national levels and remained lower than predicted by the end of 2020. Maternal health services were disrupted in approximately half of the countries, with declines ranging from 5% to 33%. Child vaccinations were disrupted for shorter periods, but we estimate that catch-up campaigns might not have reached all children missed. By contrast, provision of antiretrovirals for HIV was not affected. By the end of 2020, substantial disruptions remained in half of the countries. Preliminary data for 2021 indicate that disruptions likely persisted. Although a portion of the declines observed might result from decreased needs during lockdowns (from fewer infectious illnesses or injuries), a larger share likely reflects a shortfall of health system resilience. Countries must plan to compensate for missed healthcare during the current pandemic and invest in strategies for better health system resilience for future emergencies.