Profile: the Karonga Health and Demographic Surveillance System

Temporal changes in cause of death among adolescents and adults in six countries in eastern and southern Africa in 1995-2019: a multi-country surveillance study of verbal autopsy data

BACKGROUND

The absence of high-quality comprehensive civil registration and vital statistics systems across many settings in Africa has led to little empirical data on causes of death in the region. We aimed to use verbal autopsy data to provide comparative, population-based estimates of cause-specific mortality among adolescents and adults in eastern and southern Africa.

METHODS

In this surveillance study, we harmonised verbal autopsy and residency data from nine health and demographic surveillance system (HDSS) sites in Kenya, Malawi, Tanzania, South Africa, Uganda, and Zimbabwe, each with variable coverage from Jan 1, 1995, to Dec 31, 2019. We included all deaths to adolescents and adults aged 12 or over that were residents of the study sites and had a verbal autopsy conducted. InSilicoVA, a probabilistic model, was used to assign cause of death on the basis of the signs and symptoms reported in the verbal autopsy. Levels and trends in all-cause and cause-specific mortality rates and cause-specific mortality fractions were calculated, stratified by HDSS site, sex, age, and calendar periods.

FINDINGS

52 484 deaths and 5 157 802 person-years were reported among 1 071 913 individuals across the nine sites during the study period. 47 961 (91·4%) deaths had a verbal autopsy, of which 46 570 (97·1%) were assigned a cause of death. All-cause mortality generally decreased across the HDSS sites during this period, particularly for adults aged 20-59 years. In many of the HDSS sites, these decreases were driven by reductions in HIV and tuberculosis-related deaths. In 2010-14, the top causes of death were: road traffic accidents, HIV or tuberculosis, and meningitis or sepsis in adolescents (12-19 years); HIV or tuberculosis in adults aged 20-59 years; and neoplasms and cardiovascular disease in adults aged 60 years and older. There was greater between-HDSS and between-sex variation in causes of death for adolescents compared with adults.

INTERPRETATION

This study shows progress in reducing mortality across eastern and southern Africa but also highlights age, sex, within-HDSS, and between-HDSS differences in causes of adolescent and adult deaths. These findings highlight the importance of detailed local data to inform health needs to ensure continued improvements in survival.

FUNDING

National Institute of Child Health and Human Development of the US National Institutes of Health.

Factors associated with self-reported diagnosed asthma in urban and rural Malawi: Observations from a population-based study of non-communicable diseases

The growing burden of asthma in low- and middle-income countries has been linked to urbanisation and lifestyle changes. However, this burden has not been well characterised in adults. Therefore, we investigated the prevalence of self-reported diagnosed asthma and associated factors in urban and rural adults in Malawi, Southern Africa. Within a cross-sectional population-based survey to determine the burden and risk factors for non-communicable diseases (NCDs) in the city of Lilongwe and rural Karonga district, we collected information on self-reported previously diagnosed asthma and asthma-related symptoms using an interviewer-led questionnaire. Other data collected included: demographic characteristics, socioeconomic status indicators, NCD comorbidities, environmental exposures, and anthropometric measurements. We used multivariable logistic regression models to explore factors associated with self-reported asthma adjusting for variables associated with the outcome in univariable analysis. Findings were corrected for multiple comparisons using the Bonferroni method. We analysed data from 30,483 adult participants (54.6% urban,45.4% rural and 61.9% female). A prior asthma diagnosis was reported in 5.1% of urban and 4.5% of rural participants. In urban females, being obese (>30 kg/m2) compared to normal weight (18.5-24.9 kg/m2) was associated with greater odds of asthma (OR = 1.59, 95% CI [1.26-2.01], p<0.001), after adjusting for confounders. We observed associations between previously diagnosed heart disease and asthma in female participants which remained significant in rural females after Bonferroni correction (OR = 2.30,95%CI [1.32-4.02], p = 0.003). Among rural males, current smokers had reduced odds of diagnosed asthma (OR = 0.46,95%CI [0.27-0.79], p = 0.004) compared to those who had never smoked. In Malawi the prevalence of self-reported diagnosed asthma was greatest in females and urban dwellers. Notably, our findings indicate relationships between excess body weight as well as comorbidities and diagnosed asthma in females. Future investigations using longitudinally collected data and clinical measurements of asthma are needed to better understand these associations.

Family network and household composition: a longitudinal dataset derived from the Karonga Health and Demographic Surveillance System, in rural Malawi

Proximity to family, household composition, and structure are often studied as outcomes and as explanatory factors in a wide range of scientific disciplines. Here, we describe a large longitudinal dataset (currently including data from over 70,000 individuals from 2004 to 2017), including data on household structure, proximity to kin, population density, and other socio-demographic factors derived from data from the Karonga Health and Demographic Surveillance Site (HDSS) in Northern Malawi. We present how the dataset is generated, list some examples of how it can be used, and provide information on the limitations that affect the types of analyses that can be carried out.

Understanding the health system utilisation and reasons for avoidable mortality after fatal injury within a Three-Delays framework in Karonga, Northern Malawi: a retrospective analysis of verbal autopsy data

OBJECTIVES

To use verbal autopsy (VA) data to understand health system utilisation and the potential avoidability associated with fatal injury. Then to categorise any evident barriers driving avoidable delays to care within a Three-Delays framework that considers delays to seeking (Delay 1), reaching (Delay 2) or receiving (Delay 3) quality injury care.

DESIGN

Retrospective analysis of existing VA data routinely collected by a demographic surveillance site.

SETTING

Karonga Health and Demographic Surveillance Site (HDSS) population, Northern Malawi.

PARTICIPANTS

Fatally injured members of the HDSS.

PRIMARY AND SECONDARY OUTCOME MEASURES

The primary outcome was the proportion of fatal injury deaths that were potentially avoidable. Secondary outcomes were the delay stage and corresponding barriers associated with avoidable deaths and the health system utilisation for fatal injuries within the health system.

RESULTS

Of the 252 deaths due to external causes, 185 injury-related deaths were analysed. Deaths were predominantly among young males (median age 30, IQR 11-48), 71.9% (133/185). 35.1% (65/185) were assessed as potentially avoidable. Delay 1 was implicated in 30.8% (20/65) of potentially avoidable deaths, Delay 2 in 61.5% (40/65) and Delay 3 in 75.4% (49/65). Within Delay 1, 'healthcare literacy' was most commonly implicated barrier in 75% (15/20). Within Delay 2, 'communication' and 'prehospital care' were the most commonly implicated in 92.5% (37/40). Within Delay 3, 'physical resources' were most commonly implicated, 85.7% (42/49).

CONCLUSIONS

VA is feasible for studying pathways to care and health system responsiveness in avoidable deaths following injury and ascertaining the delays that contribute to deaths. A large proportion of injury deaths were avoidable, and we have identified several barriers as potential targets for intervention. Refining and integrating VA with other health system assessment methods is likely necessary to holistically understand an injury care health system.

Association between physical multimorbidity and common mental health disorders in rural and urban Malawian settings: Preliminary findings from Healthy Lives Malawi long-term conditions survey

In low-income Africa, the epidemiology of physical multimorbidity and associated mental health conditions is not well described. We investigated the multimorbidity burden, disease combinations, and relationship between physical multimorbidity and common mental health disorders in rural and urban Malawi using early data from 9,849 adults recruited to an on-going large cross-sectional study on long-term conditions, initiated in 2021. Multimorbidity was defined as having two or more measured (diabetes, hypertension) or self-reported (diabetes, hypertension, disability, chronic pain, HIV, asthma, stroke, heart disease, and epilepsy) conditions. Depression and anxiety symptoms were measured using the 9-item Patient Health Questionnaire (PHQ-9) and the 7-item General Anxiety Disorder scale (GAD-7) and defined by the total score (range 0-27 and 0-21, respectively). We determined age-standardized multimorbidity prevalence and condition combinations. Additionally, we used multiple linear regression models to examine the association between physical multimorbidity and depression and anxiety symptom scores. Of participants, 81% were rural dwelling, 56% were female, and the median age was 30 years (Inter Quartile Range 21-43). The age-standardized urban and rural prevalence of multimorbidity was 14.1% (95% CI, 12.5-15.8%) and 12.2% (95% CI, 11.6-12.9%), respectively. In adults with two conditions, hypertension, and disability co-occurred most frequently (18%), and in those with three conditions, hypertension, disability, and chronic pain were the most common combination (23%). Compared to adults without physical conditions, having one (B-Coefficient (B) 0.79; 95% C1 0.63-0.94%), two- (B 1.36; 95% CI 1.14-1.58%), and three- or more- physical conditions (B 2.23; 95% CI 1.86-2.59%) were associated with increasing depression score, p-trend <0.001. A comparable 'dose-response' relationship was observed between physical multimorbidity and anxiety symptom scores. While the direction of observed associations cannot be determined with these cross-sectional data, our findings highlight the burden of multimorbidity and the need to integrate mental and physical health service delivery in Malawi.

Local and long-distance migration among young people in rural Malawi: importance of age, sex and family

Background

In sub-Saharan Africa, migration of young people is common and occurs for a variety of reasons. Research focus is often on international or long-distance internal migration; however, shorter moves also affect people's lives and can reveal important information about cultures and societies. In rural sub-Saharan Africa, migration may be influenced by cultural norms and family considerations: these may be changing due to demographic shifts, urbanisation, and increased media access.

Methods

We used longitudinal data from a Health and Demographic Surveillance Site in rural northern Malawi to present a detailed investigation of migration in young people between 2004-2017. Our focus is on the cultural effects of gender and family, and separate migrations into short and long distance, and independent and accompanied, as these different move types are likely to represent very different events in a young person's life. We use descriptive analyses multi-level multinomial logistic regression modelling.

Results & conclusions

We found two key periods of mobility 1) in very young childhood and 2) in adolescence/young adulthood. In this traditionally patrilocal area, we found that young women move longer distances to live with their spouse, and also were more likely to return home after a marriage ends, rather than remain living independently. Young people living close to relatives tend to have lower chances of moving, and despite the local patrilineal customs, we found evidence of the importance of the maternal family. Female and male children may be treated differently from as young as age 4, with girls more likely to migrate long distances independently, and more likely to accompany their mothers in other moves.

Access to quality care after injury in Northern Malawi: results of a household survey

BACKGROUND

Most injury care research in low-income contexts such as Malawi is facility centric. Community-derived data is needed to better understand actual injury incidence, health system utilisation and barriers to seeking care following injury.

METHODS

We administered a household survey to 2200 households in Karonga, Malawi. The primary outcome was injury incidence, with non-fatal injuries classified as major or minor (> 30 or 1-29 disability days respectively). Those seeking medical treatment were asked about time delays to seeking, reaching and receiving care at a facility, where they sought care, and whether they attended a second facility. We performed analysis for associations between injury severity and whether the patient sought care, stayed overnight in a facility, attended a second facility, or received care within 1 or 2 h. The reason for those not seeking care was asked.

RESULTS

Most households (82.7%) completed the survey, with 29.2% reporting an injury. Overall, 611 non-fatal and four fatal injuries were reported from 531 households: an incidence of 6900 per 100,000. Major injuries accounted for 26.6%. Three quarters, 76.1% (465/611), sought medical attention. Almost all, 96.3% (448/465), seeking care attended a primary facility first. Only 29.7% (138/465), attended a second place of care. Only 32.0% (142/444), received care within one hour. A further 19.1% (85/444) received care within 2 h. Major injury was associated with being more likely to have; sought care (94.4% vs 69.8% p < 0.001), stayed overnight at a facility (22.9% vs 15.4% P = 0.047), attended a second place of care (50.3% vs 19.9%, P < 0.001). For those not seeking care the most important reason was the injury not being serious enough for 52.1% (74/142), followed by transport difficulties 13.4% (19/142) and financial costs 5.6% (8/142).

CONCLUSION

Injuries in Northern Malawi are substantial. Community-derived details are necessary to fully understand injury burden and barriers to seeking and reaching care.

Health system assessment for access to care after injury in low- or middle-income countries: A mixed methods study from Northern Malawi

BACKGROUND

Injuries represent a vast and relatively neglected burden of disease affecting low- and middle-income countries (LMICs). While many health systems underperform in treating injured patients, most assessments have not considered the whole system. We integrated findings from 9 methods using a 3 delays approach (delays in seeking, reaching, or receiving care) to prioritise important trauma care health system barriers in Karonga, Northern Malawi, and exemplify a holistic health system assessment approach applicable in comparable settings.

METHODS AND FINDINGS

To provide multiple perspectives on each conceptual delay and include data from community-based and facility-based sources, we used 9 methods to examine the injury care health system. The methods were (1) household survey; (2) verbal autopsy analysis; (3) community focus group discussions (FGDs); (4) community photovoice; (5) facility care-pathway process mapping and elucidation of barriers following injury; (6) facility healthcare worker survey; (7) facility assessment survey; (8) clinical vignettes for care process quality assessment of facility-based healthcare workers; and (9) geographic information system (GIS) analysis. Empirical data collection took place in Karonga, Northern Malawi, between July 2019 and February 2020. We used a convergent parallel study design concurrently conducting all data collection before subsequently integrating results for interpretation. For each delay, a matrix was created to juxtapose method-specific data relevant to each barrier identified as driving delays to injury care. Using a consensus approach, we graded the evidence from each method as to whether an identified barrier was important within the health system. We identified 26 barriers to access timely quality injury care evidenced by at least 3 of the 9 study methods. There were 10 barriers at delay 1, 6 at delay 2, and 10 at delay 3. We found that the barriers "cost," "transport," and "physical resources" had the most methods providing strong evidence they were important health system barriers within delays 1 (seeking care), 2 (reaching care), and 3 (receiving care), respectively. Facility process mapping provided evidence for the greatest number of barriers-25 of 26 within the integrated analysis. There were some barriers with notable divergent findings between the community- and facility-based methods, as well as among different community- and facility-based methods, which are discussed. The main limitation of our study is that the framework for grading evidence strength for important health system barriers across the 9 studies was done by author-derived consensus; other researchers might have created a different framework.

CONCLUSIONS

By integrating 9 different methods, including qualitative, quantitative, community-, patient-, and healthcare worker-derived data sources, we gained a rich insight into the functioning of this health system's ability to provide injury care. This approach allowed more holistic appraisal of this health system's issues by establishing convergence of evidence across the diverse methods used that the barriers of cost, transport, and physical resources were the most important health system barriers driving delays to seeking, reaching, and receiving injury care, respectively. This offers direction and confidence, over and above that derived from single methodology studies, for prioritising barriers to address through health service development and policy.

Maternal mental health research in Malawi: Community and healthcare provider perspectives on acceptability and ethicality

Maternal mental health (MMH) is recognised as globally significant. The prevalence of depression and factors associated with its onset among perinatal women in Malawi has been previously reported, and the need for further research in this domain is underscored. Yet, there is little published scholarship regarding the acceptability and ethicality of MMH research to women and community representatives. The study reported here sought to address this in Malawi by engaging with communities and healthcare providers in the districts where MMH research was being planned. Qualitative data was collected in Lilongwe and Karonga districts through 20 focus group discussions and 40 in-depth interviews with community representatives and healthcare providers from January through April 2021. All focus groups and interviews were audio recorded, transcribed verbatim (in local languages Chichewa and Tumbuka), translated into English, and examined through thematic content analysis. Participants' accounts suggest that biopsychosocial MMH research could be broadly acceptable within the communities sampled, with acceptability framed in part through prior encounters with biomedical and public health research and care in these regions, alongside broader understandings of the import of MMH. Willingness and consent to participate do not depend on specifically biomedical understandings of MMH, but rather on familiarity with individuals regarded as living with mental ill-health. However, the data further suggest some 'therapeutic misconceptions' about MMH research, with implications for how investigations in this area are presented by researchers when recruiting and working with participants. Further studies are needed to explore whether accounts of the acceptability and ethicality of MMH research shift and change during and following research encounters. Such studies will enhance the production of granular recommendations for further augmenting the ethicality of biomedical and public health research and researchers' responsibilities to participants and communities.

Effects of exclusive breastfeeding on educational attainment and longitudinal trajectories of grade progression among children in a 13-year follow-up study in Malawi

The benefits of exclusive breastfeeding (EBF) for infant health and survival are well documented. However, its impact on educational outcomes has been contested and poorly researched in Africa. It has been hypothesised that positive associations reported in high-income countries can be attributed to residual confounding by socioeconomic status (SES). Our study investigated whether EBF duration in infancy is associated with educational attainment and age-for-grade attainment trajectories at school-age in rural Malawi. Longitudinal data on 1021 children at the Karonga demographic surveillance site in Malawi were analysed. Breastfeeding data were collected 3 months after birth and again at age one. The school grade of each child was recorded each year from age 6 until age 13. We calculated age-for-grade based on whether a child was at, over, or under the official expected age for a grade. Generalised estimating equations estimated the average effect of breastfeeding on age-for-grade. Latent class growth analysis identified age-for-grade trajectories, and multinomial logistic regression examined their associations with EBF. Maternal-child characteristics, SES, and HIV status were controlled. Overall, 35.9% of the children were exclusively breastfed for 6 months. Over-age for grade steadily increased from 9.6% at age 8 to 41.9% at age 13. There was some evidence that EBF for 6 months was associated with lower odds of being over-age for grade than EBF for less than 3 months (aOR = 0.82, 95%CI = 0.64-1.06). In subgroup analyses, children exclusively breastfed for 6 months in infancy were less likely to be over-age for grades between ages 6-9 (aOR = 0.64, 95%CI = 0.43-0.94). Latent class growth analysis also provided some evidence that EBF reduced the odds of falling behind in the early school grades (aOR = 0.66, 95%CI = 0.41-1.08) but not later. Our study adds to the growing evidence that EBF for 6 months has benefits beyond infant health and survival, supporting the WHO's recommendation on EBF.

Novel application of multi-facility process map analysis for rapid injury care health system assessment in Northern Malawi

OBJECTIVES

We used the process mapping method and Three Delays framework, to identify and visually represent the relationship between critical actions, decisions and barriers to access to care following injury in the Karonga health system, Northern Malawi.

DESIGN

Facilitated group process mapping workshops with summary process mapping synthesis.

SETTING

Process mapping workshops took place in 11 identified health system facilities (one per facility) providing injury care for a population in Karonga, Northern Malawi.

PARTICIPANTS

Fifty-four healthcare workers from various cadres took part.

RESULTS

An overall injury health system summary map was created using those categories of action, decision and barrier that were sometimes or frequently reported. This provided a visual summary of the process following injury within the health system. For Delay 1 (seeking care) four barriers were most commonly described (by 8 of 11 facilities) these were 'cultural norms', 'healthcare literacy', 'traditional healers' and 'police processes'. For Delay 2 (reaching care) the barrier most frequently described was 'transport'-a lack of timely affordable emergency transport (formal or informal) described by all 11 facilities. For Delay 3 (receiving quality care) the most commonly reported barrier was that of 'physical resources' (9 of 11 facilities).

CONCLUSIONS

We found our novel approach combining several process mapping exercises to produce a summary map to be highly suited to rapid health system assessment identifying barriers to injury care, within a Three Delays framework. We commend the approach to others wishing to conduct rapid health system assessments in similar contexts.

Contribution of remote M.tuberculosis infection to tuberculosis disease: A 30-year population study

BACKGROUND

The importance of remote infection with M.tuberculosis as a cause of tuberculosis disease (TB) is unclear, with limited evidence of impact on TB rates beyond 10 years. Our objective was to assess rates of tuberculosis over 30 years by M.tuberculosis infection status at baseline in Karonga District, Northern Malawi.

MATERIALS AND METHODS

Population-based surveys of tuberculin skin testing (TST) from the 1980s were linked with follow-up and TB surveillance in Karonga district. We compared rates of microbiologically-confirmed TB by baseline TST induration <5mm (no evidence of M.tuberculosis infection) and those with baseline TST >17mm (evidence of M.tuberculosis infection), using hazard ratios by time since baseline and attributable risk percent. The attributable risk percent was calculated to estimate the proportion of TB in those infected that can be attributed to that prior infection. We analysed whole genome sequences of M.tuberculosis strains to identify recent transmission.

RESULTS

Over 412,959 person-years, 208 incident TB episodes were recorded. Compared to the small induration group, rates of TB were much higher in the first two years in the large induration group, and remained higher to 20 years: age, sex and area-adjusted hazard ratios (HR) 2-9 years post-TST 4.27 (95%CI 2.56-7.11); 10-19 years after TST 2.15 (1.10-4.21); ≥20 years post-TST 1.88 (0.76-4.65). The attributable risk percent of remote infection was 76.6% (60.9-85.9) 2-9 years post-TST, and 53.5% (9.1-76.2) 10-19 years post-TST. Individuals with large TST indurations had higher rates of unique-strain TB (HR adjusted for age, sex and area = HR 6.56 (95% CI 1.96-22.99)), suggesting disease following remote infection, but not of linked-strain TB (recent transmission).

CONCLUSIONS

M.tuberculosis infection can increase the risk of TB far beyond 10 years, accounting for a substantial proportion of TB occurring among those remotely infected.

Access to care following injury in Northern Malawi, a comparison of travel time estimates between Geographic Information System and community household reports

INTRODUCTION

Injuries disproportionately impact low- and middle-income countries like Malawi. The Lancet Commission on Global Surgery's indicators include the population proportion accessing laparotomy and open fracture care, key trauma interventions, within two hours. The "Golden Hour" for receiving facility-based resuscitation also guides injury care system strengthening. Firstly, we estimated the proportion of the local population able to reach primary, secondary and tertiary facility care within two and one hours using Geographic Information System (GIS) analysis. Secondly, we compared community household-reported with GIS-estimated travel time.

METHODS

Using information from a Health and Demographic Surveillance Site (Karonga, Malawi) on road network, facility location, and local staff-estimated travel speeds, we used a GIS-generated friction surface to calculate the shortest travel time from all households to each facility serving the population. We surveyed community households who reported travel time to their preferred, closest, government secondary and tertiary facilities. For recently injured community members, time to reach facility care was recorded. To assess the relationship between community household-reported travel time and GIS-estimated travel time, we used linear regression to generate a proportionality constant. To assess associations and agreement between injured patient-reported and GIS-estimated travel time, we used Kendall rank and Cohen's kappa tests.

RESULTS

Using GIS, we estimated 79.1% of households could reach any secondary facility, 20.5% the government secondary facility, and 0% the government tertiary facility, within two hours. Only 28.2% could reach any secondary facility within one hour, 0% for the government secondary facility. Community household-reported travel time exceeded GIS-estimated travel time. The proportionality constant was 1.25 (95%CI 1.21-1.30) for the closest facility, 1.28 (95%CI 1.23-1.34) for the preferred facility, 1.45 (95%CI 1.33-1.58) for the government secondary facility, and 2.12 (95%CI 1.84-2.41) for tertiary care. Comparing injured patient-reported with GIS-estimated travel time, the correlation coefficient was 0.25 (SE 0.047) and Cohen's kappa was 0.15 (95%CI 0.078-0.23), suggesting poor agreement.

DISCUSSION

Most households couldn't reach government secondary care within recognised thresholds indicating poor temporal access. Since GIS-estimated travel time was shorter than community-reported travel time, the true proportion may be lower still. GIS derived estimates of population emergency care access in similar contexts should be interpreted accordingly.

Association between household air pollution and nasopharyngeal pneumococcal carriage in Malawian infants (MSCAPE): a nested, prospective, observational study

Background

Household air pollution from solid fuels increases the risk of childhood pneumonia. Nasopharyngeal carriage of Streptococcus pneumoniae is a necessary step in the development of pneumococcal pneumonia. We aimed to assess the association between exposure to household air pollution and the prevalence and density of S pneumoniae carriage among children.

Methods

The Malawi Streptococcus pneumoniae Carriage and Air Pollution Exposure study was a nested, prospective, observational study of children participating in the cluster randomised controlled Cooking and Pneumonia Study (CAPS) in the Karonga Health and Demographic Surveillance System (HDSS) area in northern Malawi. CAPS compared the effects of a cleaner burning biomass-fuelled cookstove (intervention group) with traditional open-fire cooking (control group) on the incidence of pneumonia in children. Eligible children aged 6 weeks or 6 months (those recruited a 6 weeks were also followed up at age 6 months) were identified by the Karonga HDSS centre. Nasopharyngeal swabs were taken to detect S pneumoniae, and infant exposure to particulate matter with a diameter of ≤2·5 μm (PM_2·5) exposure was assessed by use of a MicroPEM device. The primary outcome was the prevalence of nasopharyngeal S pneumoniae carriage in all children aged 6 months, assessed in all children with valid data on PM_2·5. The effects of the intervention stoves (intention-to-treat analysis) and PM_2·5 (adjusted exposure-response analysis) on the prevalence of S pneumoniae carriage were also assessed in the study children.

Findings

Between Nov 15, 2015, and Nov 2, 2017, 485 children were recruited (240 from the intervention group and 245 from the control group). Of all 450 children with available data at age 6 months, 387 (86% [95% CI 82–89]) were positive for S pneumoniae. Geometric mean PM_2·5 exposure was 60·3 μg/m³ (95% CI 55·8–65·3) in S pneumoniae-positive children and 47·0 μg/m³ (38·3–57·7) in S pneumoniae-negative children (p=0·044). In the intention-to-treat analysis, a non-significant increase in the risk of S pneumoniae carriage was observed in intervention group children compared with control group children (odds ratio 1·36 [95% CI 0·95–1·94]; p=0·093). In the exposure-response analysis, a significant association between PM_2·5 exposure and S pneumoniae carriage was observed; a one unit increase in decile of PM_2·5 was found to significantly increase the risk of S pneumoniae carriage by 10% (1·10 [1·01–1·20]; p=0·035), after adjustment for age, sex, 13-valent pneumococcal conjugate vaccination status, season, current use of antibiotics, and MicroPEM run-time.

Interpretation

Despite the absence of effect from the intervention cookstove, household air pollution exposure was significantly associated with the prevalence of nasopharyngeal S pneumoniae carriage. These results provide empirical evidence for the potential mechanistic association between exposure to household air pollution and childhood pneumonia.

Funding

Bill & Melinda Gates Foundation.

The effect of BCG revaccination on all-cause mortality beyond infancy: 30-year follow-up of a population-based, double-blind, randomised placebo-controlled trial in Malawi

BACKGROUND

Trials of BCG vaccination to prevent or reduce severity of COVID-19 are taking place in adults, some of whom have been previously vaccinated, but evidence of the beneficial, non-specific effects of BCG come largely from data on mortality in infants and young children, and from in-vitro and animal studies, after a first BCG vaccination. We assess all-cause mortality following a large BCG revaccination trial in Malawi.

METHODS

The Karonga Prevention trial was a population-based, double-blind, randomised controlled in Karonga District, northern Malawi, that enrolled participants between January, 1986, and November, 1989. The trial compared BCG (Glaxo-strain) revaccination versus placebo to prevent tuberculosis and leprosy. 46 889 individuals aged 3 months to 75 years were randomly assigned to receive BCG revaccination (n=23 528) or placebo (n=23 361). Here we report mortality since vaccination as recorded during active follow-up in northern areas of the district in 1991-94, and in a demographic surveillance follow-up in the southern area in 2002-18. 7389 individuals who received BCG (n=3746) or placebo (n=3643) lived in the northern follow-up areas, and 5616 individuals who received BCG (n=2798) or placebo (n=2818) lived in the southern area. Year of death or leaving the area were recorded for those not found. We used survival analysis to estimate all-cause mortality.

FINDINGS

Follow-up information was available for 3709 (99·0%) BCG recipients and 3612 (99·1%) placebo recipients in the northern areas, and 2449 (87·5%) BCG recipients and 2413 (85·6%) placebo recipients in the southern area. There was no difference in mortality between the BCG and placebo groups in either area, overall or by age group or sex. In the northern area, there were 129 deaths per 19 694 person-years at risk in the BCG group (6·6 deaths per 1000 person-years at risk [95% CI 5·5-7·8]) versus 133 deaths per 19 111 person-years at risk in the placebo group (7·0 deaths per 1000 person-years at risk [95% CI 5·9-8·2]; HR 0·94 [95% CI 0·74-1·20]; p=0·62). In the southern area, there were 241 deaths per 38 399 person-years at risk in the BCG group (6·3 deaths per 1000 person-years at risk [95% CI 5·5-7·1]) versus 230 deaths per 38 676 person-years at risk in the placebo group (5·9 deaths per 1000 person-years at risk [95% CI 5·2-6·8]; HR 1·06 [95% CI 0·88-1·27]; p=0·54).

INTERPRETATION

We found little evidence of any beneficial effect of BCG revaccination on all-cause mortality. The high proportion of deaths attributable to non-infectious causes beyond infancy, and the long time interval since BCG for most deaths, might obscure any benefits.

FUNDING

British Leprosy Relief Association (LEPRA); Wellcome Trust.

Health and demographic surveillance systems in low- and middle-income countries: history, state of the art and future prospects

Health and Demographic Surveillance Systems (HDSS) have been developed in several low- and middle-income countries (LMICs) in Africa and Asia. This paper reviews their history, state of the art and future potential and highlights substantial areas of contribution by the late Professor Peter Byass.Historically, HDSS appeared in the second half of the twentieth century, responding to a dearth of accurate population data in poorly resourced settings to contextualise the study of interventions to improve health and well-being. The progress of the development of this network is described starting with Pholela, and progressing through Gwembe, Balabgarh, Niakhar, Matlab, Navrongo, Agincourt, Farafenni, and Butajira, and the emergence of the INDEPTH Network in the early 1990'sThe paper describes the HDSS methodology, data, strengths, and limitations. The strengths are particularly their temporal coverage, detail, dense linkage, and the fact that they exist in chronically under-documented populations in LMICs where HDSS sites operate. The main limitations are generalisability to a national population and a potential Hawthorne effect, whereby the project itself may have changed characteristics of the population.The future will include advances in HDSS data harmonisation, accessibility, and protection. Key applications of the data are to validate and assess bias in other datasets. A strong collaboration between a national HDSS network and the national statistics office is modelled in South Africa and Sierra Leone, and it is possible that other low- to middle-income countries will see the benefit and take this approach.

Privacy of study participants in open-access Health and Demographic Surveillance System data: A requirements analysis for data anonymisation (Preprint)

Background

Data anonymization and sharing have become popular topics for individuals, organizations, and countries worldwide. Open-access sharing of anonymized data containing sensitive information about individuals makes the most sense whenever the utility of the data can be preserved and the risk of disclosure can be kept below acceptable levels. In this case, researchers can use the data without access restrictions and limitations.

Objective

This study aimed to highlight the requirements and possible solutions for sharing health surveillance event history data. The challenges lie in the anonymization of multiple event dates and time-varying variables.

Methods

A sequential approach that adds noise to event dates is proposed. This approach maintains the event order and preserves the average time between events. In addition, a nosy neighbor distance-based matching approach to estimate the risk is proposed. Regarding the key variables that change over time, such as educational level or occupation, we make 2 proposals: one based on limiting the intermediate statuses of the individual and the other to achieve k-anonymity in subsets of the data. The proposed approaches were applied to the Karonga health and demographic surveillance system (HDSS) core residency data set, which contains longitudinal data from 1995 to the end of 2016 and includes 280,381 events with time-varying socioeconomic variables and demographic information.

Results

An anonymized version of the event history data, including longitudinal information on individuals over time, with high data utility, was created.

Conclusions

The proposed anonymization of event history data comprising static and time-varying variables applied to HDSS data led to acceptable disclosure risk, preserved utility, and being sharable as public use data. It was found that high utility was achieved, even with the highest level of noise added to the core event dates. The details are important to ensure consistency or credibility. Importantly, the sequential noise addition approach presented in this study does not only maintain the event order recorded in the original data but also maintains the time between events. We proposed an approach that preserves the data utility well but limits the number of response categories for the time-varying variables. Furthermore, using distance-based neighborhood matching, we simulated an attack under a nosy neighbor situation and by using a worst-case scenario where attackers have full information on the original data. We showed that the disclosure risk is very low, even when assuming that the attacker’s database and information are optimal. The HDSS and medical science research communities in low- and middle-income country settings will be the primary beneficiaries of the results and methods presented in this paper; however, the results will be useful for anyone working on anonymizing longitudinal event history data with time-varying variables for the purposes of sharing.

BCG re-vaccination in Malawi: 30-year follow-up of a large, randomised, double-blind, placebo-controlled trial

BACKGROUND

A large, double-blind, randomised, placebo-controlled trial of repeat BCG found 49% efficacy against leprosy but no protection against tuberculosis after 6-9 years' follow-up in 1995. We report here additional follow-up, which resulted in greatly increased tuberculosis case numbers, and allowed subgroup analysis.

METHODS

Nearly 47 000 individuals of all ages living in northern Malawi with a BCG vaccine scar were randomly assigned (1:1) between 1986 and 1989 to receive a second BCG or placebo. The investigators and project staff remained masked to all interventions. Enhanced passive surveillance ensured ascertainment of tuberculosis and leprosy to the end of 2018. Tuberculosis case definitions included rigorous microbiological or histological confirmation. Prespecified subgroup analyses were by tuberculosis type, age at vaccination, time since vaccination, previous tuberculin reactivity, HIV status and Mycobacterium tuberculosis lineage. The original trial is registered with ISRCTN registry, ISRCTN11311670.

FINDINGS

In follow-up until Dec 31, 2018, 824 participants had developed tuberculosis, including 786 with pulmonary disease, of whom 383 (63%) of 607 with known HIV status were HIV positive. There was no effect of a second BCG overall (odds ratio [OR] 0·92; 95% CI 0·80-1·05), or for pulmonary (0·93; 0·81-1·07), or lymph node tuberculosis (0·60; 0·31-1·17). The OR was lower for those with known HIV-negative tuberculosis (0·77; 0·59-1·00), for those vaccinated as children (aged <5 years, 0·74; 0·41-1·35; aged 5-14 years, 0·77; 0·60-0·99), and for cases arising at least 20 years after vaccination (0·79; 0·63-1·01). There were no differences by tuberculin status at vaccination, or lineage. There was no evidence of protection against leprosy beyond 10 years after vaccination (although there have been only nine diagnostically certain cases since 1995).

INTERPRETATION

There was no evidence that repeat BCG vaccination provides appreciable protection against overall tuberculosis in this rural African population with a high prevalence of HIV. Subgroup effects should not be overinterpreted given the multiple analyses done. However, the evidence for modest protection against HIV-negative tuberculosis, and for a delayed benefit in those vaccinated as children, is consistent with other observations in the literature.

FUNDING

LEPRA, Wellcome Trust, Bill & Melinda Gates Foundation.

Pregnancy Surveillance Methods within Health and Demographic Surveillance Systems

Background: Pregnancy identification and follow-up surveillance can enhance the reporting of pregnancy outcomes, including stillbirths and perinatal and early postnatal mortality. This paper reviews pregnancy surveillance methods used in Health and Demographic Surveillance Systems (HDSSs) in low- and middle-income countries. Methods: We searched articles containing information about pregnancy identification methods used in HDSSs published between January 2002 and October 2019 using PubMed and Google Scholar. A total of 37 articles were included through literature review and 22 additional articles were identified via manual search of references. We reviewed the gray literature, including websites, online reports, data collection instruments, and HDSS protocols from the Child Health and Mortality Prevention Study (CHAMPS) Network and the International Network for the Demographic Evaluation of Populations and Their Health (INDEPTH). In total, we reviewed information from 52 HDSSs described in 67 sources. Results: Substantial variability exists in pregnancy surveillance approaches across the 52 HDSSs, and surveillance methods are not always clearly documented. 42% of HDSSs applied restrictions based on residency duration to identify who should be included in surveillance. Most commonly, eligible individuals resided in the demographic surveillance area (DSA) for at least three months. 44% of the HDSSs restricted eligibility for pregnancy surveillance based on a woman's age, with most only monitoring women 15-49 years. 10% had eligibility criteria based on marital status, while 11% explicitly included unmarried women in pregnancy surveillance. 38% allowed proxy respondents to answer questions about a woman's pregnancy status in her absence. 20% of HDSSs supplemented pregnancy surveillance with investigations by community health workers or key informants and by linking HDSS data with data from antenatal clinics. Conclusions: Methodological guidelines for conducting pregnancy surveillance should be clearly documented and meticulously implemented, as they can have implications for data quality and accurately informing maternal and child health programs.

The ethical implications of verbal autopsy: responding to emotional and moral distress

Background

Verbal autopsy is a pragmatic approach for generating cause-of-death data in contexts without well-functioning civil registration and vital statistics systems. It has primarily been conducted in health and demographic surveillance systems (HDSS) in Africa and Asia. Although significant resources have been invested to develop the technical aspects of verbal autopsy, ethical issues have received little attention. We explored the benefits and burdens of verbal autopsy in HDSS settings and identified potential strategies to respond to the ethical issues identified.

Methods

This research was based on a case study approach centred on two contrasting HDSS in Kenya and followed the Mapping-Framing-Shaping Framework for empirical bioethics research. Data were collected through individual interviews, focus group discussions, document reviews and non-participant observations. 115 participants were involved, including 86 community members (HDSS residents and community representatives), and 29 research staff (HDSS managers, researchers, census field workers and verbal autopsy interviewers).

Results

The use of verbal autopsy data for research and public health was described as the most common potential benefit of verbal autopsy in HDSS. Community members mentioned the potential uses of verbal autopsy data in addressing immediate public health problems for the local population while research staff emphasized the benefits of verbal autopsy to research and the wider public. The most prominent burden associated with the verbal autopsy was emotional distress for verbal autopsy interviewers and respondents. Moral events linked to the interview, such as being unsure of the right thing to do (moral uncertainty) or knowing the right thing to do and being constrained from acting (moral constraint), emerged as key causes of emotional distress for verbal autopsy interviewers.

Conclusions

The collection of cause-of-death data through verbal autopsy in HDSS settings presents important ethical and emotional challenges for verbal autopsy interviewers and respondents. These challenges include emotional distress for respondents and moral distress for interviewers. This empirical ethics study provides detailed accounts of the distress caused by verbal autopsy and highlights ethical tensions between potential population benefits and risks to individuals. It includes recommendations for policy and practice to address emotional and moral distress in verbal autopsy.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12910-021-00683-7.

Age patterns of HIV incidence in eastern and southern Africa: a modelling analysis of observational population-based cohort studies

BACKGROUND

As the HIV epidemic in sub-Saharan Africa matures, evidence about the age distribution of new HIV infections and how this distribution has changed over the epidemic is needed to guide HIV prevention. We aimed to assess trends in age-specific HIV incidence in six population-based cohort studies in eastern and southern Africa, reporting changes in mean age at infection, age distribution of new infections, and birth cohort cumulative incidence.

METHODS

We used a Bayesian model to reconstruct age-specific HIV incidence from repeated observations of individuals' HIV serostatus and survival collected among population HIV cohorts in rural Malawi, South Africa, Tanzania, Uganda, and Zimbabwe, in a collaborative analysis of the ALPHA network. We modelled HIV incidence rates by age, time, and sex using smoothing splines functions. We estimated incidence trends separately by sex and study. We used estimated incidence and prevalence results for 2000-17, standardised to study population distribution, to estimate mean age at infection and proportion of new infections by age. We also estimated cumulative incidence (lifetime risk of infection) by birth cohort.

FINDINGS

Age-specific incidence declined at all ages, although the timing and pattern of decline varied by study. The mean age at infection was higher in men (cohort mean 27·8-34·6 years) than in women (24·8-29·6 years). Between 2000 and 2017, the mean age at infection per cohort increased slightly: 0·5 to 2·8 years among men and -0·2 to 2·5 years among women. Across studies, between 38% and 63% (cohort medians) of the infections in women were among those aged 15-24 years and between 30% and 63% of infections in men were in those aged 20-29 years. Lifetime risk of HIV declined for successive birth cohorts.

INTERPRETATION

HIV incidence declined in all age groups and shifted slightly to older ages. Disproportionate new HIV infections occur among women aged 15-24 years and men aged 20-29 years, supporting focused prevention in these groups. However, 40-60% of infections were outside these ages, emphasising the importance of providing appropriate HIV prevention to adults of all ages.

FUNDING

Bill & Melinda Gates Foundation.

Towards an appropriate ethics framework for Health and Demographic Surveillance Systems (HDSS): learning from issues faced in diverse HDSS in sub-Saharan Africa

Introduction

Health and Demographic Surveillance Systems (HDSS) collect data on births, deaths and migration from relatively small, geographically defined populations primarily in Africa and Asia. HDSS occupy a grey area between research, healthcare and public health practice and it is unclear how ethics guidance that rely on a research-practice distinction apply to HDSS. This topic has received little attention in the literature. In this paper, based on empirical research across sub-Saharan Africa, we map out key ethical issues for HDSS and assess the relevance of current ethics guidance in relation to these findings.

Methods

We conducted a qualitative study across seven HDSS sites in sub-Saharan Africa, including individual in-depth interviews and informal discussions with 68 research staff, document reviews and non-participant observations of surveillance activities. Qualitative data analysis drew on a framework approach led by a priori and emergent themes, drawing on the wider ethics and social science literature.

Results

There were diverse views on core ethical issues in HDSS, including regarding the strengths and challenges of community engagement, informed consent and data sharing processes. A key emerging issue was unfairness in the overall balance of benefits and burdens for residents and front-line staff when compared with other stakeholders, particularly given the socioeconomic contexts in which HDSS are generally conducted.

Conclusion

We argue that HDSS operate as non-traditional epidemiologic research projects but are often governed using ethics guidance developed for traditional forms of health research. There is a need for specific ethics guidance for HDSS which prioritises considerations around fairness, cost-effectiveness, ancillary care responsibilities, longitudinality and obligations of the global community to HDSS residents.

Population impact and effectiveness of sequential 13-valent pneumococcal conjugate and monovalent rotavirus vaccine introduction on infant mortality: prospective birth cohort studies from Malawi

Background

Pneumococcal conjugate vaccine (PCV) and rotavirus vaccine (RV) are key tools for reducing common causes of infant mortality. However, measurement of population-level mortality impact is lacking from sub-Saharan Africa. We evaluated mortality impact and vaccine effectiveness (VE) of PCV13 introduced in November 2011, with subsequent RV1 roll-out in October 2012, in Malawi.

Methods

We conducted two independent community-based birth cohort studies. Study 1, in northern Malawi (40000population), evaluated population impact using change-point analysis and negative-binomial regression of non-traumatic 14–51-week infant mortality preintroduction (1 January 2004 to 31 September 2011) and postintroduction (1 October 2011 to 1 July 2019), and against three-dose coverage. Study 2, in central Malawi (465 000 population), was recruited from 24 November 2011 to 1 June 2015. In the absence of preintroduction data, individual three-dose versus zero-dose VE was estimated using individual-level Cox survival models. In both cohorts, infants were followed with household visits to ascertain vaccination, socioeconomic and survival status. Verbal autopsies were conducted for deaths.

Results

Study 1 included 20 291 live births and 216 infant deaths. Mortality decreased by 28.6% (95% CI: 15.3 to 39.8) post-PCV13 introduction. A change point was identified in November 2012. Study 2 registered 50 731 live births, with 454 deaths. Infant mortality decreased from 17 to 10/1000 live births during the study period. Adjusted VE was 44.6% overall (95% CI: 23.0 to 59.1) and 48.3% (95% CI: −5.9 to 74.1) against combined acute respiratory infection, meningitis and sepsis-associated mortality.

Conclusion

These data provide population-level evidence of infant mortality reduction following sequential PCV13 and RV1 introduction into an established immunisation programme in Malawi. These data support increasing coverage of vaccine programmes in high-burden settings.

Development and use of clinical vignettes to assess injury care quality in Northern Malawi

BACKGROUND

It is known that outcomes after injury care in low-and-middle income countries (LMICs) are poorer than those in high income countries. However, little is known about healthcare provider competency to deliver quality injury care in these settings. We developed and used clinical vignettes to evaluate injury care quality in an LMIC setting.

METHOD

Four serious injury scenarios, developed from agreed best practice, testing diagnostic and management skills, were piloted with high and low-income setting clinicians. Scenarios were used with primary and referral facility clinicians in Malawi. Participants described their clinical course of action (assessment, diagnostic, treatment and management approaches) for each scenario, registering one point per agreed best practice response. Mean percentage total scores were calculated and univariable and multivariable comparison made across provider groups, facility types, injury care frequency and training level.

RESULTS

Fourteen Doctors, 51 Clinical Officers, 20 Medical Assistants from 11 facilities participated. Mean percentage total vignette scores varied significantly with clinician provider group (Doctors 63.1% vs Clinical Officers 49.6%, p<0.001, Clinical Officers vs Medical Assistants 39.4% p=0.001). Important care aspects most frequently included or omitted were: following chest injury, 88.2% reported chest drain insertion, 7.1% checked for tracheal deviation; following penetrating abdominal injury and shock, 98.8% secured IV access, 0% mentioned tranexamic acid; following severe head injury, 88.2% proposed CT or neurosurgical transfer, 7.1% ensured normotension; and following isolated open lower leg fracture, 90.1% arranged orthopaedic consultation, 2.4% assessed distal neurological status.

CONCLUSION

These clinical vignettes proved easy to use and collected rich data. This supports their use for assessing and monitoring clinical care quality in other similar settings.

Maternal and neonatal data collection systems in low- and middle-income countries for maternal vaccines active safety surveillance systems: A scoping review

BACKGROUND

Most post-licensure vaccine pharmacovigilance in low- and middle-income countries (LMICs) are passive reporting systems. These have limited utility for maternal immunization pharmacovigilance in LMIC settings and need to be supplemented with active surveillance. Our study's main objective was to identify existing perinatal data collection systems in LMICs that collect individual information on maternal and neonatal health outcomes and could be developed to inform active safety surveillance of novel vaccines for use during pregnancy.

METHODS

A scoping review was performed following the Arksey and O'Malley six-stage approach. We included studies describing electronic or mixed paper-electronic data collection systems in LMICs, including research networks, electronic medical records, and custom software platforms for health information systems. Medline PubMed, EMBASE, Global Health, Cochrane Library, LILACS, Bibliography of Asian Studies (BAS), and CINAHL were searched through August 2019. We also searched grey literature including through Google and websites of existing relevant perinatal data collection systems, as well as contacted authors of key studies and experts in the field to validate the information and identify additional sources of relevant unpublished information.

RESULTS

A total of 11,817 records were identified. The full texts of 264 records describing 96 data collection systems were assessed for eligibility. Eight perinatal data collection systems met our inclusion criteria: Global Network's Maternal Newborn Health Registry, International Network for the Demographic Evaluation of Populations and their Health; Perinatal Informatic System; Pregnancy Exposure Registry & Birth Defects Surveillance; SmartCare; Open Medical Record System; Open Smart Register Platform and District Health Information Software 2. These selected systems were qualitatively characterized according to seven different domains: governance; system design; system management; data management; data sources, outcomes and data quality.

CONCLUSION

This review provides a list of active maternal and neonatal data collection systems in LMICs and their characteristics as well as their outreach, strengths, and limitations. Findings could potentially help further understand where to obtain population-based high-quality information on outcomes to inform the conduct of maternal immunization active vaccine safety surveillance activities and research in LMICs.

Effect of Acute Illness on Contact Patterns, Malawi, 2017

The way persons interact when ill could profoundly affect transmission of infectious agents. To obtain data on these patterns in Africa, we recorded self-reported named contacts and opportunities for casual contact in rural northern Malawi. We interviewed 384 patients and 257 caregivers about contacts over three 24-hour periods: day of the clinic visit for acute illness, the next day, and 2 weeks later when well. For participants of all ages, the number of adult contacts and the proportion using public transportation was higher on the day of the clinic visit than later when well. Compared with the day after the clinic visit, well participants (2 weeks later) named a mean of 0.4 extra contacts; the increase was larger for indoor or prolonged contacts. When well, participants were more likely to visit other houses and congregate settings. When ill, they had more visitors at home. These findings could help refine models of infection spread.

Informing prevention of stillbirth and preterm birth in Malawi: development of a minimum dataset for health facilities participating in the DIPLOMATIC collaboration

OBJECTIVE

The global research group, DIPLOMATIC (Using eviDence, Implementation science, and a clinical trial PLatform to Optimise MATernal and newborn health in low Income Countries), aims to reduce stillbirths and preterm births and optimise outcomes for babies born preterm. Minimum datasets for routine data collection in healthcare facilities participating in DIPLOMATIC (initially in Malawi) were designed to assist understanding of baseline maternal and neonatal care processes and outcomes, and facilitate evaluation of improvement interventions and pragmatic clinical trials.

DESIGN

Published and grey literature was reviewed alongside extensive in-country consultation to define relevant clinical best practice guidance, and the existing local data and reporting infrastructure, to identify requirements for the minimum datasets. Data elements were subjected to iterative rounds of consultation with topic experts in Malawi and Scotland, the relevant Malawian professional bodies and the Ministry of Health in Malawi to ensure relevance, validity and feasibility.

SETTING

Antenatal, maternity and specialist neonatal care in Malawi.

RESULTS

The resulting three minimum datasets cover the maternal and neonatal healthcare journey for antenatal, maternity and specialist neonatal care, with provision for effective linkage of records for mother/baby pairs. They can facilitate consistent, precise recording of relevant outcomes (stillbirths, preterm births, neonatal deaths), risk factors and key care processes.

CONCLUSIONS

Poor quality routine data on care processes and outcomes constrain healthcare system improvement. The datasets developed for implementation in DIPLOMATIC partner facilities reflect, and hence support delivery of, internationally agreed best practice for maternal and newborn care in low-income settings. Informed by extensive consultation, they are designed to integrate with existing local data infrastructure and reporting as well as meeting research data needs. This work provides a transferable example of strengthening data infrastructure to underpin a learning healthcare system approach in low-income settings.DIPLOMATIC is funded by the UK National Institute for Health Research.

Prevalence and risk factors for chronic kidney disease of unknown cause in Malawi: a cross-sectional analysis in a rural and urban population

BACKGROUND

An epidemic of chronic kidney disease of unknown cause (CKDu) is occurring in rural communities in tropical regions of low-and middle-income countries in South America and India. Little information is available from Southern African countries which have similar climatic and occupational characteristics to CKDu-endemic countries. We investigated whether CKDu is prevalent in Malawi and identified its potential risk factors in this setting.

METHODS

We conducted a cross-sectional study from January-August 2018 collecting bio samples and anthropometric data in two Malawian populations. The sample comprised adults > 18 years (n = 821) without diabetes, hypertension, and proteinuria. Estimates of glomerular filtration rate (eGFR) were calculated using the CKD-EPI equation. Linear and logistic regression models were applied with potential risk factors, to estimate risk of reduced eGFR.

RESULTS

The mean eGFR was 117.1 ± 16.0 ml/min per 1.73m2 and the mean participant age was 33.5 ± 12.7 years. The prevalence of eGFR< 60 was 0.2% (95% confidence interval (95% CI) 0.1, 0.9); the prevalence of eGFR< 90 was 5% (95% CI =3.2, 6.3). We observed a higher prevalence in the rural population (5% (3.6, 7.8)), versus urban (3% (1.4, 6.7)). Age and BMI were associated with reduced eGFR< 90 [Odds ratio (OR) (95%CI) =3.59 (2.58, 5.21) per ten-year increment]; [OR (95%CI) =2.01 (1.27, 3.43) per 5 kg/m2 increment] respectively. No increased risk of eGFR < 90 was observed for rural participants [OR (95%CI) =1.75 (0.50, 6.30)].

CONCLUSIONS

Reduced kidney function consistent with the definition of CKDu is not common in the areas of Malawi sampled, compared to that observed in other tropical or sub-tropical countries in Central America and South Asia. Reduced eGFR< 90 was related to age, BMI, and was more common in rural areas. These findings are important as they contradict some current hypothesis that CKDu is endemic across tropical and sub-tropical countries. This study has enabled standardized comparisons of impaired kidney function between and within tropical/subtropical regions of the world and will help form the basis for further etiological research, surveillance strategies, and the implementation and evaluation of interventions.

Using Bayesian Latent Gaussian Graphical Models to Infer Symptom Associations in Verbal Autopsies

Learning dependence relationships among variables of mixed types provides insights in a variety of scientific settings and is a well-studied problem in statistics. Existing methods, however, typically rely on copious, high quality data to accurately learn associations. In this paper, we develop a method for scientific settings where learning dependence structure is essential, but data are sparse and have a high fraction of missing values. Specifically, our work is motivated by survey-based cause of death assessments known as verbal autopsies (VAs). We propose a Bayesian approach to characterize dependence relationships using a latent Gaussian graphical model that incorporates informative priors on the marginal distributions of the variables. We demonstrate such information can improve estimation of the dependence structure, especially in settings with little training data. We show that our method can be integrated into existing probabilistic cause-of-death assignment algorithms and improves model performance while recovering dependence patterns between symptoms that can inform efficient questionnaire design in future data collection.

Self-reported disability in relation to mortality in rural Malawi: a longitudinal study of over 16 000 adults

OBJECTIVES

We investigated whether self-reported disability was associated with mortality in adults in rural Malawi.

SETTING

Karonga Health and Demographic Surveillance Site (HDSS), Northern Malawi.

PARTICIPANTS

All adults aged 18 and over residing in the HDSS were eligible to participate. During annual censuses in 2014 and 2015, participants were asked if they experienced difficulty in any of six functional domains and were classified as having disabilities if they reported 'a lot of difficulty' or 'can't do at all' in any domain. Mortality data were collected until 31 December 2017. 16 748 participants (10 153 women and 6595 men) were followed up for a median of 29 months.

PRIMARY AND SECONDARY OUTCOME MEASURES

We used Poisson regression to examine the relationship between disability and all-cause mortality adjusting for confounders. We assessed whether this relationship altered in the context of obesity, hypertension, diabetes or HIV. We also evaluated whether mortality from non-communicable diseases (NCD) was higher among people who had reported disability, as determined by verbal autopsy.

RESULTS

At baseline, 7.6% reported a disability and the overall adult mortality rate was 9.1/1000 person-years. Adults reporting disability had an all-cause mortality rate 2.70 times higher than those without, and mortality rate from NCDs 2.33 times higher than those without.

CONCLUSIONS

Self-reported disability predicts mortality at all adult ages in rural Malawi. Interventions to improve access to healthcare and other services are needed.

Assessing the costs and efficiency of HIV testing and treatment services in rural Malawi: implications for future "test and start" strategies

BACKGROUND

Reaching the 90-90-90 targets requires efficient resource use to deliver HIV testing and treatment services. We investigated the costs and efficiency of HIV services in relation to HIV testing yield in rural Karonga District, Malawi.

METHODS

Costs of HIV services were measured over 12 months to September 2017 in five health facilities, drawing on recognised health costing principles. Financial and economic costs were collected in Malawi Kwacha and United States Dollars (US$). Costs were calculated using a provider perspective to estimate average annual costs (2017 US$) per HIV testing episode, per HIV-positive case diagnosed, and per patient-year on antiretroviral therapy (ART), by facility. Costs were assessed in relation to scale of operation and facility-level annual HIV positivity rate. A one-way sensitivity analysis was undertaken to understand how staffing levels and the HIV positivity rate affected HIV testing costs.

RESULTS

HIV testing episodes per day and per full-time equivalent HIV health worker averaged 3.3 (range 2.0 to 5.7). The HIV positivity rate averaged 2.4% (range 1.9 to 3.7%). The average cost per testing episode was US$2.85 (range US$1.95 to US$8.55), and the average cost per HIV diagnosis was US$116.35 (range US$77.42 to US$234.11), with the highest costs found in facilities with the lowest daily number of tests and lowest HIV yield respectively. The mean facility-level cost per patient-year on ART was approximately US$100 (range US$90.67 to US$115.42). ART drugs were the largest cost component averaging 71% (range 55 to 76%). The cost per patient-year of viral load tests averaged US$4.50 (range US$0.52 to US$7.00) with cost variation reflecting differences in the tests to ART patient ratio across facilities.

CONCLUSION

Greater efficiencies in HIV service delivery are possible in Karonga through increasing daily testing episodes among existing health workers or allocating health workers to tasks in addition to testing. Costs per diagnosis will increase as yields decline, and therefore, encouraging targeted testing strategies that increase yield will be more efficient. Given the contribution of drug costs to per patient-year treatment costs, it is critical to preserve the life-span of first-line ART regimens, underlining the need for continuing adherence support and regular viral load monitoring.

Direct maternal deaths attributable to HIV in the era of antiretroviral therapy: evidence from three population-based HIV cohorts with verbal autopsy

OBJECTIVE

To assess whether HIV is associated with an increased risk of mortality from direct maternal complications.

DESIGN

Population-based cohort study using data from three demographic surveillance sites in Eastern and Southern Africa.

METHODS

We use verbal autopsy data, with cause of death assigned using the InSilicoVA algorithm, to describe the association between HIV and direct maternal deaths amongst women aged 20-49 years. We report direct maternal mortality rates by HIV status, and crude and adjusted rate ratios comparing HIV-infected and uninfected women, by study site and by ART availability. We pool the study-specific rate ratios using random-effects meta-analysis.

RESULTS

There was strong evidence that HIV increased the rate of direct maternal mortality across all the study sites in the period ART was widely available, with the rate ratios varying from 4.5 in Karonga, Malawi [95% confidence interval (CI) 1.6-12.6] to 5.2 in Kisesa, Tanzania (95% CI 1.7-16.1) and 5.9 in uMkhanyakude, South Africa (95% CI 2.3-15.2) after adjusting for sociodemographic confounders. Combining these adjusted results across the study sites, we estimated that HIV-infected women have 5.2 times the rate of direct maternal mortality compared with HIV-uninfected women (95% CI 2.9-9.5).

CONCLUSION

HIV-infected women face higher rates of mortality from direct maternal causes, which suggests that we need to improve access to quality maternity care for these women. These findings also have implications for the surveillance of HIV/AIDS-related mortality, as not all excess mortality attributable to HIV will be explicitly attributed to HIV/AIDS on the basis of a verbal autopsy interview.

Investigating associations between rural-to-urban migration and cardiometabolic disease in Malawi: a population-level study

Background

The extent to which rural-to-urban migration affects risk for cardiometabolic diseases (CMD) in Africa is not well understood. We investigated prevalence and risk for obesity, diabetes, hypertension and precursor conditions by migration status.

Methods

In a cross-sectional survey in Malawi (February 2013–March 2017), 13 903 rural, 9929 rural-to-urban migrant and 6741 urban residents (≥18 years old) participated. We interviewed participants, measured blood pressure and collected anthropometric data and fasting blood samples to estimate population prevalences and odds ratios, using negative binomial regression, for CMD, by migration status. In a sub-cohort of 131 rural–urban siblings-sets, migration-associated CMD risk was explored using conditional Poisson regression.

Results

In rural, rural-to-urban migrant and urban residents, prevalence estimates were; 8.9, 20.9 and 15.2% in men and 25.4, 43.9 and 39.3% in women for overweight/obesity; 1.4, 2.9 and 1.9% in men and 1.5, 2.8 and 1.7% in women for diabetes; and 13.4, 18.8 and 12.2% in men and 13.7, 15.8 and 10.2% in women for hypertension. Rural-to-urban migrants had the greatest risk for hypertension (adjusted relative risk for men 1.18; 95% confidence interval 1.04–1.34 and women 1.17: 95% confidence interval 1.05–1.29) and were the most screened, diagnosed and treated for CMD, compared with urban residents. Within sibling sets, rural-to-urban migrant siblings had a higher risk for overweight and pre-hypertension, with no evidence for differences by duration of stay.

Conclusions

Rural-to-urban migration is associated with increased CMD risk in Malawi. In a poor country experiencing rapid urbanization, interventions for the prevention and management of CMD, which reach migrant populations, are needed.

Adolescent Fertility Is Lower than Expected in Rural Areas: Results from 10 African HDSS

The adolescent birth rate (ABR) is an important indicator of maternal health, adolescent sexual health, and gender equity; it remains high in sub-Saharan Africa. While Demographic and Health Surveys (DHS) are the main source of ABR estimates, Health and Demographic Surveillance Systems (HDSS) also produce ABRs. Studies are lacking, however, to assess the ease of access and accuracy of HDSS ABR measures. In this paper, we use birth and exposure data from 10 HDSS in six African countries to compute local ABRs and compare these rates to DHS regional rates where the HDSS sites are located, standardizing by education and place of residence. In rural HDSS sites, the ABR measure is on average 44 percent lower than the DHS measure, after controlling for education and place of residence. Strong temporary migration of childless young women out of rural areas and different capacities in capturing temporarily absent women in the DHS and HDSS could explain this discrepancy. Further comparisons based on more strictly similar populations and measures seem warranted.

Implications of HIV treatment policies on the health workforce in rural Malawi and Tanzania between 2013 and 2017: Evidence from the SHAPE-UTT study

Effective implementation of policies for expanding antiretroviral therapy (ART) requires a well-trained and adequately staffed workforce. Changes in national HIV workforce policies, health facility practices, and provider experiences were examined in rural Malawi and Tanzania between 2013 and 2017. In both countries, task-shifting and task-sharing policies were explicit by 2013. In facilities, the cadre mix of providers varied by site and changed over time, with a higher and growing proportion of lower cadre staff in the Malawi site. In Malawi, the introduction of lay counsellors was perceived to have eased the workload of other providers, but lay counsellors reported inadequate support. Both countries had guidance on the minimum numbers of personnel required to deliver HIV services. However, patient loads per provider increased in both settings for HIV tests and visits by ART patients and were not met with corresponding increases in provider capacity in either setting. Providers reported this as a challenge. Although increasing patient numbers bodes well for achieving universal antiretroviral therapy coverage, the quality of care may be undermined by increased workloads and insufficient provider training. Task-shifting strategies may help address workload concerns, but require careful monitoring, supervision and mentoring to ensure effective implementation.

Assessing the validity of and factors that influence accurate self-reporting of HIV status after testing: a population-based study

OBJECTIVES

To assess the validity of self-reported HIV status, and investigate factors that influence accurate reporting of HIV-positive status, in a population tested and informed of their HIV test result.

DESIGN

Prospective cohort study.

METHODS

We compared self-reported HIV status with biomarker-confirmed HIV test status among participants of Karonga Health and Demographic Surveillance Site in rural northern Malawi. We linked information on HIV test results to subsequent self-reported HIV status, and calculated sensitivity, specificity, positive predictive value and negative predictive value for self-reported HIV status (considered as a diagnostic test). We used Poisson regression with robust variance estimators to examine predictors of accurate self-reporting of HIV-positive status.

RESULTS

Among 17 445 adults who tested for HIV, were recorded as having received their HIV test results, and had a subsequent self-reported HIV status between 2007 and 2018: positive predictive value of self-reported HIV status was 98.0% (95% confidence interval: 97.3-98.7); negative predictive value was 98.3 (98.1-98.5); sensitivity was 86.1% (84.5-87.7); and specificity was 99.8% (99.7-99.9). Among true HIV-positive people, those who were younger, interviewed in community settings, and had tested for HIV longer ago were more likely to misreport their HIV-positive status.

CONCLUSION

In this setting, self-report provides good estimates of test-detected HIV prevalence, suggesting that it can be used when HIV test results are not available. Despite frequent HIV testing, younger people and those interviewed in community settings were less likely to accurately report their HIV-positive status. More research on barriers to self-reporting of HIV status is needed in these subgroups.

Ability of verbal autopsy data to detect deaths due to uncontrolled hyperglycaemia: testing existing methods and development and validation of a novel weighted score

OBJECTIVES

Verbal autopsy (VA) is a useful tool to ascertain cause of death where no other mechanisms exist. We aimed to assess the utility of VA data to ascertain deaths due to uncontrolled hyperglycaemia and to develop a weighted score (WS) to specifically identify cases. Cases were identified by a study or site physician with training in diabetes. These diagnoses were also compared with diagnoses produced by a standard computer algorithm (InterVA-4).

SETTING

This study was done using VA data from the Health and Demographic Survey sites in Agincourt in rural South Africa. Validation of the WS was done using VA data from Karonga in Malawi.

PARTICIPANTS

All deaths from ages 1 to 49 years between 1992 and 2015 and between 2002 and 2016 from Agincourt and Karonga, respectively. There were 8699 relevant deaths in Agincourt and 1663 in Karonga.

RESULTS

Of the Agincourt deaths, there were 77 study physician classified cases and 58 computer algorithm classified cases. Agreement between study physician classified cases and computer algorithm classified cases was poor (Cohen's kappa 0.14). Our WS produced a receiver operator curve with area under the curve of 0.952 (95% CI 0.920 to 0.985). However, positive predictive value (PPV) was below 50% when the WS was applied to the development set and the score was dominated by the necessity for a premortem diagnosis of diabetes. Independent validation showed the WS performed reasonably against site physician classified cases with sensitivity of 86%, specificity of 99%, PPV of 60% and negative predictive value of 99%.

CONCLUSION

Our results suggest that widely used VA methodologies may be missing deaths due to uncontrolled hyperglycaemia. Our WS may offer improved ability to detect deaths due to uncontrolled hyperglycaemia in large populations studies where no other means exist.

A comparison of all-cause and cause-specific mortality by household socioeconomic status across seven INDEPTH network health and demographic surveillance systems in sub-Saharan Africa

Background: Understanding socioeconomic disparities in all-cause and cause-specific mortality can help inform prevention and treatment strategies.

Objectives: To quantify cause-specific mortality rates by socioeconomic status across seven health and demographic surveillance systems (HDSS) in five countries (Ethiopia, Kenya, Malawi, Mozambique, and Nigeria) in the INDEPTH Network in sub-Saharan Africa.

Methods: We linked demographic residence data with household survey data containing living standards and education information we used to create a poverty index. Person-years lived and deaths between 2003 and 2016 (periods varied by HDSS) were stratified in each HDSS by age, sex, year, and number of deprivations on the poverty index (0–8). Causes of death were assigned to each death using the InterVA-4 model based on responses to verbal autopsy questionnaires. We estimated rate ratios between socioeconomic groups (2–4 and 5–8 deprivations on our poverty index compared to 0–2 deprivations) for specific causes of death and calculated life expectancy for the deprivation groups.

Results: Our pooled data contained almost 3.5 million person-years of observation and 25,038 deaths. All-cause mortality rates were higher among people in households with 5–8 deprivations on our poverty index compared to 0–2 deprivations, controlling for age, sex, and year (rate ratios ranged 1.42 to 2.06 across HDSS sites). The poorest group had consistently higher death rates in communicable, maternal, neonatal, and nutritional conditions (rate ratios ranged 1.34–4.05) and for non-communicable diseases in several sites (1.14–1.93). The disparities in mortality between 5–8 deprivation groups and 0–2 deprivation groups led to lower life expectancy in the higher-deprivation groups by six years in all sites and more than 10 years in five sites.

Conclusions: We show large disparities in mortality on the basis of socioeconomic status across seven HDSS in sub-Saharan Africa due to disparities in communicable disease mortality and from non-communicable diseases in some sites. Life expectancy gaps between socioeconomic groups within sites were similar to the gaps between high-income and lower-middle-income countries. Prevention and treatment efforts can benefit from understanding subpopulations facing higher mortality from specific conditions.

Lusting, learning and lasting in school: sexual debut, school performance and dropout among adolescents in primary schools in Karonga district, northern Malawi

Age at sexual debut is known to have implications for future sexual behaviours and health outcomes, including HIV infection, early pregnancy and maternal mortality, but may also influence educational outcomes. Longitudinal data on schooling and sexual behaviour from a demographic surveillance site in Karonga district, northern Malawi, were analysed for 3153 respondents between the ages of 12 and 25 years to examine the association between sexual debut and primary school dropout, and the role of prior school performance. Time to dropout was modelled using the Fine and Gray survival model to account for the competing event of primary school completion. To deal with the time-varying nature of age at sexual debut and school performance, models were fitted using landmark analyses. Sexual debut was found to be associated with a five-fold increase in rate of subsequent dropout for girls and a two-fold increase in dropout rate for boys (adjusted hazard ratio [aHR] of 5.27, CI 4.22-6.57, and 2.19, CI 1.77-2.7, respectively). For girls who were sexually active by age 16, only 16% ultimately completed primary schooling, compared with 70% aged 18 or older at sexual debut. Prior to sexual debut, girls had primary completion levels similar to those of boys. The association between sexual debut and school dropout could not be explained by prior poor school performance: the effect of sexual debut on dropout was as strong among those who were not behind in school as among those who were overage for their school grade. Girls who were sexually active were more likely to repeat a grade, with no effect being seen for boys. Pathways to dropout are complex and may differ for boys and girls. Interventions are needed to improve school progression so children complete primary school before sexual debut, and to improve sex education and contraception provision.

Pneumococcal carriage in households in Karonga District, Malawi, before and after introduction of 13-valent pneumococcal conjugate vaccination

BACKGROUND

Thirteen-valent pneumococcal conjugate vaccine (PCV13) was introduced in Malawi in November 2011 and is offered to infants at 6, 10 and 14 weeks of age as part of routine immunisation. PCV13 is expected to reduce vaccine type (VT) nasopharyngeal carriage, leading to reduced transmission and herd protection.

METHODS

We compared pneumococcal carriage in rural Karonga District, Malawi, pre-vaccine in 2009-2011 and post-vaccine in 2014 using a combination of cross-sectional and longitudinal analyses. Nasopharyngeal swabs were collected from a cohort of mother-infant pairs and household members <16 years. Pneumococci from 2009 to 2011 were serogrouped using latex agglutination and serotyped by Quellung reaction. In 2014, latex agglutination was used for both steps. Carriage prevalence ratios using prevalence data from before and after vaccine introduction were calculated by log-binomial regression, adjusted for age, seasonality and household composition. Participating infants in 2014 received PCV13 as part of routine immunisation.

RESULTS

VT carriage prior to PCV-13 introduction was 11.4%, 45.1%, 28.2%, 21.2% and 6.6% for 6-week old infants, 18-week old infants, children 1-4 years, children 5-15 years and mothers, respectively. After vaccine introduction, VT carriage decreased among vaccinated 18-week old infants (adjusted prevalence ratio 0.24 (95%CI 0.08-0.75)), vaccinated children 1-4 years (0.54 (0.33-0.88)), unvaccinated children 5-15 years (0.37 (0.17-0.78)) and mothers (0.34 (0.15-0.79)). No decrease in VT carriage was observed for 6-week old infants too young to be vaccinated (1.07 (0.38-3.02)) and PCV-13 ineligible children 1-4 years (0.84 (0.53-1.33)). Non-VT carriage increased only among vaccinated children 1-4 years (1.58 (1.21-2.06)).

CONCLUSIONS

There is evidence of reduced VT pneumococcal carriage three years after vaccine introduction in this rural Malawian population with good vaccine coverage using a 3 + 0 schedule. However carriage was sustained among 6-week-old infants and PCV13 ineligible 1-4 year olds, and there was some indication of serotype replacement in vaccinated 1-4 year olds.

Implementing electronic data capture at a well-established health and demographic surveillance site in rural northern Malawi

This article aims to assess multiple issues of resources, staffing, local opinion, data quality, cost, and security while transitioning to electronic data collection (EDC) at a long-running community research site in northern Malawi. Levels of missing and error fields, delay from data collection to availability, and average number of interviews per day were compared between EDC and paper in a complex, repeated annual household survey. Three focus groups with field and data staff with experience using both methods, and in-depth interviews with participants were carried out. Cost for each method were estimated and compared. Missing data was more common on paper questionnaires than on EDC, and a similar number were carried out per day. Fieldworkers generally preferred EDC, but data staff feared for their employment. Most respondents had no strong preference for a method. The cost of the paper system was estimated to be higher than using EDC. The existing infrastructure and technical expertise could be adapted to using EDC, but changes have an impact on data processing jobs as fewer, and better qualified staff are required. EDC is cost-effective, and, for a long-running site, may offer further savings, as devices can be used in multiple studies and perform several other functions. EDC is accepted by fieldworkers and respondents, has good levels of quality and timeliness, and security can be maintained. EDC is well-suited for use in a well-established research site using and developing existing infrastructure and expertise.

Impact of monovalent rotavirus vaccine on diarrhoea-associated post-neonatal infant mortality in rural communities in Malawi: a population-based birth cohort study

BACKGROUND

Rotavirus is a major contributor to child mortality. The effect of rotavirus vaccine on diarrhoea mortality has been estimated in middle-income but not low-income settings, where mortality is high and vaccine effectiveness in reducing admissions to hospital is lower. Empirical population-based mortality studies have not been done in any setting. Malawi introduced monovalent rotavirus vaccine (RV1) in October, 2012. We aimed to investigate the impact and effectiveness of the RV1 vaccine in reducing diarrhoea-associated mortality in infants aged 10-51 weeks.

METHODS

In this population-based cohort study, we included infants born between Jan 1, 2012, and June 1, 2015, in Mchinji, Central Malawi and analysed data on those surviving 10 weeks. Individual vaccination status was extracted from caregiver-held records or report at home visits at 4 months and 1 year of age. Survival to 1 year was confirmed at home visit, or cause of death ascertained by verbal autopsy. We assessed impact (1 minus mortality rate ratio following vs before vaccine introduction) using Poisson regression. Among vaccine-eligible infants (born from Sept 17, 2012), we assessed effectiveness (1 minus hazard ratio) using Cox regression.

FINDINGS

Between Jan 1, 2012, and June 1, 2015, we recruited 48 672 livebirths in Mchinji, among whom 38 518 were vaccine-eligible and 37 570 survived to age 10 weeks. Two-dose versus zero-dose effectiveness analysis included 28 141 infants, of whom 101 had diarrhoea-associated death before 1 year of age. Diarrhoea-associated mortality declined by 31% (95% CI 1-52; p=0·04) after RV1 introduction. Effectiveness against diarrhoea-mortality was 34% (95% CI -28 to 66; p=0·22).

INTERPRETATION

RV1 was associated with substantial reduction in diarrhoea-associated deaths among infants in this rural sub-Saharan African setting. These data add considerable weight to evidence showing the impact of rotavirus vaccine programmes.

FUNDING

Wellcome Trust and GlaxoSmithKline Biologicals.

Early school failure predicts teenage pregnancy and marriage: A large population-based cohort study in northern Malawi

Background

School dropout has been linked to early pregnancy and marriage but less is known about the effect of school performance. We aimed to assess whether school performance influenced age at sexual debut, pregnancy and marriage, and from what age school drop-out and performance were associated with these later life events.

Methods

Data from 2007–2016 from a demographic surveillance site in northern Malawi with annual updating of schooling status and grades, and linked sexual behaviour surveys, were analysed to assess the associations of age-specific school performance (measured as age-for-grade) and status (in or out of school) on subsequent age at sexual debut, pregnancy and marriage. Landmark analysis with Cox regression was used to estimate hazard ratios of sexual debut, pregnancy and marriage by schooling at selected (landmark) ages, controlling for socio-economic factors.

Results

Information on at least one outcome was available for >16,000 children seen at ages 10–18. Sexual debut was available on a subset aged ≥15 by 2011. For girls, being out of school was strongly associated with earlier sexual debut, pregnancy and marriage. For example, using schooling status at age 14, compared to girls in primary, those who had dropped out had adjusted hazard ratios of subsequent sexual debut, pregnancy and marriage of 5.39 (95% CI 3.27–8.86), 2.39 (1.82–3.12), and 2.76 (2.08–3.67) respectively. For boys, the equivalent association with sexual debut was weak, 1.92 (0.81–4.55), but that with marriage was strong, 3.74 (2.28–6.11), although boys married later. Being overage-for-grade was not associated with sexual debut for girls or boys. For girls, being overage-for-grade from age 10 was associated with earlier pregnancy and marriage (e.g. adjusted hazard ratio 2.84 (1.32–6.17) for pregnancy and 3.19 (1.47–6.94) for marriage, for those ≥3 years overage compared to those on track at age 10). For boys, overage-for-grade was associated with earlier marriage from age 12, with stronger associations at older ages (e.g. adjusted hazard ratio 2.41 (1.56–3.70) for those ≥3 years overage compared to those on track at age 14). For girls ≥3 years overage at age 14, 39% were pregnant before they were 18, compared to 18% of those who were on track. The main limitation was the use of reported ages of sexual debut, pregnancy and marriage.

Conclusions

School progression at ages as young as 10 can predict teenage pregnancy and marriage, even after adjusting for socio-economic factors. Early education interventions may reduce teenage pregnancy and marriage as well as improving learning.

Glycated haemoglobin A1c (HbA1c) for detection of diabetes mellitus and impaired fasting glucose in Malawi: a diagnostic accuracy study

OBJECTIVES

To examine the accuracy of glycated haemoglobin A_1c (HbA_1c) in detecting type 2 diabetes and impaired fasting glucose among adults living in Malawi.

DESIGN

A diagnostic validation study of HbA_1c. Fasting plasma glucose (FPG) ≥7.0 mmol/L was the reference standard for type 2 diabetes, and FPG between 6.1 and 6.9 mmol/L as impaired fasting glucose.

PARTICIPANTS

3645 adults (of whom 63% were women) recruited from two demographic surveillance study sites in urban and rural Malawi. This analysis excluded those who had a previous diagnosis of diabetes or had history of taking diabetes medication.

RESULTS

HbA_1c demonstrated excellent validity to detect FPG-defined diabetes, with an area under the receiver operating characteristic (AUROC) curve of 0.92 (95% CI 0.90 to 0.94). At HbA_1c ≥6.5% (140 mg/dL), sensitivity was 78.7% and specificity was 94.0%. Subgroup AUROCs ranged from 0.86 for participants with anaemia to 0.94 for participants in urban Malawi. There were clinical and metabolic differences between participants with true diabetes versus false positives when HbA_1c was ≥6.5% (140 mg/dL).

CONCLUSIONS

The findings from this study provide justification to use HbA_1c to detect type 2 diabetes. As HbA_1c testing is substantially less burdensome to patients than either FPG testing or oral glucose tolerance testing, it represents a useful option for expanding access to diabetes care in sub-Saharan Africa.

Sustained 10-year gain in adult life expectancy following antiretroviral therapy roll-out in rural Malawi: July 2005 to June 2014

Background: Improved life expectancy in high HIV prevalence populations has been observed since antiretroviral therapy (ART) scale-up. However, it is unclear if the benefits are sustained, and the mortality among HIV-positive individuals not (yet) on ART is not well described. We assessed temporal change in mortality over 9 years in rural Malawi.

Methods: Within a demographic surveillance site in northern rural Malawi, we combined demographic, HIV and ART uptake data. We calculated life expectancy using Kaplan-Meier estimates, and compared mortality rates and rate ratios using Poisson regression, by period of ART availability (July 2005–June 2008, July 2008–June 2011 and July 2011–June 2014).

Results: Among 32 664 individuals there were 1424 deaths; 1930 individuals were known HIV-positive, of whom 1382 started ART. Overall, life expectancy at age 15 years increased by 10 years within 5 years of ART introduction, and plateaued. Age-standardized adult mortality rates declined from 11.3/1000 to 7.5/1000 person-years between the first and last time period. In July 2011-June 2014 compared with July 2005–June 2008, mortality declined in HIV-positive individuals on ART (rate ratio adjusted (aRR) for age, sex, location and education, 0.3; 95% confidence interval (CI) 0.2–0.5) and in those not (yet) on ART (aRR 0.3; 95%CI 0.1–0.5) but not in HIV-negative individuals (aRR 1.1; 95%CI 0.7–1.9).

Conclusions: Total population adult life expectancy increased toward that of HIV-negative individuals by 2011 and remained raised. The reduction in all-cause and HIV-related mortality in HIV-positive individuals not (yet) on ART suggests ART uptake is occurring at an earlier disease stage, particularly in women.

From kitchen to classroom: Assessing the impact of cleaner burning biomass-fuelled cookstoves on primary school attendance in Karonga district, northern Malawi

Household air pollution from burning solid fuels is responsible for an estimated 2.9 million premature deaths worldwide each year and 4.5% of global disability-adjusted life years, while cooking and fuel collection pose a considerable time burden, particularly for women and children. Cleaner burning biomass-fuelled cookstoves have the potential to lower exposure to household air pollution as well as reduce fuelwood demand by increasing the combustion efficiency of cooking fires, which may in turn yield ancillary benefits in other domains. The present paper capitalises on opportunities offered by the Cooking and Pneumonia Study (CAPS), the largest randomised trial of biomass-fuelled cookstoves on health outcomes conducted to date, the design of which allows for the evaluation of additional outcomes at scale. This mixed methods study assesses the impact of cookstoves on primary school absenteeism in Karonga district, northern Malawi, in particular by conferring health and time and resource gains on young people aged 5–18. The analysis combines quantitative data from 6168 primary school students with in-depth interviews and focus group discussions carried out among 48 students in the same catchment area in 2016. Negative binomial regression models find no evidence that the cookstoves affected primary school absenteeism overall [IRR 0.92 (0.71–1.18), p = 0.51]. Qualitative analysis suggests that the cookstoves did not sufficiently improve household health to influence school attendance, while the time and resource burdens associated with cooking activities—although reduced in intervention households—were considered to be compatible with school attendance in both trial arms. More research is needed to assess whether the cookstoves influenced educational outcomes not captured by the attendance measure available, such as timely arrival to school or hours spent on homework.

How does the association of general and central adiposity with glycaemia and blood pressure differ by gender and area of residence in a Malawian population: a cross-sectional study

BACKGROUND

In high-income settings, body mass index (BMI) and measures of central adiposity, such as waist-to-hip ratio (WHR) are associated with cardiometabolic risk, but evidence from low-income settings, particularly sub-Saharan Africa (SSA), is limited. We assessed whether there are differences between central and general adiposity in their associations with fasting glucose, diabetes, systolic and diastolic blood pressures and hypertension, and whether these associations differ with gender or rural/urban setting in Malawi.

METHODS

We used data from a population-based study of 27 880 Malawian adults aged  ≥18 years, from both rural and urban areas. We used age-standardized z-scores of the means of BMI and WHR to directly compare their associations with glycaemic and blood pressure outcomes.

RESULTS

Mean fasting glucose and blood pressure values and odds of hypertension increased linearly across fifths of BMI and WHR, with stronger associations with BMI. For both BMI and WHR, the associations with outcomes were stronger in urban versus rural residents. The association with diabetes was stronger in women than men, whereas for blood-pressure related outcomes a stronger association was seen in men.

CONCLUSIONS

BMI is more strongly associated with cardiometabolic risk in SSA, and might be a more useful measure than WHR, in this population. The greater positive association of adiposity with cardiometabolic outcomes in urban residents (where rates of overweight/obesity are already high) highlights the particular importance of addressing obesity within urban SSA populations.

Unintended Childbearing and Child Growth in Northern Malawi

Objective The study aims to assess whether unintended children experience slower growth than intended children. Methods We analysed longitudinal data linked to the Karonga Health and Demographic Surveillance Site collected over three rounds between 2008 and 2011 on women’s fertility intentions and anthropometric data of children. Using the prospective information on fertility intention we assessed whether unintended children are more likely to be stunted than intended children. We applied Propensity Score Matching technique to control for endogenous factors affecting both the probability that a family has an unwanted birth and a child with poor health outcomes. Results We found that 24 % of children from unwanted pregnancies were stunted compared with 18 % of mistimed pregnancies and 17 % of those from wanted pregnancies. However, these differences in probability of children being stunted, though in the expected direction, were not significant either for large or small families, after controlling for age. The number of children in the household was associated with stunting and boys were substantially more likely to be stunted than girls. Conclusion We found no significance difference in probability of being stunted by mother’s fertility intention.