Stéphane Helleringer’s research while affiliated with New York University Abu Dhabi and other places

Background In countries with limited civil registration and vital statistics systems, assessing the impact of health crises requires precise retrospective mortality data. We tested whether calendar methods improve survey or census data on dates of recent household deaths registered in a Health and Demographic Surveillance System (HDSS). Methods Between April and June 2023, we randomized 578 households with HDSS-registered deaths in Guinea-Bissau to interviews by using (i) a standard questionnaire with close-ended questions about dates of deaths, or similar questionnaires supplemented with (ii) a physical calendar of local events printed on paper or (iii) a digital calendar implemented on tablets. We evaluated the accuracy of reported dates through record linkages to HDSS data. Results No deaths were reported in 11.8% of the 508 participating households. In other households (n = 448), informants reported 574 deaths since January 2020. Relative to the standard questionnaire, neither the physical calendar nor the digital calendar improved the proportion of deaths reported in the same month and year as recorded by using surveillance data. The physical and digital calendars reduced the share of missing data on dates of deaths (6.1% and 3.2%, respectively, versus 13.1% with the standard questionnaire). Reported dates of deaths obtained by using the digital calendar were more weakly correlated with surveillance data than those collected in other arms. Using the digital calendar also added 1.15 minutes to the data collection. Conclusion Digital calendars do not improve the reporting of dates of deaths in surveys or censuses. Further trials of the use of a physical calendar in retrospective interviews about recent household deaths are warranted.

Child mortality data in low- and middle-income countries are often derived from survey data prone to imprecision due to recall. Such imprecisions can affect the accuracy of date of birth (DOB) and age at death (AAD). Bandim Health Project runs two Health and Demographic Surveillance Systems (HDSSs) collecting prospective information on pregnancies, births, and deaths in Guinea-Bissau. In a survey, HDSS resident women were interviewed on their births and survival status of their children born within 5 years prior to the interview. We matched child records in the survey data to their corresponding records in the HDSS data based on sex, name, birth order, twinning status, and sibling names. We compared the risk and magnitude of displacement in DOB and AAD in regression models to identify risk factors for displacement. A total of 7679 out of 9960 survey-reported live births (78%) were matched to HDSS-reported births. Displacement in month or year of birth was observed for 9%, with a median displacement of 30 days backwards. Child death was associated with larger risk of displacement in DOB (adjusted risk ratio: 5.16 (95% confidence interval: 4.44–5.99)) and greater magnitude of displacement. A longer recall period was also associated with a greater risk of displacement. Among 339 children who had died, 68% had a discrepancy in AAD with 34% of these exceeding 1 week. Discrepancies were common and more likely for longer recall periods and children who had died. Hence, the interpretation of temporal mortality patterns from survey data may be compromised.

Supplementary methods of mortality surveillance are needed in settings with incomplete death registration. Local perceptions of mortality levels might be useful indicators of excess deaths. Early in the COVID-19 pandemic, we developed a survey question asking respondents to evaluate the recent frequency of funerals in their community relative to a pre-pandemic baseline. We asked this question of more than 400 residents of Karonga district in Malawi, who were interviewed up to five times during a panel survey conducted by mobile phone between June 2020 and May 2021. The proportion of respondents reporting more funerals than usual in their community increased in early 2021, indicating excess mortality not otherwise detected by case-based surveillance and only partially visible in reports of inpatient deaths at health facilities. Systems assessing changes in perceptions of mortality can be rapidly established during an epidemic and may help detect excess deaths at local levels.

We investigated behavioral responses to COVID–19 in Malawi, where a first wave of the pandemic occurred between June and August 2020. Contrary to many countries on the African continent, the Government of Malawi did not impose a lockdown or a stay-at-home order in response to the initial spread of SARS-CoV-2. We hypothesized that, in the absence of such requirements to restrict social interactions, individuals would primarily seek to reduce the risk of SARS-CoV-2 transmission during contacts, rather than reduce the extent of their social contacts. We analyzed 4 rounds of a panel survey spanning time periods before, during and after the first wave of the COVID-19 pandemic in Malawi. Five hundred and forty-three participants completed 4 survey interviews between April and November 2020. We found that the likelihood of attending various places and events where individuals work and/or socialize remained largely unchanged during that time. Over the same time frame, however, participants reported adopting on a large scale several behaviors that reduce the transmissibility of SARS-CoV-2 during contacts. The percentage of panel participants who reported practicing physical distancing thus increased from 9.8% to 47.0% in rural areas between April-May 2020 and June-July 2020, and from 11.4% to 59.4% in urban areas. The percentage of respondents who reported wearing a facial mask to prevent the spread of SARS-CoV-2 also increased, reaching 67.7% among rural residents in August-September 2020, and 89.6% among urban residents. The pace at which these behaviors were adopted varied between population groups, with early adopters of mask use more commonly found among more educated office workers, residing in urban areas. The adoption of mask use was also initially slower among women, but later caught up with mask use among men. These findings stress the importance of behavioral changes in containing future SARS-CoV-2 outbreaks in settings where access to vaccination remains low. They also highlight the need for targeted outreach to members of socioeconomic groups in which the adoption of protective behaviors, such as mask use, might be delayed.

Guidelines for conducting surveys by mobile phone calls in low- and middle-income countries suggest keeping interviews short (<20 minutes). The evidence supporting this recommendation is scant, even though limiting interview duration might reduce the amount of data generated by such surveys. We recruited nearly 2,500 mobile phone users in Malawi and randomly allocated them to 10-, 20-, or 30-minute phone interviews, all ending with questions on parental survival. Cooperation was high in all groups, and differences in completion rates were minimal. The extent of item nonresponse, age heaping, and temporal displacement of deaths in data on parental survival generally did not vary between study groups, but reports of maternal age at death were more reliable in longer interviews. Recommendations about the duration of mobile phone interviews might be too restrictive. They should not preclude additional modules, including ones on mortality, in mobile phone surveys conducted in LMICs.

La mortalité aux âges élevés reste peu documentée dans les pays où les systèmes d’enregistrement et les statistiques de l’état civil sont incomplets. Les enquêtes comportant des questions sur la survie des parents peuvent contribuer à combler cette lacune en fournissant des estimations de la mortalité des plus de 50 ans. Notre article évalue si des biais de sélection pourrait altérer la qualité de ces estimations. Nous analysons les données de trois observatoires de population et de santé regroupant 9 600 femmes et 8 500 hommes. Nous appliquons des modèles de Cox afin d’estimer la mortalité des femmes et des hommes âgés de 50 à 89 ans en fonction du nombre de leurs filles adultes (qui sont généralement les répondantes aux enquêtes). Les résultats indiquent que les personnes sans enfant vivant au moment de l’enquête font face à un risque de décès plus important. Les estimations de mortalité déduites des déclarations sur la survie des parents sont donc susceptibles de sous-estimer les niveaux de mortalité après 50 ans, même si aucun biais systématique n’apparaît clairement. La collecte de données sur la survie des parents constitue finalement un outil prometteur pour mesurer la mortalité aux âges élevés, étant donné le faible risque de biais de sélection.

Guidelines for conducting surveys by mobile phone calls in low- and middle-income countries suggest keeping interviews short (<20 minutes). The evidence supporting this recommendation is scant, even though limiting interview duration might reduce the amount of data generated by such surveys. We recruited nearly 2,500 mobile phone users in Malawi and randomly allocated them to 10-, 20-, or 30-minute phone interviews, all ending with questions on parental survival. Cooperation was high in all groups, and differences in completion rates were minimal. The extent of item nonresponse, age heaping, and temporal displacement of deaths in data on parental survival generally did not vary between study groups, but reports of maternal age at death were more reliable in longer interviews. Recommendations about the duration of mobile phone interviews might be too restrictive. They should not preclude additional modules, including ones on mortality, in mobile phone surveys conducted in LMICs.