No excess mortality detected in rural Bangladesh in 2020 from repeated surveys of a population of 81,000

Prabhat Barnwal, Yuling Yao, Yiqian Wang, Nishat Akter Juy, Shabib Raihan, Mohammad Ashraful Haque, Alexander van GeenmedRxiv 2021 · DOI

Abstract

Measuring the true impact of COVID-19 through excess mortality is challenging in low-income countries where civil registration systems are often incomplete. This study utilizes a repeated census of 16,054 households (81,164 individuals) across 135 villages in rural Bangladesh to estimate mortality changes during 2020. Using Bayesian multilevel regression and poststratification to account for survey non-response, we find that all-cause mortality in 2020 actually declined by 8% (95% CI, -21% to 7%) relative to 2019 levels. However, this lack of excess mortality occurred alongside a catastrophic economic downturn, with salaried breadwinners reporting a 40% decline in income. These findings suggest that while rural areas may have been spared the initial mortality surge seen in urban centers, the economic and nutritional costs of pandemic restrictions were substantial.

Introduction

In high-income nations, "excess mortality"—the gap between observed and expected deaths—has served as the gold standard for capturing the total toll of the COVID-19 pandemic, accounting for both direct viral deaths and indirect deaths caused by healthcare disruptions. In countries like Bangladesh, however, this metric is difficult to calculate. Only about one-third of deaths are officially recorded within the required 45-day window, creating a "scandal of invisibility" that leaves policymakers without the data necessary to balance public health interventions against economic costs.

By the end of the first year of the pandemic, official COVID-19 deaths in Bangladesh represented only 1% of typical annual mortality. Whether this low figure reflected a genuine trend or massive under-counting remained a subject of intense debate. This study provides direct evidence from a large-scale rural population to clarify the pandemic's impact during its first wave.

Methodology

The research team employed a "repeated census" approach, a methodology frequently used to estimate mortality in data-poor environments or conflict zones.

  1. Baseline Census: In early 2020, enumerators conducted an in-person, door-to-door census of 16,054 households in a 350 km² region northeast of Dhaka.
  2. Longitudinal Follow-up: The same households were contacted via telephone in May and November 2020. These surveys recorded changes in household composition (deaths), income levels, and food security.
  3. Statistical Adjustment: To address the 25% attrition rate (households unreachable by phone), the researchers used Bayesian multilevel regression and poststratification (MRP). Mortality rates were modeled by month, age, gender, and education level, then weighted to match the original population structure and compared against 2019 baseline data.

Key Findings

Contrary to the high excess mortality observed globally, the surveyed rural population in Bangladesh saw no increase in deaths. For the subset of the population contacted twice, 276 deaths were reported between February and October 2020, compared to 289 deaths for the same population and period in 2019. After adjusting for non-response, the estimated change in mortality was -8% (95% CI, -21% to 7%). Reported causes of death, such as stroke, heart disease, and cancer, remained stable, with no significant surge in respiratory-related fatalities.

The Economic-Health Paradox

While the mortality impact was minimal, the economic impact was severe. By May 2020, salaried workers reported a 40% drop in income, and business owners reported a 60% decline in profits. This economic shock translated directly into food insecurity: 25% of households reported being unable to obtain essential food items due to lost income during the initial lockdown period.

Discussion

Several factors may explain the lack of excess mortality in rural Bangladesh during 2020:

  • The "Rural Shield": The pandemic’s first wave may have been primarily an urban phenomenon, concentrated in high-density centers like Dhaka. Rural areas likely benefited from lower population density and better ventilation.
  • Mortality Substitution: The slight decline in overall mortality suggests that lockdown measures may have reduced deaths from other causes, such as road traffic accidents and the transmission of other infectious diseases like seasonal influenza.
  • Effectiveness of Restrictions: Early government-imposed restrictions on movement and trade likely limited the initial spread of the virus into the countryside.

However, these findings highlight a brutal trade-off. The measures that successfully kept the virus at bay in rural areas also pushed vulnerable populations toward extreme poverty. The data suggests that the "cost" of preventing COVID-19 deaths in these villages was a significant increase in nutritional and economic hardship.

Conclusion and Policy Implications

The study concludes that officially recorded COVID-19 deaths in 2020 likely reflected the reality of the rural population, where the pandemic's mortality impact was limited. However, the economic and nutritional toll was underestimated.

For future pandemic responses, policymakers should consider more geographically targeted restrictions—focusing on high-density urban areas—while simultaneously expanding social safety nets and financial support for rural populations to prevent humanitarian crises that may arise from lockdowns. Further research is needed to determine if subsequent waves, such as the 2021 Delta surge, eventually breached the "rural shield" observed in 2020.

Summary updated

Abstract

ABSTRACT Background Excess mortality has demonstrated under-counting of COVID-19 deaths in many countries but cannot be measured in low-income countries where civil registration is incomplete. Methods Enumerators conducted an in-person census of all 16,054 households in a sample of 135 villages within a 350 km 2 region of Bangladesh followed by a census conducted again in May and November 2020 over the phone. The date and cause of any changes in household composition, as well as changes in income and food availability, were recorded. For analysis, we stratify the mortality data by month, age, gender, and household education. Mortality rates were modeled by Bayesian multilevel regression and the strata aggregated to the population by poststratification. Results A total of 276 deaths were reported between February and the end of October 2020 for the subset of the population that could be contacted twice over the phone, slightly below the 289 deaths reported for the same population over the same period in 2019. After adjustment for survey non-response and poststratification, 2020 mortality changed by -8% (95% CI, -21% to 7%) relative to an annualized mortality of 6.1 per thousand in 2019. However, salaried breadwinners reported a 40% decline in income and businesses a 60% decline in profits in May 2020. Discussion All-cause mortality in the surveyed portion of rural Bangladesh was if anything lower in 2020 compared to 2019. Our findings suggest various restrictions imposed by the government limited the scale of the pandemic, although they need to be accompanied by expanded welfare programs. Key questions What is already known? Civil registry data from dozens of countries, where available, indicate gaps between official death counts attributed to COVID-19 and, usually, a larger increase in total mortality in 2020 compared to previous years. This approach is not available to gauge the impact of COVID-19 in countries such as Bangladesh where the civil registry system is slow and coverage incomplete. One year after the first COVID-19 case was reported in Bangladesh in 2020, the number of deaths attributed to COVID-19 was equivalent to 1% of annual mortality in previous years. Whether this low figure compared to many other countries is an accurate reflection of the situation or is distorted by massive under-counting has been much debated, albeit on the basis of little direct evidence. The lack of accurate mortality data has made it only more difficult for policy makers to balance the public health benefit of lockdowns and similar measures relative to the well-documented economic costs and hardship imposed by such measures on poor households in particular. A PubMed search conducted on May 4, 2021 under (Bangladesh[Title/Abstract]) AND (excess mortality[Title/Abstract]) limited to 2020-21 did not yield a single relevant study. What are the new findings? By conducting of repeated census of a large rural population over the course of 2020, once in person and twice over the phone, we document if anything a slight decline in mortality across a rural area of Bangladesh compared to 2019. We also place an upper limit on the level of under-reporting at the national level that is consistent with our observations. At the same time, interviewed households reported a large and sustained drop in income as well as reduced access to food. What do the new findings imply? The impact of the pandemic on mortality was thankfully limited in rural study area of Bangladesh in 2020. This suggests that officially recorded COVID-19 deaths may have been contributed largely by the urban population, about a third of the country overall. At the same time, the economic and nutritional impact of restrictions on trade and movement was substantial and probably underestimated in the rural population. As cases surge again, as they did in March–April 2021, policy makers may want to consider limiting strict restrictions to urban areas while expanding a financial support throughout the country.
BibTeX
@misc{barnwal2021no,
  author = {Prabhat Barnwal and Yuling Yao and Yiqian Wang and Nishat Akter Juy and Shabib Raihan and Mohammad Ashraful Haque and Alexander van Geen},
  title = {{No excess mortality detected in rural Bangladesh in 2020 from repeated surveys of a population of 81,000}},
  year = {2021},
  doi = {10.1101/2021.05.07.21256865},
  howpublished = {medRxiv},
}
Figure