In the early days of the COVID-19 pandemic, researchers struggled to grasp the rate of the virus’s spread and the number of related deaths. While hospitals tracked cases and deaths within their walls, the broader picture of mortality across communities remained frustratingly incomplete.
Policymakers and researchers quickly discovered a troubling pattern: Many deaths linked to the virus were never officially counted. A study analyzing data from over 3,000 U.S. counties between March 2020 and August 2022 found nearly 163,000 excess deaths from natural causes that were missing from official mortality records.
Excess deaths, meaning those that exceed the number expected based on historical trends, serve as a key indicator of underreported deaths during health crises. Many of these uncounted deaths were later tied to COVID-19 through reviews of medical records, death certificates and statistical modeling.
In addition, lack of real-time tracking for medical interventions during those early days slowed vaccine development by delaying insights into which treatments worked and how people were responding to newly circulating variants.
Five years since the beginning of COVID-19, new epidemics such as bird flu are emerging worldwide, and researchers are still finding it difficult to access the data about people’s deaths that they need to develop lifesaving interventions.
How can the U.S. mortality data system improve? I’m a technology infrastructure researcher, and my team and I design policy and technical systems to reduce inefficiency in health care and government organizations. By analyzing the flow of mortality data in the U.S., we found several areas of the system that could use updating.
A death record includes key details beyond just the fact of death, such as the cause, contributing conditions, demographics, place of death and sometimes medical history. This information is crucial for researchers to be able to analyze trends, identify disparities and drive medical advances.
Approximately 2.8 million death records are added to the U.S. mortality data system each year. But in 2022 – the most recent official count available – when the world was still in the throes of the pandemic, 3,279,857 deaths were recorded in the federal system. Still, this figure is widely considered to be a major undercount of true excess deaths from COVID-19.
In addition, real-time tracking of COVID-19 mortality data was severely lacking. This process involves the continuous collection, analysis and reporting of deaths from hospitals, health agencies and government databases by integrating electronic health records, lab reports and public health surveillance systems. Ideally, it provides up-to-date insights for decision-making, but during the COVID-19 pandemic, these tracking systems lagged and failed to generate comprehensive data.
Without comprehensive data on prior COVID-19 infections, antibody responses and adverse events, researchers faced challenges designing clinical trials to predict how long immunity would last and optimize booster schedules.
Such data is essential in vaccine development because it helps identify who is most at risk, which variants and treatments affect survival rates, and how vaccines should be designed and distributed. And as part of the broader U.S. vital records system, mortality data is essential for medical research, including evaluating public health programs, identifying health disparities and monitoring disease.
At the heart of the problem is the inefficiency of government policy, particularly outdated public health reporting systems and slow data modernization efforts that hinder timely decision-making. These long-standing policies, such as reliance on paper-based death certificates and disjointed state-level reporting, have failed to keep pace with real-time data needs during crises such as COVID-19.
These policy shortcomings lead to delays in reporting and lack of coordination between hospital organizations, state government vital records offices and federal government agencies in collecting, standardizing and sharing death records.
The U.S. mortality data system has been cobbled together through a disparate patchwork of state and local governments, federal agencies and public health organizations over the course of more than a century and a half. It has been shaped by advances in public health, medical record-keeping and technology. From its inception to the present day, the mortality data system has been plagued by inconsistencies, inefficiencies and tensions between medical professionals, state governments and the federal government.
The first national efforts to track information about deaths began in the 1850s when the U.S. Census Bureau started collecting mortality data as part of the decennial census. However, these early efforts were inconsistent, as death registration was largely voluntary and varied widely across states.
In the early 20th century, the establishment of the National Vital Statistics System brought greater standardization to mortality data. For example, the system required all U.S. states and territories to standardize their death certificate format. It also consolidated mortality data at the federal level, whereas mortality data was previously stored at the state level.
However, state and federal reporting remained fragmented. For example, states had no unifom timeline for submitting mortality data, resulting in some states taking months or even years to finalize and release death records. Local or state-level paperwork processing practices also remained varied and at times contradictory.
To begin to close gaps in reporting timelines to aid medical researchers, in 1981 the National Center for Health Statistics – a division of the Centers for Disease Control and Prevention – introduced the National Death Index. This is a centralized database of death records collected from state vital statistics offices, making it easier to access death data for health and medical research. The system was originally paper-based, with the aim of allowing researchers to track the deaths of study participants without navigating complex bureaucracies.
As time has passed, the National Death Index and state databases have become increasingly digital. The rise of electronic death registration systems in recent decades has improved processing speed when it comes to researchers accessing mortality data from the National Death Index. However, while the index has solved some issues related to gaps between state and federal data, other issues, such as high fees and inconsistency in state reporting times, still plague it.
With the Trump administration’s increasing removal of CDC public health datasets, it is unclear whether policy reform for mortality data will be addressed anytime soon.
Experts fear that the removal of CDC datasets has now set precedent for the Trump administration to cross further lines in its attempts to influence the research and data published by the CDC. The longer-term impact of the current administration’s public health policy on mortality data and disease response are not yet clear.
What is clear is that five years since COVID-19, the U.S. mortality tracking system remains unequipped to meet emerging public health crises. Without addressing these challenges, the U.S. may not be able to respond quickly enough to public health crises threatening American lives.
This article is republished from The Conversation, a nonprofit, independent news organization bringing you facts and trustworthy analysis to help you make sense of our complex world. It was written by: Dylan Thomas Doyle, University of Colorado Boulder
Read more: How records of life’s milestones help solve cold cases, pinpoint health risks and allocate public resources COVID-19 deaths in the US continue to be undercounted, research shows, despite claims of ‘overcounts’ Video: What the huge COVID-19 testing undercount in the US means
Dylan Thomas Doyle does not work for, consult, own shares in or receive funding from any company or organization that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.
