Three-year outcomes of post-acute sequelae of COVID-19

Post-acute sequelae of COVID-19 (PASC), also known as Long COVID, encompasses long-term health effects affecting nearly every organ system. While previous research has characterized PASC risks up to two years post-infection, longer-term follow-up is needed to fully understand the long-term health trajectories and inform patient care. This knowledge gap is crucial as it remains unclear whether risks persist beyond two years, or if new latent risks emerge. This study aims to comprehensively assess the risks and burdens of PASC three years after infection, distinguishing between non-hospitalized and hospitalized individuals during the acute phase. The US Department of Veterans Affairs (VA) national health care databases provide a unique opportunity to conduct this large-scale longitudinal study, given their comprehensive data and nationwide reach. The study period, spanning March to December 2020, precedes the widespread availability of COVID-19 vaccines and antivirals, allowing for an assessment of PASC in the context of the ancestral SARS-CoV-2 virus.

Numerous studies have explored the risk trajectories of various conditions following SARS-CoV-2 infection, with follow-up periods extending to one and two years. These studies revealed that while some risks diminish after the first year, many persist at two years, especially among hospitalized individuals. Approximately 25% of the two-year cumulative burden of disability and disease stems from the second year alone. However, the lack of longer-term studies leaves a significant gap in our understanding of the long-term consequences, specifically whether risks continue into the third year and if previously unseen risks manifest. This study aims to address these gaps by conducting a three-year follow-up, offering a more complete picture of PASC's long-term impact.

This retrospective cohort study utilized data from the US Department of Veterans Affairs (VA) national health care databases. The study included 135,161 US veterans who survived the initial 30 days following COVID-19 infection between March and December 2020, and 5,206,835 controls with no evidence of SARS-CoV-2 infection. The cohorts were followed longitudinally for three years. Hospitalization status during the acute phase of infection (non-hospitalized versus hospitalized) defined mutually exclusive groups for analysis. The study examined the risks and burdens of death and pre-specified PASC outcomes, including sequelae aggregated by organ system and overall PASC. The primary outcomes were death and incident PASC. A total of 80 pre-specified individual outcomes, based on previous research, were grouped into 10 organ systems: cardiovascular, coagulation and hematological, fatigue, gastrointestinal, kidney, mental health, metabolic, musculoskeletal, neurological, and pulmonary. Disability-adjusted life years (DALYs) were calculated using Global Burden of Disease (GBD) methodologies to quantify the overall health burden. Inverse probability weighting was used to address potential confounding by balancing baseline characteristics between the COVID-19 groups and the control group. The risks of death and sequelae were estimated for three one-year periods post-infection: years 1, 2, and 3. Poisson regression models were used to analyze the incidence rate ratios (IRRs) and absolute rates. Sensitivity analyses were conducted to assess the robustness of the findings. A negative outcome control analysis, examining the association between COVID-19 and incident neoplasms, was performed to mitigate concerns about bias. The study was approved by the institutional review board of the VA St. Louis Health Care System.

The study revealed distinct patterns of risk and burden across the three years of follow-up for both non-hospitalized and hospitalized individuals. In non-hospitalized individuals, the increased risk of death disappeared after the first year. The risk of PASC decreased over the three years, but a residual risk remained, contributing to 41.2 sequelae per 1,000 persons and 9.6 DALYs per 1,000 persons in the third year. In contrast, hospitalized individuals showed a significantly higher risk of death throughout all three years, with a 29% increased risk and an excess burden of 8.16 deaths per 1,000 persons in the third year. While the risk of PASC declined, a substantial residual risk persisted in the third year, resulting in 252.8 sequelae per 1,000 persons and 90.0 DALYs per 1,000 persons. The cumulative DALYs over the three years were substantially higher among hospitalized individuals (766.2 per 1,000) compared to the non-hospitalized group (91.2 per 1,000), demonstrating an 8.4-fold difference. The analysis of sequelae by organ system revealed consistent patterns. The top five organ systems contributing to DALYs in both groups were cardiovascular, mental, neurologic, coagulation and hematologic, and kidney disorders. The negative outcome control analysis showed no significant association between COVID-19 and incident neoplasms.

This study extends previous findings by providing a three-year perspective on PASC risk trajectories. The sustained elevation of mortality risk and the substantial residual risk of PASC among hospitalized individuals underscore the long-term consequences of severe COVID-19. The observed reduction in risk over time aligns with patterns seen in other post-acute infection syndromes, suggesting a gradual abatement of acute effects. However, the persistent burden in hospitalized patients highlights the significance of disease severity as a predictor of long-term outcomes. The substantial burden of PASC among the non-hospitalized group, owing to the larger population with mild disease, emphasizes the need for strategies to mitigate long-term consequences even for those experiencing milder acute infections. These findings reinforce the importance of vaccination, antiviral therapies, and early detection and management of PASC to reduce the long-term health burden of COVID-19.

This three-year longitudinal study demonstrates a significant reduction in PASC risk over time, but a residual risk remains, particularly in hospitalized individuals. While non-hospitalized patients experience a diminishing burden, those hospitalized show persistently elevated risks of death and PASC, emphasizing the lasting impact of severe acute COVID-19. Further research should focus on understanding the mechanisms behind prolonged risk, developing effective preventive and therapeutic strategies, and addressing the significant burden of PASC across all severity levels.

The study population, primarily older, White male veterans, may limit generalizability. The pre-vaccine era enrollment may not reflect the experience of vaccinated individuals. Potential biases, including misclassification and residual confounding, cannot be entirely ruled out despite the use of advanced statistical methods. The pre-specified outcome set may not encompass all possible PASC manifestations. The study’s reliance on VA data may miss undiagnosed cases or those not captured in the VA system. Additionally, the use of GBD methodology for DALY calculations assumes consistent health burden coefficients regardless of the cause of the condition.