Prevalence of SARS-CoV-2 infection and immunity in a New York county in 2022 reveals frequent asymptomatic or undiagnosed infections
For every confirmed COVID case, at least one more person was infected and never reported it. Door-to-door testing reveals how badly we've been underestimating infections—and shows a better way to track disease outbreaks.
College of Health researcher(s)
College unit(s)
Abstract
Accurate and timely surveillance of SARS-CoV-2 prevalence and immunity is critical to local and national COVID-19 pandemic responses. Representative surveillance surveys reveal more accurate estimates of COVID-19 infection than other measures based on reported test results. Our main research objectives were (i) to provide local health department officials with prevalence estimates calculated from a representative sample to better inform their decision-making efforts in response to the COVID-19 pandemic and (ii) to identify characteristics associated with COVID-19 infections among high-risk groups. Three municipalities were sampled at one timepoint (February, April, or October 2022) using a 2-stage cluster sampling design. Participants provided anterior nares swabs, which were tested for SARS-CoV-2 with a RT-PCR and for nucleocapsid protein and receptor binding domain antibodies by multiplex Luminex assay. Participants completed a survey on socio-demographics, SARS-CoV-2 prevention behaviors and attitudes, and vaccination and infection history. A total of 233 individuals from 221 households provided anterior nares swabs, and 215 samples were linked to survey data. After adjusting for study design, the household prevalence of PCR-positive tests was less than 5%, but approximately half of the population had antibodies from a prior infection and most (81% to 92%) had antibodies from either infection or vaccination. Discrepancies between self-reported positive test and vaccination status and antibody results suggested a high prevalence of asymptomatic infection and waning antibody titers. County-level infection prevalences, estimated from the county test reporting system, were 16.6% in February, 19.1% in April, and 23.8% in October, substantially lower than the prevalence of individuals with antibodies from infection in the surveys, also supporting a high prevalence of asymptomatic or unconfirmed infections. The overall small sample size precluded an analysis of characteristics associated with active or past infection. In conclusion, surveillance surveys can provide timely data on infection status and immunity to support public health responses.
Understanding True COVID-19 Spread: Key Findings from a 2022 New York County Study
What was the main purpose of this COVID-19 study?
The study had two primary objectives. The first was to provide local health department officials in a New York county with more accurate and timely estimates of SARS-CoV-2 infection and immunity. By using a representative sample of the population rather than relying on reported test data, the researchers aimed to give officials a clearer picture to inform their pandemic response decisions.
The second objective was to identify characteristics and behaviors associated with COVID-19 infection among high-risk groups. However, the study's final sample size was too small to perform this analysis with statistical confidence.
Why are official reported COVID-19 case counts often inaccurate?
Disease prevalence estimates that are based on test positivity—the proportion of all tests that come back positive—are known to be biased. The study highlights several key reasons for this inaccuracy:
- Unequal Test Access: Not everyone has the same availability or ability to access testing facilities.
- Asymptomatic Infections: Many people who are infected with SARS-CoV-2 never develop symptoms and, as a result, may not seek out testing.
- At-Home Testing: The results from at-home rapid antigen tests are often not reported to public health authorities, leaving a significant gap in the data.
- Testing Population Bias: People who get tested at a clinic are often already feeling sick. This can artificially inflate test positivity rates, making them unrepresentative of the infection rate in the general population, which includes many asymptomatic individuals.
How did the researchers get a more accurate picture of infection rates?
Instead of relying on reported test results, the researchers conducted representative surveillance surveys in three different municipalities within the county during February, April, and October of 2022. They used a 2-stage cluster sampling design, a method modeled after a World Health Organization survey design, to randomly select households for participation.
Field teams collected anterior nares (nasal) swabs from participants. These single swabs were then used for two different types of laboratory analysis:
- RT-PCR testing to detect active SARS-CoV-2 viral RNA.
- Serological testing to detect antibodies (like Nucleocapsid Protein and Receptor Binding Domain antibodies) that provide evidence of past infection or vaccination.
How can scientists distinguish between immunity from a vaccine and immunity from an actual infection?
Serologic (antibody) testing is the key to making this distinction. The study measured two specific types of antibodies, which act as markers for different kinds of immune responses:
- Receptor Binding Domain (RBD) Antibodies: The presence of these antibodies in the absence of NP antibodies is interpreted as evidence of immunity from vaccination without a prior infection.
- Nucleocapsid Protein (NP) Antibodies: The presence of NP antibodies is the key marker of a past natural infection. People who have been infected with the SARS-CoV-2 virus itself develop NP antibodies, regardless of whether they have also been vaccinated.
What was the study's most significant finding about past COVID-19 infections?
The study's most significant finding was the sheer volume of undiagnosed or asymptomatic infections it uncovered. The research revealed that the prevalence of individuals with antibodies from a past infection (seroprevalence) was two to three times higher than the cumulative prevalence estimated from the county's officially reported positive test results.
This conclusion was reinforced by another key data point: fewer than half of the participants who had antibodies indicating a past infection actually self-reported ever having a positive COVID-19 test. This major discrepancy strongly suggests a high prevalence of infections that were either asymptomatic or were never confirmed with a formal test.
How many people in the study had some form of immunity to COVID-19 by 2022?
The study found very high levels of overall immunity (from either vaccination or past infection) across the population. The results from the three survey periods were as follows:
- February 2022 (City of Ithaca): 92% of individuals had detectable antibodies.
- April 2022 (Town of Ithaca): 90% of individuals had detectable antibodies.
- October 2022 (Town of Dryden): 81% of individuals had detectable antibodies.
These results align with other U.S. studies from the same period, which estimated that over 90% of the population had antibodies by early 2022.
Did the study find many people who were actively sick with COVID-19?
No, the prevalence of active infection, as measured by positive RT-PCR tests from nasal swabs, was very low across all three surveys. The household prevalence of PCR-positive tests was less than 5% in both the February and April surveys, and it was 0% in the October survey.
What does "waning immunity" mean, and did the study find evidence of it?
"Waning immunity" refers to the gradual decrease of antibody levels over time after an infection or vaccination. The study did find evidence suggesting this was occurring. Researchers noted two key observations:
- The October 2022 survey had the lowest overall rate of seropositivity (81%). The researchers suggest this could be due to the waning of antibodies derived from vaccination, especially since this group also had the lowest rate of booster vaccination.
- The proportion of participants who reported a past positive test but no longer had detectable NP antibodies (the marker for infection) increased across the three surveys. This suggests that the antibodies from their past infections were waning over time.
What were the main limitations of this research?
Acknowledging limitations is a critical part of the scientific process. The researchers identified several for this study:
- Small Sample Size: The study enrolled fewer households than originally targeted. This limited the statistical power needed to analyze the risk factors associated with infection.
- Potential for Bias: The researchers noted two potential sources of bias. Selection bias may have occurred because high-density housing like dormitories and some apartment complexes were excluded. This could lead to an underestimation of prevalence, as SARS-CoV-2 would likely spread more rapidly in such high-density communities. Non-response bias might have occurred if an individual's level of concern about COVID-19 influenced their willingness to participate in the study.
- Recall Bias: The accuracy of the study relies in part on participants' ability to remember their vaccination dates and past test results, which may not always be perfect.
What is the key takeaway for public health from this study?
The key takeaway is that official case counts represent only the tip of the iceberg. This study proves that relying solely on reported test data leads to a massive underestimation of community infection. For public health officials, this means the true burden of disease—and the corresponding level of population immunity—can only be understood through proactive, representative surveillance.