Study population and setting
We currently have limited data on the spread of SARS-CoV-2 infection because there are many more cases than people who are tested. Surveys for SARS-CoV-2 antibodies can help provide this information. This study set out to estimate the SARS-CoV-2 seroprevalence using antibodies among people who died and had a forensic postmortem examination by the Maryland Chief Medical Examiner from May 24 to June 30, 2020. Reverse-transcription polymerase chain reaction for SARS-CoV-2 testing was performed by the medical examiner if there was suspicion of acute infection at the time of death.
Summary of Main Findings
Of the 720 eligible subjects, 500 had a viable blood specimen for testing. Among the 500 people examined, the seroprevalence of SARS-CoV-2 IgG or IgM antibodies was 10% (50 of 500; 95% CI, 9%–11%). Prevalence changed over time, from 5% (95% CI, 3%–7%) the week of June 14 to 21% (95% CI, 16%–26%) the week of June 21. The prevalence among those who died in motor vehicle crashes was similar to those who died of natural causes (adjusted prevalence ratio: 1.15; 95% CI, 0.44–3.04). A total of 12 of the 500 tested positive by RT-PCR. The distribution of race and ethnicity among decedents undergoing forensic investigation who had antibodies was similar to patterns observed in acute disease testing.
This study used a population that was easy to enumerate, and may therefore be a good source of surveillance information on SARS-CoV-2 if it can be shown that that the population was representative of the greater population to which we want to generalize. Forensic studies may also tell us about populations that are underrepresented in general population seroprevalence studies making them a potentially important source for information community spread.
The main limitation in this study is that the population that was sampled was not a representative sample of the population at risk for SARS-CoV-2 and therefore may not represent the population prevalence well (as those in car accidents may have a higher socioeconomic status and because travel and gathering restrictions were in place those in accidents might be more likely to be essential workers, etc.). In addition, this was a small study and specimens were not available for some of those eligible and had to be excluded, so some selection bias may have occurred.
It is critical to find populations where routine surveillance can be done for SARS-CoV-2. While decedents undergoing postmortems are not a perfectly representative sample of all of Maryland and differ from the full population with respect to some key factors, their seroprevalence results were very similar to those in the general population, making this a plausible candidate population for routine surveillance.
This review was posted on: 12 February 2021