By Landon Gray
To better diagnose long COVID, patients’ preexisting conditions and time of occurrence should be considered, according to a new study.
Researchers defined a set of conditions for long COVID or post-acute sequelae of COVID-19 (PASC), and described the timing of such conditions by using longitudinal data and matching a COVID-19 polymerase chain reaction (PCR)-negative population to a PCR-positive population 3-1 by age, sex, testing month and medical center (Nat Commun 2022;13[5822] https://doi.org/10.1038/s41467-022-33573-6). The researchers controlled for preexisting conditions up to four years prior to patients’ PCR tests.
“Besides looking forward to see if there were symptoms, we looked back to control for patients who had preexisting conditions, which was very novel for these PASC studies,” Michael Horberg, MD, MAS, FACP, FIDSA, the associate medical director of Research, Medical Education, Community Benefit and Medicaid Strategy for the Mid-Atlantic Permanente Medical Group, in Rockville, Md., told Infectious Disease Special Edition.
Conditions were specified as acute and persistent, which occurred initially 0 to 30 days post-test and persisted for 30 to 120 days post-test, or late (occurring initially 30-120 days post-test).
More than 100,000 patients were evaluated, and a scaled matching algorithm selected 28,118 patients who had a PCR-positive test and 70,293 who had received a negative PCR test in 2020, the first year of the COVID-19 pandemic. There was a good distribution of minorities and women in the groups: 57% were female, 40% to 53% were Black and 20% to 24% were Hispanic compared with 18% to 22% who were white. Patients younger than 65 made up approximately 87% of the study participants, and between 30% and 33% had a body mass index of at least 30 kg/m2.
The researchers found that 37.7% of the PCR-positive patients had at least one condition in the acute and persistent or late period, and 16.5% had at least one PASC-related condition in either period.
Approximately 20% of the PCR-positive patients had at least one condition and a PASC-related condition in the acute and persistent period. In the late period, the researchers found 26.1% of PCR-positive patients had at least one condition and 13.6% had a PASC-related condition.
In the PCR-negative group, 2.5% had a PASC-related condition during the acute and persistent period, while 22.1% had any conditions in this period—an increase from the PCR-positive group.
Approximately 12% of PCR-negative patients had a PASC-related condition during the late period, but 25.2% had any condition during that time frame (a decrease from the PCR group).
“We were able to demonstrate a clearly defined set of conditions for PASC and delineate them within an integrated system,” said Dr. Horberg, who is also the director of the Mid-Atlantic Permanente Research Institute, “but being able to compare that through different time periods, accounting for preexisting conditions, which again, people sometimes forget that they had these conditions or don’t want to think about them, then they manifest during an acute illness, but that still doesn’t mean that’s a PASC-related condition or a condition related to your COVID.”
The researchers concluded there was an increased risk for numerous conditions in patients who received a positive PCR test compared with PCR-negative patients. They documented a 4% higher risk for PCR-positive patients of having any condition in the late period versus PCR-negative patients (risk ratio [RR], 1.04; 95% CI, 1.01-1.07) and 8% lower in the acute and persistent period (RR, 0.92; 95% CI, 0.89-0.95). However, there was a 60% increased risk for having at least one PASC-related condition in the acute and persistent period (RR, 1.60; 95% CI, 1.48-1.72). The risk decreased to 12% in the late period (RR, 1.12; 95% CI, 1.08-1.16).
“We found a variety of conditions that were much more prevalent among the COVID-positive [patients] [compared with] the COVID-negative [patients] in both the acute and persistent period, those symptoms that occurred 30 days from the date of the test but then persisted,” Dr. Horberg explained. “And then, [conditions] that were late, those that occurred after the 30-day acute infection period, to those that were recurring up to 120 days.”
These were all diagnoses coded by physicians, Dr. Horberg added.
The PASC-related conditions included anosmia, cardiac dysrhythmias, diabetes, genitourinary conditions, malaise and fatigue, and nonspecific chest pain. Of note, the researchers added that those experiencing a PASC-related condition, versus the overall cohorts, were mostly female (62% vs. 57%), older in age (≥65 years of age; 15% vs. 12%-13%) and had a higher hospitalization rate in the 30 to 120 days after the lab test date (5.2%-6.8% vs. 1.8%-1.9%).
The research team is federally funded for additional studies to evaluate the impact that vaccination had on the incidence of COVID-19 and PASC. “I think vaccination treatment for COVID will affect later manifestations—the ‘long-hauler’ conditions. That was one reason why we kept this study to only 2020 was so that our initial studies wouldn’t have that competing issue,” Dr. Horberg said.
He also noted that the study was limited to only patients who received a positive or negative PCR test, adding that there may have been some patients who never got tested. “[Patients without a PCR test result] may have reduced some of the numbers in the positive or negative [groups], but certainly from an internal validity standpoint, we very much stand by our data.”
Dr. Horberg reported no relevant financial disclosures.