Observational cohort study of long COVID among 2 million elderly adults

In a recent study published in the journal PLoS Medicine, researchers investigated whether COVID-19 patients could fit the diagnostic criteria for long coronavirus disease (COVID-19) or post-COVID condition (PCC) after a bout of influenza (a condition referred to as long Flu).

Study: Prevalence and characteristics of long COVID in elderly patients: An observational cohort study of over 2 million adults in the US. Image Credit: p.ill.i / Shutterstock

Background

There have been wide variations in the reported incidence of PCC due to differences in the definition and measurement methods. PCC could be underreported among elderly individuals probably due to their lower likelihood of responding to surveys and their symptoms often being masked by other chronic medical conditions. Comparing long flu and PCC could provide valuable insights into the pathogenesis and treatment of PCC, a condition that is not well-characterized to date.

Multiple similarities exist between influenza and COVID-19. The two conditions have similar etiological agents, i.e., single-stranded ribonucleic acid (RNA) viruses with high affinity for respiratory organs, although generalized manifestations are observed. Both diseases are highly prevalent globally and have caused substantial socioeconomic and medical burdens. In light of the similarities between the two conditions, there could be symptomatic overlap between their post-infection syndromes.

About the study

The present observational cohort study estimated long Flu and PCC incidence in Medicare-insured COVID-19 patients based on World Health Organization (WHO) criteria. They also compared the symptomatology and use of healthcare resources between long Flu and CC patients.

The study involved Medicare (medical insurance program of the United States) beneficiaries aged >65 years, who were infected with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) between 1 April 2020 and 30 June 2021. De-identified records of the Medicare beneficiaries for the period between 2016 and 2021 were accessed through the Centres for Medicare and Medicaid Services (CMS) Virtual Research Data Centre (VRDC). The team used the international classification of diseases, tenth revision, and clinical modification (ICD-10)-CM diagnostic codes for evaluating influenza, COVID-19, and persistent symptoms.

PCC was defined based on the ICD-10 code for the condition, or the presence of any of the 11 WHO-defined PCC symptoms, after one to three months of acute infection. The team identified Long Flu among influenza patients diagnosed in 2018-2019 using the PCC symptomatic definition (influenza comparator group). Long Flu and PCC were comparatively assessed concerning the following outcomes: (i) any-cause hospitalization; (ii) PCC symptom-associated hospitalization; (iii) emergency department (ED) visits for PCC symptoms; and (iv) PCC symptom-associated outpatient visits.

Data were adjusted for gender, age, region, race, Medicare-Medicaid eligibility, chronic comorbid conditions, and previous year hospitalizations. Multiple logistic-type and log-linear-type regression modeling analyses were performed to determine the odds ratios (OR). The team excluded individuals with <12 months of Medicare insurance coverage and those without any health encounters in the year before influenza or COVID-19 diagnosis. In addition, individuals who were continually enrolled in Medicare Advantage plans, most of which were of the private health maintenance organization (HMO) type, in the year prior or 12.0 weeks post-influenza or COVID-19 diagnosis, have been excluded.

Results

Out of 2,071,532 participants who were followed up for two months, symptom-based PCC was identified among 29% (n=61,631) and 17% (n=246,154) of inpatients and outpatients, respectively. However, the estimated rates using ICD-10 codes were considerably lower, i.e., 2.6% (n=5,521) for inpatients and 0.5% (n=7,213) for outpatients. Of 933,877 individuals with influenza, 25% (n=18,824) and 17% (n=138,951) of inpatients and outpatients, respectively, satisfied the definition of long Flu.

Compared to individuals with long Flu, PCC patients had a greater incidence of fatigue, dyspnea, neurocognitive symptomatology, smell/taste loss, and palpitations. PCC outpatients had a greater likelihood of any-cause hospital admission (32%, n=74,854) vs. 27% (n=33,140), OR 1.1), and a higher number of outpatient encounters compared to long Flu patients visiting outpatient departments (2.90 vs. 2.50 visits, IRR 1.1). There were fewer ED visits among PCC patients, likely due to lower ED use during COVID-19. Despite comparable overall incidence rates, PCC patients experienced notably different symptoms in comparison to long Flu patients and had a greater likelihood of accessing outpatient and inpatient healthcare services.

Conclusions

Overall, the study findings showed that depending on particular PCC codes, underreporting could be considerable. PCC occurred in 17% and 29% of COVID-19 outpatients and inpatients, respectively, aged >65 years. The corresponding estimates for long Flu incidence, identified by similar symptoms during the 2018 and 2019 pre-pandemic influenza seasons, were 17% and 25%, respectively.

However, there were noteworthy differences in the symptoms of the two conditions. Further, PCC was related to greater utilization of healthcare resources compared to long Flu, indicating a larger impact on individual well-being and healthcare expenditures. The primary study limitation was that the PCC diagnosis was not verified independently. However, the researchers presented a technique to operationalize the clinical definition of PCC by excluding individuals based on history alone (symptoms not occurring in the previous year) or by history and comorbid conditions.

The exclusion by history alone approach yielded PCC estimates of 29% and 17% among inpatients and outpatients, respectively, compared to those published in the literature. The additional exclusion criteria of comorbidities significantly reduced PCC estimates to 8.80% and 5.70% for inpatients and outpatients, respectively. The comorbidity exclusion estimates seemed too stringent, given the prevalence of fibromyalgia with chronic pain and fatigue, cardiac failure, and chronic obstructive pulmonary disease among 54%, 43%, and 41% of the sample population, respectively.

Interesting study comparing seniors (65 and older) with #LongCovid to seniors who had long-lasting symptoms after influenza in ‘18 or ‘19. They define “long” as 1-3 months. I’d be interested to see them do a lengthier retrospective study looking at 6-18 months after Covid + flu. https://t.co/8TBKEH2cPQ

— Adina Gerver (@adinagerver) April 18, 2023