Clinical Scorecard: Patterns of Long COVID Incidence Over Time, Demographics, and Geography in Relation to SARS-CoV-2 Variant Development: Analysis from the Texas All-Payer Claims Database (TX-APCD)

At a Glance

Key Highlights

• TX-APCD covers approximately 60% of insured Texans, representing 15.3 million individuals as of September 2023.

• Long COVID was identified using ICD-10-CM code U09.9, counting each person’s first documented diagnosis.

• Weekly COVID-19 emergency department (ED) activity preceded Long COVID claims, with the strongest association at a 2-week lag during Omicron BA.1 and a 3-week lag during later variant periods.

• The median recorded interval to Long COVID diagnosis was 22 days following an ED encounter versus 26 days following a non-ED encounter.

• Crude incidence was higher among older adults, females, and Medicare Fee-for-Service beneficiaries, with geographic clustering in High Plains and Northwest Texas.

Guideline-Based Recommendations

The source is a claims-based observational study, not a clinical guideline. The following points describe its methods and the authors’ proposals.

Diagnosis

• The study identified Long COVID through U09.9 claims from any diagnosis field or provider setting.

• Claims-based identification does not establish symptom duration, clinical severity, or functional impairment.

• Recorded intervals represent documentation timelines rather than biological time to symptom onset.

Management

• The authors propose dedicated care pathways for populations with greater documented burden, including primary care referral routes, multispecialty follow-up clinics, and rehabilitation services.

• The study does not evaluate treatment effectiveness.

Monitoring & Follow-up

• Multi-payer claims data can complement other surveillance sources.

• The authors identify ED encounters and the immediate postacute period as potential opportunities for proactive screening and structured follow-up.

Risks

• Diagnosis-code reliance may underestimate burden by missing affected individuals without a U09.9 code.

• Differences in healthcare access, utilization, and coding may influence recorded incidence.

Patient & Prescribing Data

TX-APCD includes commercial insurance, Medicaid Managed Care, Medicare Advantage, and Medicare Fee-for-Service claims. Among individuals with a documented Long COVID diagnosis, 44.71% had a prior COVID-related ED visit.

The study does not assess prescribing or treatment outcomes. Vaccination history and prior infections not captured in claims could not be evaluated.

Clinical Best Practices

• Interpret claims-based incidence as documented disease burden rather than total population prevalence.

• Consider age, sex, payer, and geographic variation when planning postacute care capacity.

• Recognize that uninsured individuals, veterans, and infections managed without healthcare contact are not represented.

• The authors propose prioritizing postacute care capacity in regions with greater documented burden, including telehealth-supported specialty consultation.

Related Resources & Content

• Long COVID: major findings, mechanisms and recommendations — Davis HE, McCorkell L, Vogel JM, Topol EJ — Nature Reviews Microbiology, 2023;21:133–146.

• Surveillance for post-COVID conditions is necessary: addressing the challenges with multiple approaches — Saydah SH, Brooks JT, Jackson BR — Journal of General Internal Medicine, 2022;37:1786–1788.

• Coding long COVID: characterizing a new disease through an ICD-10 lens — Pfaff ER, Madlock-Brown C, Baratta JM, et al. — BMC Medicine, 2023;21(1):58.

• Incidence of long COVID diagnoses in 3.6 million U.S. Medicare beneficiaries with COVID-19 — Abul Y, Harris DA, Chachlani P, et al. — The Journals of Gerontology: Series A, Biological Sciences and Medical Sciences, 2025;80(7):glaf108.

• Sex differences in long COVID — Shah DP, Thaweethai T, Karlson EW, et al. — JAMA Network Open, 2025;8:e2455430.