Table 1.
Steps to identify care episodes for three study populations from the linked claims-EHR databases.
Fig 1.
Study flow diagram to identify care episodes for study populations from the linked claims-EHR databases.
CED IBM MarketScan Explorys Claims-EMR Data Set, EHR electronic health record, ICD-10-CM International Classification of Diseases, Tenth Revision, Clinical Modification, NAAT nucleic acid amplification test, P population, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2. a Episodes constructed in the claims portion of the linked claims-EHR data are of individuals aged <65 years at the index event with continuous coverage of medical benefits (allowing for a coverage gap up to 31 days) during the episodes and 6 months prior to the start of the episodes. b When multiple NAATs were present during the episode in the EHR, all must be negative. Data sources: IBM MarketScan Explorys Claims-EMR Data Set, April 1–December 31, 2020, and OneFlorida Data Trust linked Medicaid-EHR, April 1–November 30, 2020.
Fig 2.
Cohort flow diagram for the commercially insured population in CED.
CE criteria: Individuals must be continuously enrolled with medical benefits during the episode and during the 6-month baseline period prior to the episode. An enrollment gap of up to 31 days was allowed. CE continuous enrollment, CED IBM® MarketScan Explorys Claims-EMR Data Set, EHR electronic health record, NAAT nucleic acid amplification test, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2.
Fig 3.
Cohort flow diagram for the Florida Medicaid population in OneFlorida.
CE criteria: Individuals must be continuously enrolled with medical benefits during the episode and during the 6-month baseline period prior to the episode. An enrollment gap of up to 31 days was allowed. CE continuous enrollment, EHR electronic health record, NAAT nucleic acid amplification test, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2.
Table 2.
Characteristics of the study populations.
Fig 4.
Performance characteristics (PPV/NPV) of diagnosis code U07.1 in claims data among commercially insured individuals.
The PPV and NPV are not calculated for children <18 years old in study population 3 because the cell size was ≤1 in three cells of the 2X2 table for the calculation of PPV and NPV. Care setting and age were measured at the index event. Outpatient encounters include physician office, ambulatory care, outpatient hospital, and urgent care visits. Care setting was defined using a hierarchy if services at multiple settings were found on the index event date: inpatient, ED, outpatient, other. See S2 Table 2A in S2 File for the full results with exact CIs. The PPVs for study population 1 (inpatient setting at index event) and population 3 (all-cause hospitalization) are different because study population 1 episodes include diagnoses outside the hospitalization. CED IBM MarketScan Explorys Claims-EMR Data Set, CI confidence interval, ED emergency department, ICD-10-CM International Classification of Diseases, Tenth Revision, Clinical Modification, NAAT nucleic acid amplification test, NPV negative predictive value, PPV positive predictive value, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2. a Metric calculation used cell size <5. Data source: IBM MarketScan Explorys Claims-EMR Data Set, April 1–December 31, 2020.
Fig 5.
Performance characteristics (PPV/NPV) of diagnosis code U07.1 in claims data among individuals with Florida Medicaid.
Outpatient encounters include all outpatient services because specific places of service (e.g., office, hospital outpatient) cannot be identified in the data. Care setting was defined using a hierarchy if services at multiple settings were found on the index event date: inpatient, ED, outpatient, other. See S2 Table 2B in S2 File for the full results with exact CIs. The PPVs for study population 1 (inpatient setting at index event) and population 3 (all-cause hospitalization) are different because study population 1 episodes include diagnoses outside the hospitalization. CI confidence interval, ED emergency department, ICD-10-CM International Classification of Diseases, Tenth Revision, Clinical Modification, NAAT nucleic acid amplification test, PPV positive predictive value, SARS-CoV-2 severe acute respiratory syndrome coronavirus 2. a Metric calculation used cell size <5. Data source: OneFlorida Data Trust linked Medicaid-EHR, April 1–November 30, 2020.
Fig 6.
Monthly trends in performance characteristics (PPV/NPV) of diagnosis code U07.1 in claims data by population.
Month corresponds to the month of the index event. Population 1: Diagnosis of COVID-19, symptoms, or potential exposure. Population 2: Severe acute respiratory syndrome coronavirus 2 nucleic acid amplification test procedure in claims. Population 3: All-cause hospitalization. CED IBM MarketScan Explorys Claims-EMR Data Set, ICD-10-CM International Classification of Diseases, Tenth Revision, Clinical Modification, NPV negative predictive value, PPV positive predictive value. Data sources: IBM MarketScan Explorys Claims-EMR Data Set, April 1–December 31, 2020, and OneFlorida Data Trust linked Medicaid-EHR, April 1–November 30, 2020.