Recent Human Papillomavirus Vaccination is Associated with a Lower Risk of COVID-19: A US Database Cohort Study

2.1 Data Source

The data used in this study were extracted from the TriNetX network, a global health collaborative clinical research network that collects real-time electronic data from medical records [26]. This network provides access to electronic medical records (EMR; including diagnoses, procedures, medications, laboratory test results, and genomic information) of approximately 250 million individuals from 120 global health care organizations (HCOs) and is one of the largest global COVID-19 datasets. TriNetX has been used in numerous studies to assess the risks, patterns, clinical features, and consequences of COVID-19 [27,28,29,30]. In this study, we utilized the US Collaborative Network, a subset of the TriNetX database, which includes 49 HCOs in the USA. The most up-to-date and integral data that were accessible in TriNetX in June 2022, when we designed and conducted the study, contained the information as of March 2022. Therefore, data used in our primary analysis were collected and analyzed as of March 2022. The study period in our primary analysis was from January 1st 2020 through March 31st 2022.

2.2 Study Population

The study population included female patients aged 89 years and younger with at least two medical visits documented in their EMR in the TriNetX Network between January 1st 2020 and March 31st 2022. The first visit during the study period was referred to as the index date. We excluded patients who had received any dose of the COVID-19 vaccine, in order to mitigate any potential confounding effects related to COVID-19 vaccination. These patients were identified through their documented vaccination history in the EMR (Supplementary Table S1). We also excluded patients who had been diagnosed with COVID-19 or neoplasms prior to the index date. Participants were divided into two cohorts based on whether or not HPV vaccination was documented in their EMR in the TriNetX Network. The selection process was depicted in Supplementary Figure S1. The HPV-vaccinated cohort consisted of participants who had received a 2-valent, 4-valent, or 9-valent HPV vaccine prior to the index date. The HPV vaccination was identified based on the following current procedural terminology (CPT) codes: 2-valent (90650), 4-valent (90649), and 9-valent (90651). To further explore how the effect of vaccine may vary over time, we separately assessed the risk of outcomes in terms of four different lengths of look-back time of HPV vaccination, namely: (1) within 1 year before the index date, (2) within 1–2 years before the index date, (3) within 2–3 years before the index date, and (4) within 3–5 years before the index date (Supplementary Fig. S2). In comparison, the unvaccinated cohort had no HPV vaccination documented in their EMR in the TriNetX Network.

2.3 Outcomes

We identified the incidence of COVID-19 as the main outcome of our study by positive results in SARS-CoV-2-related viral RNA or IgM/IgG antibody tests, or the ICD-10 code U07.1 (COVID-19, virus identified) (details in Supplementary Table S2). We used hospitalization and all-cause mortality as our secondary outcomes. Hospitalization was identified by the International Classification of Diseases (ICD) procedure codes (1013661, 1013659, 1013668, or 1013729), the CPT codes (99221–99223, 99231–99233), or an inpatient encounter. The health care organizations can indicate whether a patient is known to be deceased at the organization. Mortality data were accurately recorded for patients who passed away during their stay in the hospital. Some health care organizations in the network have their data linked with the Datavant population mortality information, which are based on Social Security Administration data, obituary data, and some private claims data. We explored the risk of outcomes in both cohorts in terms of different follow-up periods.

2.4 Covariates

Demographic variables included age, race, problems related to housing and economic circumstances (Z59), problems related to education and literacy (Z55), problems related to employment and unemployment (Z56), and occupational exposure to risk factors (Z57). The comorbidities included hypertension (ICD-10 I10), ischemic heart diseases (ICD-10 code I20–I25), cerebrovascular diseases (ICD-10 code I60–I69), overweight and obesity (ICD-10 code E66), hyperlipidemia (ICD-10 code E78.5), type 2 diabetes mellitus (ICD-10 code E11), chronic lower respiratory diseases (ICD-10 code J40–J47), chronic kidney diseases (ICD-10 code N18), liver diseases (ICD-10 code K70–K77), nicotine dependence (ICD-10 code F17, used as a proxy variable for smoking status), sleep disorders (ICD-10 code G47), psychoactive substance use disorders (ICD-10 F10–F19), inflammatory polyarthropathies (ICD-10 code M05–M14), systemic connective tissue disorders (ICD-10 code M30–M36), spondylopathies (ICD-10 code M45–M49), non-infective enteritis and colitis (ICD-10 code K50–K52). For the presence of comorbidities, we looked back one year before the index date.

2.5 Statistical Analyses

We utilized the TriNetX platform and Advanced Analytics to conduct all data analysis. We used propensity score matching (PSM) to balance covariate patterns between the HPV-vaccinated and non-HPV-vaccinated cohorts in terms of age, race, socioeconomic status, and comorbidities, in order to reduce the impact of treatment selection bias [31]. Propensity score matching was performed based on greedy nearest neighbor matching with a caliper of 0.1 pooled standard deviations of the propensity scores in the aggregate. Covariate balance was assessed using standardized mean difference, and an absolute value of < 0.1 was considered a negligible difference in its distribution between the two cohorts. We used Kaplan–Meier analysis to estimate the probability of the outcomes among the two cohorts [32]. We used log-rank tests to indicate whether the survival curves were different between cohorts. We conducted Cox proportional hazard regression analysis to estimate hazard ratios (HRs) and 95% confidence intervals (CIs) [33]. Hazard ratios, 95% CIs, and the test for proportionality were calculated using R's Survival package v3.2–3 [34, 35]. The proportional hazard assumption was tested using the generalized Schoenfeld approach. All statistical tests were performed using the TriNetX platform with significance set at a 2-sided p < 0.05.

We further conducted subgroup analyses based on age (≤ 15 years, 16–20 years, and ≥ 21 years), race (White, Black/African American, Asian), comorbid asthma (with, without), and comorbid obesity (defined by body mass index [BMI] < 30 kg/m2, ≥ 30 kg/m2) to explore the differences between these subgroups in terms of the risk of incident COVID-19 over a 1-year follow-up period, comparing those who had received HPV vaccination within 1 year before the index date and those who had never received HPV vaccination. To verify the robustness and consistency of our findings, we further performed a sensitivity analysis using the same study design but with an extended study period between January 1st 2020 and December 31st 2022.

2.6 Ethical Statements

The TriNetX Analytics Network is compliant with the Health Insurance Portability and Accountability Act (HIPAA), the US federal law, which protects the privacy and security of health care data, and any additional data privacy regulations applicable to the contributing HCO. TriNetX is certified to the ISO 27001:2013 standard and maintains an Information Security Management System (ISMS) to ensure the protection of the health care data it has access to and to meet the requirements of the HIPAA Security Rule. The TriNetX Analytics Network was granted a waiver by the Western Institutional Review Board (WIRB) since it solely used aggregated counts and statistical summaries of de-identified data. The study protocol and survey instrument were approved by the Institutional Review Board of Chung Shan Medical University Hospital in Taiwan (CSMUH No: CS2-21176). The protocol for the research project conformed to the ethical norms and standards in the Declaration of Helsinki.

留言 (0)

沒有登入
gif