Factors Associated With Use of the Preventive Health Inventory in US Veterans

Key Points Question Which key patient, practitioner, and clinical characteristics are associated with Preventive Health Inventory (PHI) use? Findings In this cohort study of more than 4.3 million veterans, of whom 9.0% received the PHI, patients with Care Assessment Need scores and more outpatient use in the prior year were more likely to receive the PHI. Meaning These findings suggest that targeted outreach to veterans who use fewer primary care services may be needed to ensure that they receive necessary chronic disease management and preventive care.


Introduction
In response to the COVID-19 pandemic, the Veterans Health Administration's (VHA's) Office of Primary Care developed the Preventive Health Inventory (PHI) program in February 2021 1 to catch up on delayed or disrupted care This multicomponent care management intervention included development of a national dashboard of quality measures, telehealth appointments with a nurse, and completion of a templated electronic health record note that comprises a checklist of care needs.
The nurse visit and use of the templated note included screening and management for mental health, cancer prevention, and chronic disease management (eTable in Supplement 1). 1 Prior evaluations have not examined veteran receipt of PHI, a critical gap because the program targets specific chronic and preventive care needs of veterans.Understanding individual characteristics that influence PHI receipt is crucial to identify any differences in care based on specific patient, practitioner, and clinic factors.Our goal was to examine differences in individual characteristics and factors associated with PHI receipt to help the VHA address any barriers to receiving the PHI care.

Methods
We conducted a retrospective cohort study of veterans receiving primary care at the VHA between February 1, 2021, and February 1, 2022.We obtained all data from the VHA's Corporate Data Warehouse, a national repository of clinical and administrative data.To assess PHI use, we extracted specific administrative data created to track PHI template use.This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.This study was approved by the Veterans Affairs National Institutional Review Board.The institutional review board waived the need for consent because the study is considered minimal risk.
The following patient, practitioner, and clinic factors were examined in relation to their association with PHI receipt: age (years), sex (male or female), race and ethnicity (Alaska Native or Native American; Asian, Pacific Islander, or Native Hawaiian; Hispanic; non-Hispanic Black; or non-Hispanic White), 2 marital status, priority status, 3 neighborhood socioeconomic status (decile based on census data) as a surrogate marker for income, 4 rurality of residence (urban, rural, or highly rural or insular islands), drive distance to primary care (miles), Gagne comorbidity score (scores range from <0 to >9, with increased scores corresponding to increased risk of 1-year mortality), 5 Care Assessment Need (CAN) score (defined as the probability of hospital admission or mortality within 1 year, converted to a percentile), 6 count of outpatient visits in the prior year, and inpatient use in the prior year.Race and ethnicity were included as a factor in this study because there is evidence that telehealth interventions may be used less among those who identify as being in a racial or ethnic minority group.All data were extracted from administrative databases derived from the electronic health record and patient experience surveys.Practitioner variables (veteran assigned) included panel fullness (number of primary care patients cared for, adjusted for full-time equivalents), practitioner type (doctor of medicine, doctor of osteopathic medicine, physician assistant, or nurse practitioner), role of person completing the PHI reminder (physician or nurse), full-time equivalents, age (years), sex (male or female), and years of VHA tenure.Clinic variables included total clinic size (number of enrolled patients at site), staffing ratio (number of support staff for each practitioner), facility type (VHA medical center or outpatient clinic), geographic regional system of care (Veterans Integrated Service Network), and rurality (urban, rural, or highly rural or insular islands).

Statistical Analysis
We used bivariate analyses to calculate standardized mean differences (SMDs) comparing patient, practitioner, and clinic characteristics between veterans who did and did not receive the PHI.Standardized mean differences less than 0.1 were considered meaningful. 7We used binomial generalized linear models with fixed effects for clinics to estimate the association of the variables of interest and receipt of PHI.Receipt of PHI was defined as evidence of a completed templated clinic

JAMA Network Open | Public Health Factors
Associated With Use of the Preventive Health Inventory in US Veterans

Table 1 .
Unadjusted Associations Between Potential Factors Associated With PHI Use note in the veteran's electronic health record.To facilitate variable selection, we used least absolute shrinkage and selection operator procedures.For inference, we calculated marginal effects for each explanatory variable, using the Δ method for SEs.Estimated marginal effects represent differences on the probability scale.All analyses were performed using R, version 3.4.1 (R Foundation for Statistical Computing).

Table 1 .
Unadjusted Associations Between Potential Factors Associated With PHI Use (continued) Abbreviations: CAN, Care Assessment Needs; PHI, Preventive Health Inventory; SES, socioeconomic status; SMD, standardized mean difference; VA, Veterans Affairs.

Table 2 .
Average Marginal Effects for Factors Associated With Preventive Health Inventory Use