# Establishing Psychometric Properties of the Modified Barriers Experienced in Providing Healthcare Instrument

## TL;DR

This study evaluates and refines a healthcare provider barrier assessment tool to better understand challenges in rural healthcare delivery.

## Contribution

The study introduces a refined 12-item instrument (BPOC-12) with improved psychometric properties for assessing healthcare provider barriers.

## Key findings

- The original BTCPI model failed to meet model fit criteria, prompting refinement.
- The BPOC-12 model demonstrated acceptable fit and invariance across provider groups.
- Differences in perceived barriers were found between nurses and primary care providers, and between rural and urban practitioners.

## Abstract

Background: Rural healthcare providers encounter multifaceted barriers including geographic isolation, resource limitations, and provider shortages that impede optimal patient care delivery. The Barriers Experienced in Providing Healthcare Instrument (BTCPI) was designed to assess provider challenges; however, concerns regarding its psychometric properties necessitated comprehensive validation. The primary purpose of the study was to evaluate the structural validity of the instrument using confirmatory factor analysis with a sample of Idaho healthcare professionals. Because the model failed to meet criteria, the study identified a more parsimonious model that then underwent multi-group invariance testing. Methods: A survey consisting of a modified Barriers to Providing Optimal Healthcare instrument and a demographic questionnaire was distributed to Idaho healthcare providers across 22 clinical sites in the state. The structural validity of the modified 41-item, 9-factor instrument was assessed using confirmatory factor analysis (CFA), exploratory structural equation modeling (ESEM), and exploratory factor analysis (EFA). Multi-group invariance testing was also conducted to assess measurement equivalence across provider profession, practice setting (rural vs. urban), and years of experience. Results: A total of 373 healthcare providers completed the survey and were used for analysis. The proposed BTCPI model did not meet model fit criteria. An ESEM analysis was conducted and identified a 9-factor, 14-item model. However, due to fit concerns, an exploratory factor analysis was subsequently conducted and identified the 4-factor, 12-item (BPOC-12) that also met invariance criteria across groups. A group mean and variance differences were found between nurses and primary care providers as well as between rural and urban practitioners on several barrier factors. Conclusions: The BTCPI did not meet model fit criteria. Subsequent model refinement resulted in the BPOC-12, which had preliminary psychometric validity. Although the refined model offered a more condensed and preliminarily valid psychometric framework, future research should be done to assess this model. Future research should also collect responses from different healthcare professions to enhance its applicability.