2029. Barriers to Care: A New
Instrument
MS Bauer, Providence VAMC & Brown University, WO Williford,
Perry Point Cooperative Studies Coordinating Center, L McBride,
Providence VAMC, K McBride, University of Connecticut
Objectives: Access to
health care appears to be a major determinant of health outcome, and is of major
concern to the VHA. While
community-based studies have measured access—or its inverse, barriers—to
health care with a small number of questions in large surveys, we are aware of
no psychometrically valid and reliable instruments with which to measure the
construct of perceived barriers to healthcare.
Moreover, we are aware of no instruments designed to measure barriers
perceived by those currently treated in the healthcare setting who, although
they have been able to access care, may still experience significant barriers to
optimal care.
Methods: An
interview-based ordinal-scored 29-item instrument was developed based on
existing survey items and review of available clinical and theoretical
literature. It was administered to
312 veterans at the Providence VAMC presenting for a primary care or mental
health appointment. A subsample of 27 had the instrument re-administered 3-6
weeks after the initial interview to investigate test-retest reliability.
Test-retest reliability was assessed with paired t-test and Spearman’s
correlation coefficient. Principal components analysis with varimax rotation was
used to identify major factors. Validity was investigated by analyzing scores
across subgroups of subjects hypothesized a priori to differ in perceived
barriers.
Results: No
item showed significant drift by paired t-test, and 28 of 29 showed highly
significant correlations between administrations. Principal components analysis
revealed 4 main factors (with eigenvalues): Telephone (2.284), Distance (2.374),
Provider Communication (1.719), and Travel Difficulties (1.648).
Factors explained 67% of total item variance. Analysis of external validators confirmed some but not all of
the a priori hypotheses. Importantly, significant variation was seen among
individual primary care providers in Communication barriers perceived by their
patients, but not in other subscales.
Conclusions: Psychometric
analyses provide support for the reliability and validity of the Barriers to
Care Instrument in a clinical sample. Confirmation in multi-site clinical
samples, and in community-based samples, will provide further information on its
utility.
Impact: This instrument may be useful in assessing perceived barriers to care in clinical and epidemiologic investigations, and in administrative monitoring of services (e.g., to assess the impact of administrative and clinical reorganizations).