Evaluating the measurement of mental health service accessibility, acceptability, and availability in the Canadian Community Health Survey.

Given the underutilization of mental health services by those with mental health problems, growing attention has focused on barriers to utilizing services. Several researchers have used the Canadian Community Health Survey (CCHS) cycle 1.2 dataset, including measures of barriers because of accessibility, acceptability, and availability, to explore the gap between mental health service need and use. Because the psychometric properties of these barrier measures have not been evaluated, the reliability and validity of the 3 measures were examined in the present study. Confirmatory factor analyses were conducted using data from CCHS participants who had indicated unmet need regarding information on mental illness and its treatments; availability of services, medication, and psychotherapy or counseling (n = 353); as well as the full sample of participants reporting any unmet need in the past year (n = 1,784). The hypothesized 3-factor model (i.e., accessibility, acceptability, and availability) failed to converge with both samples. Exploratory factor analysis was conducted using data from the full sample (n = 1,784), and a possible 2-factor solution was obtained. Reliability analyses on this 2-factor model, as well as the 3-factor model included in the CCHS, demonstrated that internal consistency values failed to attain acceptable levels of reliability (i.e., α < .70). Based on these analyses, these barrier measures are neither reliable nor valid. These measures should not be used to examine barriers to service within the CCHS 1.2 dataset, and caution should be exercised in interpreting the findings of studies that used these measures.