Visits to PCP Among PWID at High Risk of HCV Infection
At enrolment, participants' mean age was 34 (SD = 9.6) and 80.8% were male. A minority reported having completed a college education (16.0%) and being chronically homeless (6.9%). One hundred and twelve (32.1%) reported visits to a PCP in the previous 6 months.
Table 1 presents descriptive and bivariate logistic regression analyses between PCP visits and each correlate. Among predisposing factors, male gender, chronic homelessness, cocaine injection and reporting greater unstable income were negatively associated with PCP visits. Perceived health and reports of being sick, markers of need, were not associated with the outcome. Among enabling factors, contact with street nurses and contact with food banks were positively associated with visits to PCP.
Results from the hierarchical multivariate logistic regression model are presented in Table 2. Compared to the intercept-only model, the introduction of the predisposing domain in Model A significantly improved model fit, as illustrated by the likelihood ratio test (P < 0.0001). Male gender, chronic homelessness, cocaine injection and reporting a greater proportion of the income coming through unstable sources remained independently associated with visiting a PCP. Need variables did not meet the criterion for retention into the multivariate model (i.e. P < 0.1) and thus, Model B yielded results identical to Model A. The addition of the enabling set of variables in Model C significantly improved model fit (P < 0.0001). Contact with street nurses and contact with food banks remained independently associated with visiting a PCP.
In the final multivariate model, variables that remained significantly associated with PCP visits included male gender, chronic homelessness, cocaine injection, reporting greater income through unstable sources, contact with street nurses and contact with food banks.
Results
At enrolment, participants' mean age was 34 (SD = 9.6) and 80.8% were male. A minority reported having completed a college education (16.0%) and being chronically homeless (6.9%). One hundred and twelve (32.1%) reported visits to a PCP in the previous 6 months.
Table 1 presents descriptive and bivariate logistic regression analyses between PCP visits and each correlate. Among predisposing factors, male gender, chronic homelessness, cocaine injection and reporting greater unstable income were negatively associated with PCP visits. Perceived health and reports of being sick, markers of need, were not associated with the outcome. Among enabling factors, contact with street nurses and contact with food banks were positively associated with visits to PCP.
Results from the hierarchical multivariate logistic regression model are presented in Table 2. Compared to the intercept-only model, the introduction of the predisposing domain in Model A significantly improved model fit, as illustrated by the likelihood ratio test (P < 0.0001). Male gender, chronic homelessness, cocaine injection and reporting a greater proportion of the income coming through unstable sources remained independently associated with visiting a PCP. Need variables did not meet the criterion for retention into the multivariate model (i.e. P < 0.1) and thus, Model B yielded results identical to Model A. The addition of the enabling set of variables in Model C significantly improved model fit (P < 0.0001). Contact with street nurses and contact with food banks remained independently associated with visiting a PCP.
In the final multivariate model, variables that remained significantly associated with PCP visits included male gender, chronic homelessness, cocaine injection, reporting greater income through unstable sources, contact with street nurses and contact with food banks.
Source...