Table 6.
Predictors of health service utilization.
| Independent Variable | Category | AOR (95% CI) | p-valuea |
|---|---|---|---|
| Gender | Female (ref: Male) | 1.723 (0.800-3.709) | 0.164 |
| Family Size | 4–6 persons (ref: <4 persons) | 0.721 (0.242–2.150) | 0.558 |
| More than 6 persons (ref: <4 persons) | 0.710 (0.204–2.475) | 0.591 | |
| Age | > 20 years (ref: <20 years) | 3.607 (1.123–11.590) | 0.031* |
| Study Level | 200 Level (ref: 100 Level) | 5.055 (1.895–13.484) | 0.001* |
| 300 Level (ref: 100 Level) | 4.203 (1.549–11.403) | 0.005* | |
| 400 Level (ref: 100 Level) | 13.811 (3.537–53.933) | < 0.001* | |
| 500 Level (ref: 100 Level) | 6.928 (1.289–37.225) | 0.024* | |
| #Attitude of Staff is Poor | Disagree (ref: Agree) | 2.507 (1.084–5.796) | 0.032* |
| #Distance from Facility too far | Disagree (ref: Agree) | 0.797 (0.275–2.306) | 0.675 |
| #Waiting Time too long | Disagree (ref: Agree) | 0.591 (0.239–1.458) | 0.253 |
| #Perceived incompetence of Healthcare Staff | Disagreed (ref: Agreed) | 1.299 (0.535–3.155) | 0.563 |
| #Cost of Services too expensive | Disagree (ref: Agree) | 1.057 (0.438–2.551) | 0.902 |
| #Affordability of Food affecting utilization | Disagree (ref: Agree) | 1.208 (0.556–2.624) | 0.634 |
| #High Cost of Drugs affects utilization | Disagree (ref: Agree) | 1.770 (0.772–4.057) | 0.178 |
| Family Income (Naira) | < 200,000 (ref: >1,000,000) | 39.189 (5.091-301.653) | < 0.001* |
| 200,000-500,000 (ref: >1,000,000) | 11.486 (4.788–27.559) | < 0.001* | |
| 600,000–1,000,000 (ref: >1,000,000) | 4.865 (2.440–9.700) | < 0.001* | |
| Monthly Allowance (Naira) | 60,000-100,000 (ref: 20,000–50,000) | 1.051 (0.485–2.278) | 0.899 |
| > 100,000 (ref: 20,000–50,000) | 0.806 (0.281–2.308) | 0.687 |
Key: Dependent variable = Utilization of Healthcare services; OR = Odds Ratio; CI = Confidence interval; a= Logistic regression; *= Statistically significant at p ≤ 0.05; Ref: Reference value; #= The Likert scale responses for economic barriers and perception variables were dichotomized to Agree and Disagree. Note: A cutoff p-value of < 0.2 was used to include variables into the binary logistic regression model after bivariate analysis.