Table 2.
Logistic regression models results 1,2
| Variable | OR | SE | 95% CI | p |
| Model 1. Outcome: Overall preventable hospitalization | ||||
| Log-likelihood = -194.19, χ2 = 220.14, p < 0.001, Number of observations = 488 | ||||
| Number of PCP medical visits in the previous year | 0.1 | 0.04 | 0.05–0.23 | < 0.001 |
| Number of patients for each PCP | 2.25 | 0.38 | 1.62–3.13 | < 0.001 |
| Satisfaction with PCP health services | 0.34 | 0.09 | 0.2–0.58 | < 0.001 |
| Length of hospital stay | 0.95 | 0.02 | 0.91–0.99 | 0.011 |
| Sex | 0.52 | 0.14 | 0.31–0.87 | 0.013 |
| Self-reported health status | 0.53 | 0.14 | 0.31–0.89 | 0.017 |
| Number of PCP accesses in the previous year | 0.52 | 0.15 | 0.3–0.93 | 0.027 |
| Age | 1.03 | 0.01 | 1.01–1.05 | 0.027 |
| Number of hospital admissions in the previous year | 1.76 | 0.46 | 1.06–2.93 | 0.03 |
| Additional persons in the household | 0.75 | 0.13 | 0.53–1.06 | 0.105 |
| Type of admission | 0.57 | 0.24 | 0.25–1.3 | 0.18 |
| Distance between patient's home and hospital | 0.99 | 0.01 | 0.98–1.01 | 0.399 |
| Model 2. Outcome: Preventable hospitalization for cardiovascular diseases | ||||
| Log-likelihood = -150.13, χ2 = 153.38, p < 0.001, Number of observations = 429 | ||||
| Number of patients for each PCP | 2.2 | 0.43 | 1.5–3.22 | < 0.001 |
| Satisfaction with PCP health services | 0.31 | 0.09 | 0.17–0.57 | < 0.001 |
| Number of PCP medical visits in the previous year | 0.12 | 0.06 | 0.05–0.3 | < 0.001 |
| Length of hospital stay | 0.94 | 0.02 | 0.9–0.99 | 0.013 |
| Self-reported health status | 0.5 | 0.16 | 0.27–0.93 | 0.027 |
| Number of PCP accesses in the previous year | 0.52 | 0.17 | 0.27–0.98 | 0.044 |
| Working activity | 0.37 | 0.18 | 0.13–0.97 | 0.044 |
| Sex | 0.62 | 0.18 | 0.35–1.1 | 0.106 |
| Number of hospital admissions in the previous year | 1.53 | 0.46 | 0.85–2.77 | 0.156 |
| Additional persons in the household | 0.76 | 0.15 | 0.51–1.13 | 0.172 |
| Age-adjusted Charlson et al. comorbidity index | 1.08 | 0.07 | 0.96–1.22 | 0.19 |
| Model 3. Outcome: Preventable hospitalization for respiratory diseases | ||||
| Log-likelihood = -92.36, χ2 = 116.38, p < 0.001, Number of observations = 385 | ||||
| Number of patients for each PCP | 2.86 | 0.8 | 1.66–4.94 | < 0.001 |
| Satisfaction with PCP health services | 0.27 | 0.11 | 0.13–0.58 | 0.001 |
| Number of PCP medical visits in the previous year | 0.11 | 0.07 | 0.03–0.4 | 0.001 |
| Sex | 0.27 | 0.12 | 0.11–0.64 | 0.003 |
| Number of emergency accesses in the previous year | 2.71 | 1.12 | 1.2–6.11 | 0.016 |
| Number of PCP accesses in the previous year | 0.34 | 0.15 | 0.14–0.82 | 0.017 |
| Type of admission | 0.2 | 0.17 | 0.04–1.05 | 0.058 |
| Living condition | 2.35 | 1.08 | 0.95–5.81 | 0.063 |
| Self-reported health status | 0.49 | 0.21 | 0.21–1.12 | 0.092 |
| Age | 1.03 | 0.02 | 0.99–1.07 | 0.131 |
| Age-adjusted Charlson et al. comorbidity index | 0.88 | 0.08 | 0.73–1.05 | 0.159 |
| Length of hospital stay | 0.97 | 0.03 | 0.92–1.02 | 0.232 |
| Model 4. Outcome: Preventable hospitalization for diabetes | ||||
| Log-likelihood = -21.54, χ2 = 18.77, p = 0.0277, Number of observations = 144 | ||||
| Number of patients for each PCP | 4.75 | 3.29 | 1.22–18.48 | 0.025 |
| Number of hospital admissions in the previous year | 9.74 | 10.67 | 1.14–83.39 | 0.038 |
| Number of PCP accesses in the previous year | 0.15 | 0.17 | 0.02–1.34 | 0.090 |
| Education level | 0.27 | 0.23 | 0.05–1.46 | 0.129 |
| Length of hospital stay | 0.88 | 0.09 | 0.72–1.08 | 0.216 |
| Number of specialist visits in community health services | 0.36 | 0.37 | 0.05–2.79 | 0.326 |
| Self-reported health status | 2.61 | 2.57 | 0.38–18 | 0.332 |
| Sex | 2.47 | 2.33 | 0.39–15.65 | 0.338 |
| Age | 1.04 | 0.05 | 0.95–1.14 | 0.340 |
1Significance level is set at p ≤ 0.05
2The variables are presented in order of decreasing significance.
PCP, primary care physician.