Table 2.
Multivariate Logistic Regression of Risk Factors for Hepatitis C Virus Infection
| Risk Factora | Results of Third-Generation RIBA |
Multivariate Logistic Regression Analysisb |
||
|---|---|---|---|---|
| Positive, no. (%) (n = 469) | Negative, no. (%) (n = 217) | Odds Ratio (95% CI) | P | |
| IDU | 195 (42) | 2 (1) | 35.0 (10.4–218.0) | <.0001 |
| Blood transfusion | 126 (27) | 17 (8) | 9.9 (5.6–18.3) | <.0001 |
| Intranasal cocaine use | 292 (62) | 23 (11) | 6.4 (3.8–11.2) | <.0001 |
| Intranasal cocaine use without IDU or blood transfusion | 79 (49)c | 20 (10)c | 8.5 (4.9–15.1) | <.0001 |
| Occupational exposure | 27 (6) | 5 (2) | 3.8 (1.3–12.4) | .0176 |
| Sexual promiscuity | 243 (52) | 48 (22) | 2.3 (1.4–3.7) | .0006 |
| Malesd | 148 (58) | 36 (30) | 3.3 (2.1–5.3) | <.0001 |
| Femalese | 95 (44) | 12 (13) | 5.5 (3.0–11.2) | <.0001 |
| Ear piercing | 273 (58) | 86 (40) | 1.8 (1.2–2.8) | .0088 |
| Malesd | 70 (28) | 1 (<1) | 45.9 (9.9–815) | .0002 |
| Femalese | 203 (95) | 85 (89) | … | NS |
| Male sex | 254 (54) | 121 (56) | … | NS |
| Tattooing | 102 (22) | 9 (4) | … | NS |
| Acupuncture | 29 (6) | 5 (2) | … | NS |
Data are ordered by odds ratio.
Abbreviations: CI, confidence interval; IDU, injection drug use; NS, nonsignificant; RIBA, recombinant immunoblot assay.
aOrder of entry into the model (forward selection method): (1) intranasal cocaine use, (2) blood transfusion, (3) IDU, (4) sexual promiscuity, (5) ear piercing, and (6) occupational exposure.
b Ellipses indicate that the risk factor did not meet criteria for entry into the model.
c Results from analysis of a subset of 361 subjects who denied IDU and blood transfusion (163 RIBA positive, 198 RIBA negative).
d Results from analysis of a subset of 375 males (254 RIBA positive, 121 RIBA negative).
e Results from analysis of a subset of 311 females (215 RIBA positive, 96 RIBA negative).