Table 5.
Logistic regression models for prediction of visual acuity (VA) at follow up
| Variable | Comprehensive model | Parsimonious model* | ||
| OR (95% CI) for VA<20/60 | OR (95% CI) for VA<20/400 | OR (95% CI) for VA<20/60 | OR (95% CI) for VA<20/400 | |
| Demographics | ||||
| Age (years) | 1.04 (1.02 to 1.07) | 1.03 (1.00 to 1.06) | 1.05 (1.02 to 1.07) | 1.03 (1.00 to 1.06) |
| Sex | 1.05 (0.50 to 2.19) | 0.54 (0.18 to 1.64) | 1.10 (0.54 to 2.22) | 0.59 (0.19 to 1.65) |
| Literacy | 0.20 (0.08 to 0.49) | 0.05 (0.01 to 0.41) | 0.20 (0.08 to 0.48) | 0.05 (0.01 to 0.42) |
| Place of treatment first sought: (compared to Hariaun clinic) | ||||
| Self/neighbour/shaman | 0.90 (0.25 to 3.26) | 0.57 (0.06 to 6.60) | ||
| Paramedical | 2.47 (1.17 to 5.20) | 3.75 (1.19 to 11.8) | 2.39 (1.21 to 4.73) | 4.15 (1.42 to 12.10) |
| Time to treatment first sought: (compared to day 1) | ||||
| 1st week | 1.11 (0.47 to 2.61) | 1.28 (0.36 to 4.53) | ||
| More than 1 week | 1.34 (0.38 to 4.67) | 1.14 (0.17 to 7.72) | ||
| Clinical presentation (severe compared to mild) | ||||
| Severe injury† | 1.89 (0.81 to 4.43) | 8.26 (1.04 to 65.43) | 8.04 (1.02 to 63.42) | |
*Determined using forward, backward, and stepwise regression methods with inclusion criteria of p<0.05, fixing age and sex into the model.
†Severe injury (n = 355) was defined as deep (>1/3 corneal thickness) foreign body, corneal ulcer, corneal rupture, corneal blood stain, cataract, dislocated lens, iridodialysis, iris prolapse, “other” iris damage, scleral rupture, hyphaema, macula/retinal damage.
Values in bold if confidence intervals do not contain 1.0.
“Paramedical” includes care sought at local pharmacy, local medicine shops or clinics with medical or paramedical workers, health posts, and non-eye care hospitals.