Table 3.
Accuracy of Initial Self-Screening Compared to Provider Screening for Contraindications to Use of Oral Contraceptives
|
Initial self-screen1 N, % and (95% C.I.) |
Provider | |||
|---|---|---|---|---|
| Contraindicated | Eligible for pill use |
Total | ||
| Respondent | Contraindicated |
279 22.1% (19.8-24.4%) |
325 25.7% (23.3-28.1%) |
604 47.8% (45.1-50.6%) |
| Eligible for pill use |
219 17.2% (15.1-19.3%) |
441 34.9% (32.3-37.6%) |
660 52.2% (49.4-54.9%) |
|
| Total |
498 39.4% (36.7-42.2%) |
766 60.6% (57.9-63.3%) |
N=1,264 100.0% |
|
| Sensitivity = 56.0%; 95% CI: (51.5-60.4%) Specificity = 57.6%; 95% CI: (54.0-61.1%) Positive predictive value = 46.2%, 95% CI: (42.2-50.3%) Negative predictive value = 66.8%, 95% CI: (63.1-70.4%) McNemar chi-square test=20.8, P=0.0 | ||||
1
Initial self-screen refers to women’s answer to the question whether the pill would be bad for their health. Hormonal contraceptive users were, by default, classified as deciding that they were eligible for pill use.