Periodontal treatment among mothers with mild to moderate periodontal disease and preterm birth: reanalysis of OPT trial data accounting for selective survival

Background: The Obstetrics and Periodontal Therapy (OPT) study, a randomized controlled trial, reported no effect of periodontal treatment on preterm birth. Even though there were more spontaneous abortions or stillbirths in the control group, sensitivity analyses using conventional approaches did not change the results. The development of newer epidemiological methods to assess bias caused by the truncated outcome, and the availability of OPT study data in the public domain, allowed us to reanalyse these data. Methods: We used the survivor average causal effect (SACE), also known as the principal strata effect, to correct potential bias resulting from unequal survival of fetuses in the treatment and control arms of the OPT study. Results: The risks of preterm and spontaneous abortions or stillbirths were respectively 49/413 (11.86%) and 5/413 (1.21%) in the periodontal treatment group, and 52/410 (12.68%) and 14/410 (3.33%) respectively in the control group. The risk differences (%) comparing periodontal treatment and control groups were -0.82%: 95% confidence interval (CI) -5.30% to 3.67% for preterm birth, and -2.12%: 95% CI -4.14% to -0.11% for spontaneous abortions or stillbirths before correction. Risk differences for preterm birth comparing periodontal treatment and control groups increased in magnitude, reached statistical significance and showed a beneficial effect of periodontal treatment after correction for bias using SACE. Conclusions: Periodontal treatment provided to mothers with mild to moderate periodontal disease before 21 weeks of gestation may prevent preterm births.