Author Reply

We thank Mohazzab and Masoumi for their thoughtful letter [1] and for the opportunity to clarify several methodological aspects of our study [2]. We appreciate their interest in our work and agree that transparency regarding analytical choices is important. Although our primary clinical focus was delivery within 24 h following PPROM, our analysis was based on modelling the continuous time from membrane rupture to delivery. A time-to-event framework was therefore chosen to incorporate the full latency information and to evaluate relative hazards of delivery over time rather than relying solely on a dichotomous endpoint. Importantly, women who did not deliver within 24 h were not censored at that time point but continued to contribute follow-up until delivery. In addition, the association between elevated NLR and earlier delivery persisted across extended latency horizons up to 36, 48, 60, and 72 h, suggesting that the observed effect was not driven by the choice of a single time threshold. As suggested, we now provide the results of the prespecified multivariable logistic regression analysis assessing delivery within 24 h. This analysis yielded findings consistent with the Cox model, with NLR > 10 remaining independently associated with early delivery (adjusted OR 2.57, 95% CI 1.40–4.70). These results are presented in the accompanying table (Table 1) and further support the robustness of the association across different analytical approaches. We formally evaluated the proportional hazards assumption by introducing a time-dependent interaction term between NLR > 10 and the logarithm of time to delivery within a multivariable Cox model including the same covariates as the primary analysis. This assessment suggested that the effect of NLR was strongest early after PPROM and attenuated over time, indicating a departure from strict proportionality. This time-dependent pattern is clinically plausible and aligns with the study's focus on imminent delivery rather than prolonged latency. Importantly, this finding did not materially affect the interpretation of our main results, which were consistent across complementary analyses. We agree that alternative modelling strategies, such as accelerated failure-time models, may also be appropriate in this context. However, the concordant findings across time-to-event and logistic frameworks reinforce our conclusion that elevated maternal NLR is associated with an increased likelihood of early delivery following PPROM. We thank the authors again for their constructive comments, which we believe have helped strengthen the methodological clarity of our work. All authors contributed to the conception and interpretation of the response. Daniel Gabbai drafted the initial version of the Author’s Reply. All authors critically revised the manuscript for important intellectual content and approved the final version. The authors have nothing to report. The authors declare no conflicts of interest. Clarifying Time-To-Event Methods for Delivery Timing After PPROM. https://doi.org/10.1111/1471-0528.70134. No new data were generated for this Author’s Reply. Additional analyses were conducted using the data from the original study.