Indication-dependent heterogeneity in the effect of double-suture cerclage: concerns over indication classification and a call for stratified analysis

To the Editor,

We read with great interest the retrospective cohort study by Kahraman et al., which compared perinatal outcomes following single- versus double-suture cervical cerclage using a modified Wurm technique with propensity score-based inverse probability of treatment weighting (IPTW) [1]. The standardization of suture material across both groups and the application of a doubly robust analytical framework represent commendable methodological strengths. However, we submit that the central conclusion—that suture number does not influence perinatal outcomes—is directly contradicted by the authors’ own sensitivity analysis and reflects the consequence of pooling fundamentally heterogeneous clinical phenotypes.

The three conventional cerclage indications—history-indicated (HI), ultrasound-indicated (UI), and emergency (physical examination-indicated) cerclage (ECC)—differ not merely in the timing of intervention but in the degree of cervical structural compromise. In HI cases, cerclage is placed prophylactically on an anatomically intact cervix; in UI cases, progressive shortening has been identified but the cervix remains closed; in ECC, the cervix is already dilated, effaced, and frequently associated with membrane prolapse. These are not gradations of the same condition—they represent distinct biomechanical environments in which the rationale for suture reinforcement differs fundamentally.

Critically, the indication-specific differential effect of double versus single cerclage is not a theoretical construct—it has already been suggested by published evidence. A meta-analysis by Pergialiotis et al. [2] and a subsequent systematic review by Park and Park [3] both reported that double cerclage confers comparable outcomes to single cerclage in HI cases, a modest benefit in UI cases, and a more consistent benefit in pregnancy prolongation specifically in ECC. This gradient of benefit—comparable in HI, modest in UI, most pronounced in ECC—directly implies that pooling all three indications will systematically underestimate the true effect of double cerclage in the subgroup where it matters most. The null result reported by Kahraman et al. is therefore entirely consistent with this indication-dependent pattern and cannot be interpreted as evidence against double cerclage in emergency settings [1].

Most importantly, the authors’ own Supplementary Table S3 provides direct evidence of this heterogeneity. Within the ECC subgroup, term delivery was achieved in 73.7% of single-suture versus 48.1% of double-suture cases (RR 0.65; 95% CI 0.44–0.96)—a statistically significant difference. In the prophylactic cerclage subgroup, by contrast, no significant difference was observed (RR 0.89; 95% CI 0.49–1.62). This pattern almost certainly reflects unresolved confounding by indication severity: double cerclage was preferentially applied to more advanced ECC cases, a bias that IPTW cannot fully correct when the propensity model does not account for temporal trends in practice change. Consistent with this, Supplementary Table S2 shows that the SMD for cervical dilation at diagnosis increased after IPTW weighting (0.014–0.079), suggesting residual imbalance in precisely the variable most relevant to ECC severity. The conclusion that suture number is inconsequential therefore pertains, at most, to prophylactic indications—and cannot be generalized to ECC.

A further concern arises from the classification of cerclage indication in the supplementary dataset. Among 73 cases classified as emergency cerclage (ECC), only 16 (21.9%) had documented prolapsed membranes and only 14 (19.2%) had cervical dilation at diagnosis—the two cardinal criteria that define physical examination-indicated cerclage by international consensus. Strikingly, 57 of 73 ECC cases (78.1%) lacked both cervical dilation and membrane prolapse, and 25 cases (34.2%) showed none of the three examined cervical findings—dilation, membrane prolapse, or funneling. Paradoxically, the prophylactic cerclage group demonstrated cervical dilation in 34.6% of cases—a rate nearly twice that observed in the ECC group—and prolapsed membranes were recorded in 7.7% of prophylactic cases, a finding that is clinically implausible by definition. These observations raise serious concerns about the consistency and validity of cerclage indication classification throughout the dataset, which in turn undermines the reliability of all indication-based subgroup analyses reported in this study [1].

This concern is further reinforced by the Polish COSA randomized controlled trial, which exclusively enrolled ECC cases and whose results are awaited, reflecting scientific consensus that the biomechanical rationale for double-level cerclage is most compelling in the emergency setting [4]. We have previously argued that current international guidelines uniformly lack recommendations on stitch number and configuration specifically for ECC, and that this absence should be recognized as an evidence gap rather than evidence of ineffectiveness [5]. The present study’s pooled null result risks reinforcing this gap inappropriately.

We respectfully invite the authors to present indication-stratified analyses (HI, UI, ECC separately) with appropriate confounder adjustment within each subgroup. Given that their own data already demonstrate a statistically significant difference in ECC (Supplementary Table S3), such an analysis is not merely exploratory—it tests a pre-specified biological hypothesis with direct clinical implications for the highest-risk patients.

Author contributions

Y.J.P. contributed to the conceptualization and wrote the original draft. M.I.P. contributed to the conceptualization, reviewed and edited the manuscript, and provided supervision. All authors read and approved the final manuscript.

Funding

No funding was received for the preparation of this letter.

Data availability

No datasets were generated or analyzed during the current study.

Declarations

Conflict of interest

The authors declare no competing interests.