Development and evaluation of the SPACE-Postpartum multidomain symptom framework for predicting chronic pain after caesarean delivery: a prospective cohort study

BackgroundChronic postsurgical pain after caesarean delivery impairs postpartum recovery and maternal quality of life. Existing risk models focus on demographic or procedural factors, limiting opportunities for early intervention. This study developed and prospectively evaluated a biopsychosocial predictive model, SPACE-Postpartum (Sleep, Pain, Affect, Cognition, Energy), which assesses early postpartum symptoms across five domains. MethodsIn this prospective cohort study, adults undergoing caesarean delivery at a tertiary centre completed validated patient-reported outcome measures at baseline, 2 weeks, and 3 months postpartum. Chronic pain was defined as pain at any site persisting >3 months. A five-item model with one early postpartum indicator from each SPACE domain was derived using logistic regression with ridge regularisation, supported by latent class and causal mediation analyses. ResultsOf 143 participants, 100 (70%) completed 3-month follow-up; 40% reported chronic pain. The SPACE ridge model demonstrated good discrimination (AUC 0.76, 95% CI 0.67-0.85) with internal validation and acceptable calibration. Higher acute pain intensity, pain interference, and sleep disturbance, with a trend for reduced perceived control, predicted chronic pain. Latent class analysis identified an early high-burden SPACE profile (52%) associated with greater pain interference at 3 months. Mediation analysis indicated acute pain exerted a direct effect, while sleep disturbance acted as an independent prognostic marker. Exploratory sensitivity analyses suggested potential incremental value of quadratic models (AUC 0.83-0.87), although these risked overfitting in this small dataset. ConclusionsChronic pain after caesarean delivery is common and linked to potentially modifiable early symptoms, particularly sleep disturbance and pain-related interference. Across predictive, phenotypic, and causal analyses, pain and sleep symptoms consistently demonstrated prognostic value. This single-centre proof-of-concept study provides early internal validation of the SPACE-Postpartum model. Multicentre external validation is warranted to confirm generalisability and support development of symptom-informed decision tools.