Michail Sideris, Consultant Gynaecologist and Gynaecological Oncology Surgeon at The London Clinic, shared on LinkedIn:
“Read our new systematic review, ՛Predictive Models of Morbidity and Mortality in Gynaecological Cancer Surgery՛, now published in the Journal of Clinical Medicine.
Our review confirmed that we do not yet have a validated tool to predict postoperative morbidity and mortality in women undergoing gynaecological cancer surgery.
What we did:
- Searched MEDLINE and EMBASE with no date restriction
- Included 28 studies that developed or validated models for 30-day morbidity and mortality
- Assessed every study for risk of bias and applicability using PROBAST
What we found:
- 20 of the 28 studies were at high risk of bias, mostly in the analysis
- Only 9 studies reported calibration, and no frailty-based model reported calibration or validation
- Օne gynae-oncology model (for ovarian cytoreductive surgery) has been independently externally validated, and its calibration was reduced
- Generic surgical risk calculators performed poorly
- Discrimination was moderate at best
No model can currently be recommended for routine clinical practice. Predicted risks should support clinical judgement during consent and shared decision-making, not replace it.
Importantly, a prediction model should be developed for a specific setting.
Practice and facilities vary around the world, so a model cannot simply be transferred from one healthcare system to another without local validation.”
Title: Predictive Models of Morbidity and Mortality in Gynaecological Cancer Surgery: A Systematic Review
Authors: Katie Avelina Fernandes, Sonu Krishel Vidanagamage, Mevan Gooneratne, Robert Wilkinson, Madeleine Wells, Ching Ling Pang, Shaun M May, Georgia Alexandra Elizabeth Ellis, Elly Brockbank, Arjun Jeyarajah, Alexandra Lawrence, Saurabh Phadnis, Ranjit Manchanda, Sam Oxley, Michail Sideris
You can also read: Michail Sideris: 6 Decades of Pelvic Exenterative Surgery for Gynaecological Malignancies