# Perioperative Risk Prediction in Major Gynaecological Oncology Surgery: A National Diagnostic Survey of UK Clinical Practice

## TL;DR

UK gynaecological oncology surgeons rarely use formal risk prediction tools before major surgeries, despite a strong desire for better, specialized models to improve patient outcomes.

## Contribution

The study identifies a significant gap in the adoption of perioperative risk prediction tools in gynaecological oncology and highlights the need for GO-specific models.

## Key findings

- Only 7.4% of UK gynaecological oncology surgeons use risk prediction tools routinely for all major surgeries.
- 80% of respondents expressed a desire for more accurate, gynaecological oncology-specific risk prediction models.
- Over 20% of surveyed clinicians do not use any formal risk prediction tools for perioperative diagnostics.

## Abstract

Background: Gynaecological oncology (GO) surgery involves a wide range of procedures, from minor diagnostic interventions to highly complex cytoreductive operations. Accurate perioperative diagnostics—particularly in major surgery—are critical to optimise patient care, predict morbidity, and facilitate shared decision-making. This study aimed to evaluate current practices in perioperative risk assessment amongst UK GO specialists, focusing on the use, perception, and applicability of diagnostic risk prediction tools. Methods: A national multicentre survey was distributed via the British Gynaecological Cancer Society (BGCS) to consultants, trainees, and nurse specialists. The questionnaire examined clinician familiarity with and use of existing tools such as POSSUM, P-POSSUM, and ACS NSQIP, as well as perceived reliability and areas for improvement. Results: Fifty-four clinicians responded, two-thirds of whom were consultant gynaecological oncologists. While 51.9% used morbidity prediction tools selectively, only 7.4% used them routinely for all major surgeries. The most common models were P-POSSUM (39.6%) and ACS NSQIP (25%), though over 20% did not use any formal tool. Despite this, 80% of respondents expressed a desire for more accurate, GO-specific models. Conclusions: This study reveals a gap between available perioperative diagnostics and real-world clinical use in GO surgical planning. There is an urgent need for validated, user-friendly, and GO-specific risk prediction tools—particularly for high-risk, complex surgical cases. Further research should focus on prospective validation of tools such as ACS NSQIP and their integration into routine practice to improve outcomes in gynaecological oncology.