Prediction of major complications in patients undergoing laparoscopic and open hysterectomy for benign indications

introduction

Hysterectomy is one of the most frequently performed surgical procedures. Canada has one of the highest hysterectomy rates globally, with a third of women undergoing the procedure before the age of 60. Both clinicians and patients prefer the minimal access approach, and the proportion of hysterectomies performed using the laparoscopic approach has increased substantially in many countries over the past 10 years.

The evidence-based medical paradigm for the surgical approach to hysterectomy for benign disease recommends that the surgical approach chosen is discussed with the patient by the surgeon and made on the basis of the circumstances, benefits and relative risks.

Most physicians performing a hysterectomy will intuitively identify patient characteristics that have the potential to increase the complexity and complications of surgery. A 2016 systematic review of studies reporting a significant association between patient characteristics and surgical outcome for laparoscopic hysterectomy and a 2020 population-based prospective cohort study using data from the Danish Hysterectomy Database suggested that advanced age, race, increased body mass index (BMI), diabetes mellitus, uterine weight gain, fibroids, endometriosis, and adhesions were predictors of complications in patients undergoing hysterectomy for benign indications.

However, assimilating this information to individualize and anticipate each patient’s exact risk if multiple factors are present can be challenging. A 2020 systematic review reports that risk prediction models outperform surgeons of other specialties in estimating risk and postoperative outcomes; its discriminatory capacity shows greater variation (statistical C 0.51-0.75) than other risk prediction tools.

Patients should be educated about potential risks before surgery to manage expectations. This is especially important when considering surgery for benign disease because non-surgical options are often available.

We objective is to generate predictive models that can be used in conjunction with surgeon intuition to improve preoperative patient counseling and match progress made in the technical aspects of surgery. We sought to quantify the proportion of patients undergoing hysterectomy for benign disease who would have major complications, and to develop and validate a prognostic model to individualize this risk, using national datasets.

Background:

Hysterectomy, the most common gynecological surgery, requires a surgeon to advise women about the risks of surgery. Our aim was to develop and validate a multivariable logistic regression model to predict the major complications of abdominal or laparoscopic hysterectomy for benign conditions.

Method:

We obtained routinely collected administrative health data from the UK National Health Service (NHS) from 2011 to 2018. We defined major complications based on core outcomes for postoperative complications, including ureteral, gastrointestinal, and vascular injuries, and major complications from injuries.

We determine 11 predictors a priority. We used internal-external cross-validation to assess discrimination and calibration in 7 NHS regions in the developmental cohort. We validated the final model using data from additional NHS regions.

Results:

We found that major complications occurred in 4.4% (3,037/68,599) laparoscopic hysterectomies and 4.9% (6,201/125,971) abdominal hysterectomies. Our model showed consistent discrimination in developmental groups (laparoscopic, C statistic 0.61, confidence interval [IC] 95% from 0.60 to 0.62; abdomen, C statistic 0.67, 95% CI 0.64 to 0.70 and similar or better discrimination in cohort validation (laparoscopic, C statistic 0.67, 95% CI 0.65 to 0.69, abdominal C statistic 0.67, 95% CI 0.65 to 0.69).

Adhesion was the most predictive complication in both models (laparoscopy, odds ratio [OR] 1.92, 95% CI 1.73 to 2.13; stomach, OR 2.46, 95% CI 2.27 to 2.66). Other predictors of complications included adenomyosis in the laparoscopic model and Asian ethnicity and diabetes in the abdominal model. Protective factors included age and a diagnosis of menstrual disturbances or a benign adnexal mass in both models and a diagnosis of fibroids in the abdominal model.

Calibration plots for prediction of major complications in patients undergoing (A) laparoscopy or (B) abdominal hysterectomy for benign indications. Note: AUC = area under the curve, CITL = overall calibration, E = expected, O = observed.