Case-Mix Variables and Predictors for Outcomes of Laparoscopic Hysterectomy: A Systematic Review

The assessment of surgical quality is complex, and an adequate case-mix correction is missing in currently applied quality indicators. The purpose of this study is to give an overview of all studies mentioning statistically significant associations between patient characteristics and surgical outcomes for laparoscopic hysterectomy (LH). Additionally, we identified a set of potential case-mix characteristics for LH. This systematic review was conducted according to the Meta-Analysis of Observational Studies in Epidemiology guidelines. We searched PubMed and EMBASE from January 1, 2000 to August 1, 2015. All articles describing statistically significant associations between patient characteristics and adverse outcomes of LH for benign indications were included. Primary outcomes were blood loss, operative time, conversion, and complications. The methodologic quality of the included studies was assessed using the Newcastle-Ottawa Quality Assessment Scale. The included articles were summed per predictor and surgical outcome. Three sets of case-mix characteristics were determined, stratified by different levels of evidence. Eighty-five of 1549 identified studies were considered eligible. Uterine weight and body mass index (BMI) were the most mentioned predictors (described, respectively, 83 and 45 times) in high quality studies. For longer operative time and higher blood loss, uterine weight ≥ 250 to 300 g and ≥500 g and BMI ≥ 30 kg/m(2) dominated as predictors. Previous operations, adhesions, and higher age were also considered as predictors for longer operative time. For complications and conversions, the patient characteristics varied widely, and uterine weight, BMI, previous operations, adhesions, and age predominated. Studies of high methodologic quality indicated uterine weight and BMI as relevant case-mix characteristics for all surgical outcomes. For future development of quality indicators of LH and to compare surgical outcomes adequately, a case-mix correction is suggested for at least uterine weight and BMI. A potential case-mix correction for adhesions and previous operations can be considered. For both surgeons and patients it is valuable to be aware of potential factors predicting adverse outcomes and to anticipate this. Finally, to benchmark clinical outcomes at an international level, it is of the utmost importance to introduce uniform outcome definitions.