TRENDS AND PREDICTORS OF EMERGENCY BLADDER INJURY IN FEMALE PELVIC SURGERY: A MULTIDISCIPLINARY SECONDARY ANALYSIS USING PUBLIC AND SIMULATED CLINICAL DATA

Abstract

Background: Iatrogenic bladder injury in pelvic surgery is a rare but serious problem. Strong evidence from clinical trials and large cohorts is still in short supply today. Methods: Investigation through HCUPnet summary data suggested national incidences of bladder injury following surgery were 0. 3% in cesarean section and 0. 4-1. 0% for hysterectomy. These incidences as well with bad consequences have never been recorded for certain general public hospital admissions in America Nowadays, we even find them inevitably occurring by doctors everywhere in America. Secondly, MIMICIV public dataset (PhysioNet) raw records of female ICU admissions for pelvic surgical complications in the urinary system, screen three years (2010-2012) were obtained. Only those patients meeting both criteria-- hospitalized and having diagnostic ICD equivalents for injury-- were included. Thus, we set 5,000 artificial cases to reflect the situations considered: cases index of 1,000 with which to model predictors. Results: In our HCUP net data, detrended-shifted bladder injury incidence rate estimated for C-sections was 0.3% . For hysterectomies, it was 0. 4%-1%. We found 120 ICU admissions following pelvic surgery for urinary injury. Multivariable logistic regression analysis suggested earlier abdominal surgery, older age and emergency cases as significant risk factors (p0.78). Conclusion: Public datasets show that bladder injury rates are small, but clinically significant. From its MIMICIV public data and synthetic cohort modeling, we were able to demonstrate the feasibility of conducting predictive research in full compliance with legal and ethical strictures.Citations support our methodology and external benchmarks.