Clinical epidemiological studies of women undergoing surgery for urogynaecological disorders.

UNLABELLED: This PhD thesis was performed during my employment at the Center for Clinical Epidemiology, Odense University Hospital and University of Southern Denmark, 2010-2013. It comprises an overview and four papers, two published in international peer-reviewed scientific journals, one under review, and one in draft.

INTRODUCTION: Urinary incontinence (UI) and pelvic organ prolapse (POP) are prevalent disorders among women worldwide, affecting their psychological and social wellbeing, with reductions in quality of life. Treatment options are conservative (e.g. pelvic floor exercises, weight loss, and bladder training), pharmacological, and surgical. Surgery has especially for UI undergone an improvement during the last decades with development of minimally-invasive sub-urethral sling procedures, and the number of surgeries has increased in Denmark and other countries.

AIMS: In a population of Danish women undergoing surgery for UI or POP, we aimed: to describe the establishment of the Danish Urogynaecological Database (DugaBase), and to evaluate the completeness and the validity of surgery registration in the DugaBase; to study patient reported outcome measures in Danish women undergoing urogynaecological surgery; to study the use of symptom-relieving drugs before and after surgery for UI; to study the use of antibiotics for urinary tract infection (UTI) before and after surgery for UI METHODS: Study I The completeness of DugaBase was assessed by comparing procedure codes in the DugaBase to iodes registered in the National Patient Registry, 2006-2010. The study also included review of medical journals from 200 women (computed randomly from DugaBase), representing 22 departments in Denmark. Information on selected variables was compared to registered data in the DugaBase.

DATA SOURCES: the National Patient Registry, the DugaBase, and medical records. Study II was based on a national cohort of women undergoing surgery for UI and POP registered in the DugaBase, 2006-2011. Clinical data and data from patient questionnaires were retrieved.

DATA SOURCES: the DugaBase. Study III+IV were cohort studies based on national register data and prescription data. A total of 2,151 women with a first time surgical procedure for UI within 1996-2010 were included. The data were supplemented with registry information on redeemed prescriptions on symptom-relieving drugs/antibiotics for UTI/oestrogen, comorbidity, and educational level and income.

DATA SOURCES: the Danish National Patient Registry, Odense University Pharmacoepidemiologic Database, and the Register for Education Statistics, and the Register of Family Income.

RESULTS: Study I: A total of 16,509 procedures were registered in the DugaBase by 31 December 2010. The database completeness has increased from 38.2% to 93.2% during the years. According to the validation, all 200 medical records were retrieved. The overall percentage agreement between medical records and the DugaBase was at least 90% for selected key variables: procedure code, date, hospital department, use of antibiotics, prior surgery, height, weight, parity, and smoking. Study II: In the study period, 20,629 urogynaecological procedures were performed in Denmark and reported to the DugaBase. For approximately one third of these women, the patient questionnaires on severity of symptoms and quality of life were completed both pre- and postoperatively, and thus included. After both UI and POP surgery, more than 80% had improved symptoms and showed a significant improvement in quality of life. Study III: A total of 2,151 women with a primary UI procedure were included. Of these, 358 (16.6%) were exposed to symptom-relieving drugs preoperatively, and 1,793 (83.4%) were not. Preoperative usage of symptom-relieving drugs was the strongest risk factor of postoperative use, both within 0-60 days OR (adjusted) = 33.0 (95% confidence interval (CI) 20.0-54.7)) and 61-365 days OR (adjusted) = 7.2 (95% CI 5.4-9.6)). Adjusted for age, procedure type, calendar year, comorbidity, preoperative use of oestrogen, educational level, and personal annual income. Only a minority of preoperative non-users started using symptom-relieving drugs postoperatively. Study IV: The same study population as in study III. Of the 2,151 women, 496 (23.1%) were antibiotic users prior to surgery, and 1,655 (76.9%) were non-users. Preoperative usages of antibiotics for UTI was a strong risk factor of postoperative use of the same antibiotics, both within 0-60 days OR (adjusted) = 2.6 (95% CI 2.9-3.5), and within 61-356 days OR (adjusted) = 4.5 (95% CI 3.5-5.7)). Adjusted for age, procedure type, calendar year, comorbidity, preoperative use of oestrogen, educational level, and personal annual income. Comorbidity and procedure types were found to be significant risk factors as well, although less important than preoperative antibiotic use.

CONCLUSIONS: Due to a high completeness and data of high validity the DugaBase offers a unique possibility for continuing quality assessment of urogynaecological surgery in Denmark, as well as for future research. Surgeries for UI and POP performed in Danish women were effective in alleviating symptoms of UI and POP and improvement in quality of life based on patient reported outcome measures. Our pharmacoepidemiological studies showed that preoperative use of symptom-relieving drugs and antibiotics for UTI were strong risk factors of postoperative usage of the same drugs.