Novel algorithm for management of acute epididymitis

• Published in: International journal of urology 2017-01, Vol.24 (1), p.82-87
• Main Authors: Hongo, Hiroshi, Kikuchi, Eiji, Matsumoto, Kazuhiro, Yazawa, Satoshi, Kanao, Kent, Kosaka, Takeo, Mizuno, Ryuichi, Miyajima, Akira, Saito, Shiro, Oya, Mototsugu
• Format: Article
• Language: English
• Subjects: Acute Disease - therapy; acute epididymitis; Age Factors; Aged; Algorithms; C-reactive protein; Clinical Decision-Making - methods; Decision making; Decision Support Techniques; Diabetes mellitus; diagnosis; Epididymitis; Epididymitis - diagnosis; Epididymitis - therapy; Feasibility Studies; Fever; Humans; Male; Middle Aged; Retrospective Studies; Risk Factors; Risk groups; Sensitivity and Specificity; severity; Severity of Illness Index; Statistical analysis; Urea
• ISSN: 0919-8172; 1442-2042
• DOI: 10.1111/iju.13236

Objectives To identify predictive factors for the severity of epididymitis and to develop an algorithm guiding decisions on how to manage patients with this disease. Methods A retrospective study was carried out on 160 epididymitis patients at Keio University Hospital. We classified cases into severe and non‐severe groups, and compared clinical findings at the first visit. Based on statistical analyses, we developed an algorithm for predicting severe cases. We validated the algorithm by applying it to an external cohort of 96 patients at Tokyo Medical Center. The efficacy of the algorithm was investigated by a decision curve analysis. Results A total of 19 patients (11.9%) had severe epididymitis. Patient characteristics including older age, previous history of diabetes mellitus and fever, as well as laboratory data including a higher white blood cell count, C‐reactive protein level and blood urea nitrogen level were independently associated with severity. A predictive algorithm was created with the ability to classify epididymitis cases into three risk groups. In the Keio University Hospital cohort, 100%, 23.5%, and 3.4% of cases in the high‐, intermediate‐, and low‐risk groups, respectively, became severe. The specificity of the algorithm for predicting severe epididymitis proved to be 100% in the Keio University Hospital cohort and 98.8% in the Tokyo Medical Center cohort. The decision curve analysis also showed the high efficacy of the algorithm. Conclusions This algorithm might aid in decision‐making for the clinical management of acute epididymitis.