Purpose: To evaluate prostate abscess outcomes and identify predictors for additional intervention. Additionally, assess a complexity score for clinical decision-making. Materials and methods: This is a retrospective, multicenter study including 62 consecutively admitted prostate abscess patients. Primary treatment involved antibiotics or a combination with transrectal ultrasound-guided aspiration (TRUS-GA) or transurethral resection of the prostate with abscess deroofing. Outcomes were assessed by comparing resolution with primary treatment to additional intervention. Additional intervention involved escalation to TRUS-GA, repeat TRUS-GA, or transurethral resection of the prostate with abscess deroofing. The sample was further categorized by Prostate Abscess Complexity Score (PACS) into complex and noncomplex. Risk factors, clinical features, abscess characteristics, and reintervention predictors were analyzed using logistic regression and receiver operating characteristic analysis. Results: Sixty-two patients were included, 47 (75.8%) responded well to primary treatment, while 15 (24.2%) needed additional intervention. Risk factors for additional intervention included multifocal abscess (P = 0.005), septa (P = 0.023), bilateral involvement (P = 0.008), abscess diameter ≥3 cm (P = 0.040), and volume ≥20 cc (P ≤ 0.001). The PACS, considering abscess focality, septa, laterality, polarity, abscess diameter, and volume, showed 73% sensitivity and 74% specificity in predicting additional intervention (P = 0.001) with a threshold of ≥6 points to classify as complex. Among complex cases, 80% required additional intervention, compared to 34% in noncomplex cases (P = 0.002). Conclusion: This study highlights personalized treatment for prostate abscess. Identified risk factors (multifocal abscess, septa, bilateral involvement, large abscess diameter, and increased volume) provide insights into predicting the need for additional interventions. The proposed PACS offers a valuable tool for clinical decision-making when ≥6 points are encountered.
Read full abstract