Renal abscess with low morbidity but high mortality should be paid more attention by clinician during clinical work. The early and accurate diagnosis was considered as an important factor affecting the outcome of renal abscess . In addition, advanced age, lethargy and some laboratory examinations such as higher C-reactive protein levels and blood urea nitrogen are associated with poor prognosis and high probability of mortality [8, 15]. Although some studies had been conducted to discuss the best choice of treatment for patients with different abscess size, the appropriate choice is still controversial especially on the aspects that abscess size between 3 ~ 5 cm and waiting for a response to antibiotic therapy may delay surgical intervention, thus lead to poor prognosis. In 1996, Siegel proposed that renal abscess should be treated according to the size of the abscess . In his proposal, small renal abscess (< 3 cm) were treated with antibiotics only, large renal abscess (> 5 cm) with invasive treatment, and the choice of treatment for medium renal abscess (3 ~ 5 cm) was determined by the patients’ conditions and the response to the antibiotics. Both Lee BE and Brian JL believed that percutaneous abscess drainage should be the primary treatment for renal abscess larger than 3 cm [10, 16]. In a clinical study conducted by Dalla Palma, however, he concluded that four weeks of antibiotic therapy could completely cure renal abscess with a diameter of 5 cm or less, thus avoiding invasive treatment . The similar conclusion was made by Lee SH . Hung et al in their study comparing the outcome of percutaneous and surgical treatment also found that larger abscess size appeared to have poor prognosis, but statistically insignificant . Although there is still a debate on the preference of treatment selection for renal abscess basing on size, we can’t deny that size of abscess is still an important factor that determining treatment modalities. In our study, the primary factor for the choice of initial treatment was size, thus, the difference in the size of renal abscess between the two treatment groups was significantly. But whether there are other factors can be used to guide clinical treatment, especially for the patients with renal abscess between 3 ~ 5 cm is still unknown.
In a Korean study , 56 patients with renal and perirenal abscesses were analyzed, with a mean age of 53.5 years, a 75% female representation and a 44.6% incidence of diabetes. Lee SH et al conducted a study on 51 patients with renal or perirenal abscess measured 5 cm or less, with 49 abscesses received intravenous antibiotics alone and successful outcome was presented. In their study, 91.8% patients were women, the mean age was 42.3 years and the incidence of diabetes mellitus was 46.9% . Though rare, renal abscess can also occur in children . In our study, the mean age was 48.87 ± 15.89 years, the ratio of female was 63.16% and 42.11% patients got diabetes mellitus which were conformed to the study preview. No pediatric patient was included in our study. The mean age of patients receiving conservative treatment was slightly younger than that of patients receiving invasive treatment, but no statistical difference was found. While the diabetes mellitus was significantly different between these two groups. Diabetes mellitus is more common in patients receiving invasive treatment. In a research conducted by Ko MC et al, he found that compared to individuals aged ≥ 65 years, younger diabetic patients tended to have a higher risk of renal abscess which might be attributed to the reasons that younger diabetic patients were more likely to have poor blood sugar control, poor health-related behavior, less frequent clinic visits, and irregular assessment of diabetes-related complications . It is knowledgeable that diabetes mellitus can increase the risk of infection  and the difficulty of infection control. This may be one reason that why renal abscess patients with diabetes mellitus in our study received more invasive treatment. Ko’s study may also explain the phenomenon of higher incidence of diabetes mellitus in patients receiving invasive treatment.
Recently, with the widespread use of antibiotic, the common causative organisms of renal abscess changed from S. aureus to Gram-negative bacteria represented by E. Coli and Proteus spp. [2, 20]. Yamamichi F et al also found the most causative organisms was E. Coli in renal abscess, with a rate of 23% . Our study also found the similar results.
Lin HS et al reported that thrombocytopenia was related with the poor outcome of renal abscess . In our study, the average value of PLT was 285.08 ± 129.44, with patients receiving invasive treatment had higher count of PLT. This may be explained by the risk of bleeding during the procedure of invasive treatment with low count of PLT. RDW as a predictor of severe morbidity and mortality in some chronic diseases such as congestive heart failure was well-known by ICU physicians. The role of RDW in predicting severity and mortality in patients with Gram-negative bacteremia has been reported by some scholars . In the background of Gram-negative bacteria being the foremost pathogenic bacteria in renal abscess, higher value of RDW may indicate the severity of the renal abscess and the difficulty in controlling it. In our study, we found that patients receiving invasive treatment got a higher value of RDW, which was also retained in the multivariate analysis.
Other than that, fever ≥ 38.5℃, BUN, PLR, AG and WBC in urine were also found significant difference between conservative and invasive treatment. But none of them were retained in the multivariate analysis. Only RDW, abscess size and the comorbidity of diabetes mellitus were found to be the significant factors that may affect the choice of treatment for renal abscess. The result of ROC curve revealed that 14.7% and 3.9 cm may be the cut-off for RDW and size, respectively. In general, for patients with RDW > 14.7%, abscess size > 3.9 cm and diabetes mellitus, invasive treatment should be taken into consideration, especially in patients with renal abscess between 3 ~ 5 cm.
However, there are still some limitation in our study. Firstly, our study is a retrospective clinical study. In addition to the size of the abscess and the condition of the patient, the choice of treatment may also be influenced by clinical experience at the time. Secondly, there is a limited number of cases. Since renal abscess is uncommon clinical disease, so more cases and multi-center joint study are needed. Thirdly, some factors such as C-reactive protein, erythrocyte sedimentation rate which were demonstrated to be associated with prognosis were not included in this study due to incomplete data.