Both CDC and CCI demonstrate comparable effectiveness, indicating their interchangeability. However, the CCI exhibits a marginal edge in forecasting ER admissions. This subtle discrepancy may suggest that this index performs better when multiple complications occur sequentially, in different admissions, rather than simultaneously. This interchangeability is also supported by the financial aspect, as there is no significant impact in predicting hospitalization costs using CCI as opposed to CDC. Ultimately, the decision on which scoring system to employ remains at the discretion of each service. Given these findings, transitioning from the CDC to the CCI may not be justifiable if it entails additional time costs.
Nevertheless, 2.9% of our patients exhibited multiple complications, indicating a CCI value exceeding the corresponding highest grade in the CDC. Thus, while CCI did not reveal a significant difference in accuracy compared to CDC, it demonstrated a sensitivity to detect the postoperative trajectory of these 2.9% patients that might be overlooked by the CDC.
Other authors compared CDC to CCI in ureteroscopy and in PCNL indicating no distinguishable difference between these scoring systems in relation to LOS[11]. Conversely, Boeri and colleagues did identify a disparity between CDC, r = 0.26, and CCI, r = 0.32[12]. The findings of our study are also consistent with a comparative analysis of CDC and CCI in oncological urology. Regarding radical prostatectomy and partial nephrectomy, an insignificant difference in their correlation with LOS was revealed[13]. Nonetheless, when examining radical cystectomy, an advantage of CCI over CDC in relation to LOS was observed[13]. Therefore, the utilization of CCI enhances the accuracy of estimating procedure morbidity[14].
We exclusively focused on patients aged 60 and older, contributing to a higher incidence of major complications than reported in other studies comparing these scoring systems[11, 12]. A second factor pertains to our hospital's nature—being a quaternary facility with patients generally exhibiting more severe conditions. This is evident in the distribution of ASA physical status, which tends to be higher than in those studies[12]. However, our 6.7% major complications rate (CDC III to V) is not far from the ratio of a 5,803 patients’ prospective trial for PCNL, which was 4.1%[15].
Regarding predictors of complications, although age did not emerge as a risk factor in our analysis, it is crucial to note that we deliberately focused on those aged 60 and above. Extrapolating our findings to patients of all ages might render age as a significant predictor. Despite that, we must disclaim that age itself may not be the primary risk factor, but the frailty and associated comorbidities that comes with aging[16].
The presence of a positive urine culture before surgery did not appear as a predictor of complications. This is likely attributed to our practice of administering pre-procedural antibiotics tailored to the sensitivity profile of the agent for patients with positive tests. Moreover, those with nephrostomy tubes continued antibiotics until tube removal. Nowadays, our protocol maintains antibiotics guided by urine culture prior to PCNL, but it has evolved to include only a single dose of antibiotics at the induction of the procedure with no continuation until tube removal[17–20].
Our study stands out as one of the early attempts to validate the CCI in lithiasis treatment. It distinguishes itself by comparing data from a highly homogeneous population, characterized by minimal age and ASA physical status variation. While our inclusion criteria limited to patients who underwent standard PCNL, it is unlikely that this influenced the incidence of complications[21]. It would be reasonable to extrapolate these findings to mini procedures.
However, it has limitations. As a retrospective study, even though based on prospective electronic medical records, there is always the possibility of data loss and possible detection bias. Furthermore, being a single center introduces potential patient selection bias. We do not have the precise financial cost of postoperative hospitalization for each patient, hindering our ability to correlate more accurately this information with either CDC or CCI scores.