Background: There is huge variation in Clavien-Dindo (CD) complication rates in urology. We sought to optimize the use of the CD system in kidney tumor surgery.
Methods: We retrospectively analyzed 1,286 patients undergoing kidney tumor operations in 12 Finnish hospitals during 2016-2017. Primary CD assignments were made by site urologists. Data were centrally reviewed by two authors in consensus meetings. Consistency of the primary assignments was assessed by the number of cases requiring correction. Complication load was compared as different outcome rates between five university hospital regions.
Results: The overall complication rate in primary data was 40% (517/1286) and varied significantly from 32 to 62% (p < 0.001) between the regions. The need for corrections in central review was significantly greater for CD1 (54%) compared to CD2 (16%, p < 0.001) and CD3-5 (11%, p < 0.001) categories. The final data comprised 500 CD complications after 390 surgeries. The most frequent pathologies were bleeding (8.4%), urological complications (5.9%) and postoperative fever (4.7%). The overall CD2 complications rate was statistically (p < 0.001) higher in region D and that of CD3-5 was higher (p = 0.007) in region B. In multivariable analysis, university hospital region, male sex, BMI ≥ 27, ECOG ≥ 1, partial nephrectomy type and open surgery significantly increased the risk of complications.
Conclusions: Comparative use of CD1 complications may be too inconsistent and only CD2-5 complications should be reported. Central review of the primary data and detailed guidelines are necessary.
Keywords: Clavien-Dindo classification; Kidney tumor surgery; nephrectomy; partial nephrectomy; postoperative complications; quality assessment.