Application of the Clavien-Dindo classification to a pediatric surgical network

Introduction: A comprehensive validated system to evaluate surgical complications is required in our specialty to facilitate comparison and audit. The Clavien-Dindo (CD) classification of post-surgical complications was originally described in an adult general surgical setting in 1992 and has become widely used. We aimed to apply this to a pediatric surgical setting.

Methods: Data were collected on emergency and elective surgical activity together with complications in a prospective audit over a recent 4-month period in three geographical conjoined regional pediatric surgical units (including two major trauma centres). Briefly the CD classification codes complications according to degree of harm and magnitude of intervention required [I - V (death) with III and IV sub-divided according to whether general anesthesia was needed]. Length of stay and mode of admission were recorded. Data are given as median (range). Non-parametric comparison was used, and a p value of <0.05 was regarded as significant.

Results: During the period JULY - OCT 2018 (inclusive), there were 1822 admissions (elective, n = 1186: emergency, n = 636) and 1556 operations (elective, n = 1189, and of these 393 were urological). There were 69 patient complications: CDI (n = 7), CD-II (n = 19), CD-IIIa (n = 4), CD-IIIb (n = 28), CD-IV (n = 4), CD-V (n = 7). Deaths were principally in neonates and due to NEC (n = 6) at 2.5 (1-140) days post-operatively. There was a single post-traumatic death in an adolescent. LOS was 9 (0-217) days in CD I-IV. The incidence of any complication was 4.4%, of serious complication (defined as ≥CD III) 2.6% (A = 2.1%, B = 2.0%, and C = 3.2%: p = 0.16), and of death 0.45%. The most frequent complications were wound infection (n = 12) and post-appendicectomy collections/abscess (n = 10).

Conclusions: This appears to be the 1st report of the C-D classification in a general pediatric surgery network and can be considered a benchmark. The risk of death or serious harm is very low in such a practice.

Type of study: Prospective Cohort Study.

Level of evidence: IIb.