This title appears in the Scientific Report :
2022
Please use the identifier:
http://dx.doi.org/10.1007/s00423-022-02456-1 in citations.
Please use the identifier: http://hdl.handle.net/2128/31100 in citations.
Machine learning prediction model for postoperative outcome after perforated appendicitis
Machine learning prediction model for postoperative outcome after perforated appendicitis
Purpose Appendectomy for acute appendicitis is one of the most common operative procedures worldwide in both childrenand adults. In particular, complicated (perforated) cases show high variability in individual outcomes. Here, we developedand validated a machine learning prediction model for postope...
Saved in:
Personal Name(s): | Eickhoff, Roman M. (Corresponding author) |
---|---|
Bulla, Alwin / Eickhoff, Simon B. / Heise, Daniel / Helmedag, Marius / Kroh, Andreas / Schmitz, Sophia M. / Klink, Christian D. / Neumann, Ulf P. / Lambertz, Andreas | |
Contributing Institute: |
Gehirn & Verhalten; INM-7 |
Published in: | Langenbeck's archives of surgery, 407 (2022) 2, S. 789 - 795 |
Imprint: |
Heidelberg
Springer
2022
|
DOI: |
10.1007/s00423-022-02456-1 |
Document Type: |
Journal Article |
Research Program: |
Neuroscientific Data Analytics and AI |
Link: |
OpenAccess |
Publikationsportal JuSER |
Please use the identifier: http://hdl.handle.net/2128/31100 in citations.
Purpose Appendectomy for acute appendicitis is one of the most common operative procedures worldwide in both childrenand adults. In particular, complicated (perforated) cases show high variability in individual outcomes. Here, we developedand validated a machine learning prediction model for postoperative outcome of perforated appendicitis.Methods Retrospective analyses of patients with clinically and histologically verified perforated appendicitis over 10 yearswere performed. Demographic and surgical baseline characteristics were used as competing predictors of single-patient out-comes along multiple dimensions via a random forest classifier with stratified subsampling. To assess whether complicationscould be predicted in new, individual cases, the ensuing models were evaluated using a replicated 10-fold cross-validation.Results A total of 163 patients were included in the study. Sixty-four patients underwent laparoscopic surgery, whereasninety-nine patients got a primary open procedure. Interval from admission to appendectomy was 9 ± 12 h and duration ofthe surgery was 74 ± 38 min. Forty-three patients needed intensive care treatment. Overall mortality was 0.6 % and morbid-ity rate was 15%. Severe complications as assessed by Clavien-Dindo > 3 were predictable in new cases with an accuracyof 68%. Need for ICU stay (> 24 h) could be predicted with an accuracy of 88%, whereas prolonged hospitalization (greaterthan 7–15 days) was predicted by the model with an accuracy of 76%.Conclusion We demonstrate that complications following surgery, and in particular, health care system-related outcomeslike intensive care treatment and extended hospitalization, may be well predicted at the individual level from demographicand surgical baseline characteristics through machine learning approaches. |