[Prospective Cohort Study of In-Hospital Patients Undergoing Ophthalmic Surgery for the Validation of ClassIntra: Classification of Intraoperative Adverse Events].

Background: To the best of our knowledge, there is no validated classification to assess intraoperative adverse events (iAEs) in ophthalmic surgery. ClassIntra is a newly developed classification for surgery- and anaesthesia-related iAEs that has been recently validated in various surgical disciplines, but not in ophthalmic surgery. We aim to assess the validity and practicability of ClassIntra in patients undergoing ophthalmic surgery.

Methods: A consecutive sample of in-hospital patients undergoing any type of ophthalmic surgery was included in this single-centre prospective cohort study. iAEs were classified using ClassIntra, consisting of 5 severity grades according to the symptoms of the patient and the required treatment. All patients were followed for two weeks to record all postoperative adverse events according to Clavien-Dindo. The primary endpoint was the risk-adjusted association between the most severe iAE and the weighted sum of all postoperative adverse events within the two-week follow-up using the Comprehensive Complication Index (CCI). In addition, ophthalmologists and anaesthesiologists were asked to complete an online survey assessing the severity of iAEs for 10 fictitious clinical case scenarios. Reliability was assessed by comparing the clinicians' ratings to the prespecified benchmark rating of the study team.

Results: In this study, 100 in-hospital patients with an average age of 64 years (SD 15) were included. The majority of all patients were ASA II (n = 53, 53%) or III (n = 42, 42%). Thirty-two iAEs were recorded in 22 patients (17 grade I, 12 grade II, 3 grade III). Ninety-four postoperative adverse events occurred in 50 patients (44 grade I, 36 grade II, 14 grade IIIa). We found a mean difference in CCI of 2,1 (95% confidence interval [CI] - 2,5 to 6,8) per one unit increase in severity grades of ClassIntra. Fifty ophthalmologists and anaesthesiologists completed the online survey (response rate 54%). The intraclass correlation coefficient was 0,79 (95% CI 0,64 to 0,94).

Conclusions: The application of ClassIntra during daily routine in ophthalmic surgery showed the usefulness and practicability of this classification for the standardised assessment of intraoperative adverse events. Although construct validity could not be demonstrated, the good reliability in the survey's rating underlines the criterion validity of this newly developed classification in ophthalmic surgery.