Browse the corpus
Walk the Even Hospital Database by book and chapter — the raw source passages that ground Ask, DDx, and the rest.
1 passage
Artificial intelligence-powered analysis of operating room turnover: impact of instrument burden. OBJECTIVE: Improved operating room (OR) efficiency provides greater patient throughput, reduced costs, and maximal patient care. The aim of this study was to quantify and compare OR turnover efficiency across neurosurgical and otorhinolaryngology (ENT) specialties using artificial intelligence (AI) cameras. METHODS: A prospective study was conducted after obtaining IRB approval. AI-powered cameras documented operative turnover processes for cranial, spinal, and ENT cases at a tertiary academic center during a 30-day period. The software initiated recording when a patient exited the OR, and stopped recording upon entry of the subsequent patient, ensuring patient anonymity. Parameters included instrument tray count and personnel tracking. Turnover subprocesses were classified into clearing, cleaning, waiting, and instrument setup. RESULTS: The AI model successfully itemized turnover parameters for 53 operative turnovers (6 cranial, 32 spinal, and 15 ENT cases). Case duration averaged 175.4 (SD 86.1) minutes for cranial, 120.4 (SD 50.8) minutes for spinal, and 67.6 (SD 34.6) minutes for ENT cases. The overall average was 111.7 (SD 58.1) minutes. The mean turnover durations were 56.6 (SD 9.3) minutes for cranial cases, 52.2 (SD 20.3) minutes for spinal cases, and 40.4 (SD 15.5) minutes for ENT cases (p = 0.079). Clearing, cleaning, and waiting did not reveal any significant differences between specialties. A multivariate analysis did not reach significance after comparing the different ORs or different intragroup surgeons. Instrument setup duration emerged as the greatest determinant of variability: mean 38.3 (SD 13.4) minutes for cranial cases, 27.7 (SD 11.7) minutes for spinal cases, and 17.1 (SD 8.3) minutes for ENT cases (p = 0.0005). Instrument setup was significantly correlated with the number of instrument trays (R2 = 0.33, p < 0.0001), adding 2.7 minutes per additional tray. CONCLUSIONS: AI vision systems provided automated comparisons of OR turnover parameters, highlighting distinct bottlenecks in cranial, spinal, and ENT cases. Optimizing instrument setup through tray rationalization represents a cost-effective intervention that warrants further investigation.