Background: The American Society of Anesthesiologists Physical Status (ASA-PS) classification is widely used for perioperative communication and research with known variation in agreement amongst assessors. Large language models (LLM) are increasingly considered for uniform decision making due to agreement challenges within clinicians under identical inputs. The study compared four contemporary LLMs against clinician-derived consensus and quantified within-model stability across repeated assessments. Methods: In a cross-sectional vignette study, 228 anesthesiologists from Denmark, England, and India classified 20 standardized perioperative vignettes through online survey. The modal response per vignette was considered as clinician consensus. Four LLMs (ChatGPT-5.2 Plus, Gemini 3 Pro, Perplexity AI Pro, Claude 4 Sonnet) received same set of vignettes through identical prompts. Temporal stability was assessed by repeating each vignette query nine times per model (three-time windows across three nonconsecutive days) in fresh sessions. Primary outcome was exact agreement with clinician consensus. Results: Consensus agreement for modal LLM classifications was 18/20 (90%) for ChatGPT, 17/20 (85%) for Gemini, 17/20 (85%) for Claude, and 15/20 (75%) for Perplexity. Disagreement is clustered in vignettes with weak or split clinician consensus. Stability differed by model: Claude was fully stable across all vignettes (20/20), Gemini 19/20, ChatGPT 18/20, and Perplexity 14/20; instability typically involved adjacent-class shifts. Conclusions: Contemporary LLMs often match clinician modal judgement and are largely temporally stable, with discordance concentrated in clinically ambiguous boundary cases.

Background: Post-anesthesia care unit (PACU) duration after total hip arthroplasty (THA) is relevant for ward transfer, early mobilization, and perioperative flow, but is inconsistently reported as a distinct recovery outcome. This study evaluated PACU duration after primary THA under spinal anesthesia in routine hospital-based practice and interpreted the observed duration in relation to contemporary arthroplasty literature. Methods: This retrospective observational cohort study included adult patients undergoing primary THA under spinal anesthesia within two units of a regional hospital. PACU duration was defined as time from PACU arrival until discharge to the orthopedic ward. Demographic, anesthetic, perioperative, and recovery variables were extracted from electronic medical records. Univariate analyses and multivariable linear regression were used to explore factors associated with PACU duration. A focused contextual literature comparison was performed to interpret the observed duration in relation to reported early recovery endpoints, with direct comparison anchored to conventional PACU or recovery-room intervals before ward transfer or immediate pathway progression. Results: Ninety-seven of 150 screened patients were included. Mean PACU duration was 162.3 ± 58.5 min. PACU duration was numerically longer in Unit A than Unit S, but the difference was not statistically significant. Reported early recovery endpoints in the literature were heterogeneous. In univariate analysis, male sex and exploratory spinal local anesthetic dose > 13 mg were associated with longer PACU duration. In multivariable analysis, male sex remained associated with longer PACU duration, whereas spinal dose was not independently associated when analyzed as a continuous variable. Conclusions: PACU duration after THA under spinal anesthesia should be interpreted as a multifactorial, workflow-sensitive outcome rather than a purely anesthetic recovery measure. Differences in discharge criteria, recovery pathways, and endpoint definitions limit direct comparison across studies. PACU duration may be most meaningful as an operational recovery and flow indicator within broadly comparable clinical settings.