Investigators enrolled 122 patients aged 60 years and older undergoing major surgery into a prospective cohort. CRP was measured before surgery and again on postoperative day 1. Delirium surveillance used the 3D-CAM twice daily across postoperative days 1–3. The study team analysed two primary outcomes: delirium incidence (any positive 3D-CAM during surveillance) and peak delirium severity (the maximum 3D-CAM score recorded for each patient).
Twenty of 122 patients (16.4%) developed postoperative delirium (POD). The cohort’s mean peak severity score was 2.1 (SD 2.1). Data indicated that postoperative CRP was associated with both the likelihood of delirium and with higher peak severity scores; the authors reported high predicted probabilities for these associations in their analyses.
CRP sampling occurred at two timepoints: a preoperative baseline and a single postoperative measurement on day 1. Delirium assessment relied on the 3-minute diagnostic Confusion Assessment Method (3D-CAM) administered twice daily for three days. The paper provided these measurement windows clearly; however, the source text did not include additional timing details such as exact clock times of sampling, laboratory assay methods, or analytic thresholds used for CRP categorisation in multivariable models.
The study team observed that higher postoperative CRP correlated with greater delirium severity and with higher incidence of POD. They reported predicted probabilities that linked postoperative CRP to delirium severity (reported predicted probability 94.7%) and to delirium incidence (reported predicted probability 99.9%). The investigators therefore interpreted postoperative CRP as independently associated with both outcomes in this cohort.
The abstract does not provide the full multivariable model coefficients, covariate adjustment set, or sensitivity analyses in the source text supplied here. Those methodological specifics—such as which baseline risk factors were adjusted for, how CRP was modelled (continuous or categorical), and whether preoperative CRP retained independent association—were not reported in the provided extract.
A secondary analysis examined whether perioperative anti-emetic doses of dexamethasone altered inflammatory response in this older cohort. The authors reported that dexamethasone was associated with a 41% reduction in postoperative CRP (95% credible interval 8–62%), and they reported a predicted probability for this reduction of 99.1%.
The source explicitly describes the effect size and uncertainty bounds but does not state the exact dexamethasone dose used, timing relative to induction or surgery, or whether dexamethasone allocation was randomised or observational in this cohort. Those dosing and assignment details were not available in the supplied text.
The authors confirmed institutional ethics approval from the Sydney Local Health District Ethics Committee for Royal Prince Alfred Hospital (Protocol No. X23-0165 & 2023/ETH00923) and stated that participant consent procedures were followed. The manuscript indicates that data may be available on request, contingent on ethical guidance.
The source declared no competing interests and acknowledged ANZCA Foundation funding. The report is presented as a medRxiv preprint; peer review status is not indicated in the provided text.
The authors argued that their findings support trials focused on modulating inflammation to prevent postoperative delirium. The source did not report randomized trial data testing an anti-inflammatory intervention for delirium prevention in this cohort, nor did it supply precise dexamethasone dosing or a full description of model covariates and sensitivity checks in the abstract.
Key questions that remain open in the material provided: which specific perioperative anti-inflammatory strategy (dose, timing, agent) is optimal for reducing delirium risk; whether observed CRP associations persist after adjustment for detailed clinical confounders not listed here; and whether reductions in CRP translate to meaningful, reproducible decreases in delirium incidence and patient-centred outcomes in randomised trials. The study team recommended further trials to evaluate inflammation-targeted prevention strategies; the preprint did not report those trials’ designs, registration identifiers, or timelines.
Limitations explicitly noted by the authors in the supplied text are limited; therefore readers should interpret the reported associations with caution and consult the full manuscript for detailed methods, model specifications, and any additional limitations the authors documented.