- Design
- Single-centre retrospective cohort with 1:1 propensity score matching and adjudicated outcomes
- Population
- 2,184 patients aged 65 and over having major non-cardiac surgery with perioperative high-sensitivity troponin T surveillance; 612 matched pairs analysed
- Primary outcome
- Adjudicated ischaemic myocardial injury after non-cardiac surgery
- Effect
- 14.2% vs 8.7%, odds ratio 1.75 (95% CI 1.23-2.49, p=0.002); 30-day major adverse cardiac events HR 1.68 (1.12-2.51); 1-year mortality HR 1.54 (1.07-2.22); highest hypotension quartile OR 2.84 (1.62-4.97)
Myocardial injury after non-cardiac surgery is common, mostly silent, and predicts death. This retrospective cohort asked whether patients whose autonomic reflexes are already impaired tolerate intraoperative hypotension worse, using a phenotype definable from the electronic record rather than from formal autonomic testing.
Autonomic vulnerability was defined as diabetes or hypertension plus one of documented neuropathy, orthostatic hypotension, syncope or presyncope, or unexplained resting bradycardia or chronotropic incompetence. Among 2,184 patients aged 65 or over having major non-cardiac surgery with routine troponin surveillance, 612 propensity-matched pairs were analysed.
Adjudicated ischaemic myocardial injury occurred in 87 of 612 (14.2%) with the phenotype and 53 of 612 (8.7%) without (odds ratio 1.75, 95% CI 1.23-2.49, p=0.002). Postoperative troponin elevation, 30-day major adverse cardiac events (hazard ratio 1.68, 1.12-2.51) and one-year mortality (1.54, 1.07-2.22) were all raised. Intraoperative hypotension burden modified the effect (p for interaction 0.018): in the highest hypotension quartile the odds ratio reached 2.84 (1.62-4.97).
The interaction is the clinically interesting claim and also the weakest, being an interaction term in a retrospective single-centre study - the authors call the whole thing hypothesis-generating and ask for external validation with formal autonomic testing. Propensity matching balances what was measured; the patients with this phenotype differ in ways that were not.
Still, the phenotype costs nothing to look for, and if hypotension harms these patients more, they are the ones for whom a tighter pressure target and an arterial line are worth the trouble. Reviewing the preoperative clerking for neuropathy, postural symptoms and unexplained bradycardia in an older diabetic or hypertensive patient takes a minute and changes how you plan the anaesthetic.
- Ask older diabetic or hypertensive patients about postural dizziness, syncope and neuropathy at preassessment
- Note unexplained resting bradycardia - it is part of the phenotype and easily dismissed
- Consider a tighter blood pressure target and invasive monitoring in these patients
- Ensure postoperative troponin surveillance in this group; the injury is usually silent
- The interaction with hypotension burden is retrospective and hypothesis-generating
The statistics, in plain English
An odds ratio of 1.75 translates here into 14.2% against 8.7% - about five and a half extra events per hundred patients, which is a large absolute difference for an outcome that predicts death. The interaction p value of 0.018 says hypotension burden and the phenotype are not simply additive, but interaction tests are underpowered even in large studies and are the analysis most likely to be a false positive. Propensity matching removes imbalance in measured variables only; the phenotype is defined by documentation, so patients who were asked about postural symptoms differ systematically from those who were not.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for anaesthesiology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free