- Design
- Retrospective cohort, US National Inpatient Sample 2016–2021
- Population
- 41.6 million adult surgical admissions, 2.1 million with HIV
- Primary outcome
- Perioperative organ injury and in-hospital mortality
- Effect
- Mortality 3.2% vs 2.2%; adjusted OR 1.66 (95% CI 1.54 to 1.78)
A retrospective analysis of the US National Inpatient Sample from 2016 to 2021 compared 2.1 million surgical admissions coded as HIV-positive with 39.5 million HIV-negative admissions.
People with HIV carried a much higher comorbidity burden. In-hospital mortality was 3.2% against 2.2%; after adjustment, HIV was associated with higher mortality (OR 1.66, 95% CI 1.54 to 1.78), about 2.7 extra days in hospital and less discharge home. The excess persisted, though smaller, in asymptomatic HIV, and was greatest in symptomatic disease or AIDS.
The reported 64.6% rate of perioperative organ injury in people with HIV, a third of it myocardial infarction, looks implausibly high and probably reflects coding in an administrative database. The mortality association is the more reliable signal. No CD4 count or viral load data were available.
- Record HIV status, current antiretroviral regimen, recent CD4 count and viral load in the preoperative assessment.
- Check for interactions between antiretrovirals and anaesthetic drugs, especially boosted regimens with midazolam or fentanyl.
- Continue antiretroviral therapy through the perioperative period without interruption.
- Assess cardiovascular and renal risk carefully, as comorbidity drove much of the excess.
Why it matters
India has one of the world's largest populations living with HIV, many now old enough for major surgery.
Don't overread it
Administrative coding without CD4 or viral load data; organ injury rates in particular look inflated by coding.
The statistics, in plain English
An adjusted odds ratio of 1.66 means the odds of dying in hospital were about two-thirds higher with HIV after accounting for measured factors. With 41 million records almost any difference is statistically significant; the question is whether it is real and clinically important, which coding limits make hard to judge.
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