Carbon dioxide, blood pressure, and postoperative stroke: a retrospective case-control study

Carbon Dioxide, Blood Pressure, and Perioperative Stroke: A Retrospective Case–Control Study

Carbon dioxide, blood pressure, and postoperative stroke: a retrospective case-control study

Abstract

ROMANTIC

The relationship between intraoperative physiologic changes and postoperative stroke is unclear
.
Previous data suggest that hypo/hypercapnia is associated with decreased cerebral blood flow (eg, due to hypotension) leading to ischaemia
.
This study hypothesized that intraoperative hypotension combined with hypo/hypercapnia was associated
with postoperative ischaemic stroke.

We conducted a retrospective, case-control study through a multicentre intraoperative outcome group
.
From 2004-2015, patients
(> 18 years of age) were selected from 5 major research centers to undergo non-cardiac, non-cranial and non-macrovascular surgery.
The ischemic stroke event was reviewed by manual charts, and the ischemic stroke event matched
the control 1:4.
The area under
the curve was calculated in terms of time abscissa and ordinate below the MAP threshold (below 55mmHg, less than 60mmHg, less than 65mmHg) and abnormal end-of-breath carbon dioxide (less than 30mmHg, less than 35mmHg, and above 45mmHg).
After adjusting for relevant confounders, the association
between stroke and the area under the curve was studied.

We analysed a total of 1244881 patients
.
Stroke manifestations
in 1702 patients.
126 patients were matched to 500 patients
.

The area under the curve under the set threshold was associated with stroke, with MAP below 55 mmHg (adjusted OR: 1.
17/10 mmHg-min [95% CI, 1.
10 -1.
23], P < 0.
0001) and end-of-breath carbon dioxide above 45 mmHg (adjusted OR: 1.
11 10 mmHg-min [95% CI, 1.
10 to 1.
11], P < 0.
0001)
。 Blood pressure does not interact
with carbon dioxide.

Conclusion: Intraoperative hypotension and abnormal carbon dioxide may be independent factors
increasing the risk of postoperative stroke.

Figure&Table

ROMANTIC

flow chart

Bivariate analysis

Seemingly uncorrelated regression model

Baseline characteristics of the patient