When it comes to safety, neuraxial anesthesia enjoys a slight upper hand over general anesthesia as a primary anesthetic technique in noncardiac surgery.
“Neuraxial anesthesia is associated with fewer complications after surgery, including surgical site infections, deep vein thrombosis and pneumonia,” said Gausan R. Bajracharya, MD, a clinical research fellow in the Department of Outcomes Research at the Cleveland Clinic, in Ohio. “However, the effect of neuraxial anesthesia on long-term outcomes remains unclear, as there have been no large robust trials on this topic and available studies are either small or inconclusive.”
The researchers used the American College of Surgeons–National Surgical Quality Improvement Program (NSQIP) database for the years 2011-2015. Given that major orthopedic surgery is commonly conducted with neuraxial or general anesthesia, they restricted their analysis to adults undergoing these procedures. Of 3,274,413 surgeries in the database, 323,420 patients underwent orthopedic surgery lasting at least one hour under either general or neuraxial anesthesia. The groups were matched (1:1) by propensity score and surgical type.
As reported at the 2017 annual meeting of the American Society of Anesthesiologists (abstract BOC06), Dr. Bajracharya revealed that among 72,877 matched pairs, the overall 30-day readmission rate was 4.0% in the neuraxial group and 4.4% in the general anesthesia group (P<0.001). Furthermore, neuraxial anesthesia did not reduce mortality compared with general anesthesia.
Nevertheless, neuraxial anesthesia was significantly associated with a slightly shorter hospital length of stay (2.9 vs. 3.1 days; P<0.001). Patients in the neuraxial group were also 13% more likely (hazard ratio, 1.13; 98.5% CI, 1.12-1.15) to be discharged alive from the hospital than their counterparts who received general anesthesia (P<0.001). “Again,” Dr. Bajracharya said, “the effect was statistically significant but of clinically trivial magnitude.”
The investigators examined the relative rates of VTE in the same two patient groups (abstract A1176). They used a logistic regression model to assess the effect of neuraxial anesthesia on VTE—which was defined as deep vein thrombosis or pulmonary embolism—within 30 days after surgery.
In the End, a Matter Of Preference
The researchers acknowledged several potential limitations of their analysis, most of which were born of its retrospective design. “Obviously we cannot make any statements about causality,” Dr. Bajracharya said. “What’s more, the NSQIP database comes with its own set of limitations, including a lack of intraoperative data. For example, intraoperative hypotension has been shown to be significantly associated with 30-day mortality in noncardiac surgery, but we can’t get that information from the database.”
Taking these findings into consideration, the authors concluded that patients and anesthesiologists can reasonably select general or neuraxial anesthesia, as preferred. “Our results do not strongly support neuraxial anesthesia being markedly preferable to general anesthesia,” Dr. Bajracharya concluded.
According to Daniel I. McIsaac, MD, MPH, an assistant professor of anesthesiology at the University of Ottawa, in Ontario, the findings of the analysis are comparable to those of previous research. “Your results are pretty familiar,” Dr. McIsaac said. “Most studies like this don’t tend to find a difference in mortality, as these models tend to discriminate well for death risk. But readmission and length of stay are more difficult to model. This may be because factors that influence readmissions and length of stay vary a lot more between hospitals and aren’t usually measured. One way to deal with this would be to cluster your analysis by hospital, but I assume in your data set you’re not able to do this.
“I’m wondering if you have any insights into whether some of the effect you found with length of stay and readmission might be due to your inability to adjust for differences between hospitals, and not necessarily hard data,” Dr. McIsaac added.
“A limitation of the NSQIP database is that we cannot distinguish among hospitals and thus stratify our analysis by care site,” Dr. Bajracharya replied. “A model that could account for differences between hospitals would have given us more insights. But we can only work with what the database provides us.”