Because hip fractures are a major cause of disability and mortality in the elderly population, determining what constitutes optimal surgical and postsurgical care for people who get hip fracture surgery remains a key focus for clinical research. One area of debate concerns whether patients undergoing hip fracture surgery fare better with regional (spinal or epidural) or general anesthesia.
Today, new findings released in JAMA suggest that compared with general anesthesia, regional anesthesia may be associated with a shorter hospital stay but not with a lower mortality rate.
To study the association between anesthesia type used during hip fracture surgery and surgical outcomes (namely the length of hospital stay and the risk of death within 30 days of surgery), researchers from the University of Pennsylvania collected information on more than 50 000 patients who underwent hip fracture surgery in New York state from 2004-2011. However, this question is a difficult one to study because patients who get regional anesthesia tend to be sicker than those given general anesthesia, which may bias the results (regional anesthesia is thought to be safer and have fewer adverse effects than general anesthesia).
To reduce this potential bias in their study, the investigators used a special analytic method called an instrumental variable analysis. Their analysis used the distance the patients lived from the hospital where they received surgery as an “instrument” of comparison so that instead of comparing outcomes for regional vs general anesthesia per se, the results compared outcomes for those who lived near hospitals specializing in regional anesthesia vs those who lived near hospitals specializing in general anesthesia. Comparisons using this distance as an “instrumental variable” theoretically controls for unknown imbalances in the 2 patient populations, something that is often difficult to do in observational research when randomization is not possible.
Results showed that there was no difference in mortality between individuals who lived near a hospital specializing in regional anesthesia vs one specializing in general anesthesia (5.4% vs 5.8%, respectively). However, there was a difference in length of hospital stay observed, with regional anesthesia being associated with a 0.6-day shorter length of stay than general anesthesia.
Lead author Mark Neuman, MD, MSc, discusses these findings with news@JAMA.
news@JAMA: Why did you perform this study?
Dr Neuman: The issue of what kind of anesthesia is best for patients with hip fractures getting hip surgery has been a question for a long time. A lot of people are interested in this question because over 300 000 hip fractures a year happen in the United States, and over a million happen worldwide—it’s a common problem. Furthermore, almost all hip fracture patients gets surgery, which means they need anesthesia. There has been some research suggesting that regional anesthesia may be better for hip fracture surgery, but the question remains unresolved.
news@JAMA: In your study, you used a unique method called instrumental variable analysis to look at this question. Can you briefly describe what this type of analysis is and why you decided to do it?
Dr Neuman: Instrumental variable analysis has been used in economics research in the past, and is now becoming more popular in clinical research. The basic issue that it tries to address is one of selection bias. In this case, the problem in comparing outcomes from regional versus general anesthesia is that patients who get regional anesthesia tend to be sicker than those who get general anesthesia, as it’s thought that there are less risks and side effects with regional anesthesia compared with general. So, if you were just to look at outcomes, regional anesthesia would probably be associated with worse outcomes. Instrumental variable analysis is a way of addressing that kind of selection bias, by using an unrelated “instrument” as a natural source of randomization.
In our study we used the distance from where patients lived in relation to the closest hospital as our instrument. Patients don’t choose to live in a certain neighborhood based on what kind of anesthesia the closest hospital in their neighborhood specializes in. Therefore patients who live closer to hospitals specializing in regional anesthesia should not be sicker than those living closer to hospitals specializing in general anesthesia. Using distance as an instrument for comparison thus allows for a more unbiased comparison of the 2 types of anesthesia.
news@JAMA: Using this analysis, the results of your study showed no difference in mortality associated with the 2 types of anesthesia, but a slightly decreased length of stay associated with the regional anesthesia group. Were these results surprising?
Dr Neuman: The results were a bit surprising to us. The fact that the length of stay was shorter by about half a day was interesting. Even though this may not seem like a big difference, with instrumental variable analyses, the bar for finding any difference at all is really set quite high. So the fact that there was even this amount of difference is encouraging that there may be a real difference in outcomes.
In terms of the mortality, we didn’t find a difference favoring regional anesthesia as other studies in the past have. But, I don’t think the results rule out the possibility that a difference exists, given that the direction of the results still favored regional anesthesia, and again, the bar with instrumental variable analysis is set fairly high. Therefore, this study adds a piece of information to this arena of evidence, but the evidence is far from definitive. It highlights the need for a well-designed randomized trial to further study this issue.
news@JAMA: What would your take home message be at this point for patients who need surgery for hip fracture and have questions about what type of anesthesia they should get?
Dr Neuman: I would say that the bottom line is that we still have incomplete information. The evidence from our study and other studies suggest that regional anesthesia might offer benefits, but it is not strong enough to make a definitive statement. The decision about the type of anesthesia really depends on the hospital, the surgeon, and the culture of practice. In the United States about 80% of patients get general anesthesia and 20% get regional anesthesia for hip fracture surgery. In the United Kingdom, it’s about 50/50.
news@JAMA: So whose choice should it be? What is the ultimate goal of doing a randomized trial to answer this question?
Dr Neuman: It should always be the informed patient’s choice. At the end of the day, even if we prove with a randomized trial that regional anesthesia has lower mortality than general anesthesia, some patients might still say, “I don’t feel comfortable with regional anesthesia, and I will take that mortality risk,” which is perfectly reasonable. What we’re trying to determine is the true magnitude of these risks and benefits, so we can present these numbers accurately to patients when providing them with the opportunity for informed decision making.