The first study looked at anesthetic depth during elective, non-cardiac major surgery, and mortality rates one-year post surgery. Anesthetic depth was measured using the Aspect BIS monitor, but the anesthesia provider was blinded to the BIS value. In the prospective study of 907 patients, one-year mortality tended to be higher in patients over 40 years old maintained at BIS levels below 45. Logistic regression modeling confirmed that age, comorbidity, and longer times at BIS values less than 45 (i.e. deeper anesthetic effect) were independently associated with higher mortality rates. These results were duplicated in a subsequent study of 5,057 non-cardiac surgery patients in Sweden undertaken by Drs. Rolf Sandin and Claes Lennmarken and colleagues.

Dr. Monk's second study looked at Medicare data for 1.6 million hospital stays from 4,537 institutions and found that higher institutional BIS utilization rates were associated with decreased one-year post-surgical mortality. BIS monitors and sensors were provided by Aspect Medical Systems for the first study, and Aspect collaborated with the University of Florida on the data analysis for both studies.

What led you to study anesthetic depth and its possible correlation with postoperative mortality?

Dr. Monk: Our interest arose out of work we were doing to see if anesthetic depth might be associated with cognitive decline. When you do a longitudinal study, you naturally collect a lot of data on patient outcomes, with death being one possibility. The research on anesthetic depth and mortality was a natural progression.

What, if anything, surprised you in the first study?

Dr. Monk: Two things. First, when we looked at the BIS scores, we found that older, high-risk patients were more deeply anesthetized than they needed to be. We know that older, high-risk patients generally require less anesthesia than younger, healthy patients. One would assume that older patients with co-morbidity would be more lightly anesthetized, but in fact, this study indicates that they are more deeply anesthetized than we think. In part, this may be the result of efforts by anesthesiologists to maintain perioperative hemodynamic stability even though blood pressure and heart rate are not direct indicators of anesthetic depth. The second striking thing about the study was the suggestion that depth of anesthesia may have significant long term consequences. Typically, we in anesthesia are concerned about the perioperative effects of anesthesia. Both of our studies indicate that we need to look very carefully at the possible longer term effects of excessive anesthetic depth which may, indeed, be quite serious.

Why did you look at Medicare data in the second study and what did the second study add to your understanding of this issue?

Dr. Monk: Data of the kind we were looking for is hard to obtain for many reasons, one being that without a brain monitor there is no objective way to measure depth of anesthesia, and brain monitoring is not yet the standard of care. We thought government databases, which have statistics on millions of patients, might be successfully mined for information relevant to this issue. The methodology we used is too complex to describe in detail here because we had to factor in many variables about the patients, the institutions and their patient management. We found that institutions that routinely use the BIS Monitor have lower one-year mortality rates even after adjusting for differences in case-mix and patient populations. It seems reasonable to assume that higher rates of BIS monitoring result in fewer patients being maintained at very deep anesthetic levels and thus these results are consistent with our initial finding of a correlation between anesthetic depth and one-year mortality. This finding strongly suggests that large-scale, randomized trials are needed to prove or disprove our hypothesis.

So you are not yet able to say that there is a direct correlation between excessive anesthetic depth and mortality rates one year after surgery?

Dr. Monk: No, although these studies certainly are suggestive of that. It is possible, for example, that a lower BIS value is simply a marker for patients who may be more susceptible, within one year of surgery, to die of causes unrelated to their anesthesia, or that greater sensitivity to anesthesia is related to the co-morbidities in many of the patients studied. However, it's also possible that there is an unknown mechanism by which anesthesia exerts deleterious effects on certain patient populations post-operatively and contributes to increased mortality. For example, it may be that anesthetics affect immune response. We don't know. I should point out that a study of 5,000 patients in Sweden similar to our first study yielded nearly identical results. In our first study and in the Swedish study patients who were carried deep and for longer periods of time had a higher one-year death rate, regardless of the ultimate cause of death.

"In our first study and in the Swedish study,
patients who were carried deep and for longer
periods of time had a higher one-year death rate,
regardless of ultimate cause of death."
- Dr. Terri G. Monk

You mentioned an "unknown mechanism" that might be a causal link between deep anesthesia and mortality. Any idea what that mechanism might be?

Dr. Monk: The study doesn't suggest a mechanism, but it is possible that deep anesthesia affects the early recovery period which in turn increases post-operative morbidity and mortality. It's possible, for example, that older patients recovering from deep anesthesia may be more susceptible to respiratory infections that complicate their recovery. Extremely deep anesthetic levels may also contribute to the delirium we often see in older patients post-surgery and this delirium may be connected in some way to more adverse outcomes one year later. Again, our studies don't conclusively demonstrate a causal link between deep anesthesia and mortality, but the results should certainly raise serious questions about a possible relationship between the two. After all, we have no control over the other factors that can contribute to death post-surgery such as co-morbidity, age, gender, medical history and so on. But the anesthesiologist can control depth of anesthesia. If excessive anesthetic depth is a contributing factor to post-surgical mortality, we should avoid it. Even if there is no correlation, we should avoid excessively deep levels of anesthesia simply because it may contribute to prolonged recovery times.

Have the study results changed the way you practice?

Dr. Monk: Absolutely. The study changed the way I treat older patients in the OR. For example, in the past, I routinely used anesthetic agents as a means to maintain intra-operative hemodynamic stability. Many elderly patients have hypertension and may be given excessive anesthesia to keep their blood pressure under control during surgery. I am now more likely to use beta-blockers for hemodynamic control, and to use the BIS Monitor, not hemodynamics, to assess anesthetic depth. This would make sense even if there were no correlation between over-anesthetizing and post-surgical mortality because-hemodynamics are at best an inferential method of assessing adequacy of anesthesia. The BIS Monitor gives me an independent, direct measure of anesthesia. But, if, as this study suggests, there may be a correlation between excessive anesthesia and post-surgical mortality, we should avoid deep levels of anesthesia whenever possible. Our studies and the Swedish study are very provocative in this regard and should alert the anesthesia community to the potentially adverse long term effects of excessively deep anesthesia in older, sicker patients.

Home | Signup | Feedback or Comments
© Copyright 2003. Aspect is a trademark of Aspect Medical Systems Inc and is registered in the USA. BIS, the BIS logo and Bispectral Index are trademarks of Aspect Medical Systems, Inc., and are registered in the USA, EU and other countries.

080-0104   8.13