Washington—The total time patients in surgery spend with low blood pressure may be a stronger predictor of serious postoperative complications than the depth of the hypotensive troughs they hit during the procedure, researchers have found.

The researchers, from the Cleveland Clinic in Ohio, have used their analysis to create an algorithm to guide the management of blood pressure in the operating room.

“This research attempted to address a seemingly simple question, yet one many of us might find difficult to answer: What really constitutes hypotension, especially when it comes to long-term outcomes?” said Wolf H. Stapelfeldt, MD, chair of general anesthesiology and vice chair of surgical operations at Cleveland Clinic, who led the study. “From a practical point of view, we all have to draw the line somewhere. We’ve chosen to start out with drawing the line at a mean arterial pressure [MAP] of 60 mm Hg, but what does that mean? When should we start worrying? After two minutes? Five minutes? Fifteen minutes? Quite frankly, we don’t know, and this prompted our current study.”

The Cleveland Clinic team is now in the process of validating the algorithm, including, it hopes, by comparing its data to those from Mayo Clinic and Vanderbilt University Medical Center, in Nashville, Tenn.

“Depending upon the results of these validation steps, we anticipate an evolutionary practice change assuming that not all, but many patients’ outcome might be improved by minimizing hypotensive exposures,” Dr. Stapelfeldt said. “Depending upon the magnitude of the impact on patient outcome, the [decision support system] is principally designed to allow any such algorithm to either be pushed to every provider as a mandatory alert—i.e., a matter of departmental or institutional policy—versus functionality that is made available to individual providers via subscription at their discretion.” Any alert, whether global or individual, may be disregarded by the anesthesia care team using its best judgment, he added.

Dr. Stapelfeldt and his colleagues examined the institution’s registry for adult patients undergoing non-cardiac surgery between Jan. 1, 2009 and Sept. 30, 2010. They analyzed minute-to-minute MAP readings from more than 35,000 patient records for periods of time spent below hypotensive thresholds ranging from 75 to 45 mm Hg. Patients were considered to have dropped below a certain threshold if they spent at least one minute below that level. The investigators used logistic regression to determine the relationship between cumulative time spent below the various thresholds and 30-day mortality.

As Dr. Stapelfeldt reported at the 2012 annual meeting of the American Society of Anesthesiologists (abstract 074), spending any amount of time below a certain threshold was common, decreased in incidence with progressively lower hypotensive thresholds and was associated with increased 30-day mortality (Figure 1).

Figure 1. Incidence, distribution of cumulative time spent and 30-day mortality for patients dropping below blood pressure thresholds.

Several factors independently affected patient outcomes after surgery. These included patient age, Charlson comorbidity index and cumulative blood loss. “There was a 20% increase in mortality for each additional comorbidity,” Dr. Stapelfeldt said. “And every additional 20 years of age more than doubled a patient’s risk for dying within 30 days of surgery.”

But the most striking finding involved the cumulative amount of time spent below various blood pressure thresholds, he continued. “We found that as patients started spending even just a few minutes below an MAP of 55, there was a sharp increase in the odds ratio for mortality [Figure 2]. More importantly, similar increases in mortality were seen whenever patients exceeded other, longer exposure limits for time spent below any of the less severely hypotensive MAP thresholds.”

Figure 2. Percentage increase in the odds ratio for 30-day mortality depending on the duration of hypotension below certain thresholds. (figures courtesy of Wolf H. Stapelfeldt, MD.)

Indeed, as each additional exposure limit was exceeded, there was an incremental 4% to 7% increase in the risk for 30-day mortality. “What’s most disconcerting is that many of our patients routinely exceeded a large number of these exposure limits by spending substantial amounts of time below their respective MAP thresholds, on the order of half an hour or longer in many instances.”

These results suggest that long-term outcome may not only be affected by periods of severe hypotension, but also by extended periods spent at seemingly adequate intraoperative blood pressure levels. “Unlike conventional vital sign monitors—which alert only to blood pressure dropping below a certain threshold level—decision support systems can alert to a certain level of risk attributable to significant cumulative hypotensive exposure, something apparently very critical yet difficult for humans to keep track of, unlike the detection of any sudden changes in blood pressure.

“This information would be most useful if it is provided in real time, which decision support is now able to do,” Dr. Stapelfeldt said. “The idea is that once the anesthesia care team receives any first alerts to excessive hypotensive exposure, it might be able to adjust the anesthetic in such a way that the patient’s mean arterial pressure trends higher. One can choose how aggressive one wants to be with minimizing hypotension according to what level of attributable risk one might want to attempt to mitigate.”

Christian Apfel, MD, PhD, adjunct associate professor of anesthesia, perioperative care, epidemiology and biostatistics at the University of California, San Francisco, called the research fascinating and rigorous. “We still have procedures where surgeons are asking for ‘controlled’ hypotension,” Dr. Apfel said. “And that is actually something I’ve always been uncomfortable with.

“Dr. John Drummond from the University of California, San Diego is always warning against this kind of controlled hypotension,” Dr. Apfel continued. “And until now, we only had anecdotal evidence. This decision support system might help us in those kinds of situations.”