The most comprehensive systematic review and meta-analysis on the subject to date has shown that perioperative goal-directed therapy reduces mortality, morbidity and hospital length of stay in adult patients undergoing major surgery. That did not stop researchers from advising caution in interpreting their results, however, given what they saw as the risk for bias in the randomized controlled trials (RCTs) included in the analysis.
According to Matthew A. Chong, MD, although goal-directed hemodynamic and fluid therapy during major surgery is recommended by societal and national guidelines, previous research has shown conflicting results and inconsistent benefit. “Even so, there are potential benefits that we hypothesize will occur from titrating our therapy in a goal-directed fashion,” said Dr. Chong, an anesthesia resident at the University of Western Ontario, in London, Ontario.
“So with the background of this uncertainty, we had strong justification to proceed with our systematic review and meta-analysis on perioperative goal-directed therapy in surgical patients,” he said. The investigators sought to address whether clinical outcomes differ among different surgical strata, industry-sponsored trials, and the means of measuring and achieving the goal-directed therapy.
Randomized Trials Only
Dr. Chong and his colleagues performed a comprehensive search of the literature through Dec. 31, 2016, to identify RCTs of adults undergoing major surgery. “The patients had to be randomized to either goal-directed therapy or some sort of standard of care, whether that was according to protocol or fully discretionary,” Dr. Chong said. Trauma patients and pregnant women were excluded.
The search yielded 1,945 citations, of which 95 RCTs met inclusion criteria, with 11,599 patients. Eighty-two of the studies used modern monitoring modalities, such as minimally invasive cardiac output monitors or esophageal Doppler probes; the remaining 13 used pulmonary artery catheters. “Overall, we felt that the risk of bias in the evidence base was very high. For example, only four of the 95 studies blinded their participants or personnel. Nevertheless, no evidence of publication bias was found.”
Goal-directed therapy was defined as fluid and/or vasopressor therapy titrated to hemodynamic goals or validated measures of volume responsiveness. The primary outcome was mortality; secondary outcomes included organ-specific morbidity and hospital/ICU length of stay. Two researchers independently extracted study demographics and outcomes, and assessed study quality via the Cochrane Risk of Bias Tool.
As Dr. Chong reported at the 2017 annual meeting of the Canadian Anesthesiologists’ Society (abstract 284789), goal-directed therapy was found to reduce in-study mortality compared with standard care (odds ratio [OR], 0.80; 95% CI, 0.68-0.95; P=0.009; number needed to treat for benefit [NNT]=65). Organ-specific morbidity was also reduced for patients receiving goal-directed therapy, with lower rates of pneumonia (OR, 0.77; 95% CI, 0.63-0.96; NNT=56), acute kidney injury (AKI) or renal dysfunction (OR, 0.72; 95% CI, 0.59-0.88; NNT=33), and wound infection (OR, 0.59; 95% CI, 0.48-0.72; NNT=27).
“Furthermore, hospital length of stay [0.79 days; 95% CI, 0.43-1.15 days] and ICU length of stay [0.58 days; 95% CI, 0.30-0.86 days] were both decreased with goal-directed therapy,” Dr. Chong said. In terms of negative sequelae, rates of myocardial infarction (OR, 0.97; 95% CI, 0.72-1.32), congestive heart failure (OR, 0.99; 95% CI, 0.79-1.25) and exposure to allogeneic blood transfusion (OR, 1.08; 95% CI, 0.81-1.45) were similar between groups. A sensitivity analysis by industry sponsorship revealed a larger magnitudes of effect within industry-sponsored trials for some—but not all—outcomes.
Despite the statistical strength of the findings, the analysis had limitations, as Dr. Chong discussed. “Chiefly, we found that the beneficial effects of goal-directed therapy seem to rely on the use of vasopressors. If you take the fluid-only studies, the results are not significant.”
Other limitations included the variety of monitoring devices used, the prevalence of colloid administration as part of goal-directed therapy, and the fact that 75% of included trials comprised fewer than 100 patients. “If you put that all together, the appraisal for most of the outcomes was low in terms of the evidence,” Dr. Chong said.
Nevertheless, the researchers were encouraged by their results. “We demonstrated that perioperative modern goal-directed therapy reduces mortality, with a pertinent number needed to treat,” he said. “Indeed, for every 1,000 patients we treat, 18 deaths will be prevented, and we’ll have 27 fewer cases of pneumonia, 55 fewer cases of wound infection, 35 fewer cases of AKI and 900 hospital days saved.”
Jonathan Gamble, MD, questioned the effect of various monitoring modalities on the results. “Obviously the technology between the various goal-directed therapy devices was very different,” said Dr. Gamble, assistant professor of anesthesiology, perioperative medicine and pain management at the University of Saskatchewan, in Saskatoon. “Did you worry that you would see a difference in the ability of one device over the other?”
“We didn’t do a subgroup analysis by brand, because we thought that would introduce commercial bias,” Dr. Chong said. “But we did look at the esophageal Doppler as a group and the minimally invasive cardiac monitor as a group, and those results largely held up as the same.”
Gregory M. Hare, MD, PhD, professor of anesthesia at the University of Toronto, was intrigued by the association of the results with vasopressor use. “This is a wonderful study. Can you comment on the dependence on vasopressors and how that would interact with your findings?”
“Our group hypothesized that when you pound people with fluids, once you fill the tank and they’re still not meeting their hemodynamic goals, you’re going to have to do something else,” Dr. Chong said. “Based on our results, it seems that once preload is optimized, the next step may be starting a vasopressor rather than overloading the patient with fluid.”