To the Editor:
I loved the recent Editorial1 by Martin London on the operating room of the future, not only because he referred to one of my all-time favorite movies, Back to the Future, but also because I, as an aging baby boomer, also remember going to the operating rooms early to get the single noninvasive blood pressure device at our institution. Those of us who trained with a manual cuff and paper charting have seen tremendous improvements in care and safety enabled by technology, but not without controversy. The early noninvasive blood pressure devices were criticized for not correlating with invasive arterial blood pressure.
Surprisingly, even the pulse oximeter was not accepted early on as necessary until it became mandated. The only prospective randomized trial on the use of pulse oximetry in the operating room found no difference in overall clinical outcomes.2 More recently, decision support tools have been developed that incorporate patient comorbidities, current management, and real-time vital signs with the goal of providing a decision support system analogous to the glass cockpit primary flight display in aviation,3 to which Dr. London referred.1 Our study, which evaluated such a tool, found its use was associated with less hypotension, less fluid administration, and better compliance with tidal volume protocols in both historical and parallel controls.3 Dr. London noted these findings but stated that the study “failed to document any improvement in the requisite ‘hard’ clinical outcomes.” By this I assume he was referring to the lack of improvement in postoperative myocardial and renal injury in the parallel control group.
However, in the historical control group, not only were process-of-care measures improved, but also incidences of myocardial injury, stage 1 and 2 acute kidney injuries, 30-day mortality, and hospital length of stay.3 These findings may not be as compelling since it is possible that during the 6-yr period of this study, anesthesia care improved. In the contemporaneous matched cohort analysis, the control group also had 5% greater hospital encounter charges (P < 0.001). Finally, in the sensitivity analysis, when decision support system users were compared to a parallel group who never opened the application, cases performed by decision support users were associated with 1-day reduction in hospital length of stay (P = 0.01). Hospital encounter charges and length of stay may not be as “hard” as clinical outcomes such as reductions in renal or cardiac injury, but they are not unreasonable surrogate markers of the overall clinical care required. Improvement in these outcomes does imply that use of this decision support tool was associated with improved care.
Currently, the tool provides live calculations of known variables that could, in theory, be done manually (e.g., real-time fluid balance calculated using absolute volume intake and outputs [and relative to the patient’s estimated blood volume]; cumulative duration of hypotension; or age-adjusted anesthetic dose incorporating inhaled and infusion agents). Given the likely advances in artificial intelligence in the future, more complex algorithms will be developed that will be impossible to implement manually. I think in the not-too-distant future, administration of general anesthesia without a live decision support tool will seem as archaic as manual blood pressure and charting. We may be approaching an inflection point in improved care and safety similar to that of the 1980s with the adoption of monitoring standards, this time employing live decision support.
It will be back to the future all over again, with Anesthesiology taking the lead in employing technology to improve patient safety.
Dr. Tremper is the founder of and has an equity interest in AlertWatch (Ann Arbor, Michigan), the company that developed this decision support system.