Sometimes, The Less You See, The More There Is
Adaora M. Chima, MBBS, MPH
Statements that contain the phrases “no difference” or “insignificant p-value” are dreaded ones for a researcher testing a hypothesis. To many, they are perceived to demote the value of the research study. On Saturday, April 15, the IARS 2023 Annual Meeting session, “When It’s Right to Be Wrong: Surprising and Hypothesis-Disproving Studies from the Recent Perioperative Medicine Literature” disproved this notion with multiple examples of recent studies with “insignificant” findings that counter widely held understandings of factors affecting postoperative neurocognitive effects.
Miles Berger, MD, PhD, Associate Professor of Anesthesiology at Duke University Medical Center, moderated this thought-provoking session.
The first presentation was made by Elizabeth Whitlock, MD, MS, an Assistant Professor and epidemiological researcher at the University of California, San Francisco. She shared surprising results from her analysis of data from the longitudinal Health and Retirement database, comparing cognitive decline in patients who underwent percutaneous coronary intervention (PCI) vs coronary artery bypass grafting (CABG). In the Health and Retirement study, baseline memory score and dementia probability are assessed. Patients undergo PCI or CABG when indicated, and cognitive evaluation continues thereafter.
The working hypothesis was that patients who underwent CABG would experience worse postoperative cognitive decline compared to PCI subjects, attributing the acceleration to exposure to general anesthesia, cardiopulmonary bypass exposure, and subsequent hospitalization with increased risk for delirium. This was not the case as there was no difference in postoperative cognitive decline between both groups and no difference in 5-year dementia probability either. On further stratification of the CABG group, off-pump cases demonstrated worse cognitive decline than on-pump cases. This was surprising considering that off-pump was designed to minimize exposure to general anesthesia and bypass related risks. Dr. Whitlock underscored the importance of baseline cognitive assessments, as catastrophic postoperative cognitive decline in the postoperative period may seem acceptable if the baseline is unknown.
Michael Devinney, MD, PhD, Assistant Professor at Duke University, shared findings from the SANDMAN study in his presentation titled, “Sleep Apnea and Perioperative Neurocognitive Disorders (POCD).” Sleep apnea is known to be associated with cognitive impairment in older adults, Alzheimer’s disease and worse cognition in Alzheimer’s. Thus, a working hypothesis in his study was that increased sleep apnea severity would result in decreased global cognition 6 weeks after surgery.
Patients older than 60 years of age scheduled for non-cardiac surgery underwent baseline cognitive testing and home sleep apnea tests. They were examined postoperatively for delirium on Days 1-5 and at the 6-week mark. Duration of surgery was associated with increased occurrence of delirium while baseline cognitive index was associated with delirium severity. Although age and baseline cognitive index were associated with POCD, they found no relationship between global cognitive decline and sleep apnea. Thus, untreated sleep apnea is not a risk factor for POCD or delirium.
There were limitations to the study: powering the study for disorder severity, not incidence; a study population with mostly normal cognition; and limited knowledge of sleep apnea burden or duration of the subjects. In future steps, Dr. Devinney plans to focus on the complex study of sleep, and is interested in examining sleep deficiencies, the myriad of factors that contribute to sleep disruption and the potential impact of disrupted postop sleep on neurocognitive disorders.
The third speaker, Patrick McCormick, MD, Associate Attending and Vice Chair for Informatics at Memorial Sloan Kettering Cancer Center, shared results from the TORIE study, examining the effect of age on latency to emergence from general anesthesia.
The hypothesis of the study was that older patients would recover cognitive function more slowly than younger patients after anesthesia. Patients were categorized into age groups from the 5th to 8th decades of life.
The study protocol involved rigorous cognitive tests and neuroimaging, monitoring with tests and standard anesthesia monitoring of healthy volunteers placed under general anesthesia. There was no association between age group and cognitive recovery. Participants’ performance on the cognitive tests improved over time, which was attributed to learning effects following repeated test exposure. Relevant biomarkers for neuronal damage were not significant in the study volunteers.
Older surgical patients have demonstrated slower neurocognitive recovery than younger patients, but volunteer subjects exposed to only anesthesia in this study did not show a similar effect. This shifts the spotlight to surgery and its involvement with neuroaxonal damage, possibly from inflammation.
Dr. McCormick noted that it is important to consider the possible difference that the perioperative experience makes to surgical patients compared to study volunteers who were not exposed to postoperative pain medications or hospital stays.
A lively Q&A session after the presentations generated conversations about the undetermined role of genetics in POCD; cognitive prehabilitation, which has not been found to improve POCD; and the arduous and time-consuming nature of cognitive tests for the elderly.
A combined description of the ideal neurocognitive by the speakers included routine cognitive testing almong the older population, which is easy, comprehensive, and efficient. There was a consensus amongst the speakers that although a lot of focus has been placed on postoperative cognitive decline, postoperative neurocognitive improvement is a byproduct of surgical interventions that improve quality of life and this should be examined as well.
Dr. Berger stressed the need for multicenter studies to generate robust and representative cognitive test data, and encouraged interested individuals to contact him.
This session highlights the importance of supposedly insignificant results, which can possess the power to dethrone even long held clinical convictions.