To the Editor:—
In their multicenter, cohort study of patients undergoing major surgery, Johnson et al. 1detected a fivefold increase in early postoperative cognitive dysfunction (POCD) in middle-aged patients undergoing major surgery with general anesthesia when compared with their age-matched relatives. In addition, post hoc univariant and multivariant logistic regression analyses suggested that alcohol cessation, allocation to test centers, and the additional administration of epidural analgesia were factors that significantly contributed to the risk of POCD.
The observation that epidural analgesia contributed to POCD is a surprising finding as it is in contrast to a separate publication in which investigators of the same ISPOCD2 multicenter group did not find a significant difference in POCD between general anesthesia and regional anesthesia in 525 randomized patients at 1 week and 3 months after surgery. 2Similarly, Williams-Russo et al. 3demonstrated in a prospective, randomized, controlled study of 262 older adults that epidural anesthesia does not increase the incidence of POCD at 1 week and 6 months even when continued as postoperative epidural analgesia. It is thus difficult to understand why Johnson et al. 1put such emphasis on an outcome that is derived from a post hoc and underpowered data analysis of a subgroup of 92 nonrandomized patients who probably underwent more invasive and extensive surgery than the rest of the observation group. Despite these limitations, the authors invoke a speculative hypothesis of local anesthetic toxicity to explain the association of epidural analgesia with POCD.
The risk factor analysis in this study is further weakened by the fact that those patients presenting to different centers had a highly significant variation in risk for POCD, which is shown in table 5. 1Study centers could correct for variations “with regard to patient population, data collection variations, and anesthetic and surgical practices.”1We are left to speculate as to the many reasons why patient outcome is dissimilar at the various contributing centers. This confounder could invalidate the entire study by masking the identification of other risk factors and calls for further explanation.
Hence, the authors’ summary statement that epidural analgesia is a risk factor for POCD is at best a tenuous speculation based on a post hoc analysis of a small, nonrandomized subgroup of patients. This questionable conclusion could result in a neuroaxial technique being withheld despite compelling level 1 evidence for improved cardiovascular and pulmonary outcomes and a reduction in perioperative mortality. 4Perhaps a more rational message based on the presented data should be not to caution against the practice of epidural analgesia, but rather to warn patients against having operations at the particular centers that exhibit a significantly higher incidence of POCD.