Caution in Drawing Conclusions on Long-Term Cognitive Recovery

To the Editor:

We read with great interest the article by Brown et al¹ which reported an association between delirium diagnosis and cognitive dysfunction at one month in patients following cardiac surgery. The study is notable for its rigorous delirium assessment and thorough neuropsychologic battery testing at multiple perioperative time points. The conclusion that there is no difference in cognitive change at one year between delirious and non-delirious patients, however, should be interpreted with caution. In addition to being potentially underpowered, the study measured delirium as a dichotomous yes/no despite previous findings that longer durations of delirium have greater impact on long-term cognitive impairment^2,3 and that increased delirium severity is associated with worse outcomes.^4,5 Brown et al’s findings of lower cognitive function scores, starting with baseline assessments, for those who develop postoperative delirium are in agreement with other studies demonstrating the large contribution of preoperative cognitive performance to both acute and long-term brain dysfunction after surgery.^6–8 The question of whether this signifies a continuation of a downward trajectory initiated prior to surgery, is a reflection of greater baseline comorbidities, or has another explanation has yet to be answered.

The description of resilience, or return to cognitive baseline following episode of post-operative delirium, has previously been described by Inouye and colleagues who went on to show an accelerated cognitive decline at out to 36-months.⁹ It would be interesting to see if these patients would have the same biphasic decline within domain-specific cognitive scoring. From a patient perspective, it would also be important to qualify the clinical impact of a lower overall cognitive Z-score and domain-specific cognitive scores. This could help answer whether statistically insignificant changes measured with composite cognitive scoring could potentially miss the clinical impact of delayed executive function, verbal fluency, and other domains that greatly impact patient life or whether statistically significant decline on robust in-depth testing during studies translates into appreciable deficits in the real-world.

The study by Brown et al¹ is important in adding to the growing body of literature attempting to define the inter-relationship between post-operative delirium and cognitive decline, particularly regarding domain-specific analysis. Trials aimed at reducing delirium and postoperative decline that include similar rigorous assessments but also mechanistic testing are needed. We feel that it is essential to note the potential for recovery of cognition after acute brain dysfunction, but that the extent and quality of that recovery remains to be seen.