Clarifying Baseline Cognitive Eligibility and Cross-Scale Follow-Up in a Randomized Trial of Remimazolam Tosylate for Laparoscopic Surgery [Letter]

Dear editor

Liu et al¹ deserve credit for studying young and middle-aged adults undergoing laparoscopic surgery, a group overlooked in perioperative neurocognitive research. Remimazolam was associated with less postoperative memory decline than propofol on postoperative day 1, 36.4% versus 53.2%, and day 3, 23.6% versus 37.8%, while the month-1 cognitive endpoint was similar, 6.4% versus 8.1%. However, there are two issues warrant discussion.

One concern is the mismatch between the eligibility criteria and the baseline table. The Methods exclude patients with preoperative MoCA <26. Yet Table 1 reports baseline Pre-MoCA scores of 26 (25–26) in Group R and 26 (24–27) in Group P. The interquartile range includes values below the exclusion threshold. The Methods also state that one point should be added for participants with ≤12 years of education. That matters. In 665 community residents from Western China, education had a larger effect on MoCA than age, effect size 0.272 versus 0.037, and MoCA-based cognitive impairment rates varied from 25.1% to 52.8% depending on the reference approach.² Ratcliffe et al³ analyzed 18,410 participants and concluded that repeated MoCA interpretation should use demographic adjustment and reliable change methods, not a fixed threshold alone.The present report therefore needs an explicit statement on whether Table 1 shows raw MoCA values, education-adjusted values, or both. Otherwise, readers are left uncertain whether this is a reporting ambiguity or a protocol-execution issue.

Another concern is the 1-month conclusion, because the outcome scale changed. In-hospital cognition was assessed with MoCA, but month-1 follow-up used TICS-m by telephone. Table 2 then lists “postoperative cognitive decline” at 1 month as 7/110 versus 9/111. Without a preoperative TICS-m baseline, however, this cannot strictly be called decline from baseline. It is better understood as a one-time screening result. The manuscript also provides two different TICS-m interpretive rules: page 4 states that >31 is normal and <28 indicates severe impairment, whereas page 5 states that an education-adjusted TICS-m score ≤31 is the optimal cutoff for cognitive impairment. That leaves the 28–31 interval unresolved and creates internal inconsistency. Truong et al⁴ reported good TICS-m reliability in 432 older adults, with a Person Separation Index of 0.86, but this supports careful within-instrument interpretation, not direct interchangeability with MoCA across time. A 2024 perioperative guidance likewise emphasizes baseline neurocognitive assessment when postoperative neurocognitive outcomes are inferred.⁵

Reporting both raw and adjusted baseline MoCA distributions, the number of patients below 26 before and after education correction, and a prespecified TICS-m classification scheme would materially strengthen this otherwise valuable study.