Table 1.
Cognitive impairment/decline definitions utilized for meta-analysis.
| Method | Definition utilized in meta-analysis | Sub-definition | Method reference |
|---|---|---|---|
| 1 SD | ≥1 SD decline in a participant's postoperative test score compared to their preoperative test score, on at least 20% of the tests. The SD is either calculated based on published or sample (pre-operative) norms Method reflects individual change relative to sample or population data (utilized to calculate the SD). |
Newman et al. [15] | |
| 20:20 | ≥20% decline in a participant's post-operative test score compared to their preoperative test score, on at least 20% of the tests conducted. Method reflects individual change relative to self. |
Stump [16] | |
| RCI⁎ | RCI decline of ≥1.64 in ≥20% of tests, or global decline of ≥1.64 in RCI composite score. All versions of RCI calculation methods were included in the analysis. Method reflects individual change relative to sample or population data taking practice effects into account. |
Change in participant's preoperative to postoperative test score, divided by the standard error of the difference between the two test scores (SEdifference = √ 2((SDbaseline control√(1 − rxx))2, where rxx is the test–retest reliability of the measure. | Jacobson and Truax [17] |
| Average change in preoperative to postoperative test scores of the control group is subtracted from within-participant change in preoperative to postoperative test scores. This value is then divided by the standard deviation of the control group change. | Rasmussen et al. [18] | ||
| Average change in preoperative to postoperative test scores of the control group is subtracted from within-participant change in preoperative to postoperative test scores. This value is then divided by the standard error of the difference between the test scores (SEdifference = √2((SDbaseline control√(1 − rxx))2, where rxx is the test–retest reliability of the measure. | Chelune et al. [19] | ||
| Average change in preoperative to postoperative test scores of the control group is subtracted from within-participant change in preoperative to postoperative test scores. This value is then divided by the within-participant standard deviation of the matched control group. | Mollica et al. [20] | ||
| Cut-off | Use population norms or a threshold of decline (e.g. decrease by 2 points) to define cognitive impairment in particular tests (e.g., MMSE) Method reflects the individual's performance on a test. |
Chakravarthy et al. [26], Goto et al. [27] |
Where: 1 SD — 1 standard deviation method, 20:20–20% decline method, RCI — reliable change index method, SD — standard deviation.
⁎
Both 90% (RCI = 1.64) and 95% (RCI = 1.96) intervals are used across studies to define cognitive impairment. Participants that decline by >1.64 or >1.96 (depending on the definition used) on ≥20% of tests or a composite RCI score that aggregates all test scores.