Summary of findings for the main comparison. Cognitive training compared to control immediately post intervention for people with mild to moderate dementia.
| Cognitive training compared to control immediately post intervention for people with mild to moderate dementia | ||||||
| Patient or population: people with mild to moderate dementia Setting: Community dwelling or in residential care Intervention: cognitive training Comparison: control immediately post intervention | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect (95% CI) | No. of participants (studies) | Certainty of the evidence (GRADE) | Comments | |
| Risk with control immediately post intervention | Risk with cognitive training | |||||
| Change in a global measure of cognition (composite) | Mean change in a global measure of cognition (composite) was 0 | SMD 0.42 higher (0.23 higher to 0.62 higher) | ‐ | 1389 (27 RCTs) | ⊕⊕⊕⊝ MODERATEa | Cognitive training probably has a modest effect on global cognition (based on a composite score) |
| Change in a global measure of cognition | Mean change in a global measure of cognition was 0 | SMD 0.65 higher (0.26 higher to 1.05 higher) | ‐ | 1288 (20 RCTs) | ⊕⊕⊝⊝ LOWb | Cognitive training may have a moderate effect on performance in global cognition (based on a screening measure). |
| Change in delayed memory | Mean change in delayed memory was 0 | SMD 0.81 higher (0.29 higher to 1.32 higher) | ‐ | 543 (11 RCTs) | ⊕⊝⊝⊝ VERY LOWb,c | We are unable to determine whether there is any effect on delayed memory due to the very low quality of evidence |
| Change in participants' mood | Mean change in participants' mood was 0 | SMD 0.72 higher (0.1 lower to 1.54 higher) | ‐ | 577 (8 RCTs) | ⊕⊝⊝⊝ VERY LOWb,d | We are unable to determine whether there is any effect on participants' mood due to the very low quality of evidence |
| Change in capacity for activities of daily living | Mean change in capacity for activities of daily living was 0 SD | SMD 0.12 SD higher (0.11 lower to 0.35 higher) | ‐ | 687 (10 RCTs) | ⊕⊕⊝⊝ LOWd | Cognitive training may not have an effect on capacity for activities of daily living |
| Participant burden (retention rates) | Study population | OR 0.73 (0.37 to 1.43) | 1282 (17 RCTs) | ⊕⊕⊝⊝ LOWe | Cognitive training may not be associated with an increase in participant burden as reflected in retention rates | |
| 908 per 1000 | 878 per 1000 (784 to 934) | |||||
| Change in mood and well‐being (caregiver) | Mean change in mood and well‐being (caregiver) was 0 | SMD 0.98 higher (0.27 higher to 1.68 higher) | ‐ | 36 (1 RCT) | ⊕⊕⊕⊝ MODERATEf,g | Cognitive training probably has a large effect on mood and well‐being in the caregiver |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). CI: confidence interval; OR: odds ratio; RCT: randomised controlled trial; SMD: standardised mean difference. | ||||||
| GRADE Working Group grades of evidence. High certainty: we are very confident that the true effect lies close to that of the estimate of the effect. Moderate certainty: we are moderately confident in the effect estimate: the true effect is likely to be close to the estimate of the effect, but there is a possibility that it is substantially different. Low certainty: our confidence in the effect estimate is limited: the true effect may be substantially different from the estimate of the effect. Very low certainty: we have very little confidence in the effect estimate: the true effect is likely to be substantially different from the estimate of effect. | ||||||
aInconsistency: downgraded 1 point for serious concerns regarding heterogeneity in effect size, which is moderate and statistically significant. Heterogeneity does not seem to be well explained by investigated effect moderators.
bInconsistency: downgraded 2 points for very serious concerns regarding heterogeneity in effect size, which is relatively large and statistically significant. Heterogeneity does not seem to be well explained by investigated effect moderators.
cPublication bias: downgraded 1 point for strongly suspected publication bias based on visual inspection of the funnel plot, raising the possibility that small negative studies may remain unpublished.
dImprecision: downgraded 1 point for serious concerns related to imprecision because the confidence interval crosses the no treatment threshold.
eImprecision: downgraded 2 points for very serious concerns related to imprecision because the confidence interval includes positive effect, negligible effect, and effect in the direction of the control group.
fRisk of bias: outcome estimation is based on a single study with several limitations related to unclear or high risk of bias in several domains.
gImprecision: downgraded 1 point for serious concerns related to imprecision because the analysis is based on fewer than 400 participants; however the confidence interval does not cross the no effect threshold.