Table 2.
PICO 16 (PART 2). Table reports the quality of evidence for each critical outcome in detecting the underlying molecular pathologies (e.g., amyloidosis or tauopathies) in PPA patients
| PICO 16: Detecting the underlying molecular pathologies (e.g., amyloidosis or tauopathies) in PPA patients | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Critical outcomes | No. of papers | Sample size | Gold/reference standard | FDG-PET assessment | Risk of bias | Index test imprecision | Applicability | Effect range (CI) | Effect assessment | Effect inconsistency | Outcome quality |
| Detect AD pathology | 2 | 53 PPA | 1 Pathology 1 Biomarker-based diagnosis |
1 Unclear 1 Semi-quantitative |
Serious | Very serious | Very serious | Study 1. - Accuracy: 91% (Standard error: 7) in lv-PPA, 94% (6) in av-PPA, 85% (10) sv-PPA. - PPV: 94% in lv-PPA, 92% in av-PPA, NA in sv-PPA. - NPV: 81% in lv-PPA, 96% in av-PPA, 92% in sv-PPA. - LR+: 1.91 in lv-PPA, 11.0 in av-PPA, NA in sv-PPA. - LR-: 0.09 in lv-PPA, 0.10 in av-PPA, 1.10 in sv-PPA. Study 2. - Sensitivity: 50% (CI NA). - Specificity: 100% (CI NA) |
Moderate | NA | LOW |
| Detect non-AD pathology | 2 | 31 PPA | 1 Pathology 1 Biomarker-based diagnosis |
1 Unclear 1 Semi-quantitative |
Serious | Very serious | Very serious | Study 1. - Sensitivity: 83% (CI: 36–100%). - Specificity: 100% (CI: 83–100%). - AUC: 90% (CI NA). - PPV: 100% (CI: 48–100%). Study 2. - Sensitivity: 38% (CI NA). - Specificity: 100% (CI NA) |
Moderate | NA | LOW |
| RELATIVE AVAILABILITY OF EVIDENCE: POOR | |||||||||||
Risk of bias: assessment of the study design and other methodological features (e.g., patient selection, clinical diagnostic criteria used)
Index test methods: assessment of index test methodology (e.g., technical details, image analysis methods and statistical analysis)
Applicability: representativeness of the studied population and index test reproducibility in clinical practice (semi-quantitative methods correspond to ‘serious’ indirectness, visual + semi-quantitative methods correspond to ‘not serious’ indirectness, due to partial implementation of quantitation in clinical practice)
Effect: lowest and highest values for each critical outcome; when more values were obtained for the same outcome, the highest was reported
Effect assessment: 51–70% low, 71–80% moderate, 81–100% high
Effect inconsistency: ‘Not serious’ if lowest and highest values difference was 0–20, ‘serious’ 21–40, ‘very serious’ > 40
Outcome quality: summary of evidence as from all columns