Table 4.
Item score distributions in the 13-item Methodologic Quality tool of included subtyping studies using a data-driven or hypothesis-driven approach
| Item | Score rating | Hypothesis-driven | Data-driven | p |
| (n = 8) | (n = 25) | |||
| Disease stages/duration (study population) | 0 = mixture of stages/disease duration at baseline or not reported | 5 (62.5) | 18 (72) | 0.67 |
| 1 = homogeneous disease stage/duration | 3 (37.5) | 7 (28) | ||
| Study setting (representativeness) | 0 = single-center or not reported | 7 (87.5) | 10 (40) | 0.04 |
| 1 = multi-center | 1 (12.5) | 15 (60) | ||
| Recruitment source (generalizability) | 0 = clinic-based or not reported | 8 (100) | 25 (100) | 1 |
| 1 = community or population-based | 0 (0) | 0 (0) | ||
| Diagnostic methods | 0 = not described or 1 or 2 not applicable | 2 (25) | 1 (4) | 0.04 |
| 1 = Use of formal diagnostic criteria or diagnosis by an expert neurologist | 6 (75) | 24 (96) | ||
| 2 = postmortem diagnosis | 0 (0) | 0 (0) | ||
| Sampling method | 0 = convenience or not reported | 5 (62.5) | 21 (84) | 0.32 |
| 1 = consecutive or random | 3 (37.5) | 4 (16) | ||
| Comprehensiveness of data used for subtyping (subtype definition) | 0 = single clinical or biomarker domain | 8 (100) | 9 (36) | 0.003 |
| 1 = > 1 clinical domains or biomarkers | 0 (0) | 16 (64) | ||
| Variables compared between subtypes (post hoc) | 0 = not done | 0 (0) | 2 (8) | 1 |
| 1 = single clinical domain or biomarker | 0 (0) | 0 (0) | ||
| 2 = > 1 clinical domains or biomarkers | 8 (100) | 23 (92) | ||
| Statistical methods used for subtyping | 0 = low quality | 4 (50) | 4 (16) | 0.11 |
| 1 = intermediate quality | 3 (37.5) | 10 (40) | ||
| 2 = high quality | 1 (12.5) | 11 (44) | ||
| Longitudinal follow-up | 0 = none (cross-sectional) or longitudinal < 1 year | 6 (75) | 17 (68) | 0.68 |
| 1 = short-term (1–3 years) OR longer-term but < 3 time-points | 2 (25) | 4 (16) | ||
| 2 = longer-term (> 3 year) AND > = 3 time-points | 0 (0) | 4 (16) | ||
| Completeness of follow-up | 0 = cross-sectional or ≤50% complete or not reported | 7 (87.5) | 20 (80) | 1 |
| 1 = 50–75% complete | 1 (12.5) | 2 (8) | ||
| 2 = > 75 % complete | 0 (0) | 3 (12) | ||
| Subtype stability | 0 = not assessed | 8 (100) | 23 (92) | 1 |
| 1 = assessed | 0 (0) | 2 (8) | ||
| Algorithm for classifying individual patients | 0 = not provided | 0 (0) | 24 (96) | < 0.001 |
| 1 = provided | 8 (100) | 1 (4) | ||
| Validation (internal or external) | 0 = not assessed | 8 (100) | 19 (76) | 0.57 |
| 1 = use of a test set from the same population | 0 (0) | 4 (16) | ||
| 2 = validation in an external population | 0 (0) | 2 (8) |