Table 2 Reasons for sourcing HSD
HSD use | Up to 2019 [8]: n (% of 102 trials) | 2019–2022: n (% of 52 trials) |
|---|---|---|
1. Participant recruitment | N/A | 4 (8%) |
2. Collection of baseline data | 1 (1%) | 8 (15%) |
3. Primary outcome (PO)c | ||
 3.1 PO ascertained solely from HSD | 23 (22%) | 6 (12%) |
 3.2 PO partially ascertained from HSD | N/A | 4 (8%) |
4. Secondary outcomes (SOs)c | ||
 4.1 SO(s) ascertained solely from HSD | 35 (34%) | 20 (38%) |
 4.2 SO(s) partially ascertained from HSD | N/A | 1 (2%) |
5. The use of HSD collected post-withdrawal | ||
 5.1 All outcome data can be collected from HSD | 1 (1%) | 9 (17%) |
 5.2 Partial outcome data can be collected from HSD | 17 (17%) | 23 (44%) |
6. For the feasibility study | 12 (12%) | 2 (4%) |
7. Full trial dataa to be accessed from registries | 3 (3%) | 2 (4%) |
8. Long-term follow-up (already budgeted in the current trial) | 4 (4%) | 23 (44%) |
9. Health economic (HE) analysis | ||
 9.1 HE analysis uses HSD only | 4 (4%) | 5 (10%) |
 9.2 HE analysis uses HSD alongside other sources | 7 (7%) | 11 (21%) |
10. To be used if needed | N/A | 6 (12%) |
11. Otherb | N/A | 3 (6%) |