Table 2. Summary of interrater reliability scores of skin-NTD clinical diagnoses, Liberia*.
| Disease | Total survey cases assessed | Total QC clinical diagnoses | Agreement | Verifier-only diagnoses | QC-only diagnosis | Agreement, % | Alternative diagnoses |
|---|---|---|---|---|---|---|---|
| Suspected Buruli ulcer | 15 | 11 | 10 | 5 | 1 | 62.5 | Traumatic ulcer, tropical ulcer |
| Leprosy | 12 | 7 | 7 | 5 | 0 | 58.3 | Vitiligo, tinea corporis |
| LF lymphedema | 25 | 27 | 24 | 1 | 3 | 85.7 | Non-LF edema |
| LF hydrocele | 17 | 14 | 8 | 9 | 6 | 34.8 | Hernia, non-LF hydrocele |
| Other skin disease |
105 |
115 |
95 |
10 |
20 |
76.0 |
None |
| Combined skin NTDs | 69 | 59 | 49 | 20 | 10 | 62.0† | None |
*Overall agreement; 82.8%; Cohen κ, all outcomes: 0.69 (95% CI 0.59–0.79). LF, lymphatic filariasis; NTD, neglected tropical disease; QC, quality control. †Cohen κ: 0.51 (0.39‒0.64).