Skip to main content
. 2020 Apr 10;35(4):970–978. doi: 10.3904/kjim.2019.093

Table 4.

Validation of diagnostic predictive scoring model

Variable No. of ITP patients
Internal validationcohort (n = 166) External validation cohort (n = 171)
Total score
 0 0/28 (0) 0/40 (0)
 1 4/31 (12.9) 0/50 (0)
 2 5/40 (12.5) 3/42 (7.1)
 3 26/37 (70.3) 16/25 (64)
 4 23/23 (100) 7/11 (63.6)
 5 6/7 (85.7) 3/3 (100)
Diagnosis
 Non-ITP (0–2) 9/99 (9) 3/132 (2.3)
 ITP (3–5) 55/67 (82.1) 26/39 (66.7)

Values are presented as number (%). Data are total number of patients with ITP/total number of patients with that score. Scoring model: 1 point to high hemoglobin (> 12g/dL), low ferritin (≤ 177 ng/mL), normal lactate dehydrogenase (≤ upper limit of normal) and immature platelet fraction (IPF) ≥ 7 and < 10, 2 point to IPF ≥ 10.

ITP, immune thrombocytopenic purpura.