Skip to main content

Table 5 Validity measures of the prediction algorithm

From: Identification of patients with suboptimal results after hip arthroplasty: development of a preliminary prediction algorithm

Measure Estimates in training sample Estimates with 1,000 bootstrap resamples
Sensitivity % (95 % CI) 75.0 (59.8.4–85.8) 75.0 (60.0–88.0a)
Specificity % (95 % CI) 77.8 (71.9–82.7) 77.8 (72.2–82.9a)
Positive predictive value % (95 % CI) 37.5 (27.7–48.5) 37.2 (27.2–47.2a)
Negative predictive value % (95 % CI) 94.6 (90.3–97.0) 94.7 (91.2 to 97.8a)
Positive likelihood ratio (95 % CI) 3.38 (2.49–4.57) 3.38 (2.50 to 4.63a)
Negative likelihood ratio (95 % CI) 0.32 (0.19–0.55) 0.32 (0.15 to 0.52a)
  1. • a95 % asymptotic confidence intervals
  2. • Sensitivity: number of participants classified at risk both by the PA and the postoperative WOMAC score and joint perception divided by all participants classified at risk by the postoperative WOMAC score and the joint perception (actual outcome)
  3. • Specificity: number of participants classified not at risk by the PA and the postoperative WOMAC score and joint perception divided by all participants classified not at risk by the postoperative WOMAC score and joint perception (actual outcome)
  4. • Positive predictive value: number of participants classified at risk by the PA and the post-operative WOMAC score and joint perception divided by all participants classified at risk by the PA (predicted outcome)
  5. • Negative predictive value: number of participants classified not at risk by the PA and the postoperative WOMAC score and joint perception divided by all participants classified not at risk by the PA (predicted outcome)
  6. • Positive likelihood ratio: sensitivity/ (1-specificity)
  7. • Negative likelihood ratio: (1-sensitivity)/specificity
\