Table 2.

Performance of predictive models in the task of discriminating participants between a faster (“0–2 months”) and slower (“≥3 months”) track of mobility recovery

Experiment scenario (post-/pre-operative wearable PGHD availability)	Tendon or ligament repair/reconstruction surgery
	AUROC		AUPRC
	mean (SD)	median (min.-max.)	mean (SD)	median (min.-max.)
(1) No post-op, no pre-op	0.473 (0.108)	0.489 (0.114–0.669)	0.569 (0.059)	0.563 (0.428–0.727)
(2) No post-op, 6 months pre-op	0.497 (0.096)	0.498 (0.266–0.734)	0.596 (0.068)	0.596 (0.453–0.741)
(3) 4 weeks post-op, no pre-op	0.705 (0.089)	0.701 (0.510–0.929)	0.784 (0.076)	0.799 (0.598–0.947)
(4) 4 weeks post-op, 6 months pre-op	0.724 (0.095)	0.734 (0.442–0.942)	0.795 (0.077)	0.800 (0.576–0.960)
(5) 6 months post-op, no pre-op	0.712 (0.084)	0.716 (0.542–0.929)	0.798 (0.067)	0.806 (0.628–0.937)
(6) 6 momths post-op, 6 months pre-op	0.710 (0.096)	0.721 (0.435–0.942)	0.786 (0.080)	0.791 (0.542–0.960)

Results are shown across 6 experiment scenarios in which different data availability was assumed (in each case, age and gender demographic information was used), and across surgery types considered.