Statistical analysis of arthroplasty register data

Figures & data

Cumulative revision rate of two fictitious prosthesis types characterized by non-proportional hazards.

Extract from annual report of the Swedish Knee Arthroplasty Register (SKAR) showing uncertainty indiction for ranking. The extract is published with approval from SKAR

Code	Unit	No. TKA	No. revised	RR	95% CI	Rank	95% CI
21001	Linköping	482	2	0.33	0.15–0.73	3	1–27
21014	Motala	1683	8	0.38	0.21–0.66	4	1–22
10484	Sabbatsbergs närsjh	704	5	0.38	0.20–0.72	4	1–26
52012	Alingsås	851	5	0.45	0.24–0.87	8	1–39
56010	Västerås	457	3	0.49	0.24–1.03	10	1–50
62011	Örnsköldsvik	913	8	0.49	0.28–0.87	10	1–38
22010	Jönköping	849	8	0.51	0.29–0.90	11	2–40
64011	Lycksele	352	2	0.52	0.23–1.14	12	1–58
22012	Värnamo	804	7	0.55	0.30–1.00	13	2–48
50010	östra sjukhuset	884	9	0.56	0.33–0.97	15	2–45
53011	Lidköping	730	7	0.58	0.32–1.04	15	2–52
65014	Kalix	164	1	0.58	0.24–1.39	15	1–69
12010	Enköping	1006	9	0.59	0.35–1.02	16	3–50
53010	Falköping	770	8	0.60	0.34–1.05	17	3–51
42011	Varberg	1137	13	0.60	0.37–0.98	17	4–47
50001	Sahlgrenska	447	5	0.62	0.32–1.19	18	3–59

RR: Relative risk of revision for units