Table 4.

Hazard ratios for progression to Arnold–Hilgartner stage ≥ IV

	Univariate				Multivariable
	HR	95% CI		p	HR	95% CI		p
Joint
Elbow	Reference				Reference
Knee	0.40	0.22	0.74	0.003	0.66	0.38	1.17	0.156
Ankle	1.86	1.16	2.98	0.009	0.91	0.56	1.50	0.720
Arnold–Hilgartner stage at initiation of regular prophylaxis
0	Reference				Reference
I	13.12	6.14	28.04	< 0.001	14.95	6.60	33.82	< 0.001
II	13.72	6.99	26.94	< 0.001	11.84	5.48	25.56	< 0.001
III	42.59	25.26	71.80	< 0.001	34.47	18.12	65.60	< 0.001
Age at first visit to our institute
< 6	Reference				Reference
6– < 12	2.80	0.55	14.34	0.216	2.45	0.60	9.91	0.210
12– < 18	9.42	1.95	45.55	0.005	2.63	0.68	10.23	0.162
≥ 18	16.28	3.23	82.18	0.001	2.03	0.43	9.44	0.369
Regular prophylaxis at first visit to our institute
No	Reference				Reference
Yes	0.44	0.28	0.68	< 0.001	0.54	0.32	0.90	0.019
Laterality
Right	Reference				Reference
Left	0.80	0.58	1.11	0.187	0.91	0.58	1.42	0.675

Univariate and multivariable Cox analyses of factors associated with progression to hemophilic arthropathy, defined as reaching Arnold–Hilgartner stage ≥ IV. The multivariable model included joint site, Arnold–Hilgartner stage at initiation of regular prophylaxis, age at first visit to our institute, regular prophylaxis status at first visit to our institute, and laterality. Cox models were estimated using cluster-robust standard errors with participant as the clustering variable

HA hemophilic arthropathy, HR hazard ratio, CI confidence interval