Table 5.

Competing-risks regression model of prostate cancer diagnosed in Mallorca between 2006 and 2011 before (Model 1) and after (Model 2) multiple imputation (m = 10).

	Model 1 (Original Data Set) n = 2013		Model 2 (Imputed Data Set) n = 2728
Variables	Subhazard Ratio	CI 95%	Subhazard Ratio	CI 95%
Charlson index (ref. 0)
Charlson = 1	0.80	0.59, 1.08	0.84	0.62, 1.13
Charlson = 2	1.83	1.26, 2.66	1.58	1.11, 2.26
Charlson >= 3	1.40	0.91, 2.13	0.94	0.57, 1.56
Risk (ref. medium)
High risk	2.68	1.67, 4.31	2.90	1.78, 4.71
Very high risk	17.53	11.08, 27.72	16.72	10.31, 27.09
Histology (ref. adenoc.)
Unspecified	3.40	2.50, 4.61	4.78	3.55, 6.42
Others	4.01	2.11, 7.59	2.42	1.12, 5.20
Treatment (ref. local)
Systemic	37.49	20.54, 68.42	29.83	16.83, 52.88
Mixed	10.60	5.16, 21.75	10.51	5.40, 20.46
Observation/unspecified	53.18	28.23, 100.19	35.17	19.36, 63.87