Table 5.

Association between insight and days spent hospitalised during the observation period (linear regression)

Time period evaluated	B-coefficient for the association with recorded poor insight (95% CIs, p value)
	Unadjusted	Adjusted age and gender	Adjusted age, gender, ethnicity, relationship	Adjusted age, gender, ethnicity, relationship, employment, accommodation	Adjusted age, gender, ethnicity, relationship, employment, accommodation, diagnosis
12 months n=2026	26.42 (22.09 to 30.76) <0.001	26.77 (22.42 to 31.12) <0.001	25.47 (21.05 to 29.88) <0.001	26.33 (21.85 to 30.82) <0.001	25.70 (21.17 to 30.22) <0.001
24 months n=1738	37.47 (29.07 to 45.87) <0.001	38.58 (30.17 to 47.00) <0.001	36.57 (28.04 to 45.10) <0.001	38.28 (29.62 to 46.95) <0.001)	36.78 (28.05 to 45.50) <0.001
36 months n=1461	46.84 (34.24 to 59.44) <0.001	48.55 (35.96 to 61.13) <0.001	46.04 (33.34 to 58.74) <0.001	47.32 (34.46 to 60.18) <0.001	45.33 (32.43 to 58.25) <0.001
48 months n=1185	48.64 (31.06 to 66.22) <0.001	50.77 (33.23 to 68.30) <0.001	46.09 (28.41 to 63.77) <0.001	47.55 (29.67 to 65.44) <0.001	44.77 (26.85 to 62.68) <0.001
60 months n=926	52.91 (29.89 to 75.93) <0.001	53.23 (30.19 to 76.27) <0.001	47.18 (23.99 to 70.37) <0.001	48.40 (24.79 to 72.00) <0.001	44.79 (21.15 to 68.43) <0.001