Table 2.

Crude and adjusted mean differences in communication score estimated from linear regression models^a

		Unadjusted mean score difference on 0–100 scale (95% CI)	P-value	Adjusted mean score difference on 0–100 scale (95% CI)b	P-value
Length of consultation (per minute)		0.06 (−0.15 to 0.26)	0.592	0.02 (−0.20 to 0.25)	0.841
Patient gender	Male	Reference	0.086	Reference	0.025
	Female	1.94 (−0.27, 4.16)		2.38 (0.30 to 4.47)
Patient age	18–24	Reference		Reference
	25–34	3.60 (−1.92 to, 9.13)		2.51 (−3.30 to 8.31)
	35–44	3.26 (−3.03 to 9.56)		3.14 (−3.93 to 10.22)
	45–54	4.59 (−0.80 to 9.99)		4.50 (−1.51 to 10.51)
	55–64	5.72 (0.38 to 11.07)	0.003	5.98 (−0.15 to 12.11)	<0.001
	65–74	7.01 (1.74 to 12.28)		7.18 (1.41 to 12.96)
	75–84	7.03 (1.66 to 12.40)		7.30 (1.25 to 13.35)
	≥85	8.06 (1.78 to 14.35)		7.44 (0.38 to 14.50)
Patient self-rated health	Excellent	Reference		Reference
	Very good	−3.31 (−5.18 to −1.44)		−4.02 (−5.94 to −2.10)
	Fair	−4.60 (−7.10 to −2.10)	0.001	−5.37 (−7.98 to −2.75)	<0.001
	Poor	−3.12 (−5.53 to −0.72)		−5.03 (−8.02 to −2.03)
	Very poor	−1.83 (−4.83 to 1.17)		−3.49 (−6.79 to −0.19)
Patient ethnicity	White	Reference	0.638	Reference	0.925
	Non-white	0.99 (−3.12 to 5.10)		0.20 (−3.92 to 4.32)
Doctor gender	Male	Reference	0.951	Reference	0.901
	Female	0.08 (−2.52 to 2.68)		−0.16 (−2.75 to 2.42)

^aThis is the average difference in score on a 0–100 scale between groups or attributable to an additional minute of consultation length either with or without adjustment for other factors.

^bMean score adjusted for patient gender, age, self-rated health, ethnicity, and doctor gender.