Table 4.

Distribution of Trust and Related Characteristics among Survey Responders, with Percentages Weighted to Account for Nonresponse

	Intervention Clinics		Control Clinics		χ²P value^*
	n	%^{^†}	n	%^{^†}	χ²P value^*
Total sample	935		653
Trust doctor's judgment in managing medications^{^‡}					.002
Totally disagree	15	1.9	6	0.9
Disagree	60	6.9	27	4.4
Neutral	122	13.3	59	8.4
Agree	308	33.2	218	33.9
Totally agree	426	44.7	341	52.4
Believe doctor trusts patient in managing medications^{^‡}					.002
Totally disagree	10	1.4	9	1.3
Disagree	41	4.4	13	1.9
Neutral	72	7.9	34	5.8
Agree	308	32.8	186	27.8
Totally agree	501	53.4	410	63.3
Sometimes worry doctor will stop prescribing opioids^{^‡}					.007
Totally disagree	223	23.7	196	29.8
Disagree	270	29.1	209	32.3
Neutral	163	17.9	97	16.0
Agree	175	19.1	102	15.1
Totally agree	92	10.2	42	6.7
Believe health care team made sure patient was well informed about potential problems with opioids^{^‡}					.924
Totally disagree	7	0.8	8	1.2
Disagree	15	1.6	11	1.7
Neutral	41	4.3	32	4.3
Agree	318	33.7	209	32.1
Totally agree	553	59.6	391	60.7

↵* P values are based on χ² tests of whether the distributions of responses across the 5 categories for each question differ between participants in the intervention and control clinics after weighting for nonresponse.
↵† While n's represent the raw counts, the computed percentages are based on weighting to account for nonresponse.
↵‡ The proportion with missing data for each of the four questions were 0.4%, 0.3%, 1.2%, and 0.2%, respectively.