Physician Counseling for Colorectal Cancer Screening: Impact on Patient Attitudes, Beliefs, and Behavior

Article Figures & Data

Table 1.

Patient Characteristics (n = 50)

Characteristic
Age, years (mean [SD])	57.3 (6.2)
Female	66 (33)
Race/ethnicity
Non-Hispanic white	29 (58)
African American	13 (26)
Hispanic	5 (10)
Other	3 (6)
Education
High school, GED, or less	8 (16)
Some college (no degree)	24 (48)
Graduated from college	18 (36)
Household income ($)
<20,000	12 (24)
20,00–49,999	7 (14)
50,000–84,999	16 (32)
≥85,000	12 (24)
Missing	3 (6)
Self-reported health status
Poor	1 (2)
Fair	14 (28)
Good	20 (39)
Very good	11 (22)
Excellent	4 (8)
Current smoker	18 (36)
Baseline stage of readiness with regard to colonoscopy
Precontemplation	10 (20)
Contemplation	28 (56)
Planning	12 (24)

Table 2.

Characteristics of Conversations (n = 38) Regarding Colorectal Cancer Screening

Characteristic
Who initiated CRC screening conversation
Provider	27 (71)
Patient	11 (29)
Discussion of specific tests
General discussion but no specific test mentioned	2 (5)
Specific tests discussed	36 (95)
Colonoscopy	36 (95)
Flexible sigmoidoscopy	15 (39)
Fecal occult blood testing	12 (32)
CT colonography	2 (5)
Discussion of behavioral constructs
Benefits of screening	20 (53)
Common patient-level barriers	11 (29)
Patient's risk and susceptibility to CRC	14 (37)
Patient self-efficacy to complete screening
Provider explored patient's confidence	0
Provider referred to prior successful behaviors	0
Provider reframed prior failures	0
Provider specified next steps	7 (18)
Providers expressed confidence in patient	0
One or more behavioral constructs	26 (68)
Two or more behavioral constructs	17 (45)

Table 3.

Associations Between Conversation Content and Measures of Behavioral Constructs to Colorectal Cancer Screening After the Visit (n = 50)

Behavioral Construct (Sample SD of Scale Measure After Visit)	β (95% CI)^* (Adjusted Mean Difference)
Perceived benefits (SD, 0.86)
No discussion of CRC screening	Ref
Any discussion	−0.02 (−0.35 to 0.31)
No discussion of CRC screening	Ref
Screening discussed but not benefits	−0.04 (−0.40 to 0.32)
Benefits discussed	0.00 (−0.39 to 0.38)
Perceived barriers^† (SD, 0.63)
No discussion of CRC screening	Ref
Any discussion	0.00 (−0.15 to 0.15)
No discussion of CRC screening	Ref
Screening discussed but not barriers	0.04 (−0.12 to 0.21)
Barriers discussed	−0.11 (−0.32 to 0.10)
Perceived risk/susceptibility (SD, 0.94)
No discussion of CRC screening	Ref
Any discussion	0.39 (0.09–0.68)
No discussion of CRC screening	Ref
Screening discussed but not risk/susceptibility	0.37 (0.04–0.70)
Risk/susceptibility discussed	0.42 (0.13–0.72)
Self-efficacy to undergo CRC screening (SD, 0.70)
No discussion of CRC screening	Ref
Any discussion	−0.06 (−0.42 to 0.29)
No discussion of CRC screening	Ref
Screening discussed but not next steps	−0.06 (−0.42 to 0.30)
Next steps specified	−0.08 (−0.52 to 0.37)
Intention to undergo CRC screening (SD, 0.78)
No discussion of CRC screening	Ref
Any discussion	0.42 (0.11–0.73)
No discussion of CRC screening	Ref
Screening discussed but no constructs	0.44 (0.12–0.75)
One or more constructs addressed	0.41 (0.05–0.77)

↵* β coefficients estimate adjusted mean differences (vs reference category) in construct-specific after-visit scale scores using analysis of covariance models to adjust for measures of the relevant behavioral construct, health status (fair/poor vs. good/very good/excellent), and annual household income (<$50,000 vs. ≥$50,000) before the visit. Confidence intervals are corrected for within-physician clustering by use of robust standard errors.
↵† Higher numbers on the barriers scale indicate greater perceived barriers, so negative coefficients imply lower perceived barriers after versus before the visit.
CRC, colorectal cancer.

Table 4.

Colorectal Cancer Screening Test Planning, Ordering, and Completion by Whether and How Providers Discussed Colorectal Cancer Screening

Outcome	Overall	CRC Screening Discussion			P
Outcome	Overall	No Discussion	Screening Discussed Without Addressing Behavioral Constructs	One or More Constructs Addressed	P
Advancement in stage of readiness to undergo CRC screening (N = 40)^*	22/40 (55)	1/8 (13)	8/11 (73)	13/21 (62)	.03
CRC screening ordered (N = 50)	31/50 (62)	0/12 (0)	9/12 (75)	22/26 (85)	<.001
CRC completed (N = 50)	17/50 (33)	0/12 (0)	8/12 (67)	9/26 (35)	.001

Values provide as n/N (%).
↵* Defined as more advanced stage of readiness with regard to a CRC screening test after the visit (compared with before the visit); the analysis for this outcome was restricted to the 40 patients who did not report on questionnaires before the visit that they were planning to undergo fecal occult blood testing and colonoscopy.
CRC, colorectal cancer.

Table 5.

Adjusted Association between Provider Conversation and Advancement in Stage of Readiness and Colorectal Cancer Screening Completion

Independent Variable	OR (95% CI)^†
Advancement in stage of readiness (n = 40^*)
No discussion of CRC screening	1.0 (ref)
Any discussion of CRC screening	12.7 (1.3–682.0)
No discussion of CRC screening	1.0 (ref)
Screening discussed without addressing behavioral constructs	14.5 (1.04–986.6)
One or more constructs addressed	11.0 (1.06–602.2)
CRC screening completed (n = 50)
No discussion of CRC screening	1.0 (ref)
Any discussion of CRC screening	10.0 (1.4-∞)
No discussion of CRC screening	1.0 (ref)
Screening discussed without addressing behavioral constructs	33.2 (3.6-∞)
One or more constructs addressed	5.8 (0.7-∞)

↵* Analyses for this outcome was restricted to the 40 patients who did not report on questionnaires before the visit that they were planning to undergo fecal occult blood testing and colonoscopy during. Advancement in stage of readiness is defined as more advanced stage of readiness with regard to one of these tests post-visit as compared to pre-visit.
↵† Odds ratios were estimated using exact logistic regression and adjusted for health status (fair/poor vs good/very good/excellent), annual household income (<$50,000 vs ≥$50,000), and stages of readiness before the visit with regard to fecal occult blood testing and colonoscopy. The upper 95% CI may be unbounded in the setting of zero cells (i.e., when no patients completed CRC screening in the absence of CRC screening discussions).
CRC, colorectal cancer.