Table 4.16. Screening Rate Multivariable Logistic Regression, Including Study Arm (Intervention and Control Practices): Reduced Model

Health Care Systems for Tracking Colorectal Cancer Screening Tests

Characteristics	Stool Test^a				Colonoscopy^b				Any Screening Test^c
Characteristics	OR	95% CI		p	OR	95% CI		p	OR	95% CI		p

Patient Characteristics
*Age*
50-54	—	—	—	—					—	—	—	—
55-64	1.25	0.81	1.92	0.311					1.04	0.81	1.32	0.781
65-69	1.81	1.31	2.51	0.000					1.26	1.00	1.58	0.052
70-79	1.88	1.12	3.16	0.016					1.32	1.02	1.72	0.038
Practice Characteristics
*Study arm*
Control	—	—	—	—	—	—	—	—	—	—	—	—
Card intervention	8.00	1.71	37.43	0.008	2.62	1.72	3.99	0.000	3.09	2.26	4.22	0.000
Kit intervention	19.84	3.95	99.69	0.000	1.50	0.60	3.79	0.387	6.64	4.50	9.80	0.000
*Specialty*
Family medicine					—	—	—	—	—	—	—	—
General internal medicine					1.27	1.03	1.57	0.025	1.16	0.93	1.45	0.182
*Size*
Small					—	—	—	—	—	—	—	—
Large					0.64	0.41	1.01	0.054	0.89	0.60	1.34	0.588
*Location*
Urban					—	—	—	—	—	—	—	—
Rural					2.70	1.70	4.29	0.000	1.69	1.10	2.60	0.016
Suburban					3.60	2.40	5.41	0.000	2.10	1.49	2.96	0.000
Data sources used^d
Least complete	—	—	—	—	—	—	—	—	—	—	—	—
Moderately complete	1.75	0.58	5.27	0.320	18.92	7.20	49.66	0.000	4.84	1.88	12.45	0.001
Most complete	1.38	0.65	2.92	0.395	13.85	8.05	23.83	0.000	3.31	1.79	6.12	0.000

^a Hosmer-Lemeshow Χ² = 1.96, p = 0.9619.
^b. Hosmer-Lemeshow Χ² = 10.2, p = 0.1163.
^c Hosmer-Lemeshow Χ² = 10.49, p = 0.232. Includes stool test, colonoscopy, flexible sigmoidoscopy, or barium enema x ray. Persons with screens by multiple modalities are counted only once.
^d The data sources used affects the amount of screening information available for patients of a given practice. It is not an indicator of predicting screening. Instead, this variable is used to control for the fact that some practices had more complete data than others.

Note: Confidence intervals and p values have been adjusted using cluster variance estimators.

Return to Document

Current as of December 2010

Internet Citation: Table 4.16. Screening Rate Multivariable Logistic Regression, Including Study Arm (Intervention and Control Practices): Reduced Model: Health Care Systems for Tracking Colorectal Cancer Screening Tests. December 2010. Agency for Healthcare Research and Quality, Rockville, MD. http://www.ahrq.gov/research/findings/final-reports/crcscreeningrpt/crcscreentab4-16.html