Primary Care Research

Studies examine ways to improve delivery of colorectal cancer screening in primary care practice

About 52,000 people died of colorectal cancer in the United States in 2007. However, in 2005 only about half of adults aged 50 and older had been screened in accordance with U.S. Preventive Services Task Force recommendations. The September 2008 supplement to Medical Care 46(9) describes and evaluates several promising strategies for improving colorectal cancer screening (CRC) rates within busy primary care practices and among special populations. The supplement was guest-edited by David Lanier, M.D., of the Agency for Healthcare Research and Quality (AHRQ), and Carrie N. Klabunde, Ph.D., of the National Cancer Institute (NCI). The studies were funded over the past 5 years by AHRQ and NCI. Summaries of the studies supported by AHRQ or conducted by AHRQ staff follow.

Reprints of the supplement (AHRQ Publication No. 08-0111) are available from the AHRQ Publications Clearinghouse.

Klabunde, C.N., Lanier, D., Meissner, H.I., and others. "Improving colorectal cancer screening through research in primary care settings," pp. S1-S4.

This introductory article describes the organization of the journal supplement into four thematic sections: informed decisionmaking and patient-provider communication in primary care; meeting the needs of diverse populations; provider- and health system-directed approaches to improving CRC screening delivery; and formative evaluation of strategies for increasing CRC screening use.

The supplement also contains three synthesis articles that address the importance of building data systems to evaluate CRC screening practices and outcomes at the population level; implications of new CRC screening technologies for primary care practice; and research networks as vehicles for improving CRC screening delivery in primary care.

Hawley, S.T., Volk, R.J., Krishnamurthy, P., and others. "Preferences for colorectal cancer screening among racially/ethnically diverse primary care patients," pp. S10-S16. (AHRQ grant HS11187).

This study found that primary care patients have distinct preferences for CRC screening tests that can be linked to test attributes. Thus, tailoring screening recommendations to patients' preferences may increase screening adherence. The authors asked a diverse group of patients to rate eight hypothetical CRC screening test scenarios comprised of different combinations of five attributes and six scenarios designed to depict recommended CRC screening tests (e.g., fecal occult blood test, flexible sigmoidoscopy, colonoscopy, and double-contrast barium enema), including new technology (e.g., virtual colonoscopy, and fecal immunochemical test).

Ling, B.S., Trauth, J.M., Fine, M.J., and others. "Informed decision-making and colorectal cancer screening: Is it occurring in primary care?," pp. S23-S29.(AHRQ grant HS10864).

Informed deicisionmaking is lacking during discussions of CRC screening between patients and their primary care providers, concludes this study. The authors analyzed audiotaped clinic visits between patients and primary care providers to assess the level of informed decisionmaking during discussions about CRC screening. Six of the nine informed decisionmaking elements occurred in 20 percent or fewer visits, and none of these elements were addressed in half or more of the visits. Certain decisional elements were negatively associated with screening. For example, CRC screening occurred less often for those discussing the "pros and cons" of screening (12 vs. 46 percent) and "patient preferences" (6 vs. 47 percent) compared with those who did not.

Nease Jr., D.E., Ruffin, M.T., Klinkman, M.S., and others. "Impact of a generalizable reminder system on colorectal cancer screening in diverse primary care practices," pp. S68-S73. (AHRQ/NCI cooperative agreement).

Implementing a generalizable computerized reminder system (CRS) in diverse primary care practices significantly improved CRC screening rates, according to this study. All but 1 practice increased their CRC screening rates, ranging from 41.7 to 50.9 percent across all 12 practices. Technology capabilities influenced printing days (74 percent for high-technology vs. 45 percent for low-technology practices). Also, practice cohesion influenced screening trends, with highly cohesive practices showing nearly twice the improvement in their CRC screening rate than not very cohesive practices (15.3 vs. 7.9 percent). The findings were based on pre- and post-CRS implementation CRC screening rates among patients seen at the 12 practices during the 9-month study period.

Schneider, E.C., Rosenthal, M., Gatsonis, C.G., and others, "Is the type of Medicare insurance associated with colorectal cancer screening prevalence and selection of screening strategy?" pp. S84-S90. (AHRQ grant HS11431).

The type of Medicare insurance a patient had in 2000 was linked to prevalence and type of CRC screening, found this study. The researchers compared CRC screening prevalence and type of screening among Medicare beneficiaries enrolled in Medicare managed care (MMC) plans, Medicare fee-for-service (FFS) plans with supplemental insurance, and those with Medicare FFS coverage only. Interval-appropriate CRC screening was reported by 54.7 percent of those in Medicare FFS plans with supplemental insurance, 52.9 percent in MMC plans, and 36.3 percent in the FFS group that did not have supplemental insurance. Use of fecal occult blood testing was a more common screening strategy among persons in MMC plans than those with FFS plans with or without supplemental insurance.

Lane, D.S., Messina, C.R., Cavanagh, M.F., and Chen, J.J. "A provider intervention to improve colorectal cancer screening in county health centers," pp. S109-S116. (AHRQ/NCI cooperative agreement).

Publicly funded health centers care for disadvantaged groups who typically underuse CRC screening. This study found that when health center providers received continuing medical education with a teambuilding strategic planning exercise, provider referrals, dispensing, and completion of CRC screening significantly improved. Intervention centers had a 16 percent increase in CRC referral/dispensing/completion compared with a 4 percent increase among non-intervention centers. Also, fewer patients at the intervention centers cited lack of physician recommendation as a reason for not having CRC screening.

Chan, E.C. and Vernon, S.W. "Implementing an intervention to promote colon cancer screening through E-mail over the Internet: Lessons learned from a pilot study," pp. S117-S122. (AHRQ grant HS11421).


Physician E-mails recommending CRC screening to 97 patients in their primary care practice were not feasible to implement, concluded this pilot study. However, reasons for lack of success differed for patients with home or work E-mails and those using computers at the public library to access E-mail. Patients with both types of Internet access were randomized to InterNet LETter (NetLET) or a control group. The NetLET was a personalized E-mail from the physician reminding the patient to undergo CRC screening via a fecal occult blood test (FOBT) and linked to a Web page with more information. The control group received a reminder letter from their physician. All patients were mailed an FOBT kit. Among private access patients, 26 percent of the intervention and 23 percent of the control group returned an FOBT. Among public access patients, none of the intervention group, but 3 of 9 in the control group returned an FOBT.

Lanier, D. "Practice-based research networks: Laboratories for improving colorectal cancer screening in primary care practice," pp. S147-S152.

Practice-based research networks (PBRNs), which AHRQ has supported since 1999, are a strong resource for the study of best methods for improving CRC screening in complex, time-pressured primary care offices, notes the author of this article. He focuses on two types of PBRNs currently active in the United States —those composed mostly of smaller, independent primary care practices and those composed of large integrated health systems—and the types of CRC screening-related questions that have been or can be addressed by these networks. For example, by probing the rich clinical databases of HMOs, the Cancer Research Network has been able to conduct population-based, longitudinal monitoring of the use of CRC screening services within a managed care population and study the occurrence of rare events, such as complications of screening procedures.

Return to Contents
Proceed to Next Article