- © 2005 by American Society of Clinical Oncology
Health Care Provider-Directed Intervention to Increase Colorectal Cancer Screening Among Veterans: Results of a Randomized Controlled Trial
- M. Rosario Ferreira,
- Nancy C. Dolan,
- Marian L. Fitzgibbon,
- Terry C. Davis,
- Nicolle Gorby,
- Lisa Ladewski,
- Dachao Liu,
- Alfred W. Rademaker,
- Franklin Medio,
- Brian P. Schmitt and
- Charles L. Bennett
- From the Veterans Affairs Midwest Center for Health Services and Policy Research, Hines; Veterans Affairs Chicago Health Care System; Departments of Medicine, Psychiatry, and Preventive Medicine, Center for Healthcare Studies, and Robert H. Lurie Comprehensive Cancer Center, Feinberg School of Medicine, Northwestern University, Chicago, IL; Louisiana State University Health Sciences Center-Shreveport, Shreveport, LA; and Medical University of South Carolina, Charleston, SC
- Address reprint requests to M. Rosario Ferreira, MD, MAPP, 676 N St Clair St, Ste 1400, Chicago, IL 60611; e-mail: mr-ferreira{at}northwestern.edu
Abstract
Purpose Colorectal cancer screening is the most underused cancer screening tool in the United States. The purpose of this study was to test whether a health care provider–directed intervention increased colorectal cancer screening rates.
Patients and Methods The study was a randomized controlled trial conducted at two clinic firms at a Veterans Affairs Medical Center. The records of 5,711 patients were reviewed; 1,978 patients were eligible. Eligible patients were men aged 50 years and older who had no personal or family history of colorectal cancer or polyps, had not received colorectal cancer screening, and had at least one visit to the clinic during the study period. Health care providers in the intervention firm attended a workshop on colorectal cancer screening. Every 4 to 6 months, they attended quality improvement workshops where they received group screening rates, individualized confidential feedback, and training on improving communication with patients with limited literacy skills. Medical records were reviewed for colorectal cancer screening recommendations and completion. Literacy level was assessed in a subset of patients.
Results Colorectal cancer screening was recommended for 76.0% of patients in the intervention firm and for 69.4% of controls (P = .02). Screening tests were completed by 41.3% of patients in the intervention group versus 32.4% of controls (P = .003). Among patients with health literacy skills less than ninth grade, screening was completed by 55.7% of patients in the intervention group versus 30% of controls (P < .01).
Conclusion A provider-directed intervention with feedback on individual and firm-specific screening rates significantly increased both recommendations and colorectal cancer screening completion rates among veterans.
INTRODUCTION
Colorectal cancer is the third most common cancer and the third most common cause of cancer-related deaths among men in the United States.1 In 2004, an estimated 146,940 persons will be diagnosed with colorectal cancer in the United States, and 56,730 patients will die of the disease.1 Colorectal cancer screening with fecal occult blood testing (FOBT) or flexible sigmoidoscopy reduces colorectal cancer–related mortality.2-7 Although colorectal cancer screening is recommended for individuals 50 years and older,8-10 screening is underused. In a national population-based survey conducted in 2001, only 23.5% of respondents reported having a FOBT in the preceding year, and 38.7% reported having a lower endoscopy (either flexible sigmoidoscopy or colonoscopy) in the preceding 5 years.11 In a 2003 report on the quality of care in the Veterans Affairs (VA) health care system, rates of colorectal cancer screening were the lowest for any of 17 measured quality-of-care standards.12
Nationally, low utilization rates of colorectal cancer screening have been associated with patient factors, such as poor socioeconomic status, racial and ethnic minorities, and low levels of education,13-16 as well as physician-related factors, including failure to remember to offer colorectal cancer screening or lack of time to discuss colorectal cancer screening during office visits for general medical problems.17
There have been several interventions designed to increase adherence to colorectal cancer screening, including direct mailing of FOBT kits, videos, informational leaflets, and reminders, which are given to patients to review.18 Physician-directed colorectal cancer screening interventions, including reminder systems for providers, have increased FOBT adherence in the short term.19 Prior intervention assessments have been limited because none has focused on populations characterized by high rates of individuals of lower socioeconomic status and/or who have limited literacy skills. In clinical settings with these populations, health care providers may not have time to address preventive health measures, and patients may not be able to understand or have access to information that is disseminated in written form or by telephone or mail.
The VA medical system is the largest integrated health delivery system in the country providing equal access to care. The patient population using the VA medical system has access to medical care, regardless of income level. More than half of the VA users report an income below $20,000, and only 58% have a 12th grade education level.20 Many have limited literacy skills and are unable to read or understand health-related materials.21 Physician communication to patients about the importance of colorectal cancer screening may not be well understood when patients have limited health literacy skills.21,22 In many instances, physicians are not aware of literacy barriers when communicating with patients who have limited health literacy skills and may not convey meaningful and convincing colorectal cancer screening messages to this patient population. In the VA system, a prior health maintenance study evaluated reminder systems for general medical problems, such as blood pressure control and diabetes care, and found that because of physician fatigue, any beneficial effects of this intervention quickly dissipated.23 Colorectal cancer screening was not included as one of the general medical practices in this VA randomized intervention.23
In this article, we describe a health care provider–directed intervention designed to increase the rates of colorectal cancer screening recommendations and adherence in a VA population. The health care provider–directed intervention included 1-hour meetings at 4- to 6-month intervals, during which the providers received colorectal cancer screening rates for the group, individualized confidential feedback, and instruction on effective strategies to improve communication with patients with limited literacy skills.
PATIENTS AND METHODS
Setting
We conducted a controlled trial, which was randomized by outpatient clinic firm, in two general medicine primary care outpatient firms at a VA Medical Center in Chicago, Illinois, from May 2001 to June 2003. Although the study was initially designed as a combined provider and patient intervention, this article will focus mainly on the health care provider intervention. The study was approved by the institutional review board.
Patients
Participants were male veterans who were 50 years and older and who were scheduled to be seen for a new or ongoing health problem by one of the providers from the two outpatient firms at the study medical center.
Eligibility
Patients were excluded if they had a personal or family history of colorectal cancer or polyps, a personal history of inflammatory bowel disease, or if they had had a home FOBT in the previous year or a flexible sigmoidoscopy or colonoscopy in the previous 5 years.
Health Care Providers
Health care providers in the two participating firms included residents, attendings, and nurse practitioners assigned to each firm. Each physician and nurse practitioner was assigned to one firm only, so there was no contamination. In the control arm, there were three attendings and one nurse practitioner for the duration of the study period and a total of 49 residents, with approximately 30 residents at any given time. In the intervention arm, there were three attendings, two nurse practitioners, and a total of 55 residents, with approximately 30 residents at any given time. Each firm contained similar proportions of first-, second-, and third-year residents.
Group Assignment
Randomization was performed before initiation of the study by randomly allocating one firm to the intervention arm and one firm to the control arm. Research assistants assessed eligibility by reviewing medical records of patients scheduled for appointments. Research assistants obtained written informed consent from a subset of eligible patients in both the control and the intervention firm to complete a literacy assessment and brief survey.
Intervention
Health care providers in the intervention firm were invited to attend a 2-hour workshop on rationale and guidelines for colorectal cancer screening and on improving communication with patients with low literacy skills. This initial session was given 2 months before the initiation of the study and was repeated for resident physicians who joined the firm during the study period. Every 4 to 6 months, providers were invited to attend 1-hour feedback sessions, during which they received information on the firm's colorectal cancer screening recommendation rate and patient adherence to recommended tests. The providers also received confidential information on their individual recommendation and adherence rates in a sealed envelope. These sessions included reviews of colorectal cancer screening guidelines and practical strategies to communicate with patients with low health literacy skills in busy, high-volume VA primary care clinics. Small group discussions and role-playing sessions focused on empowering providers to effectively recommend colorectal cancer screening by crafting short, powerful, and personal messages that fit individual providers and patients. The sessions also addressed provider time constraints and other preventive messages that compete with a discussion of colorectal cancer screening. Further details on these sessions are listed in the Appendix. Health care providers who were unable to participate in these sessions were contacted by one of the investigators, who briefly reviewed the highlights of the sessions and gave them their individualized feedback reports. The patient intervention was designed based on focus groups conducted with veterans. The intervention consisted of a brochure and video, which was designed and produced by the investigators. The video included cognitive information on colorectal cancer and screening and also social and emotional messages designed to motivate and empower patients to overcome barriers and increase self-efficacy. The brochure used simple language and graphics, including a simplified illustration of the colon. We also designed a simplified version of the instructions that are included with each FOBT. The purpose of this simplified version of instructions was to allow patients with limited literacy skills to follow the required directions. Patients were recruited to participate in the study when they presented to a visit with a primary care physician. Consenting patients viewed the video and received a brochure. All FOBT kits in the intervention firm contained the simplified instructions, regardless of whether the patients viewed the video or not. However, because of logistical reasons related to coordination of the patient's clinical visit, questionnaire completion, and video viewing, only 204 patients actually received the patient intervention.
Demographic Characteristics of Study Patients
Screening Recommendations and Completion of Screening Tests in the Control and Intervention Groups
Completion of Colorectal Cancer Screening Tests Among Patients With High and Limited Literacy Skills in the Control and Intervention Groups
Health Care Provider Intervention Table
Objectives
In this study, we tested the primary hypothesis that a provider-directed intervention would increase rates of colorectal cancer screening recommendation by providers and rates of completion of colorectal cancer screening tests by patients.
Outcome Measures
A research assistant examined the patient's electronic medical records to determine whether the health care provider had issued a colorectal cancer screening recommendation and whether the patient had completed any colorectal cancer screening tests within 6 to 18 months of the index visit. The index visit was defined as the first visit that each patient had with his physician or nurse practitioner after the study was initiated. Research assistants were trained in chart reviews by two of the investigators. The first 20 records reviewed by each research assistant were reabstracted for consistency of the information collected. Research assistants were given feedback on the quality of the data abstraction and received further training where needed. The main outcome measures were percentage of eligible patients who received provider recommendations for colorectal cancer screening and percentage of eligible patients who completed a colorectal cancer screening test (home FOBT, flexible sigmoidoscopy, or colonoscopy).
Because of our interest in health literacy as a potential confounding factor, we conducted exploratory analyses to address colorectal cancer screening rates according to health literacy skills. At baseline, we identified 185 patients in the control firm and 197 patients in the intervention firm who were available to complete a short interview before their visit with the primary care provider. These patients were asked to participate in a trained research assistant–administered literacy assessment using the Rapid Estimate of Adult Literacy in Medicine (REALM) instrument. The REALM is a commonly used health word recognition test that is highly correlated with other general reading tests and the Test for Functional Health Literacy in Adults.24,25 REALM raw scores range from 0 to 66 and can be converted into one of the following four reading grade levels: third grade or less (score, 8 to 18), fourth to sixth grade (score, 19 to 44), seventh to eighth grade (score, 45 to 60), and ninth grade and above (score, 61 to 66).
Statistical Analysis
Data were analyzed using a z test for comparing two independent proportions, with adjustment made for clustering of patients by provider.26 Data were analyzed with SAS Statistical Software (SAS Institute Inc, Cary, NC). The design effect is a multiplicative factor that determines the fold increase that is required in the sample size because of the intercorrelation of data within clusters.26 In this study, the design effect was 2.3 for the intervention group and 1.6 for the control group. The rate of any screening completion (FOBT or flexible sigmoidoscopy or colonoscopy, alone or in combination) in the control group was 32.4%. With these parameters, there is 80% power to detect a screening rate of 40.8% in the intervention group, assuming a two-tailed test and type I error rate of 5%.
Role of Funding Sources
Funding to develop, implement, and assess the intervention was provided by grants from the VA Health Services Research and Development Service and the National Institutes of Health. The funding sources had no role in the collection, analysis, or interpretation of the data or in the decision to submit the study for publication.
RESULTS
Enrollment
The process used to identify and include patients in the study is outlined in Figure 1. Research assistants reviewed the medical records of 5,711 patients and identified 4,318 patients who had an age of 50 years or older and who had visited the clinics between May 1, 2001, and December 31, 2002. Patients who met inclusion criteria by chart review were included in the study. In the control firm, the most common reasons for ineligibility were as follows: a personal history of colorectal cancer or polyps or inflammatory bowel disease (13.3%), a family history of colorectal cancer or polyps (3.7%), or the patient had completed colorectal cancer screening with either an FOBT in the previous year or a flexible sigmoidoscopy or colonoscopy in the previous 5 years (36.1%). Thirty-four patients died during the follow-up period and were also excluded. In the intervention firm, the most common reasons for ineligibility were as follows: a personal history of colorectal cancer or polyps or inflammatory bowel disease (13.8%), a family history of colorectal cancer or polyps (3.3%), or the patient had completed colorectal cancer screening with either an FOBT in the previous year or a flexible sigmoidoscopy or colonoscopy in the previous 5 years (35%). Thirty-four patients died during the follow-up period and were also excluded. A total of 963 patients met inclusion criteria in the control firm, and 1,015 patients met inclusion criteria in the intervention firm. One hundred eighty-five patients in the control firm and 197 patients in the intervention firm completed an interviewer-administered literacy assessment and brief survey. Chart reviews were available for all of the 1,978 patients included in the study.
Flow diagram of study enrollment.
Health Care Provider Participation in the Intervention
Of the 60 providers in the intervention firm, 15 residents did not participate in the initial workshop because they were scheduled to finish their residency in less than 2 months after the study was initiated. There were four feedback sessions with providers; 84% of the physicians and nurse practitioners attended at least one session. The number of patients included in the study varied from one to 40 (median, 19 patients) for the 60 providers in the intervention arm and from one to 46 (median, 20 patients) for the 53 providers in the control arm.
Study Population
The mean age of the patients in the study was 67.8 years; 45% patients were white, and 50% were African American. The distribution of age by decade was similar in the control and intervention groups. The mean number of general medicine clinic visits was 2.91 in the control group and 2.77 in the intervention group (P = .05; Table 1).
Approximately 20% of the patients in each arm participated in the literacy assessment and survey. In the intervention and control arms, one third of these patients had literacy levels lower than ninth grade, and 79% had completed high school. The distribution of marital status was similar in the control and intervention groups.
Recommendation and Completion Rates for Colorectal Cancer Screening
In the 6- to 18-month time period after the index visit, 69.4% of the control group patients and 76.0% of intervention group patients received a recommendation to undergo colorectal cancer screening from a health care provider (P = .02; Table 2). With respect to individual tests, FOBT was recommended to 240 veterans in the control arm (24.9%) and 576 veterans in the intervention arm (56.7%), whereas flexible sigmoidoscopy or colonoscopy was recommended for 641 veterans in the control arm (66.6%) and for 807 veterans in the intervention arm (79.5%).
In the intervention group, 41.3% of patients completed either a FOBT, flexible sigmoidoscopy, or colonoscopy compared with 32.4% of controls (P = .003; Table 2). In the control group, 165 veterans (17.1%) returned their FOBT card, and 174 (18.1%) underwent flexible sigmoidoscopy or colonoscopy. In the intervention group, 295 veterans (29.1%) returned their FOBT cards, and 190 (18.7%) underwent flexible sigmoidoscopy or colonoscopy. Among patients for whom literacy skills were measured at less than the ninth grade level, 55.7% of patients in the intervention group completed screening tests compared with 30.0% of patients in the control group (P = .002; Table 3).
DISCUSSION
We found that a health care provider–directed intervention that provided feedback on individual and firm-specific colorectal cancer screening recommendation and adherence rates resulted in a 7% absolute increase in the rates of colorectal cancer screening recommendations documented by providers in electronic medical records and a 9% absolute increase in the rates of completion of colorectal cancer screening (by FOBT, flexible sigmoidoscopy, or colonoscopy), as documented in the medical record, among veterans who received care in a general medicine VA clinic. Our study population included male veterans, approximately half of whom were African American, who were at average risk for colorectal cancer and who were currently not compliant with colorectal cancer screening. The health care providers in the intervention arm had a high participation rate in the feedback sessions, and providers who attended more sessions were more likely to recommend screening to their patients (data not shown). In interpreting our findings, several factors should be considered. The majority of prior physician-directed colorectal cancer screening interventions use reminder systems that often include other health maintenance protocols, usually implemented in large practice settings. Overall, reminder systems for providers have resulted in an absolute increase in FOBT adherence rates of up to 14%.19 In our VA center, computerized clinical reminder systems, including a reminder for FOBT, were in place in both the intervention and the control firms before the initiation of this intervention, and improvement in screening rates in the intervention compared with the control firm occurred while both firms had clinical reminders in effect. It should be noted that a prior VA study found that provider fatigue results in high overriding rates for computerized reminders for health maintenance practices, although colorectal cancer screening was not specifically evaluated in this study.23 We found that a provider-directed intervention, focusing on interactive sessions and including individualized feedback, significantly increased colorectal cancer screening adherence beyond what could be attributed to reminder systems, underscoring the benefit of personalized feedback efforts.27
Limited literacy is a recently recognized and often overlooked potential barrier to colorectal cancer screening.28 We sought to improve patient-provider communication, especially for patients with limited literacy skills, a barrier that affects many veterans. The provider feedback sessions highlighted patient communication strategies, particularly those that have proven effective in communicating health maintenance messages to individuals with limited literacy skills. During the sessions, providers were encouraged to share practical strategies that they found helpful in communicating with their VA patients. These interactive sessions provided an opportunity to craft effective colorectal cancer screening messages that could be delivered to VA patients in a short period of time during clinic visits. Given this emphasis in the feedback sessions, it was reassuring to find that, in exploratory analyses among a subset of veterans for whom literacy skills were assessed, those who had limited literacy skills in the intervention group had an almost two-fold improvement in screening rates.
The limitations of our study should be addressed. First, the patient population included only males who received care in VA general medicine clinics. Although the VA provides access to health care for all veterans, the overwhelming proportion of whom are male, it is also the largest integrated health delivery system in the country. Further studies are needed in non-VA health care settings that provide care for large numbers of persons of lower socioeconomic status. Second, the study was randomized by firm and not by patient. However, patients were randomly assigned to firms by social security number, and the patients in the two study arms were similar in terms of demographic characteristics. Moreover, among persons in the intervention and control arms who participated in the literacy assessment and brief survey, characteristics, such as marital status, education, and health literacy skills, were similar. Third, we do not know whether the presence of comorbid illnesses differentially affected participation in colorectal cancer screening, although we expect a similar distribution of comorbidities in patients in the control and intervention arms. Fourth, the patient-directed component of the intervention was not fully implemented as planned, which limits our ability to assess its effect. However, additional analyses suggest that most of the improvement in colorectal cancer screening recommendation and completion rates resulted from the provider intervention (data not shown). Finally, the quality of the review of medical records by research assistants was not evaluated, with random reabstraction of a subset of medical records. However, we expect this limitation to affect equally the intervention and the control arms, and therefore, it should not bias our interpretation of the findings of the study.
In conclusion, we found that a health care provider–directed intervention, which included educational sessions and group and individualized feedback of screening rates, significantly increased adherence to colorectal cancer screening among veterans attending a general medicine VA clinic in a large urban area. Because many of VA medical centers are located in large urban areas, our results suggest that implementation of our intervention throughout urban VA medical centers could be used to improve the system-wide performance measure for colorectal cancer screening in the VA health care system.
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
Acknowledgments
We thank Rebecca Newlin, June Lee, and Phillip Hilliker for their assistance with the implementation of this study.
Footnotes
-
Supported by grant No. PCI 99-158 from the Health Services Research Division of the Department of Veterans Affairs and by grant No. R01 CA86424-01A2 from the National Cancer Institute. M.R.F. is supported by a Research Career Development Award from the Health Services Research and Development Service of the Department of Veterans Affairs (grant No. RCD-01005-1) and by the Coleman Foundation.
Presented in part at the Veterans Affairs Health Services Research and Development’s 22nd National Meeting, Washington, DC, March 9-11, 2004; at the 28th Annual Meeting of the American Society for Preventive Oncology, Bethesda, MD, March 14-16, 2004, Bethesda, MD; at the 105th Annual Meeting of the American Gastroenterological Association, New Orleans, LA, May 15-20, 2004; and at the 40th Annual Meeting of the American Society of Clinical Oncology, New Orleans, LA, June 5-8, 2004.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
- Received July 13, 2004.
- Accepted December 2, 2004.
