Paul C. Schroy III, MD, MPH; Subodh Lal, MD; Julie T. Glick, MPH; Patricia A. Robinson, MEd; Philippe Zamor, MD; and Timothy C. Heeren, PhD
Objective: To assess patient preferences for 1 of the recommended colorectal cancer screening options or stool DNA testing (sDNA), a novel noninvasive screening test.
Study Design: Cross-sectional survey of ambula-tory-care patients in the primary care setting.
Methods: A decision aid was administered to eligible subjects, using a trained interviewer format. The decision aid described the pros and cons of colonoscopy, fecal occult blood testing (FOBT), flexible sigmoidoscopy, flexible sigmoid-oscopy plus FOBT, double-contrast barium enema, and sDNA. After reviewing the decision aid, subjects were asked to identify a preferred screening option, test features influencing their choice, and level of interest in decision making.
Results: A total of 263 subjects completed the study. Colonoscopy (50.6%), sDNA (28.1%), and FOBT (18.3%) were preferred over the other screening options. Preferences were associated with race and education but not age, sex, or prior FOBT. Subjects who preferred colonoscopy rated accuracy as the most influential test feature, whereas those who preferred sDNA or FOBT rated concerns about discomfort or frequency of testing highest. Most subjects preferred a shared (54%) or patient-dominant (34%) decision-making process.
Conclusions: Colonoscopy was the most frequently preferred screening option for average-risk individuals. Noninvasive stool-based tests, particularly sDNA, were identified by most individuals who preferred an alternative to colonoscopy. These findings affirm the need to elicit patient preferences when selecting a screening option and suggest that provider-patient decision making can be tailored to include fewer options.
(Am J Manag Care. 2007;13:393-400)
Sufficient evidence has accumulated to suggest that screening is a cost-effective strategy for reducing not only colorectal cancer mortality through early detection but also incidence through the detection and removal of premalignant adenomatous polyps.1-3
Although screening is widely endorsed, a single "best" screening test has not been universally endorsed by authoritative groups.4-7
Economic analyses to date suggest that each of the currently recommended options, including annual fecal occult blood testing (FOBT), flexible sigmoid-oscopy every 5 years, the combination of annual FOBT and flexible sig-moidoscopy every 5 years, double-contrast barium enema (DCBE) every 5 years, and colonoscopy every 10 years, exhibit comparable cost-effectiveness.1,3,8
This lack of consensus regarding an optimal screening strategy has stimulated interest in shared decision making as a potential strategy for increasing screening rates among average-risk patients. Shared decision making is a sequential, interactive process involving information exchange, values clarification, decision making, and mutual agreement between providers and their patients.9,10
Studies demonstrating that patients have distinct preferences for the different colorectal screening strategies, which reflect the relative values they place on individual test features, lend further credence to this approach.11-21
The primary purpose of this study was to provide an updated assessment of patient preferences in light of widespread media attention promoting colonoscopy and the recent introduction of stool DNA testing (sDNA). sDNA is a novel screening strategy that offers a convenient, noninvasive alternative to existing screening tests. Like FOBT, sDNA can potentially detect neoplasia anywhere in the colon with sample collection that can be performed at home. Unlike FOBT, however, no dietary or medication restrictions are required and the sample can be collected using a simple device that obviates the need for handling stool. Multitarget sDNA has reported sensitivities for detecting colorectal cancer and advanced adenomatous polyps in the range of 52% to 91% and 18% to 82%, respectively, with specificities of approximately 88% to 95%.22-29
The lowest sensitivities for both cancers (52%) and advanced adenomas (18%) were observed in the only prospective screening study of asymptomatic, mostly average-risk patients.26
Although the sensitivity was lower than previously reported, that same study found that sDNA was significantly more sensitive than FOBT for detecting cancers (52% vs 13%) but not advanced adenomas (15% vs 11%). Together, these data suggest that sDNA has medium to high sensitivity for detecting colorectal cancers but variable sensitivity for detecting advanced adenomatous polyps. In addition to being more sensitive than FOBT for detecting colorectal cancers,26
existing data also suggest that sDNA is preferred over both FOBT and colonoscopy among patients experienced with all 3 tests.30
This study addresses the important issue of whether patients lacking this experience would express similar preferences if educated about the pros and cons of the different screening modalities.
Asymptomatic, average-risk individuals between 50 and 75 years of age with no prior screening, except possibly FOBT, were deemed eligible for the study. Potential subjects unable to speak or read English were excluded. Patients with a personal history of colorectal neoplasia (cancer or polyps), inflammatory bowel disease, or a family history of colorectal neoplasia were excluded. Potential subjects were recruited from 2 sources: (1) direct patient referrals from primary care providers practicing at Boston Medical Center, an urban academic medical center, or the South Boston Community Health Center; and (2) flyers posted at various ambulatory care sites within Boston Medical Center. Participants were reimbursed $20 to cover parking and travel expenses if extra visits were required.
The study used a cross-sectional survey design similar to that used in our prior patient preference study.14
Written consent was obtained from eligible subjects by the research assistant just prior to initiating the survey. The survey was conducted using a structured interview format in which 1 of 2 research assistants verbally read the educational components of the survey instrument to the subject, who visually followed along. After concluding the educational component, subjects were asked to complete the decision aid's preference assessment, rank order test features influencing their choices, and answer a question related to decision-making autonomy. Demographic information was obtained at the end of the survey. The entire interview took approximately 20 minutes. All interviews were conducted by 1 of 2 research assistants in a private consultation room located in 1 of Boston Medical Center's outpatient clinics or the endoscopy unit. The study was reviewed and approved by Boston Medical Center's institutional review board prior to commencement.
Our survey instrument consisted of 4 main parts: (1) an educational decision aid, (2) an assessment of patient preferences and factors influencing their choices, (3) an assessment of decision-making autonomy relevant to patients' choice of a colorectal cancer screening test, and (4) an assessment of demographic information and prior screening experience. The complete survey instrument
can be viewed online at www.ajmc.com
Apart from information about sDNA, the decision aid used in this study was nearly identical to the one used in our prior study of average-risk patients. 14Patient Preferences.
After completing the educational component of the instrument, patients were asked to rank order their preferences as to the screening strategies and the importance of the various test features in determining their preference. They also were asked about willingness to pay if their preference was not covered by insurance; response categories were "yes, regardless of the cost," "maybe, depending on the cost," and "no."Decision-making Autonomy.
We assessed patients' level of desire to participate in decision making regarding the choice of screening test using a single-item, 5-point scale as described by Strull et al.31
For operational purposes, responses were categorized as "physician-dominant," "shared," or "patient-dominant" decision-making processes.Demographics and Prior Screening Experience.
Demographic information including age, sex, race/ethnicity, and education was ascertained. In addition, subjects were asked whether they had undergone prior FOBT.
Sample Size and Power Calculations
Our analyses focus on describing patient screening preferences and identifying patient characteristics and attitudes associated with those screening preferences. In a previous study, we found that nearly 50% of patients preferred colonoscopy.14
We determined that a sample of 260 subjects provided >80% power of detecting, at the 2-tailed P
<.05 level, a 20% difference in preferences between 2 subgroups of roughly equal size. Estimated percentages based on this sample are accurate to within ±6 percentage points (the width of a 95% confidence interval [CI] for a percentage).
Descriptive statistics were used to characterize the study population, the screening preferences, and the important test features used to formulate these preferences. The percentages of patients preferring each of the screening options were compared using the X2
goodness-of-fit test for equal percentages.X2
tests of independence were performed to assess associations between screening preference and subject characteristics defined by demographics, test feature influencing choice, and importance of insurance coverage. Similar analyses were performed to assess associations between desire to participate in the decision-making process and the same study group characteristics. Multiple logistic regression analysis was performed to identify independent determinants of screening preference. Significance was defined at the P
<.05 level. All statistical analyses were performed using SAS, version 8.2 (SAS Institute Inc, Cary, NC).
A total of 263 subjects, including 201 primary care patients and 62 respondents to the posted flyers, were consecutively enrolled in the study between September 2002 and August 2003. Table 1
summarizes the sample's demographics and screening experience. A majority of the subjects were 50 to 59 years of age (70.3%), female (62.4%), and white (57.8%) with a high school education or less (60.1%). Nearly 50% had undergone prior FOBT.
Screening Test Preferences and Features Influencing Choice
PDF is available on the last page.