Preventive Medicine 49 (2009) 442–448
Contents lists available at ScienceDirect
Preventive Medicine j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / y p m e d
Colorectal cancer screening mode preferences among US veterans Adam A. Powell a,b,⁎, Diana J. Burgess a,b, Sally W. Vernon c, Joan M. Griffin a,b, Joseph P. Grill a, Siamak Noorbaloochi a,b, Melissa R. Partin a,b a b c
Center for Chronic Disease Outcomes Research (CCDOR), Minneapolis Veterans Affairs Medical Center, Minneapolis, MN, USA University of Minnesota Department of Medicine, Minneapolis, MN, USA Division of Health Promotion and Behavioral Sciences, University of Texas-Houston, USA
a r t i c l e
i n f o
Available online 8 September 2009 Keywords: Colorectal cancer screening Colonoscopy Fecal occult blood testing Patient preferences
a b s t r a c t Objective. To assess colorectal cancer (CRC) screening mode preferences and correlates of these preferences among US veterans at average risk for CRC. Method. A cross-sectional survey of a nationally representative sample of VA patients was conducted between January 2005 and December 2006. We report preference distributions for screening modes among 2068 average-risk veterans and across patient subgroups based on personal, behavioral, and environmental factors. Independent predictors of preferences are identified through hierarchical logistic regression models. Results. Colonoscopy (37%) was the most preferred mode followed by fecal occult blood test (FOBT) (29%). The strongest predictors of preferences were previous screening experience, provider recommendation, and use of non-VA healthcare services. Participants in higher socioeconomic groups were more likely to choose colonoscopy and less likely to indicate no preference. Conclusion. Screening programs that offer only one mode fail to accommodate the preferences of a substantial proportion of patients. Within the VA, adding screening colonoscopy to programs currently offering only FOBT is likely to increase preferences for colonoscopy, as patients incorporate provider recommendations for and personal experience with colonoscopy into their preferences. This is likely to disproportionately benefit lower socioeconomic groups who do not currently have access to non-VA colonoscopy services. Published by Elsevier Inc.
Introduction Colorectal cancer (CRC) is the second leading cause of cancer mortality (National Cancer Institute, 2006). Screening can reduce CRC mortality by 16%–33% (Hardcastle et al., 1996; Kewenter et al., 1994; Kronborg et al., 1996; Mandel et al., 1993), and at the time of data collection for this study, the U.S. Preventive Services Task Force (2002) recommended screening using either fecal occult blood test (FOBT) annually, sigmoidoscopy or double-contrast barium enema every 5 years, or colonoscopy every 10 years. In 2003, FOBT constituted 90% of CRC screening within the VA (ElSerag et al., 2006), suggesting that few veterans were presented with other screening mode options. However, current VA policy specifies veterans be offered multiple screening modes (Perlin, 2005) and recent guidelines (Levin et al., 2008) recommend tests involving a structural exam of the colon (e.g., colonoscopy) be given priority over fecal tests because they are better able to identify pre-cancerous
⁎ Corresponding author. Center for Chronic Disease Outcomes Research (CCDOR), Minneapolis Veterans Affairs Medical Center, One Veterans Drive (111-0), Minneapolis, MN 55417, USA. Fax: +1 612 725 2118. E-mail address:
[email protected] (A.A. Powell). 0091-7435/$ – see front matter. Published by Elsevier Inc. doi:10.1016/j.ypmed.2009.09.002
polyps. Many VA facilities are therefore working to create increased colonoscopy capacity (Powell et al., 2009). The primary goal of the current research is to quantify CRC screening mode preferences among a nationally representative sample of VA patients so that facilities are better able to project demand for colonoscopy. We also examine the relationship between preferences and personal (demographic, health, cognitive), environmental (social, medical care), and behavioral (past screening) factors derived from the theory of planned behavior (Ajzen, 1985) and social cognitive theory (Bandura, 2000). These can be used to enhance demand projections and provide insight into the processes by which preferences are derived. Methods Study population/sampling frame Male and female veterans, aged 50–75 years, who had one or more primary care visits at a VA Medical Center in the past 2 years, were included. VA employees, deceased patients, and anyone enrolled in VA adult day care or nursing home facilities, or diagnosed with CRC, dementia, or Alzheimer's were excluded. To derive the study sample, the VA's 124 medical centers were grouped into 12 strata according to the size of the eligible patient population and the proportion of African-American patients within the site
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
(see Fig. 1). Two sites were then randomly selected from each stratum (24 facilities) and a simple random sample of 156 individuals was selected from each sampled site (total sample = 3744). Data collection This analysis uses data from a cross-sectional survey designed to examine veteran attitudes, beliefs and behaviors regarding colorectal cancer screening. Survey data were collected between January 2005 and December 2006. Surveys were initially mailed with a cover letter, postage-paid return envelope, and a $2 cash incentive, which has been found to encourage participation (Beebe et al., 2005). A reminder postcard was mailed after one week and a second survey (with no incentive) was sent to non-responders within 3–4 weeks of the first mailing. A minimum of six attempts were made to administer the survey via phone for those who did not return the questionnaire within three weeks of the second survey mailing. The survey instrument is available at http://www.hsrd.minneapolis.med.va.gov/PDF/ SCREEN_NationalSurvey.pdf. The study protocol was approved by the Minneapolis VA Institutional Review Board. Measures The screening mode preference question was modeled after one used by Leard et al. (1997). Participants were asked to indicate “which colon cancer test would you most want to use if your doctor recommended you be tested for colon cancer?” Response options were as follows: FOBT (fecal occult blood test), sigmoidoscopy, colonoscopy, DCBE (barium enema), “I would not want to be tested,” and “don't know.” Participants who left this question blank (n = 66) were categorized as “don't know.” Participants were given brief descriptions of each CRC screening mode (Appendix A) developed for the National Cancer Institute's colorectal cancer screening questionnaire (Vernon et al., 2004). We used these brief descriptions because most patients are not
443
provided with detailed information on screening mode alternatives during CRC screening discussions with providers (Ling et al., 2008). Therefore, our approach to eliciting preferences is similar to the approach used during timeconstrained clinical encounters. Demographic variables included race, education level, and income, all obtained from the survey, and age, obtained from administrative data. Health factors included overall health (obtained from the survey), Charlson Comorbidity Index score, and diagnoses of substance abuse and psychiatric disorders (obtained from administrative data for 100% of the sample using electronic extraction algorithms). Behavioral factors included survey questions on whether the participant had ever completed an FOBT or endoscopy (either colonoscopy or sigmoidoscopy). Cognitive factors included CRC knowledge, determined by summing correct responses to two items drawn from prior literature (Manne et al., 2002; Rutten et al., 2007), one asking the appropriate age to initiate screening (correct = 50), and the other asking participants to rate their agreement with the statement that someone can have CRC without having symptoms (correct = “strongly agree” or “agree”). Additional cognitive factors included four scales developed by Vernon et al. (1997), namely Salience/Importance of Screening (4 items, α = .89), Susceptibility to CRC (4 items, α = .76), Perceived Efficacy of Screening (2 items, α = .65), and Screening Self-efficacy (4 items, α = .80), and two scales developed by the study team, namely Test Result Anxiety (2 items, α = .78) and Endoscopy Anxiety (5 items, α = .86). Social environmental factors included marital status and Social Influence (4 items, α = .69) (Vernon et al., 1997). Medical care support factors included survey-reported use of non-VA care and physician recommendation for FOBT and for colonoscopy in the past year. All scales were dichotomized prior to analysis by assigning the value 0 to all scores at or below the median and 1 to all scores above the median. Scale items are provided in Appendix B. Analysis A total of 3025 patients completed the survey (response rate = 81%). Of these, 961 patients were high risk (family history of CRC or a personal history of polyps, or inflammatory bowel disease) and were excluded because guidelines specify colonoscopy is the only appropriate testing mode for this subgroup. The remaining 2068 average-risk participants were included in all analyses. We report preferences overall and by subgroups. Estimates were weighted to account for oversampling and stratification in our sampling plan. Because few participants chose sigmoidoscopy or barium enema, we combined these two modes in subgroup analyses. χ2 Statistics were used to test for differences in preferences across subgroups. To identify independent correlates of the three most commonly chosen screening mode preference response categories, three logistic regression analyses were performed predicting: (1) preference for colonoscopy versus all other responses, (2) preference for FOBT versus all other responses, and (3) no expressed preference versus all other responses. Each model included a random effect for facility to account for the possible interdependence of patients within each site. In order to retain participants with missing data for one or more of the independent variables included in our multivariate models, we employed multiple imputation techniques to generate values for missing data (Little and Rubin, 2003). Each model was created ten times using the imputed data sets, and parameters were constructed using averages of the estimates derived from these imputed datasets. All analyses were run using SAS version 9.1 (SAS Institute Inc., Cary, NC).
Results
Fig. 1. Subject flow diagram: United States, 2005–2006.
As shown in Table 1, participants were predominately male (96%), aged 50 to 64 years (61%), and white (72%). Forty-six percent had no college education and 39% had an income of $20,000 or less. Participants indicated good overall health and the average Charlson Comorbidity Index was 1.9; however, 50% had a psychiatric diagnosis and 39% had a substance abuse diagnosis documented in their medical record. Seventy-two percent had completed an FOBT and 58% had completed an endoscopy in the past. As shown in Fig. 2, the most commonly chosen screening mode was colonoscopy (37%), followed by FOBT (29%). Most participants that did not chose one of these two modes indicated that they did not
444
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
Table 1 Characteristics of national sample of VA patients—percent distribution or mean scale scores (N = 2068): United States, 2005–2006. Characteristic
Statistic
Demographics Gender—male Age (years) 50–64 65–75 Race/ethnicity White Black Other Education High school or less Some college College graduate Income $20,000 or less $20,001 to $40,000 $40,001 or more Health factors Overall health (mean)a Charlson Comorbidity Index (mean) Had psychiatric diagnosis Had substance abuse diagnosis Cognitive factors CRC knowledge (mean)b Salience/importance (mean)c Susceptibility (mean)c Efficacy of screening (mean)c Test result anxiety (mean)c Endoscopy anxiety (mean)c Screening self-efficacy (mean)c Social environmental factors Married Screening social influence (mean)c Medical care Support support factors Receives all healthcare at VA Physician recommended FOBT Physician recommended colonoscopy Behavioral factors Ever completed FOBT Ever completed endoscopy
96.3 61.1 38.9 72.1 14.3 13.6 46.2 35.7 18.1 39.0 32.7 28.3 3.3 1.9 49.8 38.5 1.2 4.1 2.6 4.0 2.5 2.7 3.7 58.7 3.5 36.2 55.4 27.8 71.7 58.2
CRC = colorectal cancer. a Scale: 1 (poor) to 5 (excellent). b Number correct; range: 0 to 2. c Scale: 1 (low) to 5 (high).
know which mode they preferred or left the question blank (22%). Only 4% indicated that they did not want to be tested and few identified sigmoidoscopy and barium enema as their preferred mode. As shown in Table 2, subgroups that preferred colonoscopy were older (41.8% vs 34.9%), white (40.1% vs 32.1%), college graduates
(45.8% vs 35.4%), with incomes greater than $20,000 (27.5% vs 44.2%). Younger veterans were more likely than older participants to prefer FOBT (32.3% vs 24.1%). The largest differences favoring colonoscopy (N20%) were found for individuals who had an endoscopy in the past (52.2% vs 19.1%), received a recommendation for a colonoscopy from their physician (54.0% vs 31.8%), had high screening self-efficacy (51.5% vs 25.3%), had low endoscopy anxiety (49.5% vs 24.0%), and who considered CRC screening to be important (50.2% vs 26.6%). The largest differences favoring FOBT were found among participants who had not completed an endoscopy in the past (40.9% vs 20.1%), with high endoscopy anxiety (37.9% vs 21.7%), with a physician recommendation for FOBT (36.7% vs 20.6%), and without a recommendation for colonoscopy (33.0% vs 18.5%). Multivariate models predicting preferences are shown in Table 3. Participants with no college education and participants with an income of $20,000 to $40,000 were more likely to choose colonoscopy. Older participants were less likely to choose FOBT. Blacks were less likely to choose FOBT and more likely to have no expressed preference. Those in better health were more likely to prefer colonoscopy. High perceived importance of screening, high perceived susceptibility to CRC, and low endoscopy anxiety predicted preference for colonoscopy. Participants with high test result anxiety and low screening self-efficacy scores were less likely to prefer colonoscopy and more likely to have no preference. Only one cognitive factor (high endoscopy anxiety) was an independent predictor of preference for FOBT (OR 1.69, 95% CI 1.41–2.01). No social environmental factors were predictive of preferences. Participants who received a physician recommendation for a colonoscopy or who had used non-VA services were more likely to prefer colonoscopy and less likely to prefer FOBT or have no expressed preference. Participants who received a physician recommendation for FOBT were more likely to prefer FOBT. Participants who had previously completed an FOBT were more likely to prefer FOBT and less likely to have no preference. Similarly, those who had previously completed an endoscopy were both more likely to prefer colonoscopy and less likely to have no preference. Discussion This study examined the distribution of CRC screening mode preference in a large, nationally representative sample of VA patients and drew upon established theoretical frameworks to identify predictors of these preferences. Our finding that most participants indicated a preference for either colonoscopy (37%) or FOBT (29%) is consistent with other studies indicating a fairly even preference distribution between these two modes of screening (Debourcy et al., 2008; Hawley et al., 2008; Janz et al., 2007; Leard et al., 1997; Ling et al., 2001). Schroy et al. (2007) found that nearly three times as many participants preferred colonoscopy than FOBT. However, this study
Fig. 2. Colorectal cancer screening mode preference distribution (N = 2068): United States, 2005–2006.
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
445
Table 2 Distribution of colorectal cancer screening mode preferences by subgroups (N = 2068): United States, 2005–2006. Prefer colonoscopy Demographics Age (years) 50–64 34.9⁎⁎ 65–75 41.8⁎⁎ Race/ethnicity White 40.1⁎⁎ Black 31.8⁎⁎ Other 32.4⁎⁎ Education High school or less 35.4⁎⁎ Some college 35.5⁎⁎ College graduate 45.8⁎⁎ Income $20,000 or less 27.5⁎⁎ $20,001 to $40,000 41.6⁎⁎ $40,001 or more 47.1⁎⁎ Health factors Overall health Poor to average 30.5⁎⁎ Good to excellent 42.4⁎⁎ Charlson Comorbidity index score CCI = 0 41.3⁎⁎ CCI = 1 37.7⁎⁎ CCI b 1 34.5⁎⁎ Psychiatric diagnosis No 39.8⁎ Yes 35.1⁎ Substance abuse diagnosis No 41.9⁎⁎ Yes 30.6⁎⁎ Cognitive factors CRC knowledge 0 correct answers 31.2⁎⁎ 1 correct answer 34.6⁎⁎ 2 correct answers 44.1⁎⁎ Salience/importance Low 26.6⁎⁎ High 50.2⁎⁎ Susceptibility Low 34.5 High 40.0 Efficacy of screening Low 32.1⁎⁎ High 50.7⁎⁎ Test result anxiety Low 44.7⁎⁎ High 31.4⁎⁎ Endoscopy anxiety Low 49.5⁎⁎ High 24.0⁎⁎ Screening self-efficacy Low 25.3⁎⁎ High 51.5⁎⁎ Social environmental factors Marital status Single 31.8⁎⁎ Married 41.9⁎⁎ Screening social influence Low 29.2⁎⁎ High 47.6⁎⁎ Medical care support factors Healthcare system use Receives only VA care 25.3⁎⁎ Receives non-VA care 44.7⁎⁎ Physician recommendation of FOBT No 43.5⁎⁎ Yes 32.3⁎⁎ Physician recommendation of colonoscopy No 31.8⁎⁎ Yes 54.0⁎⁎ Behavioral factors Ever completed FOBT No 36.1 Yes 37.9
Prefer FOBT
Prefer other mode
No expressed preference
Don't want to be tested
32.3⁎⁎ 24.1⁎⁎
8.1 7.3
22.3 21.0
2.4⁎⁎ 5.8⁎⁎
29.5 27.6 31.6
7.1 9.6 8.7
19.7⁎⁎ 27.8⁎⁎ 23.5⁎⁎
3.7 3.2 3.8
29.8 30.2 27.0
6.2 8.3 10.5
24.2 22.3 15.4
4.5⁎ 3.8⁎⁎ 1.2⁎
32.8 29.1 24.7
10.0⁎ 5.2⁎ 7.5⁎
25.2⁎⁎ 20.9⁎⁎ 18.0⁎⁎
4.6⁎ 3.2⁎ 2.7⁎
32.1⁎ 27.4⁎
8.1 7.6
25.1⁎ 19.4⁎
4.2 3.2
27.4 32.4 29.1
7.1 6.9 8.8
20.9 20.3 23.3
3.3 2.7 4.3
27.8 31.0
7.9 7.8
20.7 22.8
3.9 3.3
26.8 33.3
7.2 8.7
20.6 23.7
3.5 3.8
28.0 29.4 30.1
8.3 9.0 5.9
27.8⁎⁎ 23.2⁎⁎ 17.1⁎⁎
4.7 3.8 2.9
33.6⁎⁎ 24.4⁎⁎
6.6 9.3
27.0⁎⁎ 15.6⁎⁎
6.2⁎⁎ 0.5⁎⁎
32.2⁎⁎ 26.9⁎⁎
7.8 7.8
19.9 23.6
5.6⁎⁎ 1.7⁎⁎
32.0⁎⁎ 22.8⁎⁎
7.7 8.2
24.4⁎⁎ 15.4⁎⁎
3.9 2.9
27.2 31.2
8.0 7.6
16.9⁎⁎ 25.8⁎⁎
3.0 4.1
21.7⁎⁎ 37.9⁎⁎
9.6⁎ 5.8⁎
16.6⁎⁎ 27.6⁎⁎
2.5⁎ 4.8⁎
33.8⁎⁎ 24.2⁎⁎
7.1 8.6
28.0⁎⁎ 14.5⁎⁎
5.7⁎⁎ 1.2⁎⁎
31.1 28.1
9.3⁎ 6.6⁎
24.6⁎⁎ 19.5⁎⁎
3.2 3.9
34.4⁎⁎ 23.2⁎⁎
6.8⁎ 9.0⁎
24.7⁎⁎ 18.2⁎⁎
4.9⁎⁎ 2.0⁎⁎
35.1⁎⁎ 25.9⁎⁎
7.7 7.9
27.0⁎⁎ 18.6⁎⁎
4.9⁎⁎ 2.8⁎⁎
20.6⁎⁎ 36.7⁎⁎
8.7 7.1
23.8 20.2
3.5 3.7
33.0⁎⁎ 18.5⁎⁎
8.1 7.0
23.9⁎⁎ 15.3⁎⁎
3.1 5.1
21.5⁎⁎ 32.9⁎⁎
7.5 7.9
28.6⁎⁎ 18.8⁎⁎
6.3⁎⁎ 2.4⁎⁎ (continued on next page)
446
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
Table 2 (continued)
Ever completed endoscopy No Yes
Prefer colonoscopy
Prefer FOBT
Prefer other mode
No expressed preference
Don't want to be tested
19.1⁎⁎ 52.2⁎⁎
40.9⁎⁎ 20.1⁎⁎
7.1 8.4
27.7⁎⁎ 17.0⁎⁎
5.2⁎⁎ 2.3⁎⁎
Significance tests conducted using χ2 tests of difference in the proportion of participants choosing the column response option. FOBT = fecal occult blood test, CCI = Charlson Comorbidity Index, VA = United States Veteran's Affairs Healthcare System. ⁎ p b .05. ⁎⁎ p b .01.
included stool DNA testing in the screening options. When preferences for all stool tests are combined, results are similar to our findings (46% preferring stool tests and 52% preferring colonoscopy). Three studies report a majority preferring FOBT. However, each of these studies consisted of very different samples than the US veterans surveyed in our research. Almog et al. (2008) surveyed Israeli HMO members, Wolf et al. (2006) surveyed members of a New York healthcare workers union, and Nelson and Schwartz (2004) surveyed a convenience sample of adults aged 18 to 54 years, most of which had no previous CRC screening experience. Physician recommendation and previous screening behavior explained the most variance in our multivariate models. Other studies have identified physician recommendation as an important predictor in CRC screening compliance (Bejes and Marvel, 1992; Lemon et al., 2001; Myers et al., 1990; Seeff et al., 2004; Wee et al., 2005; Zapka et al., 2002). One implication of these findings is that, as sites expand their offerings, preferences are likely to evolve. Within the VA, if sites that currently offer only FOBT introduce colonoscopy into their screening program, one could expect physician colono-
scopy recommendations and colonoscopy procedures to increase. This is likely to result in a higher proportion of veterans preferring colonoscopy and fewer preferring FOBT or having no preference (both of which were associated with not having received a colonoscopy recommendation). The finding that several cognitive factors were important independent correlates of preference, especially preference for colonoscopy, is consistent with other studies indicating that colonoscopy tends to be preferred among those who place a relatively high value on test accuracy (Hawley et al., 2008; Marshall et al., 2007; Wolf et al., 2006; Ling et al., 2001) and are less concerned with discomfort associated with the test (Janz et al., 2007; Ling et al., 2001). Only one cognitive variable was a significant independent predictor of preference for FOBT (high endoscopy anxiety). Cognitive variables may be more relevant to the choice of colonoscopy than FOBT because of the greater involvement and commitment associated with preparing for and undergoing a colonoscopy. We found that preference for colonoscopy was higher among veterans that are white, higher income, college graduates, married, in better
Table 3 Multivariate odds ratios (OR) and 95% confidence intervals (CI) from hierarchical logistic regressions predicting preference for colonoscopy and preference for FOBT (N = 2068): United States, 2005–2006.
Demographics Aged 65+ Race black (vs. white) Race other (vs. white) Some college (vs. high school or less) College graduate (vs. high school or less) Income $20,000 to $40,000 (vs $20,000 or less) Income $40,000 or more (vs $20,000 or less) Health factors Overall health good to excellent CCI = 0 (vs CCI = 2) CCI = 1 (vs CCI = 2) Had psychiatric diagnosis Had substance abuse diagnosis Cognitive factors Knowledge moderate (vs low) Knowledge high (vs low) High salience/importance High susceptibility High efficacy of screening High test result anxiety High endoscopy anxiety High screening self-efficacy Social environmental factors Married High screening social influence Medical care support factors Received non-VA care Dr recommended FOBT Dr recommended colonoscopy Behavioral factors Ever completed FOBT Ever completed endoscopy
Prefer colonoscopy vs. Any any other response, OR (95% CI)
Prefer FOBT vs. any other response, OR (95% CI)
No expressed preference vs. any other response, OR (95% CI)
0.89 (0.74–1.07) 0.83 (0.63–1.10) 0.77 (0.51–1.16) 0.79 (0.64–0.99) 1.01 (0.68–1.52) ) 1.46 (1.11–1.91) 1.29 (0.97–1.70)
0.80 (0.65–0.98) 0.80 (0.68–0.94) 1.03 (0.72–1.48) 1.14 (0.71–1.82) 0.92 (0.54–1.55) 0.90 (0.66–1.23) 0.77 (0.57–1.06)
1.15 (0.79–1.68) 1.48 (1.03–2.11) 1.22 (0.96–1.55) 0.80 (0.51–1.24) 0.98 (0.57–1.68) 1.20 (0.83–1.75) 1.06 (0.89–1.26)
1.30 1.25 1.16 1.00 1.00
0.92 0.99 1.13 0.97 0.98
0.87 0.89 0.83 0.97 1.12
(1.01–1.68) (0.87–1.80) (0.86–1.57) (0.69–1.43) (0.69–1.45)
(0.76–1.12) (0.69–1.41) (0.70–1.81) (0.69–1.36) (0.71–1.36)
(0.72–1.04) (0.52–1.51) (0.58–1.20) (0.66–1.42) (0.93–1.34)
1.01 (0.67–1.52) 1.30 (0.94–1.80) 1.48 (1.07–2.04) 1.30 (1.13–1.50) 1.15 (0.93–1.41) 0.77 (0.61–0.96) 0.57 (0.42–0.77) 1.53 (1.21–1.94)
1.26 (0.80–1.98) 1.35 (0.99–1.86) 0.94 (0.78–1.13) 0.81 (0.66–1.00) 0.88 (0.68–1.14) 0.92 (0.69–1.21) 1.69 (1.41–2.01) 1.06 (0.77–1.45)
0.82 (0.55–1.22) 0.68 (0.47–1.00) 0.78 (0.61–1.00) 1.22 (0.89–1.67) 0.83 (0.63–1.11) 1.35 (1.00–1.82) 1.24 (0.77–1.98) 0.64 (0.47–0.88)
1.05 (0.78–1.41) 1.38 (0.95–2.00)
1.06 (0.87–1.29) 0.74 (0.54–1.02)
0.89 (0.65–1.23) 0.89 (0.71–1.11)
2.10 (1.59–2.76) 0.74 (0.52–1.07) 2.32 (1.57–3.43)
0.74 (0.56–0.97) 1.76 (1.30–2.38) 0.53 (0.36–0.78)
0.68 (0.50–0.93) 0.79 (0.43–1.45) 0.65 (0.48–0.86)
1.05 (0.66–1.67) 2.88 (2.06–4.02)
1.70 (1.26–2.29) 0.49 (0.38–0.64)
0.64 (0.48–0.86) 0.73 (0.52–1.03)
Bold represents statistically significant results, p b 0.05. FOBT = Fecal Occult Blood Test, CCI = Charlson Comorbidity Index, VA = United States Veteran's Affairs Healthcare System.
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
health, and users of non-VA healthcare. Debourcy et al. (2008) found that variables reflecting higher socioeconomic status (SES) were associated with greater preference for colonoscopy and less preference for FOBT. In our study, there was little difference in the appeal of FOBT across SES subgroups. Instead, subgroup differences in preference for colonoscopy tended to be offset by corresponding differences in the percent indicating no preference. One possible explanation for this is that both high and low SES subgroups have enough of an understanding of FOBT to form an opinion about this mode, but low SES groups have had less exposure to colonoscopy information. This is consistent with the fact that many veterans without access to non-VA services have only been offered FOBT. If provided with information about and the opportunity to undergo screening colonoscopy, lower SES subgroups may express preferences more similar to their higher SES counterparts. Those without access to non-VA healthcare resources are also most likely to benefit from expansion of VA screening programs to include colonoscopy. Limitations We provided patients with less detail on the screening mode options than provided in many other studies. A fully informed decision may require an assessment of each mode's sensitivity, specificity, risk profile, convenience, recommended frequency, and out-of-pocket costs. Previous research on CRC screening mode preferences indicates that information about out-of-pocket costs (Griffith et al., 2008) can affect preferences. However, VA patients do not incur co-payments for any CRC screening modes (U.S. Department of Veteran Affairs, 2008). The fact that 22% of our sample expressed no preference suggests some patients may need more information before making a choice. However, our work may better reflect patients' understandings of their options when faced with real CRC screening decisions in a busy clinic. This study only offered participants screening mode options that were guideline endorsed at the time of data collection (U.S. Preventive Services Task Force, 2002). Newly published guidelines also include CT colonography and fecal DNA testing in their options (Levin et al., 2008). These screening modes have a unique configuration of costs and benefits that appeal to some patients (Hawley et al., 2008; Schroy et al., 2007). Currently, however, these tests are not widely available within the VA. A final limitation of this research is that because our sample included only VA patients, findings may not generalize to other populations. Many participants in our study had previous experience with FOBT and/ or endoscopy. One might expect different preference distributions among populations with different prior screening experiences. Conclusion Screening programs that offer only one mode fail to accommodate the preferences of a substantial proportion of patients. Within the VA, adding screening colonoscopy to programs currently offering only FOBT is likely to increase preferences for colonoscopy as patients incorporate provider recommendations for and personal experience with colonoscopy into their preferences. This may disproportionately benefit lower socioeconomic groups who do not currently have access to non-VA colonoscopy services. Conflict of interest statement The authors have no conflicts of interest regarding the submission and publication of the manuscript.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.ypmed.2009.09.002.
447
References Ajzen, J., 1985. From intentions to actions: a theory of planned behavior. In: Kuhl, J., Beckmann, J. (Eds.), Action-control: From Cognition to Behavior. Springer, New York, pp. 11–39. Beebe, T.J., Davern, M.E., McAlpine, D.D., Call, K.T., Rockwood, T.H., 2005. Increasing response rates in a survey of Medicaid enrollees: the effect of a prepaid monetary incentive and mixed modes (mail and telephone). Med. Care 43, 411–414. Almog, R., Ezra, G., Lavi, I., Rennert, G., Hagoel, L., 2008. The public prefers fecal occult blood test over colonoscopy for colorectal cancer screening. Eur. J. Cancer Prev. 17, 430–437. Bandura, A., 2000. Health promotion from the perspective of social cognitive theory. In: Abraham, C., Norman, P., Connor, M. (Eds.), Understanding and Changing Health Behavior: From Health Beliefs to Self-regulation. Harwood Academic Publishers, Amsterdam, Netherlands, pp. 299–342. Bejes, C., Marvel, M.K., 1992. Attempting the improbable: offering colorectal cancer screening to all appropriate patients. Fam. Pract. Res. J. 12, 83–90. Debourcy, A.C., Lichtenberger, S., Felton, S., Butterfield, K.T., Ahnen, D.J., Denberg, T.D., 2008. Community-based preferences for stool cards versus colonoscopy in colorectal cancer screening. J. Gen. Intern. Med. 23, 169–174. El-Serag, H.B., Petersen, L., Hampel, H., Richardson, P., Cooper, G., 2006. The use of screening colonoscopy for patients cared for by the Department of Veterans Affairs. Arch. Intern. Med. 166, 2202–2208. Griffith, J.M., Lewis, C., Brenner, A.R.T., Pignone, M., 2008. The effect of offering different numbers of colorectal cancer screening test options in a decision aid: a pilot randomized trial. BMC Med. Inform. Decis. Mak. 8, 10. Hardcastle, J.D., Chamberlain, J.O., Robinson, M.H., et al., 1996. Randomised controlled trial of faecal-occult-blood screening for colorectal cancer. Lancet 348, 1472–1477. Hawley, S.T., Volk, R.J., Krishnamurthy, P., Jibaja-Weiss, M., Vernon, S.W., Kneuper, S., 2008. Preferences for colorectal cancer screening among racially/ethnically diverse primary care patients. Med. Care 46, S10–S16. Janz, N.K., Lakhani, I., Vijan, S., Hawley, S.T., Chung, L.K., Katz, S.J., 2007. Determinants of colorectal cancer screening use, attempts, and nonuse. Prev. Med. 44, 452–458. Kewenter, J., Brevinge, H., Engaras, B., et al., 1994. Follow-up after screening for colorectal neoplasms with fecal occult blood testing in a controlled trial. Dis. Colon Rectum 37, 115–119. Kronborg, O., Fenger, C., Olsen, J., et al., 1996. Randomised study of screening for colorectal cancer with faecal-occult-blood test. Lancet 348, 1467–1471. Leard, L.E., Savides, T.J., Ganiats, T.G., 1997. Patient preferences for colorectal cancer screening. J. Fam. Pract. 45, 211–218. Lemon, S., Zapka, J., Puleo, E., et al., 2001. Colorectal cancer screening participation: comparisons with mammography and prostate-specific antigen screening. Am. J. Public Health 91, 1264–1272. Levin, B., Lieberman, D.A., McFarland, B., et al., 2008. Screening and surveillance for the early detection of colorectal cancer and adenomatous polyps, 2008: a joint guideline from the American Cancer Society, the US Multi-Society Task Force on Colorectal Cancer, and the American College of Radiology. Ca. Cancer J. Clin. 59, 1–31. Ling, B.S., Moskowitz, M.A., Wachs, D., Pearson, B., Schroy, P.C., 2001. Attitudes toward colorectal cancer screening tests: a survey of patients and physicians. J. Gen. Intern. Med. 16, 822–830. Ling, B.S., Trauth, J.M., Fine, M.J., et al., 2008. Informed decision-making and colorectal cancer screening: is it occurring in primary care? Med. Care 46, S23–S29. Little, R.J.A., Rubin, D.B., 2003. Statistical Analysis with Missing Data. John Wiley, New York. Mandel, J.S., Bond, J.H., Church, T.R., et al., 1993. Reducing mortality from colorectal cancer by screening for fecal occult blood. N. Engl. J. Med. 328, 1365–1371. Manne, S., Markowitz, A., Winawer, S., et al., 2002. Correlates of colorectal cancer screening compliance and stage of adoption among siblings of individuals with early onset colorectal cancer. Health Psychol. 21, 3–15. Marshall, D.A., Johnson, F.R., Phillips, K.A., et al., 2007. Measuring patient preferences for colorectal cancer screening using a choice-format survey. Value Health 5, 415–430. Myers, R.E., Trock, B.J., Lerman, C., Wolf, T., Ross, E., Engstrom, P.F., 1990. Adherence to colorectal cancer screening in an HMO population. Prev. Med. 19, 502–514. National Cancer Institute, 2006. Cancer Fast Facts Sheet: Cancer of the Colon and Rectum, Bethesda, MD. Nelson, R.L., Schwartz, A., 2004. A survey of individual preferences for colo-rectal cancer screening technique. BMC Cancer 4, 76. Perlin, J.B., 2005. IL 10-2005-009: Under Secretary for Health's Information Letter: Colorectal Cancer Screening. Available at www1.va.gov/vhapublications/ ViewPublication.asp?pub_ID=1263. Accessed April 7, 2009. Powell, A.A., Gravely, A.A., Ordin, D.L ., Schlosser, J.E., Partin, M.R., 2009. Timely followup of positive fecal occult blood tests: strategies associated with improvement. Am. J. Prev. Med. 37, 87–93. Rutten, L.F., Moser, R.P., Beckjord, E.B., Hesse, B.W., Croyle, R.T., 2007. Cancer communication: Health Information National Trends Survey. NIH Pub. No. 076214. National Cancer Institute, Washington, D.C. Schroy 3rd, P.C., Lal, S., Glick, J.T., Robinson, P.A., Zamor, P., Heeren, T.C., 2007. Patient preferences for colorectal cancer screening: how does stool DNA testing fare? Am. J. Manag. Care 13, 393–400. Seeff, L.C., Nadel, M.R., Klabunde, C.N., et al., 2004. Patterns and predictors of colorectal cancer test use in the adult U.S. population. Cancer 100, 2093–2103. U.S. Department of Veteran Affairs, 2008. VA health care eligibility & enrollment. Available at: http://www.va.gov/healtheligibility/costs/. Accessed June 24, 2009. U.S. Preventive Services Task Force, 2002. Colorectal cancer screening guidelines. Rockville, MD, Agency for Healthcare Research and Quality.
448
A.A. Powell et al. / Preventive Medicine 49 (2009) 442–448
Vernon, S.W., Myers, R.E., Tilley, B.C., 1997. Development and validation of an instrument to measure factors related to colorectal cancer screening adherence. Cancer Epidemiol. Biomarkers Prev. 6, 825–832. Vernon, S.W., Meissner, H., Klabunde, C., et al., 2004. Measures for ascertaining use of colorectal cancer screening in behavioral, health services, and epidemiologic research. Cancer Epidemiol. Biomarkers Prev., 13, 898–905. Wee, C.C., McCarthy, E.P., Phillips, R.S., 2005. Factors associated with colon cancer
screening: the role of patient factors and physician counseling. Prev. Med. 41, 23–29. Wolf, R.L., Basch, C.E., Brouse, C.H., Shmukler, C., Shea, S., 2006. Patient preferences and adherence to colorectal cancer screening in an urban population. Am. J. Public Health 96, 809–811. Zapka, J.G., Puleo, E., Vickers-Lahti, M., Luckmann, R., 2002. Healthcare system factors and colorectal cancer screening. Am. J. Prev. Med. 23, 28–35.
