Abstract
One of the largest disparities in cancer mortality in the United States occurs with colorectal cancer (CRC). The objectives of this multilevel two-arm intervention trial were to compare the efficacy of two interventions to promote CRC screening (CRCS) with fecal immunochemical test (FIT) and examine sociodemographic and psychosocial predictors of FIT screening. Individuals ages 50–75 (n=326) who were not up-to-date with CRCS, could understand English or Spanish, and at average CRC risk were recruited from two federally qualified health centers (FQHCs) in Florida. Prior to intervention, CRCS rates in the FQHCs were 27.1% and 32.9%, respectively. Study enrollment occurred April 2018-November 2019. System-level intervention components included leveraging electronic medical record (EMR) systems and delivering patient reminders. Participants were randomized to C-CARES (education+FIT) or C-CARES Plus (C-CARES+personalized coaching [for those not completing FIT within 90 days]). Primary outcome was completed FIT returned <1 year. Primary outcome analyses were performed using logistic regression. 225 participants completed FIT (69.0% [95% CI: 64.0–74.0%]), with no significant difference in FIT uptake by intervention (67.3% C-CARES Plus vs. 70.8% C-CARES; p=.49). FIT uptake was significantly higher among patients who received intervention materials in Spanish (77.2%) compared to those who received materials in English (63.2%, p<.01). The personalized coaching in the C-CARES Plus arm did not appear to provide added benefit beyond the C-CARES intervention. Multilevel approaches that include EMR prompts, reminders, FIT access, and provision of low-literacy, language-concordant education can support efforts to improved community clinics’ CRCS rates. Future efforts should focus on repeat FIT screening.
Trial registration: The trial was registered at ClinicalTrials.gov (NCT03906110).
Keywords: colorectal cancer screening, randomized controlled trial, multilevel intervention, fecal immunochemical test, federally qualified health center, low literacy English and Spanish language patient education
In 2022, it is estimated that there will be 151,030 new colorectal cancer (CRC) cases diagnosed in the United States, resulting in 52,580 United States deaths.1 Only 67% of individuals ages 50–75 were up-to-date with CRC screening (CRCS) in 20182, a rate below the Healthy People 2020 goal of 70.5% and the new Healthy People 2030 goal of 74.5%.3, 4 Many factors are known to negatively influence CRC screening such as: limited electronic medical record (EMR) capabilities for facilitating scheduling, prompting, and tracking CRC screening, a lack of reminders, and a paucity of language/literacy-specific CRC educational materials available (system-level factors) as well as financial/cost concerns, fear, cancer worry and low awareness about cancer screening benefit (patient-level factors).5–7 Informed by the Preventive Health Model (PHM),8–11 this study titled Community Colorectal Cancer Awareness, Research, Education and Screening (C-CARES) examined the efficacy of a multilevel, low literacy language-specific intervention to improve screening rates in two federally qualified health center (FQHC) systems. The aims were to: 1) compare the efficacy of the C-CARES intervention C-CARES (education + FIT) or C-CARES Plus (C-CARES + personalized coaching [the latter component, for those who had not returned a FIT by 90 days]) to promote FIT uptake following intervention enrollment, and 2) examine sociodemographic and psychosocial predictors of FIT uptake. We hypothesized that C-CARES Plus would have higher FIT return rates than C-CARES up to one-year post-intervention.
METHODS
Study Overview and Setting
Our study utilized community-based participatory research approaches,12, 13 with study activities informed by a community advisory board (CAB). Study implementation occurred in two large FQHC systems in the Tampa Bay area of Florida, United States that had pre-intervention (2017) CRC screening rates of 27.1% and 32.9%, respectively (per their reported Uniform Data System [UDS] Clinical Quality Measures). The parent study was a two-arm randomized controlled trial aimed to improve initial CRCS and subsequent repeat screening (1 and 2 years post-enrollment) with targeting of non-responders to encourage FIT return.14 The sample size calculation was based upon comparing FIT use between the two study arms (C-CARES Plus vs. C-CARES) at 27 months post-intervention. Based on published studies and our team’s previous studies,15–17 initial FIT uptake rates at 3–6 months post-intervention were estimated to be approximately 60% and 80% for the C-CARES arm and the C-CARES Plus intervention, respectively. The two participating FQHCs serve a large number of English- and Spanish-speaking patients annually who are aged 50–75 years. Among participants ages 50–75 years old who receive care at the clinics at least once a year, we estimated conservatively that 35% would not be up-to-date with CRC screening and would agree to enroll in our study. We then assumed FIT completion in the C-CARES and C- CARES Plus arm would be 20% and 40% at 27 months after randomization (Time 3); a sample size of 85 per arm achieved 80% power with alpha=.05 and a two-sided test using logistic regression with potential confounds sharing 1% variance with study arm. Based on these estimates, we planned to enroll 328 individuals (164 per arm) in the study in order to achieve our final sample size goal of 170 (85 per arm) at 27 months after randomization (Time 3), anticipating that there would be some attrition over the course of the study period due to positive FIT results, loss to follow-up, and withdrawal, among other reasons. The statistical power estimate to compare FIT return differences for the two study conditions at Time 1 (the focus of the current study which included enrollment, baseline interview, randomization, delivery of study components following enrollment and up to 1-year post-enrollment, and the time window for initial FIT return [up to 1-year post enrollment]) with 328 participants (164 per arm) was 97%. Herein, we report intervention components received up to 1-year post-enrollment and findings related to initial FIT screening (up to 1-year post-intervention) (Time 1).
Study Participants
Eligible participants were ages 50–75 years, not up-to-date with CRCS recommendations (e.g., no stool blood test completed in prior year, no colonoscopy completed in prior 10 years, no virtual colonography completed in prior 5 years, no flexible sigmoidoscopy completed in prior 5 years), established patients of the FQHCs, without a personal history of CRC or colon polyps, without symptoms presumptive of CRC (e.g., lower gastrointestinal [GI] bleeding, persistent diarrhea, bloating, or lower GI pain),2 not at increased risk due to family history or personal high-risk gastrointestinal syndromes,2 could understand, read, and speak English or Spanish, and had not completed a CRC screening research study in the prior year.
Procedures
Potential participants were approached in clinic by bilingual research coordinators immediately prior to or following a scheduled visit. Individuals with a scheduled visit indicative of gastrointestinal issues, those with up-to-date CRC screening recorded in their medical chart, and those who had walk-in (acute care) appointments were not approached. Interested individuals were assessed for study eligibility via self-report by the research coordinator. In some instances where individuals were unsure of their CRC screening status, the research coordinator would confirm eligibility via EMR verification by provider/clinic staff. Eligible and interested individuals completed informed consent. Randomization allocation sequence and data management of the study was conducted using a HIPAA-complaint web-based application developed by Moffitt Cancer Center. Consented participants were randomized with 1:1 allocation following stratification by FQHC system and gender (4 strata) and using blocks of size 6 within each stratum. Next, a baseline interview was conducted in person by the research coordinator. Participants received a $15 gift card following the baseline interview. Immediately, following the baseline interview, research coordinators provided the intervention materials to the participant while they were at the clinic based upon the randomization allocation and in the language in which informed consent was performed (i.e., English or Spanish). Thus, participant recruitment, baseline interview, and intervention delivery (viewing the DVD and provision of the photonovella booklet) occurred in person during the course of the individual’s clinic visit. Study enrollment occurred April 2018-November 2019.
System-level components for both intervention arms included system prioritization of CRC screening (including provider education kick-off meeting), forming of a CAB which included members from both FQHC systems, enhanced EMR CRCS tracking and provider prompts, and patient reminders via the systems’ EMR or mail. Patient level components in both arms included Spanish- or English-language education plus FIT distribution. Specific to the CARES-Plus arm, a coaching component was added for participants who did not return their FIT kit within 90 days. Completed FIT kits were returned to the clinic in person by the patient as per clinic procedure; return date and clinical results were entered into the EMR. The primary outcome was collected via manual chart review and electronic query of the EMR by clinic personnel. Study procedures were HIPAA-compliant, approved by the Institutional Review Board of record, and conducted in accordance with the Declaration of Helsinki. The trial was registered at ClinicalTrials.gov (NCT03906110).
Intervention Components
C-CARES intervention.
Patient-level components included receipt of FIT kit and education. Participants viewed an educational DVD (in either English or Spanish) on a portable DVD player, which lasted approximately 10 minutes and received a low literacy print photonovella (in either English or Spanish) informed by PHM.15, 16 The participant viewed the DVD with the research coordinator while in clinic, and were provided with the photonovella and DVD to take home. The material development process has been described previously (i.e., extensive formative research to ensure cultural and literacy salience).15, 16 Next, participants received in person, one-on-one education from a research coordinator which included an explanation of how to complete FIT screening. Participants were encouraged to re-review the education materials at home and return a completed FIT. Individuals who did not return their FIT within approximately four weeks were sent reminders. An English/Spanish-language postcard with a CRC message was sent 90 days post-enrollment to those who had not yet returned a FIT.
C-CARES Plus intervention.
Participants received the same components as C-CARES (FIT + education) with the addition of personalized coaching via phone by a trained research coordinator for those not completing a FIT (non-responders) within 90 days. During this structured telephone interaction, participants were guided in a discussion about any obstacles and strategies to FIT completion. Common points addressed during the coaching sessions included misplaced FIT kit or forgetting to complete the FIT. Strategies were suggested by the research coordinator to overcome these barriers (e.g., picking up a new FIT kit from the clinic or placing the FIT kit somewhere visible in the bathroom to serve as a reminder to complete the FIT). In addition, the coaching component frequently included additional teaching such as reviewing the steps for FIT sample collection and completion, specifying a timeline to drop off a completed FIT at the clinic, and suggesting that the participants review the educational materials again (i.e., photonovella and DVD).
Measures
FIT uptake.
FIT uptake was determined by return of a completed FIT to the clinic within 1 year of enrollment.
PHM variables.
PHM variables were collected with seven subscales, including salience and coherence (4 items), perceived susceptibility (3 items), cancer worry (2 items), response efficacy (2 items), self-efficacy (6 items), social influence (4 items), and religious beliefs (5 items).8–11
CRC awareness/knowledge.
CRC awareness/knowledge was assessed with nine items.18 A total score was calculated by summing the items answered correctly.
Health literacy.
Three single self-reported items assessed health literacy, including: 1) the frequency with which someone needs to get help with written materials from doctor, 2) one’s confidence in filling out forms without help, and 3) self-rated ability to read.19, 20
Numeracy.
Numeracy was self-reported and assessed with a single adapted item21 that asked participants to rate their ability to understand numbers, charts, or tables on a five-point scale from very good to not very good.
Sociodemographics.
Sociodemographic characteristics were self-reported and included age, gender, race, ethnicity, foreign-born status, years in the United States (if foreign-born), nativity of parents (born in the United States vs. foreign-born), marital status, education, insurance status, type of insurance, employment status, annual household income, income stability,22 and current living situation.22
Statistical Analysis
Sociodemographic and psychosocial variables were summarized using descriptive statistics. Intervention group differences were examined via chi-square test or t-test. Those with p-values<.10 were reviewed as potential covariates in primary analyses. Logistic regression was used to evaluate the effect of the intervention on FIT uptake, with alpha=.05. Post-hoc exploratory analyses evaluated sociodemographic, health-related, and study variables as predictors of FIT uptake using univariate logistic regression. Descriptive statistics were used to summarize days for FIT return and coaching (the latter component, for the C-CARES Plus arm only). Statistical analyses were conducted using the SAS statistical software package (version 9.4; SAS Institute Inc, Cary, NC). Statistical analyses were conducted in 2020 and 2021.
RESULTS
Participant characteristics are displayed in Tables 1 and 2. Most were female (59.2%), Hispanic/Latino ethnicity (53.1%), born in the United States (51.5%), and had an annual income less than $25,000 (74.9%). Among those marking ‘White’ for race, 19.7% marked Hispanic ethnicity. Among those marking ‘Other’ for race, 74.6% marked Hispanic ethnicity. By chance, C-CARES Plus participants had a significantly higher proportion self-reporting difficulty understanding numbers, charts or tables (numeracy) (p=.02) and a marginally significant trend for greater years of schooling (p=.08). Fifty-eight percent of the intervention materials provided were in English, with the remaining 42% provided in Spanish (p=.19).
Table 1.
Sociodemographics characteristics
| Variable | Level | Overall N=326 n (%) | C-CARES Plus N=165 n (%) | C-CARES N=161 n (%) | p-value |
|---|---|---|---|---|---|
| FQHC Clinic | FQHC 1, Site 1 | 80 (24.5) | 40 (24.2) | 40 (24.8) | 0.77 |
| FQHC 1, Site 2 | 56 (17.2) | 31 (18.8) | 25 (15.5) | ||
| FQHC 2, Site 1 | 124 (38.0) | 59 (35.8) | 65 (40.4) | ||
| FQHC 2, Site 2 | 66 (20.2) | 35 (21.2) | 31 (19.3) | ||
| Gender | Male | 132 (40.5) | 62 (37.6) | 70 (43.5) | 0.36 |
| Female | 193 (59.2) | 102 (61.8) | 91 (56.5) | ||
| Other | 1 (0.3) | 1 (0.6) | 0 (0.0) | ||
| Married/partnered | No | 178 (54.8) | 97 (58.8) | 81 (50.6) | 0.14 |
| Yes | 147 (45.2) | 68 (41.2) | 79 (49.4) | ||
| Employed (full- or part-time) | No | 188 (57.7) | 101 (61.2) | 87 (54.0) | 0.19 |
| Yes | 138 (42.3) | 64 (38.8) | 74 (46.0) | ||
| Racea | American Indian or Alaskan Native | 3 (0.9) | 1 (0.6) | 2 (1.2) | 0.52 |
| Asian | 2 (0.6) | 2 (1.2) | 0 (0.0) | ||
| Black/African American | 18 (5.5) | 13 (7.9) | 5 (3.1) | ||
| Native Hawaiian or Pacific Islander | 2 (0.6) | 0 (0.0) | 2 (1.2) | ||
| White | 152 (46.8) | 74 (45.1) | 78 (42.9) | ||
| Other | 133 (40.9) | 65 (39.6) | 68 (42.2) | ||
| More than one race | 15 (4.6) | 9 (5.5) | 6 (3.7) | ||
| Hispanic ethnicity | No | 153 (46.9) | 84 (50.9) | 69 (42.9) | 0.15 |
| Yes | 173 (53.1) | 81 (49.1) | 92 (57.1) | ||
| Education | Less than HS diploma | 133 (40.9) | 60 (36.6) | 73 (45.3) | 0.08 |
| HS diploma | 84 (25.8) | 40 (24.4) | 44 (27.3) | ||
| Beyond HS diploma | 108 (33.2) | 64 (39.0) | 44 (27.3) | ||
| Born in the United States | No | 158 (48.5) | 77 (46. 7) | 81 (50.3) | 0.51 |
| Yes | 168 (51.5) | 88 (53.3) | 80 (49.7) | ||
| Annual household incomeb | Less than $10K | 99 (30.4) | 53 (32.1) | 46 (28.6) | 0.71 |
| $10K–$25K | 145 (44.5) | 73 (44.2) | 72 (44.7) | ||
| $25K–$35K | 46 (14.1) | 21 (12.7) | 25 (15.5) | ||
| $35K–$50K | 21 (6.4) | 9 (5.4) | 12 (7.5) | ||
| $50K–$75K | 6 (1.8) | 2 (1.2) | 4 (2.5) | ||
| $75K–$100K | 3 (0.9) | 2 (1.2) | 1 (0.6) | ||
| Do not know | 5 (1.5) | 5 (3.0) | 0 (0) | ||
| Prefer not to answer | 1 (0.3) | 0 (0) | 1 (0.6) | ||
| Household income stability | Very unstable | 41 (12.6) | 24 (14.6) | 17 (10.6) | 0.28 |
| Moderately unstable | 24 (7.4) | 11 (6.7) | 13 (8.1) | ||
| Slightly unstable | 38 (11.7) | 21 (12.7) | 17 (10.6) | ||
| Slightly stable | 34 (10.4) | 18 (10.9) | 16 (9.9) | ||
| Moderately stable | 77 (23.6) | 44 (26.7) | 33 (20.5) | ||
| Very stable | 112 (34.4) | 47 (28.5) | 65 (40.4) | ||
| Current living situationc | Own home | 134 (41.1) | 70 (42.4) | 64 (39.8) | 0.66 |
| Live with family | 74 (22.7) | 40 (24.2) | 34 (21.1) | ||
| Rent apartment | 103 (31.6) | 49 (29.7) | 54 (33.5) | ||
| Rent room | 11 (3.4) | 4 (2.4) | 7 (4.3) | ||
| Lack usual place | 2 (0.6) | 0 (0) | 2 (1.2) | ||
| Homeless | 2 (0.6) | 2 (1.21) | 0 (0.0) | ||
| Medical insurance | No | 201 (61.7) | 100 (60.6) | 101 (62.7) | 0.69 |
| Yes | 125 (38.3) | 65 (39.4) | 60 (37.3) | ||
| Type of medical insurance | Medicare | 14 (11.2) | 7 (10.8) | 7 (11.67) | 0.43 |
| Medicaid | 35 (28.0) | 19 (29.2) | 16 (26.67) | ||
| Private | 37 (29.6) | 21 (32.3) | 16 (26.67) | ||
| ACA | 14 (11.2) | 9 (13.9) | 5 (8.33) | ||
| Other | 25 (20.0) | 9 (13.9) | 16 (26.67) | ||
| Primary language spoken at home | English | 173 (53.7) | 93 (56.7) | 80 (50.6) | 0.44 |
| Spanish | 146 (45.3) | 69 (42.1) | 77 (48.7) | ||
| Other | 3 (0.9) | 2 (1.2) | 1 (0.6) | ||
| Preferred language for health information | English | 189 (58.0) | 101 (61.2) | 88 (54.7) | 0.28 |
| Spanish | 136 (41.7) | 63 (38.2) | 73 (45.3) | ||
| Parents born outside of United States | Other | 1 (0.3) | 1 (0.6) | 0 (0.0) | 1.00 |
| No | 152 (46.6) | 77 (46.7) | 75 (46.6) | ||
| Yes | 172 (52.8) | 87 (52.7) | 85 (52.8) |
Note: FQHC = federally qualified health center
The p-value reported is from an analysis of White versus all other races listed.
The p-value reported is from an analysis excluding ‘prefer not to answer’ and combining the three highest income groups into a single group due to low frequency counts.
The p-value reported is from an analysis excluding ‘Lack usual place’ and ‘Homeless’ due to low frequency counts.
Table 2.
CRC knowledge, health beliefs, literacy, and numeracy characteristics
| Variable | Level | Overall N=326 n (%) or M (SD) | C-CARES Plus N=165 n (%) or M (SD) | C-CARES N=161 n (%) or M (SD) | p-value |
|---|---|---|---|---|---|
| Intervention language | Spanish | 136 (31.7) | 63 (38.2) | 73 (45.4) | 0.19 |
| English | 190 (58.3) | 102 (61.8) | 88 (54.7) | ||
| Self-reported get help with written materials from doctor (… of the time) | None | 158 (48.5) | 77 (46.7) | 81 (50.3) | 0.21 |
| A little | 66 (20.2) | 39 (23.6) | 27 (16.8) | ||
| Some | 68 (20.9) | 37 (22.4) | 31 (19.3) | ||
| Most | 18 (5.5) | 7 (4.2) | 11 (6.8) | ||
| All | 16 (4.8) | 5 (3.0) | 11 (6.8) | ||
| Self-related confidence in filling out forms without help | Very | 174 (55.9) | 85 (54.1) | 89 (57.8) | 0.90 |
| Quite a bit | 44 (14.1) | 23 (14.7) | 21 (13.6) | ||
| Somewhat | 46 (14.8) | 24 (15.3) | 22 (14.3) | ||
| A little bit | 36 (11.6) | 18 (11.5) | 18 (11.7) | ||
| Not at all | 11 (3.5) | 7 (4.5) | 4 (2.6) | ||
| Self-rated ability to read | Very good | 11 (3.5) | 4 (2.6) | 7 (4.6) | 0.47 |
| Good | 41 (13.2) | 22 (14.0) | 19 (12.3) | ||
| OK | 41 (13.2) | 17 (10.8) | 24 (15.6) | ||
| Fair | 83 (26.7) | 40 (25.5) | 43 (27.9) | ||
| Not very good | 135 (43.4) | 74 (47.1) | 61 (39.6) | ||
| Self-rated understanding of numbers, charts or tables | Very good | 25 (8.0) | 12 (7.6) | 13 (8.4) | 0.02 |
| Good | 49 (15.8) | 29 (18.5) | 20 (13.0) | ||
| OK | 50 (16.1) | 16 (10.2) | 34 (22.1) | ||
| Fair | 95 (30.5) | 45 (28.7) | 50 (32.4) | ||
| Not very good | 92 (29.6) | 55 (35.0) | 37 (24.0) | ||
| PHM salience/coherence (range 4–20, α=0.53) | 18.6 (1.8) | 18.6 (1.9) | 18.6 (1.7) | 0.80 | |
| PHM susceptibility (range 3–15, α=0.75) | 8.7 (2.6) | 8.6 (2.5) | 8.7 (2.6) | 0.90 | |
| PHM cancer worry (range 2–10, α=0.80) | 5.6 (2.9) | 5.4 (2.8) | 5.7 (2.9) | 0.35 | |
| PHM response efficacy (range 2–10, α=0.66) | 8.8 (1.5) | 8.7 (1.5) | 8.9 (1.4) | 0.24 | |
| PHM social influence (range 4–20, α=0.53) | 16.7 (3.1) | 16.8 (3.2) | 16.7 (3.1) | 0.84 | |
| PHM religious beliefs (range 5–20, α=0.69) | 12.7 (5.4) | 12.3 (5.3) | 13.2 (5.6) | 0.17 | |
| PHM self-efficacy (range 6–30, α=0.77) | 28.2 (2.9) | 28.0 (3.0) | 28.3 (2.8) | 0.38 | |
| Cancer awareness score (range 0–6, α=0.36) | 3.6 (13) | 3.6 (1.3) | 3.5 (1.3) | 0.71 |
Boldface indicates statistical significance (p<0.05).
α is Cronbach’s alpha for this sample.
FIT Return
Two-hundred twenty-five participants (69.0%, 95% CI [64.0–74.0%]) returned the FIT: 111 (67.3%) from C-CARES Plus and 114 (70.8%) from C-CARES. Group differences in FIT uptake were evaluated using logistic regression in a model that included numeracy and educational attainment as covariates. Intervention was not a significant predictor of FIT uptake (AOR=0.82, 95% CI [0.50–1.33], p=.42).
To better understand FIT return, sociodemographic, health-related, and study variables were evaluated individually as predictors, with results presented in Table 3. The five variables that predicted FIT return highly covaried conceptually and statistically. For example, those who received materials in Spanish were more likely to report Hispanic ethnicity, Spanish as the language spoken at home, a parent born outside the US, and preference for health information in Spanish (p’s<.001). Spanish-language intervention receipt was selected to represent the other four variables given this covariation and its direct relevance to the intervention.
Table 3.
Predictors of FIT uptake in univariate analysis
| Variable | Level | N | Odds Ratio (95% CI) | p-value |
|---|---|---|---|---|
| Intervention language | Spanish | 136 | 1.93 (1.20–3.26) | 0.007 |
| English | 190 | - | - | |
| Gender | Male | 132 | 0.92 (0.57–1.49) | 0.74 |
| Female | 193 | - | - | |
| Parent born outside of United States | Yes | 172 | 1.75 (1.09–2.81) | 0.02 |
| No | 152 | - | - | |
| Married/partnered | No | 178 | 0.83 (0.51–1.33) | 0.44 |
| Yes | 147 | - | - | |
| Employed (full- or part-time) | No | 188 | 1.14 (0.71–1.83) | 0.59 |
| Yes | 138 | - | - | |
| White race | No | 174 | 1.49 (0.93–2.39) | 0.10 |
| Yes | 152 | - | - | |
| Hispanic ethnicity | Yes | 173 | 1.74 (1.08–2.80) | 0.02 |
| No | 153 | |||
| Educationb | Less than HS diploma | 133 | 1.31 (0.77–2.25) | 0.32 |
| HS diploma | 84 | 1.59 (0.85–2.98) | 0.14 | |
| Beyond HS diploma | 108 | - | ||
| Living in owned home | Yes | 134 | 1.39 (0.86–2.26) | 0.18 |
| No | 192 | - | - | |
| Medical insurance | Yes | 125 | 0.77 (0.48–1.25) | 0.29 |
| No | 201 | - | - | |
| Primary language spoken at home | Spanish | 146 | 1.86 (1.14–3.04) | 0.01 |
| English | 173 | - | - | |
| Preferred language for health information | Spanish | 136 | 2.08 (1.26–3.43) | 0.004 |
| English | 189 | - | - | |
| Annual household incomec | 320 | 1.19 (0.93–1.53) | 0.16 | |
| Household income stabilityc | 326 | 1.08 (0.95–1.23) | 0.26 | |
| Worry about getting cancer | 326 | 0.99 (0.75–1.30) | 0.91 | |
| Worry about getting CRC | 326 | 1.15 (0.78–1.68) | 0.48 | |
| Cancer awareness score | 326 | 0.98 (0.82–1.18) | 0.86 | |
| PHM salience/coherence | 326 | 1.04 (0.92–1.19) | 0.52 | |
| PHM susceptibility | 326 | 1.00 (0.91–1.09) | 0.96 | |
| PHM response efficacy | 326 | 1.11 (0.95–1.30) | 0.18 | |
| PHM social influence | 326 | 1.04 (0.96–1.12) | 0.34 | |
| PHM religious beliefs | 326 | 0.99 (0.95–1.03) | 0.65 | |
| PHM self-efficacy | 326 | 1.03 (0.96–1.12) | 0.42 | |
| Get help with written materials | 326 | 0.94 (0.77–1.15) | 0.55 | |
| Confidence in filling out forms | 311 | 0.99 (0.81–1.20) | 0.91 | |
| Self-rated ability to read | 311 | 1.03 (0.84–1.26) | 0.79 | |
| Self-rated understanding of numbers, charts or tables | 311 | 0.97 (0.80–1.17) | 0.74 |
Boldface indicates statistical significance (p<0.05). P-values less than alpha are in bold with three decimal places.
For Education, Type 3 p-value equals 0.32.
Levels presented in Table 1 were coded 1 to 5 and variable was evaluated as a continuous predictor of FIT kit uptake.
These predictors were further explored with the focus on intervention material language. Those receiving Spanish-language materials returned 77.2% of FIT kits provided, whereas those receiving English-language materials returned 63.2% (p=.007). To better understand this observed difference, these groups were compared across all other sociodemographics. Analyses found that those receiving materials in Spanish were more likely to be married/living with a partner, employed full-or part-time, uninsured, and had fewer years of education (p’s<01).
The mean number of days to FIT return was 24.6 days for the overall study (range: 0–284 days; mode: 2 days; median: 7 days). Among C-CARES Plus participants, the mean number of days to FIT return was 21.7 days (range: 0–278; mode: 1 day; median: 7 days). Among C-CARES participants, the mean number of days to FIT return was 27.4 days (range: 0–284 days; mode: 7 days; median: 7 days). The vast majority of participants in the overall study who returned a FIT did so within 90 days or less (91.1%), with no significant differences by study arm (93.7% in C-CARES Plus and 88.6% in C-CARES, p=0.18).
A total of 59 C-CARES Plus intervention participants were eligible for the coaching as they had not returned their FIT within 90 days. A total of 36 participants were reached for coaching (range of dates coaching received: 105–200 days), and 10 returned a FIT.
Clinical Findings
Forty-six participants who returned their FIT (20.4%) had an abnormal result. The group differences were not significant (p=0.82) with 22 (19.8%) in C-CARES Plus and 24 (21.1%) in C-CARES. Among those with an abnormal result, 20 individuals completed a colonoscopy, 5 individuals refused a colonoscopy, 13 individuals were not reachable for navigation to a colonoscopy/lost to follow-up, and 8 are pending/in progress to receive a colonoscopy. Among those completing a colonoscopy, no cancers were identified.
DISCUSSION
In the year prior to intervention implementation, the clinic organizations had CRC screening rates of 27.1% and 32.9%, respectively.14 Although there were no statistically significant differences in FIT screening between intervention arms, the current trial resulted in an overall initial FIT uptake of 69%, a rate approaching the national CRC screening initiative goal at the time of the study (70.5% by 2020).3 The observed CRC screening uptake result in our study is higher than many prior CRC screening interventions conducted among patients receiving care at FQHCs.23–27 Findings highlight the importance of providing language-specific, literacy-relevant education and access to easy and convenient screening.15–17, 28 Intervention language, representing four highly-related sociodemographics (i.e., Hispanic/Latino ethnicity, Spanish as one’s primary language, a preference for receiving medical information in Spanish, and having parents born outside the United States), predicted FIT uptake. It may be that the culturally-targeted Spanish-language education materials (photonovella and DVD) were especially salient, engaging, and motivating for the intended audience (i.e., individuals of Hispanic/Latino ethnicity) and this promoted higher FIT uptake among Spanish speakers, whereas the English language materials featured racially- and ethnically-diverse individuals, but were not targeted to a single racial/ethnic group. Yet, as related over 30 years ago, the effect of language on screening practices may be less of a cultural factor but rather more of an access factor (our intervention provided all participants with a free FIT).29 This suggests that language may be a critical determinant and a bridge to open the door to health service utilization, which was evident in our study due to the higher return of FIT kits among Spanish-speaking participants. Indeed, language may be an important factor to understand the use of screening services. Additional outreach among Latinos who are monolingual or prefer to receive their health information in Spanish as well as greater provider education about language as an access factor may boost CRC screening utilization. Recently, Castañeda and colleagues further endorsed the importance of improved institutional access to care among Latinos as a key determinant to reaching higher CRC screening rates.30 Further, Huguet et al. (2021) found that community health centers with a higher proportion of Hispanic/Latino patients were more likely to have met metrics for the percentage of their patients who were up-to-date with cervical cancer screening, CRC screening, and tobacco-cessation intervention compared to those with a lower proportion.31 However, it is unclear whether the finding from that study relates to those community health centers with more Hispanic/Latino patients perhaps also having a higher proportion of Spanish-speaking providers and/or greater availability of educational materials in Spanish.
The C-CARES Plus coaching component did not appear to provide added benefit beyond the C-CARES intervention. It may be that individuals who did not complete a FIT had decided against completing CRC screening, or other factors may have impeded their CRC screening decision-making. The decision to provide coaching at 90 days was based upon our prior research in which we found that the majority of FIT kits returned were completed within 90 days; it is possible that a different timeline (e.g., coaching at 30 days or 45 days) might have been more effective at a time closer to receipt of the FIT kit to reinforce screening importance and serve as a reminder. Further research is needed to better disentangle if a different dose, timing, or frequency of coaching might be more beneficial and whether there might be other possible addressable obstacles. Or perhaps, as described by Lam and colleagues, the routine and expanded use of technology (such as mobile messaging) might further improve patient engagement and boost screening rates.32
The current study has multiple strengths. First, the primary outcome was the return of a FIT as captured by the EMR. Second, the study provided low literacy, language-specific education and access to CRC screening for all participants, important steps in promoting equity. Individuals who had a positive FIT result were supported to access a diagnostic colonoscopy. Finally, this study engaged and reached a racially and ethnically-diverse multilingual patient population with 53% reporting Hispanic/Latino ethnicity.
Limitations
Study limitations should also be noted. First, data were collected in a single geographic region in urban/suburban FQHCs, which may limit generalizability. Second, enrollment of the current study was limited to those who speak, read, and understand either English or Spanish. Third, enrollment was limited to those patients attending a scheduled clinic visit; patients with a scheduled clinic visit may differ in multiple ways (e.g., more regular and frequent care, more preventive visits) from patients who attend the clinic as walk-ins (e.g., acute health issue) in terms of their health and cancer screening behaviors. However, for calculation of UDS measures all age-eligible patients seen during a calendar year would be included in the UDS measures, without adjusting for multiple visits or frequency/regularity of care (potentially reducing the overall system screening rates). Fourth, we were only able to reach 36 of the 59 eligible participants for the CARES Plus arm coaching component, which may have impacted the potential effect of that component. Finally, the findings presented are limited to data from initial CRC screening uptake (up to 1-year post-intervention).
CONCLUSIONS
Both interventions included low literacy education materials in either English or Spanish plus provision of a FIT and were successful in promoting FIT uptake among patients receiving care at FQHCs. The overall initial screening rate of 69% in this study far exceeds the prevailing UDS rates of the clinics, thus, supporting potential utility of these strategies. In addition to harnessing system-level EMR tracking, prompt components and reminders, clinics should consider language preference of patients and ensure that clear health education materials are available in patients’ preferred languages. Such multilevel strategies should be part of routine health care and organizational practices as they have the potential to promote health equity and improve health outcomes for one of the most detectable, treatable, and beatable cancers.
Supplementary Material
Figure 1.
CONSORT diagram
Highlights.
Colorectal cancer screening rates are suboptimal in community clinics.
Systems- and patient-level colorectal cancer screening components were provided.
Intervention resulted in overall fecal immunochemical test (FIT) uptake of 69%.
There were no differences in FIT uptake by intervention arm.
ACKNOLWEDGEMENTS
Disclosure of funding:
This research was supported by the Florida Department of Health Bankhead-Coley Research Program under grant #7BC04 (PIs: C.K. Gwede & C.D. Meade) and in part by Florida Blue under grant #5868 (PIs: C.K. Gwede & C.D. Meade). The efforts of Dr. Christy were supported by the National Cancer Institute (R25CA090314; PI: T.H. Brandon) when she was a postdoctoral fellow. The efforts of Drs. Cousin and Ewing were supported by the National Cancer Institute (T32CA090314; PIs: T.H. Brandon & S.T. Vadaparampil). Support was provided by the Biostatistics and Bioinformatics Shared Resources at the H. Lee Moffitt Cancer Center & Research Institute, an NCI-designated Comprehensive Cancer Center (P30-CA076292: PI: J. L. Cleveland). The funders had no role in the study design, collection, analysis, and interpretation of data, the writing of the manuscript, or the decision to submit the manuscript for publication. The content is solely the responsibility of the authors and does not necessarily represent the official views of the State of Florida Department of Health, National Institutes of Health, or Florida Blue.
The manuscript authors wish to thank Cindy Burcham for her assistance formatting the Figure for the manuscript. Preliminary study results were presented as a poster presentation at the 2019 American Academy of Nursing annual meeting in Washington D. C. Study results were presented as an oral presentation at the 2021 International Cancer Education Conference (virtual).
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Declaration of interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
- 1.American Cancer Society. Cancer Facts & Figures 2022 2022. Available https://www.cancer.org/research/cancer-facts-statistics/all-cancer-facts-figures/cancer-facts-figures-2022.html. Accessed 2/28/2022.
- 2.American Cancer Society. Colorectal Cancer Facts and Figures Available: https://www.cancer.org/research/cancer-facts-statistics/colorectal-cancer-facts-figures.html. Accessed 11/2/2020, 2020.
- 3.U.S. Department of Health and Human Services. Healthy People 2020. Available: https://www.healthypeople.gov/2020/topics-objectives/objective/c-16. [DOI] [PubMed]
- 4.U.S. Department of Health and Human Services. Health People 2030: Objectives: Browse Objectives: Cancer: Increase the proportion of adults who get screened for colorectal cancer-C-07 Available: https://health.gov/healthypeople/objectives-and-data/browse-objectives/cancer/increase-proportion-adults-who-get-screened-colorectal-cancer-c-07/data.
- 5.Honein-AbouHaidar GN, Kastner M, Vuong V, et al. Systematic Review and Meta-study Synthesis of Qualitative Studies Evaluating Facilitators and Barriers to Participation in Colorectal Cancer Screening. Cancer epidemiology, biomarkers & prevention : a publication of the American Association for Cancer Research, cosponsored by the American Society of Preventive Oncology. 2016;25(6):907–917. [DOI] [PubMed] [Google Scholar]
- 6.Kim J, Wang H, Young L, et al. An Examination of Multilevel Factors Influencing Colorectal Cancer Screening in Primary Care Accountable Care Organization Settings: A Mixed-Methods Study. Journal of public health management and practice : JPHMP. 2019;25(6):562–570. [DOI] [PubMed] [Google Scholar]
- 7.Wang H, Roy S, Kim J, et al. Barriers of colorectal cancer screening in rural USA: a systematic review. Rural and remote health. 2019;19(3):5181. [DOI] [PubMed] [Google Scholar]
- 8.Myers RE, Sifri R, Hyslop T, et al. A randomized controlled trial of the impact of targeted and tailored interventions on colorectal cancer screening. Cancer. 2007;110(9):2083–2091. [DOI] [PubMed] [Google Scholar]
- 9.Tiro JA, Vernon SW, Hyslop T, et al. Factorial validity and invariance of a survey measuring psychosocial correlates of colorectal cancer screening among African Americans and Caucasians. Cancer Epidemiol Biomarkers Prev. 2005;14(12):2855–2861. [DOI] [PubMed] [Google Scholar]
- 10.McQueen A, Tiro JA, Vernon SW. Construct validity and invariance of four factors associated with colorectal cancer screening across gender, race, and prior screening. Cancer Epidemiol Biomarkers Prev. 2008;17(9):2231–2237. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Vernon SW, Myers RE, Tilley BC. Development and validation of an instrument to measure factors related to colorectal cancer screening adherence. Cancer Epidemiol Biomarkers Prev. 1997;6(10):825–832. [PubMed] [Google Scholar]
- 12.Israel BA, Eng E, Schulz AJ, et al. Methods for community-based participatory research for health: John Wiley & Sons 2012. [Google Scholar]
- 13.Wallerstein NB, Duran B. Using community-based participatory research to address health disparities. Health Promot Pract. 2006;7(3):312–323. [DOI] [PubMed] [Google Scholar]
- 14.Health Resources and Services Administration. Health Center Program Data Available: https://data.hrsa.gov/tools/data-reporting/program-data?grantNum=H80CS00541. Accessed March 26, 2021, 2021.
- 15.Davis SN, Christy SM, Chavarria EA, et al. A randomized controlled trial of a multicomponent, targeted, low-literacy educational intervention compared with a nontargeted intervention to boost colorectal cancer screening with fecal immunochemical testing in community clinics. Cancer. 2017;123(8):1390–1400. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Gwede CK, Sutton SK, Chavarria EA, et al. A culturally and linguistically salient pilot intervention to promote colorectal cancer screening among Latinos receiving care in a Federally Qualified Health Center. Health Educ Res. 2019;34(3):310–320. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Christy SM, Davis SN, Williams KR, et al. A community-based trial of educational interventions with fecal immunochemical tests for colorectal cancer screening uptake among blacks in community settings. Cancer. 2016;122(21):3288–3296. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Finney Rutten LJ, Blake KD, Skolnick VG, et al. Data Resource Profile: The National Cancer Institute’s Health Information National Trends Survey (HINTS). Int J Epidemiol. 2020;49(1):17–17j. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Chew LD, Bradley KA, Boyko EJ. Brief questions to identify patients with inadequate health literacy. Fam Med. 2004;36(8):588–594. [PubMed] [Google Scholar]
- 20.Bishop WP, Craddock Lee SJ, Skinner CS, et al. Validity of Single-Item Screening for Limited Health Literacy in English and Spanish Speakers. Am J Public Health. 2016;106(5):889–892. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Fagerlin A, Zikmund-Fisher BJ, Ubel PA, et al. Measuring numeracy without a math test: Development of the Subjective Numeracy Scale (SNS). Medical Decision Making. 2007;27:672–680. [DOI] [PubMed] [Google Scholar]
- 22.Kushel MB, Gupta R, Gee L, et al. Housing instability and food insecurity as barriers to health care among low-income Americans. J Gen Intern Med. 2006;21(1):71–77. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Khankari K, Eder M, Osborn CY, et al. Improving colorectal cancer screening among the medically underserved: a pilot study within a federally qualified health center. J Gen Intern Med. 2007;22(10):1410–1414. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Lasser KE, Murillo J, Lisboa S, et al. Colorectal cancer screening among ethnically diverse, low-income patients: a randomized controlled trial. Arch Intern Med. 2011;171(10):906–912. [DOI] [PubMed] [Google Scholar]
- 25.Menon U, Szalacha LA, Kue J, et al. Effects of a Community-to-Clinic Navigation Intervention on Colorectal Cancer Screening Among Underserved People. Ann Behav Med. 2020;54(5):308–319. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Mojica CM, Almatkyzy G, Morales-Campos D. A Cancer Education-Plus-Navigation Intervention Implemented Within a Federally Qualified Health Center and Community-Based Settings. J Cancer Educ. 2019. [DOI] [PubMed] [Google Scholar]
- 27.Singal AG, Gupta S, Skinner CS, et al. Effect of Colonoscopy Outreach vs Fecal Immunochemical Test Outreach on Colorectal Cancer Screening Completion: A Randomized Clinical Trial. Jama. 2017;318(9):806–815. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sudore RL, Schillinger D. Interventions to Improve Care for Patients with Limited Health Literacy. J Clin Outcomes Manag. 2009;16(1):20–29. [PMC free article] [PubMed] [Google Scholar]
- 29.Solis JM, Marks G, Garcia M, et al. Acculturation, access to care, and use of preventive services by Hispanics: findings from HHANES 1982–84. American journal of public health. 1990;80 Suppl(Suppl):11–19. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Castañeda SF, Gallo LC, Nodora J, et al. Colorectal cancer screening among Hispanics/Latinos in the HCHS/SOL sociocultural ancillary study. Preventive medicine reports. 2019;15:100947. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Huguet N, Hodes T, Holderness H, et al. Community Health Centers’ Performance in Cancer Screening and Prevention. Am J Prev Med 2021:S0749–3797(0721)00433–00435. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Lam TYT, Wu PI, Tang RSY, et al. Mobile messenger-initiated reminders improve longitudinal adherence in a community-based, opportunistic colorectal cancer screening program: A single-blind, crossover randomized controlled study. Cancer. 2021;127(6):914–921. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.