Despite evidence that screening for colorectal cancer (CRC) reduces the incidence of and mortality from the disease, only 69.7% of the eligible population in the United States is up-to-date with CRC screening (National Cancer Institute & Centers for Disease Control and Prevention, n.d.). Studies have consistently shown that provider recommendation to be screened, an evidence-based intervention (EBI) recommended by The Community Guide, is strongly associated with screening completion (Brawarsky et al., 2004; Honein-AbouHaidar et al., 2016; Lobchuk et al., 2012; Wee et al., 2005). Because providers may not recommend CRC screening during an office visit, provider reminder systems have been developed that prompt providers to offer screening to eligible adults. Patients may also not schedule or complete their ordered CRC screening test. Patient reminder systems encourage patients to complete the test and may contain additional educational information about the importance of being tested.

Through its Colorectal Cancer Control Program (CRCCP), the Centers for Disease Control and Prevention has funded awardees to implement interventions to increase CRC screening uptake, particularly among underserved and uninsured populations. (Additional detail on the CRCCP is provided in a companion article in this journal by Tangka et al.). Since January 2016, The University of Chicago Center for Asian Health Equity, a CRCCP awardee, has worked with a designated federally qualified health center (FQHC), Heartland Health Centers, to implement multiple EBIs. Multilevel interventions to increase CRC screening have been shown effective, and there is some evidence to indicate that they are cost-effective (Community Preventive Services Task Force, 2016). The intention of this article is to further contribute to the limited literature on the cost of CRC interventions by determining the effectiveness and cost of implementing a provider reminder system, the key EBI at Heartland, supplemented with patient reminders and provider assessment and feedback (PAF).

During the pre-implementation phase (January 2015–December 2015), 75.8% (N = 2,453) of the eligible population was not up to date with CRC screenings (Table 1). Of those not up-to-date at the time of the visit, 18.8% (N = 460) received a CRC screening order—50.6% (N = 233) of these were for FITs and FOBTs—from their physician, and 2.4% (N = 11) completed the order. During the implementation phase (January 2016–September 2017), 69.6% (N = 3,610) of the age-eligible population was not up-to-date with their CRC screening; 37.0% (N = 1,334) received an order—77.4% (N = 1,033) of these were for FITs or FOBTs—and 23.6% (N = 315) completed the screening, representing an overall increase in orders completed of 21.2 percentage points.

Table 2 displays the characteristics of patients who received a CRC screening order and completed it. The table shows both the pre-implementation (January 2015–December 2015) and implementation (January 2016–September 2017) periods. Screening orders and screenings completed increased across all sociodemographic categories. The percentage of male patients who received orders and completed screens increased by 19.1 percentage points each, while the percentage points for female patients who received orders and completed screens increased by 17.4 and 23.3 percentage points, respectively. By race and ethnicity, percentage point increases were highest among Hispanic and Asian patients: CRC screening tests ordered for Hispanic patients increased by 23.4 percentage points, and screens completed increased by 24.2 percentage points, while screens ordered for Asian patients increased by 21.1 percentage points, and screens completed increased by 28.2 percentage points. By age, the largest percentage point increases for screens completed were among older patients: patients 65 to 69 years of age increased by 28.0 percentage points, and patients 70 to 75 years of age increased by 29.8 percentage points. By insurance group, the largest percentage point increases of screens completed were 31.3 for uninsured patients and 19.8 for patients with Medicare.

From January 2016 through September 2017, the costs to implement each of the EBIs were as follows: provider reminders, $23,649; PAF, $9,336; and patient reminders, $7924.08 (Table 3). Components of provider reminders included identifying age-eligible patients and manually inputting flags for CRC screening ($21,597), and monitoring and tracking patient CRC screening status ($2,052). Overall, we estimated that an additional 283 screens were conducted (based on 21.2% increase in screening and 1,334 orders during the implementation period) as a result of EBI implementation at an implementation cost of $145 per screen completed.

This study has important implications for the implementation of EBIs to increase CRC screening among a large urban FQHC with a diverse population. First, the pre-implementation process allowed the awardee to determine the infrastructure needs and their capacity to implement EBIs. Second, each EBI was modified to fit the needs and staffing model for this health system, which may have improved efficiency and impact. Third, the effectiveness of these EBIs on CRC order and screening uptake was significant across all populations served and independent of insurance status. Furthermore, this study adds to the growing body of literature on the cost of implementing multicomponent interventions for increasing CRC screening uptake in an FQHC setting (Community Preventive Services Task Force, 2016; Kemper et al., 2018; Lara et al., 2018). While The Community Guide (Community Preventive Services Task Force, 2016) found the median incremental cost per additional person screened to be $582.44 for various multilevel, multicomponent interventions, Kemper et al. (2018) found the cost per FIT kit returned to be $39.81 and Lara et al. (2018) found the incremental cost per person screened for multiple EBIs implemented to be in the range of $23.78 to $29.16 in two FQHCs.

In this study, CRC screening completion was facilitated by using EBIs recommended by The Community Guide for provider reminders, patient reminders, and PAF (Sabatino et al., 2012). CRC screening uptake increased for patients of all races/ethnicities and demographic backgrounds at the FQHC. On national surveys, uptake of CRC screening tests is much lower among people 50 to 64 years of age compared with those 65 to 75 years of age, and among people who are Hispanic or Asian compared with other race/ethnicity groups (National Cancer Institute, n.d.). In this study, CRC screening tests ordered and completed increased substantially among patients 50 t 64 years of age, indicating that this was an effective intervention to increase screening in this population. Similarly, Hispanic and Asian patients had significant increases in CRC screening uptake (24.2 and 28.2 percentage point increases, respectively) such that by the end of the implementation period, these groups had the highest percentage of screens ordered and completed compared with other race/ethnicity groups.

This study contributes to the sparse literature around EBI influence on screens ordered for CRC screening. Specifically, the study demonstrates that both screens ordered and screens completed increased with these interventions. Screens ordered is an important measure to evaluate provider recommendations and engagement in the CRC screening process. The combination of all three EBIs might have led to an increase in CRC screening uptake, where screens ordered increased from 18.8% to 37.0%, and screens completed increased from 2.4% to 23.6% among those not up to date with screening. This study demonstrates that EBIs to increase CRC screening may be effective in clinics with fewer resources that serve populations who have poor access to or under-utilize health care services. Tailoring the EBIs to the needs and capacity of participating clinics may have been key to their effectiveness.

The total implementation cost for all EBIs was calculated to be $40908.97, and the cost per additional screen was $144.65. Although the incremental cost is higher for this health system than for some other FQHCs that implemented multicomponent interventions (Lara et al., 2018; Tangka et al., 2019), we believe that this is in part caused by manual entry of provider and patient reminders into the EMR. As this health system moves toward a more automated system of reminders, we expect that this incremental cost will decrease.

This study has several limitations. Although we would have liked to implement and measure the effect of each EBI separately, the goal of the health system was to see an increase in CRC screening uptake; therefore, we initiated a multicomponent approach to EBI implementation. Tracking process data that might isolate each EBI could help future implementation analyses. Second, our study population is a diverse urban, underserved population that may not represent a typical clinic population in another FQHC. However, this study does demonstrate that CRC screening rates increased in an underserved population when interventions were implemented effectively and with appropriate support. Third, this study presents the results on changes in screening coverage, but we may have underreported screenings performed and we were unable to collect and report data on diagnostic colonoscopy follow-up for those with positive FITs, a necessary step before screening can be considered complete (Nadel et al., 2019). Last, the cost data were collected retrospectively, which may have introduced recall bias. Future efforts will attempt to collect time and cost data contemporaneously as the interventions are being planned and implemented.

This study reports on the effectiveness of EBI implementation on both screens ordered as well as CRC screens completed across a health system of safety net primary care clinics. The introduction of a provider reminder system entered manually and supplemented with patient reminders, and PAF, increased CRC screening. Lessons learned from the Heartland Health Center’s demonstrated that effectiveness in increasing screening among Hispanic, Asian, and uninsured patients may inform other clinic systems working to eliminate cancer screening disparities among priority populations. As provider reminders become more automated in EMRs, the cost of labor for manually entering reminders may decrease and improve cost-effectiveness of the multicomponent interventions.