The Effect of Sports Team Branding With Discount Incentive on Mailed Colorectal Cancer Screening: A Randomized Quality Improvement Study

Despite improvements, colorectal cancer (CRC) screening remains suboptimal. Mailing fecal immunochemical test (FIT) kits directly to patients is an evidence-based approach to increase screening rates, but the effectiveness is limited by response rate.^1,2 A partnership with a local professional sports team created an opportunity to co-brand kits and offer a limited number of merchandise discounts. While there have been no studies evaluating the effect of branding on FIT completion, there is mixed evidence regarding financial incentives for CRC screening.³ While there is evidence of benefit for incentives improving uptake of stool-based testing in a clinic setting,⁴ there has been no demonstrated benefit in mailed FIT.⁵ Co-branding with merchandise discounts could enhance the effect of incentives by invoking social norms through association with a recognized local sports team. Given this opportunity, we conducted a randomized quality improvement study to evaluate whether branding FIT kits in addition to offering a discount incentive could improve CRC screening uptake.

Patients who were aged 45–75 years at average risk for CRC and not up-to-date with screening in 7 community primary care practices within Chester County, PA, were included. These patients were texted with an opportunity to opt out of receiving a mailed FIT kit. If they replied “yes” or did not reply, inclusive of patients without textable phone numbers, they were randomized to receive either 1) a FIT kit branded with the Philadelphia 76ers professional basketball team in addition to a 15% discount code on online merchandise if they were among the first 100 kits to be returned (branded kit), or 2) the standard mailed FIT kit (standard kit) that included a letter describing the importance of screening and instructions on how to complete testing. All patients were texted with follow-up reminders at 42 days and 49 days after randomization. This project was reviewed and determined to qualify as quality improvement by the University of Pennsylvania’s Institutional Review Board and was determined to not require board approval.

The primary outcome was the difference in FIT completion between the 2 arms from May 3, 2024, to September 16, 2024. Secondary outcomes included a subgroup analysis of demographics, percent of abnormal FIT results, time to FIT completion, and incentive redemption. Patient demographic characteristics were summarized using proportions and compared using χ² tests. A 2-proportion Z-test was used to evaluate the primary outcome. A Kaplan-Meier survival analysis and a log-rank test were used to evaluate time to FIT completion.

A total of 1843 patients were sent a text message about receiving a FIT kit. Of these, a total of 1657 patients were randomized and included in the study, including 16 patients per arm who had phone numbers incapable of receiving texts. There was no significant difference in demographics between the 2 arms (Supplementary Table 1). There was no significant increase in FIT completion for those receiving branded kits compared to those receiving standard kits (13.3% vs 15.0%, respectively, P = .31, Table 1). A subgroup analysis showed no significant difference in FIT completion between arms by age, gender, and race/ethnicity (Table 1). There was no significant difference in FIT positivity (standard 2.5% vs branded 3.8%, P = .56) or time to FIT completion (Figure 1, P = .85). There were no redemptions of the discount incentive.

Table 1.

FIT Completion Rates of Patients Assigned to Receive Standard or Branded FIT Kits

	Standard (n = 827)	Branded (n = 830)	P value
Overall	124 (15.0%)	110 (13.3%)	.31
Gender			.19
Female	63 (15.9%)	47 (11.4%)
Male	61 (14.2%)	64 (15.3%)
Age			.54
45–54	53 (11.45%)	43 (9.5%)
55–64	40 (17.0%)	33 (14.6%)
65+	31 (24.0%)	35 (23.2%)
Race/ethnicity			.57
Asian	8 (21.6%)	5 (14.7%)
Black	13 (11.3%)	11 (9.7%)
White	94 (16.0%)	90 (15.1%)
Other	5 (9.8%)	1 (2.5%)
Unknown	4 (10.5%)	4 (9.3%)

Figure 1.

Cumulative incidence of FIT completion over time, compared by control (Standard) and intervention (Branded) arms. There is no significant difference in time to FIT completion between arms (P = .85).

This study did not find evidence of increased response from the branding FIT kits with a possible discount incentive. There are multiple reasons that may contribute to this. The effectiveness of branding depends on the strength and relevance of social norms. The impact of branding preventive health materials, such as a FIT kit, with a professional sports team is dependent on whether patients identify with the team and associate it with positive health behaviors. In this context, not all patients may have affinity for the team or perceive the association as credible. Additionally, there are many strategies for branding implementation, and this study represents only one approach. Though difficult to compare due to variation across health systems, patient populations, and implementation,¹ the rates of return in both arms are consistent with mailed FIT return rates, approximately 15%–20%, in a separate study evaluating behavioral interventions on mailed FIT completion in our health system.⁶

The incentive amounted to a possible merchandise discount, and its weakness was highlighted by the fact it was not redeemed by a single patient. Though prior evidence on financial incentives for cancer screening is mixed, those studies offered larger cash incentives.³ Small incentives may not be sufficient to motivate patients who are not up-to-date to complete screening. Additionally, framing of incentives matter; the incentive was not guaranteed and could only be redeemed for a discount on merchandise that may not be salient for patients. A qualitative study in an insured population associated with an urban academic hospital suggests financial incentives may not increase screening for most patients, and those for whom it may incite behavior change may have a unique set of barriers, such as a lack of belief in screening.⁷ Patients who remain unscreened likely represent a unique group with complex, multifaceted barriers for which financial incentives may not be effective. For these patients, branding and modest incentives may not meaningfully change behavior in the absence of more targeted or intensive interventions. Additionally, the median income approximated by patient ZIP codes was high, likely weakening the relative value of the discount incentive. Further research is required to identify specific incentives that may overcome unique barriers to screening in this population.

Strengths of this study include the fact that it was embedded in routine clinical operations and our ability to randomize at an individual level. This study demonstrates the feasibility of conducting randomized evaluations in partnership with health care systems and stakeholders. In clinical medicine, providers frequently make adjustments to their clinical workflow without formally evaluating the associated impact. Pragmatic evaluations embedded in real-world care can help systematically assess these routine variations. One limitation of this study was the fact that our findings may be difficult to generalize. This was a predominately white population in a community setting, relatively higher income, and not up-to-date with screening. Branding was specific to one local professional sports team and represented one implementation of many possibilities. Additionally, we did not have qualitative data to contextualize why patients in each group may or may not be completing screening. However, a study performed in the same health system with a different population described barriers with discomfort handling stool, difficulties remembering to complete the test, failure to receive a kit, and preconceived screening beliefs as contributing to incomplete testing.⁸ The study was also limited in what we could offer in terms of incentives.

In summary, we did not find evidence that branding FIT kits and offering a chance for a weak merchandise discount significantly increased FIT completion rates. Further work could explore what did or did not work about this branding partnership and how to optimize strategies to improve screening. Future research should explore other ways to leverage social norms with branding in CRC screening, as well as how such strategies might be adapted to engage populations less likely to complete CRC screening.