1 Introduction

Colorectal cancer (CRC) screening is a critical preventative service recommended by the United States Preventive Services Task Force (USPSTF), Centers for Disease Control and Prevention, American Cancer Society, and the American College of Gastroenterology (ACG), part of the US Multi-Society Task Force. Currently, CRC is the second leading cause of cancer death in the United States (US), and yet a third of the eligible population between ages 45 to 75 does not undergo routine recommended screening [1, 2]. CRC is largely preventable. Between 1997 and 2017, screening rates in Rhode Island nearly doubled, and with this increase in screening came a subsequent > 50% reduction in new CRC diagnoses, from 70 to 31.5 per 100,000 [3]. Endoscopy centers have been strained by a period of limited access, decreased staffing, and other challenges related to the COVID-19 pandemic [4]. The onset of the pandemic in 2020 caused CRC screening to decline by 86–94% compared to previous years [5]. The recent change in guidelines reducing the recommended age at which to begin screening from 50 to 45 further worsened the backlog of patients requiring screening. There are approximately 100 million people in the US between the ages of 50 and 75, roughly 33 million of whom remain unscreened for CRC [6]. There are an additional 20 million people between the ages of 45–49, who now require screening per the updated guidelines. This leaves at least 53 million unscreened Americans, at risk for CRC [7].

CRC screening rates for average-risk individuals aged 50–75 in the US have reached around 68%. By race, screening rates are 71% for whites, 70% for African Americans, 64% for Asians, and 56% for Hispanics. Regarding insurance status, 71% of insured Americans are screened, while only 40% of the uninsured are screened. When examining whether a patient is connected to a primary care provider (PCP) and receiving regular care or not, 73% of patients with a regular PCP are up-to-date on their screening, while only 36% of patients without a PCP have been screened [8]. Area deprivation has also been linked to trends in cancer screening. Using the area deprivation index (ADI), a composite measure of social determinants of health incorporating neighborhood rankings by socioeconomic disadvantage, one study investigated screening rates based on census block groups and reported that individuals living in the 20% most deprived areas were almost half as likely to undergo recommended CRC screening as those living in the 20% least deprived areas [9]. Healthcare disparities are dramatically apparent when examining screening rates for CRC.

The Center for Primary Care (CPC), Rhode Island Hospital’s (RIH) main safety-net outpatient primary care clinic for adult patients, sends many screening colonoscopy referrals to the RIH Gastroenterology Fellows Clinic (GIFC). The GIFC is the primary gastroenterology (GI) group providing care for the uninsured and underinsured in the state. Rhode Island has a population of 1 million people, over 40,000 of whom are without health insurance [10]. Due to limited resources, time between referral and completion of procedure can be many months, particularly for average-risk screening colonoscopies. One potential solution to reduce this disparity and improve access is to utilize fecal immunochemical testing (FIT), one of seven first-line CRC screening tests recommended by USPSTF guidelines.

The implementation and addition of screening FIT boasts numerous advantages, including the lack of requirement for bowel preparation, anesthesia or sedation, as well as the general ease of being able to conduct the test from the comfort of one’s home. It is not surprising that yearly FIT is generally preferred to colonoscopy every 10 years; 68.9% of 40–49 year olds and 77.4% of patients 50 or older prefer yearly FIT [11]. With the advantages of FIT comes the caveat that success relies heavily on patient compliance. Unfortunately, patients inconsistently complete the test for analysis, or return the test after the kit has expired. To improve patient adherence, we implemented a quality improvement initiative using standardized interventions for patients who had outstanding kits, via a dedicated patient navigator. There have been several studies investigating various methodologies to improve CRC screening rates, including mailed outreach and distribution of health literacy materials, but to our knowledge this is the first study examining the utility and cost-effectiveness of a dedicated patient navigator to enhance the rate of FIT completion [12,13,14].

The updated 2021 ACG Clinical Guidelines for CRC screening state, “We recommend organized screening programs to improve adherence to CRC screening compared with opportunistic screening” (strong recommendation, low quality evidence). They go further to say, “We suggest the following strategies to improve adherence to screening: patient navigation, patient reminders, clinician interventions, provider recommendations, and clinical decision support tools” (conditional recommendation, very low-quality evidence) [15]. This study provides a critical piece of evidence to support this recommendation and endorse the use of patient navigators as a cost-effective intervention to increase screening rates.

2 Methods

We performed a retrospective analysis on 3542 unique individuals at a single internal medicine clinic with 5211 FIT kits ordered as part of USPSTF recommended care from January 2017 through December 2021. Starting in January 2021 we instituted a dedicated patient navigator to support patients in completing FIT.

2.1 Study design

The study was reviewed and approved by RIH’s Lifespan Health System Institutional Review Board on 4/6/2022 (package #1869781–1). Utilizing a REDCap database, we analyzed increases in return rates on FIT kits with the addition of the dedicated patient navigator [16]. We tracked statistics for FIT kits on the CPC study population starting from January 2017. The addition of a dedicated language-concordant patient navigator in January 2021 was the result of a plan-do-study-act (PDSA) analysis which suggested that return rates on FIT kits could be improved. To analyze the impact of the patient navigator and to allow for direct comparison, we compared statistics of the study population from January 1, 2017 through December 31, 2020 to those from January 1, 2021 through December 31, 2021. Baseline demographic and outcome measures were obtained from electronic medical records. We compared the average cost of CRC screening per patient before and after our intervention; the total costs taken into consideration included FIT kit, FIT analysis, patient navigator, and colonoscopy. The allowed charge amounts on the components of colonoscopy fees (provider, anesthesia, facility, and outpatient) were obtained from Rhode Island Department of Health claims data and Fair Health.

2.2 Setting

The CPC primarily serves individuals from deprived and medically underserved populations, [17] and includes patients of all racial and ethnic backgrounds, many of whom speak a language other than English, and the majority of whom are of low socioeconomic status. Deprived areas are defined based on ADI, according to rankings of socioeconomic disadvantage [18]. Medically underserved populations identify geographic regions with a shortage of primary health care services, with individuals facing economic, cultural, or language barriers to health care [17]. Patients at the CPC range in insurance status; the highest proportion are Medicaid beneficiaries (48%), many utilize Medicare (34%), some are privately insured with commercial insurance (10%), while others are uninsured or undocumented—“self-pay” (7%).

2.3 Screening program

A variety of modalities are available as options for CRC screening. Stool-based tests include yearly high-sensitivity guaiac fecal occult blood test (gFOBT), yearly FIT, and multi-target stool DNA (sDNA)-FIT every 1–3 years [2]. Direct visualization tests include colonoscopy every 10 years, CT colonography or flexible sigmoidoscopy every 5 years, or flexible sigmoidoscopy every 10 years with the addition of yearly FIT [2].

For many years, screening at the CPC was accomplished primarily via colonoscopy. In 2016, RIH’s Lifespan Laboratories adopted FIT. FIT is a sensitive and specific screening modality, which may be used to reach and risk-stratify patients to determine who most urgently needs to undergo colonoscopy. Starting in January 2017 the CPC, via a quality improvement pilot program, began actively promoting FIT in an effort to increase the total number of patients screened. The goal was to increase the yield for detecting advanced neoplasia and cancer, improve triage for CRC screening via colonoscopy, and reduce the wait time for colonoscopy. Patients were enrolled in FIT by their PCP during regular clinic visits, with positive tests referred for colonoscopy. Beginning in January 2021, a dedicated culturally competent, language congruent patient navigator was implemented to improve the yield of screening FIT.

2.4 Framework

The patient navigator program involved several outreach steps, outlined as follows (Fig. 1).

Fig. 1
figure 1

diagram of the fecal immunochemical testing patient navigator intervention program. CPC Center for Primary Care; FIT fecal immunochemical testing; GIFC Gastroenterology Fellows Clinic; RIH Rhode Island Hospital

Full size image

Following initial distribution of FIT kits to patients, the navigator received a weekly distribution list. At this step in the process, they made their first outreach attempt, documenting returned kits through EPIC, RIH’s electronic medical record, and noting any potential patient barriers [19]. They made a total of three outreach attempts via three telephone calls over the subsequent three weeks, each time documenting success of returned kits as well as any barriers to compliance. Prior to reaching out to any patients, our navigator was trained in the role and importance of CRC screening via FIT. The navigator was also given information about common demographics of patients at the CPC, in order to ensure a culturally sensitive approach to all patient communication. Our navigator was bilingual (English and Spanish) and utilized a Language Line telephone interpreter for patients who spoke other languages. The patient navigator worked from a pre-formulated script, but also offered personalized assistance. If the patient needed guidance, they talked them through the instructions to help with the process of stool collection and return of the kit. If the patient refused FIT, they reiterated that CRC is the second leading cause of cancer death in the US yet can be prevented if detected at an early stage. They sent information on screening and CRC to the patient, and asked if they’d like to speak with their PCP further about screening options, setting up a telehealth meeting if they agreed. For patients who had misplaced or lost their original FIT kit, they provided instructions on obtaining a replacement, and called them to follow up after two weeks.

2.5 Screening cohort

The patient cohort included the 3542 unique patients at the CPC between the ages of 45–75 who were sent FIT kits as part of routine USPSTF recommended care from 2017 to 2021. The subsequent analysis included data on the 4030 total kits sent between 2017 and 2020 (before the intervention) as well as the 1181 total kits sent throughout 2021 (during which time the patient navigator was active). There were more FIT kits than patients because some of the patients received kits yearly, as per current guidelines.

2.6 Data collection

Demographic data collected on each patient included gender, age, ethnicity, race, insurance status, language, and whether they were from deprived and/or medically underserved areas.

Patients were analyzed by pre-/post-intervention categorization to assess impact of the patient navigator. Patients were identified as requiring follow up via an internal report of outstanding FIT kits that had been picked up from the lab but not yet returned. Each month the patient navigator documented the total number of patients who had been sent FIT kits, the number of kits returned with their drop-off and collection dates, and the number of outstanding kits requiring follow up. For each outstanding kit, they contacted the patient and determined and documented barriers to successful completion.

Colonoscopy costs were broken down into provider fees ($299.86), cost of anesthesia ($369.00), facility fees ($812.14), and related outpatient costs ($775.76) [20].

The median cost of each FIT kit itself is $3.04, and the average cost of analysis per returned kit is less than $19 [21, 22]. Hiring a dedicated patient navigator costs $4.52 per patient, based on the average 10-min phone call allotted to each patient with an outstanding kit.

2.7 Statistical analyses

Data recorded by the patient navigator was analyzed retrospectively. Information documented included primary reason cited for completion failure, return rates of FIT kits prior to the intervention versus following the addition of the patient navigator, and time to successful completion of FIT kits pre-/post-intervention. Chi-square, Fisher exact test, and Student’s t-tests were performed for descriptive analyses on demographics and cost components. Multivariable (Cox hazard model) logistic regression was used to compare FIT kit completion rates pre-/post-intervention, and multivariable linear regression was used to compare average total costs associated with screening. The model was adjusted by age, gender, race, ethnicity, language, insurance status, deprived and medically underserved areas.

Sensitivity analysis was performed to reinforce and validate the findings where sub-analyses were done: 1) to consider only patients older than 50 years in both study groups, to address the potential confounding factor of age, given the overlap with the post-intervention period of the 2021 change in CRC screening guidelines which decreased the recommended age to initiate screening from 50 to 45; 2) to analyze the pre-intervention period starting after March of 2020, to address the effect of Covid-19 by limiting the pre-intervention population to that group which would have been impacted by the pandemic similarly to the post-intervention population; and 3) to use variable ratio matching (from 1:1 to a maximum of 1:5) of propensity scores to reduce selection bias due to differences between the two study periods resulting in unbalanced patient distribution – this was performed for age, race, insurance type, and preferred language [23, 24]. We used standardized mean difference values in the range of -0.25 to 0.25 to ensure that the goal of matching was achieved. We used gender, age groups, race, ethnicity, preferred language, insurance, ADI, and medically underserved areas as covariates for matching. Analysis was performed in SAS© software, version 9.4.

3 Results

3.1 Demographics

There were 3542 patients analyzed in the study (Table 1).

Table 1 baseline characteristics of patients who underwent fecal immunochemical testing, comparing pre- and post-intervention periods, from 2017 through 2021
Full size table

Pre-intervention included 2458 (69.4%), and post-intervention included 1084 (30.6%). We analyzed racial and ethnic demographics to better identify and illuminate the existence of disparities prevalent in our patient population. Over 45% were Hispanic/Latino and only 33% identified as White/Caucasian. Close to 75% were on government insurance with less than 25% utilizing commercial insurance. More than 6% were uninsured or undocumented. Approximately 50% of the patients were non-English speaking. Over 95% were from medically deprived areas, and over 90% were from medically underserved areas.

3.2 Primary results

The predominant reasons cited for failure to complete testing were “forgot to complete” (25%), “too busy” (13%), and “lost kit” (11%). Other patients had been unable to complete testing secondary to illness/hospitalization, required additional guidance, or did not have solid stools for the kit; some preferred to undergo colonoscopy, while others stated a general refusal to complete CRC screening.

There was no significant difference in completion rates between the two groups at 2 weeks. The intervention improved completion rates from 64.7% to 74.5% at 1 month (hazard rate (HR) 1.12, 95% confidence interval (CI) [1.003–1.243]), 73.7% to 90.1% at 3 months (1.28[1.156–1.410]), and 87.8% to 98.8% at 1 year (1.36[1.239–1.498]). Overall, the intervention improved FIT kit completion rates by 38.5% (1.39[1.260–1.522]) (Figs. 2, 3 and 4).

Fig. 2
figure 2

fecal immunochemical testing kit orders and return rates at CPC, from 2017 to 2021. CPC Center for Primary Care; FIT fecal immunochemical testing

Full size image
Fig. 3
figure 3

fecal immunochemical testing kit completion rates during pre- vs post-intervention period, from 2017 to 2021. CI confidence interval; HR hazard rate

Full size image
Fig. 4
figure 4

fecal immunochemical testing kit completion rates during pre- vs post-intervention period, from 2017 to 2021, represented on a cumulative survival curve. FIT fecal immunochemical testing

Full size image

FIT was positive in 4.9%.

Compared to the original analysis (Figs. 2, 3 and 4), which had an overall improvement of 38.5%, as above:

When considering only patients aged 50 or older: the intervention improved completion rates from 64.6% to 74.6% at 1 month (1.12[0.997–1.251]), 73.6% to 89.2% at 3 months (1.256[1.129–1.394]), and 87.8% to 98.6% at 1 year (1.34[1.241–1.484]). Overall, the intervention improved FIT kit completion rates by 37.1% (1.37[1.241–1.515]).

When examining orders after March 1, 2020, in order to consider only patients affected by Covid-19: the intervention improved completion rates from 69.9% to 74.5% at 1 month (1.28[1.050–1.548]), 74.3% to 90.1% at 3 months (1.41[1.181–1.681]), and 93.2% to 98.8% at 1 year (1.44[1.226–1.702]). Overall, the intervention improved FIT kit completion rates by 47.3% (1.47[1.251–1.735]).

After propensity score matching, the results show that the intervention improved completion rates from 64.5% to 74.7% at 1 month (1.12[1.006–1.248]), 73.5% to 90.1% at 3 months (1.283[1.162–1.417]), and 87.6% to 98.8% at 1 year (1.37[1.245–1.506]). Overall, the intervention improved FIT kit completion rates by 39.2% (1.39[1.266–1.530]).

3.3 Cost analysis

We analyzed the costs associated with CRC screening at our clinic, examining average fees for each component of screening. In the year 2021, during which we utilized our navigator, the costs broke down as follows. The average cost per patient for each colonoscopy was $2281.80; this made up 98.85% of the total cost associated with CRC screening for the year. This amount encompassed the facility fee ($812.14, 36%), outpatient fee ($775.76, 34%), anesthesia fee ($369.00, 16%), and provider fee ($299.86, 13%) [20]. Non-colonoscopy fees made up just over 1% of the total CRC screening costs for the year and included the cost of the FIT kit itself ($3.04, 0.13%), the cost of FIT analysis ($19.00, 0.82%), and the cost of our patient navigator ($4.52, 0.20%) [21, 22]. The cost per patient of hiring our patient navigator was determined by examining the average amount of time spent on each patient; the cost was $4.52 per 10-min phone call.

We compared the average cost of CRC screening per patient before and after our intervention of patient navigator (Fig. 5). In the post-intervention period, the total number of patients being screened with a much lower cost modality was higher, and the use of the higher cost modality (colonoscopy) was much reduced. As such, the average total cost during the pre-intervention period was $946.1 ± 1873.5, compared to the average total cost during the post-intervention period of $72.0 ± 383.3. With the addition of our intervention, the total cost of CRC screening per patient was decreased by $874.18 ($72.0 ± 383.3 vs. $946.1 ± 1873.5, p < 0.0001).

Fig. 5
figure 5

average total cost per patient associated with CRC screening per month, from April 2017 to December 2021

Full size image

Compared to the original analysis (Fig. 5), which had an overall savings of $874.18, as above:

When considering only patients aged 50 or older: with the addition of our intervention, the total cost of CRC screening per patient was decreased by $868.00 ($73.0 ± 389.6 vs. $941.1 ± 1875.6, p < 0.0001).

When examining orders after March 1st 2020 in order to consider only patients affected by Covid-19: with the addition of our intervention, the total cost of CRC screening per patient was decreased by $327.00 ($72.0 ± 383.3 vs. $399.00 ± 986.3, p < 0.0001).

After propensity score matching, the total cost of CRC screening per patient was decreased by $894.16 ($64.34 ± 366.8 vs. $958.5 ± 1885.4, p < 0.0001).

4 Discussion

FIT can increase access to CRC screening and timely screening rates, particularly in resource-limited settings, and may decrease the burden on endoscopy centers nationwide by improving the efficiency of colonoscopy in the average risk screening population. The addition of a dedicated patient navigator is a relatively straightforward intervention that, by providing culturally competent care and personalized attention, improves completion rates and return times, allowing FIT to be a reliable method of screening. In a severely strained healthcare system with nationwide delays in referral times, one outcome of this intervention is the ability to prioritize patients for diagnostic colonoscopies compared to routine screening colonoscopies. Increasing screening rates and prioritizing high-risk patients for urgent diagnostic colonoscopies will lead to earlier detection and treatment of CRC.

It is important to note that our study overlapped with COVID-19, which was declared a global pandemic by the World Health Organization on March 11, 2020, and it is critical to address the influence of the pandemic on our findings. In Fig. 5, one can observe a downward trend in overall CRC screening costs in the year leading up to the implementation of our intervention. While there may have been other factors contributing, this is certainly related to the decline in numbers of colonoscopies performed as endoscopy centers closed due to pandemic precautions. As the pandemic continued and the public became aware of the importance of social distancing, the appeal of an at-home testing modality increased, making FIT an appealing alternative to colonoscopy. Our intervention in January 2021 thus came at a critical time, enabling FIT to become a reliable screening modality for our population. As shown above, even when limiting analysis of our intervention to the time of the pandemic (limiting pre-intervention analysis to start after March of 2020) we were still able to demonstrate both an improvement in FIT kit completion rates and a significant financial savings.

We have demonstrated that patient navigators are a cost-effective screening intervention. The additional downstream savings are worth mentioning. Given that FIT allows physicians to stratify their patient population to expedite those with highest risk to the endoscopy suite, the colonoscopies being performed are more likely to identify cancerous or pre-cancerous lesions, thereby increasing the efficiency of colonoscopy in the overall population. Triage by FIT has been shown to identify 1 CRC case in every 33 colonoscopies, as opposed to 1 CRC case in every 200 procedures for patients not pre-screened with FIT [25]. By establishing a more effective population-wide screening program via prioritization of positive FIT, it becomes easier to catch neoplasms in their pre-malignant stage, thereby saving the healthcare system from the exponential costs of treating CRC. The national cost of CRC care (adjusted to 2021 USD using U.S. Bureau of Labor and Statistics Medical Consumer Price Index Estimates) is approximately $18.3 billion, though the economic burden of CRC varies remarkably by stage, with each case of stage 1 CRC costing $37,872, stage 2 $53,709, stage 3 $76,447, and stage 4 $111,241 [26,27,28]. Further, lost productivity from CRC deaths is estimated to cost over $15 billion annually [29].

While the benefits of our intervention are clear, the critical next step is ensuring that patients with a positive FIT go on to colonoscopy; here again there is a clear and present role for patient navigators. Non-compliance with colonoscopy after positive FIT doubles the risk of dying from CRC [30]. There are a variety of reasons why patients might originally opt for FIT over colonoscopy. While the theoretical risk of colon cancer is frightening, an imminent colonoscopy may also be intimidating, particularly if patients have had prior negative experiences with the preparation or anesthesia. Many individuals also harbor an innate distrust of the healthcare system; it is widely believed that the degree of this distrust has been worsening over the past few decades [31]. This is particularly evident among racial and ethnic minority groups, where a history of adverse treatment, marginalization, and persistent disparities in health outcomes supports the lack of faith in a just medical system, fueling the desire to avoid procedures and other interactions with the healthcare system [31, 32]. Another potential factor dissuading patients from completing necessary colonoscopies is inconvenience; many patients prefer the ease of a private at-home testing modality. This is especially relevant for an un-/under-insured and underserved population, as many patients can’t afford to take a day off work or don’t have anyone to transport them to/from the procedure. A positive FIT result does nothing to eradicate these numerous barriers. We expect that engagement with a patient navigator would play a critical role in assisting patients from positive FIT to completion of diagnostic colonoscopy [33, 34].

Patients opt for FIT because it is quick, private, convenient, and simple. By providing personalized attention and individualized answers to patients’ questions, the addition of a culturally competent, language congruent patient navigator ensures that those qualities hold true. While smartphone application-based systems have recently emerged in an attempt to aid in screening compliance, the lack of personalized care and the requirement for technological acumen will limit these interventions in ways that the patient navigator is able to supersede. Patient navigator programs are inherently reliant on the individual. Sick days, maternity leave, and general human error are all rate-limiting factors that we acknowledge may restrict efficiency and results. With an appropriate funding mechanism to support the upfront costs of hiring the necessary manpower and providing adequate training and support, these barriers can be partially overcome. A patient navigator is essential to improve health equity. If the patient navigator program can be successfully replicated and implemented broadly, we suspect that CRC screening rates would increase further and rates of CRC mortality would decrease.