Talk to the Veterans Crisis Line now
U.S. flag
An official website of the United States government

Health Services Research & Development

Go to the ORD website
Go to the QUERI website

PPO 14-369 – HSR&D Study

PPO 14-369
Reducing Colorectal Cancer Death through Mailed Outreach Screening
E. Carter Paulson MD MSCE
Corporal Michael J. Crescenz VA Medical Center, Philadelphia, PA
Philadelphia, PA
Funding Period: October 2016 - December 2017

The VA has made substantial advances in colorectal cancer (CRC) prevention and control for Veterans. Despite this progress, several opportunities exist to enhance the delivery of screening as the primary approach to reduce disease burden. The long-term goal of this project was to optimize access to and use of effective screening among all Veterans, particularly those hard to reach, by improving the screening process.

This project sought to establish the foundation to systematically study and improve the delivery and effectiveness of CRC screening in the VA by developing, implementing, and disseminating a pilot mailed-fecal immunochemical test (FIT) screening outreach program that does not rely on having a clinical office visit.

The objective of this study was to develop, implement, and disseminate a pilot population-based mailed FIT outreach screening program in the Corporal Michael J Crescenz VA Medical Center (CMCVAMC) that does not rely on having an office visit, by conducting a proof-of-concept three-arm parallel-design pragmatic randomized trial to: 1) compare the effects of usual care (UC), screening invitation + reminder (invitation-reminder), or screening invitation + mailed FIT kit + reminder (mailed-FIT); 2) explore whether the FIT completion rate varies by age or race/ethnicity; and 3) explore barriers to use of mailed outreach screening for Veterans.

Study Design and Patients: This was a 3-arm parallel-design randomized trial.
The study population was composed of Veterans in the southeastern Pennsylvania and Southern New Jersey who, during the funding year, were 50 to 75 years old, had received care in CMCVAMC in the 18-48 months prior to selection for the study but who had not been seen by primary care in the 18 months prior to selection, and were due for CRC screening based on CPRS screening reminder. We excluded, via chart review, patients with documented gastrointestinal (GI) symptoms such as rectal bleeding, unexplained weight loss or change in bowel habits, first degree relative(s) with CRC, and a personal history of inflammatory bowel disease (IBD), or colon or rectal resection. We also excluded patients with evidence of prior colonoscopy within 10 years, sigmoidoscopy within 5 years, and fecal occult blood test (FOBT)/FIT within 1 year.

Interventions: The patients were blocked randomized in 9 blocks of 87 patients 1:1:1 to usual care, invitation-reminder, or mailed-FIT. Those randomized to usual care continued to receive the current practice at CMCVAMC of offering screening during an office visit. Those randomized to screening invitation-reminder received usual care, a mailed invitation letter to engage in CRC screening, 3 reminder phone attempts, and a mailed reminder. The invitation letter included lay-audience descriptions of screening tests such as FIT and endoscopy, and symptoms that should prompt diagnostic work-up such as change in bowel habit with rectal bleeding. The packet had instructions to contact the study team if the patient believed he/she was not eligible for screening at that time. The letter informed patients that a telephone reminder would follow in 4 weeks from invitation letter. They received three weekly reminders at weeks 4, 5, and 6 from the invitation letter. For the purposes of this intervention, Week 1 will be the week the screening invitation letter was sent (time zero). Those randomized to mail-FIT received a pre-notification letter with the CRC screening information, a mailed FIT kit, 3 reminder phone attempts, and a mailed reminder. Patients who did not return their kit after 4 weeks of the date the pre-notification letter was mailed received a live telephone reminder, followed by 2 additional calls at the end of weeks 5 and 6, if needed.

Outcomes: The primary outcomes were the rate of return of the mailed-FIT kits and the completion rate of any CRC screening within 6 months of the mailing of the screening invitation. We also collected data on a number of other outcomes, including the proportion of participants whose letters were returned, the proportion of patients who responded that they were actually up-to-date with CRC screening outside the VA, and the returned FIT positivity rate.

Power and sample size calculation: Studies have shown that among underserved patients 40% of those assigned to mailed-FIT completed screening compared to 12% in usual care (UC). We assumed 15% screening rate in usual care alone, 25% with invitation-reminders, and 40% with mailed-FIT. With those assumptions, a total sample size of 500 (250 per arm) will give 80% power for UC vs. screening invitation-reminder, and 152 patients per arm will give a power of 80% for the screening invitation-reminder vs. mailed-FIT. Accounting for the possibility that up to 20% of participants could be ineligible post-randomization, we concluded that we should enroll 783 patients, 261 per arm, for this trial.

Analysis: We used the Chi-squared statistic to test the null hypothesis that the completion rates among the arms were equal using an "intention-to-treat" approach, based on a 2-sided alpha of 0.05. We also compared the CRC screening completion rates among the arms excluding patients whose mailed invitations and/or FIT kits were returned to the study, as these patients never actually received the study interventions. We also compared the 6 month screening rate among arms excluding patients who reported that they were actually up to date with CRC screening at the start of the study (as determined by a combination of returned screening reminder cards, patient phone calls, and patient report during reminder phone calls and 6-month follow-up phone calls and review of the electronic medical record). Finally, the rate of CRC screening and mailed FIT return rate was compared based on age, gender, and race.

We reviewed 3,108 patient charts and identified 782 patients who were eligible for the study. One patient was identified and randomized twice which is why we randomized only 782 participants instead of 783. The eligible participant was randomized in two separate blocks into the usual care arm, thus the usual care arm only had 260 patients, compared to 261 randomized to mailed invitation and 261 randomized to mailed FIT test. There was no difference in age, race or gender among the arms. The median age in the usual care and mailed-invitation arms was 60, and in the mailed FIT arm was 61. In those respective arms, 62.5%, 57.1%, and 61.3% of the participants were African American (p=0.79). There was similarly no difference in gender breakdown with 3.1% of usual care, and 2.7% of both the mailed invitation and mailed FIT arms being female.

In the mail FIT arm, overall, 62/261 (23.8%) of participants returned the mailed FIT card. An additional 6 patients in this arm did not return the mailed FIT test, but completed CRC screening for a total CRC screening rate within 6 months of 68/261 (26.05%) In the usual care group, overall, 15/260 (5.8%) of the patient had CRC screening confirmed during 6 months' follow-up. In the mailed invitation arm, 20/261 (7.7%) completed CRC screening. The difference among arms was significant (p<0.001). The differences in completed CRC screening rate was significant between usual care and mailed FIT arms (p<0.001) and between mailed invitation and mailed FIT arms (p<0.001). There was no significant difference, however, between the mailed invitation and usual care arms (p=0.39).

In the mailed invitation arm, 21/261 (8.1%) of the mailed invitations were returned to the study office as undeliverable. If these participants were excluded from analysis, the rate of CRC screening completion in the mailed invitation arm was 20/240 (8.3%) In the mailed FIT arm, 27/261 (10.3%) were returned. If these participants were excluded, the rate of mailed FIT return was 62/234 (26.5%) and the rate of any CRC screening completion was 67/234 (28.6%). Again, the mailed FIT arm had a significantly higher rate of completed CRC screening in 6 months than the other two arms. The mailed invitation arm (8.3%), while higher than the usual care arm (5.8%), was not significantly different from it (P=0.26).

We identified patients via returned screening reminder cards, reports from reminder calls and 6-month follow-up calls, and 6-month review of the electronic medical record, who confirmed that, at the time the study began, they were actually up-to-date with CRC screening. These are patients who appropriately would not get screening during our 6-month study window as they were not due for screening during that period. In the usual care arm, mailed invitation arm and mailed FIT arm 6/260 (2.3%), 15/261 (5.8%) and 9/261 (3.5%) (p=0.11), reported being up to date with their screening, respectively. If these patients were excluded from analysis, along with the participants whose mailed was returned, the screening rates per the respective arms above were 5.5%, 8.9% and 28.0% (p<0.001). Interestingly, the rate of overall screening in the mailed FIT arm went down slightly (28.6 to 28.0%) with this exclusion. This is because 4 patients who reported being up to date with their CRC screening at the start of the study still returned their study mailed FIT cards.

Finally, we identified 5 patients who died; 4 in the mailed-FIT card were confirmed, by phone call reminder or 6-month follow-up, to have been dead at the start of the study. One patient in the usual care arm died during the 6-month follow-up period. In our final analysis, we excluded these patients who were already dead at the start of the study or who died during follow-up, as they were not eligible for CRC screening. Excluding these patients, along with the patients who reported they were already up-to-date with CRC screening at the start of the study and the patients whose mailed invitation or mailed FIT tests was returned, the completed 6 months CRC screening rates were 14/253 (5.5%) in the usual care arm, 20/225 (8.9%) in the mailed invitation arm, and 63/222(28.4%) in the mailed FIT arm (p<0.001).

In the overall, intention-to-treat analysis, there was no difference in rates of completed CRC screening at 6 months by race, age or gender. 13.1% of African Americans had completed CRC screening at 6 months compared to 13.3% of white patients and 12.2 % of other/unknown patients (p=0.98). By age group, 12% of participants 50-59 received screening, while 13.0% of those 60-65 and 14.9% of those 66-75 years old did (p=0.55). Finally, 9.1% of the 22 female participants received screening, compared to 13.3% of the male participants (p=0.57).

By evaluating the feasibility of a mailed FIT outreach CRC screening program, we can enhance the VA's capacity to, and efficiency at, providing screening for this preventable condition among all Veterans irrespective of their ability to make an office visit, and advance current efforts aimed at increasing or maintaining high CRC screening rates. We may enable randomized trials of interventions to increase screening use and enhance understanding of screening practices at VAMCs.

External Links for this Project

NIH Reporter

Grant Number: I21HX001747-01A1

Dimensions for VA

Dimensions for VA is a web-based tool available to VA staff that enables detailed searches of published research and research projects.

If you have VA-Intranet access, click here for more information

VA staff not currently on the VA network can access Dimensions by registering for an account using their VA email address.
    Search Dimensions for this project


None at this time.

DRA: Health Systems, Cancer
DRE: Treatment - Observational
Keywords: none
MeSH Terms: none

Questions about the HSR&D website? Email the Web Team.

Any health information on this website is strictly for informational purposes and is not intended as medical advice. It should not be used to diagnose or treat any condition.