Few studies have directly targeted nonparticipants in colorectal cancer screening to identify effective engagement strategies. We undertook a randomized controlled trial that targeted nonparticipants in a previous trial of average-risk subjects which compared participation rates for mailed invitations offering a fecal test, a blood test or a choice of either. Nonparticipants (n = 899) were randomized to be offered a kit containing a fecal immunochemical test (FIT), directions on how to arrange a blood DNA test, or the option of doing either. Screening participation was assessed 12 weeks after the offer. To assess the cognitive and attitudinal variables related to participation and invitee choice, invitees were surveyed after 12 weeks, and associations were investigated using multinomial logistic regression. Participation rates were similar between groups (P = 0.88): 12.0% for FIT (35/292), 13.3% for the blood test (39/293), and 13.4% for choice (39/290). Within the choice group, participation was significantly higher with FIT (9.7%, 28/290) compared with the blood test (3.8%, 11/290, P = 0.005). The only variable significantly associated with participation was socioeconomic status when offered FIT, and age when offered choice but there was none when offered the blood test. Survey respondents indicated that convenience, time-saving, comfort, and familiarity were major influences on participation. There was no clear advantage between a fecal test, blood test, or choice of test although, when given a choice, the fecal test was preferred. Differences in variables associated with participation according to invitation strategy warrant consideration when deciding upon an invitation strategy for screening nonparticipants.

Prevention Relevance:

This trial of screening for those at average risk for colorectal cancer targeted past fecal-test nonparticipants and compared participation rates for mailed invitations offering a fecal test, blood test, or choice of either. Although there was no clear advantage between strategies, factors associated with participation differed between each strategy.

Addressing barriers to participation in colorectal cancer screening remains a major public health priority given less than optimal uptake and reparticipation in current screening programs. Until strategies are identified that maximize participation rates, the potential benefits of screening at the population level will not be recognized (1, 2). The nature of the barriers to screening has been well described in the literature. They include problems with the accessibility of the necessary health care services (e.g., urban, rural, and remote location of residence); belief in the value of screening, which is associated with stage of readiness to screen; exposure to trusted advocacy; personal responses to the process involved, such as a distaste for certain types of testing, and inconvenience associated with undertaking the screening test and associated processes; and difficulties with access to follow-up (3–16). Willingness to screen can also be linked to specific demographic features, namely, economic status, age, gender, and education (17–20).

Despite the identification of these barriers and implementation of efforts appropriate to a target population to overcome them, the issue of how to engage more people, especially established nonparticipants, has not been adequately addressed. The majority of strategies demonstrated to improve participation have been tested in an environment that has included a substantial number of people who are already prepared to screen, or are close to being willing to screen. They have not focused on those who have not taken up a previous offer of screening.

Existing evidence shows that an important barrier to participation is the nature of the screening test (9, 10, 21) and the demands it places on the potential participant. For example, it is now widely accepted that people find the guaiac fecal occult blood tests less acceptable than fecal immunochemical tests (FIT) because of the dietary demands and extra sampling demands of the former (22). Even the FIT requires a significant commitment; the very nature of a fecal test and the distaste of feces among some people are established barriers (10). Thus, an alternative biological sample might improve participation, especially in nonparticipants previously offered FIT (13). A difficulty with this approach might be inconvenience—a blood test could pose a logistic barrier because it cannot be undertaken at home.

Targeting people known to have refused to screen with FIT, and providing a blood-test option, provide data on the potential to “rescue” nonparticipants through the offer of a different form of screening. The addition of a choice option, where participants choose between a blood test and FIT, further empowers the health consumer and facilitates personalization in mass population screening programs (23). Some research already suggests that people are likely to prefer a blood test to a fecal test when presented with a hypothetical choice (9). In a recent randomized controlled trial targeting a typical average risk (age 50–75 years without symptoms or personal or family history of neoplasia) screening population with noninvasive tests, we found that mailing an offer of a choice between a blood test and FIT did not lead to better participation compared with a group offered only FIT (13). However, that study did not objectively test in a randomized manner how nonparticipants might respond to a later, “rescue” screening offer that directly compared the options for engaging them and included a choice.

Consequently, we undertook a randomized controlled trial (RCT) of screening that targeted nonparticipants in a previous trial of average-risk people and which compared participation rates for mailed invitations offering a fecal test, a blood test, or a choice of either. To better understand the demographic, cognitive and attitudinal variables underlying the choices that previous nonparticipants make when deciding to take up this subsequent offer we surveyed invitees 12 weeks after the offer of screening regardless of their participatory status.

Study design and population

The study population comprised nonparticipants in a previous RCT of individuals at average risk for colorectal cancer. Results of the first round of that trial have been reported in detail (13). Initially, 1,800 study invitees ages 50–74 years were randomly selected from six South Australian electorate regions from the Australian electoral roll and provided with an opportunity for colorectal cancer screening with either a FIT or a circulating tumor DNA blood test. As detailed below, 899 individuals who did not participate in the screening offer, and who had not otherwise withdrawn, were available 22 months later for this study. These nonparticipants were randomized (via a random number generator) to be mailed an invitation to screen with one of three interventions: a kit containing a FIT, directions on how to arrange a blood test, or the option of doing either (with advice relevant to both so that they could choose).

All invitees were sent an advanced notification letter (24), informing them of the purpose of the study and reminding them about their nonparticipation in the first round; this letter did not identify the group to which they had been randomized. Two weeks later, they were invited to screen depending on their assigned study group and provided they had not indicated a desire to withdraw. Six weeks after the invitation, a reminder letter was sent to nonparticipants (5). After 12 weeks, individuals, regardless of participatory status, were mailed a follow-up survey (details below) and participation rates for the different groups were determined (Fig. 1).

Figure 1.

Disposition of study subjects commencing with nonparticipants from the previous trial of colorectal cancer screening in average-risk subjects. FIT, fecal immunochemical test.

Figure 1.

Disposition of study subjects commencing with nonparticipants from the previous trial of colorectal cancer screening in average-risk subjects. FIT, fecal immunochemical test.

Close modal

Ethical approval for the study was obtained from the Southern Adelaide Clinical Human Research Ethics Committee (reference number 483.14). The trial was registered with the Australian and New Zealand Clinical Trials Registry (ACTRN 12615000972527) prior to enrolment of the first participant. Written informed consent was obtained from all subjects and the study was conducted in accordance with the Declaration of Helsinki.

Screening tests

All screening test offers were posted to the invitee's home address. Screening test kits included an invitation letter, information sheet about the study and option for withdrawing, a consent form and a form for collection of participant details including demographics. Demographic data sought were: age, gender, marital status, education, employment status, country of birth, cultural identity, health insurance status, and postcode. The latter was used to allow area-based determination of an index of relative socioeconomic disadvantage (IRSD) information (25). The IRSD is an area-based assessment of disadvantage by comparing residential area with census data on income.

The FIT kit included two sample collection tubes (OC Sensor, Eiken Chemical Company) and collection sheets, instructions on how to complete the test, and a reply-paid return envelope for return of the sample tubes and forms to the laboratory (Flinders Centre for Innovation in Cancer, Bedford Park, SA) by mail, as previously described (26) and consistent with Australian policy for screening. Invitees were instructed to sample from two different bowel motions and return the samples within 10 days. Samples collected incorrectly were not analyzed, and a replacement kit was sent for the opportunity to repeat the test, but receipt of any completed test was still classed as participation in FIT screening. Analysis of FIT collection tubes for hemoglobin concentration followed manufacturer instructions, and results were reported to participants as previously described (26).

The blood test kit included a referral form for the test, and a list of blood collection center locations with directions on how to contact one that was conveniently located. Invitees were informed that the test was new and accuracy for colorectal cancer detection was similar performance to the FIT in that it would detect at least 60% of colon and rectum cancers (14). To be classed as a participant for screening with the blood test, the study invitee had to attend a blood collection center where blood was collected (K3-EDTA). Plasma was separated from the blood sample and couriered to Clinical Genomics for assay of methylated BCAT1 and IKF1 (Colvera, Clinical Genomics Pty Ltd) as previously described (27).

The choice group received both sets of materials, i.e., a FIT kit plus directions on how to arrange the blood test, and a clear indication that they could choose whichever suited them.

Survey

The survey was provided in paper form 12 weeks after the invitation, along with a reply-paid envelope for return. Invitees were encouraged to complete the survey regardless of whether or not they participated in screening. It included a series of statements covering four issues. Items addressing why nonparticipants chose not to screen, such as lack of interest, recent colonoscopy, or other bowel test or examination, perceived inconvenience (test process or logistics), need for doctors recommendation, embarrassment, lack of symptoms or of family history of disease, were adapted from Todorov and colleagues (28). Those addressing the decision to screen, including decision regret, were adapted from Brehaut and colleagues (29). Factors contributing to choice of test and intention to screen in future studies were user defined based on our previous experience with participation studies and published literature on barriers. Participants indicated extent to which these factors contributed with response options including “Very Much, Slightly, Not at All, or Not Applicable.” For the Choice group, the survey also addressed items relevant to why participants might have chosen a particular test, specifically; convenience, comfort, time-saving, familiarity, perceived doctor's or peer's recommendation, pleasantness or disgust, fear of clinics or needles and dislike of blood or feces. These responses were assessed on a 5-point Likert scale (Very Large Extent, Large Extent, Moderate Extent, Small Extent and Not At All).

Endpoint assessment and analyses

Screening participation rates at 12 weeks (return of FIT kit or attendance for blood sampling) were compared blinded between groups using X2 tests. The association between screening outcome and sociodemographic characteristics in each study group was examined using a Fisher exact test given the small sample size in certain subgroups. Three key sociodemographic characteristics included in the analyses were age (in 2016), gender, and IRSD. For IRSD, a higher score indicates a relative lack of disadvantage (30). IRSD scores were categorized into two groups (high or low relative to the median level). The associations between participants' sociodemographic characteristics as well as the test invitation strategy in the previous trial and screening outcome in each study group were also studied using binary logistic regression analyses (with odds ratios reported). A probability of P < 0.05 (two-tailed) was considered statistically significant. The sample size (n≈300) was based on a FIT participation rate of 12% and gives 80% power to detect a difference (5%; two-sided) if participation in a test group is 20%, where this magnitude of difference would be considered a worthwhile gain. All analyses were performed with Stata Statistical Software (version 16).

Study population

Disposition of study invitees is shown in Fig. 1. Withdrawals were similar between the groups and were either at the request of the invitees, because recent medical events rendered them unsuitable for screening, or because inability to locate them was confirmed. Demographics of invitees in each study group are shown in Table 1. There were no statistically significant differences between the groups in age (χ2(8) = 5.058, P = 0.751), gender distribution (χ2(2) = 0.854, P = 0.652), or socioeconomic status (χ2(2) = 0.829, P = 0.661).

Table 1.

Demographic characteristics of invitees (all of whom were previous screening nonparticipants) according to study group to which they were randomized.

Randomized to fecal immunochemical test (n = 297)Randomized to blood test (n = 301)Randomized to choice (n = 301)
Intervention(%, 95% CI)(%, 95% CI)(%, 95% CI)
Age range (years) 
 50–55 71 (23.9%; 19.4%–29.1%) 76 (25.2%; 20.6%–30.5%) 74 (24.6%; 20.0%–29.8%) 
 56–60 71 (23.9%; 19.4%–29.1%) 71 (23.6%; 19.1%–28.7%) 64 (21.3%; 17.0%–26.2%) 
 61–65 60 (20.2%; 16.0%–25.2%) 58 (19.3%; 15.2%–24.1%) 62 (20.6%; 16.4%–25.6%) 
 66–70 41 (13.8%; 10.3%–18.2%) 54 (17.9%; 14.0%–22.7%) 56 (18.6%; 14.6%–23.4%) 
 71–74 54 (18.2%; 14.2%–23.0%) 42 (14.0%; 10.5%–18.4%) 45 (15.0%; 11.3%–19.5%) 
Gender 
 Males 164 (55.2%; 49.4%–61.0%) 155 (51.5%; 45.7%–57.3%) 159 (52.8%; 47.0%–58.6%) 
 Females 133 (44.8%; 39.2%–50.5%) 146 (48.5%; 42.9%–54.2%) 142 (47.2%; 41.6%–52.8%) 
Relative socioeconomic disadvantage 
 Most disadvantaged 112 (37.7%; 32.4%–43.4%) 115 (38.2%; 32.9%–43.8%) 105 (34.9%; 29.7%–40.5%) 
 Least disadvantaged 185 (62.3%; 56.6%–67.6%) 186 (61.8%; 56.2%–67.2%) 196 (65.1%; 59.5%–70.3%) 
Randomized to fecal immunochemical test (n = 297)Randomized to blood test (n = 301)Randomized to choice (n = 301)
Intervention(%, 95% CI)(%, 95% CI)(%, 95% CI)
Age range (years) 
 50–55 71 (23.9%; 19.4%–29.1%) 76 (25.2%; 20.6%–30.5%) 74 (24.6%; 20.0%–29.8%) 
 56–60 71 (23.9%; 19.4%–29.1%) 71 (23.6%; 19.1%–28.7%) 64 (21.3%; 17.0%–26.2%) 
 61–65 60 (20.2%; 16.0%–25.2%) 58 (19.3%; 15.2%–24.1%) 62 (20.6%; 16.4%–25.6%) 
 66–70 41 (13.8%; 10.3%–18.2%) 54 (17.9%; 14.0%–22.7%) 56 (18.6%; 14.6%–23.4%) 
 71–74 54 (18.2%; 14.2%–23.0%) 42 (14.0%; 10.5%–18.4%) 45 (15.0%; 11.3%–19.5%) 
Gender 
 Males 164 (55.2%; 49.4%–61.0%) 155 (51.5%; 45.7%–57.3%) 159 (52.8%; 47.0%–58.6%) 
 Females 133 (44.8%; 39.2%–50.5%) 146 (48.5%; 42.9%–54.2%) 142 (47.2%; 41.6%–52.8%) 
Relative socioeconomic disadvantage 
 Most disadvantaged 112 (37.7%; 32.4%–43.4%) 115 (38.2%; 32.9%–43.8%) 105 (34.9%; 29.7%–40.5%) 
 Least disadvantaged 185 (62.3%; 56.6%–67.6%) 186 (61.8%; 56.2%–67.2%) 196 (65.1%; 59.5%–70.3%) 

Abbreviation: CI, confidence interval.

Participation in screening

Prior to sending the screening test, 24 individuals opted out of the study, leaving 875 to be invited. On an intention-to-screen basis (including all 899 randomized individuals). Overall, a low proportion of those who were randomized completed screening in this offer (12.6%, 113/899; Fig. 1). The screening participation rate in those who were invited did not differ significantly between the groups: offer of FIT [12.0%; 95% CI 8.7–16.3% (35/292)], blood test [13.3%; 95% CI 9.8–17.7% (39/293)], or choice of test [13.4%; 95% CI 10.0–17.9% (39/290); P = 0.884]. Within the choice group, significantly more individuals participated with the FIT (9.7%; 95% CI 6.8–13.6%, 28/290) compared with the blood test (3.8%; 95% CI 2.1–6.7%, 11/290, P = 0.005; Supplementary Fig. S1). All who participated with FIT returned both samples.

Demographics for participants in each study group who received a screening invitation are shown in Table 2. Overall, participation rates regardless of randomization group increased significantly with age; the unadjusted OR compared with those ages 50–55 years were 2.17 (95% CI 1.18–4.01, P = 0.013) for 61–65 years and rose progressively to 2.44 (95% CI 1.29–4.61, P = 0.006) for 71 years and older. Gender did not differ between participants and nonparticipants, but socioeconomic status did (least disadvantaged unadjusted OR 1.73, 95% CI 1.11–2.69, P = 0.015 compared with most disadvantaged).

Table 2.

Demographics for participants in each study group. Screening rates are for the n = 875 that received a screening test invitation.

No. of participants (%)
InterventionOffered FIT (n = 292)Offered blood test (n = 293)Offered choice (n = 290)
Age range (years) 
 50–55 5/71 (7.0%) 10/74 (13.5%) 4/72 (5.6%) 
 56–60 8/71 (11.3%) 5/71 (7.0%) 9/61 (14.8%) 
 61–65 12/58 (20.7%) 9/55 (16.4%) 9/61 (14.8%) 
 66–70 4/40 (10.0%) 7/51 (13.7%) 5/53 (9.4%) 
 71–74 6/52 (11.5%) 8/42 (19.0%) 12/43 (27.9%)b 
Gender 
 Male 17/161 (10.5%) 24/151 (15.9%) 21/152 (13.8%) 
 Female 18/131 (13.7%) 15/142 (10.6%) 18/138 (13.0%) 
Relative socioeconomic disadvantage 
 Most disadvantaged 6/110 (5.5%)a 11/114 (9.6%) 13/99 (13.1%) 
 Least disadvantaged 29/182 (15.9%)a 28/179 (15.6%) 26/191 (13.6%) 
No. of participants (%)
InterventionOffered FIT (n = 292)Offered blood test (n = 293)Offered choice (n = 290)
Age range (years) 
 50–55 5/71 (7.0%) 10/74 (13.5%) 4/72 (5.6%) 
 56–60 8/71 (11.3%) 5/71 (7.0%) 9/61 (14.8%) 
 61–65 12/58 (20.7%) 9/55 (16.4%) 9/61 (14.8%) 
 66–70 4/40 (10.0%) 7/51 (13.7%) 5/53 (9.4%) 
 71–74 6/52 (11.5%) 8/42 (19.0%) 12/43 (27.9%)b 
Gender 
 Male 17/161 (10.5%) 24/151 (15.9%) 21/152 (13.8%) 
 Female 18/131 (13.7%) 15/142 (10.6%) 18/138 (13.0%) 
Relative socioeconomic disadvantage 
 Most disadvantaged 6/110 (5.5%)a 11/114 (9.6%) 13/99 (13.1%) 
 Least disadvantaged 29/182 (15.9%)a 28/179 (15.6%) 26/191 (13.6%) 

Abbreviation: FIT, fecal immunochemical test.

aTwo statistically significant relationships were found (Fisher exact test): Between relative socioeconomic disadvantage and the screening outcome in the FIT group (P = 0.009).

bTwo statistically significant relationships were found (Fisher exact test): Between age and the screening outcome in the Choice group (P = 0.018).

Comparison of demographics and survey variables between screening participants and nonparticipants, regardless of the study group, is shown in Table 3. Age, socioeconomic status, and initial testing strategy offered in the first trial were significantly associated with participation. Randomized group membership showed a trend (P = 0.085), while higher education, fulltime work, marital status, private medical insurance and born in Australia were not significant (Supplementary Table S1).

Table 3.

Comparison of demographics and survey variables between screening participants and nonparticipants, regardless of study group.

Participation status
VariableParticipant (n = 113)Nonparticipant (n = 762)P value (χ2)
Age range (years) P = 0.018 
 50–55 19 (16.8%) 198 (26.0%)  
 56–60 22 (19.5%) 181 (23.8%)  
 61–65 30 (26.5%) 144 (18.9%)  
 66–70 16 (14.2%) 128 (16.8%)  
 71–74 26 (23.0%) 111 (14.6%)  
Gender P = 0.675 
 Males 62 (54.9%) 402 (52.8%)  
 Females 51 (45.1%) 360 (47.2%)  
Socioeconomic status P = 0.014 
 Most disadvantaged 30 (26.5%) 293 (38.5%)  
 Least disadvantaged 83 (73.5%) 469 (61.5%)  
Initial test offered in the previous trial P = 0.012 
 FIT only 61 (54.0%) 504 (66.1%)  
 Choice of FIT or blood test 52 (46.0%) 258 (33.9%)  
Group of randomization P = 0.0845 
 FIT 35 (31.0%) 257 (33.7%)  
 Blood test 39 (34.5%) 254 (33.3%)  
 Choice 39 (34.5%) 251 (32.9%)  
Participation status
VariableParticipant (n = 113)Nonparticipant (n = 762)P value (χ2)
Age range (years) P = 0.018 
 50–55 19 (16.8%) 198 (26.0%)  
 56–60 22 (19.5%) 181 (23.8%)  
 61–65 30 (26.5%) 144 (18.9%)  
 66–70 16 (14.2%) 128 (16.8%)  
 71–74 26 (23.0%) 111 (14.6%)  
Gender P = 0.675 
 Males 62 (54.9%) 402 (52.8%)  
 Females 51 (45.1%) 360 (47.2%)  
Socioeconomic status P = 0.014 
 Most disadvantaged 30 (26.5%) 293 (38.5%)  
 Least disadvantaged 83 (73.5%) 469 (61.5%)  
Initial test offered in the previous trial P = 0.012 
 FIT only 61 (54.0%) 504 (66.1%)  
 Choice of FIT or blood test 52 (46.0%) 258 (33.9%)  
Group of randomization P = 0.0845 
 FIT 35 (31.0%) 257 (33.7%)  
 Blood test 39 (34.5%) 254 (33.3%)  
 Choice 39 (34.5%) 251 (32.9%)  

Abbreviation: FIT, fecal immunochemical test.

Table 4 shows demographic comparisons between those who completed a FIT test and those who did a blood test, irrespective of which group they were randomized to; there were no obvious differences in age, gender, IRSD or test exposure in the previous trial between FIT and blood test participants.

Table 4.

Comparison of demographics in those who participated by FIT vs. those who participated by blood test, regardless of the study group.

VariableFIT participant (n = 63)Blood test participant (n = 50)P value (χ2)
Age range (years) 0.545 
 50–55 8 (12.7%) 11 (22.0%)  
 56–60 15 (23.8%) 7 (14.0%)  
 61–65 17 (27.0%) 13 (26.0%)  
 66–70 8 (12.7%) 8 (16.0%)  
 71–74 15 (23.8%) 11 (22.0%)  
Gender 0.830 
 Males 34 (54.0%) 28 (56.0%)  
 Females 29 (46.0%) 22 (44.0%)  
Socioeconomic status 0.459 
 Most disadvantaged 15 (23.8%) 15 (30.0%)  
 Least disadvantaged 48 (76.2%) 35 (70.0%)  
Initial test offered in the previous trial 0.701 
 FIT only 33 (52.4%) 28 (56.0%)  
 Choice of FIT or blood test 30 (47.6%) 22 (44.0%)  
VariableFIT participant (n = 63)Blood test participant (n = 50)P value (χ2)
Age range (years) 0.545 
 50–55 8 (12.7%) 11 (22.0%)  
 56–60 15 (23.8%) 7 (14.0%)  
 61–65 17 (27.0%) 13 (26.0%)  
 66–70 8 (12.7%) 8 (16.0%)  
 71–74 15 (23.8%) 11 (22.0%)  
Gender 0.830 
 Males 34 (54.0%) 28 (56.0%)  
 Females 29 (46.0%) 22 (44.0%)  
Socioeconomic status 0.459 
 Most disadvantaged 15 (23.8%) 15 (30.0%)  
 Least disadvantaged 48 (76.2%) 35 (70.0%)  
Initial test offered in the previous trial 0.701 
 FIT only 33 (52.4%) 28 (56.0%)  
 Choice of FIT or blood test 30 (47.6%) 22 (44.0%)  

Demographic associations with participation varied according to the test invitation strategy (FIT, blood, or choice) are shown in Table 5. Overall, age and IRSD were associated with likelihood of screening but associations differed within a test invitation strategy. For FIT invitees only IRSD was significant, whereas only oldest age was a significant association for those offered choice. No associations were seen in blood test invitees.

Table 5.

Relative likelihood of screening according to demographic characteristics, within each randomization group.

Randomization group
AllFITBlood testChoice
ORSEORSEORSEORSE
Age range, reference 50–55 years 
 56–60 1.255 [0.417] 1.282 [0.789] 0.492 [0.287] 3.054 [1.938] 
61–65 2.012 [0.636]a 2.604 [1.521] 1.096 [0.558] 2.798 [1.765] 
 66–70 1.269 [0.457] 1.077 [0.775] 1.035 [0.559] 1.819 [1.273] 
71–74 2.494 [0.817]b 1.511 [0.992] 1.214 [0.650] 7.246 [4.546]b 
Gender: Male 1.100 [0.226] 0.771 [0.292] 1.586 [0.572] 1.111 [0.397] 
Socioeconomic score (IRSD) high (least disadvantaged) 1.721 [0.394]a 3.317 [1.582]a 1.694 [0.652] 1.323 [0.513] 
Initial test offered in previous trial, reference FIT 
 Blood 1.198 [0.303]       
 Choice 1.209 [0.307]       
N 875  292  293  290  
Randomization group
AllFITBlood testChoice
ORSEORSEORSEORSE
Age range, reference 50–55 years 
 56–60 1.255 [0.417] 1.282 [0.789] 0.492 [0.287] 3.054 [1.938] 
61–65 2.012 [0.636]a 2.604 [1.521] 1.096 [0.558] 2.798 [1.765] 
 66–70 1.269 [0.457] 1.077 [0.775] 1.035 [0.559] 1.819 [1.273] 
71–74 2.494 [0.817]b 1.511 [0.992] 1.214 [0.650] 7.246 [4.546]b 
Gender: Male 1.100 [0.226] 0.771 [0.292] 1.586 [0.572] 1.111 [0.397] 
Socioeconomic score (IRSD) high (least disadvantaged) 1.721 [0.394]a 3.317 [1.582]a 1.694 [0.652] 1.323 [0.513] 
Initial test offered in previous trial, reference FIT 
 Blood 1.198 [0.303]       
 Choice 1.209 [0.307]       
N 875  292  293  290  

Note: Binary logit model estimates. Dependent variable is a dummy variable which equals to 1 if screened and 0 otherwise.

Odds ratios (ORs) reported.

Abbreviation: IRSD, Index of Relative Socioeconomic Disadvantage.

aP < 0.05.

bP < 0.01.

Factors contributing to the choice of test

The survey included questions seeking to establish factors contributing to test choice among those offered choice, noting that follow-up surveys were received from only 20 of the 39 participants and 6 of the 251 nonparticipants in the choice group. Supplementary Table S2 shows the degree to which a factor influenced the decision to choose one or the other test (FIT or blood) or to not screen at all. For FIT and blood participants, convenience, comfort, familiarity and time-saving were a large or very large influence in 50% or more. Pleasantness was, however, rated as large or very large in just 18.2% of FIT participants compared with 62.5% of blood test participants. In addition, 25% of blood test participants cited a dislike of feces as a reason for their choice in test. Physician recommendation was a large or very large influence in only 33.3% of FIT participants, and in none of the blood test participants. Response rate was low (n = 6) in nonparticipants but two-thirds considered convenience and lack of physician recommendation as large or very large influences on the decision not to screen.

Reasons for nonparticipation in screening

In the survey, there were a series of questions addressing reasons for nonparticipation among those who chose not to screen. Relatively few nonparticipants completed the survey (only 25/786, 3.2%, compared with 71/113, 62.8% participants). The most frequent reasons were relatively recent colonoscopy (7/25, 28%), or other bowel examination (7/25). Five respondents said that their lack of symptoms contributed to some extent, and five said that they were busy or lacked time. Interestingly, 21/25 (84%) agreed or strongly agreed that they made the right decision by not participating and 17/25 (68.0%) of these nonparticipants indicated that it was likely that they would do a FIT if offered one in the future, with 20/25 (80.0%) indicating that it was likely that they would do a blood test for screening if offered one in the future. If offered a choice between the two tests, 13 (52%) said it was likely they would choose FIT, and 19 (76%) said it was likely that they would choose blood test.

Engaging nonparticipants in colorectal cancer screening remains a public health challenge when seeking to optimize population participation rates. Strategies aiming to “rescue” this often-sizable proportion of the population have received little attention. To test strategies that might deal with barriers to screening in the context of mailed invitations from a central screening coordinating center, we invited nonparticipants in a previous trial of FIT-based screening to participate in a study to determine if a strategy avoiding fecal sampling (i.e., a blood test) or providing a choice of either blood test or FIT, rather than a single test option, might successfully engage these noncompliers.

There was no participatory advantage for any strategy, with rates falling in the range 12.0% to 13.3%. This overall success rate of 1 in 8 in noncompliers is of the order that might have been expected from results in studies where the entire population was invited to screen in the second screening round, whether or not they had participated in the first (31–33). Even though fecal aversion is an established barrier to FIT-based screening, provision of a blood test did not clearly overcome the barrier. This might have been because the logistics of a blood test is more complex in this screening context; the test cannot be done at home. It might also be because FIT-based screening is well known and is public health policy in South Australia, where the study was conducted, whereas the blood test is a novel idea for colorectal cancer screening in this population and awareness of a blood test for cancer screening is likely much lower (13). These confounding factors could explain why those given a choice were more willing to do FIT. If screening were to be offered at the time a person was seeing a health practitioner with conveniently located venesection services, the result could be quite different. Certainly, those who did participate rated convenience as an important factor in undertaking screening, whether they returned a FIT or a blood test.

Older age and socioeconomic advantage were associations with participation, consistent with many previous studies (17–20, 34). Although the demographics of the study groups were equivalent, the association of the measured variables with participation was found to differ between invitation strategies. Only socioeconomic advantage was significantly associated with participation when offered FIT, while neither age nor socioeconomic advantage was associated with participation when offered the blood test. Age was an association when offered a choice. Such differences seem likely to reflect the logistic, cognitive and behavioral differences between these test types and they warrant consideration when deciding upon an invitation strategy for prior nonparticipants.

Offering a choice of screening test is supported in some guidelines; for example, the American Cancer Society advocates for invitees to be given the opportunity to make an “informed” choice of test (23, 35). Nonetheless, we saw no advantage in providing choice to noncompliers, consistent with several previous studies of unselected populations (13, 36). Neither the provision of choice nor the nature of the alternative option (designed to circumvent fecal aversion) achieved better participation in these screening noncompliers. It is inevitable that different screening strategies present differing challenges. Here, the convenience of the home fecal test compared with the time and expenses associated with the requirement to attend for venesection are competing influences and they are obviously viewed differently by population subgroups. The blood test added complexity to screening logistics when screening is offered by mail, whereas it would not do so if screening were to be offered during attendance at a clinic. Likewise, an older established test would have more credibility than a new test, although others might respond positively to the idea of a new technology. Unfortunately, assessing the relative impact of these factors in noncompliers is very difficult since survey responses were low in this group, especially for nonparticipants.

A trial such as this presents certain weaknesses and results are clearly confined to the context of mailing an invitation to a mass population selected only because of age. The observed effect on participation is not necessarily relevant to personalized screening delivered through direct discussion with a health professional. Test logistics differed greatly and nothing is known about why an individual did not screen in the first place. But in practice, neither of these can be overcome. Survey responses were also low and power to detect meaningful associations was severely limited, but this is a difficulty consistently observed when targeting noncomplying populations.

Strengths of the study were the focus on noncompliers, the comparability of study groups, and the lack of bias that might have resulted from prior exposure to a test in the previous trial. Results are directly relevant to mass screening by mail, a standard practice in many countries (1, 2).

In conclusion, the provision of different screening options to FIT nonparticipants “rescued” 1 in 8 invitees. There was no clear advantage between a FIT, blood test, or choice of test although, when given a choice, FIT was preferred in this population familiar with a mailed invitation to screen with FIT. Demographic and attitudinal variables associated with participation differed between those invited to screen with a fecal test, a blood test or a choice of either and these differences warrant consideration when choosing between invitation strategies. Although the provision of choice did not improve overall screening participation, it did provide the opportunity for nonparticipants to select the test that they prefer, which may result in improved ongoing participation.

G.P. Young reports personal fees from Clinical Genomics Inc NJ USA and grants from Eiken Chemical Co. Japan during the conduct of the study. G. Chen reports grants from National Health and Medical Research Council (NHMRC) of Australia during the conduct of the study. E.L. Symonds reports grants from NHMRC during the conduct of the study; grants from Eiken Chemical Company and grants from Clinical Genomics outside the submitted work. No disclosures were reported by the other authors.

G.P. Young: Conceptualization, supervision, funding acquisition, writing–review and editing. G. Chen: Conceptualization, formal analysis, writing–review and editing. C.J. Wilson: Conceptualization, funding acquisition, writing–review and editing. E. McGrane: Formal analysis, writing–original draft, writing–review and editing. D.L. Hughes-Barton: Data curation, validation, writing–review and editing. I.H.K. Flight: Conceptualization, writing–review and editing. E.L. Symonds: Conceptualization, data curation, formal analysis, funding acquisition, writing–original draft, writing–review and editing.

Financial support was provided by NHMRC Australia Project Grant (APP1101837; to E.L. Symonds, G.P. Young, G. Chen, and C.J. Wilson. Eiken Chemical Company Tokyo, Japan, provided institutional support for E.L. Symonds.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

1.
Schreuders
EH
,
Ruco
A
,
Rabeneck
L
,
Schoen
RE
,
Sung
JJ
,
Young
GP
, et al
Colorectal cancer screening: a global overview of existing programmes
.
Gut
2015
;
64
:
1637
49
.
2.
Young
GP
,
Rabeneck
L
,
Winawer
SJ
. 
The global paradigm shift in screening for colorectal cancer
.
Gastroenterology
2019
;
156
:
843
51
.
3.
AIHW
.
National bowel cancer screening program monitoring report: 2018
.
Canberra
:
Australian Institute of Health and Welfare Canberra
; 
2018
.
4.
Cole
SR
,
Gregory
T
,
Whibley
A
,
Ward
P
,
Turnbull
D
,
Wilson
C
, et al
Predictors of re-participation in faecal occult blood test-based screening for colorectal cancer
.
Asian Pac J Cancer Prev
2012
;
13
:
5989
94
.
5.
Cole
SR
,
Young
GP
,
Esterman
A
,
Cadd
B
,
Morcom
J
. 
A randomised trial of the impact of new faecal haemoglobin test technologies on population participation in screening for colorectal cancer
.
J Med Screen
2003
;
10
:
117
22
.
6.
Cole
SR
,
Zajac
I
,
Gregory
T
,
Mehaffey
S
,
Roosa
N
,
Turnbull
D
, et al
Psychosocial variables associated with colorectal cancer screening in South Australia
.
Int J Behav Med
2011
;
18
:
302
9
.
7.
May
FP
,
Almario
CV
,
Ponce
N
,
Spiegel
BM
. 
Racial minorities are more likely than whites to report lack of provider recommendation for colon cancer screening
.
Am J Gastroenterol
2015
;
110
:
1388
94
.
8.
Moons
L
,
Mariman
A
,
Vermeir
P
,
Colemont
L
,
Clays
E
,
Van Vlierberghe
H
, et al
Sociodemographic factors and strategies in colorectal cancer screening: a narrative review and practical recommendations
.
Acta Clin Belg
2002
;
75
:
33
41
.
9.
Osborne
J
,
Wilson
C
,
Moore
V
,
Gregory
T
,
Flight
I
,
Young
G
. 
Sample preference for colorectal cancer screening tests: blood or stool?
OJPM
2012
;
2
:
326
31
.
10.
Osborne
JM
,
Flight
I
,
Wilson
CJ
,
Chen
G
,
Ratcliffe
J
,
Young
GP
. 
The impact of sample type and procedural attributes on relative acceptability of different colorectal cancer screening regimens
.
Patient Prefer Adherence
2018
;
12
:
1825
36
.
11.
Sheeran
P
. 
Intention-behavior relations: a conceptual and empirical review
.
Eur Rev Soc Psychol
2002
;
12
:
1
36
.
12.
Symonds
EL
,
Cock
C
,
Meng
R
,
Cole
SR
,
Fraser
RJL
,
Young
GP
. 
Uptake of a colorectal cancer screening blood test in people with elevated risk for cancer who cannot or will not complete a faecal occult blood test
.
Eur J Cancer Prev
2018
;
27
:
425
32
.
13.
Symonds
EL
,
Hughes
D
,
Flight
I
,
Woodman
R
,
Chen
G
,
Ratcliffe
J
, et al
A randomized controlled trial testing provision of fecal and blood test options on participation for colorectal cancer screening
.
Cancer Prev Res
2019
;
12
:
631
40
.
14.
Symonds
EL
,
Pedersen
SK
,
Baker
RT
,
Murray
DH
,
Gaur
S
,
Cole
SR
, et al
A blood test for methylated BCAT1 and IKZF1 vs. a fecal immunochemical test for detection of colorectal neoplasia
.
Clin Transl Gastroenterol
2016
;
7
:
e137
.
15.
Taber
JM
,
Aspinwall
LG
,
Heichman
KA
,
Kinney
AY
. 
Preferences for blood-based colon cancer screening differ by race/ethnicity
.
Am J Health Behav
2014
;
38
:
351
61
.
16.
de Klerk
CM
,
Gupta
S
,
Dekker
E
,
Essink-Bot
ML
Expert Working Group ‘Coalition to reduce inequities in colorectal cancer screening’ of the World Endoscopy O. Socioeconomic and ethnic inequities within organised colorectal cancer screening programmes worldwide
.
Gut
2018
;
67
:
679
87
.
17.
Klabunde
C
,
Blom
J
,
Bulliard
JL
,
Garcia
M
,
Hagoel
L
,
Mai
V
, et al
Participation rates for organized colorectal cancer screening programmes: an international comparison
.
J Med Screen
2015
;
22
:
119
26
.
18.
Lo
SH
,
Halloran
S
,
Snowball
J
,
Seaman
H
,
Wardle
J
,
von Wagner
C
. 
Colorectal cancer screening uptake over three biennial invitation rounds in the English bowel cancer screening programme
.
Gut
2015
;
64
:
282
91
.
19.
Singh
H
,
Bernstein
CN
,
Samadder
JN
,
Ahmed
R
. 
Screening rates for colorectal cancer in Canada: a cross-sectional study
.
CMAJ Open
2015
;
3
:
E149
57
.
20.
Smith
SG
,
McGregor
LM
,
Raine
R
,
Wardle
J
,
von Wagner
C
,
Robb
KA
. 
Inequalities in cancer screening participation: examining differences in perceived benefits and barriers
.
Psychooncology
2016
;
25
:
1168
74
.
21.
Worthley
DL
,
Cole
SR
,
Esterman
A
,
Mehaffey
S
,
Roosa
NM
,
Smith
A
, et al
Screening for colorectal cancer by faecal occult blood test: why people choose to refuse
.
Intern Med J
2006
;
36
:
607
10
.
22.
Young
GP
,
Symonds
EL
,
Allison
JE
,
Cole
SR
,
Fraser
CG
,
Halloran
SP
, et al
Advances in fecal occult blood tests: the FIT revolution
.
Dig Dis Sci
2015
;
60
:
609
22
.
23.
Liang
PS
,
Wheat
CL
,
Abhat
A
,
Brenner
AT
,
Fagerlin
A
,
Hayward
RA
, et al
Adherence to competing strategies for colorectal cancer screening over 3 years
.
Am J Gastroenterol
2016
;
111
:
105
14
.
24.
Cole
SR
,
Smith
A
,
Wilson
C
,
Turnbull
D
,
Esterman
A
,
Young
GP
. 
An advance notification letter increases participation in colorectal cancer screening
.
J Med Screen
2007
;
14
:
73
5
.
25.
Walker
R
,
Hiller
JE
. 
The index of relative socio-economic disadvantage: general population views on indicators used to determine area-based disadvantage
.
Aust N Z J Public Health
2005
;
29
:
442
7
.
26.
Symonds
EL
,
Osborne
JM
,
Cole
SR
,
Bampton
PA
,
Fraser
RJ
,
Young
GP
. 
Factors affecting faecal immunochemical test positive rates: demographic, pathological, behavioural and environmental variables
.
J Med Screen
2015
;
22
:
187
93
.
27.
Pedersen
SK
,
Baker
RT
,
McEvoy
A
,
Murray
DH
,
Thomas
M
,
Molloy
PL
, et al
A two-gene blood test for methylated DNA sensitive for colorectal cancer
.
PLoS One
2015
;
10
:
e0125041
.
28.
Todorov
K
,
Wilson
C
,
Sharplin
G
,
Corsini
N
. 
Faecal occult blood testing (FOBT)-based colorectal cancer screening trends and predictors of non-use: findings from the South Australian setting and implications for increasing FOBT uptake
.
Aust Health Rev
2018
;
42
:
45
52
.
29.
Brehaut
JC
,
O'Connor
AM
,
Wood
TJ
,
Hack
TF
,
Siminoff
L
,
Gordon
E
, et al
Validation of a decision regret scale
.
Med Decis Making
2003
;
23
:
281
92
.
30.
Australian Bureau of Statistic. 2016 Socio-economic indexes for areas: Technical Pape
r
, 
2016
.
cat. no. 2033.0.55.001
.
31.
Lindholm
E
,
Brevinge
H
,
Haglind
E
. 
Survival benefit in a randomized clinical trial of faecal occult blood screening for colorectal cancer
.
Br J Surg
2008
;
95
:
1029
36
.
32.
Steele
RJ
,
McClements
PL
,
Libby
G
,
Black
R
,
Morton
C
,
Birrell
J
, et al
Results from the first three rounds of the Scottish demonstration pilot of FOBT screening for colorectal cancer
.
Gut
2009
;
58
:
530
5
.
33.
Weller
D
,
Coleman
D
,
Robertson
R
,
Butler
P
,
Melia
J
,
Campbell
C
, et al
The UK colorectal cancer screening pilot: results of the second round of screening in England
.
Br J Cancer
2007
;
97
:
1601
5
.
34.
Gregory
TA
,
Wilson
C
,
Duncan
A
,
Turnbull
D
,
Cole
SR
,
Young
G
. 
Demographic, social cognitive and social ecological predictors of intention and participation in screening for colorectal cancer
.
BMC Public Health
2011
;
11
:
38
.
35.
Volk
RJ
,
Leal
VB
,
Jacobs
LE
,
Wolf
AMD
,
Brooks
DD
,
Wender
RC
, et al
From guideline to practice: new shared decision-making tools for colorectal cancer screening from the American Cancer Society
.
CA Cancer J Clin
2018
;
68
:
246
9
.
36.
Inadomi
JM
,
Vijan
S
,
Janz
NK
,
Fagerlin
A
,
Thomas
JP
,
Lin
YV
, et al
Adherence to colorectal cancer screening: a randomized clinical trial of competing strategies
.
Arch Intern Med
2012
;
172
:
575
82
.