Endoscopic follow-up of positive fecal occult blood testing in the Ontario FOBT Project

Abstract

BACKGROUND:

The Ontario FOBT Project is a pilot study of fecal occult blood testing (FOBT) for colorectal cancer screening conducted among age-eligible volunteers (50 to 75 years) in 12 of 37 public health regions in Ontario.

METHODS:

Volunteers responded to invitations from primary care practitioners (PCPs) in six regions, and from public health programs in the remaining regions. FOBT collection kits were distributed from routine laboratory specimen collection sites, to which completed kits were returned. Results were sent to PCPs in all 12 regions, with copies sent to the study office at Cancer Care Ontario (Toronto, Ontario). Follow-up of positive results was at the discretion of the PCPs. The study files contained the unique Ontario Health Insurance Numbers, the date of the analyses, the number of satisfactory slides and the results for each slide. The Ontario Health Insurance Numbers were encrypted for each participant, and along with the study file, were linked to medical billing claims, hospital records and aggregate demographic data.

RESULTS:

Among participants with positive results (men 3.5% and women 2.2%), the median time from date of FOBT analysis to date of colonoscopy was 121 days among men and 202 days among women. At the end of follow-up, after positive FOBT (six to 17 months), 73% of men and 56% of women had proceeded to colonoscopy.

CONCLUSION:

Although colonoscopy appeared to be acceptable to the majority of participants with positive FOBT, accessibility problems was the likely explanation for lengthy intervals between the date of positive FOBT and its performance. Differences between the experiences of men and women require further investigation.

Fecal occult blood testing (FOBT) is the only screening test for colorectal cancer (CRC) proven, in large randomized trials (1), to reduce mortality. In 2001, the Canadian Task Force on Preventive Health Care (2) endorsed CRC screening with FOBT (level A – good evidence to recommend) or with periodic flexible sigmoidoscopy (level B – fair evidence to use), for people at average risk of CRC. The Task Force did not endorse colonoscopy (level C – insufficient evidence to recommend for or against use).

Compliance of age-eligible individuals is necessary (but not sufficient) to benefit from CRC screening. The members of the eligible population must be compliant with FOBT, and if the test detects occult blood then they must be compliant with follow-up colonoscopy (3).

Population-based cancer screening is not recommended if resources are not available to deal with abnormalities possibly associated with an underlying cancer or precancerous lesion (4). Although the utilization of large bowel endoscopy dramatically increased between 1992 and 2001 in Ontario (5), the uptake of all colorectal evaluations, including colonoscopy for any clinical or screening reason, was low among age-eligible residents of Ontario (6). Canadian provinces may not have sufficient capacity for large bowel endoscopy for the follow-up of positive FOBT results if the utilization of CRC screening increases from the current low level.

METHODS

The Ontario Ministry of Health and Long-Term Care funded the present pilot project to provide information on the process of policy development for CRC screening by FOBT. The primary objectives were to assess the rate of positive FOBT results, the rate of compliance with colonoscopy following positive FOBT and the time interval from positive FOBT to colonoscopy.

Printed health educational materials on CRC and FOBT were developed and distributed to primary care practitioners (PCPs) in six regions and to public health units in six other regions. PCPs in the six primary care regions were encouraged to distribute the materials directly to age-eligible patients during routine office visits scheduled for other purposes. Potential participants received these materials, along with a laboratory requisition and a consent form, between March 1, 2004, and February 28, 2005, directly from their PCP. In the public health units, materials were distributed through a wider network of community partners that also included PCPs. Identical materials were provided in all 12 regions. Consent was obtained for the use of the unique Ontario Health Insurance Number (OHIN) for data linkage and observation. Potential participants (50 to 75 years of age), who took the requisition and a signed consent form to the laboratory, received a collection kit for FOBT to be returned to the laboratory after completion. The FOBT requisitions were prestamped with the words ‘Ontario FOBT Project’ for purposes of special reimbursement of the tests (not out of the regular Ontario Health Insurance Plan pool for laboratory testing) and to ensure transmittal of results to the study office.

Results of each FOBT kit analysis were sent to the participants’ PCPs in all 12 regions and to the study office at Cancer Care Ontario (Toronto, Ontario). Results of kits returned to laboratories by May 31, 2005, were eligible for analysis. This allowed 12 weeks after the final date of recruitment for participants to complete and return the FOBT collection kits. Follow-up of the FOBT results with participants was not conducted in the present study and PCPs in all 12 regions followed their usual practice for positive FOBT results.

A database of the FOBT results was prepared at Cancer Care Ontario and included the OHIN of each participant, the date of the FOBT analysis, the number of completed slides per participant and the results of each slide. The database was transmitted to the Institute for Clinical Evaluative Sciences (Toronto, Ontario), where the OHINs were encrypted and linked to the Medicare billing claims, and hospital and demographic databases. Demographic variables, including residential descriptors (region, urban versus rural, and regionally adjusted neighbourhood income quintiles) during the study period, were extracted, and persons for whom one PCP provided the majority of office visits were labelled as having a regular PCP. Episodes of colonoscopy occurring up to November 30, 2005, were extracted, providing a minimum observation period of six months after the date of all positive FOBT kits. Results of colonoscopy, apart from the performance of polypectomy at the time of the colonoscopy, were not available.

To minimize the study burden on PCPs and public health programs, eligibility ascertainment was not included in the study (persons outside the age range and persons with a prior history of CRC were excluded before analysis). The study was approved by the Research Ethics Board of the Sunnybrook and Women’s College Health Sciences Centre (Toronto, Ontario).

RESULTS

The randomly selected regions were representative of the province of Ontario as a whole. Both groups of regions included a metropolitan area with an academic health sciences centre, other large urban areas and rural settings. The total number of age-eligible participants submitting informed consent forms and completing FOBT collection kits was 6972. Participants were more likely to be women, of older age, living in higher income neighbourhoods and having a regular PCP, and were less likely to have a history of comorbid illnesses than nonparticipants. The proportions of urban and rural residents were similar among participants and nonparticipants (Table 1). Participants were less likely to have had FOBT, a barium enema, a colonoscopy or a polypectomy during the year before recruitment (March 1, 2003, to Febrary 28, 2004) than nonparticipants (Table 2).

TABLE 1.

Comparison of participants to nonparticipants among the 12 health units

Variable	Participants	Nonparticipants
Age (years), n
50 to 54	1326	239,835
55 to 59	1519	190,505
60 to 64	1546	144,969
65 to 69	1403	116,938
70 to 75	1178	112,200
Sex, n (%)
Male	2827 (40.6)	395,860 (49.2)
Female	4145 (59.4)	408,387 (50.8)
Neighbourhood income quintiles, n (%)
1 (lowest)	986 (14.1)	140,912 (17.5)
2	1208 (17.3)	158,001 (19.7)
3	1495 (21.4)	162,706 (20.2)
4	1612 (23.1)	161,670 (20.1)
5 (highest)	1515 (21.7)	164,592 (20.5)
Missing	156 (2.2)	16,366 (2.0)
Residence, n (%)
Urban	5325 (76.4)	611,406 (76.0)
Rural	1647 (23.6)	192,841 (24.0)
Regular primary care*, n (%)
Yes	6388 (91.6)	640,331 (79.6)
No	584 (8.4)	163,916 (20.4)

^*Provided by a primary care practitioner

TABLE 2.

Large bowel procedures during the 12 months before the start of recruitment

Procedure	Participants n (%)	Nonparticipants n (%)
FOBT	51 (0.73)	26,029 (3.24)
Barium enema	59 (0.85)	7907 (0.98)
Colonoscopy	122 (1.75)*	21,075 (2.62)
Polypectomy	23 (0.33)	5952 (0.74)

^*Thirty-five participants were not billed to cecum or terminal ileum colonoscopy. FOBT Fecal occult blood testing

The percentage of participants who used all three collection slides was 99.3%. The analysis was negative for occult blood in 96% of participants and positive among 2.8%, and slides among 1.2% were not assessable. Men and older participants were more likely to have positive FOBT results than women and younger participants (Table 3).

TABLE 3.

Positive fecal occult blood testing results by age and sex

Variable	Frequency, n	n/Total	%
Age, years
50 to 54	24	24/1326	1.81
55 to 59	41	41/1519	2.70
60 to 64	35	35/1546	2.26
65 to 69	37	37/1403	2.64
70 to 75	55	55/1178	4.67
Sex
Female	93	93/4145	2.24
Male	99	99/2827	3.50

Colonoscopy and polypectomy were more frequent among participants than nonparticipants (Table 4) between March 1, 2004, and November 30, 2005. Overall, 124 of 192 (64.6%) participants with positive FOBT had a completed colonoscopy by November 30, 2005. Among participants with positive results, men were more likely to proceed to colonoscopy than women (73% of men had colonoscopy compared with 56% of women). Men proceeded to colonoscopy faster than women; the median time among male participants with positive FOBT was 121 days, but among females participants, it was 202 days. The number of FOBT-positive cases is inadequate to warrant comparisons of colonoscopy utilization in this setting among the 12 regions or among income quintiles. In addition, 270 of 6780 (4.0%) participants with negative or nonassessable FOBT proceeded to colonoscopy by November 30, 2005 (Table 4).

TABLE 4.

Comparison among large bowel procedures performed, between March 1, 2004, and November 30, 2005, for participants and nonparticipants

Large bowel procedure	Participants n (%)	Nonparticipants n (%)
Positive FOBT result in the Ontario FOBT Project	192 (2.8)	–
Colonoscopy	462 (6.6)	39,965 (4.9)
Polypectomy	140 (2.01)*	12,841 (1.60)
Large bowel resection	11 (0.16)	1822 (0.23)

^*Forty-seven participants had a positive fecal occult blood test (FOBT)

DISCUSSION

Although CRCs detected after positive FOBT or other screening procedures among asymptomatic individuals are more likely to be stage A (stage 1) (1) than those diagnosed after the onset of symptoms, many investigators have concluded that the mortality reduction observed in population trials of FOBT is largely due to the removal of polyps discovered in the course of colonoscopy prompted by positive FOBT (3).

The acceptability and accessibility of colonoscopy are determinants of the effectiveness of CRC screening by FOBT. Our finding that 27% of men with positive FOBT and 44% of women with positive FOBT had not proceeded to colonoscopy by November 30, 2005 – 6.5 months after the test date of the last positive FOBT among participants – suggests that a significant minority of participants may not deem it to be acceptable. Differences in colonoscopy acceptability between men and women has been observed in other publications (7–9).

However, it is possible that PCPs are less likely to refer women for colonoscopy. The uptake of colonoscopy among men in the present study is similar to that observed by Grazzini et al (10). The higher uptake among men was also seen in studies described by Meissner et al (11) and Church et al (12). However, in the National Health Service pilot study, uptake did not differ between men and women (13).

The length of intervals from date of positive FOBT to date of colonoscopy suggests that the current capacity of services in Ontario is not sufficient for prompt access for FOBT screening participants with positive test results. During the National Health Service pilot, the interval from referral to colonoscopy for symptomatic or ill patients rose from 10 weeks to 17 weeks at one participating hospital, whereas FOBT-positive patients waited only two weeks near the beginning of the pilot and six weeks by the end of the pilot (14).

CONCLUSION

Screening for CRC by FOBT requires appropriate measures to ensure adequate capacity for prompt colonoscopy following positive FOBT, and these measures must avoid making sick patients wait even longer than usual for diagnostic colonoscopy. In addition, there is a need for research to improve the acceptability of FOBT among men, and to improve the acceptability and accessibility of colonoscopy among women testing positive by FOBT.