Download ž / Screening with faecal occult blood test FOBT for colorectal cancer

European Journal of Cancer Pre¨ ention 2001, 10, 251᎐256
Screening with faecal occult blood test ž FOBT/
for colorectal cancer: assessment of two methods
that attempt to improve compliance
L Ore, L Hagoel, I Lavi, G Rennert
(Recei¨ ed 18 July 2000; accepted 8 February 2001)
Screening with the faecal occult blood test (FOBT) has been shown in randomized control trials to be
effective in reducing mortality from colorectal cancer. Compliance to this test recommendation, however, by
the general population is usually low. To evaluate different methods of increasing compliance with FOBT,
using mailed test kits or order cards, with or without information leaflets, subjects were randomly assigned to
receive a test kit or a kit request card. An information leaflet was included in half of the mailings. All
participants were contacted for interview. Compliance was evaluated through the central computer system of
the study’s FOBT laboratory. Self-initiated compliance with FOBT in the year preceding the study was 0.6%
of the study participants. The overall compliance rate with the programme invitation was 17.9%, with a
somewhat higher, though non-significant response to the mailed kit (19.9%) over the kit request card (15.9%).
Women complied with the test significantly more than men, older participants more than younger. Compliance to FOBT is low among the Israeli population aged 50–74 who receive a formal invitation to carry out
this screening. Mailing a kit request card within the framework of a screening programme can achieve a
substantial increase (to 17.9%) in the level of compliance for the relatively low cost of postage. More effort is
needed to study additional means of convincing the non-responders to take part in this potentially life saving
activity.
䊚 2001 Lippincott Williams & Wilkins
Key words: Colorectal cancer, FOBT, Israel, screening.
Introduction
Cancer of the colon or rectum is a leading cause of
cancer-related death in the western world ŽGloeckler
Ries et al., 1996.. The incidence rates in the Israeli
population in 1995 were 24.4 per 100 000 in ŽJewish.
women and 32.5 per 100 000 in ŽJewish. men ŽIsrael
Cancer Registry, 1998.. The disease is usually detected at an advanced stage when prognosis is poor,
and carries a very unfavourable 5-year survival rate
of about 50% ŽMerrill et al., 1998.. Of all suggested
tests ᎐ faecal occult blood testing ŽFOBT., various
endoscopic examinations and double-contrast barium enema tests ŽWinawer and Shike, 1995. ᎐ FOBT
is the most suitable technology for high-volume
population-based screening, due to its simplicity, low
cost and non-invasiveness ŽSimon, 1987..
Three randomized controlled trials have thus far
been published, showing an 18᎐33% decrease in
mortality among groups periodically carrying out
FOBT, compared with the controls that do not carry
out this screening ŽHardcastle et al., 1996; Mandel et
al., 1993; Kronborg et al., 1996.. As with every
population screening activity, in order to translate
the qualities of FOBT into a population benefit, it is
of major importance to assure a high compliance
rate ŽFowler and Austoker, 1997.. The current health
services basket in Israel includes an annual FOBT
free of charge for men and women in the age group
50᎐74. Former experience with breast cancer screening activities in this country has marked a wide array
of health beliefs as predictors of future screening
behaviour ŽOre et al., 1997; Hagoel et al., 1999..
Compliance with colorectal cancer screening recom-
Department of Community Medicine and Epidemiology, Carmel Medical Center and The Faculty of Medicine, Technion, 7 Michal Street, Haifa
34362, Israel. Correspondence to: L Hagoel. Fax: (q972) 4 8344 358. E-mail: [email protected]
0959-8278 䊚 2001 Lippincott Williams & Wilkins
European Journal of Cancer Prevention. Vol 10. 2001
251
L Ore et al.
mendations is usually lower than that recorded for
breast or cervical cancer screening throughout the
world ŽHardcastle et al., 1996; Mandel et al., 1993;
Winawer et al., 1993; Kronborg et al., 1996..
This study was undertaken in order to assess the
compliance to FOBT colorectal cancer screening in
the Israeli population. The focus was a comparison
between two modes of approaching participants
Žmailing a test kit versus mailing a kit request card.,
as well as the impact of including an information
leaflet with each.
Methods
A random sample of 1000 women and 1000 men, all
aged 50᎐74 years and residents of Haifa Žthe third
largest city in Israel ., was selected from the computerized member list of the Kupat Holim Clalit ŽKHC,
the largest Health Maintenance Organization
wHMOx in the country..
All participants were randomly assigned to receive, free of charge, one of the following:
Ža. An envelope containing an FOBT kit ŽHaemoccult II SENSA, three faecal samples. to be sent
by mail to a central laboratory located in
Carmel Medical Centre as an act of compliance. Envelopes included: Ž1. a letter inviting
recipients to perform the test annually; Ž2. a
detailed explanation of how the test should be
carried out; Ž3. dietary restrictions for 48 hours
before test performance Žavoiding meat and
certain fresh fruit and vegetables, as well as
consumption of vitamin C, ferrum and aspirincontaining medications..
Žb. An envelope containing a letter as described
above, and a kit request card to be returned to
the same laboratory Žconsidered as the initial
response required in this approach. which then
would respond by mailing the test kit to the
participants.
Half of all envelopes in each of the above mentioned groups also included a leaflet describing colon
cancer risk and the importance of early detection of
cancer. This was the only effort made to enhance
compliance, in addition to the test being free of
charge. Test results were delivered to participants’
family physicians.
About 2 months after the initial mailing, a 10minute telephone interview was conducted by trained
interviewers, who were blinded to the approach to
which each interviewee was assigned. The interview
included questions about demographic characteristics and compliance to the test reported in this
paper. Participants were asked about their reasons
for compliancernon-compliance to FOBT testing
without any reference to its potential benefit, nor
any encouragement to perform the test. Other questions included in the interview were related to health
behaviours, as well as personal and family medical
history and knowledge and attitudes towards colon
cancer and its early detection. These data will be
reported separately.
The compliance to FOBT was assessed about
5 months after the initial mailing using the KHC
National FOBT Screening Program database. Participants could not receive the study kits elsewhere.
Statistical analysis
Respondents were categorized into two groups: compliers and non-compliers.
Table 1. Demographic characteristics of the study population Ž N s 1940 a .
Variable
Categories
N
%
b
50᎐64
65᎐74
Women
Men
Married
Other
EuroperAmerica
Israel
AsiarAfrica
0᎐8
9᎐12
13 q
1009
931
974
966
1466
270
835
380
313
364
567
411
52.0
48.0
50.2
49.8
84.4
15.6
54.6
20.5
24.9
27.1
42.3
30.6
Age
Gender
Marital status
Ethnic origin
Educationc Žyears.
a
Due to missing values, N may change in specific variables.
Mean " SD: 63.5" 6.88.
c
Mean " SD: 11.4" 4.31.
b
252
European Journal of Cancer Prevention. Vol 10. 2001
Colorectal cancer screening with FOBT
Figure 1. Study design.
Compliers were participants who actually performed the test Žboth kit and card receivers.. Noncompliers were divided into two subgroups: card
receivers who showed intention to perform the test
and returned the kit request card, but did not perform the test, and kit receivers who ignored the
request to perform the test.
The statistical procedures used included univariate analyses, as well as cross tabulations and chisquare tests. Compliance was used as the dependent
variable, whereas approach, leaflet, interview and
selected demographic and other characteristics were
used as the explanatory variables. Significance level
was set at P- 0.05.
Results
Out of the 2000 potential participants, 1940 Ž97%.
were eventually studied ŽFigure 1.. Sixty patients
were excluded either because they had moved out of
the area, or they were hospitalized or passed away.
As planned, half of the participants received a card
and half received a kit. Half of each of these groups
received a leaflet as well. The mean age of the
population studied was 63.5" 6.88. About 55% were
of EuropeanrAmerican origin, 84.4% were married
and 30.6% had 13 or more years of education ŽTable
1.. No differences in distribution of the demographic
characteristics of the study population by mode of
approach Žtest kits or kit request cards. were
observed.
According to the programme’s database 12 participants Ž0.6%. were found to have performed the test
in the year prior to the current mailing, on their own
initiative. In response to the present study, 347 participants Ž17.9%. complied with the test recommendation within 4 months, with kit receivers complying
to a higher extent than card receivers Ž19.9% and
15.9% respectively, Ps 0.02..
There were no significant differences in overall
compliance by gender: 18.0% of women versus 17.8%
of men. However, an interaction was observed by
Table 2. FOBT performance by gender and by approach
Women
Men
Total
Kit
Card
Kit
Card
Kit
Card
Number of tests
performed
Total number
approached
% Performed
P-value
103
72
89
83
192
155
484
490
480
486
964
976
21.3
14.7
18.5
17.1
19.9
15.9
0.007
0.55
0.02
European Journal of Cancer Prevention. Vol 10. 2001
253
L Ore et al.
Table 3. FOBT performance by age and by approach
Age
50᎐64
65᎐74
Total
Kit
Card
Kit
Card
Kit
Card
Number of kits
performed
Total number
approached
% Performed
P-value
79
78
113
77
192
155
500
509
464
467
964
976
15.8
15.3
24.4
16.5
19.9
15.9
0.83
gender and approach. While men adhered similarly
to both approaches Žabout 18%., women kit receivers had a statistically significant higher rate of
compliance than women who received a card Ž21.3%
versus 14.7%, respectively; Table 2..
Elderly subjects Žaged 65᎐74. were more compliant than younger subjects Ž20.4% versus 15.5%.. By
design, gender distribution was even in both age
groups. An interaction was observed between age
and approach in relation to FOBT performance: in
the younger age group there was no difference in
performance by approach, while in the 65᎐74 age
group there was a higher compliance rate among kit
receivers ŽTable 3..
The leaflet had no observable impact overall
Ž17.9% compliance among both leaflet receivers and
non-receivers., nor did it show an impact by gender
or age Ždata not included..
European Journal of Cancer Prevention. Vol 10. 2001
0.02
The rate of performance of the test is presented
in Figure 2. The figure displays 2-week intervals
beginning from the date the study materials were
posted. As seen, 58% of the study’s test performers
complied with the test prior to the beginning of the
interviews. The response pattern following the interviews fluctuated. There were no differences among
those responding before the interview and those
responding thereafter in terms of gender, age and
type of approach.
Discussion
Although malignant colorectal tumours are a leading cancer among the Israeli population, baseline
compliance with FOBT performance is very low
Ž0.6%.. This rate is much lower than the 19.8%
reported in the US ŽMMWR, 1999.. No major stud-
Figure 2. Faecal occult blood test ŽFOBT. performance rate by time Ž n s 347..
254
0.003
Colorectal cancer screening with FOBT
ies have been carried out thus far in the Israeli
population in order to characterize and explain the
barriers to performance of this important, non-invasive procedure. This study, although very slightly
over-representing the elderly and men Žcomparison
to national demographic data is not presented. is, to
the best of our knowledge, the first of its kind in the
country.
The study was undertaken in order to examine the
impact of different methods on improving compliance, as well as to enhance our understanding of the
characteristics of performers versus non-performers.
The present analysis focuses on the first of these two
objectives.
Both approaches studied, a one-step versus twostep, did increase the FOBT compliance rate Žfrom
0.6% to about 18%.. Compared with compliance
rates in clinical trials ŽMandel et al., 1993; Hardcastle et al., 1996; Kronberg et al., 1996. this is a low
result, partly explained by the strong follow-up efforts invested in clinical trials. Although considered
simple by medical professionals, this is an embarrassing and inconvenient test, the performance of
which requires a high motivational level ŽMitchellBeren et al., 1989..
The translation of the compliance to a significant
favourable effect upon the population’s health may
require an increased cooperative effort on the part
of both health care providers and the public. It has
been shown that an explanation from the family
physician addressed personally to the patient with an
enclosed FOBT kit can enhance compliance rates
ŽKing et al., 1992.. It should be noted that at the
time our study was undertaken, the extent of local
professional support for FOBT in Israel was almost
non-existent.
A small advantage of about 4% was observed Žin
terms of test performance. for the direct mailing of
an FOBT kit, rather than the two-step approach. It
is possible that mailing a kit request card first, and
later receiving the kit in the return mail, is more
cumbersome than direct receipt of the kit. The reluctance or delay in responding to the two-step
approach may be due not only to the additional
effort of ordering the kit, but also to the requirement of another decision-making step in the face of
uncertainty as to the test’s concrete requirements
and feasibility.
On the other hand, from an economic point of
view, the relatively low overall compliance Ž18%.
and the relatively small difference between the two
approaches points to the advantage of the two-step
approach in saving the high costs associated with
dispersing kits that remain unused. It emerges from
the data that health policy decision-makers should
adopt the more economic two-step approach. However, in health care organizations where the option
of the direct mailing of a kit seems more relevant,
limiting its use to women and to the elderly appears
to be more prudent economically. This is based
upon the observed interaction between gender and
approach and between age and approach, leaving
younger men as the target population for a mailed
card intervention. Higher compliance rates among
women and elderly individuals were also reported
elsewhere ŽBat et al., 1986; Hardcastle et al., 1996..
The lack of impact of the inclusion of an informational leaflet on compliance rates among women,
men, the younger and the elderly participants alike
is consistent with previous findings reported in Israel
ŽBiger et al., 1994.. However, the latter findings were
based on an Israeli study that examined the use of
leaflets within the context of a programme for early
detection of cancer in general, not of colorectal
cancer in particular. Higher compliance rates were
reported in men who received a leaflet, as opposed
to no effect on women ŽHart et al. 1997..
Our findings support the notion that compliance
with cancer early detection programmes is not solely
a rational behaviour. It raises the question, from a
budgetary point of view, of whether the effort made
to increase compliance through information delivery
on colorectal cancer is justified without targeting
emotional barriers. When investing efforts in the use
of updated informational leaflets in large intervention programmes, the expected outcome of this activity must be stressed: such leaflets may enhance
knowledge and understanding; there is no guarantee, however, that they will affect compliance.
An additional strategy, the interview, may be considered a potential means of improving the low
overall compliance rate. Indeed, 42% of the test
performers responded only after they were interviewed. Nonetheless, this study did not confirm a
distinct impact on the compliance rate following the
interviews. Less expensive and time-consuming
means must be developed and implemented to further encourage compliance.
In conclusion, while proven to be effective in
reducing colorectal mortality, FOBT is still significantly underused in Israel. Measures designed to
improve compliance can affect about one-fifth of the
population, and should be employed as an initial
activity. Reaching higher proportions of the population must also involve significant efforts on behalf of
European Journal of Cancer Prevention. Vol 10. 2001
255
L Ore et al.
the primary care physician, and the building of a
supportive professional and social environment.
Acknowledgements—This study was funded by a research grant from the Israeli Cancer Association.
References
Bat L, Pines A, Ron E, et al. Ž1986.. A community-based program
of colorectal screening tests and compliance. Am J Gastroenterol 81: 647᎐51.
Biger C, Epstein LM, Hagoel L, Tamir A, Robinson E Ž1994.. An
Evaluation of an education programme for prevention and
early diagnosis of malignancy in Israel. Eur J Cancer Pre¨ 3:
305᎐12.
Fowler G, Austoker J Ž1997.. Screening. In Oxford Textbook of
Public Health. Žeds Detels R, Holland WW, McEwen J, Omenn
GS.; Oxford University Press, New York, pp. 1587.
Gloeckler Ries LA, Hankey BF, Harras A, Devesa SS Ž1996..
Cancer incidence: mortality and patient survival in the United
States. In Cancer Epidemiology and Pre¨ ention, 2nd edn Žeds
Schottenfeld D, Fraumeni JF.; Oxford University Press, New
York, pp. 173᎐5.
Hagoel L, Ore L, Neter E, Shifroni G, Rennert G Ž1999.. The
gradient in mammography screening behavior ᎐ a lifestyle
marker. Soc Sci Med 48: 1281᎐90.
Hardcastle JD, Chamberlain JO, Robinson MHE, et al. Ž1996..
Randomised controlled trial of fecal occult blood screening for
colorectal cancer. Lancet 348: 1472᎐7.
Hart AR, Barone TL, Gay SP, et al. Ž1997.. The effect on
compliance of a health education leaflet in colorectal cancer
screening in general practice in central England. J Epidemiol
Community Health 51: 187᎐91.
256
European Journal of Cancer Prevention. Vol 10. 2001
Israel Cancer Registry Ž1998.. Trends in Cancer Incidence and
Mortality in Israel, 1970᎐1995. Publication No. 207, Ministry of
Health, Jerusalem.
King J, Fairbrother G, Thompson C, Morris DL Ž1992.. Colorectal cancer screening: optimal compliance with postal faecal
occult blood test. Aust N Z J Surg 62: 714᎐19.
Kronborg O, Fenger C, Olsen J, Jorgensen OD, Sondergaard O
Ž1996.. Randomised study of screening for colorectal cancer
with fecal occult blood test. Lancet 348: 1467᎐71.
Mandel JS, Bond JH, Church TR, et al. Ž1993.. Reducing mortality from colorectal cancer by screening for fecal occult blood.
N Engl J Med 328: 1365᎐71.
Merrill RM, Henson DE, Ries LA Ž1998.. Conditional survival
estimates in 34,963 patients with invasive carcinoma of the
colon. Dis Colon Rectum 41: 1097᎐106.
Mitchell-Beren ME, Dodds ME, Choi KL, Waskerwitz TR Ž1989..
A colorectal cancer prevention, screening, and evaluation program in community black churches. CA Cancer J Clin 39:
115᎐18.
Ore L, Hagoel L, Shifroni G, Rennert G Ž1997.. Compliance with
mammography screening in Israeli women: the impact of a
pre-scheduled invitation and of letter type. Isr J Med Sci 33:
103᎐11.
MMWR Ž1999.. Screening for colorectal cancer ᎐ US 1997.
MMWR Weekly 48: 116᎐21.
Simon JB Ž1987.. The pros and cons of fecal occult blood testing
for colorectal neoplasm. Cancer Metastasis Re¨ 6: 397᎐411.
Winawer SJ, Flehinger BJ, Schottenfeld D, Miller DG Ž1993..
Screening for colorectal cancer with fecal occult blood testing
and sigmoidoscopy. J Natl Cancer Inst 85: 1311᎐18.
Winawer SJ, Shike M Ž1995.. Prevention and control of colorectal
cancer. Cancer Pre¨ ention and Control Žeds Greenwald P,
Kramer BS, Weed DL.; Marcel Dekker, New York, pp. 537᎐60.