Critics of cancer screening have suggested that the publicity associated with it can provide people with a new health worry.1 After a mass media heart disease campaign in Norway a national survey showed that 17% of those who had seen the campaign materials were worried about heart disease and that these people were most likely to make behaviour changes.2 No comparable data on publicity about cancer screening are available.3
We evaluated the impact of publicity about a new bowel cancer screening programme, comparing a group who had been sent information about the programme with a control group who had not.
Participants, methods, and results
Participants aged 55-64 were identified from family health services authority registers and confirmed by their general practitioner to be suitable for screening for bowel cancer. This study group is part of a randomised controlled trial of flexible sigmoidoscopy for the prevention of bowel cancer.4
Altogether 2961 adults in Leicester were randomised on a 2:1 ratio, with computer generated random allocations, to receive brief information by post about the flexible sigmoidoscopy screening test (n=1974) or not (n=987). Couples were randomised together to avoid contamination. People in the information group were also asked whether they would be interested in having the test. All participants were sent a letter from their general practitioner requesting their cooperation with the questionnaire study and a questionnaire, which included items on worry about bowel cancer, perceived risk, minor bowel symptoms, and anxiety (shortened version of the state trait anxiety inventory). Demographic and health status items were also included (table). Participants were unaware that they were participating in a study of the effect of information. The study was powered to detect a 0.6 difference in the primary outcome, the mean state anxiety score.
Response rates were high, particularly in the control group (80.4% (794/987) for the controls v 72.2% (1425/1974) for the information group; χ21=24.5, P<0.0001). More women than men responded (79.1% (1169/1477) v 70.8% (1051/1484); χ2(1)=27.3, P<0.0001). There were no significant demographic differences between groups or any differences in health status or visits by the general practitioner (table).
Reported worry about bowel cancer was lower in the information group than the control group (35.6% v 29.1% said they were not at all worried; 95% confidence interval for difference 2.4% to 11%), as was perceived risk (4.4% v 6.8% perceived their risk to be higher than average (0.3% to 4.6%)). There was a trend for anxiety to be lower in the intervention group (10.5 v 10.8; t=1.7, P=0.09; −0.66 to 0.05). The information group reported more minor bowel symptoms in the previous three months than the control group (32.1% v 39.0% reported no symptoms; −11% to −2.7%). The pattern of results was the same when analysed separately for men and women.
Comment
Publicity about screening did not increase either worry about bowel cancer or subjective risk; indeed, the information seemed to be reassuring, with the informed group reporting less worry and lower subjective risk. The only evidence of any adverse effect was that patients in the information group were more likely to report minor bowel symptoms. This might be attributable to the recognised phenomenon of an increased perception of symptoms while thinking about a particular disease.5
An alternative explanation for the apparently reassuring effect of screening is that the information caused denial—respondents were more worried but would not admit it. This seems unlikely, given that the information group reported more symptoms than the control group. The response rate was slightly lower in the information group; this raises the possibility of a small subset of anxious non-responders, whose experiences are not reflected in studies that use postal questionnaires to collect the data.
Overall, these results are encouraging. They give no support to the idea that publicity about cancer screening produces widespread alarm, at least in this age group, and even suggest that screening publicity can have a positive effect.
Table.
Demographic characteristics of participants in study, and attitudes to bowel cancer
| No (%) of participants
|
χ2 (for trend where appropriate) | ||
|---|---|---|---|
| Information group (n=1425) | Control group (n=795) | ||
| Demographics | |||
| Age (years): | χ2=1.63, P=0.20 | ||
| 55-59 | 759 (53.3) | 401 (50.4) | |
| 60-64 | 666 (46.7) | 394 (49.6) | |
| Marital status: | χ2=1.04, P=0.31 | ||
| Married | 1048 (79.0) | 612 (80.8) | |
| Not married | 279 (21.0) | 145 (19.2) | |
| Employment status: | χ2=2.38, P=0.12 | ||
| Employed | 714 (54.2) | 436 (57.7) | |
| Unemployed | 604 (45.8) | 320 (42.3) | |
| Housing tenure: | χ2=0.120, P=0.73 | ||
| Owns home | 1120 (85.9) | 640 (85.3) | |
| Rents home | 184 (14.1) | 110 (14.7) | |
| Car ownership: | χ2=0.018, P=0.89 | ||
| Yes | 1104 (83.4) | 627 (83.2) | |
| No | 220 (16.6) | 127 (16.8) | |
| Educational qualifications: | χ2=1.99, P=0.16 | ||
| Yes | 501 (38.4) | 262 (35.3) | |
| No | 802 (61.6) | 480 (64.7) | |
| Ethnic group: | χ2=0.078, P=0.78 | ||
| White | 1273 (97.4) | 732 (97.6) | |
| Other | 34 (2.6) | 18 (2.4) | |
| Health status | |||
| Current health: | χ2=0.722, P=0.40 | ||
| Excellent | 164 (12.2) | 89 (11.7) | |
| Good | 816 (60.9) | 446 (58.6) | |
| Fair | 313 (23.3) | 204 (26.8) | |
| Poor | 48 (3.6) | 22 (2.9) | |
| Frequency of visits to GP in previous 3 months: | χ2=0.776, P=0.38 | ||
| Have not been | 543 (40.4) | 331 (43.6) | |
| Have been once | 418 (31.1) | 215 (28.3) | |
| Have been twice | 187 (13.9) | 110 (14.5) | |
| Have been ⩾3 times | 195 (14.5) | 104 (13.7) | |
| Psychological variables | |||
| Worry about bowel cancer: | χ2=37.7, P<0.0005 | ||
| Not worried | 477 (35.6) | 220 (29.1) | |
| A bit worried | 687 (51.3) | 353 (46.6) | |
| Quite worried | 128 (9.6) | 122 (16.1) | |
| Very worried | 47 (3.5) | 62 (8.2) | |
| Perceived risk of bowel cancer: | χ2=9.37, P=0.002 | ||
| Lower than average | 164 (12.9) | 70 (9.5) | |
| Same as average | 1049 (82.7) | 613 (83.6) | |
| Higher than average | 56 (4.4) | 50 (6.8) | |
| No of bowel symptoms in previous 12 months: | χ2=19.3, P<0.0005 | ||
| None | 457 (2.1) | 310 (39.0) | |
| 1 | 402 (28.2) | 245 (30.8) | |
| >1 | 565 (39.7) | 240 (30.2) | |
| Mean (SD) state anxiety score | 10.5 (3.9) | 10.8 (4.1) | |
GP=general practitioner.
Footnotes
Funding: Financial support for the flexible sigmoidoscopy trial was provided by the Imperial Cancer Research Fund and the Medical Research Council.
Competing interests: None declared.
References
- 1.Skrabanek P. False premises and false promises of breast cancer screening. Lancet. 1985;ii:316–320. doi: 10.1016/s0140-6736(85)90362-9. [DOI] [PubMed] [Google Scholar]
- 2.Søgaard AJ, Fønnebø V. Self-reported change in health behaviour after a mass media-based health education campaign. Scand J Psychol. 1992;33:125–134. doi: 10.1111/j.1467-9450.1992.tb00892.x. [DOI] [PubMed] [Google Scholar]
- 3.Wardle J, Pope R. The psychological costs of screening for cancer. J Psychosom Res. 1992;36:609–624. doi: 10.1016/0022-3999(92)90051-3. [DOI] [PubMed] [Google Scholar]
- 4.Atkin WS, Hart A, Edwards R, McIntyre P, Aubrey R, Wardle J, et al. Uptake, yield of neoplasia, and adverse effects of flexible sigmoidoscopy screening. Gut. 1998;42:560–565. doi: 10.1136/gut.42.4.560. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Woods SM, Natterson J, Silverman J. Medical students’ disease: hypochondriasis in medical education. J Med Educ. 1966;41:785–790. [PubMed] [Google Scholar]