Prevalence of cancer in the elderly: discrepancies between self-reported and registry data

Abstract

Summary Data on self-reported cancer by a sample of 3349 elderly persons in the south-west of France were validated against registry data in the same region: only 21% of the persons on the cancer registry reported occurrence of cancer. Breast cancer was found to be most frequently accurately reported.Keywords: elderly; validity; questionnaire; registry recordsIn both large cohort and genetic studies, the absence of a population based cancer register for convenient identification of individuals with cancer often means that self-reported data are used. In such cases, only positive responses tend to be verified. However, relatively little attention has been focused on the real possibility of significant numbers of false-negative responses. In view of the high incidence of cancer in the elderly (Yancik and Ries, 1994;Coleman and Lutz, 1996), an increasing number of elderly people are likely to be included in future epidemiological studies. These considerations prompted us to examine the validity of information reported by a population of elderly persons with respect to the occurrence of cancer. Data from a cross-sectional epidemiological survey carried out in the French department of the Tarn (SW France) were cross-checked against data from a cancer registry of the same department used as reference. MATERIALS AND METHODSA cross-sectional epidemiological survey was carried out in 1994 on persons 75 years of age and older, living either at home or in an institution in the French department of the Tarn (SW France). After stratification with respect to the (distribution of the) population of the canton (county) by residence, age and gender, 5161 persons were selected at random from the electoral registers of 41 communes (parishes) in the department. The subjects were informed by mail of the forthcoming visit of an investigator (physician or nurse) to their home, and the data were recorded on a questionnaire. Data from some individuals were recorded by telephone interview, and in certain cases, information was obtained from a near relative. Out of this population, 1129 persons could not be contacted owing to errors in electoral rolls from death (n= 158), incorrect age (n=10), incorrect or change of address (n=542) or absence from home (n=419). Among those contacted, 664 did not respond owing to health problems (n=185) or other reasons (n=479). The overall participation was 65.3% (3368/5161). Correspondence to: F Berthier Information about cancer was recorded from responses to the following question: 'Have you had a polyp, tumour, cyst, nodule or cancer?' and if 'yes', specify which: polyp, tumour, cyst, nodule or cancer, year of diagnosis and anatomical localization of the lesion(s), coded according to the International Classification of Diseases (ICD 9th revision). The 19 persons for whom the response to the last question was incomplete were put into the nonrespondent group. The survey region was covered by a cancer registry set up in 1982, which recorded all cases of cancer apart from basal cell skin carcinoma.The data...

The validity of self‐reported cancer in an Australian population study

“…There are several reasons that may explain the reporting patterns observed in our study. It has been suggested that cancers with clear‐cut diagnostic criteria, such as breast cancer, are more likely to be self‐reported than cancers whose diagnostic procedures are ambiguous 14,15 . For example, an elevated prostate surface antigen (PSA) does not give a clear‐cut diagnosis of prostate cancer, which may be confusing to patients.…”

“…Further, advancing age at the time of questionnaire completion was positively associated with false negative and false positive reporting in our study. Previous studies 14,15 reported that the main reason for false negative and false positive reporting among older persons was declining cognitive functioning. For some, cancer may also be perceived as a ‘taboo’ subject that is associated with fear and stigma, thus making it difficult for doctors to acknowledge and discuss their diagnosis 22 .…”

Agreement between Self-Reported and Confirmed Cancer Diagnoses in New York City Firefighters and EMS Workers, 2001–2011

“…For example, sensitivities ranged from 15.5% to 90.0% for self-reported prostate cancer, and from 36.9% to 85.3% for self-reported melanoma. [5][6][7][8] However, the findings from these studies may not be generalizable to other populations, possibly including FDNY firefighters and emergency medical service (EMS) workers, given the recent focus on World Trade Center (WTC)-related illnesses, including cancer.…”

Bayesian modelling of imperfect ascertainment methods in cancer studies

“…However, this method of cancer ascertainment may introduce inaccuracies, because of under-or over-reporting by patients or proxies [1][2][3]. For example, errors in self-report may arise from lack of knowledge of tumour type, or because pre-malignant lesions may be confused with actual cancers, such as in cervical disease and skin lesions [2].…”