Original Article
Self-reported information on the diagnosis of colorectal cancer was reliable but not necessarily valid

https://doi.org/10.1016/j.jclinepi.2007.05.018Get rights and content

Abstract

Objective

Self-report is commonly used in epidemiologic studies; however, few data exist on the reliability and validity of this method for eliciting information related to the diagnosis of colorectal cancer. We examined the test–retest reliability and validity of colorectal cancer patients reporting on the process of their diagnosis.

Study Design and Setting

One hundred and sixteen participants completed two telephone interviews, 1 month apart, and 95 general practitioners (GPs) completed a written questionnaire, to elicit information relating to key elements of the process of diagnosis of colorectal cancer.

Results

Acute symptoms such as rectal bleeding had higher reliability and validity than more general symptoms. Colonoscopy was the most accurately recalled diagnostic test. Recall of diagnosis date, and date of colonoscopy, had high test–retest reliability. There were considerable differences between dates of diagnostic tests given by participants and GPs, but there was no evidence of a bias in a particular direction. Accuracy of recall did not diminish as time from diagnosis increased.

Conclusion

This study confirms that self-reported symptoms, tests, and dates in the colorectal cancer diagnostic pathway are generally reliable; however, the validity of reported symptoms and tests can be moderate to poor.

Introduction

In the United States of America, Europe, and most of the industrialized world, colorectal cancer is the third most common cancer and the second most common cause of cancer deaths [1], [2], [3]. Risk increases progressively with age, and the incidence of colorectal cancer is expected to increase as worldwide trends in population aging continue [1].

The process of diagnosis for colorectal cancer is often characterized by multiple symptoms and a number of diagnostic tests. Symptoms may range from the acute, such as rectal bleeding, to more general, constitutional complaints, such as lethargy and loss of appetite. Symptoms may not be apparent until late in the disease course, and as a result, colorectal cancer is often diagnosed at an advanced stage [4]. Patients sometimes undergo a number of tests before a conclusive diagnosis is made.

Self-report is a common method of obtaining data pertaining to symptoms and diagnostic tests for epidemiologic studies, primarily because it is easy to implement and is inexpensive [5], [6], [7], [8]. However, self-reported data may be limited by a number of factors, such as recall error, biases such as social desirability and acquiesce, and by poorly constructed measures [7], [9], [10], [11].

There is a scarcity of data on the reliability of self-reported colorectal cancer symptoms. Malats et al. [12] compared hospital records with interview data to assess agreement on the initial symptom experienced by patients with cancer of the digestive tract (just over half had colorectal cancer). They found that the concordance between hospital records and interview data was 61% for type of first symptom, with only 46% agreement for the date of first symptom (within 30 days).

A number of studies have assessed the validity of self-reported testing for colorectal cancer [5], [13], [14], [15], [16]. However, these studies focused on colorectal cancer screening in the general population. There are no published data on the validity of self-reported diagnostic testing among colorectal cancer patients. Most studies of the validity of self-reported screening tests found at least moderate agreement between self-report and other, more objective data [13], [14], [15], [16]. Colonoscopy and sigmoidoscopy were more accurately recalled than fecal occult blood testing; recall was generally better for more invasive tests [5], [14]. A common finding was that participants tended to report having had screening tests more recently than had been the case.

We examined the test–retest reliability and the validity of patient self-report on key elements of the process of diagnosis of colorectal cancer.

Section snippets

Sample

One hundred and eighty-one adults were randomly selected from participants in a population-based, longitudinal study of the diagnosis of colorectal cancer and the quality of life following diagnosis. All had been diagnosed with a first, primary diagnosis of colorectal cancer between January 1, 2003 and December 31, 2004, were aged between 20 and 80 years, and had been recruited through the Queensland Cancer Registry. During a telephone interview, these 181 potential participants were asked

Results

The mean and median time between Interview 1 and Interview 2 was 34 days (range = 27–44 days). The sociodemographic and disease-related characteristics of the study participants are presented in Table 2.

Discussion

To our knowledge, this is the first study to assess the test–retest reliability and validity of self-reported cancer diagnostic tests. Such information is important because self-report will continue to be widely used in cancer epidemiology studies. Our findings suggest that the accuracy of self-reported diagnostic information is variable, depending on the type of information being elicited. Overall, we found the test–retest reliability to be moderate to high, and the validity moderate to poor.

Acknowledgments

The authors wish to acknowledge the general practitioners who kindly completed the questionnaires for this study. This project was funded by The Cancer Council Queensland.

References (26)

  • A. Jemal et al.

    Cancer statistics, 2005

    CA Cancer J Clin

    (2005)
  • A. Redaelli et al.

    Screening, prevention and socioeconomic costs associated with the treatment of colorectal cancer

    Pharmacoeconomics

    (2003)
  • L. Caplan et al.

    Validity of women's self-reports of cancer screening test utilization in a managed care population

    Cancer Epidemiol Biomarkers Prev

    (2003)
  • Cited by (0)

    View full text