The cumulative false-positive rate in colorectal cancer screening: a Markov analysis

Eur J Gastroenterol Hepatol. 2020 May;32(5):575-580. doi: 10.1097/MEG.0000000000001669.

Abstract

Background: Faecal occult blood testing is widely used in colorectal cancer screening. However, there is little empirical long-term evidence on the accumulation of false-positive test results over several screening rounds. We aimed to systematically explore and quantify the cumulative false-positive rate for various scenarios of colorectal cancer screening.

Methods: Using a Markov analysis, we estimated the lifetime cumulative number of false-positive test results (cumFP) per 100 000 50-year-old persons. We varied the screening interval and the specificity of a single screening test and the starting age of screening.

Results: For a test with a specificity of 98% used from 50 to 74 years, the cumFP at age 74 was 26 260 (1-year interval), 15 102 (2-year interval), and 10 819 (3-year interval), respectively. For a test with a specificity of, respectively, 95 and 92% used at a 2-year interval, the cumFP at age 74 was 2.2 times and 3.0 times higher as compared to a test with a specificity of 98%. The cumFP at age 74 was 18% lower for screening persons aged 54-74 years vs. 50-74 years.

Conclusion: Our findings quantitatively illustrate the large variation of the cumFP in colorectal cancer screening between screening strategies, which is relevant to informed decision making and adequate resource planning.

MeSH terms

  • Aged
  • Colonoscopy
  • Colorectal Neoplasms* / diagnosis
  • Colorectal Neoplasms* / epidemiology
  • Early Detection of Cancer*
  • False Positive Reactions
  • Humans
  • Markov Chains
  • Mass Screening
  • Middle Aged
  • Occult Blood