Deaths and cardiopulmonary events following colorectal cancer screening—A systematic review with meta-analyses

Research output: Contribution to journalJournal articleResearchpeer-review


  • Fulltext

    Final published version, 1.05 MB, PDF document

Colorectal cancer screening programmes (CRCSPs) are implemented worldwide despite recent evidence indicating more physical harm occurring during CRCSPs than previously thought. Therefore, we aimed to review the evidence on physical harms associated with endoscopic diagnostic procedures during CRCSPs and, when possible, to quantify the risk of the most serious types of physical harm during CRCSPs, i.e. deaths and cardiopulmonary events (CPEs).

Systematic review with descriptive statistics and random-effects meta-analyses of studies investigating physical harms following CRCSPs. We conducted a systematic search in the literature and assessed the risk of bias and the certainty of the evidence.

We included 134 studies for review, reporting findings from 151 unique populations when accounting for multiple screening interventions per study. Physical harm can be categorized into 17 types of harm. The evidence was very heterogeneous with inadequate measurement and reporting of harms. The risk of bias was serious or critical in 95% of assessments of deaths and CPEs, and the certainty of the evidence was very low in all analyses. The risk of death was assessed for 57 populations with large variation across studies. Meta-analyses indicated that 3 to 23 deaths occur during CRCSPs per 100,000 people screened. Cardiopulmonary events were assessed for 55 populations. Despite our efforts to subcategorize CPEs into 17 distinct subtypes, 41% of CPE assessments were too poorly measured or reported to allow quantification. We found a tendency towards lower estimates of deaths and CPEs in studies with a critical risk of bias.

Deaths and CPEs during CRCSPs are rare, yet they do occur during CRCSPs. We believe that our findings are conservative due to the heterogeneity and low quality of the evidence. A standardized system for the measurement and reporting of the harms of screening is warranted.
Original languageEnglish
Article numbere0295900
JournalPLoS ONE
Issue number3
Number of pages23
Publication statusPublished - 2024

ID: 385271249