Based on solid evidence, NSAIDs reduce the risk of adenomas, but the extent to which this translates into a reduction of CRC is uncertain.
|Study Design: No adequate studies with CRC outcome.|
|Internal Validity: Not applicable (N/A).|
|External Validity: N/A.|
Based on solid evidence, harms of NSAID use are relatively common and potentially serious, and include upper gastrointestinal bleeding, chronic kidney disease, and serious cardiovascular events such as myocardial infarction, heart failure, and hemorrhagic stroke.
Magnitude of Effect: The estimated average excess risk of upper gastrointestinal complications in average-risk people attributable to NSAIDs is 4/1,000 to 5/1,000 people per year. The excess risk varies with the underlying gastrointestinal risk, however, it likely exceeds ten extra cases per 1,000 people per year in more than 10% of users. Serious cardiovascular events are increased by 50% to 60%.
|Study Design: Evidence obtained from RCTs and high quality systematic reviews and meta-analyses.[13,14]|
|Internal Validity: Good.|
|External Validity: Good.|
Based on solid evidence, daily aspirin (acetylsalicylic acid [ASA]) for at least 5 years reduces CRC incidence and mortality. This is based on two reports of extended follow-up of two RCTs [16,17] and meta-analysis of observational studies. A third report that adds extended follow-up of an additional two RCTs (with a meta-analysis of all four RCTs) adds certainty to this conclusion.
Magnitude of Effect: After 23 years of follow-up, incidence of CRC in the placebo group was 3.8% and in the ASA group 2.5% (hazard ratio [HR] = 0.63; 95% CI, 0.47–0.85). In the report from all four RCTs, the 20-year risk of death due to CRC in trials that allocated ASA for at least 5 years was reduced by about 40% (HR = 0.60; 95% CI, 0.45–0.81), absolute risk reduction was from about 3.1% to 1.9%. The primary effect was on mortality from proximal colon cancer.
|Study Design: Extended follow-up of four RCTs.|
|Internal Validity: Fair, some data from registries and death certificates, some loss to follow-up; variation in ASA dose; adherence to ASA unknown after end of trials (5–9 years).|
|External Validity: Fair, most data from British men; fewer data from women.|