Previous studies have found evidence suggesting paracetamol, as compared to ibuprofen, increases the risk of several health conditions, namely stomach ulcers, heart attack, stroke, kidney failure and death. Given an apparent relationship between paracetamol and subsequent adverse health events, one question is whether paracetamol causes these conditions or whether it reflects the characteristics of patients before getting the drug. An important possible reason for the latter pattern occurs when patients are either much more or much less likely to be prescribed certain medications because of a pre-existing health condition. An existing health condition may cause physicians not to prescribe a medication because of side effects that have been associated with the condition. For example, people with a stomach ulcer might not be prescribed ibuprofen because it can cause stomach bleeding and they are already at risk. Instead, they would get acetaminophen if needed for mild pain and would likely continue to be treated for their ulcer. Then the use of acetaminophen would appear to be related to subsequent stomach ulcer, too. The objective of this study is to determine whether there is evidence of physician prescribing for paracetamol versus ibuprofen in a medical records database based on pre-existing conditions.
Previous studies have found an association between paracetamol and several adverse events, including upper gastrointestinal bleeding, myocardial infarction, stroke, acute renal failure and death. Some of these studies used ibuprofen as a comparator1,2 and when discussing the results it was common to raise the question of channeling but rare to offer evidence. Given an association is identified and is reproducible; the question is whether it is causal or reflects bias. One important candidate for a possible source of bias is channeling.
The primary objective is to determine whether there is channeling bias in the first prescription of paracetamol vs. ibuprofen in an electronic medical records database. A second objective is to repeat the analysis using negative control conditions as a way of calibrating the amount of residual variation inherent in the data source or study design. A third objective is to assess the prospects to reliably control for bias through adjustment methods. The findings would have bearing on any studies that compare adverse event rates among subjects exposed to paracetamol and subjects exposed to ibuprofen.
We will use the CPRD to form a cohort of patients given a first prescription of single-ingredient paracetamol or ibuprofen. Frequencies of prior gastrointestinal bleeding, myocardial infarction, stroke, ant acute renal failure, each in turn, will be calculated to determine whether channeling was present. In validation, we will examine frequencies of prior diagnoses of approximately 36 negative control conditions which are known not to be associated with use of paracetamol versus ibuprofen. Additionally, a propensity score analysis will be performed to determine the proportion of overlap of factors that predict first prescription of paracetamol or ibuprofen.