Introduction NSAIDs consequences are amongst the commonest iatrogenic problems in medicine,resulting in >40% of peptic ulcers1and thrombosis prompting guidance on their use.2Although aspirin has an established role in secondary prevention of vascular diseases,3guidelines recommend aspirin should not be used in primary prevention of vascular disease in those with hypertension or diabetes.4PPI play an important role in reducing the risk of peptic ulceration in NSAIDs with guidelines stating PPI should be used in all taking NSAIDs.5Guidelines recommend PPI use to prevent peptic ulceration with clopidogrel6though with PPI observed to reduce clopidogrel effects on platelet aggregation prompted guidance change, advising avoiding PPI use in those taking clopidogrel.7We performed a population-based study to observe how these changing recommendations have affected prescribing practice.
Method Using the CPRD, prescribing rates were calculated in 6 month periods from 2003 to 2012 in a repeated cross-sectional study. For every observational period, we included patients between 45–100 years old, considering them exposed to a drug if they had at least 2 prescriptions for that drug. For prescribing rates the total number of patients was used as denominator, while for co-prescribing rates the number of patients exposed to NSAIDs or aspirin was used. Segmented regression was used to estimate changes in prescribing levels and in trends before and after guidelines were issued.
Results The proportion of patients prescribed aspirin increased from 10.7% in 2003 (n = 127221) to 12.1% in 2012 (n = 169288). The trend change between observation periods before and after guidelines (4) was from a 37% (95% CI 32%, 42%) increase to a 35% (95% CI 23%, 48%) decrease. The trend change between observation periods before and after advice (2) was a 12% (95% CI 2%, 21%) increase to a 4% (95% CI 15%, 23%) decrease. PPI co-prescribing rates increased from 15% to 36% in aspirin users and from 12% to 46% in NSAID users, growing both before and after recommendations but without any statistically significant change in trend. A reduction (–11.7% [95% CI –12.6, –10.7]) was observed in co-prescription of PPI and clopidogrel in the year following recommendations of avoidance such practice.
Conclusion Aspirin prescribing guidelines were associated with a significant impact on its prescription for primary prevention of vascular disease. PPI co-prescription with NSAIDs fall far short of the universal use that NICE recommends. Concerns raised about a harmful interaction between PPI and clopidogrel caused a substantial fall in PPI co-prescription.
Disclosure of interest None Declared.