Dose variations associated with formulations of NSAID prescriptions for children: a descriptive analysis of electronic health records in the UK.

Background: NSAIDs, particularly ibuprofen, are commonly prescribed for children but there is limited published research on real-life prescribed doses for this class of drugs.

Objective: The aim of the study was to investigate if variations in NSAID doses prescribed to children can be explained by patient age, indication, dosage form, type of NSAID or year of prescription.

Study Design: Recorded daily doses for drugs within the 'Anti-rheumatics, non-steroidal plain' anatomical classification were studied. First prescriptions of a distinct NSAID substance within 13-month time periods in a patient's history were included. To enable grouping and comparison of NSAIDs, doses were analysed as prescribed daily doses (PDDs) relative to the adult defined daily dose, stated as the relative PDD (rPDD) in this study. Multiple regression analysis was performed with the rPDD as the response variable, and age, indication, dosage form, NSAID substance and year of prescription as the explanatory variables.

Setting: Prescriptions from the Intercontinental Medical Statistics (IMS) Health Disease Analyzer database containing electronic health records of general practitioners in the UK issued from 1988 to December 2005.

Patients: Data for children aged 2-11 years with NSAID prescriptions including daily dose information.

Results: A total of 21 473 first prescriptions for 19 695 patients were studied. The vast majority of prescriptions were for ibuprofen (n = 20 855), which were therefore analysed separately. The other NSAID prescriptions were grouped (n = 618), containing diclofenac, indometacin, mefenamic acid, naproxen and piroxicam ('NSAID group'). The rPDD varied considerably with dosage form in both the ibuprofen and NSAID groups. In particular, tablets/capsules were prescribed at higher doses than liquid dosage forms. In the NSAID group, naproxen was prescribed at noticeably higher doses. The rPDD varied only slightly with age in both groups. Prescriptions indicated for rheumatic disease were associated with lower doses than other indications in the NSAID group. The rPDD was not influenced by year of prescription.

Conclusions: This study shows a correlation between higher prescribed NSAID doses and tablet/capsule formulation, and highlights the need for careful choice of dose formulation when prescribing medicines for children.