Use of drug combinations in patients with osteoarthritis: a population-based cohort study

Abstract

Background Patients with osteoarthritis (OA) use different drugs in search for relief, with potential impact on both safety and costs. There is a scarcity of data on drug utilization in these patients. Objectives We used a large database including primary care medical records and pharmacy invoice data to explore use of the following drugs in OA patients in the period 2006-2010: non-steroidal anti-inflammatory drugs (NSAIDs), Symptomatic Slow Acting Drugs for OA (SYSADOA), COX-2 inhibitors (COX2i), opioids, and other analgesics (paracetamol and metamizol). Methods SIDIAP contains anonymised clinical information from primary care records and pharmacy invoice data for a representative >5 million people in Catalonia (North-East Spain). We screened this database (www.sidiap.org) to identify those aged 40 years or older with a new diagnosis of OA using ICD-10 codes. Among these, we ascertained prevalence of use of oral NSAIDs, COX2i, SYSADOA, opioids, and analgesics (all alone and in combination) in the period 2006-2011. We estimated prevalence (and 99% Confidence Intervals) of use of these drugs and their combinations, and the proportion of occasional (MPR≥25%) and regular (MPR≥50%) users for each of them. Results We identified 281,356 patients with an incident clinical diagnosis of OA. Among these, 128,314(45.6%) were treated with 3 or more drugs, and 57,835(20.6%) with none of the drugs studied. The 3 most common combinations of two drugs were: oral NSAID+analgesic (12.2%); 2.topical NSAID+analgesic (2.5%); and oral NSAID+SYSADOA (1.8%). Besides, 34,672(12.3%) patients received only 1 drug, the 3 most common being, in descending order: oral NSAID, analgesics, and SYSADOA. The proportions of occasional and regular users were highest for SYSADOA (24.5% occasional and 6.9% regular users), and lowest for opioids (7.3% occasional and 1.9% regular users). Conclusions About 75% of OA patients are treated with at least 2 drugs, and above half of them with 3 or more. However, compliance with these drugs is very low, and hence both clinical benefits and safety issues are difficult to predict using data from clinical trials, where compliance is high. These data are of interest for health care provision and clinical management of patients with OA. Disclosure of Interest D. Prieto-Alhambra Grant/research support from: Unrestricted grant from BIOIBERICA SA, N. Wilson: None Declared