Opioid Prescriptions Analysis in Liverpool

The following data were extracted from patient Egton Medical Information System (EMIS) records: anonymised ID, age, ethnicity, gender, GP practice code, GP partial postcode, name of opioid, dose and quantity prescribed, date prescription was added to patient record, most recent issue date, course status (past or current) and any reported problems linked to the opioid prescription. Liverpool CCG (LCCG) acted as the gatekeeper and obtained verbal consent from GP practices to share patient information. Sixty-two of the 88 GP (70.5%) practices located across LCCG agreed to share patient data. An extract report was uploaded onto EMIS web, the data was extracted and then saved onto a secure network in an Excel spreadsheet. The data was pre-processed using Microsoft Excel, after which 93,236 prescriptions written for 30,474 patients remained (see Fig 1).
All prescriptions were cross-referenced with the British National Formulary (BNF) and re-coded according to their active opioid ingredient. This resulted in 12 groups including: oxycodone, tramadol, matazinol, methadone, morphine, tapentadol, pethidine, fentanyl, codeine, buprenorphine, dihydrocodeine and hydromorphine. Opioids commonly indicated for cancer or drug dependence (including dextropropoxphene, diamorphine, alfentanil, coproxomol, galenphol, oxylan and pavacol) were excluded. Dosage instructions were recoded to facilitate calculation of MED; if missing, maximal possible daily dose provided by the BNF was used.