Inventi Impact: Clinical Research

Inventi:pcr/19045/16

Examining Variations in Prescribing Safety in UK General Practice: Cross Sectional Study Using the\nClinical Practice Research Datalink

STUDY QUESTION\nWhat is the prevalence of different types of potentially\nhazardous prescribing in general practice in the\nUnited Kingdom, and what is the variation between\npractices?\nMETHODS\nA cross sectional study included all adult patients\npotentially at risk of a prescribing or monitoring error\ndefined by a combination of diagnoses and\nprescriptions in 526 general practices contributing to\nthe Clinical Practice Research Datalink (CPRD) up to 1\nApril 2013. Primary outcomes were the prevalence of\npotentially hazardous prescriptions of anticoagulants,\nanti-platelets, NSAIDs, �² blockers, glitazones,\nmetformin, digoxin, antipsychotics, combined\nhormonal contraceptives, and oestrogens and\nmonitoring by blood test less frequently than\nrecommended for patients with repeated prescriptions\nof angiotensin converting enzyme inhibitors and loop\ndiuretics, amiodarone, methotrexate, lithium, or\nwarfarin.\nSTUDY ANSWER AND LIMITATIONS\n49 927 of 949 552 patients at risk triggered at least one\nprescribing indicator (5.26%, 95% confidence interval\n5.21% to 5.30%) and 21 501 of 182 721 (11.8%, 11.6% to\n11.9%) triggered at least one monitoring indicator. The\nprevalence of different types of potentially hazardous\nprescribing ranged from almost zero to 10.2%, and for\ninadequate monitoring ranged from 10.4% to 41.9%.\nOlder patients and those prescribed multiple repeat\nmedications had significantly higher risks of triggering\na prescribing indicator whereas younger patients with\nfewer repeat prescriptions had significantly higher risk\nof triggering a monitoring indicator. There was high\nvariation between practices for some indicators.\nThough prescribing safety indicators describe\nprescribing patterns that can increase the risk of harm\nto the patient and should generally be avoided, there\nwill always be exceptions where the indicator is\nclinically justified. Furthermore there is the possibility\nthat some information is not captured by CPRD for\nsome practicesâ��for example, INR results in patients\nreceiving warfarin.\nWHAT THIS STUDY ADDS\nThe high prevalence for certain indicators emphasises\nexisting prescribing risks and the need for their\nappropriate consideration within primary care,\nparticularly for older patients and those taking\nmultiple medications. The high variation between\npractices indicates potential for improvement through\ntargeted practice level intervention.\nFUNDING, COMPETING INTERESTS, DATA SHARING\nNational Institute for Health Research through the\nGreater Manchester Primary Care Patient Safety\nTranslational Research Centre (grant No\nGMPSTRC-2012-1). Data from CPRD cannot be shared\nbecause of licensing restrictions.