Trends and variation in unsafe prescribing of methotrexate: a cohort study in English NHS primary care.

Background

Prescribing high doses of methotrexate increases the potentially fatal risk of toxicity. To minimise risk, it is recommended that only 2.5 mg tablets are used.

Aim

To describe trends in GP prescribing of methotrexate over time; the harm associated with methotrexate errors at a national level; ascertain variation between practices and clinical commissioning groups (CCGs) in their implementation of the safety guidance; and map current variations at CCG and practice level.

Design And Setting

A retrospective cohort study of English GP prescribing data (August 2010–April 2018), and data acquired via freedom of information (FOI) requests.

Method

The main outcome measures were: variation in ratio of non-adherent/adherent prescribing, geographically and over time, between practices and CCGs; and description of responses to FOI requests.

Results

Of 7349 practices in England, 1689 prescribed both 2.5 mg and 10 mg tablets to individual patients in 2017, breaching national guidance. In April 2018, 697 practices (≥90th percentile) prescribed >14.3% of all methotrexate as 10 mg tablets, likewise breaching national guidance. The 66 practices at ≥99th percentile gave >52.4% of all prescribed methotrexate in the form of 10 mg tablets. The prescribing of 10 mg tablets fell during the study period, with 10 mg tablets as a proportion of all prescribed methotrexate tablets falling from 9.1% to 3.4%. Twenty-one deaths caused by methotrexate poisoning were reported from 1993–2017 in England and Wales.

Conclusions

The prevalence of unsafe methotrexate prescribing has reduced but remains common, with substantial variation between practices and CCGs. The authors recommend investment in better strategies around implementation. As 21 deaths that occurred from 1993–2017 in England and Wales were attributed to methotrexate poisoning, the coroners’ reports for these deaths should be reviewed to identify recurring themes.