Background National guidelines endorse prescription of medications in anticipation of symptoms in the last days of life. Little is known about current practice in anticipatory prescribing (AP). A service evaluation was undertaken to evaluate AP across one county in the UK.

Method Electronic records were searched of patients known to the Bedfordshire PEPS (Partnership for Excellence in Palliative Support) Co-ordination Centre, who had died during a one-year period and had been prescribed injectable medications for symptom control.

Results Out of 392 eligible patients, the records of 132 were selected at random. 486 injectable medications were prescribed: 77% were in anticipation of symptoms and 23% in response to existing symptoms. Diamorphine and midazolam were most commonly prescribed in anticipation, with 43/88 (49%) and 41/93 (44%) prescriptions leading to drug administration respectively. The corresponding data for glycopyrronium, cyclizine, haloperidol and metoclopramide were 25/78 (32%), 15/67 (22%), 3/11 (27%) and 3/14 (21%). Overall, only 37% of all medications prescribed in anticipation were administered. Recognition of the need for AP came from palliative care nurses (50%), GPs (32%) and District Nurses (14%).

Most patients had malignant disease (87%). The median time between prescription and first drug administration was 9 days in patients with cancer (range 0–368), and 61 days in those with non-malignant disease (range 3–298).

Conclusion Most prescribing of injectable medications was in anticipation, rather than in response to, symptoms. About a third of anticipatory prescribing led to drug administration. Although diamorphine and midazolam were both commonly prescribed and administered, antiemetics were less so. Further work is needed to evaluate the cost-effectiveness of AP, particularly in relation to antiemetics, and to investigate the unexpectedly long interval between drug prescription and administration in patients with non-malignant disease.