Background A treatment exchange protocol (TEP) contains three ways of intervening: a) suspend the prescribed treatment (not very useful), b) keep the patient on his home medicines during the hospitalisation, c) exchange the treatment (TE) to another drug included in the hospital formulary.

The protocol used not to be applied to multidose drugs commonly used by patients.

Purpose To assess the prescription of drugs not included in the hospital formulary (NIDHF) and to know the acceptability of recommendations for a change of treatment.

Materials and methods Observational, prospective, two-month study in a General Hospital. Every day The authors recorded new NIDHF prescriptions, age, sex and diagnosis. TEP was applied to everyone. When the recommendation wasn't accepted, The authors recorded the reason (eg, allergy). NIDHFs without a recommendation because of a lack of evaluation or agreement were quantified, knowing they were a therapeutic void in the hospital formulary (HF) and TEP.

Results The authors identified 251 NIDHFs from 209 patients (average age 66, 46% male). The authors obtained an average of six new prescriptions daily; half of them were replaced according to the TEP.

The analysis of the non-replaced NIDHF drugs was:

• ▶ 46% of NIDHF drugs were kept during hospitalisation. They were mainly oral antidiabetic and antidementia agents.

• ▶ In 12% the drug was acquired occasionally for a justified reason.

• ▶ In 7% they were multidose drugs.

• ▶ In 18% the TE was rejected.

• ▶ In 9% the TE wasn't available.

• ▶ In the remainder, the reason for rejection wasn't specified and patients provided treatment.

Clinical services with more NIDHF prescriptions were: Internal Medicine (40%) and Traumatology (10%). The lowest acceptance of recommendations was in Home Hospitalisation (45%) and Surgery (33%). The greatest number of prescriptions for drugs without agreed therapeutic exchange was in Pneumology (23%) and Otorhinolaryngology (28%).

Conclusions The adherence to the HF and acceptance of the TEP recommendations were high. Many of the requirements of NIDHF are solved with TEP. The study has enabled us to detect therapeutic areas in which the HF and TEP could be improved. Any changes must be reviewed and agreed with medical services to reduce the likelihood of adverse events and promote good-quality pharmacotherapy.