Background Drug dosage errors may be found during pharmaceutical validation in the treatment of Alzheimer’s disease. An appropriate maintenance dosage must be determined to optimise drug therapy.

Purpose To determine the incidence of drug dosage errors in the treatment of Alzheimer’s disease in patients admitted to a tertiary hospital.

To analyse the causes of drug dosage errors and to evaluate the rate of acceptance by the physician of the pharmaceutical intervention (PI) recommending a correct dosage.

Material and methods A 3 month prospective, interventional, analytic study (July to September) was performed.

All inpatients taking any of the drugs for the treatment of Alzheimer’s disease were included.

Patients were selected using the computer prescription order entry (CPOE). The pharmacist, advised by a geriatrician, reviewed the dosage of these drugs on a daily basis.

In order to verify the correct dosage and to identify the possible cause of the error, the pharmacist reviewed the clinical history for every selected patient.

Whenever a drug dosage error was identified, a PI took place, with the pharmacist sending a dosage recommendation to the physician through the CPOE.

Anthropometric data (age, gender) as well as prescribed drugs and drug dosage were collected.

The incorrect doses, causes of the dosage error and degree of acceptance of the PI were counted.

Results 64 patients were included. Average age: 83.4 years, 64% women.

We reviewed 74 prescriptions with the following drugs: rivastigmine 37.9% (28), donepezil 25.6% (19), galantamine 9.5% (7) and memantine 27% (20).

There was a dosage error in 28.4% of prescriptions, all due to lower doses than recommended.

The causes of the errors were: 52.4% wrong dosage prior to admission, 28.6% incorrect reconciliation of home treatment and 19.0% incorrect record in the CPOE by the physician.

PI was performed in 85.7% of prescriptions with dosage errors. 16.6% of PIs were accepted. All of the accepted PIs were concerned with reconciliation errors.

Conclusion More than a quarter of the reviewed prescriptions were wrong. The low acceptance of PI may be due to the physician’s belief that long term treatment does not affect the clinical course of the acute process that caused admission to hospital.

References and/or Acknowledgements Thanks to the co-authors.

No conflict of interest.