The contribution of a clinical pharmacist to the improvement of medication at a geriatric hospital unit in Norway

It is well documented that the elderly often use many drugs and do not usually have sufficient knowledge concerning the correct use of the drugs. The result often becomes polypharmacy, poor compliance, and consequently, increased morbidity and hospitalisation. A more correct use of drugs will be cost-saving for the community and may improve the patient’s quality of life.1-4 On this background, the geriatric unit at Telemark Hospital wished to attach a clinical pharmacist to the therapeutic team for the evaluation of patients’ medication.

The unit, with a total of 18 beds, has elective admissions only. The patients are admitted for two weeks for rehabilitation and for the evaluation of functional impairment, loss of skills and drug treatment. These are not normally patients with severe mental disorders or patients who should rather have been admitted to a long term stay unit at a nursing home.

A clinical pharmacist often has a somewhat different approach to the use of drugs and may give valuable supplementary information about for example interactions, during the physician’s decision-making process concerning potential changes of and the follow-up of the medication. Outside the Nordic countries, several studies have been performed describing the benefit of clinical pharmacists in hospitals.5,6,7,8 However, in the Nordic countries, very few studies have been published.9,10,11

In the period April 2000 to June 2004 a clinical pharmacist participated at the pre-round meeting and ward round in the geriatric unit on a randomly chosen day each month. The pre-round meeting is an interdisciplinary meeting where plans for the patient’s treatment and for the ward rounds are made. The hospital had granted the pharmacist access to the patients’ records in advance. The method of the study has been assessed and accepted by the national ethical committee under the assumption that the patients give their consent. At the ward round, the physician informed the patients of the study and the patients gave their oral consent to the pharmacist’s assistance in the evaluation of their medication.

Approximately one hour before the pre-round meeting the pharmacist went through the medication and other relevant information in the patient record for one of the two patient groups in the unit. This was chosen at random. The other patient group was not used as a control group to record changes in medication without assistance of the pharmacist. None of the patients in the study group were to be counted more than once in the case of rehospitalisation at a later stage. The current medication of the patient, both that which was previously initiated and any changes made so far during the hospitalisation, was evaluated with regard to potential drug-related problems, table 1.12,13 Problems concerning drug information were not part of the study. Proposals concerning monitoring (measurements of serum concentrations, laboratory samples and other examinations to be able to optimise the drug treatment, directly or indirectly) were made in cases where the medical records did not give any information on whether the samples or examinations had already been ordered by the physician, or in cases where the results from these examinations had not been assessed.

Table 1
The physician’s decision about changes in medication based on drug-related problems proposed by the pharmacist.

The results of the physician’s decisions on the current day in relation to potential changes in medication proposed by the pharmacist were continuously recorded by the pharmacist according to a classification system prepared for this purpose, table 1. The further follow-up of the decisions on consequent days or after the patient’s discharge from the hospital was not a part of the study. The relevant drug was recorded and classified according to the Anatomical Therapeutic Chemical (ATC) Classification System.

Altogether, the pharmacist was present at the geriatric unit for 27 days during the study period, and the medication of 250 patients was evaluated for potential drug-related problems. None of the patients expressed any doubt in allowing the pharmacist to participate in the evaluation of their medication. An average of six minutes was used for each patient to record drug-related problems, before the pre-round meeting. The pharmacist used all relevant information about each patient which was systematically collected in the medical record: the card with current and previously changed fixed medication and on demand, the results of examinations and laboratory samples, the medical history, clinical risk factors, diseases and symptoms. A medication history was not performed.

For 188 of the patients (75%), at least one drug-related problem was identified. The mean age of the 188 patients was 81.6 years (range 59–95 years); 136 females and 52 males. At all except the three last occasions, the same physician, the geriatrist, participated. In addition, a pre-registration house officer and also occasionally a senior house officer were present. The physicians had the same understanding of the drug-related problems as the pharmacist. For these 188 patients, the physician made 606 decisions (mean 3.2 per patient) regarding potential changes to the patient’s medication based on the proposals from the pharmacist (table 2). Table 3 shows the 20 most frequently involved drug groups.

Table 2
Physician’s decision (%) to change the patient’s medication on the current day (n=606), based on drug-related problems proposed by the pharmacist

Table 3
Numbers of drugs evaluated by the physician, shown as the 20 most frequent ATC groups (n=606), based on drug-related problems proposed by the pharmacist

The absence of a control group in the study did not allow us to know how many changes in medication were made without the pharmacist’s participation, and if the clinical pharmacist actually increased the number of changes in medication. However, there was no doubt from the physician’s point of view that the contribution by a clinical pharmacist actually improved the patient’s medication.

Our study has focused on the physician’s decisions on the current day based on drug-related problems proposed by the pharmacist, rather than on the types and numbers of drug-related problems. Behind every single decision there may be several different drug-related problems. For example, high dosages, interactions, adverse effects, laboratory measurements and examinations may lead to the decision of dosage reduction. A classification system was prepared for this study, owing to the lack of an appropriate method (table 1). Australian scientists have published a somewhat different classification of physicians’ decision of drug-related problems.14 No attempt has been made to assess the clinical significance of changes for a single drug. Further follow-up of the decisions on consequent days or after the patient’s discharge from the hospital was not included in the study. The study will not give answers to the essential patient outcomes: to prevent, cure, arrest or slow diseases or symptoms. But the pharmacist contributes to the general patient outcomes by improvement of the drug therapy.

At least one drug-related problem was identified in 75% of the patients and a mean of 3.2 potential changes in medication were undertaken per patient. This is in accordance with a multi-centre study performed at six departments of internal medicine and two departments of rheumatology at five other Norwegian hospitals.11 In this study, at least one drug-related problem was identified in 81% of the patients. In the study by Pretsch and co-workers, drug-related problems were identified in 80% of the patients at a lung department.9 Most proposals of changes were made for central nervous system drugs and cardiovascular drugs (table 3). This is in accordance with other studies15,1617,18, and reflects the frequency of use of these drugs in the elderly.

Nearly half (42%, 255/606) of all the pharmacists proposals on the current day have resulted in either the discontinuation of medication or in dosage reduction, or a decision to make this assessment at a later stage during the hospitalisation (table 2). A direct comparison of these results with data from other studies is difficult as there are differences in the processing and presentation of the results. In the study by Viktil and co-workers, 24.7% of the patients in a geriatric unit received drugs unnecessarily and a non-optimal dosage was the most frequently reported problem.17

Nearly one third (184/606) of the physician’s decisions resulted in an increased need for further monitoring and the follow-up of clinical data, e.g. measurements of drug concentrations in serum, measurements of electrolytes, creatinine, uric acid, INR, blood glucose, iron status, and blood pressure. The cardiovascular drugs (ATC groups B and C) dominated (82%, 151/184), with diuretics on top with 35 of 184 decisions, followed by renin-angiotensin drugs, digitalis and antithrombotic drugs.

Drugs with anticholinergic effects are particularly unfavourable in the elderly, in particular with regard to cognitive function.19 These drugs, such as tricyclic antidepressants, sedative antihistamines and conventional neuroleptics, were therefore to a great extent withdrawn.

Most proposals were made concerning benzodiazepines (hypnotics and anxiolytics), table 3. These drugs, as well as tramadol and codeine-containing analgesics, were the drugs for which discontinuation or dosage reduction was most frequently performed or planned to be performed at a later stage during the hospitalisation (91 of 129 decisions, 71%). The discontinuation of medication was performed in cases where the patient was using low doses or in case of concomitant use of several benzodiazepines. These are drugs with a potential of addiction which lead to a general overconsumption, but also give increased tiredness, increased tendency to fall and the risk of cognitive failure. The pain regimen was also evaluated with regard to the improvement of pain relief.

Even if there initially was an interdisciplinary agreement concerning the drug-related problems, 11.6% of the physician’s decisions still resulted in an unchanged prescription (table 2). This can be explained by new information about the patient being forthcoming during the pre-round meeting and ward round. But the main reason was that many patients chose to maintain their original prescription themselves in spite of the information concerning adverse effects. The number of patients was unfortunately not registered in the study. The drugs involved are known to have an addiction potential, with hypnotic drugs as the most frequent (17/70), as well as anxiolytics and analgesic drugs. Hospitalisation is in itself a new situation for the patient making it even more difficult to change this type of medication in the elderly who may suffer from complex diseases and have problems concerning their social network. Further work with a gradual dosage reduction and a possible withdrawal was therefore addressed in the physician’s final report to the family physician. In the study by Pretsch and co-workers, 10% of the drug-related problems were not taken into account at the pre-round meeting, mainly because the physician did not agree with the pharmacist.9 In that study, however, the pharmacist did not participate in the ward round, and thus any possible objections by the patient against changes should be added.

In summary we conclude that a clinical pharmacist contributes to a more correct medication of the elderly, even with the modest contribution such as participation in the pre-round meeting and the ward round one day per month. The study may support the argument for the gradual introduction of a clinical pharmacy service, as this is not widely accepted in most hospitals in Norway.

Nurses and physicians at the Geriatric Unit, Telemark Hospital, Norway.