Preventable and Non-Preventable Risk Factors for Adverse Drug Events Related to Hospital Admission in the Elderly

J. Doucet, A. Jego, D. Noel, C.E. Geffroy, C. Capet, A. Coquard, E. Couffin, A.L. Fauchais, P. Chassagne, D. Mouton-Schleifer and E. Bercoff


Clin Drug Invest. 2002;22(6) 

In This Article


Many circumstances, which are often connected, lead to increased prevalence of ADEs in the elderly, but it is useful to emphasise the preventable factors to improve effective avoidance of ADEs.[21,22] To our knowledge, no previous study has evaluated the importance of the role of each risk factor in ADEs, or if these risk factors are preventable.

McElnay et al.[23] developed a risk model for ADEs in the elderly from several data using physiological, social and pathological status of patients. We only considered the pharmacological conditions of the drugs (excess dose, DDI, interference with acute or chronic disease), and it is not therefore possible to compare our results.

Our results showed that ADEs do not occur preferentially with recently prescribed drugs; they occurred both with drugs prescribed for less than 1 month (43.7%) and with drugs prescribed for 1 month or more (55.3%). However, when an ADE occurred as a result of a drug prescribed for 1 month or more, an interfering acute disease, usually acute dehydration, was statistically more frequent (p < 0.05). Data on drug monitoring usually concern short-term, not long-term, prescriptions, which may explain why the monitoring of drugs is discontinued when they are well tolerated for a long time, even though the risk of ADEs persists.[3] Our results suggest that monitoring of elderly patients should be continued over a long period to include the occurrence of an interfering acute disease and possibly stop a drug with an increased risk of an ADE from being prescribed.

One or more DDIs led to 60.6% of ADEs. This prevalence corresponds to previous results.[14] DDIs have been shown to be related to the number of drugs administered per patient.[3,4,24] In our study, the mean number of drugs was seven per patient, which was similar to the number of drugs reported in several previous studies.[5,14,25,26] According to Beers' definition, 46.5% of DDIs were considered to be preventable:[27] contraindicated drug combinations, potentialisation of lateral effects (e.g. two anticholinergic drugs), lack of dose adaptation in each drug (e.g. combination of a sulphonylurea with an oral anticoagulant). The other DDIs (53.5%) were considered to be nonpreventable because they were a result of validated combinations. In these cases, ADEs were a consequence of pharmacological modifications that were not available at the time of the prescription.

In the present study, one or more interfering chronic diseases were not taken into account by the physician at the time of the prescription of a new drug in 11% of the ADEs. The interfering chronic disease was usually chronic renal failure, but several other conditions were found in previous studies, namely: orthostatic hypotension (when prescribing antihypertensive or psychotropic drugs), prostatic hypertrophy (when prescribing anticholinergics drugs), cardiac failure or arteritis (for -blockers), chronic respiratory failure (for hypnotics or opioids) or malnutrition (anticoagulants).[9,21]

An excess drug dosage was administered in 14.8% of ADEs. These results are probably underestimated because reference dosages available to physicians are not always adapted to the elderly population. This occurs for low molecular weight heparins and for several psychotropic drugs.

We based our evaluation on risk factors of ADEs that were preventable at the time of the drug prescription or by adequate monitoring: preventable DDIs, no dose adaptation taking into account associated chronic or acute diseases, and excess doses. A total of 40.2% of the ADEs were considered to be preventable because all the risk factors were preventable. Previous studies have estimated that 30 to 55% of ADEs were preventable.[6,7,21,28,29] However, these studies did not consider combinations of preventable risk factors, and removing only one of two risk factors does not remove the risk of an ADE.

Although ADEs in the elderly might be prevented if the risk factors were better known, these risk factors are numerous and not always preventable. Our study only evaluated some preventable risk factors. We have stated earlier why poor compliance was not evaluated. On the other hand, we did not evaluate other situations of inappropriate prescriptions (defined as prescriptions with no demonstrated efficacy or failure to change to a less dangerous drug), which have been reported for psychotropic drugs for example.[4,27,30,31] However, reference data available to French physicians are not always accurate.