Multimorbidity and Polypharmacy: How to Reduce the Harmful Drug Load and Yet Add Needed Drugs in the Elderly? Proposal of a New Drug Classification: Fit for the Aged

To the Editor: Drug therapy is the most relevant therapeutic intervention in medicine; elderly patients tend to suffer from multiple diseases (multimorbidity) and thus are likely to receive multiple drug treatments (polypharmacy). It has been demonstrated that patients aged 65 and older take five or more drugs in 44% (male) and 57% (female) of cases and 10 or more drugs in 12% of cases.1 Taking 10 and more drugs is unpredictable and expensive and seems to cause more harm than good, given that up to 100,000 deaths in the United States annually are attributed to medications.2 One of the first formal attempts to improve drug safety in the aged population was the establishment of criteria for drugs to be avoided by Beers in 1997, which was updated in 2003.3 Applying the Healthcare Effectiveness Data and Information Set 2006 measure to nearly 1 million U.S. veterans, potentially inappropriate prescribing (PIPE) occurred in 19.2% of older men and 23.3% of older women.4 The evidence for the effectiveness of the Beers list is not compelling; a recent study showed no correlation with adverse effects if a patient received PIPE.5 For example, it does not allow for amiodarone treatment, although this may be necessary in some patients with atrial fibrillation and thus needs to be specified in the exceptions. The other shortcoming of the Beers approach is that an important aspect remains uncovered, namely, that there should also be a positive labeling of drugs that are indispensable in elderly people as data on morbidity, mortality, and safety are available or emerging for this particular treatment group. Although insufficient in major therapeutic areas at present, there is increasing clinical evidence for beneficial action of, for example, antihypertensive drugs (e.g., Systolic Hypertension in Europe Trial,6 Hypertension in the Very Elderly Trial7) and lipid-lowering agents (e.g., Prospective Study of Pravastatin in the Elderly at Risk Trial8) in elderly people. It is therefore proposed that drugs be labeled taking into consideration their usefulness for elderly people (Table 19) as does the Food and Drug Administration classification for the safety of drugs in pregnancy, ranking drugs from A through D (plus X for high risk). Thus, the proposed classification would be an extension of the Beers approach into the positive listing of valuable drugs. This seems necessary, because overtreatment and undertreatment are both typical problems of the aged population. Undertreatment, for example, relates to the poor control of arterial hypertension or anticoagulation in elderly people10 and may leave more than half of the patients un- or undertreated. This positive drug assessment (much more than the negative one, see above) is based on evidence from controlled trials. The assessment must take the real-life situation into account. Because compliance, dosing, application, and use errors are important especially in elderly people and in controlled trials, adherence is carefully enforced, thus not reflecting the situation in practice. The labeling of drugs requires a two-step process similar to the ones employed by health technology assessment institutes such as the National Institute for Clinical Excellence in the United Kingdom. The first step is a systematic review of available evidence and data from the literature or company repositories, and the second step is a consensus process based on those data, involving major stakeholders in relation to the treatment (e.g., scientific societies, regulatory authorities, patient organizations, drug industry, healthcare insurers). If no or only little direct evidence for elderly patients is available, a consensus-driven classification would be available. For new drugs, it should become part of the market authorization process. For marketed drugs, it should become a mandatory addition to the drug information documents. The limitations to this process will only show if the classification is widely accepted and introduced; it is anticipated that consensus for classification could vary for particular drugs, indications, and interest groups. Thus, endorsement by regulatory authorities would be required to settle controversial cases. A pilot study in a limited setting (e.g., in a pharmacovigilance study) needs to precede the regulatory action (labeling); this study should validate the utility of the proposed classification as demonstrated by less PIPE and greater use of necessary drugs. The main advantage of this proposal over the Beers approach is thought to be the protection or even promotion of drugs that are known to be efficacious and safe in elderly patients that is based on evidence. The typical use of the scheme would address general practitioners who receive multiple medication advice from different medical specialities for their patients. They would then synthesize the recommendations into a rank order of drugs that they could—with some rational judgment on top of the scheme—use to cut the list short. If important drugs are lacking, they could even add important drugs. The utility of the classification could be limited to outpatients in general practices as the most prevalent patient group, and alternate classifications may be derived for other groups based on frailty, mobility, and other determinants. Conflict of Interest: The editor in chief has reviewed the conflict of interest checklist provided by the author and has determined that the author has no financial or any other kind of personal conflicts with this letter. The author was employed by AstraZeneca R&D, Mölndal, as director of discovery medicine (translational medicine) from 2003 to 2006, while on sabbatical leave from his professorship at the University of Heidelberg. Since returning to this position in January 2007, he has received lecturing and consulting fees from Sanofi-Aventis, Novartis, Takeda, Roche, Pfizer, Bristol-Myers, Lilly, and Novo-Nordisk.