|Home | About | Journals | Submit | Contact Us | Français|
Editor—For almost 50 years American Hospital Formulary System Drug Information (AHFS DI) has had the most widely vetted, evidence based process for content development of any drug compendium in the US—a process that is not limited to the professional labelling (package insert) approved by the Food and Drug Administration (FDA) but includes extensive ongoing review and professional analysis of the medical literature identified through resources such as PubMed (Medline).1,22
Vidal et al's criticisms should have been directed principally at the inadequacies of the drug regulatory processes and pharmaceutical manufacturers in researching and publishing such information rather than at the drug information publishers (p 263).3 In the US the FDA does not require manufacturers to conduct research establishing the impact of renal function on dosing. Instead an FDA guidance document merely provides opinion and advice and carries no regulatory force in stimulating this research.1 Thus, whether studies should be conducted in renally impaired patients remains at the manufacturer's discretion,4 and we can confirm that little specific primary information is currently published.
AHFS DI is in fact fully documented.1,5 References are not available in the print version because of space limitations, but the users' guide to AHFS DI clearly explains how they can be accessed.1,5 Since 1984, reference citations have been accessible as part of the electronic versions of AHFS DI (www.ahfsdruginformation.com).1,5 Currently, electronic access to 72 000 specific reference citations is available for around 480 000 statements, covering 81% of AHFS DI monographs. Documentation for the remainder is in the AHFS DI archives, from 1959.
Such access shows that the principal source of AHFS DI commentary on drug use for renally impaired patients originates in the manufacturers' FDA approved labelling, simply because it is often the only readily accessible source. Thus, Vidal et al's criticisms that the qualitative and undefined terms found in studied drug information sources are ill suited for practical use rightfully belong with the FDA and manufacturers, not AHFS DI. We cannot create such precise and specific information without adequately documented data, which are generally limited.
Roughly 30% of the total number of pertinent (systemically administered) drugs in AHFS DI have recommendations based quantitatively on specific creatinine clearances compared with the 20% reported for the subset of drugs studied by Vidal et al. We have no expeditious way of determining the number of additional drugs that would meet their second criterion for a “precise recommendation” defined by use of a term such as “avoid.”
We cannot comment specifically about the discrepancies they observed among the four drug information sources studied. Generally, AHFS DI would defer to precautionary statements on use in renal impairment found in FDA approved labelling, which is typically more comprehensive than similar product literature in other countries. Exceptions would include a notable body of evidence from the primary literature modifying or contradicting the FDA labelled information and compelling evidence from foreign regulatory sources. Therefore, some observed discrepancies may simply reflect different international regulatory processes and conclusions.
We are concerned by Vidal et al's method of simply choosing the top 100 prescribed drugs from a single hospital over potentially more clinically relevant drugs for which both narrow therapeutic ranges and the importance of renal function in elimination have been established. Regulatory issues in the US work against clearly defining such a subset of drugs. The lack of mandated renal research and the current FDA practice of encouraging clinical trials that study maximally tolerated dosages in otherwise healthy adults contribute to this difficulty. As a result, there is little incentive to establish minimally effective dosages in renally competent adults, let alone in special populations such as those with renal, hepatic, or age related impairment.
These important factors substantially impair any attempt by publishers to establish clear dose-effect (both efficacy and toxicity) relations and make developing precise recommendations extremely difficult.
We agree that the dose and dosing interval, contraindications, and expected adverse effects should be no less evidence based than the efficacy and effectiveness of a drug. Unfortunately, few such data support this evidence based evaluation by publishers for most drugs. A focus for future change should be stimulating and publishing research on the impact of renal function on dosing.
Competing interests: None declared.
See Papers p 263