Rationale

In a specialized area, such as ophthalmology, it is not common for a practitioner to see the patient volume necessary to make a correlation between possible cause and effect of medication-related ocular disease. Postmarketing observational studies from multiple sources permit the evaluation of drug safety in a real-world setting where off-label use and various practice patterns occur. There is no question that this has limited ability to determine causation, but it can detect “signals” that alert the clinician as to adverse drug events. In subspecialty areas of medicine with comparatively limited markets, sometimes this is all that we have. A national registry specifically interested in a specialized area of medicine has filled a need, as shown by the three-plus decades of the National Registry of Drug-Induced Ocular Side Effects (NRDIOSE).

The National Registry, which is based at the Casey Eye Institute in Portland, Oregon, USA (www.eyedrugregistry.com), is a clearinghouse of spontaneous reports collected mostly from ophthalmologists from around the world. It is the only database that collects only eye-related adverse drug reactions (ADRs). The MedWatch program run by the US FDA (www.fda.gov/medwatch/index.html) collects ADRs on all organ systems in the USA and is another source for reporting data and requesting data. The Uppsala Monitoring Center, a branch of the World Health Organization (WHO), Uppsala, Sweden (www.who-umc.org) collects spontaneous reports on all organ systems from around the world and has more than 70 national centers that report to them, including the FDA. Finally, clinicians and patients frequently report an ADR directly to the drug company, who in turn periodically submits these spontaneous reports to the FDA.

Regardless of where an ADR is submitted, the various organizations mentioned above can be contacted with questions about an ADR or how many types of reports exist for specific drug–ADR combinations. The National Registry provides this information free of charge to ophthalmologists, and the FDA is required to provide this information to the public through the Freedom of Information Act. The WHO may charge a fee, depending on the type of information requested. The information from pharmaceutical companies should eventually end up in the FDA’s MedWatch database.

Recently, spontaneous reporting databases have adopted statistical analyses methods of interpreting ADRs. At the Uppsala Monitoring Center, for instance, a quantitative method for data mining the WHO database is part of the signal detection strategy. Their method is called the Bayesian Confidence Propagation Neural Network (BCPNN). An Information Component (IC) number is calculated based on a statistical dependency between a drug and an ADR calculated on the frequency of reporting. The IC value does not give evidence of causality between a drug and an ADR; it is only an indication or signal that it may be necessary to study the individual case reports in the WHO database. The IC value calculation is a tool that can guide the WHO to create a hypothesis of association between drugs and ADRs among the over three million case reports in the WHO database.

This method of analysis is also being adopted within the pharmaceutical industry and at the FDA. The National Registry is also able to use the IC values because its staff are consultants to the WHO. If a clinician suspects an ADR, especially if it may be a new drug-induced ocular side effect, he/she is encouraged to report this to the National Registry. Membership on the website is free, and it only takes 1–2 minutes to register. Frequent updates are also available on the website, informing clinicians of the latest in ocular pharmacovigilance.