advertisement
Editorial IGR 11-4
Editorial:
Do randomized clinical trial results need to be validated in the clinical setting?
Jeffrey M. Liebmann and Carlos Gustavo de Moraes
Prospective, randomized clinical trials provide extraordinarily useful information that can be used to advance patient care. This has been particularly true for the recent, large, multi-center, glaucoma clinical trials, each of which has contributed to our understanding of disease pathophysiology, progression, and responsiveness to treatment. In general, these research studies confirmed long-standing medical consensus regarding the diagnosis and treatment of glaucoma across the spectrum of glaucomatous disease, ranging from ocular hypertension to moderately advanced glaucomatous optic neuropathy and visual field loss. Clinicians, researchers and patients have all benefited from these projects.
One of the most important limitations of these prospective, carefully monitored projects is that the study population does not always resemble the patients that are seen in daily clinical practice. For example, subjects in the Ocular Hypertension Treatment Study were primarily of European or African ancestry, had refractive errors less than six diopters, intraocular pressures at the time of study entry between 24 and 32 mmHg in one eye and 21 and 32 mmHg in the fellow eye, and were adept at achromatic perimetry. Therefore, the results of the study may or may not be directly applicable to the wider, more diverse United States and global populations with disorders such as angle closure, secondary open-angle glaucoma mechanisms, high myopia, intraocular pressures less than 21 mmHg or greater than 32 mmHg, and patients who have difficulty with perimetry. These types of challenges, present in all research, suggest that it might be worthwhile to design and organize clinical research studies to confirm (or deny) these results in our clinical practices. One way to assess whether the results from prospective trials are applicable to clinical practice is to develop and analyze ever-larger databases of information gleaned from clinical practice. While the selection bias inherent using this approach is likely greater than that of a closely-monitored, randomized clinical trial with its strict inclusion and exclusion criteria, an appropriately large number of subjects and associated data (optic nerve imaging, intraocular pressure, corneal thickness, perimetry, treatment interventions, etc) would mitigate this effect. The need for this type of research, sometimes called 'Practical Clinical Trials', has been recognized in other fields of medicine to be an important link between the rarefied world of the controlled trial and the vagaries of day-to-day patient management by physicians with dissimilar patients and training.1 Data derived from clinical practice has the additional advantage of reflecting the adoption of new technologies, testing paradigms and treatments as they are incorporated into clinical practice.
Exploration of existing databases in ophthalmology is not new and much information has been gleaned from post-hoc analyses of prospective study databases, including governmental and insurance company databases, even when these studies were not initially designed with specific ancillary questions in mind. Yet, numerous challenges exist when developing these types of databases. Patient data, from demographics to diagnostic testing, is not always uniform, reproducible, or appropriately collected. Our own experience with a clinical database consisting of over 40,000 patients and 130,000 visual field tests, however, suggests that given large enough databases, it is possible to both confirm the validity and importance of the various glaucoma risk factors identified by the randomized trials, but also to identify, confirm or enhance our understanding of risk factors and disease progression.
In summary, research projects that take advantage of large, existing clinical databases can be used to link clinical trial results to clinical practice and potentially broaden the applicability of randomized clinical trials to wider groups of patients. Ultimately, the validity of the research will depend on the collection and analysis of high quality data. Collaboration among clinicians and/or research centers with appropriate biostatistical support can use this clinical information to expand our understanding of glaucoma and better identify better treatment modalities and interventions.