To the Editor:
We appreciate the insightful comments of Dr. Muller, et al1 that highlight the uses and limitations of hospital data and registry resources. We agree wholeheartedly with Dr. Muller, et al regarding the potential for hospital data, specialist care registries, and similar databases to address issues of healthcare use in different healthcare settings. Critical to the usefulness of the resource is the validity of the disease and procedure coding of records from all relevant sources and the representativeness of the studied population for interrogation of important clinical questions.
We also agree with Dr. Muller, et al regarding the importance of primary care medical records for research purposes. We are fortunate to have access to the Rochester Epidemiology Project, a unique resource that identifies the full medical record of all patients seen in the defined geographic area of Olmsted County, Minnesota2. This resource provides access to every medical provider contact, including primary care, subspecialty care, all surgeries and procedures, all hospitalization records, and all laboratory and prescription drug information of individual patients. This overcomes the biases inherent, for example, in studying referral patients and patients seen only in specific clinical settings such as oncology clinics, medical and pediatric subspecialty clinics, and surgery clinics.
REFERENCES
- 1.
- 2.