Today's News

Thursday, February 24, 2005

Registry data delivers important lessons

From 1995 to 2002, the NAOR initiative attempted to evaluate hip and knee joint replacement outcomes in the United States and Canada. However, NAOR, which was developed by 10 clinicians, faced several problems,including physician and patient enrollment, follow-up issues and cost.

Between 1995 and 2001, more than 1,000 surgeons enrolled in the registry. But only 489 (40 percent) actually entered patient data. Additionally, surgeon enrollment steadily declined, from 489 surgeons in 1995 to 133 surgeons in 2001.

During the seven-year period, a total of 38,910 patients were registered. Here, too, there was a gradual annual decline in the number of registrations. Approximately 1,500 patients were registered in the hip registry in 1997; by 2001, the number of registrations had declined to 1,200. More than 3,000 knee replacements were registered in 1997; however this dropped to 1,600 by 2001.

Completion of one-year follow-up forms improved from 24 percent in 1998 to 30 percent in 2001. Two-year follow-up completion rates only improved from 6 percent in 1998 to 11 percent in 2001. The cost for maintaining this registry was approximately $750,000 a year.

The NAOR experience demonstrated substantial interest among surgeons to enroll patients, but much less commitment to submit the data. Additionally, the submission of patient data decreased markedly over time, and it was difficult to obtain one- and two-year follow-up data.

The AJRR will need to recognize and address these issues to be successful. Researchers recommend limiting data collection and not depending on surgeon input. Registries should have more defined and limited goals than those of physicians' research databases.

The research team was headed by John J. Callaghan, MD, of Iowa City, Iowa, and included James P. Waddell, MD, of Toronto, Ontario and Frederick Anderson, PhD, of Worcester, Mass.