September 24, 2004

Dear Professor Moran:

Thank you so much for your recent letter regarding our article from the September issue of the journal on 30-day Mortality Following Hip Arthroplasty For Acute Fracture. Beginning in 1969 with the first total hip arthroplasty performed at our institution Dr. Mark Coventry then chair of the Department of Orthopedic Surgery here at Mayo initiated a Total Joint Registry. Since that time we have gathered in prospective fashion information on all patients undergoing hip arthroplasty and subsequently other total joint arthroplasties at the time of the index procedure and at regular follow-up intervals subsequent to the operation. Details of this more than three decade endeavor was recently documented in an article by Berry et al. (Berry, D.J.; Kessler, M.; Morrey, B.F.: Maintaining a Hip Registry for 25 Years. Mayo Clinic Experience. Clin Orthop Rel Res 344:61-68, Nov. 1997) As a result of these early and continuing efforts, each and every patient is followed at 6 to 12 weeks following surgery, again at one year, at 2 years and then at 5, 10 and 15 years post surgery with follow-up continuing at five year increments thereafter. We now have patients with greater than 30 years follow-up of procedures performed in the early 1970’s. Each follow-up interval represents an opportunity to track the patient and determine that death has occurred if in fact they have died since an earlier contact or appointment. When patients found to be lost to follow-up, extensive efforts are undertaken by our regular staff who are employed full time in this endeavor to find them. If there is no contact at their last known residence, then telephone contact is made with country courthouses, and other public sources of information regarding residences, and deaths in the area of the patient’s last known address. Because of this effort we feel confident of the data regarding mortality on our arthroplasty patients during the first 30 days post surgery and indeed over the ensuing years post procedure.

The ability to generate reliable follow-up information from a large single institutional data base, has helped serve as the basis for earlier publications from our group on perioperative mortality in both hip and knee arthroplasty populations compared to the demographic characteristics in other areas. (Parvizi J, Johnson BG, Rowland C, Ereth MH, Lewallen DG.: Thirty-day mortality after elective total hip arthroplasty. J Bone Joint Surg Am. 2001 Oct;83-A(10):1524-8; Parvizi J, Sullivan TA, Trousdale RT, Lewallen DG. : Thirty-day mortality after total knee arthroplasty. J Bone Joint Surg Am. 2001 Aug;83-A(8):1157-61. )

In interpreting the data from our recent study of acute fracture patients it is important to note that despite a significant referral practice, a large portion of our patients, and especially those with fractures, are drawn from the nearby upper mid west region. This area of the United States has a low incidence of indigent patients and an under- representation of minority populations when compared to the demographics other areas of the United States and large urban centers both here and abroad. These factors may result in a difference in observed mortality over the first 30 days post hip arthroplasty for acute fracture when compared to other institutions serving a very different patient demographic.

The authors wish to thank Professor Moran for his interest and questions as these have helped to bring out several important points of potential interest to other readers.

Sincerely,

David G. Lewallen, M.D.

Dear Sir

Thirty-day mortality following hip arthroplasty for acute fracture.

I read with great interest this article by Drs Parvizi, Ereth and Lewallen (JBJS 86-A; 1983-1987: September 2004). I should be grateful for some more detail on the calculation of thirty-day mortality.

Thirty-day mortality should only be calculated when the outcome (i.e. dead or alive) of the entire cohort or population is known thirty-days after the index event. Complete follow-up of the entire cohort is required for this calculation. A thirty-day mortality of 2.4% is remarkably low for the elderly, hip fracture population, particularly as the background mortality (death by natural causes) is 0.8% per month. Are the authors calculating in-hospital mortality, rather than true thirty-day mortality? This is an important issue if mortality rates are to be compared between institutions as the low mortality rate quoted could be due to a relatively short hospital stay with early rehabilitation in the community. Reductions in length of stay may also explain the decrease in mortality over the past three decades.

Yours sincerely

Prof. Christopher G Moran MD FRCS

Consultant Trauma & Orthopaedic Surgeon