The value of maintaining outcomes in an individual practice for 25 years.

D. Collis
DOI: https://doi.org/10.1097/00003086-199711000-00008
IF: 4.755
1997-11-01
Clinical Orthopaedics and Related Research
Abstract:Decision making in orthopaedic surgery is facilitated by outcome data, not only from the literature, but also from one's own practice. Between 1971 and 1996, 2602 primary and 587 revision total hip replacements were performed. All had prospective collection of data, which altered surgical practice. A posterior approach in 147 primary total hip replacements resulted in 10 dislocations (6.8%) leading to the use of the anterolateral approach in 1788 subsequent hips with only eight dislocations (0.45%). The initial 298 hips were done in a standard operating room with restricted access and were covered with perioperative antibiotics. No infections occurred permitting ongoing use of inexpensive infection prevention. From 1971 to 1975, 168 classic polished Charnley stems were implanted with only seven stems requiring revision to date. In contrast, use of the matte finished Müller prosthesis in the early 1970s in 58 hips has resulted in nine revised stems. From 1980 to 1990, 789 second generation design matte finished Iowa stems were implanted. Prospective data collected disclosed that a small number (1.5%) of early loosenings occurred and all were associated with extensive bone lysis which made revision difficult. Only polished stems are now cemented by the author. The author firmly believes that his practice patterns have been promptly and appropriately altered by the availability of a prospectively collected hip data file.
What problem does this paper attempt to address?