[Infectious complications of total shoulder arthroplasty]

To evaluate our experience with the therapy of infected total shoulder arthroplasty and to compare the treatment methods used. Although infected total shoulder arthroplasty is not a frequent finding at the present time, the necessity of treating this complication may become more urgent with the continually increasing number of arthroplasty procedures performed.

From 1992 till the beginning of 2005, eleven patients were treated for infected total shoulder arthroplasty. Of them, seven underwent the primary surgery in an outside hospital and four were initially treated at our department. The average age of the patients at the time of infection diagnosis was 61 years. The right shoulder was infected in nine and the left in two patients. An acute infection occurred in one patient (9 %), sub-acute in three (27 %) and late in seven patients (64 %). The average period between the primary operation and infection manifestation was 19.3 months.

The group of 181 patients operated on for shoulder replacement between 1992 and 2005 was evaluated, and a deep infection of total shoulder arthroplasty was found in 11 patients (2.2 %). An antibiotic therapy alone was sufficient to eradicate the infection in only 20 % of the infected patients, but these showed good Constant scores (average, 42 points). Revision surgery, debridement and suction therapy had a low success rate (33 %) and good Constant scores (average, 45 points) in the cured patients. A two-stage reimplantation was 100 % successful but had a poor outcome, with an average Constant score of 26 points. However, a two-stage reimplantation involving a spacer had both a 100 % success rate and a good outcome with an average Constant score of 49 points. On statistical evaluation using the unpaired t-test, there was a significant difference in the Constant scores (T 4.35 p=0.005) between the patients undergoing exchange arthroplasty with (n=40) and without (n=4) the spacer.The cement spacer inserted for the period between the operations was well tolerated by the patients.

There is a consensus that an antibiotic therapy is indicated only in exceptional situations. Similarly, debridement and suction drainage are successful only if the infection is diagnosed early. Poor function scores after resection arthroplasty are not surprising, because a sharp residual proximal humerus is likely to irritate soft tissues and, in addition, it is not possible to reconstruct a rotator cuff to match it. An unexpected finding, however, is the fact that, in contrast to hip joint arthroplasty, resection shoulder arthroplasty shows poor outcomes also in terms of infection eradication. Comparing the results of one-stage with two-stage reimplantation is a complex issue. Attention should be paid to a relationship between the methods routinely used to treat an infected total shoulder arthroplasty and those preferred by the given hospital for treatment of other joints. If the therapy is well established in that hospital and gives good long-term results, it is optimal to use it also for the treatment of infected total shoulder arthroplasty.

Early diagnosis and immediate therapy can prevent more serious damage to soft tissues. The method of treating infected total shoulder arthroplasty is not different from other big joint therapies. The use of a spacer will allow us to remodel soft tissues satisfactorily even after extensive debridement. The functional results of treatment involving a spacer are significantly better. Key words: infection, shoulder, shoulder arthroplasty, one-stage revision, two-stage revision, spacer.