Can the presence of an infection be predicted before a revision total hip arthroplasty? Preliminary study to establish an infection score

详细信息    查看全文

• 作者：J.-Y. Jenny^a ; ^{jean-yves.jenny@chru-strasbourg.fr" class="auth_mail" title="E-mail the corresponding author} ; B. Adamczewski^a ; E. De Thomasson^b ; J. Godet^a ; H. Bonfait^c ; C. Delaunay^c ; French Hip ; Knee Society (SFHG)
• 关键词：Infection ; Revision ; Total hip arthroplasty ; Preoperative diagnosis
• 刊名：Orthopaedics & Traumatology: Surgery & Research
• 出版年：2016
• 出版时间：April 2016
• 年：2016
• 卷：102
• 期：2
• 页码：161-165
• 全文大小：356 K

文摘

The diagnosis of periprosthetic joint infection can be challenging, in part because there is no universal diagnostic test. Current recommendations include several diagnostic criteria, and are mainly based on the results of deep microbiological samples; however, these only provide a diagnosis after surgery. A predictive infection score would improve the management of revision arthroplasty cases. The purpose of this study was to define a composite infection score using standard clinical, radiological and laboratory data that can be used to predict whether an infection is present before a total hip arthroplasty (THA) revision procedure.

Hypothesis

The infection score will make it possible to differentiate correctly between infected and non-infected patients in 75% of cases.

Material and methods

One hundred and four records from patients who underwent THA revision for any reason were analysed retrospectively: 43 with infection and 61 without infection. There were 54 men and 50 women with an average age of 70 ± 12 years (range 30–90). A univariate analysis was performed to look for individual discriminating factors between the data in the medical records of infected and non-infected patients. A multivariate analysis subsequently integrated these factors together. A composite score was defined and its diagnostic effectiveness was evaluated as the percentage of correctly classified records, along with its sensitivity and specificity.

Results

The score consisted of the following individually weighed factors: body mass index, presence of diabetes, mechanical complication, wound healing disturbance and fever. This composite infection score was able to distinguish correctly between the infected patients (positive score) and non-infected patients (negative score) in 78% of cases; the sensitivity was 57% and the specificity 93%.

Discussion

Once this score is evaluated prospectively, it could be an important tool for defining the medical – surgical strategy during THA revision, no matter the reason for revision.

Level of evidence

Level IV – retrospective study.