Abstract
Deep infection is a devastating complication of total joint arthroplasty. In a significant proportion of cases it remains a diagnostic challenge. Haematological tests are not specific, particularly in chronic cases, and radiological investigations such as bone scan and radiographs are of only limited value. The most common infective organisms are staphylococcus and some streptococcus species. Acidity is a well established occurrence in infective processes and is caused by the direct production of acid by the organism or by enzymatic degradation of tissues 1,2. In wound infections, peritonitis and some other conditions pH is used as an indicator of infection in clinical practice3. The aim was to assess whether fluid biochemistry (pH, pCO2, pO2, Lactate and Glucose) is altered in infected total knee replacements and whether it could be used as a diagnostic test.
Nineteen consecutive patients undergoing either revision total knee replacement (TKR) or arthroscopic synovial biopsy were included in the study. All had had their primary joint replacement within the previous 3 years. All had a painful total knee replacement and some had evidence of loosening of the prosthesis on radiological investigations. The following investigations were performed on each patient, White cell count (WCC), Erythrocyte Sedimentation Rate (ESR), C-Reactive Protein (CRP), interface synovial biopsy for histology and microbiology and a synovial fluid aspiration from the affected joint prior to application of a tourniquet. A blood gas analyser was used to measure pH, pCO2, pO2, Lactate and Glucose in all synovial fluid specimens.
Seven patients were diagnosed as having an infected TKR on clinical and laboratory investigations. The mean synovial fluid biochemistry results were pH = 7.09, pO2= 5.08kPa, pCO2=10.40kPa, Lactate = 5.33 mmol/l, Glucose = 2.30 mmol/l. In the non-infected group the results were pH = 7.23, pO2 = 7.72kPa, pCO2 = 8.41kPa, Lactate = 4.03 mmol/l, Glucose = 3.42 mmol/l. The differences in pH, pCO2/pO2 ratio, and glucose levels were statistically significant (t-Test p < 0.05) between the two groups. Lactate levels were not significantly different. There was no correlation between high WCC’s and synovial fluid biochemistry or laboratory results for infected cases. Using laboratory results as a gold standard and a synovial fluid pH of less than 7.20, the sensitivity was 85% and specificity 77% for diagnosing an infection. Using a combination of the synovial fluid biochemistry results these values were greater.
Synovial fluid biochemistry is significantly altered in infected total knee replacements. pH levels below 7.2, pCO2/pO2 ratio above 2.5 and Glucose levels below 2.5 mmol/l are strong indicators of an infected TKR. Synovial pH assessment may prove to be a quick, cheap and effective method of diagnosing an infected TKR and may also apply to other joints. Further studies using non-problematic TKR’s as controls are required.
Abstracts prepared by Dr P E Watkins, Hodgkin Building, Guys Campus, King’s College London.
1 Simmon et al. (1994) Infection22(6):386–9. Google Scholar
2 Matsumoto et al. (1998). Google Scholar
3 de Soet et al. (2000) Caries Res.34(6):486–90. Google Scholar
