Diagnostic utility of pericardial fluid pH in diagnosing infectious pericardial effusions among patients with moderate and large effusions undergoing pericardiocentesis at Groote Schuur Hospital: a subs-study of the IMPI trial

Master Thesis


Permanent link to this Item
Journal Title
Link to Journal
Journal ISSN
Volume Title
Diagnosis of infectious pericardial disease has been challenging in the developing world despite improvement of treatment modalities. The diagnostic utility of pH in diagnosing infectious pericardial fluid is unknown, yet this concept is well studied in pleural fluid. This cross-sectional diagnostic study evaluated the diagnostic utility of pH in infectious compared to non-infectious pericardial effusions in a high-burden setting. Methods: Patients of 18 years with moderate to large effusion between the 1st February 2016 and 31st May2018 were enrolled at Groote Schuur Hospital in Cape Town, South Africa. After safe pericardiocentesis, pH was measured with a blood gas analyzer. Mycobacterium tuberculosis culture and/or gene Xpert for TB and/or bacteria culture and/or microscopy served as the reference standard for definite infectious pericardial effusions. We calculated sensitivity, specificity, positive and negative predictive values, negative and positive likelihood ratios for an a priori pH cut off of 7.35. Receiver operating characteristic curve analysis was used for selection of ideal pH cut off. RESULTS Using a set sensitivity of 70% we estimated that we needed to recruit a sample size of 149 subjects for a 95% confidence interval and power of 80%. We screened 200 patients, and excluded 60 because they did not meet the appropriate exclusion criteria. The prevalence of infectious pericarditis was 27.1% (n/N=34/140) as confirmed by the reference standard. We found the median pH (IQR) was 7.30(7.20-7.30) for definite infection, 7.30(7.30-7.35) for probable infection and 7.50(7.40-7.55) for non-infectious effusions p value <0.01 (test for trend). At a cut off or <7.35, the sensitivity was 89.5(95%CI: 75%.5-97.1%) and the specificity was 72.5% (95% CI: 62.8%-80.9%). The ideal ROC- determined cut off for pH that would give maximum sensitivity and specificity was ≤7.30 and the maximum sensitivity and specificity at optimum cut off are 86.8% (95% CI:71.9 - 95.6) and 86.8% (95% CI:71.9 - 95.6), respectively. The area under the curve at this cut-off point is 0.86 (95% CI 0.79 to 0.9), p<0. 001. CONCLUSION: In conclusion, pericardial PH offers diagnostic utility for infectious causes of pericardial effusions using both a PH of 7.35 and an ideal cut-off of 7.30. We recommend that given the simplicity of the test it should be adopted in evaluation of patients with pericardial effusions.