INFLA-score: A new diagnostic paradigm to identify pericarditis

BACKGROUND

Diagnosis of pericarditis may be challenging because not all patients meet the conventional criteria. An overlooked diagnosis implies a longer course of symptoms and an increased risk of recurrences. C-reactive protein (CRP), widely used as an inflammation marker, has some limitations. This study aimed to assess the usefulness and prognostic value of INFLA-score, a validated index assessing low-grade inflammation, in the definite diagnosis of pericarditis.

METHODS

Patients with suspected pericarditis were included. The INFLA-score was computed based on white blood cells and platelet count, neutrophil-to-lymphocyte ratio, and CRP, ranging from -16 to +16. An INFLA-score > 0 was considered positive for the presence of pericardial inflammation. The primary end point was the association of INFLA-score with diagnosis of pericarditis according to conventional criteria. The recurrence of pericarditis at 6 months was the secondary end point.

RESULTS

A total of 202 patients were included, aged 47 ± 17 years, and 57% were females. Among 72 (36%) patients with a diagnosis of pericarditis, an INFLA-score > 0 was observed in 86% (vs. 36%, p < 0.001), abnormal CRP in 42% (vs. 10%, p < 0.001), pericardial effusion in 44% (vs. 19%, p < 0.001), abnormal electrocardiogram in 56% (vs. 24%, p < 0.001), and rubs in 5% (vs. 0.1%, p = 0.072). INFLA-score > 0 had the strongest predictive value for the diagnosis of pericarditis (hazard ratio 8.48, 95% confidence interval [CI] 3.39-21.21), with 86% sensitivity and 64% specificity, as opposed to CRP (hazard ratio 1.72, non-significant 95% CI 0.69-4.29). Recurrent pericarditis at 6 months was more frequent in patients with a positive INFLA-score (37% vs. 8%, p < 0.001, rate ratio 4.15, 95% CI 2.81-6.12). In patients with normal CRP, INFLA-score-confirmed ongoing inflammation in 78% of the cases. Compared with the conventional criteria, the INFLA-score had the highest accuracy (area under the curve = 0.82). Different cutoffs were valuable to rule out (INFLA-score > 0, sensitivity 86%, and negative likelihood ratio 0.22) or rule in (INFLA-score ≥ 10, specificity 97%, and positive likelihood ratio 13) the diagnosis.

CONCLUSIONS

The INFLA-score is a useful diagnostic tool to assess the probability of pericarditis, with a strong prognostic value for further recurrences, outperforming CRP.