NEUTROPHILE/LYMPHOCYTE INDEX (IN/L), CREATININE (Cr), AND PROCALCITONIN (PROCAL) AS PREDICTORS OF AMEBIAN LIVER ABSCESS.

Abstract

A liver abscess (HA) is the accumulation of purulent material in the liver parenchyma that can be bacterial, parasitic, fungal, or mixed. The incidence ranges from 2.3 to 22 per 100,000 people. In Mexico, the annual incidence of amoebic HA is 6.7 per 100,000 inhabitants. Determine the cut-off points for the neutrophil/lymphocyte index (IN/L), creatinine (Cr), and procalcitonin (Procal) to predict the etiology of liver abscess. Research design: cross-sectional. We analyzed medical records of patients admitted during 2019 with HA diagnosis and amoeba PCR. The qualitative variables were expressed in frequencies and percentages. The numerical variables in means and standard deviation. We use X2, Fisher's exact, Student's t, and Mann-Whitney U to compare groups as appropriate. ROC curve was used to determine sensitivity (S), specificity (E), positive predictive value (PPV), negative predictive value (NPV), and likelihood value (+ LR). The p-value <0.05 was considered statistically significant. Out of a total of 32 patients diagnosed with HA during 2019, 20 patients treated with drainage and a PCR test for amoeba from the abscess fluid were included. Of these, 85%(17) were men, with a mean age of 45.33±10.93 years. 45%(9) were of amoebic etiology. In the latter group, the etiology can be predicted with the neutrophil/lymphocyte index with a cohort point of ≥10.34 with an AUC of 0.838, S: 100%, E: 81%, PPV: 81%, NPV: 100%. (9/11 vs 0/0 [81.8% vs 0.0%] + LR: 5.49; 95%CI:1.50-14 p =0.000). The creatinine value of ≥1.02 with an AUC of 0.818, S: 66.7%, E: 90.9%, PPV: 85.7%, NPV: 76.9%, (6/7 vs 3/13 [85.7% vs 23.1%] + LR: 7.33;95% CI:1.07-50 p =0.017) and with a procalcitonin cohort point of ≥11.1 with an AUC: 0.808, S: 77.8%, E: 90.9%, PPV: 85.7%, NPV: 87.5%, (7/8 vs 2/12 [87.5% vs 16.7%] + LR: 8.56;95% CI:1.28-57 p =0.005), with these cut-off points a significant difference was evidenced between the amoebic vs bacterial etiology, for IN/L: p =0.000, for Cr: p =0.017 and for procalcitonin: p =0.005, which are shown in figure 1. Amebic HA is etiologically more frequent in the West and generally in countries with poor infrastructure and development. It reports high mortality with conservative treatment and multiple abscesses, so it is crucial to identify their etiology. In the present study, we propose the cut-off points of biochemical markers for the diagnosis of amoebic HA through IN/L, Cr, and procal that are accessible in units where there is no amoeba CRP. We were able to determine an adequate AUC and good sensitivity, specificity, positive and negative predictive value; therefore, we could use these biochemical markers to predict the etiology of liver abscesses. The authors declare that there is no conflict of interest.