Complete blood count derived inflammation indexes predict outcome in COVID-19 patients: a study in Indonesia.

INTRODUCTION: Inflammation plays a vital role in the pathophysiology of COVID-19. Complete blood count (CBC) is a routine test performed on patients. It provides information regarding the inflammatory process and can be used as a predictor of outcome. This study aimed to explore the correlation between different complete blood count (CBC)-derived inflammation indexes at hospital admission, such as neutrophil to lymphocyte ratio (NLR), derived NLR (dNLR), platelet to lymphocyte ratio (PLR), monocyte to lymphocyte ratio (MLR), neutrophil to lymphocyte × platelet ratio (NLPR), aggregate index of systemic inflammation (AISI), systemic inflammation response index (SIRI), and systemic immune-inflammation index (SII), to in-hospital mortality in confirmed COVID-19 patients. METHODOLOGY: A retrospective observational study was performed at Ulin Referral Hospital of South Kalimantan with 445 COVID-19 patients from April to November 2020. The patients were divided into two groups, non-survivor and survivor. A receiver operating characteristic (ROC) curve was used to determine the cut-off values. Bivariate analysis was performed using the Chi Square test, the risk ratio was calculated, and logistics regression was determined. RESULTS: Increase of NLR, dNLR, PLR, MLR, NLPR, MLR, AISI, SIRI, and SII from cut-off values were significantly correlated with patient survival outcome. The cut off values were 6.90, 4.10, 295, 0.42, 0.037, 1,422, 1.80, and 2,504 respectively. NLPR was dominant in predicting in-hospital mortality (OR: 6.668, p = 0.000) with a 28.1% sensitivity and 95.9% specificity. CONCLUSIONS: CBC-derived inflammation indexes were associated with the survival outcome of confirmed COVID-19 patients and NLPR was a dominant variable.

Klasifikasi persepsi pengguna Twitter terhadap kasus COVID-19 menggunakan metode logistic regression

Dunia dilanda penyakit Coronavirus (COVID-19) yang menyerang sistem pernafasan pada manusia. Virus tersebut berasal dari Wuhan China. Dan saat ini sudah ditetapkan sebagai pandemi karena sudah menyebar hampir di seluruh Negara. Hal tersebut memicu gagasan dan opini masyarakat Amerika Serikat di media sosial Twitter. Cuitan tersebut dimanfaatkan untuk mengetahui emosi seseorang dengan mengelompokan dalam 5 label diantaranya extreme positive positive neutral negative dan extreme negative. Pada hal ini penulis mengelompokan label menjadi 3 label kelas diantaranya positive neutral dan negative. Penulis menguji menggunakan metode Logistic Regression dengan memberi variasi hyperparameter L2 dan None. Pada hyperparameter L2 diperoleh nilai akurasi 77% dan F1 score sebesar 74%. Dan pada variasi hyperparameter None diperoleh nilai akurasi 74% dan F1 Score 70%. Dalam demikian pada nilai hyperparameter L2 merupakan variasi terbaik pada metode Logistic Regression.