Karakteristik Pasien Demam Berdarah Dengue pada Instalasi Rawat Inap RSUD Kota Prabumulih Periode Januari–Mei 2016

Abstract Dengue hemorrhagic fever is an acute epidemic disease that requires a good and complete clinical examination accompanied by an accurate laboratory examination if clinical symptoms are inadequate. The delays in diagnosis results in an increased risk of death. Hospital facilities and health workers are very influential in the recovery of patients with dengue fever. . This scientific paper presents the characteristics of patients (gender, age, temperature, leukocytes, hematocrit, platelets, hemoglobin) in patient of dengue fever in Prabumulih City Hospital. Data analysis was performed on medical record data of patients with dengue fever who were hospitalized in January-May 2016 at Prabumulih City Hospital. This type of research is descriptive analytic with retrospective design. The data discussed is the result of observing the patient’s clinical condition from the first day to the eighth day. The majority of DHF patients are female with the most age groups at 0-4 years. The body temperature of the highest DHF patients on day 1 was 39.80 C and on the eighth day showed a normal temperature of 36.50 C. The lowest hematological value on day 1 was leukocytes of 1,600 cells/mm3 , hematocrit was 27.9%, platelets were 8,000 cells / mm3 , hemoglobin 9.4 gram / dL. The lowest hematological value on the last six days of treatment is 5,600 cell / mm3 leukocytes, 27.9% hematocrit, 74,000 cell / mm3 platelets, 9.7 gram/dL hemoglobin. Hematological values for normal leukocytes and platelets while platelets and hemoglobin did not approach normal values. Abstrak Demam berdarah dengue (DBD) adalah penyakit epidemik akut yang memerlukan pemeriksaan klinis yang baik dan lengkap disertai pemeriksaan laboratorium yang akurat jika gejala klinis tidak memadai. Keterlambatan dalam diagnosis mengakibatkan peningkatan risiko kematian. Fasilitas rumah sakit dan tenaga kesehatan sangat berpengaruh dalam kesembuhan pasien demam berdarah. Naskah ilmiah ini menyajikan karakteristik pasien (jenis kelamin, umur, suhu, leukosit, hematokrit, trombosit, dan hemoglobin) rawat inap demam berdarah di RSUD Kota Prabumulih. Analisa data dilakukan terhadap data rekam medis pasien demam berdarah yang rawat inap bulan Januari-Mei 2016 di RSUD Kota Prabumulih. Jenis penelitian ini adalah deskriptif analitik dengan desain retrospektif. Data yang dibahas adalah hasil pengamatan keadaan klinis pasien mulai pada hari pertama hingga hari ke delapan. Mayoritas pasien DBD berjenis kelamin perempuan dengan kelompok umur terbanyak pada 0-4 tahun. Suhu tubuh pasien DBD tertinggi pada hari ke 1 adalah 39,80 C dan pada hari ke delapan menunjukkan suhu normal yaitu 36,50 C. Nilai hematologi terendah pada hari ke 1 yaitu leukosit sebesar 1.600 sel/mm3 , hematokrit 27,9%, trombosit 8.000 sel/mm3 , hemoglobin 9,4 gram/dL. Nilai hematologi terendah pada hari ke enam terakhir perawatan yaitu leukosit 5.600 sel/mm3 , hematokrit 27,9%, trombosit 74.000 sel/mm3 , hemoglobin 9,7 gram/dL. Nilai hematologi untuk leukosit dan trombosit normal sedangkan trombosit dan hemoglobin tidak mendekati nilai normal.

Download Full-text

Adolescent Tri-ponderal Mass Index Growth Trajectories and Incident Diabetes Mellitus in Early Adulthood

The Journal of Clinical Endocrinology & Metabolism ◽

10.1210/clinem/dgab235 ◽

2021 ◽

Author(s):

Yi-Fan Wu ◽

Hsien-Yu Fan ◽

Yang-Ching Chen ◽

Kuan-Liang Kuo ◽

Kuo-Liong Chien

Keyword(s):

Early Adulthood ◽

Growth Mixture Modeling ◽

The Body ◽

Growth Trajectory ◽

Incident Diabetes ◽

Growth Trajectories ◽

City Hospital ◽

Z Score ◽

Increased Risk ◽

Adult Diabetes

Abstract Purpose Studies have reported the influence of adolescent obesity on development of adult diabetes, but the effect of the growth pattern during this period has rarely been explored. Also, the tri-ponderal mass index (TMI) was thought to be a better estimation of adolescent body fat levels than the body mass index (BMI), so we sought to investigate whether growth trajectories derived by these two indices could predict incident diabetes. Methods We conducted a study by using the Taipei City Hospital Radiation Building Database, a longitudinal cohort established from 1996 until now. Physical exam results including blood test results were collected annually and the BMI z-score/TMI growth trajectory groups during 13–18 years of age were identified using growth mixture modeling. A Cox proportional hazard model for incident diabetes was used to examine the risk of baseline obese status and different BMI/TMI growth trajectories. Results Five growth trajectory groups were identified for the BMI z-score and the TMI. During approximately 20,400 person-years follow-up, 33 of 1,387 participants developed diabetes. Baseline obesity defined by the BMI z-score and the TMI were both related to adult diabetes. The persistent increase TMI growth trajectory exhibited a significantly increased risk of diabetes after adjusting for baseline obese status and other correlated covariates (hazard ratio: 2.85, 95% confidence interval (CI): 1.01–8.09). There was no association between BMI growth trajectory groups and incident diabetes. Conclusions A specific TMI growth trajectory pattern during adolescence might be critical for diabetes prevention efforts.

Download Full-text

The lineage of coronavirus SARS-CoV-2 of Russian origin: Genetic characteristics and correlations with clinical parameters and severity of coronavirus infection

The Siberian Journal of Clinical and Experimental Medicine ◽

10.29001/2073-8552-2021-36-4-132-143 ◽

2022 ◽

Vol 36 (4) ◽

pp. 132-143

Author(s):

O. S. Glotov ◽

A. N. Chernov ◽

A. I. Korobeynikov ◽

R. S. Kalinin ◽

V. V. Tsai ◽

...

Keyword(s):

Blood Analysis ◽

Viral Rna ◽

Early Warning Score ◽

Cdna Libraries ◽

City Hospital ◽

Genetic Characteristics ◽

S Protein ◽

Risk Of Death ◽

Increased Risk ◽

D Dimer

The identification of new SARS-CoV-2 and human protein and gene targets, which may be markers of the severity and outcome of the disease, are extremely important during the COVID-19 pandemic. The goal of this study was to carry out genetic analysis of SARS-CoV-2 RNA samples to elucidate correlations of genetic parameters (SNPs) with clinical data and severity of COVID-19 infection.Material and Methods. The study included viral RNA samples isolated from 56 patients with COVID-19 infection who received treatment at the City Hospital No. 40 of St. Petersburg from 04/18/2020 to 04/18/2021. Patients underwent physical examination with the assessments of hemodynamic and respiratory parameters, clinical risk according to National Early Warning Score (NEWS), computed tomography (CT) of the chest, and laboratory studies including clinical blood analysis, assessment of ferritin, C-reactive protein (CRP), interleukin-6 (IL-6), lactate dehydrogenase (LDH), D-dimer, creatinine, and glucose levels. All patients tested positive for SARS-CoV-2 RNA by polymerase chain reaction (PCR). Single nucleotide polymorphisms (SNPs) in viral RNA were identified through the creation of cDNA libraries by targeted sequencing (MiSeq Illumina). Bioinformatic analysis of viral samples was performed using the viralrecon v2 pipeline with the further annotation via Pangolin and Nextlade. Sampled genomes were visualized using the Integrative Genomics Viewer (IGV) software. Statistical data processing (descriptive statistics and graphical analysis of data relationships from diff erent tables) was performed using a GraphPad device on the Prism 8.01 platform.Results. A comparative analysis of SNP frequencies in the virus genome in samples from deceased and discharged patients was carried out. The SNPs associated with risk of death (OR > 1), neutral SNPs (OR = 1), and protective SNPs (OR < 1) were identifi ed. Patient samples were infected with 14 lines of SARS-CoV-2, fi ve of which (B.1.1.129, B.1.1.407, B.1.1.373, B.1.1.397, and B.1.1.152) were of Russian origin. The SNPs in the samples infected with the strains of non-Russian origin were associated with an increased risk of mortality (OR = 2.267, 95% confi dence interval 0.1594-8.653) compared to the SNPs in the samples obtained from the group of patients infected with the strains of Russian origin. Positive correlations were identifi ed between the average SNP number, nonsynonymous SNPs, and S-protein SNPs with the degree of respiratory failure, total NEWS score, CT-based form of disease, duration of treatment with mechanical ventilation, disease outcome, levels of LDH, glucose, D-dimer, and ferritin, and RNA amount in the PCR test. S-protein SNPs negatively correlated with the leukocyte and neutrophil counts.

Download Full-text

Benchmarking of Machine Learning Models to Assist the Prognosis of Tuberculosis

10.20944/preprints202103.0284.v2 ◽

2021 ◽

Author(s):

Maicon Herverton Lino Ferreira da Silva Barros ◽

Geovanne Oliveira Alves ◽

Lubnnia Morais Florêncio Souza ◽

Élisson da Silva Rocha ◽

João Fausto Lorenzato de Oliveira ◽

...

Keyword(s):

Machine Learning ◽

Clinical Symptoms ◽

Treatment Decision ◽

Gradient Boosting ◽

Original Form ◽

Learning Models ◽

Data Set ◽

Risk Of Death ◽

Increased Risk ◽

Machine Learning Models

Tuberculosis (TB) is an airborne infectious disease caused by organisms in the Mycobacterium tuberculosis (Mtb) complex. In many low and middle-income countries, TB remains a major cause of morbidity and mortality. Once a patient has been diagnosed with TB, it is critical that healthcare workers make the most appropriate treatment decision given the individual conditions of the patient and the likely course of the disease based on medical experience. Depending on the prognosis, delayed or inappropriate treatment can result in unsatisfactory results including the exacerbation of clinical symptoms, poor quality of life, and increased risk of death. This work benchmarks machine learning models to aid TB prognosis using a Brazilian health database of confirmed cases and deaths related to TB in the State of Amazonas. The goal is to predict the probability of death by TB thus aiding the prognosis of TB and associated treatment decision making process. In its original form, the data set comprised 36,228 records and 130 fields but suffered from missing, incomplete, or incorrect data. Following data cleaning and preprocessing, a revised data set was generated comprising 24,015 records and 38 fields, including 22,876 reported cured TB patients and 1,139 deaths by TB. To explore how the data imbalance impacts model performance, two controlled experiments were designed using (1) imbalanced and (2) balanced data sets. The best result is achieved by the Gradient Boosting (GB) model using the balanced data set to predict TB-mortality, and the ensemble model composed by the Random Forest (RF), GB and Multi-layer Perceptron (MLP) models is the best model to predict the cure class.

Download Full-text

P1828Aortic calcification detected by computed tomography and aortic vulnerable plaques: aortic angioscopy study

European Heart Journal ◽

10.1093/eurheartj/ehz748.0580 ◽

2019 ◽

Vol 40 (Supplement_1) ◽

Author(s):

K Kojima ◽

Y Ebuchi ◽

S Migita ◽

T Morikawa ◽

T Mineki ◽

...

Keyword(s):

Computed Tomography ◽

Cardiovascular Disease ◽

Suspected Coronary Artery Disease ◽

The Body ◽

Aortic Calcification ◽

Plaque Instability ◽

Risk Of Death ◽

Vulnerable Plaques ◽

Increased Risk ◽

Atheromatous Plaques

Abstract Background Aortic calcification is associated with atherosclerotic risk factors and an increased risk of death and cardiovascular disease. However, the relationships aortic calcification and aortic plaque instability are not yet elucidated. Recently, some reports showed non-obstructive aortic angioscopy seemed to visualize atherosclerotic changes of aortic wall more clearly compared with computed tomography (CT). The purpose of this study was to evaluate whether aortic calcification is associated with aortic vulnerable plaques in patients with cardiovascular disease. Methods We investigated 60 consecutive patients with confirmed or suspected coronary artery disease who underwent both aortic angioscopy and CT. The AC volume (ACV) was measured using the volume-rendering method by extracting the area >130 HU within the whole aorta. ACV index (ACVI) was defined as ACV divided by the body surface area. We evaluated the number of ruptured plaque (RP), ulceration and fissure by aortic angioscopy in the whole aorta. We excluded 4 hemodialysis patients. All patients were divided into the median value of ACVI. Results The mean age of patients was 68±10. The median of ACVI was 10.7 ml/m2 [3.9–22.7]. High ACVI patients had significantly greater number of RP, ulceration and atheromatous plaques detected by aortic angioscopy compared with those of low ACVI (2.2±2.7 vs 0.8±1.1, p=0.033, 1.6±1.2 vs 0.9±1.0, p=0.041, 4.0±3.1 vs 1.9±1.8, p=0.009, respectively). Furthermore, the patients without aortic calcification did not have RP at all. In a multivariate model, the number of the atheromatous plaques was independently associated with high ACVI (odds ratio 1.57, 95% confidence interval 1.07–2.69, p=0.018) Conclusions Aortic calcification detected by CT was related to aortic vulnerable plaques in patients with cardiovascular disease.

Download Full-text

Role of Vitamin C (Ascorbic Acid) on Human Health- A Review

African Journal of Food, Agriculture, Nutrition and Development ◽

10.18697/ajfand.8.1155 ◽

2005 ◽

Vol 5 (8) ◽

pp. 01-12

Author(s):

Mary K Walingo ◽

Keyword(s):

Blood Pressure ◽

Ascorbic Acid ◽

Vitamin C ◽

Human Health ◽

The Body ◽

Water Soluble ◽

Plasma Vitamin ◽

Risk Of Death ◽

Increased Risk

Vitamin C, also known as ascorbic acid, abounds in nature and is highly labile. It is a water-soluble vitamin that is lost in large amounts during food processing. It is a vitamin whose prescribed requirement across cultures is not uniform. For example , the prescribed requirement of vitamin C in Great Britain is 30mg/day, while in the U.S.A., it is 60mg/day and 100mg/day in Japan. These variations are unusual and point to the need for further research to establish the acceptable RDAs for diverse populations. The RDA for vitamin C should be more than the amount needed to prevent the occurrence of disease. Vitamin C plays significant functions in the body that enhance its role in the health status of the human body. The biochemical functions of vitamin C include: stimulation of certain enzymes, collagen biosynthesis, hormonal activation, antioxidant, detoxification of histamine, phagocytic functions of leukocytes, formation of nitrosamine, and proline hydroxylation amongst others. These functions are related to the health effects of vitamin C status in an individual. In human health, vitamin C has been associated with reduction of incidence of cancer, blood pressure, immunity, and drug metabolism and urinary hydroxyproline excretion, tissue regeneration. This vitamin is needed for the proper metabolism of drugs in the body through adequate hepatic mixed function oxidase system. Epidemiological data have revealed the preventive and curative role of vitamin C on certain disease conditions in the body though controversies still persist. Vitamin C is effective in protecting against oxidative damage in tissues and also suppresses formation of carcinogens like nitrosamines. There is an inverse relationship with blood pressure and both plasma vitamin C and Vitamin C. Vitamin C has a lowering effect on blood pressure, especially on systolic pressure more than a diastolic pressure. Low levels of plasma vitamin C are associated with stroke and with an increased risk of all cause mortality. Increased consumption of ascorbic acid raises serum ascorbic levels and could decrease the risk of death.

Download Full-text

PREMATURE OVARIAN FAILURE, ITS CONSEQUENCES, MORTALITY AND IMPACT ON THE CARDIOVASCULAR SYSTEM

Nauka molodykh (Eruditio Juvenium) ◽

10.23888/hmj202191147-156 ◽

2021 ◽

Vol 9 (1) ◽

pp. 147-156

Author(s):

S.V. Lopukhov ◽

◽

E.V. Filippov ◽

Keyword(s):

Cardiovascular Diseases ◽

Premature Ovarian Failure ◽

Correct Diagnosis ◽

Vascular Diseases ◽

Ovarian Failure ◽

Pharmacological Therapy ◽

The Body ◽

Risk Of Death ◽

Increased Risk

This review focuses on the topic of premature ovarian failure (POF) as highly relevant in modern medicine (up to 2% of women in the population suffer from this disease). However, patients with premature ovarian failure not only are still not receiving any treatment, but even making this diagnosis is very difficult. Even after a correct diagnosis is made, these patients are not followed up, despite the fact they have already developed a hormonal imbalance. These women develop two groups of complications: short-term complications associated with a rapid estrogen deficiency in the body, and much more dangerous long-term complications affecting multiple organs and even systems. But in the meanwhile, women with premature ovarian failure are under increased risk of death from all causes, in particular from coronary heart disease (CHD), respiratory diseases, genitourinary diseases and from external causes. And this is despite the fact that cardio-vascular diseases (CVD) are already the leading cause of death among women worldwide. It is women with POF that are at the highest risk of development of cardiovascular diseases, compared to women with normal menopause. These patients, therefore, constitute one of the most important groups to be targeted by screening and prevention strategies primarily for cardiovascular diseases. These strategies should include the use of risk stratification tools to identify women that need lifestyle modifying and pharmacological therapy to prevent development of such diseases in them. This is the only way to maintain a high quality of life in these women over the long term.

Download Full-text

A Cohort Study Assessing the Association Between Body Composition Parameters and Symptomatic Hypocalcemia After Total Thyroidectomy

The American Surgeon ◽

10.1177/0003134820979578 ◽

2020 ◽

pp. 000313482097957

Author(s):

Suleyman U. Celik ◽

Can Konca ◽

Volkan Genc

Keyword(s):

Body Composition ◽

Total Thyroidectomy ◽

Clinical Symptoms ◽

The Body ◽

Postoperative Hypocalcemia ◽

Parathyroid Autotransplantation ◽

Asymptomatic Patients ◽

Increased Risk ◽

Symptomatic Hypocalcemia ◽

The Relationship

Background Postoperative hypocalcemia is one of the major concerns following thyroidectomy and the most frequent cause of prolonged hospital stay. The aim of this study was to evaluate the relationship between body composition parameters and symptomatic hypocalcemia following total thyroidectomy. In addition, the effects of disease- and patient-related factors on hypocalcemia were investigated. Methods A total of 144 patients were prospectively included between March 2014 and September 2017. Patients were divided into 2 groups according to the presence or absence of clinical symptoms of hypocalcemia. Subsequently, the relationship between body composition parameters and hypocalcemia was evaluated. Results Postoperative hypocalcemia-related symptoms occurred in 28 patients (19.4%). Permanent hypocalcemia was not encountered in any patient. Patients with hypocalcemic symptoms were more likely to have nodules ≥40 mm (39.3% vs. 17.2%, P = .011), retrosternal goiters (25.0% vs. 7.8%, P = .017), central lymph node dissection (LND) (32.1% vs. 11.2%, P = .015), and parathyroid autotransplantation (28.6% vs. 3.4%, P < .001) than those without symptoms. However, no differences were observed in the body composition parameters between symptomatic and asymptomatic patients. On multivariate analysis, lower preoperative intact parathyroid hormone (iPTH) levels (odds ratios (ORs) .96, 95% confidence intervals (CIs) .93-.99), the presence of retrosternal goiters (OR 10.26, 95% CI 2.23-47.14), central LND (OR 16.05, 95% CI 3.90-66.07), and parathyroid autotransplantation (OR 36.22, 95% CI 6.76-194.13) predicted hypocalcemia. Discussion This study demonstrates that patients with lower preoperative iPTH levels, retrosternal goiters, central LND, and parathyroid autotransplantation are at an increased risk of developing clinical symptoms of hypocalcemia. Body composition parameters have no effect on the incidence of hypocalcemia after total thyroidectomy.

Download Full-text

Rheumatoid arthritis-associated interstitial lung disease – an underestimated, complex and interdisciplinary problem in everyday clinical practice

Medycyna Faktów ◽

10.24292/01.mf.0420.7 ◽

2020 ◽

Vol 13 (4) ◽

pp. 431-441

Author(s):

Sebastian Majewski ◽

Joanna Makowska

Keyword(s):

Rheumatoid Arthritis ◽

Interstitial Lung Disease ◽

Lung Disease ◽

Clinical Symptoms ◽

Future Research ◽

Clinical Aspects ◽

Scientific Review ◽

Disease Modifying Drugs ◽

Risk Of Death ◽

Increased Risk

Rheumatoid arthritis (RA) is a chronic, systemic inflammatory connective tissue disease affecting 0.5–1% of the general population. Interstitial lung disease (ILD) is a serious complication of RA, leads to a deterioration in the quality of life, increases the risk of hospitalization and premature death. The clinical course of RA-associated ILD (RA-ILD) varies from interstitial lesions involving little areas of the lung which do not lead to clinical symptoms, to progressive interstitial lesions that may lead to acute respiratory exacerbation, development of respiratory failure, and death. The main risk factor for the development of ILD in the course of RA, apart from older age and male sex, is the high activity of the underlying disease. Effective treatment of RA, aimed at the goal of remission, may therefore be a preventive method for the development of RA-ILD. Due to non-specific symptoms, the diagnosis of RA-ILD is often overlooked. Despite significant morbidity and increased risk of death due to RA-ILD, currently, we do not have any guidelines on the management of this clinical situation. Moreover, the problem is even more complicated due to the possibility of potential pneumotoxicity of many disease-modifying drugs and their unclear effectiveness regarding lung involvement in RA. Taken together, optimal management of a patient with RA-ILD is a clinical challenge. There is an urgent need to clarify several clinical aspects concerning the early diagnosis, monitoring, establishing indications for treatment, and choice of appropriate therapeutic management. In this work, we performed a scientific review of the current state of knowledge on RA-ILD and indicated the directions of future research needed aiming at improving patients’ care.

Download Full-text

Dengue Hemorrhagic Fever in North America: A Case Report

PEDIATRICS ◽

10.1542/peds.66.4.631 ◽

1980 ◽

Vol 66 (4) ◽

pp. 631-633

Author(s):

Linda Spigelblatt ◽

Robert Rosenfeld ◽

Yvette Bonny ◽

Michel Laverdiere

Keyword(s):

World War Ii ◽

North America ◽

Dengue Fever ◽

Dengue Hemorrhagic Fever ◽

Pacific Islands ◽

Hemorrhagic Fever ◽

Maculopapular Rash ◽

Hemorrhagic Diathesis ◽

Epidemic Disease ◽

The Pacific

Dengue hemorrhagic fever, a severe, often fatal, illness, occurs mostly in children and is characterized by a hemorrhagic diathesis, fever, vomiting, a maculopapular rash, liver involvement, and occasionally, a protein-losing shock syndrome.1 This disease is to be differentiated from dengue fever, a relatively benign disease occurring primarily in adults and manifested by myalgia, arthralgia, bone pain, and leukopenia. Cases of dengue fever in North America have been described among travellers from the Carribean.2-6 Dengue hemorrhagic fever is an epidemic disease described after World War II and limited to areas of Southeast Asia, India, and the Pacific islands.7-8 We believe this to be the first reported case in North America of dengue hemorrhagic fever with disseminated intravascular coagulation in a child of Southeast Asian origin.

Download Full-text

Benchmarking Machine Learning Models to Assist in the Prognosis of Tuberculosis

Informatics ◽

10.3390/informatics8020027 ◽

2021 ◽

Vol 8 (2) ◽

pp. 27

Author(s):

Maicon Herverton Lino Ferreira da Silva Barros ◽

Geovanne Oliveira Alves ◽

Lubnnia Morais Florêncio Souza ◽

Elisson da Silva Rocha ◽

João Fausto Lorenzato de Oliveira ◽

...

Keyword(s):

Machine Learning ◽

Clinical Symptoms ◽

Treatment Decision ◽

Gradient Boosting ◽

Original Form ◽

Learning Models ◽

Data Set ◽

Risk Of Death ◽

Increased Risk ◽

Machine Learning Models

Tuberculosis (TB) is an airborne infectious disease caused by organisms in the Mycobacterium tuberculosis (Mtb) complex. In many low and middle-income countries, TB remains a major cause of morbidity and mortality. Once a patient has been diagnosed with TB, it is critical that healthcare workers make the most appropriate treatment decision given the individual conditions of the patient and the likely course of the disease based on medical experience. Depending on the prognosis, delayed or inappropriate treatment can result in unsatisfactory results including the exacerbation of clinical symptoms, poor quality of life, and increased risk of death. This work benchmarks machine learning models to aid TB prognosis using a Brazilian health database of confirmed cases and deaths related to TB in the State of Amazonas. The goal is to predict the probability of death by TB thus aiding the prognosis of TB and associated treatment decision making process. In its original form, the data set comprised 36,228 records and 130 fields but suffered from missing, incomplete, or incorrect data. Following data cleaning and preprocessing, a revised data set was generated comprising 24,015 records and 38 fields, including 22,876 reported cured TB patients and 1139 deaths by TB. To explore how the data imbalance impacts model performance, two controlled experiments were designed using (1) imbalanced and (2) balanced data sets. The best result is achieved by the Gradient Boosting (GB) model using the balanced data set to predict TB-mortality, and the ensemble model composed by the Random Forest (RF), GB and Multi-Layer Perceptron (MLP) models is the best model to predict the cure class.

Download Full-text