Employing supervised machine learning algorithms for classification and prediction of anemia among youth girls in Ethiopia.

In developing countries, one-quarter of young women have suffered from anemia. However, the available studies in Ethiopia have been usually used the traditional stastical methods. Therefore, this study aimed to employ multiple machine learning algorithms to identify the most effective model for the prediction of anemia among youth girls in Ethiopia. A total of 5642 weighted samples of young girls from the 2016 Ethiopian Demographic and Health Survey dataset were utilized. The data underwent preprocessing, with 80% of the observations used for training the model and 20% for testing. Eight machine learning algorithms were employed to build and compare models. The model performance was assessed using evaluation metrics in Python software. Various data balancing techniques were applied, and the Boruta algorithm was used to select the most relevant features. Besides, association rule mining was conducted using the Apriori algorithm in R software. The random forest classifier with an AUC value of 82% outperformed in predicting anemia among all the tested classifiers. Region, poor wealth index, no formal education, unimproved toilet facility, rural residence, not used contraceptive method, religion, age, no media exposure, occupation, and having more than 5 family size were the top attributes to predict anemia. Association rule mining was identified the top seven best rules that most frequently associated with anemia. The random forest classifier is the best for predicting anemia. Therefore, making it potentially valuable as decision-support tools for the relevant stakeholders and giving emphasis for the identified predictors could be an important intervention to halt anemia among youth girls.

Evaluation of Dermatoglyphic Features of Type 2 Diabetic Patients as Compared to Non-Diabetics Attending Hospitals in Southern Ethiopia.

Background: Diabetes mellitus (DM) is a complex metabolic disorder characterized by persistent hyperglycemia resulting from defects in insulin secretion, insulin action, or both. Type 2 DM has a strong hereditary background. A study of the dermatoglyphic features can help in the early identification of people with a genetic predisposition to develop type 2 DM. Purpose of the Study: The study was aimed to compare the finger and palmar dermatoglyphics features in type 2 diabetic and non-diabetic patients and to evaluate the association with other variables. Patients and Materials: Institution-based cross-sectional study was conducted in one referral hospital and three primary hospitals located in Gedeo Zone. Three hundred ninety study participants (130 type 2 diabetics and 260 non-diabetics) from adult OPD and diabetic clinics were included in the study. The association among different dermatoglyphics variables was analyzed using the chi-square test. An independent t-test was conducted to analyze the mean difference and to test the significance of numerical variables. A statistically significant association was declared at P-value <0.05. Results: In the present study, a total of 3900 digits (1300 from type 2 diabetic and 2600 non-diabetic) were analyzed for distribution of digital dermatoglyphic patterns. Loop pattern is found to be the commonest in both study groups with a prevalence of 65.8% and 60.7% among type 2 diabetics and non-diabetics, respectively. Both the right-left axial tri-radius angle was significantly higher in type 2 diabetes mellitus patients compared to non-diabetic study participants. Total finger ridge counts and absolute finger ridge counts showed no significant difference in both groups. Conclusion: The study revealed Loop pattern was more frequent in type 2 DM compared to non-diabetic study participants. Whorl type was more common than arch type in non-diabetic patients compared to type 2 DM groups. The result also showed tri-radius angle is significantly wider in diabetic groups in both hands.

Patterns of superficial veins in the cubital fossa and its clinical implications among southern Ethiopian population.

Cubital fossa is the site where the venous accesses are frequently made. Superficial veins at this site display variations in their pattern among different populations. Knowledge of different venous pattern in the cubital fossa is important for diagnostic, surgical and therapeutic procedures. The purpose of this study was to report variations of the cubital superficial vein patterns in the southern Ethiopian subjects. An institution based cross-sectional study design was employed among 401 randomly selected patients presented at the triage room of Arba Minch General Hospital from January 15 to February 15, 2021. A questionnaire was used to collect socio-demographic data and images of the common and variant superficial venous patterns were recorded. Descriptive statistical analysis was performed. P<0.05 was considered as statistical significance. In the present study, a total of 802 cubital fossae from 401 study participants were examined. Five patterns of superficial veins were identified. Type 2 was the most common pattern and observed in 55.0% of cubital fossae (42.1% right and 67.8% left cubital fossae). The least common, type 5 variant was detected in 2.6% cubital fossae (2.7% right and 2.5% left). Statistically significant association based on sex and laterality was noted. The current study concluded that type 2 and type 3 patterns were more frequent superficial venous patterns in the cubital fossa and more common in males than female. Awareness of these uncommon cubital venous patterns and their incidence is very useful for those performing venipuncture or venisection especially under emergency conditions.

Factors associated with preterm birth at Wachemo University Nigist Eleni Mohammed memorial hospital, southern Ethiopia: case-control study.

BACKGROUND: Preterm birth is defined as the birth of a baby before 37 completed weeks of gestation. Worldwide, prematurity is the second foremost cause of death in children under the age of 5 years. Preterm birth also gives rise to short and long term complications. Therefore, the primary aim of this study was to identify the factors associated with preterm birth in Wachemo University Nigist Eleni Mohammed Memorial referral hospital, Hadiya Zone, Southern Ethiopia. METHODS: An institution-based unmatched case-control study was conducted from July 01, 2018 to June 30, 2019 among mothers who gave birth in Wachemo University Nigest Eleni Mohammed Memorial referral hospital. A retrospective one-year data was retrieved from medical records of mothers with their index neonates. Simple random sampling technique was employed to recruit study participants. SPSS version 20 software was used for data entry and computing statistical analysis. Both bivariable and multivariable logistic regression analyses were used to determine the association of each independent variable with the dependent variable. Odds ratio with their 95% confidence intervals was computed to identify the presence and strength of association, and statistical significance was affirmed if p < 0.05. RESULT: The current study evaluated 213 medical records of mothers with index neonates (71 cases and 142 controls). Urban residency [AOR = 0.48; 95% Cl; 0.239, 0.962], antenatal care follow up [AOR = 0.08; 95 Cl; 0.008, 0.694], premature rupture of membranes [AOR = 3.78; 95% Cl; 1.467, 9.749], pregnancy induced hypertension [AOR = 3.77; 95% Cl; 1.408, 10.147] and multiple pregnancies [AOR = 5.53; 95% Cl; 2.467, 12.412] were the factors associated with preterm birth. More than one-third (36.6%) preterm neonates died in the present study. CONCLUSIONS: The present study found that urban residency, antenatal care follow up, premature rupture of membranes, pregnancy induced hypertension and multiple pregnancies were factors associated with preterm birth. The mortality among preterm neonates is high. Enhancing antenatal care follow up and early detection and treatment of disorders among pregnant women during antenatal care and undertaking every effort to improve outcomes of preterm birth and reduce neonatal mortality associated with prematurity is decisive.

Xpert MTB/RIF assay for diagnosis of pulmonary tuberculosis in sputum specimens in remote health care facility.

BACKGROUND: Xpert MTB/RIF assay is considered as a great advance over conventional smear and culture in the diagnosis of TB and MDR-TB by simultaneously detecting M.tuberculosis and rifampicin resistance bacilli. However, very little information regarding the performance characteristics of Xpert MTB/RIF assay is available in Ethiopia. Therefore, the purpose of this study was to evaluate the performance of Xpert MTB/RIF assay compared to conventional sputum smear and culture methods for the diagnosis of pulmonary tuberculosis in remote health care facility. METHODS: A paired expectorated sputum samples were obtained from 227 consecutively recruited patients with signs and symptoms suggestive of tuberculosis at Karamara hospital during December 2013 to May 2014. One of the sputum specimen was tested directly by Ziehl-Neelsen staining and Xpert MTB/RIF assay without NALC-NaOH decontamination. The other of pair of sputa specimen was cultured for isolation of TB bacilli by conventional methods. Diagnostic performance of Xpert MTB/RIF assay and AFB smear microscopy were calculated against culture as the gold standard. RESULTS: Overall 25.5% (58/227) samples were positive for Mycobacterium tuberculosis complex (MTBC) by MGIT and/or LJ media of which 36.2% (21/58) and 65.5% (35/58) were positive by AFB smear microscopy and Xpert MTB/RIF respectively. The sensitivity, specificity, as well as the positive and negative predictive value of Xpert MTB/RIF assay were 65.5% (95% CI: 53.3-77.7%), 96.3% (95% CI: 93.4-99.2%), 86.4% (95% CI: 76.2-96.5%), and 88.6% (95% CI: 83.9-93.3%) respectively. Eighteen of 58 (31%) cases that were smear microscopy negative, were positive by Xpert MTB/RIF assay. CONCLUSIONS: Although Xpert MTB/RIF assay demonstrated high sensitivity in detecting MTBC in sputum specimens compared with conventional AFB smear microscopy, it demonstrated suboptimal sensitivity in smear negative patients compared to conventional culture.