SARS-CoV-2 variant typing using real-time reverse transcription-polymerase chain reaction-based assays in Addis Ababa, Ethiopia.

Objectives: This study aimed to determine the SARS-CoV-2 variants in the first four COVID-19 waves using polymerase chain reaction (PCR)-based variant detection in Addis Ababa, Ethiopia. Methods: A cross-sectional study was conducted using repository nasopharyngeal samples stored at the Ethiopian Public Health Institute COVID-19 testing laboratory. Stored positive samples were randomly selected from the first four waves based on their sample collection date. A total of 641 nasopharyngeal samples were selected and re-tested for SARS-CoV-2. RNA was extracted using nucleic acid purification instrument. Then, SARS-CoV-2 detection was carried out using 10 µl RNA and 20 µl reverse transcription-PCR fluorescent mix. Cycle threshold values <38 were considered positive. Results: A total of 374 samples qualified for B.1.617 Lineage and six spike gene mutation variant typing kits. The variant typing kits identified 267 (71.4%) from the total qualifying samples. Alpha, Beta, Delta, and Omicron were dominantly identified variants from waves I, II, III, and IV, respectively. From the total identified positive study samples, 243 of 267 (91%) of variants identified from samples had cycle threshold values <30. Conclusions: The study data demonstrated that reverse transcription-PCR-based variant typing can provide additional screening opportunities where sequencing opportunity is inaccessible. The assays could be implemented in laboratories performing SARS-CoV-2 molecular testing.

Association between helminth infection and allergic disorders among children in Batu, Ethiopia.

BACKGROUND: Helminths are potent immunomodulators and in their chronic infection state they may protect against allergy-related disease and atopy. However, they are also known for inducing allergic conditions. This study aimed to assess the association between helminths,  atopy and allergic conditions. METHODS: A total of 461 school children participated in this cross-sectional study. Data on allergic symptoms and a range of confounding variables was gathered from parents via an interviewer-led questionnaire. Skin sensitization to house dust mite and cockroaches was analyzed, and a stool sample was collected for helminth analysis. Serum total Immunoglobulin E using enzyme-linked immunosorbent assay and eosinophil count were also measured. RESULTS: Overall sensitivity to both allergens was 2.4%. Self-reported allergic outcomes in the last 12 months for the 461 participants had been : wheezing 3.7%, asthma 2.2%, eczema 13.2% and hay fever 6.9%. Overall, the prevalence of helminth infection was 11.9% (53/444). A borderline significant association was found between atopy and any allergy symptoms (odds ratio [OR]: 3.32, 95% confidence interval [95% CI: 0.99, 11.1], p = .052). There was no significant association between helminths and atopy (OR: 0.64 [95% CI: 0.29, 1.41], p = .268) and also between helminths and allergic symptoms (OR: 0.64 [95% CI: 0.29, 1.41], p = .268). Bivariate analysis showed keeping an animal in the house increases the risk of atopy while maternal and paternal history of allergy increases the risk of developing allergic symptoms in the children. CONCLUSION AND CLINICAL RELEVANCE: This study found a non-significant inverse association between helminths infection and atopy and allergic disorders, likely due to reduced statistical power, resulting in a lower prevalence of atopy and allergic conditions. A high powered longtitudinal study is necessary to explore the casuality and potential therapeutic benefits of helminths for allergic disorders.

Prevalence and antibiotic resistance pattern of bacteria from sepsis suspected neonates at St. Paul's Hospital Millennium Medical College, Addis Ababa, Ethiopia.

BACKGROUND: Neonatal sepsis is the major cause of neonatal mortality and morbidity, especially in low and middle-income countries. Continuous monitoring of pathogens and their antibiotic resistance pattern is crucial for managing neonatal sepsis. This study aimed to determine neonatal sepsis due to bacteria, antibiotic resistance patterns, associated risk factors and patient outcomes at St. Paul's Hospital Millennium Medical College. METHOD: An institutional-based cross-sectional study was conducted on 400 neonates suspected of sepsis at St. Paul's Hospital Millennium Medical College from March 2020 to July 2020. A questionnaire was used to collect socio-demographic information, clinical parameters and potential risk factors from study participants. About 2ml of blood was drawn aseptically and inoculated into Tryptone Soya Broth at the patient's bedside. Bacterial identification was performed by using standard microbiological techniques. The disk diffusion method was used to determine the antibiotic susceptibility patterns of each isolated bacteria. Data entry and analysis were done using Statistical Package for Social Sciences (SPSS) version 20 software. Bivariate and multivariable logistic regressions were used to assess associated risk factors of neonatal sepsis. A p-value less than 0.05 was considered statically significant with a 95% confidence interval. RESULTS: The overall prevalence of neonatal septicemia was 21% (84/400). Of these, 67 (79.8%) and 17 (20.2%) were gram-negative and gram-positive bacteria, respectively. Klebsiella spp, 37 (44%), E. coli 19 (21.6%) and Coagulase negative Staphylococci 13 (15.47%) were the leading cause of neonatal sepsis. Ciprofloxacin and amikacin were the most effective antibiotics for gram-negative and gram-positive bacteria. Multidrug resistance was observed in 84% of the bacterial isolates. Low birth weight and preterm were associated with neonatal septicemia (AOR = 49.90, 95% CI = 15.14-123.081, P = 0.002) and (AOR = 18.20, 95% CI = 6.835-27.541, P = 0.004) respectively. CONCLUSION: Klebsiella spp and E. coli were frequently isolated bacteria in our study. The proportion of multidrug-resistance was significantly high. Most isolated bacteria were resistant to ampicillin, ceftazidime, cefotaxime and gentamycin, which indicates the necessity of continuous evaluation of antibiotic resistance rate.

Evaluation of the Diagnostic Performance of the LINEAR Cromatest and Laboquick URS 10-T Dipsticks Among Urinary Tract Infection Suspects in Addis Ababa, Ethiopia: A Cross-Sectional Study.

Background: Urinary tract infections (UTIs) bring a significant and serious health-related problem. Repeated infections may lead to the development of renal scarring and end-stage renal dysfunction. Therefore, balancing the choices of UTI diagnostic tools depending on the costs versus accuracy can minimize false results that may subject patients to wrong treatments. Objective: The objective of the study was to evaluate the diagnostic performance of LINEAR cromatest and Laboquick URS 10-T dipsticks against conventional urine culture at Arsho Advanced Medical Laboratory (AAML), Addis Ababa, Ethiopia. Methods: A similar cohort of UTI-suspected patients from our previous study, who were prospectively enrolled from Arsho Advanced Medical Laboratory, Addis Ababa, Ethiopia, were considered for this analysis. Data analysis was performed using SPSS version 26. The sensitivity, specificity, and predictive value of different dipsticks were calculated using culture as a gold standard. ROC curve analysis was also used to determine diagnostic performance. A test with a P-value of <0.05 was considered statistically significant. Results: Out of 446 urine samples processed, bacterial growth was observed in 141 (31.6%). Of this figure, 105/141 (74.5%) and 36/141 (25.5%) were from female and male participants, respectively. The sensitivity and specificity of the LINEAR Cromatest dipstick were 83.7% and 67.9%, respectively (P-value <0.001). The Laboquick URS 10-T dipstick showed sensitivity and specificity of 87.9% and 68.5%, respectively (P-value <0.001). The ROC analysis showed an AUC of >0.7 for both dipsticks. Conclusion: In a setting where there is no access to urine culture, the urine dipstick may be considered an alternative diagnostic tool in the diagnosis of UTIs.

Microbiological Investigations of Fine Needle Aspirates from Newly Suspected and Previously Treated Tubercular Lymphadenitis Patients.

Background: Extrapulmonary tuberculosis (EPTB), particularly tubercular lymphadenitis (TBLN), remains to pose a huge public health problem in Ethiopia. A significant number of TBLN patients who completed a full course anti-TB treatment regimen were reported to have enlarged lymph nodes and other TB-like clinical presentations. This could either be from a paradoxical reaction or microbiological relapse, possibly due to mono/multi-drug resistance. Objective: To investigate the rate of mono and multidrug resistance patterns of Mycobacterium tuberculosis as a cause of the observed treatment failures in clinically diagnosed and anti-TB treatment (newly or previously)-initiated LN patients. Methods: A cross-sectional study was conducted on 126 TBLN-suspected and previously treated patients between March and September 2022. Data were analyzed using SPSS (Version 26.0). Descriptive statistics were used to determine the frequency, percentage, sensitivity, specificity, and positive and negative predictive values. The level of agreement was determined using Cohen's kappa and a Chi-square test was used to measure the association between risk factors and laboratory test outcomes. A P-value <0.05 was considered statistically significant. Results: Mycobacterium tuberculosis was confirmed in 28.6% (N=36) of the 126 cases using BACTEC MGIT 960 culture detection method. Approximately, 13% (N=16) of the samples were collected from previously treated TBLN patients, of which 5/16 (31.3%) were multi-drug resistant, 7/16 were drug-sensitive and 4/16 were culture negative. To rule out other non-tuberculous agents, all samples were grown on blood and Mycosel agar plates, and no growth was detected. Conclusion: The emergence of drug resistant (DR) TB seems to not just be limited to pulmonary form but also to TBLN. In this study we observed a considerable number of microbiologically confirmed relapses among previously treated cases, possibly indicating the need for confirmation of drug resistance using rapid molecular methods or phenotypical methods during treatment follow up.

Prevalence of Hepatitis B Virus, Human Immune Deficiency Virus and Associated Risk Factors Among Individuals with Presumptive Pulmonary Tuberculosis Attending at Saint Peter's Specialized Hospital, Addis Ababa, Ethiopia.

Background: Hepatitis B virus (HBV), human immunodeficiency virus (HIV) and tuberculosis are the causes of widely spread infectious disease, especially in resource-limited countries. The extent of HBV infection and its contributing factors among people with suspected pulmonary tuberculosis (PTB) were not adequately addressed. Objective: To assess the prevalence of HBV, HIV & their associated risk factors and the magnitude of TB among individuals with presumptive pulmonary tuberculosis attending at St. Peter's Specialized Hospital, Addis Ababa, Ethiopia. Methods: A cross-sectional study was conducted among 387 individuals with presumptive PTB from October to December 2020. A standard questionnaire was used to collect socio-demographic data and associated risk factors. Sputum samples were analyzed by GeneXpert, Florescent Microscopy and Ziehl-Nelson staining technique. HBsAg test was carried out using Murex Version 3 ELISA test kit from serum/Plasma samples, HIV testing was performed by rapid HIV test kits and data were analyzed using SPSS version 23. Results: The mean age of study participants was 44.2 years. Overall, 14 (3.6%), 28 (7.2%) and 37 (9.6%) of them were positive for HBV, HIV & TB, respectively. Only single patient was HBV-HIV co-infected (0.3%). The TB-HIV co-infection was identified in 6 (1.6%). In multivariate analysis, being partner separated, alcohol consumption, body piercing and having multiple sexual partners were significantly associated with HBV infection. Having a spouse, who is divorced, widowed, sharing scissors, alcohol consumption and contact with multiple sexual partners also significantly associated with HIV infection. Conclusion: This study showed that HBV, HIV and TB are still public health issues that need awareness and health education on the risky behaviors and transmission of HBV, HIV & TB among individuals with presumptive TB suspects. Further large-scale study is necessary.

HIV rapid test performance among health facilities enrolled in HIV rapid test quality improvement initiative (RTQII) in Ethiopia.

BACKGROUND: As the Human Immunodeficiency Virus (HIV) rapid testing services expanded to reach the global target that 95% of people living with the virus will know their status by 2030, ensuring the quality of those services becomes critical. This study was conducted to assess the performance of HIV Rapid testing at sites in health facilities that were enrolled in the Rapid Test Quality Improvement Initiative (RTQII) in Ethiopia. METHODS: Characterized HIV proficiency testing (PT) panels of Dried Tube Specimen (DTS) were prepared, verified, and distributed to testing sites from August to December 2019. In addition on-site evaluation of HIV testing sites (HTSs) was conducted using a checklist to assess testing conditions. For proficiency testing, the study included 159 HIV testing sites (HTSs) in 41 Health facilities (HFs) in five administrative regions and two city administrations. The collected data was analyzed by SPSS version 20 and chi-square test was applied to identify the association between acceptable performance and contributing factors. Testing sites with 100% PT score as well as conducting the test with adherence to the National HIV Testing Algorithm were considered acceptable. RESULTS: The overall acceptable performance (100% PT score with the correct algorithm followed) was found to be 62% while 12% scored 80% and 11% scored between 20 and 60%. The rest 15% were not considered as acceptable due to failure to adhere to the National HIV Testing Algorithm. Testing sites that participated in External Quality Assessment/Proficiency Testing schemes have shown better performance than those that did not participate with 70% and 56% performance respectively (p = 0.057).

Incidence of surgical site infection, bacterial isolate, and their antimicrobial susceptibility pattern among patients who underwent surgery at Dessie Comprehensive Specialized Hospital, Northeast Ethiopia.

Objective: Surgical site infection is the most common postoperative complication worldwide, representing a major burden for patients and health systems. The aim of the study is to determine the incidence and bacterial profile that cause surgical site infection among patients who underwent surgery in parts of Northeast Ethiopia. Methods: A health facility-based cross-sectional study was conducted in Dessie Comprehensive Specialized Hospital from July 22 to October 25, 2016. A total of 338 patients from the obstetrics and gynecology and general surgical wards were included, through consecutive sampling technique. The specimens were collected aseptically on the first day when the patients had presented with clinical evidence of infection and then sent to the microbiology laboratory. The data were entered and analyzed by SPSS version 20, and the results were explained by frequency distribution in tables and figures. Results: The majority of participants were female (74.3%) and more than half (61.2%) of the surgeries were performed in the gynecology and obstetrics ward. Clinically, 49 patients (14.5%) were diagnosed as developing surgical site infection, and wound swabs were taken for bacteriological study. About 41 (83.7%) swabs showed bacterial growth, indicating 12.13% overall prevalence of bacterial surgical site infection. Out of 48 bacterial isolates, more than half (56.25%) of them were Gram negative. The most frequent isolate was Staphylococcus aureus, 14 (66.67%), followed by Escherichia coli, 9 (33.33%). Out of the total bacterial isolates, 38 (79.2%) isolates were found to be multidrug resistant, and the rate of multidrug resistant was higher among Gram-negative isolates. Conclusion: An average rate of surgical site infection was found to be reported and significant numbers of bacterial isolates were also detected. The highest rate of surgical site infection was reported in prostate surgery, followed by small bowel, vaginal hysterectomy, and exploratory laparotomy surgical procedures. Periodic surveillance on the incidence rate and bacterial profile along with the determination of their antibiotic susceptibility should be performed.

Prevalence of Bacterial Urinary Tract Infection, Associated Risk Factors, and Antimicrobial Resistance Pattern in Addis Ababa, Ethiopia: A Cross-Sectional Study.

Background: Urinary tract infections (UTIs) brought a significant and serious health-related problem that may lead to the subsequent development of serious indications with the challenge of the emergence of antibiotic resistance. Therefore, the choice of antibiotics depends on the accuracy of the diagnostic tool of UTIs to minimize false results that may subject patients to wrong treatments. This study aimed to determine the prevalence of bacteriuria, associated factors, and AMR pattern of UTI-suspected patients. Methods: A cross-sectional study was conducted from March to May 2022, at Arsho Advanced Medical Laboratory (AAML), Addis Ababa, Ethiopia. Species identification and isolation from bacterial colonies were characterized by gram stain and biochemical properties followed by antibiotic susceptibility testing using the Kirby-Bauer method on Muller-Hinton agar. Logistic regression analysis was carried out to determine the association between the independent variables and significant bacterial growth to identify factors that affect the prevalence of UTI. A test is considered statistically significant that has a P value less than 0.05. Results: Out of 141 (31.6%) which yielded significant bacteriuria, 16 different species of bacterial uropathogens were identified. A total of 105/446 (91 Gram-negative and 14 Gram-positive) of bacterial growth in the female gender and 36/446 (33 Gram-negative and 3 Gram positive) in male were observed with a P value of 0.03. The most predominant bacteria were E. coli followed by Klebsiella pneumoniae. Amoxicillin had shown the highest resistance rate (100%) followed by Ampicillin (98.9%). Females and participants with previous infections were shown to be associated with significant bacterial growth. Conclusion: Based on our study finding, the ordinarily used antibiotics seem to face emerging resistant strains, which needs considerable and due attention to the impact of UTI in developing countries including Ethiopia. History of previous UTIs and female gender were shown to be possible risk factors associated with UTIs.

Drug Resistance in Tuberculous Lymphadenitis: Molecular Characterization.

Background: Drug-resistant tuberculosis (TB) epidemic in high-TB-incidence countries, particularly Ethiopia, remains a significant challenge. As a result, we investigated the drug resistance, common gene mutation, and molecular characterization of mycobacterial isolates from patients with suspected tuberculous lymphadenitis (TBLN). Methodology. A cross-sectional study of 218 FNA samples from TBLN patients inoculated on Lowenstein-Jensen media was carried out. The culture isolates were identified as MTB by polymerase chain reaction (PCR) and the difference-9 (RD9) test region. In addition, the GenoType MTBDRplus assay tested the first and second-line MTB drugs, and the spoligotyping strain-dependent polymorphism test was determined. Results: Among the 50 culture-positive isolates, 14% (7/50) had drug resistance caused by a gene mutation. Out of these, 4 (8%) isolates were mono-resistant to isoniazid drug, which is caused by a gene mutation in katG in the region of interrogated at codon 315 in the amino acid sequence of S315T1, and 3 (6%) isolates were resistant to both rifampicin and isoniazid drugs. The mutation was observed for katG (at codon 315 with a change in the sequence of amino acid S315T) and rpoB (at codon 530-533 with a change in the sequence of amino acid S531L (S450L)) genes. The most prevalent spoligotypes were orphan and SIT53 strains. Conclusion: The predominance of INH mono-resistance poses a critical risk for the potential development of MDR-TB, as INH mono-resistance is a typical pathway to the occurrence of MDR-TB. The orphan and SIT53 (T) strains were the most common in the study area, and a drug-resistant strain caused by a common gene mutation could indicate the transmission of clonal-resistant strains in the community.

Phenotypic characterization of Peripheral B cells in Mycobacterium tuberculosis infection and disease in Addis Ababa, Ethiopia.

BACKGROUND: Mortality and morbidity from tuberculosis (TB) remain one of the most important public health issues. Although cell-mediated immunity is the main immune response against Mycobacterium tuberculosis (MTB), the role of B-cells during MTB infection and disease is unclear. METHODS: Peripheral blood mononuclear cells (PBMC) were isolated from treatment naïve Pulmonary TB patients (TB, n = 16), latent TB-infected participants (LTBI, n = 17), and healthy controls (HC, n = 19). PBMCs were stained with various fluorescently labeled antibodies to define B-cell subsets using multicolor flow cytometry. RESULTS: Atypical memory B cells (CD19+CD27-CD21-) and circulating marginal zone B-cells (CD19+CD27+CD21+IgM+IgD+CD23-) were significantly higher in active TB when compared to LTBI and HC. CD5+ regulatory B cells (Breg, CD19+CD24hiCD38hiCD5+) and resting B-cells (CD19+CD27+CD21+) in Active TB patients were significantly lower compared to HC and LTBI. Overall, there were no differences in B cell percentages (CD19+), naïve B cells (CD19+CD27-CD21+), Breg (CD19+CD24hiCD38hi), and activated memory B cells (CD19+CD27+CD21-) among the three study groups. CONCLUSIONS: These results indicated that multiple subsets of B cells were associated with TB infection and disease. It will be useful to examine these cell populations for their potential use as biomarkers for TB disease and LTBI.

Prevalence of Dyslipidemia among School-Age Children and Adolescents in Addis Ababa, Ethiopia.

Objective Lipid abnormalities during childhood might be associated with a higher risk of atherosclerosis development in adulthood. In Ethiopia, there were no data on this area. The present work was aimed at assessing the lipid profile abnormalities among children and adolescents aged between 5 and 17 years, in Addis Ababa, Ethiopia. Materials and Methods The present school-based cross-sectional study was done from March 2019 to October 2019 in the capital city, Addis Ababa. A total of 504 students were randomly recruited for this study. The total cholesterol, triglyceride, low-density lipoprotein, and high-density lipoprotein-cholesterol levels were determined using the Cobas c501 automated chemistry analyzer. The prevalence of lipid abnormalities was described in percentages. Bivariate and multivariate analyses were performed, and data with p -value less than 0.05 was considered statistically significant. Results Dyslipidemia in at least one of the lipid profiles was observed in 322 (63.9%) school children, whereas only one of the participants had an abnormality in all four lipid profile tests. The prevalence of total cholesterol 200 mg/dL or higher, low-density lipoprotein 130 mg/dL or higher, high-density lipoprotein less than 40 mg/dL, and triglycerides 130 mg/dL or higher was in 14 (2.8%), 16 (3.2%), 294 (58.4%), and 46 (19.1%) participants, respectively. Conclusion In the current work, the majority of the study participants had a normal lipid profile except increased prevalence of a reduced concentration of high-density lipoprotein-cholesterol. The overall prevalence of serum lipid abnormalities was 63.9%. Serum lipid levels did not show significant differences with sex, age, fasting habits, or obesity.

Performances of four Nucleic Acid Amplification Tests for the identification of SARS-CoV-2 in Ethiopia.

Since Coronavirus Disease-2019 (COVID-19) outbreak was reported, many commercial Nucleic Acid Amplification Tests (NAAT) have been developed all over the world, and it has been the standard method. Even though several assays were rapidly developed and applied to laboratory diagnostic testing, the performance of these assays was not evaluated in different contexts. Thus, this study aimed to assess the performance of Abbott SARS-CoV-2, Daan Gene, BGI and Sansure Biotech assays by using Composite Reference Standard (CRS). The study was conducted at the Ethiopian Public Health Institute (EPHI) from December 1 to 30/2020. Of the 164 nasopharyngeal samples were extracted by using a QIAamp RNA mini kit and Abbott DNA sample preparation system. Out of 164 samples, 59.1% were positive and 40.9% were negative by CRS. Sansure Biotech positivity was significantly low compared to CRS (p < 0.05). The overall agreement of the four assays compared to CRS was 96.3-100%. The performance of the four assays had almost comparable diagnostic performance, except for a low positive rate of Sansure Biotech assay. Hence, Sansure Biotech assay [Research Use Only (RUO)] needs further verification on its use in Ethiopia. Finally an additional study should be considered for evaluating assays with respective manufacturer claims.

Magnitude of Mycobacterium tuberculosis, drug resistance and associated factors among presumptive tuberculosis patients at St. Paul's Hospital Millennium Medical College, Addis Ababa, Ethiopia.

BACKGROUND: Mycobacterium tuberculosis (M. tuberculosis) remains one of the most significant causes of death and a major public health problem in the community. As a result, the aim of this study was to determine magnitude of Mycobacterium tuberculosis, its drug resistance, and associated factors among presumptive tuberculosis (TB) patients at St. Paul's Hospital Millennium Medical College, Addis Ababa, Ethiopia. METHODS: Cross-sectional study was conducted at St. Paul's Hospital Millennium Medical College (SPHMMC), Addis Ababa, Ethiopia from January to July 2019. Demographic and clinical data were collected by structured questionnaire through face to face interview. Using microscopic examination and GeneXpert MTB/RIF assay and culturing in the Lowenstein-Jensen (LJ) culture media, we collected and analyzed both pulmonary and extra-pulmonary clinical samples. Data were analyzed by SPSS version 23. Binary logistic regression was done to identify the associated risk factors and p-value less than 0.05 was taken as significant association. RESULTS: Of the total 436 respondents, 223(51%) were male. The mean ±SD age of the participants was 38±17years. Overall, 27/436(6.2%) of the participants had confirmed Mycobacterium tuberculosis using the GeneXpert MTB/RIF assay and LJ culture media, and two isolates were resistant to RIF and one to INH medication, with two (0.5%) being MDR-TB. MTB infection was associated with previous TB contact history, patient weight loss, and CD4+ T-cell counts of 200-350/mm3 of blood. CONCLUSION: The magnitude of M. tuberculosis and MDR-TB in this study underscores the need for improved early case detection and management of MDR-TB in order to reduce transmission and patient suffering.

Helicobacter pylori (H. pylori) risk factor analysis and prevalence prediction: a machine learning-based approach.

BACKGROUND: Although previous epidemiological studies have examined the potential risk factors that increase the likelihood of acquiring Helicobacter pylori infections, most of these analyses have utilized conventional statistical models, including logistic regression, and have not benefited from advanced machine learning techniques. OBJECTIVE: We examined H. pylori infection risk factors among school children using machine learning algorithms to identify important risk factors as well as to determine whether machine learning can be used to predict H. pylori infection status. METHODS: We applied feature selection and classification algorithms to data from a school-based cross-sectional survey in Ethiopia. The data set included 954 school children with 27 sociodemographic and lifestyle variables. We conducted five runs of tenfold cross-validation on the data. We combined the results of these runs for each combination of feature selection (e.g., Information Gain) and classification (e.g., Support Vector Machines) algorithms. RESULTS: The XGBoost classifier had the highest accuracy in predicting H. pylori infection status with an accuracy of 77%-a 13% improvement from the baseline accuracy of guessing the most frequent class (64% of the samples were H. Pylori negative.) K-Nearest Neighbors showed the worst performance across all classifiers. A similar performance was observed using the F1-score and area under the receiver operating curve (AUROC) classifier evaluation metrics. Among all features, place of residence (with urban residence increasing risk) was the most common risk factor for H. pylori infection, regardless of the feature selection method choice. Additionally, our machine learning algorithms identified other important risk factors for H. pylori infection, such as; electricity usage in the home, toilet type, and waste disposal location. Using a 75% cutoff for robustness, machine learning identified five of the eight significant features found by traditional multivariate logistic regression. However, when a lower robustness threshold is used, machine learning approaches identified more H. pylori risk factors than multivariate logistic regression and suggested risk factors not detected by logistic regression. CONCLUSION: This study provides evidence that machine learning approaches are positioned to uncover H. pylori infection risk factors and predict H. pylori infection status. These approaches identify similar risk factors and predict infection with comparable accuracy to logistic regression, thus they could be used as an alternative method.

High Levels of Methicillin-Resistant Staphylococcus aureus Carriage Among Healthcare Workers at a Teaching Hospital in Addis Ababa Ethiopia: First Evidence Using mecA Detection.

Background: Staphylococcus aureus is a major human pathogen and causes healthcare and community-acquired infection. Data on the extent of MRSA colonization among health-care workers (HCWs) in sub-Saharan Africa are limited. Hence, we determined the burden of MRSA colonisation among HCWs and administrative staff in Tikur Anbessa Specialised Hospital (TASH), College of Health Sciences (CHS), Addis Ababa University, Ethiopia. Methods: Using a cross-sectional study design, participants were screened for MRSA colonisation between June 2018 and August 2019 using nasal swabs. The swabs were analysed using standard laboratory methods including antibiotic resistance gene, mecA. Anonymised sociodemographic data were collected by pretested questionnaires to evaluate HCWs factors associated with MRSA carriage. Results: A total of 588 HCWs and 468 administrative staff were screened for MRSA. Women were over-represented. Overall, 49.1% (289/588) of HCWs were nurses and 25% (117/468) of the administrative staff were cleaners or laundry workers. Overall, 138 S. aureus isolates were retrieved from the nasal swabs of both groups (16.3%, 96/588 from HCWs). The burden of MRSA colonisation was 4.8% (28/580, 95% CI: 3.1-6.5%) among HCWs compared to 0.2% (1/468, 95% CI: 0.18-0.6%) of administrative staff (p value <0.05). The majority of S. aureus and all MRSA isolates were resistant to penicillin. Isolates from HCWs were more resistant to tested antibiotics than administrative staff (P-value <0.05). Conclusion: This is the first report in Ethiopia on MRSA colonization using mecA and revealed that; (i) overall carriage rates of MRSA in HCWs are comparable with observations reported in some other countries and (ii) HCWs exhibit a higher burden of MRSA carriage than administrative staff. Our data support strategic screening of MRSA and antimicrobial stewardship for better intervention measures.

Machine learning-based risk factor analysis and prevalence prediction of intestinal parasitic infections using epidemiological survey data.

BACKGROUND: Previous epidemiological studies have examined the prevalence and risk factors for a variety of parasitic illnesses, including protozoan and soil-transmitted helminth (STH, e.g., hookworms and roundworms) infections. Despite advancements in machine learning for data analysis, the majority of these studies use traditional logistic regression to identify significant risk factors. METHODS: In this study, we used data from a survey of 54 risk factors for intestinal parasitosis in 954 Ethiopian school children. We investigated whether machine learning approaches can supplement traditional logistic regression in identifying intestinal parasite infection risk factors. We used feature selection methods such as InfoGain (IG), ReliefF (ReF), Joint Mutual Information (JMI), and Minimum Redundancy Maximum Relevance (MRMR). Additionally, we predicted children's parasitic infection status using classifiers such as Logistic Regression (LR), Support Vector Machines (SVM), Random Forests (RF) and XGBoost (XGB), and compared their accuracy and area under the receiver operating characteristic curve (AUROC) scores. For optimal model training, we performed tenfold cross-validation and tuned the classifier hyperparameters. We balanced our dataset using the Synthetic Minority Oversampling (SMOTE) method. Additionally, we used association rule learning to establish a link between risk factors and parasitic infections. KEY FINDINGS: Our study demonstrated that machine learning could be used in conjunction with logistic regression. Using machine learning, we developed models that accurately predicted four parasitic infections: any parasitic infection at 79.9% accuracy, helminth infection at 84.9%, any STH infection at 95.9%, and protozoan infection at 94.2%. The Random Forests (RF) and Support Vector Machines (SVM) classifiers achieved the highest accuracy when top 20 risk factors were considered using Joint Mutual Information (JMI) or all features were used. The best predictors of infection were socioeconomic, demographic, and hematological characteristics. CONCLUSIONS: We demonstrated that feature selection and association rule learning are useful strategies for detecting risk factors for parasite infection. Additionally, we showed that advanced classifiers might be utilized to predict children's parasitic infection status. When combined with standard logistic regression models, machine learning techniques can identify novel risk factors and predict infection risk.

Utility of line probe assay in detecting drug resistance and the associated mutations in patients with extrapulmonary tuberculosis in Addis Ababa, Ethiopia.

Introduction: Molecular tests allow rapid detection of Mycobacterium tuberculosis and drug resistance in a few days. Identifying the mutations in genes associated with drug resistance may contribute to the development of appropriate interventions to improve tuberculosis control. So far, there is little information in Ethiopia about the diagnostic performance of line probe assay (LPA) and the M. tuberculosis common gene mutations associated with drug resistance in extrapulmonary tuberculosis. Thus, this study aimed to assess the frequency of drug resistance-associated mutations in patients with extrapulmonary tuberculosis (EPTB) and to compare the agreement and determine the utility of the genotypic in the detection of drug resistance in Addis Ababa, Ethiopia. Methods: A cross-sectional study was conducted on stored M. tuberculosis isolates. The genotypic and phenotypic drug susceptibility tests were performed using LPA and BACTEC-MGIT-960, respectively. The common mutations were noted, and the agreement and the utility of the LPA were determined using the BACTEC-MGIT-960 as a gold standard. Results: Of the 151 isolates, the sensitivity and specificity of MTBDRplus in detecting isoniazid resistance were 90.9% and 100%, respectively. While for rifampicin, it was 100% and 99.3% for sensitivity and specificity, respectively. The katG S315Tl was the most common mutation observed in 85.7% of the isoniazid-resistant isolates. In the case of rifampicin, the most common mutation (61.9%) was observed at position rpoB S531L. Mutations in the gyrA promoter region were strongly associated with Levofloxacin and Moxifloxacin resistance. Conclusion: Line probe assay has high test performance in detecting resistance to anti-TB drugs in EPTB isolates. The MTBDRplus test was slightly less sensitive for the detection of isoniazid resistance as compared to the detection of rifampicin. The most prevalent mutations associated with isoniazid and rifampicin resistance were observed at katG S315Tl and rpoB S531L respectively. Besides, all the fluoroquinolone-resistant cases were associated with gyrA gene. Finally, a validation study with DNA sequencing is recommended.

Integration of cytopathology with molecular tests to improve the lab diagnosis for TBLN suspected patients.

BACKGROUND: Tuberculosis lymphadenitis (TBLN) diagnosis is often challenging in most resource poor settings. Often cytopathologic diagnosis of TBLN suspected patients is inconclusive impeding timely clinical management of TBLN suspected patients, further exposing suspected patients either for unnecessary use of antibiotics or empirical treatment. This may lead to inappropriate treatment outcome or more suffering of suspected patients from the disease. In this study, an integrated diagnostic approach has been evaluated to elucidate its utility in the identification of TBLN suspected patients. METHODS: A cross-sectional study was conducted on 96 clinically diagnosed TBLN suspected patients, where fine needle aspirate (FNA) samples were collected at the time of diagnosis. FNA cytology, Ziehl-Neelsen (ZN), Auramine O (AO) staining, GeneXpert MTB/RIF and Real time PCR (RT-PCR) were performed on concentrated FNA samples. Considering culture as a gold standard, the sensitivity, specificity, positive and negative predictive values were calculated. Cohen's Kappa value was used to measure interrater variability and level of agreement and a P-value of <0.05 was considered as statistically significant. RESULT: Out of the 96 FNA sample, 12 (12.5%) were identified to have Mycobacterium tuberculosis (Mtb) using ZN staining, 27 (28.1%) using AO staining, 51 (53.2%) using FNAC, 43 (44.7%) using GeneXpert MTB/RIF, 51 (53.1%) using Real time PCR (RT-PCR) and 36 (37.5%) using Lowenstein-Jensen (LJ) culture. Compared to LJ culture, the sensitivities of GeneXpert MTB/RIF, RT-PCR, and FNAC were 91.7%, 97.2%, and 97.2%, respectively and the specificities were 83.3%, 73.3%, and 68.3%, respectively. GeneXpert MTB/RIF and RT-PCR when combined with FNAC detected 61 (63.5%) cases as having Mtb, and the sensitivity and specificity was 100% and 58.3%, respectively. CONCLUSION: FNA cytology and RT-PCR detected more TBLN cases compared to other Mtb detection tools and the detection sensitivity even improved when FNA cytology was combined with GeneXpert MTB/RIF, performed on concentrated FNA sample, suggesting the combined tests as an alternative approach for improved diagnosis of TBLN.

Effect of heat inactivation for the detection of severe acute respiratory syndrome-corona virus-2 (SARS-CoV-2) with reverse transcription real time polymerase chain reaction (rRT-PCR): evidence from Ethiopian study.

BACKGROUND: Coronavirus disease 2019 (COVID-19) has been a major public health importance and its specimen needs to be handled safely due to concerns of potential transmissibility to health care workers. Heat inactivation of the sample before nucleic acid isolation might permit safe testing processes. Hence, it is important to assess the effect of heat inactivation on SARS-CoV-2 RT-PCR detection in resource limited settings. METHODS: An experimental study was conducted at Ethiopian Public Health Institute (EPHI) from September 25 to October 15, 2020. A total of 188 Oro-pharyngeal swabs were collected from COVID-19 suspected cases, referred to EPHI for SARS COV-2 testing. One batch of the sample was inactivated at 56 °C heat for 30 min, and the other batch was stored at 4 °C for a similar period of time. RNA extraction and detection were done by DAAN Gene kit protocols. Abbott m2000 RT-PCR was used for amplification and detection. Data analysis was done by using SPSS version 23.0; Chi-square and Pearson correlation test for qualitative and semi-quantitative analysis were used. p-value < 0.05 was considered as statistically significant. RESULTS: Out of 188 total samples, 119 (63.3%) were positive and 69 (36.7%) were negative in the non-inactivated group. While, 115 (61.2%) of samples were positive and 73 (38.8) were negative in heat inactivated sample batch. Rate of positivity between groups did not have statistically significant difference (p > 0.05). The mean Cycle threshold (Ct) value difference between the two groups of ORF1a/b gene and N gene was 0.042 (95% CI - 0.247-0.331; t = 0.28; p = 0.774) and 0.38 (95% CI 0.097-0.682; t = 2.638; p = 0.010) respectively. CONCLUSION: Heat inactivation at 56 °C for 30 min did not affect the qualitative rRT-PCR detection of SARS-CoV-2. However, the finding showed that there was statistically significant Ct value increment after heat inactivation compared to untreated samples. Therefore, false-negative results for high Ct value (Ct > 35) samples were found to be the challenge of this protocol. Hence alternative inactivation methods should be investigated and further studies should be considered.