-
BMC Public Health Jun 2024Coronavirus disease 2019 (COVID-19), a global public health crisis, continues to pose challenges despite preventive measures. The daily rise in COVID-19 cases is...
BACKGROUND
Coronavirus disease 2019 (COVID-19), a global public health crisis, continues to pose challenges despite preventive measures. The daily rise in COVID-19 cases is concerning, and the testing process is both time-consuming and costly. While several models have been created to predict mortality in COVID-19 patients, only a few have shown sufficient accuracy. Machine learning algorithms offer a promising approach to data-driven prediction of clinical outcomes, surpassing traditional statistical modeling. Leveraging machine learning (ML) algorithms could potentially provide a solution for predicting mortality in hospitalized COVID-19 patients in Ethiopia. Therefore, the aim of this study is to develop and validate machine-learning models for accurately predicting mortality in COVID-19 hospitalized patients in Ethiopia.
METHODS
Our study involved analyzing electronic medical records of COVID-19 patients who were admitted to public hospitals in Ethiopia. Specifically, we developed seven different machine learning models to predict COVID-19 patient mortality. These models included J48 decision tree, random forest (RF), k-nearest neighborhood (k-NN), multi-layer perceptron (MLP), Naïve Bayes (NB), eXtreme gradient boosting (XGBoost), and logistic regression (LR). We then compared the performance of these models using data from a cohort of 696 patients through statistical analysis. To evaluate the effectiveness of the models, we utilized metrics derived from the confusion matrix such as sensitivity, specificity, precision, and receiver operating characteristic (ROC).
RESULTS
The study included a total of 696 patients, with a higher number of females (440 patients, accounting for 63.2%) compared to males. The median age of the participants was 35.0 years old, with an interquartile range of 18-79. After conducting different feature selection procedures, 23 features were examined, and identified as predictors of mortality, and it was determined that gender, Intensive care unit (ICU) admission, and alcohol drinking/addiction were the top three predictors of COVID-19 mortality. On the other hand, loss of smell, loss of taste, and hypertension were identified as the three lowest predictors of COVID-19 mortality. The experimental results revealed that the k-nearest neighbor (k-NN) algorithm outperformed than other machine learning algorithms, achieving an accuracy of 95.25%, sensitivity of 95.30%, precision of 92.7%, specificity of 93.30%, F1 score 93.98% and a receiver operating characteristic (ROC) score of 96.90%. These findings highlight the effectiveness of the k-NN algorithm in predicting COVID-19 outcomes based on the selected features.
CONCLUSION
Our study has developed an innovative model that utilizes hospital data to accurately predict the mortality risk of COVID-19 patients. The main objective of this model is to prioritize early treatment for high-risk patients and optimize strained healthcare systems during the ongoing pandemic. By integrating machine learning with comprehensive hospital databases, our model effectively classifies patients' mortality risk, enabling targeted medical interventions and improved resource management. Among the various methods tested, the K-nearest neighbors (KNN) algorithm demonstrated the highest accuracy, allowing for early identification of high-risk patients. Through KNN feature identification, we identified 23 predictors that significantly contribute to predicting COVID-19 mortality. The top five predictors are gender (female), intensive care unit (ICU) admission, alcohol drinking, smoking, and symptoms of headache and chills. This advancement holds great promise in enhancing healthcare outcomes and decision-making during the pandemic. By providing services and prioritizing patients based on the identified predictors, healthcare facilities and providers can improve the chances of survival for individuals. This model provides valuable insights that can guide healthcare professionals in allocating resources and delivering appropriate care to those at highest risk.
Topics: Humans; COVID-19; Machine Learning; Ethiopia; Male; Female; Middle Aged; Adult; Algorithms; Aged; SARS-CoV-2; Hospitalization; Electronic Health Records; Young Adult; Adolescent
PubMed: 38943093
DOI: 10.1186/s12889-024-19196-0 -
BMC Microbiology Jun 2024Mangroves are complex and dynamic coastal ecosystems under frequent fluctuations in physicochemical conditions related to the tidal regime. The frequent variation in...
BACKGROUND
Mangroves are complex and dynamic coastal ecosystems under frequent fluctuations in physicochemical conditions related to the tidal regime. The frequent variation in organic matter concentration, nutrients, and oxygen availability, among other factors, drives the microbial community composition, favoring syntrophic populations harboring a rich and diverse, stress-driven metabolism. Mangroves are known for their carbon sequestration capability, and their complex and integrated metabolic activity is essential to global biogeochemical cycling. Here, we present a metabolic reconstruction based on the genomic functional capability and flux profile between sympatric MAGs co-assembled from a tropical restored mangrove.
RESULTS
Eleven MAGs were assigned to six Bacteria phyla, all distantly related to the available reference genomes. The metabolic reconstruction showed several potential coupling points and shortcuts between complementary routes and predicted syntrophic interactions. Two metabolic scenarios were drawn: a heterotrophic scenario with plenty of carbon sources and an autotrophic scenario with limited carbon sources or under inhibitory conditions. The sulfur cycle was dominant over methane and the major pathways identified were acetate oxidation coupled to sulfate reduction, heterotrophic acetogenesis coupled to carbohydrate catabolism, ethanol production and carbon fixation. Interestingly, several gene sets and metabolic routes similar to those described for wastewater and organic effluent treatment processes were identified.
CONCLUSION
The mangrove microbial community metabolic reconstruction reflected the flexibility required to survive in fluctuating environments as the microhabitats created by the tidal regime in mangrove sediments. The metabolic components related to wastewater and organic effluent treatment processes identified strongly suggest that mangrove microbial communities could represent a resourceful microbial model for biotechnological applications that occur naturally in the environment.
Topics: Microbiota; Bacteria; Wetlands; Phylogeny; Heterotrophic Processes; Carbon Cycle; Carbon; Methane; Autotrophic Processes; Metabolic Networks and Pathways
PubMed: 38943070
DOI: 10.1186/s12866-024-03390-6 -
AMB Express Jun 2024Rhizopus arrhizus is a saprotrophic, sometimes clinically- and industrially-relevant mold (Mucorales) and distributed worldwide, suggesting it can assimilate a broad...
Rhizopus arrhizus is a saprotrophic, sometimes clinically- and industrially-relevant mold (Mucorales) and distributed worldwide, suggesting it can assimilate a broad spectrum of substrates. Here, 69 strains of R. arrhizus were investigated by using the Biolog FF MicroPlate for the profiles of utilizing 95 carbon and nitrogen substrates. The study showed that most R. arrhizus strains were similar in average well color development (AWCD) and substrate richness (SR). Nevertheless, 13 strains were unique in principal component analyses, heatmap, AWCD, and SR analyses, which may imply a niche differentiation within R. arrhizus. The species R. arrhizus was able to utilize all the 95 carbon and nitrogen substrates, consistent with the hypothesis of a great metabolic diversity. It possessed a substrate preference of alcohols, and seven substrates were most frequently utilized, with N-acetyl-D-galactosamine and L-phenylalanine ranking at the top of the list. Eight substrates, especially L-arabinose and xylitol, were capable of promoting sporulation and being applied for rejuvenating degenerated strains. By phenotyping R. arrhizus strains in carbon and nitrogen assimilation capacity, this study revealed the extent of intra-specific variability and laid a foundation for estimating optimum substrates that may be useful for industrial applications.
PubMed: 38942930
DOI: 10.1186/s13568-024-01733-0 -
Scientific Reports Jun 2024This work presents experimental results on the energy efficiency in hydrogen production using atmospheric microwave plasma (915 MHz) through steam reforming of ethanol....
This work presents experimental results on the energy efficiency in hydrogen production using atmospheric microwave plasma (915 MHz) through steam reforming of ethanol. Ethanol was chosen as a liquid hydrogen carrier due to its high hydrogen atom content, low cost, and wide availability. The experimental work began with the maximization of an energy efficiency of the used microwave plasma source. The process of maximization involved determining a position of a movable plunger that ensures the most efficient transfer of microwave energy from a microwave source to the generated plasma in the microwave plasma source. The aim of the investigations was to test the following working conditions of the microwave plasma source: absorbed microwave power P by the generated plasma (up to 5.4 kW), the carrier gas volumetric flow rate (up to 3900 Nl/h), and the amount of the introduced ethanol vapours on the efficiency of hydrogen production (up to 2.4 kg/h). In the range of tested working conditions, the highest energy yield for hydrogen production achieved a rate of 26.9 g(H)/kWh, while the highest hydrogen production was 99.3 g(H)/h.
PubMed: 38942801
DOI: 10.1038/s41598-024-65874-9 -
Scientific Reports Jun 2024HIV/AIDS is one of the most devastating infectious diseases affecting humankind all over the world and its impact goes beyond public health problems. This study was...
Determinants of hemoglobin level and time to default from Highly Active Antiretroviral Therapy (HAART) for adult clients living with HIV under treatment; a retrospective cohort study design.
HIV/AIDS is one of the most devastating infectious diseases affecting humankind all over the world and its impact goes beyond public health problems. This study was conducted to investigate the joint predictors of hemoglobin level and time to default from treatment for adult clients living with HIV/AIDS under HAART at the University of Gondar Comprehensive and Specialized Hospital, North-west Ethiopia. The study was conducted using a retrospective cohort design from the medical records of 403 randomly selected adult clients living with HIV whose follow-ups were from September 2015 to March 2022. Hemoglobin level was projected using Sahli's acid-hematin method. Hence, the hemoglobin tube was filled with N/10 hydrochloric acid up to 2 g % marking and the graduated tube was placed in Sahli's hemoglobin meter. The blood samples were collected using the finger-pick method, considering 22 G disposable needles. The health staff did this. From a total of 403 adult patients living with HIV/AIDS included in the current study, about 44.2% defaulted from therapy. The overall mean and median estimated survival time of adult clients under study were 44.3 and 42 months respectively. The patient's lymphocyte count (AHR = 0.7498, 95% CI: (0.7411: 0.7587), p-value < 0.01), The weight of adult patients living with HIV/AIDS (AHR = 0.9741, 95% CI: (0.9736: 0.9747), p-value = 0.012), sex of adult clients (AHR = 0.6019, 95% CI: (0.5979, 0.6059), p-value < 0.01), WHO stages III compared to Stage I (AHR = 1.4073, 95% CI: (1.3262, 1.5078), p-value < 0.01), poor adherence level (AHR = 0.2796, 95% CI: (0.2082, 0.3705) and p-value < 0.01), bedridden patients (AHR = 1.5346, 95% CI: (1.4199, 1.6495), p-value = 0.008), and opportunistic infections (AHR = 0.2237, 95% CI: (0.0248, 0.4740), p-value = 0.004) had significant effect on both hemoglobin level and time to default from treatment. Similarly, other co-morbidity conditions, disclosure status of the HIV disease, and tobacco and alcohol addiction had a significant effect on the variables of interest. The estimate of the association parameter in the slope value of Hgb level and time default was negative, indicating that the Hgb level increased as the hazard of defaulting from treatment decreased. A patient with abnormal BMI like underweight, overweight, or obese was negatively associated with the risk of anemia (lower hemoglobin level). As a recommendation, more attention should be given to those patients with abnormal BMI, patients with other co-morbidity conditions, patients with opportunistic infections, and low lymphocytes, and bedridden and ambulatory patients. Health-related education should be given to adult clients living with HIV/AIDS to be good adherents for medical treatment.
Topics: Humans; Male; Adult; Female; Retrospective Studies; HIV Infections; Antiretroviral Therapy, Highly Active; Hemoglobins; Middle Aged; Ethiopia; Young Adult; Anti-HIV Agents; Undertreatment
PubMed: 38942753
DOI: 10.1038/s41598-024-62952-w -
The Lancet. Public Health Jul 2024Alcohol container labels might reduce population-level alcohol-related harms, but investigations of their effectiveness have varied in approach and quality. A systematic... (Review)
Review
Alcohol container labels might reduce population-level alcohol-related harms, but investigations of their effectiveness have varied in approach and quality. A systematic synthesis is needed to adjust for these differences and to yield evidence to inform policy. Our objectives were to establish the effects of alcohol container labels bearing one or more health warnings, standard drink information, or low-risk drinking guidance on alcohol consumption behaviour, knowledge of label message, and support for labels. We completed a systematic review according to Cochrane and synthesis without meta-analysis standards. We included all peer-reviewed studies and grey literature published from Jan 1, 1989, to March 6, 2024, in English, French, German, or Spanish that investigated the effects of alcohol container labels compared with no-label or existing label control groups in any population on alcohol consumption behaviour, knowledge of label message, or support for labels. Data were synthesised narratively as impact statements and assessed for risk of bias and certainty in the evidence. A protocol was preregistered (PROSPERO CRD42020168240). We identified 40 publications that studied 31 labels and generated 17 impact statements. 24 (60%) of 40 publications focused on consumption behaviour and we had low or very low certainty in 13 (59%) of 22 outcomes. Alcohol container labels bearing health warnings might slow the rate of alcohol consumption (low certainty), reduce alcoholic beverage selection (moderate certainty), reduce consumption during pregnancy (low certainty), and reduce consumption before driving (moderate certainty). Interventions with multiple types of rotating alcohol container labels likely substantially decrease alcohol use (moderate certainty) and reduce alcohol sales (high certainty). To the best of our knowledge, this is the first systematic review on multiple types of alcohol container labels assessing their effects for certainty in the evidence. Limitations included heterogeneity in label designs and outcome measurements. Alcohol container labels probably influence some alcohol consumption behaviour, with multiple rotating messages being particularly effective, although effects might vary depending on individual health literacy or drinking behaviour. Alcohol container labels might therefore be effective components of policies designed to address population-level alcohol-related harms.
Topics: Humans; Alcohol Drinking; Alcoholic Beverages; Product Labeling; Health Knowledge, Attitudes, Practice
PubMed: 38942559
DOI: 10.1016/S2468-2667(24)00097-5 -
The Lancet. Public Health Jul 2024Alcohol health-warning labels are a policy option that can contribute to the reduction of alcohol-related harms, but their effects and public perception depend on their...
Effect of alcohol health warning labels on knowledge related to the ill effects of alcohol on cancer risk and their public perceptions in 14 European countries: an online survey experiment.
BACKGROUND
Alcohol health-warning labels are a policy option that can contribute to the reduction of alcohol-related harms, but their effects and public perception depend on their content and format. Our study aimed to investigate the effect of health warnings on knowledge that alcohol causes cancer, the perceptions of three different message topics (responsible drinking, general health harm of alcohol, and alcohol causing cancer), and the role of images included with the cancer message.
METHODS
In this online survey experiment, distributed in 14 European countries and targeting adults of the legal alcohol-purchase age who consumed alcohol, participants were randomly allocated to one of six label conditions using a pseudorandom number generator stratified by survey language before completing a questionnaire with items measuring knowledge and label perceptions. Effect on knowledge was assessed as a primary outcome by comparing participants who had increased knowledge after exposure to labels with the rest of the sample, for the six label conditions. Label perceptions were compared between label conditions as secondary outcomes.
FINDINGS
19 110 participants completed the survey and were eligible for analysis. Our results showed that a third of the participants exposed to the cancer message increased their knowledge of alcohol causing cancer (increase for 1131 [32·5%, 95% CI 29·8 to 35·2] of 3409 participants [weighted percentage] for text-only message; increase for 1096 [33·3%, 30·4 to 36·2] of 3198 [weighted percentage] for message inlcuding pictogram; and increase for 1030 [32·5%, 29·6 to 35·4] of 3242 [weighted percentage] for message including graphic image), compared with an increase for 76 (2·4%, -1·2 to 6·0) of 3018 participants who viewed the control message. Logistic regression showed that cancer messages increased knowledge compared with the control label (odds ratio [OR] 20·20, 95% CI 15·88 to 26·12; OR 21·16, 16·62 to 27·38; OR 20·61, 16·19 to 26·68). Cancer messages had the highest perceived impact and relevance, followed by general health harm and responsibility messages. Text-only and pictogram cancer messages were seen as clear, comprehensive, and acceptable, whereas those including an image of a patient with cancer had lower acceptability and the highest avoidance rating of all the labels. The only identified interaction between perceptions and experimental conditions (with gender) indicated higher comprehensibility and acceptability ratings of cancer labels than responsibility messages and control labels by women, with the results reversed in men.
INTERPRETATION
Health warnings are an effective policy option to increase knowledge of alcohol causing cancer, with a generalisable effect across several countries. Europeans consider alcohol health-warning labels to be comprehensible and acceptable, with cancer-specific health warnings having the highest perceived impact and relevance.
FUNDING
EU4Health.
Topics: Humans; Europe; Male; Female; Health Knowledge, Attitudes, Practice; Adult; Neoplasms; Middle Aged; Product Labeling; Surveys and Questionnaires; Young Adult; Alcohol Drinking; Adolescent; Public Opinion; Alcoholic Beverages; Aged
PubMed: 38942558
DOI: 10.1016/S2468-2667(24)00102-6 -
The Lancet. Public Health Jul 2024Even though alcohol consumption is an established risk factor for cancer, evidence regarding the effect of a reduction or cessation of alcohol consumption on cancer...
BACKGROUND
Even though alcohol consumption is an established risk factor for cancer, evidence regarding the effect of a reduction or cessation of alcohol consumption on cancer incidence is scarce. Our main study aim was to assess the effect of alcohol rehabilitation and abstinence on cancer incidence in people with alcohol dependence.
METHODS
We conducted a nationwide hospital retrospective cohort study which included all adults residing in mainland France and discharged in 2018-21. Multivariable Cox proportional hazards models were used to estimate the effect of rehabilitation treatment at hospital or a history of abstinence versus alcohol dependence without rehabilitation or abstinence on the risk for incident alcohol-associated cancers by sex, controlled for potential confounding risk factors.
FINDINGS
10 260 056 men and 13 739 369 women were discharged from French hospitals in 2018-21. Alcohol dependence was identified in 645 720 (6·3%) men and 219 323 (1·6%) women. Alcohol dependence was strongly related to alcohol-associated cancer sites in both sexes (hepatocellular carcinoma and oral, pharyngeal, laryngeal, oesophageal, and colorectal cancers), except for breast cancer. Rehabilitation treatment or abstinence was associated with significantly lower risks compared with alcohol dependence without rehabilitation or abstinence (adjusted hazard ratios: 0·58, 99·89% CI 0·56-0·60 in men and 0·62, 0·57-0·66 in women). Relative risk reductions were significant for each alcohol-associated cancer site in both sexes and supported by all subgroup and sensitivity analyses.
INTERPRETATION
Our study results support the clear benefits of alcohol rehabilitation and abstinence in reducing the risk for alcohol-associated cancers. As only two in five patients with alcohol dependence were recorded with a history of rehabilitation treatment or abstinence, a large untapped potential exists for reducing cancer incidence.
FUNDING
European Union's EU4Health programme.
Topics: Humans; Male; Female; France; Neoplasms; Middle Aged; Retrospective Studies; Alcoholism; Adult; Aged; Risk Factors; Incidence
PubMed: 38942557
DOI: 10.1016/S2468-2667(24)00107-5 -
The Lancet. Public Health Jul 2024
Topics: Humans; France; Alcohol Drinking
PubMed: 38942549
DOI: 10.1016/S2468-2667(24)00124-5 -
The Lancet. Public Health Jul 2024
Topics: Humans; Prisoners; Prisons; Mortality
PubMed: 38942547
DOI: 10.1016/S2468-2667(24)00105-1