Applicability of machine learning algorithm to predict the therapeutic intervention success in Brazilian smokers.

Smoking cessation is an important public health policy worldwide. However, as far as we know, there is a lack of screening of variables related to the success of therapeutic intervention (STI) in Brazilian smokers by machine learning (ML) algorithms. To address this gap in the literature, we evaluated the ability of eight ML algorithms to correctly predict the STI in Brazilian smokers who were treated at a smoking cessation program in Brazil between 2006 and 2017. The dataset was composed of 12 variables and the efficacies of the algorithms were measured by accuracy, sensitivity, specificity, positive predictive value (PPV) and area under the receiver operating characteristic curve. We plotted a decision tree flowchart and also measured the odds ratio (OR) between each independent variable and the outcome, and the importance of the variable for the best model based on PPV. The mean global values for the metrics described above were, respectively, 0.675±0.028, 0.803±0.078, 0.485±0.146, 0.705±0.035 and 0.680±0.033. Supporting vector machines performed the best algorithm with a PPV of 0.726±0.031. Smoking cessation drug use was the roof of decision tree with OR of 4.42 and importance of variable of 100.00. Increase in the number of relapses also promoted a positive outcome, while higher consumption of cigarettes resulted in the opposite. In summary, the best model predicted 72.6% of positive outcomes correctly. Smoking cessation drug use and higher number of relapses contributed to quit smoking, while higher consumption of cigarettes showed the opposite effect. There are important strategies to reduce the number of smokers and increase STI by increasing services and drug treatment for smokers.

Machine learning models to predict traumatic brain injury outcomes in Tanzania: Using delays to emergency care as predictors.

Constraints to emergency department resources may prevent the timely provision of care following a patient's arrival to the hospital. In-hospital delays may adversely affect health outcomes, particularly among trauma patients who require prompt management. Prognostic models can help optimize resource allocation thereby reducing in-hospital delays and improving trauma outcomes. The objective of this study was to investigate the predictive value of delays to emergency care in machine learning based traumatic brain injury (TBI) prognostic models. Our data source was a TBI registry from Kilimanjaro Christian Medical Centre Emergency Department in Moshi, Tanzania. We created twelve unique variables representing delays to emergency care and included them in eight different machine learning based TBI prognostic models that predict in-hospital outcome. Model performance was compared using the area under the receiver operating characteristic curve (AUC). Inclusion of our twelve time to care variables improved predictability in each of our eight prognostic models. Our Bayesian generalized linear model produced the largest AUC, with a value of 89.5 (95% CI: 88.8, 90.3). Time to care variables were among the most important predictors of in-hospital outcome in our best three performing models. In low-resource settings where delays to care are highly prevalent and contribute to high mortality rates, incorporation of care delays into prediction models that support clinical decision making may benefit both emergency medicine physicians and trauma patients by improving prognostication performance.

Applicability of machine learning technique in the screening of patients with mild traumatic brain injury.

Even though the demand of head computed tomography (CT) in patients with mild traumatic brain injury (TBI) has progressively increased worldwide, only a small number of individuals have intracranial lesions that require neurosurgical intervention. As such, this study aims to evaluate the applicability of a machine learning (ML) technique in the screening of patients with mild TBI in the Regional University Hospital of Maringá, Paraná state, Brazil. This is an observational, descriptive, cross-sectional, and retrospective study using ML technique to develop a protocol that predicts which patients with an initial diagnosis of mild TBI should be recommended for a head CT. Among the tested models, he linear extreme gradient boosting was the best algorithm, with the highest sensitivity (0.70 ± 0.06). Our predictive model can assist in the screening of mild TBI patients, assisting health professionals to manage the resource utilization, and improve the quality and safety of patient care.

Antivenom access impacts severity of Brazilian snakebite envenoming: A geographic information system analysis.

BACKGROUND: Snakebite envenoming (SBE) is a neglected tropical disease capable of causing both significant disability and death. The burden of SBE is especially high in low- and middle-income countries. The aim of this study was to perform a geospatial analysis evaluating the association of sociodemographics and access to care indicators on moderate and severe cases of SBE in Brazil. METHODS: We conducted an ecological, cross-sectional study of SBE in Brazil from 2014 to 2019 using the open access National System Identification of Notifiable Diseases (SINAN) database. We then collected a set of indicators from the Brazil Census of 2010 and performed a Principal Component Analysis to create variables related to health, economics, occupation, education, infrastructure, and access to care. Next, a descriptive and exploratory spatial analysis was conducted to evaluate the geospatial association of moderate and severe events. These variables related to events were evaluated using Geographically Weighted Poisson Regression. T-values were plotted in choropleth maps and considered statistically significant when values were <-1.96 or >+1.96. RESULTS: We found that the North region had the highest number of SBE cases by population (47.83/100,000), death rates (0.18/100,000), moderate and severe rates (22.96/100,000), and proportion of cases that took more than three hours to reach healthcare assistance (44.11%). The Northeast and Midwest had the next poorest indicators. Life expectancy, young population structure, inequality, electricity, occupation, and more than three hours to reach healthcare were positively associated with greater cases of moderate and severe events, while income, illiteracy, sanitation, and access to care were negatively associated. The remaining indicators showed a positive association in some areas of the country and a negative association in other areas. CONCLUSION: Regional disparities in SBE incidence and rates of poor outcomes exist in Brazil, with the North region disproportionately affected. Multiple indicators were associated with rates of moderate and severe events, such as sociodemographic and health care indicators. Any approach to improving snakebite care must work to ensure the timeliness of antivenom administration.

Using the Pan American Health Organization Digital Conversational Agent to Educate the Public on Alcohol Use and Health: Preliminary Analysis.

BACKGROUND: There is widespread misinformation about the effects of alcohol consumption on health, which was amplified during the COVID-19 pandemic through social media and internet channels. Chatbots and conversational agents became an important piece of the World Health Organization (WHO) response during the COVID-19 pandemic to quickly disseminate evidence-based information related to COVID-19 and tobacco to the public. The Pan American Health Organization (PAHO) seized the opportunity to develop a conversational agent to talk about alcohol-related topics and therefore complement traditional forms of health education that have been promoted in the past. OBJECTIVE: This study aimed to develop and deploy a digital conversational agent to interact with an unlimited number of users anonymously, 24 hours a day, about alcohol topics, including ways to reduce risks from drinking, that is accessible in several languages, at no cost, and through various devices. METHODS: The content development was based on the latest scientific evidence on the impacts of alcohol on health, social norms about drinking, and data from the WHO and PAHO. The agent itself was developed through a nonexclusive license agreement with a private company (Soul Machines) and included Google Digital Flow ES as the natural language processing software and Amazon Web Services for cloud services. Another company was contracted to program all the conversations, following the technical advice of PAHO staff. RESULTS: The conversational agent was named Pahola, and it was deployed on November 19, 2021, through the PAHO website after a launch event with high publicity. No identifiable data were used and all interactions were anonymous, and therefore, this was not considered research with human subjects. Pahola speaks in English, Spanish, and Portuguese and interacts anonymously with a potentially infinite number of users through various digital devices. Users were required to accept the terms and conditions to enable access to their camera and microphone to interact with Pahola. Pahola attracted good attention from the media and reached 1.6 million people, leading to 236,000 clicks on its landing page, mostly through mobile devices. Only 1532 users had a conversation after clicking to talk to Pahola. The average time users spent talking to Pahola was 5 minutes. Major dropouts were observed in different steps of the conversation flow. Some questions asked by users were not anticipated during programming and could not be answered. CONCLUSIONS: Our findings showed several limitations to using a conversational agent for alcohol education to the general public. Improvements are needed to expand the content to make it more meaningful and engaging to the public. The potential of chatbots to educate the public on alcohol-related topics seems enormous but requires a long-term investment of resources and research to be useful and reach many more people.

The development of the Pan American Health Organization digital health specialist on alcohol use.

Introduction: On 19 November 2021 the Pan American Health Organization (PAHO) developed and deployed the first-ever digital health worker dedicated to alcohol-related topics, named Pahola. This paper describes this developmental process and the first results of its uptake and interactions with the public. Methods: PAHO secured a non-exclusive worldwide license with a technology company to use their Human OS ecosystem, which enables human-like interactions between digital people and users via an application. Google Digital flow ES was used to develop the conversations of Pahola on topics related to alcohol and health, screening of alcohol risk using the AUDIT and providing a quit/cut back plan to users, along with additional treatment services and resources in each country of the Americas. A communication campaign was also implemented from launching date until 31 December 2021. Results: Pahola attracted good attention from the media, and potentially reached 1.6 million people, leading to 236,000 sessions on its landing page, mostly through mobile devices. The average time people effectively spent talking to Pahola was five minutes. Major dropouts were observed in different steps of the conversation flow. Discussion: Pahola was quickly able to connect to a large worldwide population with reliable alcohol information. It could potentially increase the delivery of SBI and improve alcohol health literacy. However, its preliminary results pointed to much needed changes to its corpus and on its accessibility, which are being currently implemented.

Corticosteroid Randomization after Significant Head Injury and International Mission for Prognosis and Clinical Trials in Traumatic Brain Injury Models Compared with a Machine Learning-Based Predictive Model from Tanzania.

Hospitals in low- and middle-income countries (LMICs) could benefit from decision support technologies to reduce time to triage, diagnosis, and surgery for patients with traumatic brain injury (TBI). Corticosteroid Randomization after Significant Head Injury (CRASH) and International Mission for Prognosis and Clinical Trials in Traumatic Brain Injury (IMPACT) are robust examples of TBI prognostic models, although they have yet to be validated in Sub-Saharan Africa (SSA). Moreover, machine learning and improved data quality in LMICs provide an opportunity to develop context-specific, and potentially more accurate, prognostic models. We aim to externally validate CRASH and IMPACT on our TBI registry and compare their performances to that of the locally derived model (from the Kilimanjaro Christian Medical Center [KCMC]). We developed a machine learning-based prognostic model from a TBI registry collected at a regional referral hospital in Moshi, Tanzania. We also used the core CRASH and IMPACT online risk calculators to generate risk scores for each patient. We compared the discrimination (area under the curve [AUC]) and calibration before and after Platt scaling (Brier, Hosmer-Lemeshow Test, and calibration plots) for CRASH, IMPACT, and the KCMC model. The outcome of interest was unfavorable in-hospital outcome defined as a Glasgow Outcome Scale score of 1-3. There were 2972 patients included in the TBI registry, of whom 11% had an unfavorable outcome. The AUCs for the KCMC model, CRASH, and IMPACT were 0.919, 0.876, and 0.821, respectively. Prior to Platt scaling, CRASH was the best calibrated model (χ2 = 68.1) followed by IMPACT (χ2 = 380.9) and KCMC (χ2 = 1025.6). We provide the first SSA validation of the core CRASH and IMPACT models. The KCMC model had better discrimination than either of these. CRASH had the best calibration, although all model predictions could be successfully calibrated. The top performing models, KCMC and CRASH, were both developed using LMIC data, suggesting that locally derived models may outperform imported ones from different contexts of care. Further work is needed to externally validate the KCMC model.

Emergency Care Gap in Brazil: Geographical Accessibility as a Proxy of Response Capacity to Tackle COVID-19.

Background: The new coronavirus disease (COVID-19) has claimed thousands of lives worldwide and disrupted the health system in many countries. As the national emergency care capacity is a crucial part of the COVID-19 response, we evaluated the Brazilian Health Care System response preparedness against the COVID-19 pandemic. Methods: A retrospective and ecological study was performed with data retrieved from the Brazilian Information Technology Department of the Public Health Care System. The numbers of intensive care (ICU) and hospital beds, general or intensivist physicians, nurses, nursing technicians, physiotherapists, and ventilators from each health region were extracted. Beds per health professionals and ventilators per population rates were assessed. A health service accessibility index was created using a two-step floating catchment area (2SFCA). A spatial analysis using Getis-Ord Gi* was performed to identify areas lacking access to high-complexity centers (HCC). Results: As of February 2020, Brazil had 35,682 ICU beds, 426,388 hospital beds, and 65,411 ventilators. In addition, 17,240 new ICU beds were created in June 2020. The South and Southeast regions have the highest rates of professionals and infrastructure to attend patients with COVID-19 compared with the northern region. The north region has the lowest accessibility to ICUs. Conclusions: The Brazilian Health Care System is unevenly distributed across the country. The inequitable distribution of health facilities, equipment, and human resources led to inadequate preparedness to manage the COVID-19 pandemic. In addition, the ineffectiveness of public measures of the municipal and federal administrations aggravated the pandemic in Brazil.

Machine Learning and Precision Medicine in Emergency Medicine: The Basics.

As machine learning (ML) and precision medicine become more readily available and used in practice, emergency physicians must understand the potential advantages and limitations of the technology. This narrative review focuses on the key components of machine learning, artificial intelligence, and precision medicine in emergency medicine (EM). Based on the content expertise, we identified articles from EM literature. The authors provided a narrative summary of each piece of literature. Next, the authors provided an introduction of the concepts of ML, artificial intelligence as an extension of ML, and precision medicine. This was followed by concrete examples of their applications in practice and research. Subsequently, we shared our thoughts on how to consume the existing research in these subjects and conduct high-quality research for academic emergency medicine. We foresee that the EM community will continue to adapt machine learning, artificial intelligence, and precision medicine in research and practice. We described several key components using our expertise.

Mortality associated with alternative policy options for primary care and the Mais Médicos (More Doctors) Program in Brazil: forecasting future scenarios

[ABSTRACT]. Objective. To forecast the impact of alternative scenarios of coverage changes in Brazil’s Family Health Strategy (Estratégia Saúde da Família) (ESF)—due to fiscal austerity measures and to the end of the Mais Médicos (More Doctors) Program (PMM)—on overall under-5 mortality rates (U5MRs) and under-70 mortality rates (U70MRs) from ambulatory care sensitive conditions (ACSCs) up through 2030. Methods. A synthetic cohort of 5 507 Brazilian municipalities was created for the period 2017-2030. A municipal-level microsimulation model was developed and validated using longitudinal data. Reductions in ESF coverage, and its effects on U5MRs and U70MRs from ACSCs, were forecast based on two probable austerity scenarios, as compared to the maintenance of current ESF coverage. Fixed effects longitudinal regression models were employed to account for secular trends, demographic and socioeconomic changes, variables related to health care, and program duration effects. Results. In comparison to maintaining stable ESF coverage, with the decrease in ESF coverage due to austerity measures and PMM termination, the mean U5MR and U70MR would be 13.2% and 8.6% higher, respectively, in 2030. The end of PMM would be responsible for a mean U5MR from ACSCs that is 4.3% higher and a U70MR from ACSCs that is 2.8% higher in 2030. The reduction of PMM coverage due only to the withdrawal of Cuban doctors who have been working in PMM would alone be responsible for a U5MR that is 3.2% higher, and a U70MR that is 2.0% higher in 2030. Conclusions. Reductions in primary health care coverage due to austerity measures and the end of the PMM could be responsible for many avoidable adult and child deaths in coming years in Brazil.

[RESUMEN]. Objetivo. Proyectar el impacto de los distintos escenarios alternativos de cambios en la cobertura de la estrategia de salud familiar de Brasil (Estratégia Saúde da Família o ESF) —motivados por medidas de austeridad y la desaparición del programa Mais Médicos (PMM)— en las tasas generales de mortalidad de menores de 5 años y menores de 70 años debidas a trastornos sensibles al cuidado ambulatorio hasta el año 2030. Métodos. Se formó una cohorte sintética de 5 507 municipios brasileños para el período 2017-2030. Se elaboró un modelo de microsimulación a escala municipal y se lo validó empleando datos longitudinales. Se proyectaron las reducciones de la cobertura de la ESF y sus efectos sobre la tasa de mortalidad de menores de 5 años y menores de 70 años debida a trastornos sensibles al cuidado ambulatorio con base en dos contextos probables de austeridad, en comparación con el mantenimiento de la actual cobertura de la ESF. Se emplearon modelos de regresión longitudinal con efectos fijos para dar cuenta de las tendencias históricas, los cambios demográficos y socioeconómicos, las variables relacionadas con la atención de salud y los efectos de la duración del programa. Resultados. En comparación con el mantenimiento de una cobertura estable de la ESF, ante su disminución por medidas de austeridad y la desaparición del PMM, las tasas medias de mortalidad de menores de 5 años y menores de 70 aumentarían en 13,2 % y 8,6 % respectivamente para el año 2030. La desaparición del PMM sería responsable de una tasa media de mortalidad de menores de 5 años debida a trastornos sensibles al cuidado ambulatorio que será un 4,3 % mayor y, en el caso de los menores de 70 años, un 2,8 % mayor para el año 2030. Tan solo la reducción de la cobertura de PMM exclusivamente a raíz de la retirada de los médicos cubanos que han trabajado en este programa daría cuenta de un incremento del 3,2 % de la tasa de mortalidad de menores de 5 años y del 2,0 % en el caso de los menores de 70 años para el año 2030. Conclusiones. Las reducciones de la cobertura de la atención primaria de salud debidas a medidas de austeridad y la desaparición del PMM serían responsables de muchas muertes evitables de niños y adultos en los próximos años en Brasil.

[RESUMO]. Objetivo. Fazer uma projeção da repercussão de cenários alternativos, com a mudança na cobertura da Estratégia de Saúde da Família (ESF) no Brasil decorrente de medidas de austeridade fiscal e do fim do Programa Mais Médicos, nas taxas de mortalidade em crianças menores de 5 anos (TM-5) e taxas de mortalidade em indivíduos até 70 anos (TM-70) por causas sensíveis à atenção ambulatorial até 2030. Métodos. Esta análise se baseou em uma coorte sintética de 5.507 municípios brasileiros criada para o período 2017–2030. Um modelo de microssimulação ao nível municipal foi desenvolvido e validado com dados longitudinais. A diminuição da cobertura da ESF e sua repercussão nas TM-5 e TM-70 por causas sensíveis à atenção ambulatorial foram projetadas em dois cenários prováveis de austeridade comparados à continuidade da cobertura atual da ESF. Modelos de regressão com efeitos fixos para dados longitudinais foram usados para levar em consideração as tendências seculares, as variações populacionais e socioeconômicas, as variáveis relacionadas à assistência de saúde e os efeitos da continuidade do programa. Resultados. Comparando-se à continuidade da cobertura estável da ESF, com a diminuição da cobertura decorrente de medidas de austeridade e do fim do Programa Mais Médicos, as TM-5 e TM-70 médias seriam 13,2% e 8,6% maiores em 2030. O fim do Programa Mais Médicos resultaria em um aumento de 4,3% na TM-5 média e de 2,8% na TM-70 média por causas sensíveis à atenção ambulatorial em 2030. A diminuição da cobertura do Programa Mais Médicos decorrente exclusivamente da saída dos médicos cubanos do programa estaria associada a um aumento de 3,2% na TM-5 e de 2,0% na TM-70 em 2030. Conclusões. A diminuição na cobertura da atenção primária à saúde decorrente de medidas de austeridade e do fim do Programa Mais Médicos teria como resultado muitas mortes evitáveis em adultos e crianças no Brasil nos anos que estão por vir.

Addressing inequalities in medical workforce distribution: evidence from a quasi-experimental study in Brazil.

BACKGROUND: Brazil faces huge health inequality challenges since not all municipalities have access to primary care physicians. The More Doctors Programme (MDP), which started in 2013, was born out of this recognition, providing more than 18 000 doctors in the first few years. However, the programme faced a restructuring at the end of 2018. METHODS: We construct a panel municipality-level data between 2008 and 2017 for 5570 municipalities in Brazil. We employ a difference-in-differences empirical approach, combined with propensity score matching, to study the impacts of the programme on hospitalisations for ambulatory care sensitive conditions and its costs. We explore heterogeneous impacts by age of the patients, type of admissions, and municipalities that were given priority. FINDINGS: The MDP reduced ambulatory admissions by 2.9 per cent (p value <0.10) and the costs by 3.7 per cent (p value <0.01) over the mean. The reduction was driven by infectious gastroenteritis, bacterial pneumonias, asthma, kidney and urinary infections, and pelvic inflammatory disease. The results held on the subsample of municipalities targeted by the programme. By comparing the benefits of the programme from the reduction in the costs of ambulatory admissions to the total financial costs of the MDP, the impacts allowed the government to save at least BRL 27.88 (US$ 6.9 million) between 2014 and 2017. CONCLUSION: Addressing inequalities in the distribution of the medical workforce remains a global challenge. Our results inform the discussion on the current strategy adopted in Brazil to increase access to primary healthcare in underserved areas.

Geospatial modeling of microcephaly and zika virus spread patterns in Brazil.

Microcephaly and Zika Virus infection (ZIKV) were declared Public Health Emergencies of International Concern by the World Health Organization in 2016. Brazil was considered the epicenter of the outbreak. However, the occurrence of both ZIKV and microcephaly in Brazil was not evenly distributed across the country. To better understand this phenomenon, we investigate regional characteristics at the municipal level that can be associated with the incidence of microcephaly, our response variable, and its relationship with ZIKV and other predictors. All epidemiological data in this study was provided by the Ministry of Health official database (DATASUS). Microcephaly was only confirmed after birth and the diagnostic was made regardless of the mother's ZIKV status. Using exploratory spatial data analysis and spatial autoregressive Tobit models, our results show that microcephaly incidence is significantly, at 95% confidence level, related not only to ZIKV, but also to access to primary care, population size, gross national product, mobility and environmental attributes of the municipalities. There is also a significant spatial autocorrelation of the dependent variable. The results indicate that municipalities that show a high incidence of microcephaly tend to be clustered in space and that incidence of microcephaly varies considerably across regions when correlated only with ZIKV, i.e. that ZIKV alone cannot explain the differences in microcephaly across regions and their correlation is mediated by regional attributes.

A traumatic brain injury prognostic model to support in-hospital triage in a low-income country: a machine learning-based approach.

OBJECTIVE: Traumatic brain injury (TBI) is a leading cause of death and disability worldwide, with a disproportionate burden of this injury on low- and middle-income countries (LMICs). Limited access to diagnostic technologies and highly skilled providers combined with high patient volumes contributes to poor outcomes in LMICs. Prognostic modeling as a clinical decision support tool, in theory, could optimize the use of existing resources and support timely treatment decisions in LMICs. The objective of this study was to develop a machine learning-based prognostic model using data from Kilimanjaro Christian Medical Centre in Moshi, Tanzania. METHODS: This study is a secondary analysis of a TBI data registry including 3138 patients. The authors tested nine different machine learning techniques to identify the prognostic model with the greatest area under the receiver operating characteristic curve (AUC). Input data included demographics, vital signs, injury type, and treatment received. The outcome variable was the discharge score on the Glasgow Outcome Scale-Extended. RESULTS: The AUC for the prognostic models varied from 66.2% (k-nearest neighbors) to 86.5% (Bayesian generalized linear model). An increasing Glasgow Coma Scale score, increasing pulse oximetry values, and undergoing TBI surgery were predictive of a good recovery, while injuries suffered from a motor vehicle crash and increasing age were predictive of a poor recovery. CONCLUSIONS: The authors developed a TBI prognostic model with a substantial level of accuracy in a low-resource setting. Further research is needed to externally validate the model and test the algorithm as a clinical decision support tool.

Gestão de recursos humanos em saúde e sua necessária interface ao mundo organizacional: um olhar ampliado / Management of human resources in health and its necessary interface to the organizational world: an extended look

A gestão de recursos humanos, envolta em um debate que, de um lado privilegia sua posição estratégica nas organizações e, de outro, se atém à multifacetada composição de suas variáveis ditas operacionais, normalmente expressas nos clássicos subsistemas organizacionais, encontra no campo da saúde um espaço privilegiado de discussão e aplicabilidade. No caso brasileiro, a complexidade do Sistema Único de Saúde, que tem na Atenção Primária à Saúde sua estratégia central e na Saúde da Família sua expressão mais visível de atuação, aumenta sua relevância, considerando o perfil do atendimento priorizado e a necessidade de se alinhar ações públicas àquelas voltadas à prática cotidiana de gerenciamento de pessoas. Nesta perspectiva, o presente ensaio, de natureza teórica, procura ampliar o olhar da gestão de recursos humanos na saúde, expandindo o debate para além de estereótipos usuais neste campo associados a dimensões como capacitação e quantidade de profissionais. Para tanto, foi feita uma reflexão que procura aproximar gestão de recursos humanos e saúde, o que permitiu a sistematização de múltiplos indicadores que servem como elementos de análise e atuação que podem ampliar o entendimento deste campo do conhecimento em suas especificidades dadas pelo contexto da saúde. (AU)