RESUMO
OBJECTIVE: Deterioration index (DI) is a computer-generated score at a specific frequency that represents the overall condition of hospitalized patients using a variety of clinical, laboratory and physiologic data. In this paper, a contrastive transfer learning method is proposed and validated for early prediction of adverse events in hospitalized patients using DI scores. METHODS AND PROCEDURES: An unsupervised contrastive learning (CL) model with a classifier is proposed to predict adverse outcome using a single temporal variable (DI scores). The model is pretrained on an unsupervised fashion with large-scale time series data and fine-tuned with retrospective DI score data. RESULTS: The performance of this model is compared with supervised deep learning models for time series classification. Results show that unsupervised contrastive transfer learning with a classifier outperforms supervised deep learning solutions. Pretraining of the proposed CL model with large-scale time series data and fine-tuning that with DI scores can enhance prediction accuracy. CONCLUSION: A relationship exists between longitudinal DI scores of a patient and the corresponding outcome. DI scores and contrastive transfer learning can be used to predict and prevent adverse outcomes in hospitalized patients. CLINICAL IMPACT: This paper successfully developed an unsupervised contrastive transfer learning algorithm for prediction of adverse events in hospitalized patients. The proposed model can be deployed in hospitals as an early warning system for preemptive intervention in hospitalized patients, which can mitigate the likelihood of adverse outcomes.
Assuntos
Serviços de Laboratório Clínico , Pacientes , Humanos , Estudos Retrospectivos , Algoritmos , Aprendizado de MáquinaRESUMO
INTRODUCTION: Stomach cancer is one of the most common causes of cancer worldwide, especially in the population over 65 years. The survival rate of the elderly is lower in comparison with young people, and they are underrepresented in clinical trials and research in general. The evaluation of Multidimensional Geriatric Assessment (MGA) would be key for assessing the prognosis of these patients and therefore having a more informed decision-making process when considering one of the most vulnerable populations. METHODS: A search was performed in the OVID, Embase, and PubBMed databases. There was no restriction on publication time, language, or study design. Eligible studies were those that included geriatric patients with a diagnosis of nonmetastatic stomach cancer who receive oncospecific and surgical management, used Multidimensional/Comprehensive Geriatric Assessment (MGA), and which outcomes included at least overall survival, morbidity, and mortality. RESULTS: Four studies were included, and the MGA battery was not implemented, but rather easily measurable scales such as nutritional status, functional status, cognitive and behavioral disorders, comorbidities, and polypharmacy. Some authors proposed that the assessment of overall survival is not explicit among the included studies; patients with gastric cancer and mild, moderate, severe, and total dependence had higher mortality than independent patients (39% [HR 1.39; 95% CI: 1.09-1.7], 68% [95% CI: 1.46-1.93], 187% [HR 2.87 95% CI: 2.47-3.34], and 234% [95% CI: 2.81-3.97]), respectively. The Zhou study showed an association between sarcopenia, assessed by imaging studies, and a longer hospital stay in days (16 [9] vs. 13 [6], p 0.004). The study by Pujara found that polypharmacy (OR 2.36 CI: 1.08-5.17) and weight loss greater than 10% in the past 6 months were associated with greater postoperative morbidity at 90 days (OR 2.36 CI: 1.08-5.17, OR 11.21 CI: 2.16-58.24). CONCLUSION: MGA was not broadly implemented. Geriatric assessment dependency appears to be a prognostic marker of survival in patients with gastric cancer. Sarcopenia appears to be an important prognostic marker for short- and long-term outcomes. Higher quality studies in this specific population are required to support the systematic use of this assessment for the choice of appropriate therapy according to the patient.
RESUMO
The Deterioration Index (DI) is an automatic early warning system that utilizes a machine learning algorithm integrated into the electronic health record and was implemented to improve risk stratification of inpatients. Our pilot implementation showed superior diagnostic accuracy than standard care. A score >60 had a specificity of 88.5% and a sensitivity of 59.8% (PPV 0.1758, NPP 0.9817). However, acceptance in the clinical workflow was divided; nurses preferred standard care, while providers found it helpful.
Assuntos
Algoritmos , Registros Eletrônicos de Saúde , Humanos , Pacientes Internados , Aprendizado de Máquina , Fluxo de TrabalhoRESUMO
Prolonged QT interval is an independent risk factor for all-cause mortality. However, evaluation of mortality associated to the implementation of a clinical decision support system to increase awareness and provide management recommendations has been challenging. Here we present our attempt to develop a model using only electronic data and different control groups.
Assuntos
Sistemas de Apoio a Decisões Clínicas , Humanos , Grupos Controle , Pacientes , Fatores de RiscoRESUMO
Severe hypercholesterolemia/possible familial hypercholesterolemia (FH) is relatively common but underdiagnosed and undertreated. We investigated whether implementing clinical decision support (CDS) was associated with lower low-density lipoprotein cholesterol (LDL-C) in patients with severe hypercholesterolemia/possible FH (LDL-C ≥ 190 mg/dL). As part of a pre-post implementation study, a CDS alert was deployed in the electronic health record (EHR) in a large health system comprising 3 main sites, 16 hospitals and 53 clinics. Data were collected for 3 months before ('silent mode') and after ('active mode') its implementation. Clinicians were only able to view the alert in the EHR during active mode. We matched individuals 1:1 in both modes, based on age, sex, and baseline lipid lowering therapy (LLT). The primary outcome was difference in LDL-C between the two groups and the secondary outcome was initiation/intensification of LLT after alert trigger. We identified 800 matched patients in each mode (mean ± SD age 56.1 ± 11.8 y vs. 55.9 ± 11.8 y; 36.0% male in both groups; mean ± SD initial LDL-C 211.3 ± 27.4 mg/dL vs. 209.8 ± 23.9 mg/dL; 11.2% on LLT at baseline in each group). LDL-C levels were 6.6 mg/dL lower (95% CI, -10.7 to -2.5; P = 0.002) in active vs. silent mode. The odds of high-intensity statin use (OR, 1.78; 95% CI, 1.41-2.23; P < 0.001) and LLT initiation/intensification (OR, 1.30, 95% CI, 1.06-1.58, P = 0.01) were higher in active vs. silent mode. Implementation of a CDS was associated with lowering of LDL-C levels in patients with severe hypercholesterolemia/possible FH, likely due to higher rates of clinician led LLT initiation/intensification.
RESUMO
Background: Threshold-based early warning systems (EWS) are used to predict adverse events (Aes). Machine learning (ML) algorithms that incorporate all EWS scores prior to an event may perform better in hospitalized patients. Methods: The deterioration index (DI) is a proprietary EWS. A threshold of DI >60 is used to predict a composite AE: all-cause mortality, cardiac arrest, transfer to intensive care, and evaluation by the rapid response team in practice. The DI scores were collected for adult patients (≥18 y-o) hospitalized on medical or surgical services during 8-23-2021 to 3-31-2022 from four different Mayo Clinic sites in the United States. A novel ML model was developed and trained on a retrospective cohort of hospital encounters. DI scores were represented in a high-dimensional space using random convolution kernels to facilitate training of a classifier and the area under the receiver operator characteristics curve (AUC) was calculated. Multiple time intervals prior to an AE were analyzed. A leave-one-out cross-validation protocol was used to evaluate performance across separate clinic sites. Findings: Three different classifiers were trained on 59,617 encounter-derived DI scores in high-dimensional feature space and the AUCs were compared to two threshold models. All three tested classifiers improved the AUC over the threshold approaches from 0.56 and 0.57 to 0.76, 0.85 and 0.94. Time interval analysis of the top performing classifier showed best accuracy in the hour before an event occurred (AUC 0.91), but prediction held up even in the 12 h before an AE (AUC 0.80 at minus 12 h, 0.81 at minus 9 h, 0.85 at minus 6 h, and 0.88 at minus 3 h before an AE). Multisite cross-validation using leave-one-out approach on data from four different clinical sites showed broad generalization performance of the top performing ML model with AUC of 0.91, 0.91, 0.95, and 0.91. Interpretation: A novel ML model that incorporates all the longitudinal DI scores prior to an AE in a hospitalized patient performs better at outcome prediction than the currently used threshold model. The use of clinical data, a generalized ML technique, and successful multisite cross-validation demonstrate the feasibility of our model in clinical implementation. Funding: No funding to report.
RESUMO
Interdisciplinary research teams can be extremely beneficial when addressing difficult clinical problems. The incorporation of conceptual and methodological strategies from a variety of research disciplines and health professions yields transformative results. In this setting, the long-term goal of team science is to improve patient care, with emphasis on population health outcomes. However, team principles necessary for effective research teams are rarely taught in health professional schools. To form successful interdisciplinary research teams in cardio-oncology and beyond, guiding principles and organizational recommendations are necessary. Cardiovascular disease results in annual direct costs of $220 billion (about $680 per person in the US) and is the leading cause of death for cancer survivors, including adult survivors of childhood cancers. Optimizing cardio-oncology research in interdisciplinary research teams has the potential to aid in the investigation of strategies for saving hundreds of thousands of lives each year in the United States and mitigating the annual cost of cardiovascular disease. Despite published reports on experiences developing research teams across organizations, specialties and settings, there is no single journal article that compiles principles for cardiology or cardio-oncology research teams. In this review, recurring threads linked to working as a team, as well as optimal methods, advantages, and problems that arise when managing teams are described in the context of career development and research. The worth and hurdles of a team approach, based on practical lessons learned from establishing our multidisciplinary research team and information gleaned from relevant specialties in the development of a successful team are presented.
RESUMO
To evaluate the seasonal variation and isotopic composition of phytoplankton, water samples were collected monthly between October 2007 and November 2008 in Lake Catalão, a floodplain lake at the confluence between rivers Negro and Amazon. Analyses of total chlorophyll concentration and ð13C and ð15N isotopic abundances were made from particulate size fractions of 30-60, 10-30 and < 10 µm in the littoral, pelagic, and floating meadows regions. Chlorophyll concentration was found to be inversely associated to lake depth, and high concentrations of chlorophyll in the floating meadows zone were significant. The fraction < 10 µm was the most abundant representing in average more than 40 % of the particulate matter. The ð13C values were relatively constant during the study (-25.1 % ∼ -34.0 %), whereas the ð15N values showed strong variability (15.6 % ∼ 2.4 %), which has been attributed to the resuspension of sediments during mixing of the water column. Mixing associated to the sudden drop in temperature during the rising water period was an important event in the trophic and isotopic dynamics of the lake. Variations in chlorophyll content were generally associated with the dilution process, in which concentration was inversely correlated to the water level, whereas abundance was directly correlated to the water level.
Con el propósito de evaluar la variación estacional de la abundancia isotópica (ð13C e ð15N) del fitoplancton, muestreos mensuales fueron realizados entre octubre de 2007 y noviembre de 2008 en el lago Catalão, un lago de inundación en la zona de confluencia de los ríos Negro y Solimãµes, ubicado frente a la ciudad de Manaus (AM, Brasil). Análisis de la clorofila total y evaluaciones de la abundancia natural de ð13C y ð15N fueron realizados en las fracciones partículadas de 30-60, 10-30 y < 10 µm en las zonas litoral, pelágica y de macrófitas acuáticas. La concentración de clorofila presentó una relación inversa con la profundidad del lago, siendo relevantes las altas concentraciones encontradas dentro del tapete de macrófitas acuáticas. La fracción < 10µm fue la que presentó la mayor concentración de clorofila, representando más del 40% del material particulado. Los valores de ð13C fueron relativamente constantes durante el período de estudio (-25,1 % ∼ - 34 %), mientras que la abundancia natural de ð15N presentó una amplia variación (15,6 % ∼ – 2,4 %), que fue atribuida al proceso de resuspensión de los sedimentos en los procesos de mezcla de la columna de agua. En este contexto, el fenómeno de la mezcla asociada con la friaje durante el período de aguas altas, fue un evento importante en la dinámica trófica e isotópica del lago. En general las variaciones de la concentración de clorofila fueron asociadas al proceso de dilución, en el cual la concentración es inversa y la abundancia es directamente relacionada con el nivel del agua.
RESUMO
Con el propósito de evaluar la participación de bacterias heterotróficas en el flujo de carbono en la cadena trófica de un lago de inundación amazónico, recolectas mensuales de estos microorganismos fueron realizadas durante el año hidrológico de diciembre de 2007 hasta noviembre de 2008, en las regiones litoral, pelágica y de macrófitas acuáticas del lago Catalão, en la Amazonia central brasilera. Las bacterias fueron multiplicadas in vitro, usando como sustrato carbono orgánico disuelto (COD) del lago en cada una de las regiones estudiadas, para posterior análisis de la abundancia de isotopos estables de carbono y nitrógeno. Estos datos fueron confrontados con los valores de COD de los cuatro periodos limnológicos del lago (seca, inundación, llena y bajada de aguas). En general se encontró que la fuente principal de carbono de las bacterias heterotróficas fue aquella de origen C4, que presentó un aporte mínimo de 75%, en la medida en que el d13C de las bacterias presentó valor medio de -17,72‰ ± 2,25. Confrontando ese valor, con el d13C del zooplancton en el mismo período (- 33,04‰ ± 3,81) se concluye que el aporte de las bacterias heterotróficas al flujo de carbono hacia niveles tróficos superiores en el lago estudiado es mínimo.
In order to evaluate the role of heterotrophic bacteria on carbon flow in food chains of an Amazonian floodplain lake, monthly collections of these organisms were made during the hydrological year from December 2007 to November 2008. Littoral, pelagic, and aquatic macrophyte regions of the Catalão Lake in Central Amazonia were sampled and bacteria were multiplied in vitro, using dissolved organic carbon (COD) of each one of the regions studied as a substrate. The bacterial biomass obtained was used for stable isotope analyses of carbon and nitrogen. These data were confronted with COD values of the four hydrological periods of the lake (dry, rising, flood and fall). In general, it was found that the main source of carbon for heterotrophic bacteria was that of C4 origin, which presented a minimum contribution of 75% of bacterial biomass, to the extent that the bacteria d13C average value was -17.72‰ ± 2.25. Comparing this value with the d13C of zooplankton in the same period (-33.04‰ ± 3.81) permit concludes that the contribution of heterotrophic bacteria in the carbon flow to higher trophic levels in the Catalão lake is minimal.
RESUMO
Para definir o efeito do seston como fonte de alimento para Ceriodaphnia cornuta eDiaphanosoma spinolosum durante a enchente no lago Catalão (Amazônia Central) foramdesenvolvidos experimentos in situ, utilizando a variação natural de δ13C e δ15N comotraçadores de alimento. As duas espécies tiveram acesso a três diferentes frações de seston (< 10,< 30 e < 60 μm) sendo mantidas em garrafas plásticas de 1,1 L submersas, durante nove dias. Asatividades foram iniciadas com uma população de dez indivíduos em cada frasco, com três frascospor cada fração e quatro intervalos de tempo, para um total de 27 frascos por espécie. A cada trêsdias, foram coletados três frascos e os organismos foram fixados para contagem e observação deparâmetros populacionais. Os organismos nos últimos três frascos de cada fração foram fixadoscom formalina e usados para análise de isótopos estáveis de carbono e nitrogênio. As duasespécies cresceram em todas as frações de alimento, sendo melhor seu desempenho populacionalna fração < 30 μm. D. spinolosum e C. cornuta apresentaram enriquecimento nos valores de δ13C eδ15N, sendo maior na fração < 10 μm. Estes resultados sugerem que as frações do seston testadasproduzem diferentes taxas de crescimento populacional e assinaturas isotópicas nos cladóceros.(AU)
In order to define the effect of seston size fraction as a food source forCeriodaphnia cornuta and Diaphanosoma spinolosum, in situ experiments using natural abundance ofδ13C and δ15N were carried out during the rising water period in Catalão Lake. Both species werefed with three different fractions of seston from the lake (< 10, < 30 and < 60 μm) andmaintained in submerged 1.1 L plastic bottles for nine days. Ten individuals were put in eachflask, maintaining three flasks for each fraction, for a total of 27 flasks per species. Every threedays we collected three flasks and fixed the individuals for subsequent counting and observationof demographic population parameters. The organisms in the last three flasks of each fractionwere fixed for stable isotope analysis of δ13C and δ15N. Both species grew in all of the foodfractions, with the best the performance coming in the < 30 μm fraction. D. spinolosum andC. cornuta showed enrichment in δ13C and δ15N, which was highest in the < 10 μm fraction forboth species. These results show that the seston size fractions tested produce different populationgrowth rates and isotopic signatures in cladocerans.(AU)
Assuntos
Cladocera , Eucariotos , EucariotosRESUMO
Para definir o efeito do seston como fonte de alimento para Ceriodaphnia cornuta eDiaphanosoma spinolosum durante a enchente no lago Catalão (Amazônia Central) foramdesenvolvidos experimentos in situ, utilizando a variação natural de δ13C e δ15N comotraçadores de alimento. As duas espécies tiveram acesso a três diferentes frações de seston (< 10,< 30 e < 60 μm) sendo mantidas em garrafas plásticas de 1,1 L submersas, durante nove dias. Asatividades foram iniciadas com uma população de dez indivíduos em cada frasco, com três frascospor cada fração e quatro intervalos de tempo, para um total de 27 frascos por espécie. A cada trêsdias, foram coletados três frascos e os organismos foram fixados para contagem e observação deparâmetros populacionais. Os organismos nos últimos três frascos de cada fração foram fixadoscom formalina e usados para análise de isótopos estáveis de carbono e nitrogênio. As duasespécies cresceram em todas as frações de alimento, sendo melhor seu desempenho populacionalna fração < 30 μm. D. spinolosum e C. cornuta apresentaram enriquecimento nos valores de δ13C eδ15N, sendo maior na fração < 10 μm. Estes resultados sugerem que as frações do seston testadasproduzem diferentes taxas de crescimento populacional e assinaturas isotópicas nos cladóceros.
In order to define the effect of seston size fraction as a food source forCeriodaphnia cornuta and Diaphanosoma spinolosum, in situ experiments using natural abundance ofδ13C and δ15N were carried out during the rising water period in Catalão Lake. Both species werefed with three different fractions of seston from the lake (< 10, < 30 and < 60 μm) andmaintained in submerged 1.1 L plastic bottles for nine days. Ten individuals were put in eachflask, maintaining three flasks for each fraction, for a total of 27 flasks per species. Every threedays we collected three flasks and fixed the individuals for subsequent counting and observationof demographic population parameters. The organisms in the last three flasks of each fractionwere fixed for stable isotope analysis of δ13C and δ15N. Both species grew in all of the foodfractions, with the best the performance coming in the < 30 μm fraction. D. spinolosum andC. cornuta showed enrichment in δ13C and δ15N, which was highest in the < 10 μm fraction forboth species. These results show that the seston size fractions tested produce different populationgrowth rates and isotopic signatures in cladocerans.
Assuntos
Cladocera , EucariotosRESUMO
Para definir o efeito do seston como fonte de alimento para Ceriodaphnia cornuta e Diaphanosoma spinolosum durante a enchente no lago Catalão (Amazônia Central) foram desenvolvidos experimentos "in situ", utilizando a variação natural de δ13C e δ15N como traçadores de alimento. As duas espécies tiveram acesso a três diferentes frações de seston (< 10, < 30 e < 60 µm) sendo mantidas em garrafas plásticas de 1,1 L submersas, durante nove dias. As atividades foram iniciadas com uma população de dez indivíduos em cada frasco, com três frascos por cada fração e quatro intervalos de tempo, para um total de 27 frascos por espécie. A cada três dias, foram coletados três frascos e os organismos foram fixados para contagem e observação de parâmetros populacionais. Os organismos nos últimos três frascos de cada fração foram fixados com formalina e usados para análise de isótopos estáveis de carbono e nitrogênio. As duas espécies cresceram em todas as frações de alimento, sendo melhor seu desempenho populacional na fração < 30 µm. D. spinolosum e C. cornuta apresentaram enriquecimento nos valores de δ13C e δ15N, sendo maior na fração < 10 µm. Estes resultados sugerem que as frações do seston testadas produzem diferentes taxas de crescimento populacional e assinaturas isotópicas nos cladóceros.
In order to define the effect of seston size fraction as a food source for Ceriodaphnia cornuta and Diaphanosoma spinolosum, in situ experiments using natural abundance of δ13C and δ15N were carried out during the rising water period in Catalão Lake. Both species were fed with three different fractions of seston from the lake (< 10, < 30 and < 60 µm) and maintained in submerged 1.1 L plastic bottles for nine days. Ten individuals were put in each flask, maintaining three flasks for each fraction, for a total of 27 flasks per species. Every three days we collected three flasks and fixed the individuals for subsequent counting and observation of demographic population parameters. The organisms in the last three flasks of each fraction were fixed for stable isotope analysis of δ13C and δ15N. Both species grew in all of the food fractions, with the best the performance coming in the < 30 µm fraction. D. spinolosum and C. cornuta showed enrichment in δ13C and δ15N, which was highest in the < 10 µm fraction for both species. These results show that the seston size fractions tested produce different population growth rates and isotopic signatures in cladocerans.