Unsupervised natural language processing in the identification of patients with suspected COVID-19 infection

Patients with post-COVID-19 syndrome benefit from health promotion programs. Their rapid identification is important for the cost-effective use of these programs. Traditional identification techniques perform poorly especially in pandemics. A descriptive observational study was carried out using 105,008 prior authorizations paid by a private health care provider with the application of an unsupervised natural language processing method by topic modeling to identify patients suspected of being infected by COVID-19. A total of 6 models were generated: 3 using the BERTopic algorithm and 3 Word2Vec models. The BERTopic model automatically creates disease groups. In the Word2Vec model, manual analysis of the first 100 cases of each topic was necessary to define the topics related to COVID-19. The BERTopic model with more than 1,000 authorizations per topic without word treatment selected more severe patients - average cost per prior authorizations paid of BRL 10,206 and total expenditure of BRL 20.3 million (5.4%) in 1,987 prior authorizations (1.9%). It had 70% accuracy compared to human analysis and 20% of cases with potential interest, all subject to analysis for inclusion in a health promotion program. It had an important loss of cases when compared to the traditional research model with structured language and identified other groups of diseases - orthopedic, mental and cancer. The BERTopic model served as an exploratory method to be used in case labeling and subsequent application in supervised models. The automatic identification of other diseases raises ethical questions about the treatment of health information by machine learning.

Regional myocardial sympathetic denervation precedes the development of left ventricular systolic dysfunction in chronic Chagas' cardiomyopathy

BACKGROUND

Regional myocardial sympathetic denervation is a conspicuous and early disorder in patients with chronic Chagas' cardiomyopathy (CCC), potentially associated to the progression of myocardial dysfunction OBJECTIVE: To evaluate in a longitudinal study the association between the presence and the progression of regional myocardial sympathetic denervation with the deterioration of global and segmental left ventricular dysfunction in CCC.

METHODS

18 patients with CCC were submitted at initial evaluation and after 5.5 years to rest myocardial scintigraphy with 123Iodo-metaiodobenzylguanidine and 99mTc-sestamibi and to two-dimensional echocardiography to assess myocardial sympathetic denervation, extent of fibrosis, and the left ventricular ejection fraction (LVEF) and wall motion abnormalities.

RESULTS

In the follow-up evaluation, compared to the initial one, we observed a significant decrease in LVEF (56 ± 11 to 49% ± 12; P = .01) and increased summed defects scores in the myocardial innervation scintigraphy (15 ± 10 to 20 ± 9; P < .01). The presence of regional myocardial sympathetic denervation in ventricular regions of viable non-fibrotic myocardium presented an odds ratio of 4.25 for the development of new wall motion abnormalities (P = .001).

CONCLUSION

Regional and global myocardial sympathetic denervation is a progressive derangement in CCC. In addition, the regional denervation is topographically associated with areas of future development of regional systolic dysfunction in patients with CCC.

Pulmonary Vascular Volume Estimated by Automated Software is a Mortality Predictor after Acute Pulmonary Embolism

Fundamento:

A embolia pulmonar aguda (EPA) tem desfecho clínico variável. A angiotomografia computadorizada (angio-CT) é considerada o padrão-ouro para o diagnóstico.

Objetivo:

Avaliar se o volume vascular pulmonar (VVP) quantificado por software automatizado é um preditor de mortalidade após EPA.

Métodos:

Estudo de coorte retrospectivo no qual a imagem da angio-CT de 61 pacientes com EPA foi reanalisada. O VVP e o volume pulmonar (VP) foram estimados automaticamente pelo software Yacta. Calculamos o VVP ajustado pela razão: VVP(cm³)/VP(litros). Parâmetros prognósticos clássicos da angio-CT (carga embólica; razão do diâmetro do ventrículo direito/ventrículo esquerdo; razão do diâmetro da artéria pulmonar/aorta; desvio do septo interventricular; infarto pulmonar e refluxo de contraste na veia hepática) foram avaliados. A mortalidade em 1 mês foi o desfecho analisado. Consideramos um valor de p <0,05 como estatisticamente significativo.

Resultados:

Sete mortes (11%) ocorreram entre os 61 pacientes durante 1 mês de seguimento. O VVP ajustado <23cm³/L foi um preditor independente de mortalidade na análise univariada (odds ratio [OR]: 26; intervalo de confiança de 95% [IC95%]: 3-244; p=0,004) e na análise multivariada (OR ajustado: 19 [IC95%: 1,3-270]; p=0,03). Os parâmetros clássicos da angio-CT não foram associados à mortalidade em 1 mês nesta amostra. O VVP ajustado <23cm³/L apresentou sensibilidade de 86%, especificidade de 82%, valor preditivo negativo de 94% e valor preditivo positivo de 64% para identificação dos pacientes que morreram.

Conclusão:

VVP ajustado <23cm³/L foi um preditor independente de mortalidade após EPA. Esse parâmetro mostrou melhor desempenho prognóstico do que os outros achados clássicos da angio-CT. (Arq Bras Cardiol. 2020; 115(5):809-818)

Prognostic Value of MMP-9 -1562 C/T Gene Polymorphism in Patients With Sepsis

Introduction:

Matrix metalloproteinase-9 (MMP-9) plays an important role in the pathophysiology of sepsis. A single-nucleotide polymorphism (SNP) at position -1562 (C/T) in the MMP-9 gene has been associated with differential MMP-9 expression, being higher when the -1562 T allele is present. We evaluated the association of the SNP MMP9 -1562 C/T with severity and mortality in patients with sepsis to establish whether the prognosis of the disease is affected.

Materials and Methods:

A case-control study exploratory was carried out in a cohort of infected patients. 540 individuals were selected in total, 270 patients with sepsis and 270 controls (infected but non-septic), classified according to the 2016 consensus (Sepsis-3). The presence of the single-nucleotide polymorphism (SNP; allele T and/or allele C) was determined through analyses of restriction fragment length polymorphism and plasma levels of MMP-9 were determined through enzyme-linked immunosorbent assay immunoassay.

Results:

SNP MMP-9 -1562 has two known alleles (T and C), with predominance of the C over the T allele; in the group of patients with sepsis, T allele was found in 7.2% of cases, while C allele in the rest (92.8%); in comparison, in the group of infected but non-septic patients, frequencies were 9.4% for T allele and 90.6% for the C allele (P = .33). Also, the presence of the polymorphic T allele was not related to the levels of MMP-9 in patients with sepsis in comparison with infected but non-septic patients 780 (397-1375) ng/mL vs 646 (172-1249) ng/mL (P = .64). There was also no association between the SNP and sepsis mortality (P = .78).

Conclusions:

We concluded that there was no association between the SNP MMP9 -1562 C/T and sepsis or between the SNP MMP9 -1562 C/T and sepsis mortality in the Northeastern Colombian septic patient cohort. Further research is needed to clarify the correlation among sepsis, genetic factors with allele T and MMP-9 plasma concentration.

[Evaluation and incorporation of health technologies: process and methodology adopted by a high-complexity care university]

The demographic and epidemiological transitions tend to increase the role of hospitals in medical care. Within such organizations, effective, safe, and cost-effective health technologies ensure better quality of care and increase users' survival, thus emphasizing the importance of evaluation of such hospital-based technologies. This article aims to present a model for the evaluation and incorporation of technology in a teaching hospital that provides high-complexity care. The article describes an approach to methods/processes assessment that can be used easily by any hospital. The model allowed proper health technology assessment (HTA), thereby legitimizing decisions on technology incorporation by the hospital administration with high levels of acceptance and adoption by the clinical staff, suggesting that hospital-based HTA (provided that it is well-structured, with the support of institutional administration) can be a powerful tool for dissemination and valorization of HTA culture in an environment with the widest use and the greatest impact on the health system's budget.

The need to improve health care in prisons

OBJECTIVE

To analyze physical structure, working conditions of health professionals and outline of the procedures established in prisons.

METHODS

We analyzed 34 provisional detention centers and 69 male and six female prison units in the state of Sao Paulo, Southeastern Brazil, in 2009. A self-applied instrument was developed to collect quantitative data on the characteristics of health care structure, equipment and personnel in prisons. Analysis of variance (ANOVA) or equivalent non-parametric tests and Chi-square or Fisher’s tests were used to compare categorical and continuous variables, respectively, between the groups.

RESULTS

The main problems were delays in the results of laboratory tests and imaging. With respect to the teams, it was observed that a large majority were in conditions close to those proposed by the Bipartite Commission 2013 but without improvement being reflected in the indicators. With respect to the process, more than 60.0% of prisons located in small towns do not have the structural conditions to ensure secondary or tertiary health care for the continuity of treatment.

CONCLUSIONS

This profile of prisons in the country can be used for planning and monitoring future actions for the continuous improvement of healthcare processes.

Assessment of MMP-9, TIMP-1, and COX-2 in normal tissue and in advanced symptomatic and asymptomatic carotid plaques

Background

Mature carotid plaques are complex structures, and their histological classification is challenging. The carotid plaques of asymptomatic and symptomatic patients could exhibit identical histological components.

Objectives

To investigate whether matrix metalloproteinase 9 (MMP-9), tissue inhibitor of MMP (TIMP), and cyclooxygenase-2 (COX-2) have different expression levels in advanced symptomatic carotid plaques, asymptomatic carotid plaques, and normal tissue.

Methods

Thirty patients admitted for carotid endarterectomy were selected. Each patient was assigned preoperatively to one of two groups: group I consisted of symptomatic patients (n = 16, 12 males, mean age 66.7 ± 6.8 years), and group II consisted of asymptomatic patients (n = 14, 8 males, mean age 67.6 ± 6.81 years). Nine normal carotid arteries were used as control. Tissue specimens were analyzed for fibromuscular, lipid and calcium contents. The expressions of MMP-9, TIMP-1 and COX-2 in each plaque were quantified.

Results

Fifty-eight percent of all carotid plaques were classified as Type VI according to the American Heart Association Committee on Vascular Lesions. The control carotid arteries all were classified as Type III. The median percentage of fibromuscular tissue was significantly greater in group II compared to group I (p < 0.05). The median percentage of lipid tissue had a tendency to be greater in group I than in group II (p = 0.057). The percentages of calcification were similar among the two groups. MMP-9 protein expression levels were significantly higher in group II and in the control group when compared with group I (p < 0.001). TIMP-1 expression levels were significantly higher in the control group and in group II when compared to group I, with statistical difference between control group and group I (p = 0.010). COX-2 expression levels did not differ among groups. There was no statistical correlation between MMP-9, COX-2, and TIMP-1 levels and fibrous tissue.

Conclusions

MMP-9 and TIMP-1 are present in all stages of atherosclerotic plaque progression, from normal tissue to advanced lesions. When sections of a plaque are analyzed without preselection, MMP-9 concentration is higher in normal tissues and asymptomatic surgical specimens than in symptomatic specimens, and TIMP-1 concentration is higher in normal tissue than in symptomatic specimens.

Histological composition and progression of carotid plaque

Background

To analyse histological composition and progression of carotid plaque.

Methods

Thirty-one patients (22 males, mean age 68.03 ± 7.3 years) admitted for carotid endarterectomy for extracranial high-grade internal carotid artery stenosis (≥ 70% luminal narrowing) were enrolled. The patients were divided into 2 groups according to symptomatology (group I, 17 symptomatic patients; and group II, 14 asymptomatic patients). A histological analysis and inflammatory cell quantification of each excised carotid plaque was made. Nine carotid arteries were removed from human cadavers that were not preselected for carotid artery disease. These specimens were used as a control tissue without any macroscopic signs of atherosclerotic plaques.

Results

Fifty eight percent of all carotid plaques were classified as complex plaque with possible surface defect, hemorrhage or thrombus. The inflammatory cells concentration did not differ between the two groups. All specimens from human cadavers were classified as preatheroma with extracellular lipid pools.

Conclusion

Asymptomatic and symptomatic patients could have the same histological components on their carotid plaques. Fibrotic and calcific plaques could become vulnerable as complex plaques with surface defect, hemorrhage and thrombus could remain silent. Asymptomatic carotid stenosis should be followed close with no invasive diagnostic methods and clinical evaluation.

Ultrasonic tissue characterization of vulnerable carotid plaque: correlation between videodensitometric method and histological examination

BACKGROUND

To establish the correlation between quantitative analysis based on B-mode ultrasound images of vulnerable carotid plaque and histological examination of the surgically removed plaque, on the basis of a videodensitometric digital texture characterization.

METHODS

Twenty-five patients (18 males, mean age 67 +/- 6.9 years) admitted for carotid endarterectomy for extracranial high-grade internal carotid artery stenosis (> or = 70% luminal narrowing) underwent to quantitative ultrasonic tissue characterization of carotid plaque before surgery. A computer software (Carotid Plaque Analysis Software) was developed to perform the videodensitometric analysis. The patients were divided into 2 groups according to symptomatology (group I, 15 symptomatic patients; and group II, 10 patients asymptomatic). Tissue specimens were analysed for lipid, fibromuscular tissue and calcium.

RESULTS

The first order statistic parameter mean gray level was able to distinguish the groups I and II (p = 0.04). The second order parameter energy also was able to distinguish the groups (p = 0,02). A histological correlation showed a tendency of mean gray level to have progressively greater values from specimens with < 50% to > 75% of fibrosis.

CONCLUSION

Videodensitometric computer analysis of scan images may be used to identify vulnerable and potentially unstable lipid-rich carotid plaques, which are less echogenic in density than stable or asymptomatic, more densely fibrotic plaques.

An automatic temperature-control system for solutions in free flow

We describe a temperature-control system for solutions in free flow, suitable for electrophysiological or optical studies of isolated cells, natural epithelia or cell culture monolayers. The system is small enough to be located close to the preparation and was designed specifically to be coupled to the inlets of a modified, continuous-flow Ussing chamber, allowing rapid change of the solutions bathing tissue surfaces. The system consists of a highly compact monoblock heating unit and a control circuit. Solutions from different reservoirs, kept at room temperature or lower (from an ice bath), can be rapidly switched at the inlet of the heating unit by manually or electrically actuated microvalves without affecting the temperature of the fluid leaving the heating unit. The control unit consists of a bead thermistor firmly placed close to the heating unit outlet and an electronic circuit which is basically a proportional controller. This unit continuously regulates the electric current through the Ni-Cr heater, keeping the temperature of the fluid leaving the heating unit constant at a preset value. The system allows control of fluid temperature (normally 37 degrees C) for flow rates in the range of 1.0 ml/min to 12 ml/min. However, the temperature can be set at any value above that of the incoming fluid.

Intramedullary cysticercosis. Case report, literature review and comments on pathogenesis

A 60-year-old woman developed progressive spastic crural paraplegia and global anesthesia and global anesthesia below T10, succumbing to urinary infection and bronchopneumonia 8 months after the onset of symptoms. At necropsy, the spinal cord was completely destroyed by a necrotic intramedullary cysticercus at T8. Six cysticerci were found in the brain. A review of the 17 published cases of intramedullary cysticercosis (including this one) showed this condition to be clinically indistinguishable from spinal cord tumours. The surgical prognosis was fair in 8 of 11 operated patients. The role of hematogenous and ventriculo-ependymal pathways in the pathogenesis of intramedullary cysticercosis was examined. The topographical distribution of intramedullary cysticerci (5 cervical, 12 thoracic, 2 lumbar, none sacral) was found to be statistically proportional to the blood flow to each of these regions; this favours the hematogenous route of infestation. The low spinal cord blood flow (100 times less than that to the brain), the type of vascularization of the cord (low calibre vessels under low pressure) and peculiarities of the cord tissue (such as its harder consistency) are all thought to be contributory factors accounting for the scarcity of intramedullary cysticerci. No evidence for an ependymal route of spread could be adduced.