Sergio Terrasa

Evaluation of HUGE equation (hematocrit, urea, gender) performance for screening chronic kidney disease in clinically stable cirrhotic patients

for HUGE equation validation patient’s evaluation by two nephrologists blind between them, as gold standard for renal health status, was used [3–5]. CKD is an entity frequently diagnosed in cirrhotic patients, and this kidney–liver alteration may be caused by diseases that can affect both organs (e.g., chronic virus C infection with cryoglobulinemia), or renal conditions induced by cirrhosis (e.g., Ig A nephropathy or pre-renal insufficiency induced by an increased vasomotor tone on renal circulation [6]. An equation like HUGE could be useful for screening CKD in cirrhotic patients, but this equation is based on serum parameters that can be altered by cirrhosis: Hematocrit may be low due to erythropoietin resistance (chronic disease anemia), and serum urea values may be lower because of a reduced urea biosynthesis (reduced hepatic conversion of ammonium to urea) [7, 8]. Thus, we decided to originally evaluate whether HUGE equation could be an accurate tool for detecting CKD in stable cirrhotic patients despite the influence that this condition can have on its constituents. With this objective, we performed a retrospective observational study to assess the operational characteristics of HUGE equation for screening CKD in 75 patients suffering from stable liver cirrhosis (Child–Pugh A) mostly secondary to hepatitis C (43 %), selected from a population of 750 cirrhotic patients who were on follow-up during 1 year (January 2014–January 2015) by the Hepatology and Transplantation Section of the Internal Medicine Division in the Hospital Italiano de Buenos Aires (Argentina). Inclusion criteria were as follows: to have information, every 4 months, during the year of the study (2014–2015) regarding patients’ serum and urine electrolytes, urea, creatinine, uric acid, hematocrit, hemoglobin, glucose, intact parathyroid hormone, urinalysis, and renal ultrasound. Editor,

Development and validation of various phenotyping algorithms for Diabetes Mellitus using data from electronic health records

BACKGROUND AND OBJECTIVE Recent progression towards precision medicine has encouraged the use of electronic health records (EHRs) as a source for large amounts of data, which is required for studying the effect of treatments or risk factors in more specific subpopulations. Phenotyping algorithms allow to automatically classify patients according to their particular electronic phenotype thus facilitating the setup of retrospective cohorts. Our objective is to compare the performance of different classification strategies (only using standardized problems, rule-based algorithms, statistical learning algorithms (six learners) and stacked generalization (five versions)), for the categorization of patients according to their diabetic status (diabetics, not diabetics and inconclusive; Diabetes of any type) using information extracted from EHRs. METHODS Patient information was extracted from the EHR at Hospital Italiano de Buenos Aires, Buenos Aires, Argentina. For the derivation and validation datasets, two probabilistic samples of patients from different years (2005: n = 1663; 2015: n = 800) were extracted. The only inclusion criterion was age (≥40 & <80 years). Four researchers manually reviewed all records and classified patients according to their diabetic status (diabetic: diabetes registered as a health problem or fulfilling the ADA criteria; non-diabetic: not fulfilling the ADA criteria and having at least one fasting glycemia below 126 mg/dL; inconclusive: no data regarding their diabetic status or only one abnormal value). The best performing algorithms within each strategy were tested on the validation set. RESULTS The standardized codes algorithm achieved a Kappa coefficient value of 0.59 (95% CI 0.49, 0.59) in the validation set. The Boolean logic algorithm reached 0.82 (95% CI 0.76, 0.88). A slightly higher value was achieved by the Feedforward Neural Network (0.9, 95% CI 0.85, 0.94). The best performing learner was the stacked generalization meta-learner that reached a Kappa coefficient value of 0.95 (95% CI 0.91, 0.98). CONCLUSIONS The stacked generalization strategy and the feedforward neural network showed the best classification metrics in the validation set. The implementation of these algorithms enables the exploitation of the data of thousands of patients accurately.

Deep Bidirectional Recurrent Neural Networks as End-To-End Models for Smoking Status Extraction from Clinical Notes in Spanish

Although natural language processing (NLP) tools have been available in English for quite some time, it is not the case for many other languages, particularly for context specific texts like clinical notes. This poses a challenge for tasks like text classification in languages other than English. In the absence of basic NLP tools, manually engineering features that capture semantic information of the documents is a potential solution. Nevertheless, it is very time consuming. Deep neural networks, particularly deep recurrent neural networks (RNN), have been proposed as End-to-End models that learn both features and parameters jointly, thus avoiding the need to manually encode the features. We compared the performance of two classifiers for labeling 14718 clinical notes in Spanish according to the patients’ smoking status: a bag-of-words model involving heavy manual feature engineering and a bidirectional long-short-term-memory (LSTM) deep recurrent neural network (RNN) with GloVe word embeddings. The RNN slightly outperforms the bag-of-words model, but with 80% less overall development time. Such algorithms can facilitate the exploitation of clinical notes in languages in which NLP tools are not as developed as in English.

Translation, transcultural adaptation, and validation of two questionnaires on shared decision making

To translate, transcultural adapt, and validate the “CollaboRATE” measure and the “Ask 3 Questions” intervention in Argentina, allowing us to quantify the degree of use and implementation of shared decision making (SDM).

The HUGE formula (hematocrit, urea, gender) for screening for chronic kidney disease in elderly patients: a study of diagnostic accuracy

Chronically reduced glomerular filtration rate (GFR) in old people does not always mean that they suffer from chronic kidney disease (CKD) since their GFR can just be reduced by aging. The HUGE equation has been recently described and validated in Spain for screening CKD without taking into account the patient’s GFR value. This equation is based on patient’s hematocrit, plasma urea levels and gender. The present study documented that the HUGE equation had and acceptable performance for screening CKD in elderly Argentine patients.

Validación psicométrica en español de la versión corta brasileña del cuestionario Primary Care Assessment Tools: usuarios para la evaluación de la orientación de los sistemas de salud hacia la atención primaria

OBJECTIVE To validate the Brazilian short version of the PCAT for adult patients in Spanish. DESIGN Analysis of secondary data from studies made to validate the extended version of the PCAT questionnaire. LOCATION City of Córdoba, Argentina. Primary health care. PARTICIPANTS The sample consisted of 46% of parents, whose children were enrolled in secondary education in three institutes in the city of Cordoba, and the remaining 54% were adult users of the National University of Cordoba Health Insurance. MAIN MEASURES Pearsons correlation coefficient comparing the extended and short versions. Goodness-of-fit indices in confirmatory factor analysis, composite reliability, average variance extracted, and Cronbachs alpha values, in order to assess the construct validity and the reliability of the short version. RESULTS The values of Pearsons correlation coefficient between this short version and the long version were high .818 (P<.001), implying a very good criterion validity. The indicators of good global adjustment to the confirmatory factor analysis were good. The value of composite reliability was good (.802), but under the variance media extracted: .3306, since 3 variables had weak factorials loads. The Cronbachs alpha was acceptable (.85). CONCLUSIONS The short version of the PCAT-users developed in Brazil showed an acceptable psychometric performance in Spanish as a quick assessment tool, in a comparative study with the extended version.

Excesivo rastreo de osteoporosis en mujeres menores de 65 años: estudio de corte transversal

Overuse of osteoporosis screening in women at low risk of fracture may lead to overdiagnosis, inappropriate treatment and medicalization. The objective of this work was to estimate the proportion of women aged 45 to 64 enrolled in a private health insurance plan in Buenos Aires undergoing hip dual-energy x-ray absorptiometry (DXA) in 2011 without meeting osteoporosis screening criteria. In this cross-sectional study, 4310 women of this age range that had undergone a hip DXA were identified. A randomly selected sub-group of 401 women was then assessed for the presence of risk factors for osteoporosis and complete data were retrieved for 178 women. Appropriate screening was defined by two criteria: 1) having a 10-year fracture risk higher than that of a 65-year old woman (estimated using the FRAX® tool); 2) having at least one risk factor for fracture. It was found that 86.5% of the women who underwent hip DXA did not exceed the minimum 10-year fracture risk threshold required for screening; 5.8% of them had osteoporosis and 61.0% osteopenia. According to the second criterion, 49.4% had no risk factors, 3.4% of these women had osteoporosis and 62.5% osteopenia. The results show that at least half the women screened did not meet osteoporosis screening criteria.