A decision support system to facilitate management of patients with acute gastrointestinal bleeding

Objective: To develop a model to predict the bleeding source and identify the cohort amongst patients with acute gastrointestinal bleeding (GIB) who require urgent intervention, including endoscopy. Patients with acute GIB, an unpredictable event, are most commonly evaluated and managed by non-gastroenterologists. Rapid and consistently reliable risk stratification of patients with acute GIB for urgent endoscopy may potentially improve outcomes amongst such patients by targeting scarce health-care resources to those who need it the most. Design and methods: Using ICD-9 codes for acute GIB, 189 patients with acute GIB and all. available data variables required to develop and test models were identified from a hospital medical records database. Data on 122 patients was utilized for development of the model and on 67 patients utilized to perform comparative analysis of the models. Clinical data such as presenting signs and symptoms, demographic data, presence of co-morbidities, laboratory data and corresponding endoscopic diagnosis and outcomes were collected. Clinical data and endoscopic diagnosis collected for each patient was utilized to retrospectively ascertain optimal management for each patient. Clinical presentations and corresponding treatment was utilized as training examples. Eight mathematical models including artificial neural network (ANN), support vector machine (SVM), k-nearest neighbor, linear discriminant analysis (LDA), shrunken centroid (SC), random forest (RF), logistic regression, and boosting were trained and tested. The performance of these models was compared using standard statistical analysis and ROC curves. Results: Overall the random forest model best predicted the source, need for resuscitation, and disposition with accuracies of approximately 80% or higher (accuracy for endoscopy was greater than 75%). The area under ROC curve for RF was greater than 0.85, indicating excellent performance by the random forest model Conclusion: While most mathematical models are effective as a decision support system for evaluation and management of patients with acute GIB, in our testing, the RF model consistently demonstrated the best performance. Amongst patients presenting with acute GIB, mathematical models may facilitate the identification of the source of GIB, need for intervention and allow optimization of care and healthcare resource allocation; these however require further validation. (c) 2007 Elsevier B.V. All rights reserved.

Enterohormonal changes after digestive adaptation: Five-year results of a surgical proposal to treat obesity and associated diseases

Background Recent physiological knowledge allows the design of bariatric procedures that aim at neuroendocrine changes instead of at restriction and malabsorption. Digestive adaptation is a surgical technique for obesity based in this rationale. Methods The technique includes a sleeve gastrectomy, an omentectomy and a jejunectomy that leaves initial jejunum and small bowel totaling at least 3 m (still within normal variation of adult human bowel length). Fasting ghrelin and resistin and fasting and postprandial GLP-1 and PYY were measured pre- and postoperatively. Results Patients: 228 patients with initial body mass index (BMI) varying from 35 to 51 kg/m(2); follow-up: I to 5 years; average EBMIL% was 79.7% in the first year; 77.7% in the second year; 71.6% in the third year; 68.9% in the fourth year. Patients present early satiety and major improvement in presurgical comorbidities, especially diabetes. Fasting ghrelin and resistin were significantly reduced (P<0.05); GLP-1 and PYY response to food ingestion was enhanced (P<0.05). Surgical complications (4.4%) were resolved without sequela and without mortality. There was neither diarrhea nor detected malabsorption. Conclusions Based on physiological and supported by evolutionary data, this procedure creates a proportionally reduced gastrointestinal (GI) tract that amplifies postprandial neuroendocrine responses. It leaves basic GI functions unharmed. It reduces production of ghrelin and resistin and takes more nutrients to be absorbed distally enhancing GLP-1 and PYY secretion. Diabetes was improved significantly without duodenal exclusion. The patients do not present symptoms nor need nutritional support or drug medication because of the procedure, which is safe to perform.

Reliability assessment of Cobb angle measurements using manual and digital methods

BACKGROUND CONTEXT: The vertebral spine angle in the frontal plane is an important parameter in the assessment of scoliosis and may be obtained from panoramic X-ray images. Technological advances have allowed for an increased use of digital X-ray images in clinical practice. PURPOSE: In this context, the objective of this study is to assess the reliability of computer-assisted Cobb angle measurements taken from digital X-ray images. STUDY DESIGN/SETTING: Clinical investigation quantifying scoliotic deformity with Cobb method to evaluate the intra- and interobserver variability using manual and digital techniques. PATIENT SAMPLE: Forty-nine patients diagnosed with idiopathic scoliosis were chosen based on convenience, without predilection for gender, age, type, location, or magnitude of the curvature. OUTCOME MEASURES: Images were examined to evaluate Cobb angle variability, end plate selection, as well as intra- and interobserver errors. METHODS: Specific software was developed to digitally reproduce the Cobb method and calculate semiautomatically the degree of scoliotic deformity. During the study, three observers estimated the Cobb angle using both the digital and the traditional manual methods. RESULTS: The results showed that Cobb angle measurements may be reproduced in the computer as reliably as with the traditional manual method, in similar conditions to those found in clinical practice. CONCLUSIONS: The computer-assisted method (digital method) is clinically advantageous and appropriate to assess the scoliotic curvature in the frontal plane using Cobb method. (C) 2010 Elsevier Inc. All rights reserved.