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.

Endoscopic Anatomy of the Pterygopalatine Fossa and the Transpterygoid Approach: Development of a Surgical Instruction Model

Introduction: The pterygopalatine fossa (PPF) is a narrow space located between the posterior wall of the antrum and the pterygoid plates. Surgical access to the PPF is difficult because of its protected position and its complex neurovascular anatomy. Endonasal approaches using rod lens endoscopes, however, provide better visualization of this area and are associated with less morbidity than external approaches. Our aim was to develop a simple anatomical model using cadaveric specimens injected with intravascular colored silicone to demonstrate the endoscopic anatomy of the PPF. This model could be used for surgical instruction of the transpterygoid approach. Methods: We dissected six PPF in three cadaveric specimens prepared with intravascular injection of colored material using two different injection techniques. An endoscopic endonasal approach, including a wide nasoantral window and removal of the posterior antrum wall, provided access to the PPF. Results: We produced our best anatomical model injecting colored silicone via the common carotid artery. We found that, using an endoscopic approach, a retrograde dissection of the sphenopalatine artery helped to identify the internal maxillary artery (IMA) and its branches. Neural structures were identified deeper to the vascular elements. Notable anatomical landmarks for the endoscopic surgeon are the vidian nerve and its canal that leads to the petrous portion of the internal carotid artery (ICA), and the foramen rotundum, and V2 that leads to Meckel`s cave in the middle cranial fossa. These two nerves, vidian and V2, are separated by a pyramidal shaped bone and its apex marks the ICA. Conclusion: Our anatomical model provides the means to learn the endoscopic anatomy of the PPF and may be used for the simulation of surgical techniques. An endoscopic endonasal approach provides adequate exposure to all anatomical structures within the PPF. These structures may be used as landmarks to identify and control deeper neurovascular structures. The significance is that an anatomical model facilitates learning the surgical anatomy and the acquisition of surgical skills. A dissection superficial to the vascular structures preserves the neural elements. These nerves and their bony foramina, such as the vidian nerve and V2, are critical anatomical landmarks to identify and control the ICA at the skull base.

Programmed hypertension in rats treated with a NF-kappa B inhibitor during nephrogenesis: renal mechanisms

Suppression of the renin-angiotensin system (RAS) during murine lactation causes progressive renal injury, indicating a physiological action of angiotensin II on nephrogenesis. The nuclear factor NF-kappa B system is one of the main intracellular mediators of angiotensin II. We investigated whether inhibition of this system with pyrrolidine dithiocarbamate (PDTC) during rat nephrogenesis would lead to similar hypertension and renal injury as observed with RAS suppressors. Immediately after delivery, 32 Munich-Wistar dams, each nursing 6 male pups, were divided into 2 groups: C, untreated, and PDTC, receiving PDTC, 280 mg kg(-1) day(-1) orally, during 21 days. After weaning, the offspring were followed until 10 months of age without treatment. Adult rats that received neonatal PDTC exhibited stable hypertension and myocardial injury, without albuminuria. To gain additional insight into this process, the renal expression of RAS components and sodium transporters were determined by quantitative real-time PCR (qRT-PCR) at 3 and 10 months of life. Renal renin and angiotensinogen were upregulated at 3 and downregulated at 10 months of age, suggesting a role for early local RAS activation. Likewise, there was early upregulation of the proximal sodium/glucose and sodium/bicarbonate transporters, which abated later in life, suggesting that additional factors sustained hypertension in the long run. The conclusions drawn from the findings were as follows: (1) an intact NF-jB system during nephrogenesis may be essential to normal renal and cardiovascular function in adult life; (2) neonatal PDTC represents a new model of hypertension, lacking overt structural injury or functional impairment of the kidneys; and (3) hypertension in this model seems associated with early temporary activation of renal RAS and sodium transporters. Hypertension Research (2011) 34, 693-700; doi: 10.1038/hr. 2011.4; published online 17 February 2011

Right adolescent idiopathic thoracic curve (Lenke 1 A and B): does cost of instrumentation and implant density improve radiographic and cosmetic parameters?

In adolescent idiopathic scoliosis (AIS) there has been a shift towards increasing the number of implants and pedicle screws, which has not been proven to improve cosmetic correction. To evaluate if increasing cost of instrumentation correlates with cosmetic correction using clinical photographs. 58 Lenke 1A and B cases from a multicenter AIS database with at least 3 months follow-up of clinical photographs were used for analysis. Cosmetic parameters on PA and forward bending photographs included angular measurements of trunk shift, shoulder balance, rib hump, and ratio measurements of waist line asymmetry. Pre-op and follow-up X-rays were measured for coronal and sagittal deformity parameters. Cost density was calculated by dividing the total cost of instrumentation by the number of vertebrae being fused. Linear regression and spearman`s correlation were used to correlate cost density to X-ray and photo outcomes. Three independent observers verified radiographic and cosmetic parameters for inter/interobserver variability analysis. Average pre-op Cobb angle and instrumented correction were 54A degrees (SD 12.5) and 59% (SD 25) respectively. The average number of vertebrae fused was 10 (SD 1.9). The total cost of spinal instrumentation ranged from $6,769 to $21,274 (Mean $12,662, SD $3,858). There was a weak positive and statistically significant correlation between Cobb angle correction and cost density (r = 0.33, p = 0.01), and no correlation between Cobb angle correction of the uninstrumented lumbar spine and cost density (r = 0.15, p = 0.26). There was no significant correlation between all sagittal X-ray measurements or any of the photo parameters and cost density. There was good to excellent inter/intraobserver variability of all photographic parameters based on the intraclass correlation coefficient (ICC 0.74-0.98). Our method used to measure cosmesis had good to excellent inter/intraobserver variability, and may be an effective tool to objectively assess cosmesis from photographs. Since increasing cost density only improves mildly the Cobb angle correction of the main thoracic curve and not the correction of the uninstrumented spine or any of the cosmetic parameters, one should consider the cost of increasing implant density in Lenke 1A and B curves. In the area of rationalization of health care expenses, this study demonstrates that increasing the number of implants does not improve any relevant cosmetic or radiographic outcomes.

Modulation of B-1 and B-2 kinin receptors expression levels in the hippocampus of rats after audiogenic kindling and with limbic recruitment, a model of temporal lobe epilepsy

Epileptic seizures are hypersynchronous, paroxystic and abnormal neuronal discharges. Epilepsies are characterized by diverse mechanisms involving alteration of excitatory and inhibitory neurotransmission that result in hyperexcitability of the central nervous system (CNS). Enhanced neuronal excitability can also be achieved by inflammatory processes, including the participation of cytokines, prostaglandins or kinins, molecules known to be involved in either triggering or in the establishment of inflammation. Multiple inductions of audiogenic seizures in the Wistar audiogenic rat (WAR) strain are a model of temporal lobe epilepsy (TLE), due to the recruitment of limbic areas such as hippocampus and amygdata. In this study we investigated the modulation of the B-1 and B-2 kinin receptors expression levels in neonatal WARs as well as in adult WARs subjected to the TLE model. The expression levels of pro-inflammatory (IL-1 beta) and anti-inflammatory (IL-10) cytokines were also evaluated, as well as cyclooxygenase (COX-2). Our results showed that the B-1 and B-2 kinin receptors mRNAs were up-regulated about 7- and 4-fold, respectively, in the hippocampus of kindled WARs. On the other hand, the expressions of the IL-1 beta, IL-10 and COX-2 were not related to the observed increase of expression of kinin receptors. Based on those results we believe that the B, and B2 kinin receptors have a pivotal role in this model of TLE, although their participation is not related to an inflammatory process. We believe that kinin receptors in the CNS may act in seizure mechanisms by participating in a specific kininergic neurochemical pathway. (c) 2007 Elsevier B.V. All rights reserved.