Using data mining for prediction of hospital length of stay: an application of the CRISP-DM Methodology

Hospitals are nowadays collecting vast amounts of data related with patient records. All this data hold valuable knowledge that can be used to improve hospital decision making. Data mining techniques aim precisely at the extraction of useful knowledge from raw data. This work describes an implementation of a medical data mining project approach based on the CRISP-DM methodology. Recent real-world data, from 2000 to 2013, were collected from a Portuguese hospital and related with inpatient hospitalization. The goal was to predict generic hospital Length Of Stay based on indicators that are commonly available at the hospitalization process (e.g., gender, age, episode type, medical specialty). At the data preparation stage, the data were cleaned and variables were selected and transformed, leading to 14 inputs. Next, at the modeling stage, a regression approach was adopted, where six learning methods were compared: Average Prediction, Multiple Regression, Decision Tree, Arti���cial Neural Network ensemble, Support Vector Machine and Random Forest. The best learning model was obtained by the Random Forest method, which presents a high quality coe���cient of determination value (0.81). This model was then opened by using a sensitivity analysis procedure that revealed three in���uential input attributes: the hospital episode type, the physical service where the patient is hospitalized and the associated medical specialty. Such extracted knowledge con���rmed that the obtained predictive model is credible and with potential value for supporting decisions of hospital managers.

Contributos dos doutores palha��os da Opera����o Nariz Vermelho para a qualidade da adapta����o e do desenvolvimento em contexto pedi��trico: o olhar da crian��a e seus pais

Tese de doutoramento em Ci��ncias da Educa����o (��rea Especialidade em Psicologia da Educa����o)

Lewis Score - Prognostic value in patients with isolated small bowel Crohn's disease

Background and aims: Small bowel capsule endoscopy (SBCE) allows mapping of small bowel inflammation in Crohn���s disease (CD). We aimed to assess the prognostic value of the severity of inflammatory lesions, quantified by the Lewis score (LS), in patients with isolated small bowel CD. Methods: A retrospective study was performed in which 53 patients with isolated small bowel CD were submitted to SBCE at the time of diagnosis. The Lewis score was calculated and patients had at least 12 months of follow-up after diagnosis. As adverse events we defined disease flare requiring systemic corticosteroid therapy, hospitalization and/or surgery during follow-up. We compared the incidence of adverse events in 2 patient subgroups, i.e. those with moderate or severe inflammatory activity (LS =790) and those with mild inflammatory activity (135 = LS < 790). Results: The LS was =790 in 22 patients (41.5%), while 58.5% presented with LS between 135 and 790. Patients with a higher LS were more frequently smokers (p = 0.01), males (p = 0017) and under immunosuppressive therapy (p = 0.004). In multivariate analysis, moderate to severe disease at SBCE was independently associated with corticosteroid therapy during follow-up, with a relative risk (RR) of 5 (p = 0.011; 95% confidence interval [CI] 1.5���17.8), and for hospitalization, with an RR of 13.7 (p = 0 .028; 95% CI 1.3���141.9). Conclusion: In patients with moderate to severe inflammatory activity there were higher prevalences of corticosteroid therapy demand and hospitalization during follow-up. Thus, stratifying the degree of small bowel inflammatory activity with SBCE and LS calculation at the time of diagnosis provided relevant prognostic value in patients with isolated small bowel CD.

Aplica����o do paradigma Lean em processos cl��nicos do Centro Hospitalar do Alto Ave

Disserta����o de mestrado em Engenharia Industrial