Master's Dissertation
DOI
https://doi.org/10.11606/D.5.2018.tde-01112018-103744
Document
Author
Full name
Ana Cláudia Rodrigues Lopes Amaral de Souza
E-mail
Institute/School/College
Knowledge Area
Date of Defense
Published
São Paulo, 2018
Supervisor
Committee
Francisco, Rossana Pulcineli Vieira (President)
Brizot, Maria de Lourdes
Calderon, Iracema de Mattos Paranhos
Moisés, Elaine Christine Dantas
Title in Portuguese
Modelo de predição para o uso de insulina em gestantes diagnosticadas com diabetes gestacional pela glicemia de jejum
Keywords in Portuguese
Diabetes gestacional
Glicemia de jejum
Gravidez
Insulina
Modelos logísticos
Nomogramas
Abstract in Portuguese
Objetivo: Avaliar os fatores de risco e propor um modelo para a predição da necessidade de insulina durante o tratamento de Diabetes Mellitus Gestacional (DMG) diagnosticado precocemente. Métodos: coorte retrospectiva de gestantes que foram diagnosticadas com DMG pela glicemia de jejum (GJ) anormal na primeira visita pré-natal e que receberam atendimento pré-natal em um hospital terciário de ensino em São Paulo, Brasil, entre 2012 e 2015. De acordo com a necessidade de insulinoterapia para atingir os alvos glicêmicos, as gestantes foram divididas em dois grupos (grupo Insulina ou grupo Dieta) e comparadas quanto a variáveis clínicas e laboratoriais. O desempenho dessas variáveis na predição da necessidade de insulina para o tratamento do DMG foi identificado por modelo de regressão logística, e um nomograma foi criado para facilitar a interpretação clínica. Resultados: No total, foram incluídas 408 mulheres para análise. Entre elas, 135 (33%) necessitaram de terapia com insulina. No modelo de regressão logística, idade materna, índice de massa corporal prégestacional, valor da GJ, história de DMG anterior e história familiar de diabetes foram variáveis independentes significativas para a predição da necessidade de insulina. Conclusão: O modelo de predição elaborado permitiu a construção de um nomograma e uma calculadora digital de fácil uso clínico para avaliar a necessidade de insulinoterapia em mulheres com diagnóstico precoce de DMG
Title in English
Prediction model for insulin need in women diagnosed with gestational diabetes by fasting blood glucose
Keywords in English
Fasting blood glucose
Gestational diabetes
Insulin
Logistic
models
Nomograms
Pregnancy
Abstract in English
Objective: To evaluate risk factors and to propose a model for the prediction of insulin requirement during the treatment of early-diagnosed gestational diabetes mellitus (GDM). Methods: Retrospective cohort with chart review of pregnant women receiving antenatal care at a tertiary teaching hospital in São Paulo, Brazil, who were diagnosed with GDM by abnormal fasting blood glucose (FBG) at the first prenatal visit between 2012 and 2015. Groups were divided according to the requirement for insulin therapy to achieve blood glucose targets (insulin or diet group), the subjects were compared regarding clinical and laboratory variables. The performance of these variables in predicting insulin need for GDM treatment was identified by a logistic regression model, and a nomogram was created to facilitate clinical interpretation. Results: In total, 408 women were included for analysis. Among them, 135 (33%) needed insulin therapy. In the logistic regression model, maternal age, pre-pregnancy body mass index, FBG value, prior GDM and family history of diabetes were significant independent variables for the prediction of insulin need. Conclusion: The prediction model found allowed the construction of a nomogram and digital calculator that is easy to use in order to evaluate the need for insulin therapy in women with early diagnosis of GDM
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
2018-11-01