Developing and validating a prognostic index predicting re-hospitalization of patients with Hyperemesis Gravidarum

Zohar H. Morris, Abed N Azab, Shlomit Harlev, Ygal Plakht

Research output: Contribution to journalArticlepeer-review

4 Scopus citations

Abstract

Objective: Hyperemesis Gravidarum (HG) can have adverse effects on pregnant women's quality of life and is the main cause for hospitalization in the first half of the pregnancy. Between 19%–30% of the women admitted to hospital return for a second hospitalization during their pregnancy. These hospitalizations create a burden on the health system and are associated with high health care costs. The aim of this study was to develop a prediction tool that predicts re-hospitalization of women with HG and to validate it. Study design: A retrospective cohort study was conducted. The data was retrieved from the computerized information systems of two medical centers in Israel and included all women who were hospitalized between 2010 and 2012 with HG. Two thirds formed the training set and one third formed the validation set. Risk factors for re-hospitalization were determined based on a logistic regression model. Using the regression coefficient of each risk factor, a total score was calculated for each woman. The index was validated using logistic regression, c-statistic and Chi-square for trend test. Results: 282 women were included in the study, 37.6% of which were re-hospitalized. Three risk factors were included in the final index and were given a score: gestational age (2); length of hospitalization (1) and HG during previous pregnancies/first pregnancy (1). Each subject received a total score between 0 and 4. Re-hospitalization rates in the training set increased from 13.6% to 65% for scores 0 and 4, respectively. Validation testing showed: c-statistic-0.678 and Chi-square for trend, p < 0.001 (training set); c-statistic-0.622 and Chi-square for trend p = 0.036 (validation set). Conclusions: The developed index is a simple tool, based on few parameters and has a high level of accuracy. It potentially helps reduce hospitalization costs, assists in identifying women at risk of re-hospitalization, improves their treatment plan and prevents another admission to the hospital.

Original languageEnglish
Pages (from-to)113-117
Number of pages5
JournalEuropean Journal of Obstetrics, Gynecology and Reproductive Biology
Volume225
DOIs
StatePublished - 1 Jun 2018

Keywords

  • Hyperemesis Gravidarum
  • Prediction model
  • Re-hospitalization

Fingerprint

Dive into the research topics of 'Developing and validating a prognostic index predicting re-hospitalization of patients with Hyperemesis Gravidarum'. Together they form a unique fingerprint.

Cite this