Labor estimation by informational objective assessment (LEIOA) for preterm delivery prediction
- 50 Downloads
To introduce LEIOA, a new screening method to forecast which patients admitted to the hospital because of suspected threatened premature delivery will give birth in < 7 days, so that it can be used to assist in the prognosis and treatment jointly with other clinical tools.
From 2010 to 2013, 286 tocographies from women with gestational ages comprehended between 24 and 37 weeks were collected and studied. Then, we developed a new predictive model based on uterine contractions which combine the Generalized Hurst Exponent and the Approximate Entropy by logistic regression (LEIOA model). We compared it with a model using exclusively obstetric variables, and afterwards, we joined both to evaluate the gain. Finally, a cross validation was performed.
The combination of LEIOA with the medical model resulted in an increase (in average) of predictive values of 12% with respect to the medical model alone, giving a sensitivity of 0.937, a specificity of 0.747, a positive predictive value of 0.907 and a negative predictive value of 0.819. Besides, adding LEIOA reduced the percentage of incorrectly classified cases by the medical model by almost 50%.
Due to the significant increase in predictive parameters and the reduction of incorrectly classified cases when LEIOA was combined with the medical variables, we conclude that it could be a very useful tool to improve the estimation of the immediacy of preterm delivery.
KeywordsPremature delivery Quantitative diagnosis Approximate entropy Generalized Hurst exponent Logistic regression
This work was supported by the Basque Government (Grants PRE-2015-1-194 and IT974-16) and the Ferring S.A. Laboratories (Grant BC/A/16/004). Authors would also like to thank the Archive Service of Cruces University Hospital for their continuous help and effort on the sample acquisition.
IM: protocol/project development, data collection or management, data analysis, and manuscript writing/editing. LA: protocol/project development, data collection or management, data analysis, and manuscript writing/editing. LM: protocol/project development, data analysis, and manuscript writing/editing. LF-L: protocol/project development, data collection, or management. CB: data analysis and manuscript writing/editing. IMdlF: protocol/project development and manuscript writing/editing. MBP: data collection or management and data analysis. LG: data collection or management. IA: data collection or management. RM: protocol/project development, data collection or management, data analysis, and manuscript writing/editing.
Compliance with ethical standards
Conflict of interest
Authors report that they have no conflicts of interest.
- 1.Behrman R, Butler A (2007) Preterm birth. National Academies Press, Washington, D.C.Google Scholar
- 9.Hess WL, Mccaul JF, Perry KG, Howard PR, Morrison JC (1990) Correlation of uterine activity using the term guard monitor versus standard external tocodynamometry compared with the intrauterine pressure catheter. Obstet Gynecol 76:52–55Google Scholar
- 15.Malaina I, Matorras R, Martínez L, Fernandez-Llebrez L, Bringas C, Aranburu L et al (2016) Montevideo units vs autoregressive models on preterm labor detection. In: International work-conference on time series. ITISE proceedings 2016. http://itise.ugr.es/proceedings2016/Proceedings_ITISE2016.pdf. Accessed 27 June 2017
- 17.Hurst H (1951) Long term storage capacity of reservoirs. Trans Am Soc Civil Eng 6:770–799Google Scholar
- 23.Esposti F, Signorini M, Ferrario M, Magenes G (2006) Self-similarity behavior characterization of fetal heart rate signal in healthy and intrauterine grow retardated fetuses. In: 28th annual international conference of the IEEE Engineering in Medicine and Biology Society IEEE EMBS’06, pp 6157–6160Google Scholar
- 27.Verhulst PF (1838) Notice sur la loi que la population suit dans son accroissement. Corresp Math Phys 10:113–121Google Scholar