Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia

Introduction. Iron deficiency anemia and thalassemia are the most common causes of microcytic anemia. Powerful statistical computer programming enables sensitive discriminant analyses to aid in the diagnosis. We aimed at investigating the performance of the multiple discriminant analysis (MDA) to th...

Full description

Bibliographic Details
Main Authors: Eloísa Urrechaga, Urko Aguirre, Silvia Izquierdo
Format: Article
Language:English
Published: Hindawi Limited 2013-01-01
Series:Anemia
Online Access:http://dx.doi.org/10.1155/2013/457834
id doaj-0aef8ff7c807480db25f10d0acabccd4
record_format Article
spelling doaj-0aef8ff7c807480db25f10d0acabccd42020-11-24T22:44:11ZengHindawi LimitedAnemia2090-12672090-12752013-01-01201310.1155/2013/457834457834Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic AnemiaEloísa Urrechaga0Urko Aguirre1Silvia Izquierdo2Laboratorio, Hospital Galdakao-Usansolo, 48960 Galdakao, Vizcaya, SpainUnidad de Investigación CIBER Epidemiología y Salud Pública, 48960 Galdakao, Vizcaya, SpainGenética Clínica, Servicio de Bioquímica Clínica, Hospital Universitario Miguel Servet, 50009 Zaragoza, SpainIntroduction. Iron deficiency anemia and thalassemia are the most common causes of microcytic anemia. Powerful statistical computer programming enables sensitive discriminant analyses to aid in the diagnosis. We aimed at investigating the performance of the multiple discriminant analysis (MDA) to the differential diagnosis of microcytic anemia. Methods. The training group was composed of 200 β-thalassemia carriers, 65 α-thalassemia carriers, 170 iron deficiency anemia (IDA), and 45 mixed cases of thalassemia and acute phase response or iron deficiency. A set of potential predictor parameters that could detect differences among groups were selected: Red Blood Cells (RBC), hemoglobin (Hb), mean cell volume (MCV), mean cell hemoglobin (MCH), and RBC distribution width (RDW). The functions obtained with MDA analysis were applied to a set of 628 consecutive patients with microcytic anemia. Results. For classifying patients into two groups (genetic anemia and acquired anemia), only one function was needed; 87.9% β-thalassemia carriers, and 83.3% α-thalassemia carriers, and 72.1% in the mixed group were correctly classified. Conclusion. Linear discriminant functions based on hemogram data can aid in differentiating between IDA and thalassemia, so samples can be efficiently selected for further analysis to confirm the presence of genetic anemia.http://dx.doi.org/10.1155/2013/457834
collection DOAJ
language English
format Article
sources DOAJ
author Eloísa Urrechaga
Urko Aguirre
Silvia Izquierdo
spellingShingle Eloísa Urrechaga
Urko Aguirre
Silvia Izquierdo
Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
Anemia
author_facet Eloísa Urrechaga
Urko Aguirre
Silvia Izquierdo
author_sort Eloísa Urrechaga
title Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
title_short Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
title_full Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
title_fullStr Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
title_full_unstemmed Multivariable Discriminant Analysis for the Differential Diagnosis of Microcytic Anemia
title_sort multivariable discriminant analysis for the differential diagnosis of microcytic anemia
publisher Hindawi Limited
series Anemia
issn 2090-1267
2090-1275
publishDate 2013-01-01
description Introduction. Iron deficiency anemia and thalassemia are the most common causes of microcytic anemia. Powerful statistical computer programming enables sensitive discriminant analyses to aid in the diagnosis. We aimed at investigating the performance of the multiple discriminant analysis (MDA) to the differential diagnosis of microcytic anemia. Methods. The training group was composed of 200 β-thalassemia carriers, 65 α-thalassemia carriers, 170 iron deficiency anemia (IDA), and 45 mixed cases of thalassemia and acute phase response or iron deficiency. A set of potential predictor parameters that could detect differences among groups were selected: Red Blood Cells (RBC), hemoglobin (Hb), mean cell volume (MCV), mean cell hemoglobin (MCH), and RBC distribution width (RDW). The functions obtained with MDA analysis were applied to a set of 628 consecutive patients with microcytic anemia. Results. For classifying patients into two groups (genetic anemia and acquired anemia), only one function was needed; 87.9% β-thalassemia carriers, and 83.3% α-thalassemia carriers, and 72.1% in the mixed group were correctly classified. Conclusion. Linear discriminant functions based on hemogram data can aid in differentiating between IDA and thalassemia, so samples can be efficiently selected for further analysis to confirm the presence of genetic anemia.
url http://dx.doi.org/10.1155/2013/457834
work_keys_str_mv AT eloisaurrechaga multivariablediscriminantanalysisforthedifferentialdiagnosisofmicrocyticanemia
AT urkoaguirre multivariablediscriminantanalysisforthedifferentialdiagnosisofmicrocyticanemia
AT silviaizquierdo multivariablediscriminantanalysisforthedifferentialdiagnosisofmicrocyticanemia
_version_ 1725692455425146880