Variability of Prognostic Results Based on Biological Parameters in Sickle Cell Disease Cohort Studies in Children: What Should Clinicians Know?
Open Access
- 12 February 2021
- Vol. 8 (2), 143
- https://doi.org/10.3390/children8020143
Abstract
Background: Many pediatric studies describe the association between biological parameters (BP) and severity of sickle cell disease (SCD) using different methods to collect or to analyze BP. This article assesses the methods used for collection and subsequent statistical analysis of BP, and how these impact prognostic results in SCD children cohort studies. Methods: Firstly, we identified the collection and statistical methods used in published SCD cohort studies. Secondly, these methods were applied to our cohort of 375 SCD children, to evaluate the association of BP with cerebral vasculopathy (CV). Results: In 16 cohort studies, BP were collected either once or several times during follow-up. The identified methods in the statistical analysis were: (1) one baseline value per patient (2) last known value; (3) mean of all values; (4) modelling of all values in a two-stage approach. Applying these four different statistical methods to our cohort, the results and interpretation of the association between BP and CV were different depending on the method used. Conclusion: The BP prognostic value depends on the chosen statistical analysis method. Appropriate statistical analyses of prognostic factors in cohort studies should be considered and should enable valuable and reproducible conclusions.Keywords
This publication has 33 references indexed in Scilit:
- Impact of early transcranial Doppler screening and intensive therapy on cerebral vasculopathy outcome in a newborn sickle cell anemia cohortBlood, 2011
- Endogenous nitric oxide synthase inhibitors in sickle cell disease: abnormal levels and correlations with pulmonary hypertension, desaturation, haemolysis, organ dysfunction and deathBritish Journal of Haematology, 2009
- G6PD deficiency, absence of α-thalassemia, and hemolytic rate at baseline are significant independent risk factors for abnormally high cerebral velocities in patients with sickle cell anemiaBlood, 2008
- Joint modelling of repeated measurements and time‐to‐event outcomes: The fourth Armitage lectureStatistics in Medicine, 2007
- A Joint Latent Class Changepoint Model to Improve the Prediction of Time to Graft FailureJournal of the Royal Statistical Society Series A: Statistics in Society, 2007
- Sickle cell leg ulcers: associations with haemolysis and SNPs in Klotho, TEK and genes of the TGF‐β/BMP pathwayBritish Journal of Haematology, 2006
- WHO Child Growth Standards based on length/height, weight and ageActa Paediatrica, 2006
- Modeling the Relationship of Survival to Longitudinal Data Measured with Error. Applications to Survival and CD4 Counts in Patients with AIDSJournal of the American Statistical Association, 1995
- Fetal haemoglobin and early manifestations of homozygous sickle cell disease.Archives of Disease in Childhood, 1992
- Percentile curves for hemoglobin and red cellvolume in infnacy and childhoodThe Journal of Pediatrics, 1979