Predictors of nocturnal oxyhaemoglobin desaturation in children with sickle cell disease

Kirkham, Fenella, Wilkey, O, Hewes, D.K.M and Evans, J.P.M (2008) Predictors of nocturnal oxyhaemoglobin desaturation in children with sickle cell disease. 13th Congress of the European Hematology Association, Copenhagen, Denmark. 11 - 14 Jun 2008.

Abstract

Background. Nocturnal oxyhaemoglobin desaturation (NOD, mean saturation <96% on overnight pulse oximetry) appears to be a predictor of central nervous system events and frequent pain in children with sickle cell disease (SCD). The risk factors are poorly understood.
Aims. To undertake a secondary analysis of potential risk factors (age, genotype, haematocrit previous chest crisis, comorbid asthma or central nervous system disease and upper airway infection) for NOD in the East London cohort.
Method. Of an unselected hospital cohort of children, 69 (median age 7.8 (range 1-16.5) years; 39 male) had overnight pulse oximetry at home between 1990 and 1995. Those over the age of 6 and any with CNS events were invited to have magnetic resonance imaging (MRI) and angiography (MRA). Clinical details, including emergency department attendances and hospital admissions, were obtained from the hospitalnotes for the period preceding the overnight study. A sleep questionnaire was administered to a subset. Logistic regression was used to look at predictors of NOD.
Results. For mean overnight saturation (SpO2), the range was 84-99.7% (median 96.1%) and values were significantly lower in patients with HbSS (median 95.4, range 84-99.7%) than in those with HbSC (median 97.9, range 94.8-98.9%) or Sb thalassaemia (median 97.8, range 93.5-99.5%). 25 children (36%) had mean saturation <96%. 13 had dips in saturation suggesting OSA, of whom 10 also had mean saturation <96%. 36 completed the sleep questionnaire, of whom 30 snored (14 sometimes, 3 often, 13 always); there was no evidence that NOD was associated with concurrent snoring (Fisher’s exact test, p=0.8). Of 11 children with abnormal overnight studies who had repeat overnight pulse oximetry after adenotonsillectomy, 8 were still abnormal. In univariate logistic regression, genotype (OR 9.00, 95% ci 1.08, 74.7; p=0.04), haemogloin at the time of the study (odd ratio, OR 0.64 95% confidence intervals, ci 0.47, 0.87; p=0.005), greater age at first Casualty attendance for an upper airways infection (p=0.02) and abnormal MRA (OR 4.5, 95%ci 1.24, 16.3; p=0.02) were associated with OD and there were trends for white cell count at the time of the study (OR 1.10, 95% ci 1.00, 1.21; p=0.06) and infarction on MRI (OR 3.14, 95% ci 0.67, 14.6; p=0.15). There was no effect of age at sleep study, haemoglobin F, doctor-diagnosed asthma, number of admissions with chest infection or any other measure of complication rate. Both haemoglobin at the time of the study (OR 0.68, 95% ci 0.48, 0.95; p=0.02) and age at first upper airway infection (OR 1.44, 95% ci 1.04, 1.99; p=0.02) remained in the multiple logistic regression model.
Summary and Conclusions. Nocturnal oxyhemoglobin desaturation is common in patients with HbSS, but is difficult to predict clinically and may not improve with adenotonsillectomy. Prospective studies of aetiology should commence in infancy and ideally include serial measurements of cerebral as well as lung function. If NOD is confirmed as a predictor of complications in SCD, treatment options might include hydroxyurea or overnight continuous positive airways pressure.

This record has no associated files available for download.

Contributors

Download statistics