Prevalence and predictors of hypoxaemia in hospitalised children at the emergency unit of a resource constrained cen

Abstract

Abstract:Background: Hypoxaemia is often poorly detected and treated in emergently-ill children in resource-poor centres because of the non-availability of pulse oximeters and similar facilities to detect it. This study sets out to determine the prevalence and simple predictors of hypoxaemia among children with or without respiratory features at the emergency unit of the Wesley Guild Hospital, Ilesa, Nigeria Methods: Children aged one month to 14 years were consecutively recruited and prospectively studied over an eight month period. All the children had their peripheral oxygen saturation (SpO 2 ) measured at presentation using a portable pulse oximeter (Nellcor (R) N-200, USA) and hypoxaemia was defined as SpO 2 < 90%. Relevant history and examination findings were compared among hypoxaemic and nonhypoxaemic children. Multivariate analysis was used to predict the presence of hypoxaemia. Results: Four hundred and two children were recruited with male to female ratio of 1.3:1 and105 (26.1) presented with respiratory features. Eighty three (20.6%) were hypoxaemic including 40 (38.1%) of those with respiratory features at admission. Infancy, chest in-drawing, cyanosis and grunting were associated with hypoxaemia (p < 0.05) among those with respiratory features, while infancy, pallor and tachycardia were significant among those with no respiratory features. Grunting (OR = 7.875; 95% CI=1.029-15.797; p = 0.045) and Cyanosis (OR =13.579; 95% CI = 1.360-14.379; p = 0.009) independently predict hypoxaemia among the children with respiratory features. Conclusion: Hypoxaemia occurred in approximately one out of five ill children admitted to the emergency unit of the WGH, Ilesa and was significantly associated with mortality. Emergently ill children with cyanosis and grunting especially infants should preferentially be placed on oxygen therapy even when hypoxaemia cannot be confirmed.

Hypoxaemia in Nigerian Children Presenting to the Children Emergency Ward (CHEW) of a Tertiary Hospital

“…The prevalence of hypoxaemia in this study was similar to that of 20.6% in a study in Ife, Nigeria [9] and 23.8% in Kanpur, India. [10] It is however higher than the 5.8%, 11.9% and 13% found in studies done in The Gambia, [11] Chandigarh, India [12] and Enugu, Nigeria [13] respectively, and lower than 41.5% and 73% seen in Ilorin, Nigeria 2 and Papua New Guinea respectively.…”

“…The significantly higher prevalence of hypoxaemia among infants (35.8%) compared to children above one year in this study is corroborated by similar studies among children in Enugu, [13] Ife, [9] The Gambia [11] and India. [12] This may be due to the fact that infants have a lower tidal volume and relative inefficient compensatory mechanisms to improve ventilation.…”

“…Chest in drawing was also significantly associated with presence of hypoxaemia in the study by Kuti and colleagues in Ile Ife, Nigeria. [9] In this study, grunting and wheezing had the highest positive predictive value in occurrence of hypoxaemia, also a study in Ibadan, Nigeria by Adebola and colleagues [16] showed that flaring and chest in drawing were predictive of hypoxaemia. The use of clinical signs in determination of hypoxaemia in ill children has met diverse views.…”

Epidemiology of Hypoxic Community-Acquired Pneumonia in Children Under 5 Years of Age: An Observational Study in Northern India

“…In our study, additional clinical features of pallor and wheezing were significantly associated with presence of hypoxic pneumonia. Pallor was also found to be significantly associated with hypoxic pneumonia in a study conducted in India (12) and also in a Nigerian study (41). Therefore, pallor must be included as an additional general danger sign of pneumonia.…”

Clinical emergency care quality indicators in Africa: a scoping review and data summary

“…The 53 studies containing ‘AFEM-CC quality indicators near match’ evidence were split between two groups of publications. The first group included 38 publications, which were identified as containing only ‘AFEM-CC quality indicators near match’ data 57–94. The second group included 15 publications, which were cited above as containing ‘AFEM-CC quality indicators exact match’ data, but which contained additional ‘AFEM-CC quality indicators near match’ data 22 25 26 28 29 31 33 34 39 41 45 46 54–56.…”