Success and time implications of SpO₂ measurement through pulse oximetry among hospitalised children in rural Bangladesh: Variability by various device-, provider- and patient-related factors

Abstract

Background Hypoxaemia is one of the strongest predictors of mortality among children with pneumonia. It can be identified through pulse oximetry instantaneously, which is a non-invasive procedure but can be influenced by factors related to the specific measuring device, health provider and patient. Following WHO's global recommendation in 2014, Bangladesh decided to introduce pulse oximetry in paediatric outpatient services, ie, the Integrated Management of Childhood Illness (IMCI) services in 2019. A national committee updated the existing IMCI implementation package and decided to test it by assessing the pulse oximetry performance of different types of assessors in real-life inpatient settings. Methods We adopted an observational design and conducted a technology assessment among children admitted to a rural district hospital. Eleven nurses and seven paramedics received one-day training on pulse oximetry as assessors. Each assessor performed at least 30 pulse oximetry measurements on children with two types of handheld devices. The primary outcome of interest was obtaining a successful measurement of SpO2, defined as observing a stable (±1%) reading for at least 10 seconds. Performance time, ie, time taken to obtain a successful measurement of SpO 2 was considered the secondary outcome of interest. In addition, we used Generalized Estimating Equation to assess the effect of different factors on the pulse oximetry performance. Results The assessors obtained successful measurements of SpO 2 in all attempts (n = 1478) except one. The median time taken was 30 (interquartile range (IQR) = 22-42) seconds, and within 60 seconds, 92% of attempts were successful. The odds of obtaining a successful measurement within 60 seconds were 7.3 (95% confidence interval (CI) = 3.7-14.2) times higher with a Masimo device than a Lifebox device. Similarly, assessors aged >25 years were 4.8 (95% CI = 1.2, 18.6) times more likely to obtain a successful measurement within 60 seconds. The odds of obtaining a successful measurement was 2.6 (95% CI = 1.6, 4.2) times higher among children aged 12-59 months compared to 2-11 months. Conclusions Our study indicated that assessors could achieve the necessary skills to perform pulse oximetry successfully in real-life inpatient settings through a short training module, with some effect of device-, provider- and patient-related factors. The National IMCI Programme of Bangladesh can use these findings for finalising the national IMCI training modules and implementation package incorporating the recommendation of using pulse oximetry for childhood pneumonia assessment.

Global Emergency Medicine: A Scoping Review of the Literature from 2022

“…An observational study from Bangladesh found that after a brief training, health care personnel can obtain a successful pulse oximetry measurement in 60 seconds in children aged <5 years 62 . Razzak et al 63 .…”

“…An observational study from Bangladesh found that after a brief training, health care personnel can obtain a successful pulse oximetry measurement in 60 seconds in children aged <5 years. 62 Razzak et al 63 predicted that implementing a complete trauma system in LMICs could save more than 200,000 lives annually. A systematic review of 2174 articles by Sessions et al 64 evaluated the impact of continuous positive airway pressure on mortality and adverse effects among pediatric patients with respiratory distress and found 20 studies of variable evidence.…”

Exploring the Feasibility of Vision-Based Non-Contact Oxygen Saturation Estimation: Considering Critical Color Components and Individual Differences

“…To remove these unwanted Using the RGB signals and the transformed YCrCgCb and HSV signals, a dataset of size N × 10 was generated, where N represents the number of frames in the video. In this paper, considering that it takes a minimum of 30 s to successfully measure oxygen saturation [23], signals from 10, 20, and 30 s were used, resulting in dataset lengths of 300, 600, and 900, respectively.…”