Feasibility of introducing pulse oximetry for identifying hypoxaemia among children with pneumonia in paediatric outpatient settings in Bangladesh: Generating evidence and synthesising knowledge for influencing policy, programme planning and practice
Rahman, Ahmed Ehsanur
BACKGROUND: Pneumonia is the leading cause of childhood mortality, accounting for 16% of all under-5 deaths globally. Hypoxaemia is common among children with pneumonia and one of the strongest predictors of mortality. Since 2014, the World Health Organization has recommended introducing pulse oximetry for hypoxaemia identification and pneumonia classification in the Integrated Management of Childhood Illness (IMCI) services, which is a global strategy developed explicitly for outpatient management of common childhood illnesses, including pneumonia, in low-resource and high-burden settings by minimally trained health care providers. Unfortunately, there are few experiences of introducing pulse oximetry in paediatric outpatient settings and integrating it with IMCI services by adopting a health system strengthening approach. Bangladesh is one of the South Asian countries with high burdens of childhood pneumonia and hypoxaemia. Although Bangladesh has adopted the IMCI strategy and scaled up it nationally, pulse oximetry is neither recommended nor routinely used in IMCI services in Bangladesh. Successful introduction of a generic recommendation, technology, or device, like pulse oximetry, in routine services, demands an in-depth understanding of the problem and the context, followed by context-specific adaptations, demonstrations, and feasibility assessments. Also, it requires strategic and extensive engagement with policymakers and stakeholders to promote country ownership and government leadership, which are prerequisites for scalability and sustainability. OBJECTIVES AND METHOD: The overall goal of my PhD is to improve the management of childhood pneumonia by introducing and integrating pulse oximetry in routine IMCI services in Bangladesh. Furthermore, the aim is to support the Government of Bangladesh in taking an evidence-based decision in this regard. Hence, I was engaged in a series of discussions with the policymakers of the Ministry of Health and Family Welfare of the Government of Bangladesh to understand their perspectives on the existing evidence gaps and research priorities for making informed decisions regarding pulse oximetry integration. Based on these consultations, I identified my PhD objectives. RESULTS: A. Estimating the burden of hypoxaemia among children with pneumonia: I conducted a systematic review and meta-analysis by searching 11 bibliographic databases and citation indices. I reported pooled prevalence of hypoxaemia (SpO2<90%) by classification of clinical severity and by clinical settings by using the random-effects meta-analysis models. I identified 2,825 unique records from the databases, of which 57 studies met the eligibility criteria: 26 from Africa, 23 from Asia, four from South America, and four from multiple continents. The prevalence of hypoxaemia was 31% (95% CI, 26 to 36; 101,775 children) among all children with WHO-defined pneumonia, 41% (95% CI, 33 to 49; 30,483 children) among those with very severe or severe pneumonia, and 8% (95% CI, 3 to 16; 2,395 children) among those with non-severe pneumonia. The prevalence was much higher in studies conducted in emergency and inpatient settings than those conducted in outpatient settings. In 2019, we estimated that over 7 million children (95% UR, 5 to 8 million) were admitted to the hospital with hypoxaemic pneumonia. I also conducted a secondary analysis of data obtained from icddr,b-Dhaka Hospital, a secondary level referral hospital located in Dhaka, Bangladesh. I included 2,646 children aged 2-59 months admitted with WHO-defined severe pneumonia during 2014-17. On admission, the prevalence of hypoxaemia among children hospitalised with pneumonia was approximately 40% (95% CI, 38 to 42). Hypoxaemia was the strongest predictor of mortality (AOR = 11.1; 95% CI, 7.3 to 16.9) and referral (AOR = 5.9; 95% CI, 4.3 to 17.0) among other factors such as age, sex, history of fever and cough or difficulty in breathing, and severe acute malnutrition. Among those who survived, the median duration of hospital stay was 7 days (IQR, 4 to 11) in the hypoxaemic group and 6 days (IQR, 4 to 9) in the non-hypoxaemic group, and the difference was significant at p<0.001. B. Understanding the context of managing children with pneumonia, including hypoxaemia in Bangladesh: I conducted a secondary analysis using data from the 2017-18 round of the Bangladesh Demographic and Health Survey (BDHS), which adopts a nationally representative sample of households. I included 456 deaths among children under 5 years of age in our analysis. Descriptive statistics were used to present the causes, timing, and places of death with uncertainty ranges (UR). Pneumonia is the major killer (19%, 95% CI, 15.3 to 22.7), accounting for approximately 24,268 (UR, 21,626 to 26,695) under-5 deaths per year. Among children aged 1-11 months, pneumonia accounts for approximately 43% of deaths. I further conducted a secondary analysis of the Bangladesh Health Facility Survey 2017, which was conducted with a nationally representative sample including all administrative divisions and types of health facilities. More than 90% of the district hospital and sub-district hospitals and three-fourths of primary level health centres provide IMCI-based pneumonia management services. Pulse oximetry was available in 27% of the district hospitals, 18% of the sub-district level hospitals and none of the primary level health centres. Around 72% of the sub-district hospitals had the availability of one of any of the four oxygen sources (oxygen concentrators, filled oxygen cylinder with flowmeter, filled oxygen cylinder without flowmeter, and oxygen distribution system), followed by district hospitals (66%). Almost none of the primary level health centres had oxygen sources available on the day of the visit. C. Assessing the feasibility of introducing pulse oximetry in routine IMCI services: Based on literature review and expert consultations, I developed a conceptual framework, which guided the planning and implementation of a 4-step stakeholder engagement process for introducing pulse oximetry in routine IMCI services in Bangladesh. In the first step, a comprehensive desk review and key informant interviews were conducted to identify stakeholder organisations and score them based on their power and interest levels regarding IMCI implementation in Bangladesh. In the second step, two national level, two district level and five sub- district level sensitisation workshops were organised to orient all stakeholder organisations having high power or high interest regarding the importance of using pulse oximetry for pneumonia assessment and classification. In the third step, national and district level high power-high interest stakeholder organisations were involved in developing a joint action plan for introducing pulse oximetry in routine IMCI services. In the fourth step, led by a formal working group under the leadership of the Ministry of Health, we updated the National IMCI Implementation Package, including all guidelines, training manuals, services registers and referral forms in English and Bangla. Our engagement process contributed to the national decision to introduce pulse oximetry in paediatric outpatient settings and update the National IMCI Implementation Package demonstrating country ownership, government leadership and multi-partner involvement, which are steppingstones towards scalability and sustainability. However, our experience clearly delineates that stakeholder engagement is a context-driven, time-consuming, resource-intensive, iterative, and mercurial process that demands meticulous planning, prioritisation, inclusiveness, and adaptability. Based on WHO’s global recommendation in 2014, the National IMCI Programme of Bangladesh decided to introduce pulse oximetry in routine IMCI services in 2019 and developed a short training package for IMCI service providers. They decided to test the package in a relatively controlled setting for finalising the content and choice of pulse oximetry device before the demonstration in routine outpatient settings and subsequent scale-up. A cross-sectional study was conducted among children admitted to a rural district hospital. We employed 11 nurses and seven paramedics as assessors who received a one-day training on pulse oximetry. Each assessor performed at least 30 pulse oximetry measurements on children with two types of handheld devices. The assessors successfully established a stable SpO2 reading in all attempts (n=1478) except one. The median time taken was 30 seconds (IQR, 22 to 42), and within 60 seconds, 92% of attempts were successful. The median time was significantly (p<0.0001) higher among assessments conducted with a Lifebox device (36 seconds, IQR, 25 to 50) than those with a Masimo device (27 seconds, IQR, 20 to 35). Similarly, assessors aged >25 years are 4.8 (95% CI, 1.2 to 18.6) times more likely to obtain a stable reading within 60 seconds. Regarding patient-related factors, the odds of obtaining a stable SpO2 reading was 2.6 (95% CI, 1.6 to 4.2) times higher among children aged 12-59 months than among children aged 2-11 months. The National IMCI Programme of Bangladesh designed and developed a district implementation model for introducing pulse oximetry in IMCI services through stakeholder engagement and demonstrated the model in the Kushtia district by adopting a health system strengthening approach. Between December 2020 and June 2021, two assessment rounds were conducted based on WHO’s implementation research framework and outcome variables in 12 facilities involving 22 IMCI service providers and 1860 children presenting with cough/difficulty in breathing in the IMCI consultation rooms. WE OBSERVED THAT IMCI SERVICE PROVIDERS PERFORMED PULSE OXIMETRY ON ALMOST ALL ELIGIBLE CHILDREN, OF WHICH 99% OF ASSESSMENTS WERE SUCCESSFUL; 85% (95% CI, 83 TO 87) IN ONE ATTEMPT AND 69% (95% CI, 67 TO 71) WITHIN ONE MINUTE. The adherence to standards of procedures related to pulse oximetry was 92% (95% CI, 91 to 93), and agreement regarding identifying hypoxaemia was 96% (95% CI, 95 to 97). The median performance time was 36 seconds (IQR, 20 to 75), which was longer among younger children (2-11 months: 44 seconds, IQR, 22 to 78; 12-59 months: 30 seconds, IQR 18 to 53, p<0.001) and among those classified as pneumonia/severe pneumonia than as no pneumonia (41 seconds, IQR, 22 to 70; 32 seconds, IQR, 20 to 62, p<0.001). We observed improvements in all indicators in the second round of assessments. Caregivers showed positive attitudes towards using this novel technology for the assessment of children. CONCLUSION: Based on context-specific experience generated through these studies, the Government of Bangladesh decided to integrate pulse oximetry into routine IMCI services throughout Bangladesh. Furthermore, the learnings synthesised through these studies can also help convince the policymakers and managers of other LMICs with similar burdens and contexts to introduce pulse oximetry in routine settings providing outpatient-based paediatric services and contribute to achieving the target of averting all preventable childhood pneumonia deaths by 2025.