The study assesses and compares the effect of 'pulse oximetry' (PO) used by Lady Health Workers (LHWs) at household level on increasing hospital referral acceptance rates in intervention clusters (district Jamshoro) for 0-59 months old children with severe pneumonia with the effect of LHWs using clinical signs alone in non-intervention clusters of the same district.

Pneumonia accounts for an estimated 18% of under-5 mortality across the globe. Majority of these pneumonia-specific deaths occur in 15 countries, in which Pakistan ranks fifth. Failure to seek early care and delays in hospital referral are commonly acknowledged determinants of mortality in childhood pneumonia with a higher proportion reported from rural settings than urban. Acceptance rates of 'facilitated' hospital referral advice have been reported as low as 8-23% for sick young infants in peri-urban Karachi. They are even lower for non-facilitated referral in rural settings in children under five with severe pneumonia in rural settings.

Hypoxemia is an important sign of cardio-respiratory compromise in acutely ill children with a reported prevalence of 5.9%-58.9% in children aged 0-59 months from facilities and 16.1%-38.7% from community settings. Pulse oximetry (PO) is a rapid, portable, non-invasive, and accurate method of measuring arterial hemoglobin oxygenation (Sp02), and has therefore been used in the trial and clinical settings to detect hypoxemia. According to literature, mortality increases 4.3 times in children with pneumonia and hypoxemia than in those without hypoxemia.

The study hypothesizes that hypoxemia is a valuable sign of severe and very severe pneumonia in children 0-59 months and early identification at the community level followed by appropriate management with supplemental oxygen and antibiotics at the facility will improve acceptance of hospital referral and clinical outcomes

Study aims and design:

The overall aim of the study is to assess if detection of hypoxemia and/or severe pneumonia in children 0-59 months by LHWs during their monthly home visits will increase hospital referral acceptance among families in District Jamshoro, Sindh, Pakistan.

Primary objectives are:

1. To assess and compare the impact of 'pulse oximetry' used by LHWs at household level on increasing hospital referral acceptance rates in intervention clusters (district Jamshoro) for 0-59 months old children with severe pneumonia with the impact of LHWs using clinical signs alone in non-intervention clusters of the same district

2. To investigate the likely predictors (demographic, clinical) of hospital referral acceptance in both the groups

   Secondary objective is:

3. To compare clinical outcomes (treatment completion, treatment failure, hypoxemia) of children 0-59 months who accepted referral to those who refused admission and were treated at home

Sample size estimation:

Defining a cluster as (the catchment area of ) one Lady Health Worker (LHW), and assuming a power of 80% in detecting 15% increase in referral acceptance from a baseline 10% to 25% among 0-59 month old children with severe pneumonia (pneumonia prevalence at 2-week recall (MICS Sindh):7.5% (18% of which is assumed severe) a total of 116 clusters/LHWs will be selected from both intervention and control areas to capture 4160 children overall with severe pneumonia.

Study methods: An enhanced acute respiratory infections training module will be developed with training videos emphasizing classification of acute respiratory infections according to the new WHO algorithm, use of pulse oximeters to identify hypoxemia, identification of danger signs with addition of hypoxemia as a danger sign in intervention clusters, case management of pneumonia at home with oral amoxicillin and severe pneumonia with stat dose of antibiotic before referral to hospital. The faculty members of the Department of Pediatrics and Child Health at Aga Khan University Karachi, Pakistan will lead these training sessions with trainers of the LHW program.

Intervention delivery Children aged 0-59 months with cough and/or difficult breathing during regular home visit will be assessed by LHWs for signs and symptoms of severe pneumonia (fast breathing/chest in-drawing and one or more danger sign (unable to eat/drink, vomiting, convulsion and lethargy/unconsciousness) and/or Stridor)) and hypoxemia (SpO2 <92%) using a handheld pulse oximeter (Masimo Rad-5v) to measure blood oxygen level. LHWs will also do case management of children with pneumonia and severe pneumonia. A 3-day course of oral amoxicillin will be given to children with pneumonia at home, whereas children with severe pneumonia and or hypoxemia (eligible for recruitment) will be requested for informed consent and offered stat dose of oral amoxicillin and referral to nearest referral hospital (DHQ Kotri).

The study investigators will have provided these pulse oximeters to the LHW Program in advance and highlighted which ones will receive them. Physicians at the referral center serving the intervention clusters will also receive handheld pulse oximeters. All the LHWs and staff will be trained on the use, and maintenance of these pulse oximeters. Children with severe pneumonia with or without hypoxemia will be advised to go to hospital for antibiotics and oxygen, using the PO reading as a tool to convince parents. Name of the pre-designated health facility with available oxygen and study physician will be provided to all the LHWs so that follow up visits can be made to ensure the patient receives safe and recommended care at referral facility. Project staff will pretest and regularly monitor PO accuracy and quality of readings.

Hypoxemia will be defined as an arterial oxygen saturation (SpO2) <92%. SpO2 measurement will be recorded after 1 minute of stable observation. If the SpO2 comes 92% or less, the child will first be assessed for nasal obstruction with readings repeated after applying nasal saline drops. If repeat reading shows hypoxemia, the child will be referred to nearest designated referral hospital and admitted for oxygen via nasal or nasopharyngeal route and intravenous antibiotics, as per recommendations.

Implementation of active control: Clinical Signs assessment Children aged 0-59 months with cough and/or difficult breathing during regular home visit will be assessed by LHWs for signs and symptoms of severe pneumonia (fast breathing/chest in-drawing and one or more danger sign (unable to eat/drink, vomiting, convulsion and lethargy/unconsciousness) and/or Stridor)). A 3-day course of oral amoxicillin will be given to children with pneumonia at home, whereas children with severe pneumonia (eligible for recruitment) will be requested for informed consent and offered stat dose of oral amoxicillin and referral to nearest referral hospital.

Procedure at referral facility:

Children who accept hospital referral in both intervention and control clusters, and reach hospital premises with LHW referral slip will be assessed by study nurse/physician at the referral center. An SMS notification with brief details of referred child will have been provided to trained study personnel (study physician/nurse) in advance at time of referral at both the referral facilities. Children with severe pneumonia and/or hypoxemia as per LHWs who reach referral hospital premises will be examined and subjected to pulse oximetry again by the study physician at referral facility. If signs and symptoms of severe pneumonia is present, the child will be admitted for further appropriate treatment (Oxygen therapy via nasal or nasopharyngeal route and intravenous antibiotics etc.) and if the symptoms are not severe (absence of danger sign), the child will be treated in outpatient care as per the standard of referral facility. All the children admitted at referral facility will undergo 12 hourly monitoring by study personnel and filling of case reporting form (CRF) and hospital physician form (HPF) at day 1 of visit. Those children who refused the referral will be visited by study community health workers after 24 hours to confirm referral refusal and to fill CRF. All the children will be again be visited by the study staff at day 7 and 14 to collect data of follow-up form.

Inclusion Criteria:

1. Children of 0-59 months age
2. Signs of severe pneumonia (clinical signs only in control clusters, clinical signs and/or hypoxemia in intervention clusters)
3. Written informed consent by legal guardian

Exclusion Criteria:

1. Residents of out of catchment areas