Randomised Exposure Study of Pollution Indoors and Respiratory Effects (RESPIRE): the effect of reducing exposure to smoke from traditional woodstoves on child pneumonia in rural Guatemala

Physician-diagnosed pneumonia in children through 18 months of age. Pneumonia case finding and management were based on the Integrated Management of Childhood Illness (IMCI) Strategy, developed by World Health Organization (WHO)/United Nations Children's Fund (UNICEF). This approach was selected because:
1. Experience has demonstrated that locally recruited field staff can be trained to assess ALRI and other common conditions using this method
2. In terms of case definition, it is sufficiently valid and consistent with most other recent population-based field studies of ALRI
3. There are well developed training materials and procedures for both case finding and management
4. It integrates case finding with clinically effective management of sick children

Child health outcomes were assessed through weekly home visits by trained field workers, with referral to study physicians working in local community centres to maintain blindness to interventions status for clinical examination. Follow-up of children were from age at initial recruitment (birth to 4 months) until they reached 18 months of age. The fieldworkers, trained in IMCI, assessed and identified all children with rapid breathing, in addition to other signs of ALRI and the WHO-defined general danger signs. The rationale for carrying out the home visits at weekly intervals was based on the duration of pneumonia, in particular the duration of visible signs (principally fast breathing and chest indrawing). Estimates of this vary and depend on definitions of when an episode finishes. Weekly visits allowed almost all cases of ALRI to be identified through direct observation of the sick child, with a minority by mother's recall of the most relevant signs. Maternal recall of these signs does not have a high sensitivity and specificity, and was kept to a minimum.

Children meeting criteria for possible ALRI (or other significant illness) were referred to one of the study doctors who carried out a clinical examination including pulse oximetry. All children with a diagnosis of pneumonia were offered a direct antigen test for RSV, and a chest X-ray). Chest X-rays were carried out initially at San Marcos Hospital and subsequently at a private clinic to ensure quality control, normally within 24 - 48 hours of the initial assessment by the study doctor, subject to parental consent. All X-rays images were stored (scanned), and read by independent reviewers, blind to intervention status, using WHO training software and procedures. Children meeting the criteria for referral to hospital were admitted (subject to parental consent) to San Marcos hospital. All children admitted with suspected pneumonia received a chest X-ray, as per usual practice at this hospital. All children diagnosed with ALRI or other severe conditions by study MDs were offered free treatment under normal medical protocols.

Definition of new episodes:
New episodes were defined, drawing on current practice for ALRI field studies, as follows:
1. A sick child meeting the criteria for ALRI less than 15 days after apparent recovery from an earlier episode is regarded as a relapse of the first episode
2. A sick child meeting the ALRI criteria 15 or more days after apparent recovery from an earlier episode is regarded as a new episode

Severity:
The severity of the illness as determined by field workers was categorised according to the symptoms and signs included in the IMCI algorithm. Severe FW-cases were those with chest indrawing and/or the most reliably described general danger sign, that is inability to breast-feed or drink. Severity of all physician-diagnosed outcomes was determined by the level of oxygen saturation measured by pulse oximetry. A cut off of 87% was taken, based on the established definition of this being 2 standard deviations below the mean level for a representative group of well children in the study sample. The relatively low level reflects the altitude of the study site (mean 2,600 metres).

Derived from the case finding methods described above:
1. ARI outcomes based on field worker assessments in the home:
1.1. Acute upper respiratory infections
1.2. All ALRI
1.3. Severe ALRI
2. ARI outcomes based on physician diagnosis and investigations:
2.1. Severe (hypoxaemic) physician-diagnosed pneumonia
2.2. All RSV positive physician-diagnosed pneumonia
2.3. Severe (hypoxaemic) RSV positive physician-diagnosed pneumonia
2.4. All RSV negative physician-diagnosed pneumonia
2.5. Severe (hypoxaemic) RSV negative physician-diagnosed pneumonia
2.6. All chest X-ray positive physician-diagnosed pneumonia
2.7. Severe (hypoxaemic) chest X-ray positive physician-diagnosed pneumonia
3. Outcomes based on verbal autopsies:
3.1. All deaths
3.2. Deaths due to pneumonia
4. Other child health outcomes:
4.1. Diarrhoea
4.2. Birth weight
4.3. Burns and scalds