The investigators Study Hypothesis is introduction of modified IMCI guideline for managing severe pneumonia in first level health facilities will result in 40% increase in the appropriate management (appropriate case management at the first level facility and referral compliance by the caregivers) of severe pneumonia cases in the intervention arm compared to the comparison arm.

The IMCI was implemented in Bangladesh in 2001. Bangladesh adapted the generic IMCI clinical guidelines which recommend that children with one or more danger signs, or any other severe condition, as identified by the health worker of the first level facility will be referred urgently. Although the quality of care achieved in ten IMCI facilities in Matlab, Bangladesh was assessed as high ( El Arifeen 2004) the rates of compliance with referral found to be low. In 2002-03 the compliance rates were less than 37% and had decreased to 17% by the end of 2003 (El Arifeen S, unpublished data). In many settings, children who are referred do not actually attend hospital, which severely limits their appropriate care. This is mainly because the higher level facilities are geographically inaccessible; financial and social constraints also prevent care givers to comply with the referral advice.

The Govt of Bangladesh is presently planning to modify the current IMCI guideline. The modified guideline recommends that treatment of all severe pneumonia cases without danger signs unless there is another severe classification will be done at the first level facility. In 2004 the Govt. of Bangladesh in collaboration with ICCDR,B carried out a pilot study in rural Bangladesh (Chowdhury, El Arifeen et al. 2008) to assess the safety and effectiveness of a modified IMCI clinical management guideline for children with severe pneumonia aged 2months - 5 years. The study concluded that appropriate management of severe pneumonia at local level could not only save lives but represent a major cost decrease through reduction of referral, treatment and admission costs. Another research conducted in Pakistan, a randomized equivalency trial revealed that community management of pneumonia through home treatment with high-dose oral amoxicillin was just as effective as current hospital based treatments for severe pneumonia (Hazir et al. 2008). Although all of these findings are very encouraging, none of the study really focused on the immediate outcomes of the IMCI such as improved care seeking, adherence of the health providers and household caregivers to the algorithmic protocol and understanding the barriers and facilitators to the use of this protocol. Moreover, these studies were conducted in controlled research settings and therefore do not provide adequate understanding of the challenges and potential of implementing IMCI in routine health systems. So there is need to carry out research in actual community settings to understand what components of this modified guideline will work in 'real life situations' taking into account all the challenges and constraints faced by health systems of developing countries like Bangladesh.

The primary objectives of our study are to assess adherence of service providers to the modified IMCI guideline for managing severe pneumonia among under-5 children and care-givers' compliance with referral advice given by the health providers after initial care-seeking for under-5 children with severe pneumonia. The secondary objectives will be to measure the resulting changes in the care-seeking pattern for childhood pneumonia due to the introduction of modified IMCI guideline

Two MNCS programme upazilas will be purposively selected for this study in consultation with GoB and partners. The likely districts from where the two upazilas will be selected are Sherpur and Gopalganj. Twenty unions (average population of about 20,000) will be identified from the selected upazilas and randomly assigned to intervention or comparison arms. The intervention and evaluation will continue for 15 months. This will be a cluster randomized trial. Of the 20 selected study unions, 10 unions will be randomly assigned to intervention and 10 to comparison, that is, about 200,000 population (~40,000 households) in each arm. Each geographically delimited union has a Union Health and Family Welfare Centre (UHFWC) and family welfare centre (FWC), servicing the population residing within it. Health providers on the intervention arm will be trained regarding modification in the IMCI guideline. The training will not address any other aspect of the IMCI Training. Training on standard IMCI guideline will only be given to the health providers of both the arms, if not received before. Figure 1 (source: Chowdhury et al 2008) describes the modified and standard IMCI guideline. Under the modified guideline a sick child aged 2-59 months with severe pneumonia should be referred only if it accompanies with danger signs, stridor or any other sign or symptoms of other severe classification. All other children with severe pneumonia should be treated at the first level health facility with oral antibiotic (amoxicillin).

The unions belonging to the comparison arm will continue following the standard IMCI guideline. To measure the outcome of the interventions, the study will involve the following methods: Health facility survey including health providers' survey; extraction and review of health facility service records for provider compliance; routine surveillance for collection of information on care-seeking for childhood pneumonia; tracking and structured interview of severe pneumonia cases seen at facilities Details of the MNCS intervention package and delivery systems can be found in Annex I. All the MNCS upazilas have also already implemented facility-based IMCI.

All drugs in the study protocol are listed drugs of Bangladesh Government. The Study will ensure adequate and timely supplies of all drugs and other minimum essential logistics in all health facilities in the study area. The health facilities will be monitored on an ongoing basis by field supervisors, who will visit each facility once a month.

Inclusion Criteria:

• Severe pneumonia with patients of 2 months to 5 years
• Attending first level health care facilities
• In Selected study areas

Exclusion Criteria:

• Any other categories of pneumonia other than severe pneumonia.