In each study site, information will be systematically collected and summarized to describe the implementation of IMCI, the functioning of the health system, and specific events or activities that may affect child health or system functioning. Annex 1 lists proposed information to be collected; it will be incorporated into the data-collection tools at each level of the health system, or collected through special monitoring activities at study sites.

The evaluation impact indicators include measures of child mortality and nutritional status. Data to support the measurement of these indicators will be collected either through existing demographic surveillance systems or through specially-designed demographic surveys.

Sample surveys of households will be used to collect indicators of family behaviour related to nutrition (including breastfeeding), the use of insecticide-treated nets to prevent malaria, home care, care-seeking, compliance with recommended treatment and health services utilization. Other indicators measured through these surveys will include vaccination status of children under five, vitamin A supplementation coverage, assessment of weight for age, and determination of anaemia prevalence.

Descriptive information collected through household surveys includes socioeconomic information, access and utilization of health services, and costs to the household of child health services and drugs (see Section V). Statistical analyses will be performed within and across sites where appropriate.

The survey is composed of a set of data-collection activities that together will provide the information needed for the evaluation of first-level-facility services. These activities include an entry screening of presenting children, the observation of case management by a trained surveyor, a gold-standard re-examination of the child, an exit interview with the care-taker of the children who were observed during case management, an inventory of facility supplies, drugs and equipment, and a series of interviews with facility staff to obtain information on facility operations and costs.

Indicators measured at the first-level facility include those related to the quality of case management, the availability of health system support for IMCI, care-taker satisfaction, and selected indicators of utilization, care-seeking behaviour and clinic organization.

Descriptive information collected at first-level facilities includes details of facility services and service hours, availability of patient registers and drug records, information on the number and types of facility visits by children, staff time allocation, and additional information needed to estimate costs. Statistical analyses will be performed within and across sites where appropriate.

V. Measuring costs

The Economic Component of the Multi-Country Evaluation of IMCI Effectiveness, Cost and Impact (MCE) has two main goals. The first is to provide evidence on whether IMCI is of high, moderate, or low cost-effectiveness compared with other ways of using scarce health resources. This can be done either by determining the total costs and total health effects of providing IMCI services to under-five children or by estimating the additional costs of adding IMCI to current treatment practices and comparing them with to the additional health benefits that accrue.

The second goal is to provide health planners and donors with information on the cash expenditures (the financial costs) that were needed to introduce IMCI in the first place and then to keep it running.

The first goal will provide information to assist decision-makers in countries that are considering whether to implement or continue IMCI, by showing the extent to which IMCI is an efficient use of scarce health resources. The second type of analysis (called financial analysis) is useful in monitoring, planning or budgeting purposes in those countries.