Supporting information for: Evidence-Based Estimates of the Expected Duration of Initial Consultations for Sick Children in Low- and Middle-income Countries Using the Integrated Management of Childhood Illness Clinical Algorithm

Few studies have examined the time required for primary care consultations; none have focused on sick child visits in low- and middle-income countries (LMICs). This project begins to fill that gap by providing evidence-based estimates of the time needed for initial visits with under-five infants and children at public or not-for-profit facilities in countries using the Integrated Management of Childhood Illness (IMCI) clinical algorithm.

Estimates of the mean expected duration of IMCI consultations require (a) classification profiles, i.e., tabulations of the gold standard health issues presented by patients less than 5 years old; (b) lists of the tasks included in applicable versions of the IMCI algorithm and the conditions that elicit them, and (c) an estimate of the time needed to perform tasks with no pre-defined minimum duration. The latter requires, in addition to classification profiles, information on rates of task performance and the mean observed duration of consultations.

The IMCI clinical algorithm and the research surrounding it provide unusually rich sources of such information. Developed in the mid 1990s by the World Health Organization and the United Nations Children’s Fund, the IMCI algorithm seeks to reduce child mortality in LMICs by improving the technical quality of primary care services. For infants less than 2 months old, the algorithm focuses on bacterial infections, feeding problems, low weight, and, in some versions, jaundice. For children 2-59 months old, the foci include acute respiratory infections, especially pneumonia; diarrhea; fevers, especially malaria and measles; malnutrition, and anemia. Immunization status is a concern for both age groups. The algorithm provides a scheme to classify the health issues with which infants and children present, an array of tasks providers may be expected perform, and criteria by which tasks are elicited. Research on the design and utility of the algorithm, its effects on provider performance, and related topics furnishes data on the prevalence of gold standard IMCI classifications in a variety of patient populations. In some cases, it also enables one to calculate the time required to perform tasks.

I found such information by searching MEDLINE, the database of the International Network for Rational Use of Medicines, the websites of the WHO and its regional offices, GOOGLE, and GOOGLE SCHOLAR using search terms such as ‘Integrated Management of Childhood Illness’, ‘observational’, ‘prospective’, ‘classification’, ‘clinical signs’, ‘health facility survey’, and ‘validity’. I also reviewed studies that cited a qualified study and, conversely, material included in the bibliographies of qualified studies.

The supplemental information files contain the following:

WORKBOOK S1_STUDIES USED

Lists features of, and sources for, the studies used to construct classification profiles and to estimate the time required to perform the average task with no predefined minimum duration. With 2 exceptions (see below, DATA S1 and DATA S2), all the studies have been published or are readily available on the internet. None of the data can be used to identify individuals.

DATA S1_REPORT OF THE HEALTH FACILITY SURVEY IN BOTSWANA, 2007-08 and DATA S2_REPORT OF THE HEALTH FACILITY SURVEY IN TANZANIA, 2003

PDF files of Health Facility Survey reports that were found on the internet but have since been taken down.

WORKBOOK S2_CLASSIFICATION PROFILES: INFANTS; WORKBOOK S3_CLASSIFICATION PROFILES: CHILDREN IN UPPER MIDDLE-INCOME COUNTRIES; WORKBOOK S4_CLASSIFICATION PROFILES: CHILDREN IN LOWER MIDDLE-INCOME COUNTRIES (I); WORKBOOK S5_CLASSIFICATION PROFILES: CHILDREN IN LOWER MIDDLE-INCOME COUNTRIES (II), and WORKBOOK S6_CLASSIFICATION PROFILES: CHILDREN IN LOW INCOME COUNTRIES                                                                                                          

Columns A-F of the worksheets in WORKBOOKS S2-S6 contain classification profiles for the populations used in this project.

Beginning with column S and continuing to the right, the worksheets in WORKBOOKS S2-S6 apply the rules governing the performance of clinical tasks from the relevant version of the algorithm to each combination of classifications resulting from the classification profile in columns A-F. The cumulative total of combinations for each classification profile is calculated in columns G-I. The probability of each combination appears in column Q.

Using the geometric mean from WORKBOOK S8 (see below), Column R of the worksheets in WORKBOOKS S2-S6 shows the mean expected duration of consultations for patients with each combination of classifications.

The composition of population-specific mean consultations by task category is shown columns O and P of the worksheets in WORKBOOKS S2-S6.

The distribution of consultations by expected duration for each classification profile is shown in columns J-N of the worksheets in WORKBOOKS S2-S6.

WORKBOOK S7_IMCI CLINICAL TASKS

Lists the clinical tasks provided by relevant IMCI algorithms for the care of infants and children. Consists of 6 worksheets covering mandatory tasks, conditional assessments, and treatment and counseling tasks for infants and children.

WORKBOOK S8_MINUTES PER TASK WITH NO MINIMUM DURATION

Provides estimate of the mean time required to perform a task with no minimum duration for each of 7 populations for which the required data are available, corrected, where necessary, for the effect of an observer on the rate and pace of task performance. Also provides a geometric mean for all 7 populations.

TEXT S1_NOTES OF THE CONSTRUCTION OF CLASSIFICATION PROFILES

WORD document describing the steps involved in constructing each profile, problems encountered, and how they were solved.

TEXT S2_NOTES ON THE IDENTIFICATION OF IMCI CLINICAL TASKS

WORD document describing the standards used in identifying clinical tasks in IMCI algorithms.

TEXT S3_NOTES ON THE ESTIMATION OF MINUTES PER TASK WITH NO MINIMUM DURATION

WORD document describing the steps involved in estimating the mean time required to perform a task with no predefined minimum duration, problems encountered, and how they were solved.