A Trial of Improved Financial Access to Healthcare in Ghana

Objectives

General:

To assess by means of a randomized trial, the impact of improving financial access to primary health care on morbidity due to severe malaria and utilization of health services among children six months to five years of age.

Specifically:

1. To compare rates of severe anaemia in children under five years of age from households randomized to those with improved financial access and those without for the first year.
2. To compare the number of admissions to hospital with malaria among children under five years of age from households with improved access and households without.
3. To compare the health service utilization rate among children under five years of age from households with improved access and households without.
4. To compare the reported median period between the onset of febrile (presumed malaria) illness and reporting at health facilities among children under five years of age from households with improved access and households without.
5. To document community knowledge, attitude and practice with regards to severe malaria user fees, exemptions pre-payment schemes and indicators of poverty.

Study design and methods

Study Design:

The proposed study is a two-arm randomized controlled trial to assess the impact of improved financial access to primary health care on morbidity due to severe malaria and utilization of services among children 6 months to 5 years. It will have a quantitative element based on pre-determined outcomes as well as a qualitative element.

Intervention The intervention group will be enrolled into a pre-payment scheme operating in the area which will allow them free access to primary care including diagnosis and drugs whenever they are ill with no limit, and limited access to secondary health care.

The control group will have to pay user fees for their health care as pertains normally.

Recruitment and Enrolment Procedures:

Study participants will be selected using the process outlined in the trial profile in the annexe. The actual study participants will be children between 6 months and five years of age in the households selected.

Inclusion Criteria:

• Child in household aged between 6 months and 5 years
• Informed consent from parents
• Household to which child belongs resident in the Dodowa sub-district and intending to remain resident until the end of the next two years

Exclusion Criteria

• Parents refuse to participate in the study
• Household due to emigrate from study area within two years
• Household currently enrolled

Randomization Procedure A stratified randomization procedure will be used. Households will be divided into three strata based on residence being ≤5km, 5-10km, and >10km from the health centre respectively since distance from health facility is known to be a major determinant to its use and may not be adequately taken care of with simple random allocation. Allocation of the intervention will be carried out separately within each stratum.

Household heads within each stratum will be invited to meetings at which random numbers will be allocated to each household. The numbers on raffle tickets will then be dropped into a rotating barrel and mixed up thoroughly in the view of all after which the barrel will be rotated. A child will be invited to pick the papers which will be allocated to intervention or control as they are drawn, alternately. Those in the intervention group will have pre-payment made for year one; the others will have it in year 2. After each draw the barrel will be rotated once more. This process will be used to make the trial more acceptable to community members by showing them the lack of favouritism and randomness of the allocation (Adapted from [43]. Since registration would have closed for the year, households will not be able to change their group until the study ends.

Outcome Measures

Primary:

• Prevalence of severe anaemia among children under five years of age in both groups after one peak malaria transmission season.

Secondary

• Mean haemoglobin count among the two groups at study end
• Number of admissions to hospital with malaria among children under five years of age in both groups
• Average number of visits to per child under five years to primary care facility among both groups
• Anthropometric measurements of children in the two groups
• All cause mortality in both groups

All households in the Dodowa sub-district with at least one child 6 months to 5 years who do not register in the pre-payment scheme for the year will be eligible to participate in the study. The information will be obtained from the data base of households at the DHIMT offices. This sub-district, which is mainly forest with an estimated 2003 midyear population of 30,154, has been selected because earlier studies have shown that it has the highest incidence of malaria in the district. Household heads or their representatives will be invited to either of three locality meetings based on distance from the health facility, where the study will be explained to them and consent sought for their participation in the study. Those agreeing to participate will have their particulars recorded and enrolled into the study. They will then be randomized into two groups using the procedure already described. Baseline data collection including haematological, anthropometric assessments and interviews will then be carried out on all participants and slides for parasites in a random subset. Background variables about the household including known risk factors for malaria such as type of housing, use of insecticide treated materials, use of other preventive measures and some indicators of household wealth will be documented. Children found to have haemoglobin count less than 6g/dl, fever or a history of fever and patent parasitaemia will be treated according to local guidelines; this will not be a reason for withdrawal from the study.

Differences between those who voluntarily registered in the pre-payment scheme and those who did not, in terms of health service utilization, health status, household wealth and other variables will be documented as a subsidiary part of the study.

Follow-Up:

Morbidity monitoring will be carried out through a hospital/clinic based passive case detection system throughout the malaria transmission season from May to October. This will be complemented by one cross-sectional household survey at the end of the malaria transmission season in November during which interviews, haematological, parasitological and anthropometric assessments will be repeated. Children found to have severe anemia during the second survey will have a haematological work-up comprising Hb electrophoresis, full blood count, and stool examination. All study participants who visit the clinic will be treated by the health workers as usual in accordance with the current standard treatment guidelines in Ghana. In the case of mortality, at a culturally appropriate time, a trained field worker will visit the home and a standard questionnaire will be administered to the child's mother or caretaker. The results will be reviewed by three independent physicians to arrive at a possible diagnosis. If two out of three agree on a diagnosis of malaria, this diagnosis will be accepted. For children who visited hospital or died in the hospital, hospital case notes will be reviewed and where necessary, the attending physician will be interviewed for more complete information on the child's illness.

Plan for Data Collection

Interviews:

Standard clinic record forms will be used to collect data from the primary care clinic or hospital through a combination of direct interviews and abstraction of information from their case notes by trained research health staff. Any unclear information in the case notes will be verified with the clinician concerned immediately. Field data will also be collected by means of standard pre-coded questionnaires administered to household heads. Individual in-depth interviews will also be administered to respondents using an interview guide. Focus groups will be organized by contacting the participants and arranging for a suitable meeting date, time and venue, the previous week. Focus groups will consist of 8-12 discussants with a moderator and an observer/notetaker. Discussions will be recorded by means of a tape recorder with permission from discussants.

Laboratory Measurements:

A finger-prick sample of blood will be obtained using a haemocue® microcuvette for haemoglobin count estimation. Thick and thin blood films will be obtained from the same finger prick sample for parasitological measurements for both surveys.

Anthropometric Measurements:

All children will be weighed using infant weighing scales or hanging scales that are usually used for weighing children during child welfare clinics for children unable to stand alone and a mechanical stand-on scale for older children who can stand alone. Supine length of children will be recorded using a measuring board whilst a height measure will be used to take the standing height of all children 2 years and above. The mid upper arm circumference (MUAC) of each child will be measured with a MUAC tape (TALC Institute of Child Health).

Data Management and Analysis:

Quantitative data will be recorded on pre-coded questionnaires or clinical data record forms whilst qualitative data will be recorded both on tape recorder and small notebooks. Data will be checked daily for accuracy and consistency, correctly labelled, and stored for data entry.

For quantitative data, data entry clerks will carry out double entry of the raw data from questionnaires and clinical record forms in an electronic form into Microsoft Access. After data entry, data cleaning will be carried out to identify errors and inconsistencies for correction. Analysis will be carried out using Stata version 8. Summary statistics, odds ratios, confidence limits and p-values will all be calculated to compare outcomes between the two groups for the primary end point and secondary endpoints. Stratified analysis will be carried out based on age, sex, distance from health facility and household wealth. Means of continuous variables such as rise in haemoglobin count will be compared using the Student t-test. Geometric means will be used to compare parasite densities. Qualitative data will be analysed using Textbase Alpha after converting the transcribed text file into an ASCII file.