Study 13: Site choice

The study was conducted in three countries and four sites:

  • Bangladesh (Chittagong)
  • Ghana (Navrongo)
  • Tanzania (Kilosa, Handeni)


In Bangladesh the reported clinical cases in 1996 were 104,500, about 50% being P falciparum, and about 80% of cases from Chittagong division in the south-East of Bangladesh with a population of about 30 million. In Chittagong, the study was implemented in Cox’s Bazar, Ramu, Naikyongchari and Ukhiya Upazillas (sub districts). These are very rural areas with seasonal variations in perennial transmission, predominantly falciparum malaria, a regular influx of migrants and a high degree of morbidity and mortality from malaria.

There is a district hospital at Cox’s bazaar and in addition two non-government hospitals. There were about 150 Primary Health Care Workers living in the villages to cover this population. Transport facilities from patient’s home to near by health facilities is usually difficult particularly in rainy reason and at night. There was no formal patient transport (e.g. ambulance service) available. Studies indicated long hours-required before access to appropriate care, up to 18 hours at times. The number of severe malaria cases from Chittagong and Chittagong Hill Districts was very high in 1997 with more than 50% hospital admissions for malaria. Hospital malaria mortality in the district was 234 in the year 1996 (3) (4).


The trial was carried out in the Kassena-Nankana District adjacent to the border with Burkina Faso in the northern part of Ghana. There is a district hospital with satellite health centres. The District hospital in Navrongo is a 140-bed facility with 32 pediatric beds/cots and serves the entire district for secondary care. The four health centres are run by medical assistants who provide static clinical services and limited outreach services, referring complicated cases to the district hospital. Trained traditional birth attendants and village health workers and community health officers support the health centres. There is a large informal sector of health providers - the traditional healers and soothsayers who tend to be the first point of call for cases of febrile convulsions and several other conditions (altered consciousness) considered to be caused by evil spirits.

Parasite rates were highest in the oldest children (5-7 years), but parasite densities, low haemoglobin levels and rates of febrile illness were highest in 6-11 month olds (1). Malaria was estimated to account for 23% of child deaths in the Kassena-Nankana district in northern Ghana(2). An analysis of routine health facility records from the three health centres for 1997 was carried out specifically to produce some baseline data. The analysis was restricted to out-patient attendants among the under-five year olds at the health centres and the paediatric ward admissions at the district hospital. The analyses showed that health facilities were under-utilized (perhaps related to distances patients had to travel to the facilities). Outpatient attendance by month in the health centres showed 3,399 out-patient visits, malaria was diagnosed all year round and accounted for 48% of patients. The paediatric ward of the district hospital recorded 1,842 admissions in 1997. An analysis of the patient records of the patients diagnosed as malaria showed the average distance travelled by these patients was 11.5 kilometres (range 1-48 kilometres), 68% of these malaria admissions were under 2 years of age, admitted to the hospital, 1-3 days after the onset of the illness. The main reasons for admission were persistent vomiting (38%) and convulsions (30%). Blood films were taken in only 14% of children diagnosed as malaria on admission and 92% of these slides were reported as positive. Most of the children recovered from their illness and were discharged home but 5% are reported to have died on admission. There was no record in the case histories about referrals either from the health centres or the informal sector, represented by the traditional healers.


After careful review of alternatives, Morogoro region was chosen; it was a new area without ongoing research on malaria, most of the villages did not have village health posts and therefore there were likely to be delays in reaching injectable treatment. A study that had followed up all deaths within three divisions of Morogoro during an observational period of 3.5 years indicated that there were 1795/6294 deaths in children below the age of 5 years, of which 84% occurred at home, half dying without having had any contact with medical care(5).

Kilosa district, Morogoro Region is situated 300 km west of Dar es Salaam with 135 registered villages. The total population in 2000 was estimated at 448 830. The district had two hospitals, one government (Kilosa Hospitals with 150 beds) and mission hospital (Berega Hospital with 100 beds). Admissions to Kilosa hospital (January 1998 to June 1999) showed that 75.7% (2816/3718) paediatric admissions were malaria and anaemia related, 93% in children below 5 and 90% of malaria deaths were in this age group.


One year after the clinical trial had begun in Kilosa, the trial area was expanded to Handeni District, Tanga Region, which is situated 274 kms from Dar es Salaam with 112 registered villages and a total population of 248,633 according to 2002 census. The district has one government hospital (Handeni District Hospital) with 110 beds and a mission hospital (Kwamkono) with 70 beds. The main sources of anti-malarial medications are shops. These mostly treat mild malaria.


  • Binka et al (1994) Trans. of the Royal Society of Tropical Medicine and Hygiene 88 381-385.
  • Binka et al. (1995) International Journal of Epidemiology 24:127-135.
  • Bin-Yunus E, Faiz MA, Rahman MR et al. Study to document pre-admission risk factors for development of severe malaria and the spectrum of it and the outcome in different categories of hospitals in malaria endemic zone of Bangladesh. J Bang Coll Phys Surg 2004; 22: 83-88.
  • Montanari RM, Bangali AM, Talukder KR et al. Three case definitions of malaria and their effect on diagnosis, treatment and surveillance in Cox's Bazaar district, Bangladesh. Bull WHO 2001; 79: 648-56.
  • 5.Ministry of Health, Adult Morbidity and Mortality Project, 1997. The policy implications of adult morbidity and mortality: end of phase 1 report. Dar es Salaam, Tanzania: United Republic of Tanzania, 1997. 197-207.