Care Seeking During Fatal Childhood Illnesses: Siaya District, Kenya, 1998

In developing countries, the majority of the approximately 12 million fatal illnesses that occur each year among children younger than 5 years can be prevented or treated effectively by means of simple interventions.1 In Nyanza Province, Kenya, 1 of every 5 children dies before reaching 5 years of age.2 CARE Kenya's Community Initiatives for Child Survival in Siaya project trained community health workers in approximately 200 villages and established community pharmacies in an effort to reduce childhood mortality in Siaya District, one of the least developed areas of Nyanza Province.

The community health workers were trained to treat children with fever, cough or difficult breathing, and diarrhea; refer seriously ill children; and promote illness prevention behaviors. A year after the health workers were trained, we investigated care-seeking itineraries during terminal illnesses to identify opportunities to further reduce childhood mortality.

We identified deaths that occurred among children younger than 5 years during the period May 1 through August 30, 1998, by reviewing community health workers' monthly household registers and asking the health workers, at the end of each month, whether any children who had died in their village were not recorded in the registers. We interviewed caregivers about the deceased children's symptoms and duration of terminal illness, types of health providers consulted, and number, chronology, and timeliness of visits.

Overall, 99 deaths among children younger than 5 years were identified in the project area; we interviewed caregivers of 97 children. The median interval between date of death and interview was 29 days (range: 7–152 days). Seventy percent of the children were infants, and 44% were female. During their terminal illness, most of the children received care outside the home, yet 90% died at home, and only 6% received inpatient care at any time during the illness (Table 1 ▶).

TABLE 1—

Sources of Health Care During Terminal Illnesses of 97 Deceased Children Younger Than 5 Years: Siaya District, Western Kenya, 1998

Source of Care	No. (%)
Treatment at home
Home care (massage, fluids)	58 (60)
Tried medicines available at home	39 (40)
Consultation outside the home
Drug vendor	42 (43)
Traditional healer	49 (51)
Untrained practitioner of Western medicine	20 (21)
Community health worker	25 (26)
Health worker at a health facility	46 (47)
Inpatient care	6 (6)
Trained provider of Western medical care (community health worker or health worker at a health facility)	56 (58)
Referred to higher level care	10 (10)
Returned for follow-up care
Returned to a traditional healer	37 (76)
Returned to a trained provider of Western medical care	18 (32)
Returned to a drug vendor	6 (14)
Returned to an untrained practitioner of Western medicine	11 (55)

The median duration of terminal illnesses was 7 days (range: 4 hours to 92 days). Ninety percent of the children had fever, cough or difficult breathing, or diarrhea. Caregivers usually consulted multiple health providers (mean: 3.8; range: 1–15). Caregivers consulted traditional healers most frequently (51%); 76% of the caregivers who consulted traditional healers returned to them a second time. In all, 46% of the caregivers consulted a health worker at a health facility, and 26% consulted a community health worker.

In all, 58% of the 97 children were seen by a trained provider of Western medical care (either a community health worker or a health worker at a health facility). Only 32% (18/56) of the children seen by a trained provider of Western medical care returned to such a provider for follow-up care.

No predominant reason was given for not consulting a community health worker or a health worker in a health facility. The most frequently mentioned reason for not consulting a community health worker was not knowing about such individuals (26%), whereas the reason most frequently given for not consulting a worker in a health facility was preference for traditional healers (16%).

Only 10% of the 97 children were referred for higher level care. The median delay in consulting a health provider after onset of symptoms was 2 days; promptness of care seeking did not differ by type of provider consulted.

This investigation of fatal childhood illnesses involved important limitations, including possible underreporting of childhood deaths and potential recall bias in information obtained from caregivers. Because we included in the study only children who had died, we were not able to identify risk factors for death or measure the effectiveness or coverage of project interventions. However, the findings point to opportunities for further reducing childhood deaths.

First, follow-up care and referral were infrequent. Only 32% of the children seen by a trained Western medical provider returned to such a provider for follow-up care, and only 10% were referred for higher level care. Only 6% of the children were hospitalized. Better use of referral facilities might be achieved by offering community health workers and health facility staff additional in-service training in recognizing severe illnesses and in counseling caregivers to return for follow-up care.

Second, traditional healers are a potential resource for improving child survival.3,4 Traditional healers saw half of the children who died, and 76% of the children they saw returned to a traditional healer. Dialogue with traditional healers might lead to an understanding of how they can help ensure that the severely ill children they see are also seen at a health facility prepared to offer them appropriate treatment.

Finally, community health workers, although easily accessible, are underused by caregivers of terminally ill children; only 26% of the children in our study were seen by a community health worker. The main reason given by caregivers for not using community health workers was not knowing they existed. Traditional gatherings and village meetings can enhance dissemination of information in rural Kenyan communities to increase awareness of community health workers.