Sidu Biai

Biai 2011 Biai, Sidu. 2011. Evaluation of Interventions to Reduce Hospital and Community Mortality. Bandim Health Project, Statens Serum Institut. Faculty of Health Sciences, University of Copenhagen.
The aim of this thesis is to evaluate the impact of community, primary and tertiary health care interventions on child mortality, and to determine risk factors for hospitalisation.

The mortality rates both in the community and in the hospital are very high. Neonatal mortality is 55 per 1000 live births, infant mortality 138 per 1000 live birth, and under-five mortality is 223 per 1000 live births (127). The overall mortality rate in the paediatric ward of the national hospital Simao Mendes was 12% in 1991-96 (146). 35% of all deaths among hospitalised children were caused by malaria, 15% by both acute respiratory infection (ARI) and diarrhoea diseases. Most developing countries are implementing the WHO immunisation programme. Although vaccines reach most children, many modifications of the recommended schedule are observed in practice. We investigated the association between vaccination status and hospitalisation in Guinea-Bissau.

Community and Primary health care intervention: deployment of rectal artesunate
We investigated the impact on overall mortality of a community-based administration of rectal artesunate (rArs) suppositories to non per os (NPO) children with suspected malaria.

Study area and population
Children between 3 months and 5 years of age, deployment of rArs by three types of community dispensers, health agents, traditional healers and maternal coordinators in rural villages in the two regions Biombo and Oio in Guinea-Bissau.

Since 1989-1990, BHP has followed 56 villages in Oio and Biombo as part of the rural HDSS. BHP covered 16 of the villages where rArs was dispensed, and 40 villages where rArs was not deployed and these were used as controls.

Risk factors for hospitalisation
We investigated risk factors for hospitalisation and impact of vaccination status and sequence of vaccination on risk of hospitalisation for children.

Study area and population
From May 2003 to May 2004, we registered all children less than 5 years of age consulting at the Out-Patient Department (OPD) of the paediatric department. Information on age, sex, diagnosis, hospitalisation, vaccines received and socio-economic facts was collected.

In-hospital intervention: standardised protocol and financial incentive
We investigated the impact of a standardised protocol and a financial incentive on case-fatality for children hospitalised with severe malaria.

At the paediatric department at the national hospital, Simao Mendes, children aged 3 months to 5 years admitted with severe malaria were randomised to “normal” procedures in one ward or a ward where a standardised protocol had to be followed, and the personnel was financially motivated to follow the protocol.

Community and Primary health care intervention: deployment of rArs
From May 2005 to end July 2006, 4,458 children aged 3 months and less than 5 years of age lived in the BHP HDSS villages in Biombo and Oio. Of these 1,327 (29.8%) lived in villages where rArs was deployed, with 749 (56%), 301 (23%) and 277 (21%) children in the three dispenser-arms: health agents, traditional healers and maternal coordinators, respectively.

Villages where rArs had been dispensed had an overall mortality rate ratio (MRR) of 1.27 (95%CI 0.68-2.38) compared with the control villages, and controlled for age and village clustering. Following the intervention the MRR was 0.64 (0.27-1.52), 1.60 (0.67-3.79) and 2.52 (1.48-4.32) for the three dispenser-arms, respectively (p=0.01). In the year prior to the intervention, there had been no significant difference in mortality (p=0.28) though mortality in villages later assigned to the health agent arm had a slightly increased relative mortality (annex 1, table 2).

Risk factors for hospitalisation
From May 2003 to May 2004, 18,609 consultations of children less than 5 years of age were registered at the OPD at the national hospital Simão Mendes in Bissau. Included in the analyses were 11,949 (64%) with information on vaccination, and of these 2219 (19%) were hospitalised (annexe 2, Figure 3).

In the interval 9 to 17 months of age, having MV only as the latest vaccine was significantly associated both with lower risk of hospitalisation compared with children who were delayed and had only received DTP (HRR=0.71 (0.60-0.85)), and a lower risk compared with having received DTP after MV (HRR=0.63 (0.45-0.88) or MV and DTP simultaneously (HRR=0.65 (0.50-0.85)).

The female-male HRR was significantly different for children who had MV (HRR=0.85 (0.72-1.00)) or DTP (HRR=1.09 (0.97-1.23)) as most recent vaccination (p=0.02).

In-hospital intervention: standardised protocol and financial incentive
During the study period, we randomised 951 children between 3 months and 5 years of age: 460 (48%) to the intervention ward and 491 (52%) to the control ward.
In the 12 months before the trial, overall in-hospital mortality had been 13% (413/3076). During the intervention in-hospital mortality was 5% for the intervention group and 10% in the control group (RR=0.48 (0.29-0.79)). For children with positive malaria slides maybe even stronger (RR=0.36 (0.16-0.80)). The cumulative mortality 4 weeks after discharge was also lower in the intervention group (RR=0.61 (0.40 to 0.95)).

These results suggest that convenient, accessible anti-malarial drug deployment using health agents can reduce overall child mortality in the community. However, they do not demonstrate a beneficial effect in villages using traditional healers and maternal coordinators. We do not know the reasons for the negative impact on children’s survival.

Following the recommended vaccination schedule for BCG and MV is associated with a reduced risk of hospitalisation but this is not the case for DTP and booster DTP. Receiving DTP simultaneously with MV or after MV is associated with increased risk of hospitalisation.  Vaccines have sex-differential effects on the risk of hospitalisation.

Supervising healthcare workers to adhere to a standardised treatment protocol was associated with greatly reduced in-hospital mortality. Financial incentives may be important for the dedication and compliance of staff members.