Is IMCI an effective mechanism for delivery of child survival interventions in a high HIV prevalence setting? : a study to determine the effectiveness of the Intergrated Management of Childhood Illness (IMCI) strategy in management of sick children in routine practise in primary health care clinics in South Africa
Introduction: Integrated management of childhood illness (IMCI) is a child survival strategy that has been adopted in South Africa (SA) as the standard of care for managing sick children in the primary health care setting. IMCI includes guidelines for management of paediatric HIV. This study aimed to investigate effectiveness of IMCI as a vehicle to deliver essential child survival interventions, particularly HIV interventions, in routine practise in a high HIV prevalence setting, and to investigate barriers and enabling factors for IMCI implementation. Methods: The study was conducted in Limpopo and KwaZulu-Natal provinces, SA. In the qualitative component, focus group discussions were conducted with IMCI trained health workers and carers of children under 5 years, to explore experiences of IMCI implementation, particularly the HIV component, from the perspective of both target groups. A comparative survey was then conducted. Randomly selected IMCI trained nurses were observed for up to 20 consultations with sick children presenting consecutively to the facility, and their findings compared to those of an IMCI expert who subsequently assessed the child. Observed children were tested for HIV. Results: IMCI trained nurses found IMCI training informative and empowering, and there was agreement among nurses that their skills in managing sick children improved after training. Barriers to IMCI implementation included increased time required for IMCI consultations and lack of support from colleagues. IMCI trained nurses expressed reluctance to implement the HIV component of IMCI, believing it to be unnecessary, unacceptable to mothers and that they lacked the skills to implement HIV care. In total, 77 IMCI trained nurses were observed for a total of 1357 consultations between May 2006 and January 2007; nurses were observed for a mean of 17.7 consultations. Components of the IMCI assessment were frequently omitted; 14/77(18%) nurses asked about all main symptoms in every child. IMCI classifications were often incorrect; 52/112 (46.4%) children with a general danger sign were correctly classified. The HIV component was poorly implemented, 342/1357 (25.2%) children were correctly classified for HIV, although the HIV algorithm performed well when implemented by IMCI experts. Conclusion: IMCI implementation is fragmented and incomplete. Interventions are urgently needed to achieve and maintain high quality health worker performance in implementing IMCI.