The use of computerised clinical decision support system in maternal and neonatal care in rural Africa : enthusiasm and concerns

Sammanfattning: Background: Maternal and neonatal mortality rates remain high in sub-Saharan Africa. The shortage of skilled Healthcare Workers (HCWs) and poor access to guidelines are plausible reasons for the reported low quality in maternal and neonatal services in rural Africa. One strategy to support HCWs at point of care is to provide easy access to guidelines using computerised Clinical Decision Support Systems (CDSS). Still, data are lacking on how such systems are optimally designed and implemented to be useful for HCWs. Aims: The aims were to develop a CDSS to be used in rural Africa (I), to understand perceived needs and attitudes among HCWs to the use of a CDSS based on the WHO guidelines in maternal and neonatal care in Burkina Faso (II), and to explore the reasons why the HCWs failed to use the CDSS as expected (III). Methods: The CDSS was programmed in java software language to be able to run on any hardware (I). The CDSS was part of an intervention to improve quality of care in six rural healthcare facilities in Burkina Faso for 24 months. A total of 45 HCWs were interviewed to capture perceived needs and attitudes to the CDSS. Data were analysed with content analysis (II). To understand any reduced use of the CDSS a workshop was organised (group discussions and a plenary session) with 13 HCWs with data analysed thematically (III). Socio-demographic data were analysed descriptively (II-III). Results: The CDSS was designed with a user interface and an XML database for storing patient data (signs and symptoms) as well as an algorithm to provide advice on recommended care and actions based on WHO guideline information. The CDSS was developed with limited input from HCWs (I). The HCWs expressed willingness to use new technologies such as a CDSS and computers but reported a fear of extra workload as well as a fear that the CDSS should be complicated to use (II). After 12 months, the decreased use of the CDSS was partly explained by unreliable power supply and poor fit between the CDSS and the daily workflow. Still, the HCWs were enthusiastic to learn more by using the CDSS (III). Conclusions: The CDSS was successfully developed and tested in rural Burkina Faso. Despite this, its use was unexpectedly low. It was found that: 1. the design and the implementation of a CDSS have to be contextualised; 2. the usage of the CDSS and the software performance need to be continuously monitored and 3. the HCWs need to be actively consulted during all phases of design and testing of a CDSS to enhance its use.