Determinants of scaled implementation for electronic laboratory information systems in 27 clinical laboratories in Côte d'Ivoire: A mixed-methods study applying the Consolidated Framework for Implementation Research (CFIR)

Background Electronic laboratory information systems (LIS) are critical for promoting data quality and improving laboratory performance, but sustained implementation following system deployment in resource-limited settings can be challenging. The OpenELIS project in Côte d’Ivoire seeks to promote routine use of the OpenELIS software at 108 clinical laboratories. This study assessed the sustainment of OpenELIS routine use at laboratories, identified implementation facilitators and barriers, and explained how high-sustainment laboratories differed from low-sustainment ones. Methods We conducted a convergent parallel mixed-methods study at 27 clinical laboratories across 15 health regions in Côte d’Ivoire in 2021. We collected and analyzed de-identified routine testing data from OpenELIS servers to describe the sustained use of OpenELIS at each laboratory by calculating the percentage of calendar days during the implementation period when test results were recorded in OpenELIS. We conducted interviews informed by the Consolidated Framework for Implementation Research with laboratory staff to identify implementation determinants and to summarize differences between laboratories of high-, medium-, and low-sustainment. We used deductive and inductive approaches to code the interview data and conduct thematic analysis. Results OpenELIS was used during 9.0-99.3% of calendar days, and high-sustainment sites performed more routine tests. Common facilitators of sustainment included receiving positive comments about OpenELIS from external high-level leaders and partners; appreciating advantages of OpenELIS; and perceiving high accessibility of knowledge and information. Common barriers included low adaptability, power outages, server failure, space limitations, and printing costs. Participants of low-sustainment laboratories reported lower motivation and capability in using OpenELIS, perceived more complexity and lower software design quality, expressed more concerns about staffing challenges for data entry, received fewer positive comments about OpenELIS from internal leads, and had less reflection and feedback about implementation. Conclusions Ensuring infrastructure stability may benefit all laboratories. Refresher training, increased frequency of support visits, and staffing plans may help low-sustainment laboratories to increase their routine LIS use. Further research that tests implementation strategies may be beneficial for addressing the barriers and extend the facilitators to other settings.