Quality Improvement Interventions Targeting Immunization Data from Urban Lower-level Private-for-profit Health Service Providers in Kampala Capital City: Processes and Implementation Experiences

"The participatory approach resulted in co-creating interventions targeting modification of health workers' and health managers' behaviors towards practices that promoted accurate and timely immunization data at the private health facility level."

In Uganda, delivery of immunisation services within urban contexts is predominantly by private-for-profit service providers and follows urban market demand and supply forces. Lower-level private providers tend to establish in densely populated suburbs located in informal settlements. These urban immunisation service providers are faced with challenges including high staff turnover that impedes skills retention and a lack of a direct incentive for data quality improvement. This paper documents the implementation processes and experiences of piloting a co-created intervention whose aim was to contribute to improvements in the accuracy and timeliness of immunisation service data among lower-level private-for-profit immunisation service providers in Kampala Capital City Authority (KCCA), Uganda.

A quasi-experimental design was adopted, with a participatory baseline data collection process leading to the identification of two poor-performing city divisions (Kawempe and Makindye) where the intervention was implemented. The researchers adopted a human-centred design to develop the intervention in a participatory manner together with urban immunisation stakeholders. Participatory methodologies (e.g., co-creation workshops and collective identification of Behavioural Change Techniques (BCT) likely to alter the behaviours of health workers delivering immunisation services) were adopted to agree on an appropriate intervention that would respond to identified underlying bottlenecks.

The COM-B model was used to link intervention functions in response to particular bottlenecks. It represents Capability (C), Opportunity (O), Motivation (M) and Behaviour (B). In order to effect positive changes in the desired behaviours and in this study - that is, the accurate recording and timely submission of immunisation data from service providers - the researchers determined that both Capabilities and Opportunities had to be targeted. For example, the Capability aspect of the framework targeted behavioural change, focusing on health workers' skills and knowledge to improve immunisation data quality, while the Opportunities aimed at targeting interventions that improved the work environment affecting data quality. In addition, Motivation aimed at changing health worker practices that impeded data quality and promoting reflexive processes for them to develop a positive attitude towards quality data generation.

The proposed interventions were aligned to the middle layer of the Behavioural Change Wheel (BCW), as reflected in the Figure above. They were translated into the Theoretical Framework for Acceptability (TFA) in response to challenges associated with the local translation of theoretically derived interventions that would not have been compatible with the local context. The goal was to gain consensus on the practices that service providers would adopt into their work routine to lead to improvements in data quality and timely submission.

Forty private health facilities participated in the implementation research, with 20 assigned to the intervention and 20 assigned to the control. Performance measurements were assessed at baseline and endline to compare two key outcomes between the intervention and control groups: (i) accurate submission of immunisation data from the private (assessed by computing monthly monitoring data using five proxy indicators at the health facility level, such as oral polio vaccine (OPV) dose 1 and "fully immunised"); and (ii) timely submission of data (assessed based on the date the data was submitted into the District Health Information Software 2 (DHIS2) system).

Through the above-described theory-driven design with the COM-B as the guiding model, the behavioural change intervention functions targeted to cause the desired change leading to improvements in data quality among private providers were: (i) targeted training, (ii) modelling through continuous quality improvement (CQI), (iii) persuasion through onsite mentorships and enhanced monitoring and support supervision, (iv) coercion through onsite spot checks, enforced supply of vaccines upon data accountability, and enforced timely submission of data, and (v) environmental restructuring through support that ensured availability of data tools and vaccines.

For example, the first intervention function that was targeted according to the BCW was education in the form of training. Specifically, capacity building through focus group training is a participatory research approach that is common in sub-Saharan Africa settings, where it has been applied in fostering research capacities. The training was augmented with site visits and follow-up communication exchanges between health workers and supervisors. Adopting the training approach was appropriate, as it responded to the need among private providers where high rates of staff turnover left the facilities with diminished skills related to data quality and timely submission.

The researchers conclude that identifying barriers and collaboratively working with private-for-profit immunisation service providers to address problems and boost motivation to improve immunisation data quality "is paramount. Participatory approaches were found to be effective in addressing this gap. Strengthening the health worker capacities, service delivery improvements and playing an active oversight function to the private providers were found to be a step in the right direction towards responding to the complex challenges impeding quality immunization data among lower-level private-for-profit urban immunization service providers in Kampala Capital City of Uganda."