Our study showed that health workers across different cadres and with different levels of pre-service training were able to learn how to effectively perform the ICT RDT, from a one day training session. Nursing assistants experienced some difficulties in pricking the finger with a lancet and filling a pipette with blood. Similar difficulties have been reported in Tanzania [18
], South Africa and Zambia [19
] among low level cadres although these are generally overcome with practice. Apart from the practical skill of performing the RDT, it was realised that the informally trained health workers (nursing assistants) require additional training to improve their knowledge on aspects of prevention and specific treatment of malaria and non-malaria fever cases. These results imply a potentially important role of RDTs in the implementation of parasite-based malaria diagnosis by health workers at LLHFs in Uganda where trained laboratory technicians or the infrastructure to support routine microscopy are not readily available.
The one-day training given at a central location seems sufficient to transfer the skills necessary to roll-out the use of RDTs for malaria diagnosis at peripheral health facilities in Uganda. Immediate utilisation of RDTs (within 6 weeks of training) appears to be recommended as evidenced by the finding that 75% of febrile patients who presented to the health facilities thereafter were offered a diagnostic test. It is important to design studies to evaluate the long-term retention of knowledge and skills attained as well as to assess the need and timing of re-fresher training to maintain the high utilization and quality in performance of RDTs for parasite-based malaria diagnosis at lower level health care facilities.
It was interesting to demonstrate that health workers who received the one-day training in RDT use were able to multiply this effect by adequately training their counterparts at their respective health facilities (cascade model training). This finding is consistent with reports from the malaria case-management team at the Infectious Diseases Institute in Uganda, which showed that cascade training by trainees after a trainer of trainers (TOT) course was comparable to the initial first-hand training [21
]. However, this was an incidental finding in our study and there is not enough data to compare it with the one-day first-hand training. There is a need to further explore this strategy given the fact that over half of the training costs were spent on participants’ transport and per-diem. Peer training may be a potential mechanism to maintain use of RDTs in the context of high health worker turn over in addition to the fact that periodic training may be limited after country-wide implementation of RDTs. Future training programmes should systematically evaluate the utility and cost-effectiveness of the cascade training model in the implementation of parasite-based malaria diagnosis.
This is one of the first studies to document the cost of the training process on the use of RDTs in a resource-limited setting. The one-day training cost up to US $101 to train one health worker to use RDTs. The documented costs included the cost of facilitating the core trainers from the centre (Ministry of Health and Malaria Consortium) to the districts, transport and operational costs. This study did not document capital costs including preparation of training materials since this is an intervention that should be integrated into the current health care system to improve the efficiency of service delivery. We recommend cost-effectiveness studies to understand the utility and long-term impact of one-day on the scale up of parasite-based malaria case management in resource-limited settings.
Lessons learnt: Given the limited numbers of health workers, there is need for strategic phasing of the training sessions to avoid interruption of services at the health facilities. It is for this reason that we performed two one-day trainings to have half the trainees attend independently while the other half offered patient care and vise-versa. Within the context of the health system where the majority of health care providers are not formally trained, the training content should be tailored to suit the lower academic levels and should include more on background knowledge in addition to the practical skills.
This study did not cost the preparation of the training materials since these materials are readily available from the NMCP. Generic training materials have also been made available through collaborations with international agencies like WHO [22
]. It is likely that the observed performance could have been biased since the health workers were aware that they were being observed and scored. This could amounts to a Hawthorne effect, but it does not affect the outcomes because it was the same bias introduced during the two assessments; during the training and six weeks post-training. In addition, the study design did not include health facilities to receiving RDTs without training as a comparative arm to facilities with training + RDTs. Such a design would not be feasible for an operational study. Therefore it was not possible to attribute the antimalarial saved to the training alone, or even extrapolate the impact in this setting. Routine health facility data was used to document patients tested without being able to quality control its acquisition. However the main outcome variables of training were collected by our study terms.