The impact of long term institutional collaboration in surgical training on trauma care in Malawi

Sven Young, Leonard Banza, Nyengo Mkandawire

Attempts to address the huge, and unmet, need for surgical services in Africa by training surgical specialists in well established training programmes in high-income countries have resulted in brain drain, as most trainees do not return home on completion of training for various reasons. Local postgraduate training is key to retaining specialists in their home countries. International institutional collaborations have enabled Kamuzu Central Hospital (KCH) in Lilongwe, Malawi, to start training their own surgical specialists from 2009. The direct impact of this has been an increase in Malawian staff from none at all to 12 medical doctors in 2014 in addition to increased foreign faculty. We have also seen improved quality of care as illustrated by a clear reduction in the amputation rate after trauma at KCH, from nearly every fourth orthopaedic operation being an amputation in 2008 to only 4 % in 2014. Over the years the training program at KCH has, with the help from its international partners, merged with the College of Medicine in Blantyre, Malawi, into a national training programme for surgery. Our experiences from this on-going international institutional collaboration to increase the capacity for training surgeons in Malawi show that long-term institutional collaboration in the training of surgeons in low-income countries can be done as a sustainable and up-scalable model with great potential to reduce mortality and prevent disability in young people. Despite the obvious and necessary focus on the rural poor in low-income countries, stakeholders must start to see the value of strengthening teaching hospitals to sustainably meet the growing burden of trauma and surgical disease. Annual operating data from Kamuzu Central Hospital's Main Operating Theatre log book for the years 2008-2014 was collected. Observed annual numbers were presented as graphs for easy visualization. Linear regression curve estimations were calculated and plotted as trend lines on the graphs.