Blood culture collection technique and pneumococcal surveillance in Malawi during the four year period 2003–2006: an observational study

A prospective observational study.Following the introduction of a specialised blood culture team in 2005, the proportion of contaminated cultures decreased (19.6% in 2003 to 5.0% in 2006), blood volume cultured increased and pneumococcal recovery increased significantly from 2.8% of all blood cultures to 6.1%. With each extra 1 ml of blood cultured the odds of recovering a pneumococcus increased by 18%.Standardisation and assessment of blood culture performance (blood volume and contamination rate) should be incorporated into pneumococcal disease surveillance activities where routine blood culture practice is constrained by limited resources.Blood cultures are an essential component of good clinical care in the diagnosis and management of blood stream infections (BSI) which are frequent in hospitalised patients in Malawi and the rest of Africa [1-6]. Information from blood culture surveillance is also an important tool for establishing public health priorities, assessing the impact of interventions – particularly vaccines – and for providing information on antimicrobial resistance patterns to help formulate prescribing guidelines for empirical therapy. Blood culture surveillance has been the key tool used in the USA for recognising the enormous potential of childhood pneumococcal conjugate vaccination for decreasing disease in adults[7].In Malawi as in much of the rest of sub-Saharan Africa, blood culture facilities are available in only a limited number of centres. As a consequence single reports are often extrapolated as representative of disproportionately large regions and time periods, and inaccuracies in these reports are more likely to be perpetuated as a result of the lack of alternative data. Blood cultures need to be specific and sensitive and therefore as representative as possible of the true BSI disease burden and particularly so where BSI surveillance is restricted to very few sites.Isolation of bacteria from blood is usually taken as definitive proof