RESUMO
BACKGROUND: Blood culture collection practice in low-resource settings where routine blood culture collection is available has not been previously described. METHODOLOGY: We conducted a secondary descriptive analysis of children aged 2-23 months enrolled in the Malawi Childhood Acute Illness and Nutrition (CHAIN) study, stratified by whether an admission blood culture had been undertaken and by nutritional status. Chi-square test was used to compare the differences between groups. RESULTS: A total of 347 children were included, of whom 161 (46%) had a blood culture collected. Children who had a blood culture collected, compared to those who did not, were more likely to present with sepsis (43% vs. 20%, p < 0.001), gastroenteritis (43% vs. 26%, p < 0.001), fever (86% vs. 73%, p = 0.004), and with poor feeding/weight loss (30% vs. 18%, p = 0.008). In addition, hospital stay in those who had a blood culture was, on average, 2 days longer (p = 0.019). No difference in mortality was observed between those who did and did not have a blood culture obtained. CONCLUSION: Blood culture collection was more frequent in children with sepsis and gastroenteritis, but was not associated with mortality. In low-resource settings, developing criteria for blood culture based on risk factors rather than clinician judgement may better utilize the existing resources.
Blood culture is key to investigating bloodstream infections, but in-hospital decisions to perform blood culture in a low-resource setting have not been previously described. We linked blood culture data to the Childhood Acute Illness and Nutrition (CHAIN) cohort at a Malawi tertiary hospital and compared clinical characteristics and outcomes of children between those who did and did not have a blood culture done on admission. Of those hospitalized, 46% of the children had a blood culture collected at admission. Only 3% of blood cultures had significant growth of pathogenic bacteria. There were significant differences in nutritional status, presenting symptoms, clinical diagnoses and hospital length of stay between those who received blood culture collection on admission and those who did not, but there was no difference in mortality. Clinical judgement used to determine blood culture collection may not best identify children most at risk.