Impact of age and renal function on usefulness of NT-proBNP to diagnose heart failure
.

AIMS: NT-proBNP is a useful biomarker for heart failure (HF) diagnosis. We aimed to determine NT-proBNP's ability to diagnose HF by age and renal function. MATERIALS AND METHODS: We analyzed 3,699 consecutive and unique adults admitted for dyspnea at the Emergency Unit of St. Joseph St. Luc Hospital, Lyon, France, from December 1, 2012 to June 30, 2016, who had concomitant measurement of NT-proBNP and serum creatinine. We excluded patients with acute coronary syndrome and dialysis patients. Receiving operating characteristic (ROC) analysis assessed ability and cut-off points of NT-proBNP to diagnose HF. RESULTS: Mean age was 79.1 ± 13.0 years. Mean estimated glomerular filtration rate (eGFR, CKD EPI formula) was 64 ± 26 mL/min/1.73m2. The ROC area under the curve (AUC) was 0.813 on average, optimal NT-proBNP cut-off point was 1,896 ng/L. AUC decreased (0.882, 0.813, 0.767) by age class (18 - 69, 70 - 84, 85+ years, respectively), and optimal cut-off points increased (1,041, 1,902, 2,321 ng/L). AUC decreased (0.881, 0.830, 0.783, 0.781, 0.705) by eGFR class (≥ 90, 60 - 89, 45 - 59, 30 - 44, < 30 mL/min/1.73m2), and cut-off points increased (757, 1,362, 2,283, 4,108, 7,288 ng/L). The lowest value of cut-off points associated with highest sensitivity and specificity was detected in young patients with eGFR ≥ 90 (597 ng/L) while the worst value was found in age 85+ patients with eGFR < 30 (7,288 ng/L). AUC decreased below 0.8 in age 70+ patients with eGFR < 45 mL/min/1.73m2;. CONCLUSION: The ability of NT-proBNP to diagnose HF decreased strongly with age and renal function. NT-proBNP's usefulness in diagnosing HF in age 70+ patients with eGFR < 45 mL/min/1.73m2 remains uncertain.

Establishment of indicators to improve clinical analysis prescribing patterns.

An appropriate medical analysis prescribing pattern is part of the medical biologists' work as it enhances patient care and reduces costs. In this study, we use four indicators to aim to evaluate the relevance of clinical analysis prescription. We confronted clinical data and medical analysis prescribed in June 2013 in the emergency department (ED) and found that prescriptions were justified in 73% of TnT prescriptions but only in less than 50% of NTproBNP (27%), APTT (37%), PR (33%) or INR (23%) prescriptions. We noted that staff training, an improved communication between biologists and clinical physicians, and better computing devices, have led to better prescribing patterns. From 2013 to 2015, inappropriate associations of PR and APTT have significantly declined in the intensive care unit. At the same period, amounts of medical analysis as well as department spendings decreased in the ED. The use of indicators is essential to evaluate and monitor the relevance of medical analysis patterns. In this work, we propose to combine a global indicator (cost/day of hospitalization or medical analysis amount/month) with a regular follow up on inadequate prescribed analysis associations. These indicators will need to be adjusted to each clinical department.