Background A substantial proportion of microbiological screening in diagnostic laboratories is due to suspected urinary tract infections (UTIs), yet approximately two thirds of urine samples typically yield negative culture results. By reducing the number of query samples to be cultured and enabling diagnostic services to concentrate on those in which there are true microbial infections, a significant improvement in efficiency of the service is possible. Methodology Screening process for urine samples prior to culture was modelled in a single clinical microbiology laboratory covering three hospitals and community services across Bristol and Bath, UK. Retrospective analysis of all urine microscopy, culture, and sensitivity reports over one year was used to compare two methods of classification: a heuristic model using a combination of white blood cell count and bacterial count, and a machine learning approach testing three algorithms (Random Forest, Neural Network, Extreme Gradient Boosting) whilst factoring in independent variables including demographics, historical urine culture results, and clinical details provided with the specimen. Results A total of 212,554 urine reports were analysed. Initial findings demonstrated the potential for using machine learning algorithms, which outperformed the heuristic model in terms of relative workload reduction achieved at a classification sensitivity > 95%. Upon further analysis of classification sensitivity of subpopulations, we concluded that samples from pregnant patients and children (age 11 or younger) require independent evaluation. First the removal of pregnant patients and children from the classification process was investigated but this diminished the workload reduction achieved. The optimal solution was found to be three Extreme Gradient Boosting algorithms, trained independently for the classification of pregnant patients, children, and then all other patients. When combined, this system granted a relative workload reduction of 41% and a sensitivity of 95% for each of the stratified patient groups. Conclusion Based on the considerable time and cost savings achieved, without compromising the diagnostic performance, the heuristic model was successfully implemented in routine clinical practice in the diagnostic laboratory at Severn Pathology, Bristol. Our work shows the potential application of supervised machine learning models in improving service efficiency at a time when demand often surpasses resources of public healthcare providers. Electronic supplementary material The online version of this article (10.1186/s12911-019-0878-9) contains supplementary material, which is available to authorized users.

Early Warning Scores are not sufficiently accurate to rule in or rule out mortality in patients with sepsis, based on the evidence available, which is generally poor quality.

No abstract

The interaction between colistin and tigecycline against eight well-characterized NDM-1-producing Enterobacteriaceae strains was studied. Time-kill methodology was employed using a 4-by-4 exposure matrix with pharmacokinetically achievable free drug peak, trough, and average 24-h serum concentrations. Colistin sulfate and methanesulfonate alone showed good early bactericidal activity, often with subsequent regrowth. Tigecycline alone had poor activity. Addition of tigecycline to colistin does not produce increased bacterial killing; instead, it may cause antagonism at lower concentrations. New Delhi metallo-␤-lactamase 1 (NDM-1)-producing Enterobacteriaceae strains are being reported all over the world (10,11,13). The chemotherapeutic options for treating NDM-1-producing Enterobacteriaceae infection are limited (5). Colistin and tigecycline (TGC) are both agents for which many strains still have MICs below the clinical breakpoint. Tigecycline and colistin act on bacterial cells by different mechanisms: tigecycline by inhibition of protein synthesis and colistin on the outer cell membrane. Also, tigecycline is bacteriostatic by nature, whereas colistin is bactericidal. Therefore, there is a potential scope for both antagonism and synergy. Also, there are compelling reasons why clinicians may choose to use these drugs in combination, especially given the recent controversies regarding the efficacy and safety of tigecycline monotherapy (1,14).We studied the bactericidal activities of tigecycline (TGC), colistin sulfate (CS), and colistin methanesulfonate (CMS) alone as well as in various combinations against NDM-1-producing Enterobacteriaceae.The bactericidal activities of TGC, CS, and CMS were assessed using time-kill methodology (12). For each antimicrobial agent, we used pharmacokinetically achievable free drug serum concentrations. We used the following concentrations (mg/liter) reflecting peak (C max ), 24-h average (C ss ), and trough (C min ) concentrations: 0.17, 0.04, and 0.025, respectively, for TGC (4); 0.29, 0.16, and 0.1, respectively, for CS (3, 9); and 8.5, 2.7, and 2.1, respectively, for CMS (2, 6, 7). A 4-by-4 drug exposure matrix of TGC with CS and CMS was used along with an antibiotic-free growth control (GC).Eight well-characterized strains of NDM-1-producing Enterobacteriaceae (2 of Escherichia coli, 2 of Klebsiella oxytoca, and 4 of Klebsiella pneumoniae) stored at Ϫ70°C were selected for the study. The MICs for all the isolates were measured by the gradient strip method using the Etest (bioMérieux). The tigecycline MICs of the two E. coli strains were 0.25 and 0.38 mg/liter, and both K. oxytoca strains had MICs of 0.25 mg/liter. Of the four K. pneumoniae strains, two had a tigecycline MIC of 0.38 mg/liter and the remaining two strains had MICs of 1.5 and 3.0 mg/liter. The colistin MIC of both E. coli strains was 0.38 mg/liter, and the MICs of the two K. oxytoca strains were 0.094 and 0.125 mg/liter. Two out of the four K. pneumoniae strains had a colistin MIC of 0.38 mg/ liter, and the remain...