Acute critical illness is often preceded by deterioration of routinely measured clinical parameters, e.g., blood pressure and heart rate. Early clinical prediction is typically based on manually calculated screening metrics that simply weigh these parameters, such as early warning scores (EWS). The predictive performance of EWSs yields a tradeoff between sensitivity and specificity that can lead to negative outcomes for the patient. Previous work on electronic health records (EHR) trained artificial intelligence (AI) systems offers promising results with high levels of predictive performance in relation to the early, real-time prediction of acute critical illness. However, without insight into the complex decisions by such system, clinical translation is hindered. Here, we present an explainable AI early warning score (xAI-EWS) system for early detection of acute critical illness. xAI-EWS potentiates clinical translation by accompanying a prediction with information on the EHR data explaining it.
Improving diagnosis of prosthetic joint infections (PJIs) has become an increasing challenge due to a steadily rising number of patients with prosthetic implants. Based on a systematic literature search we have ascertained the evidence base for improvement of culture diagnosis. We searched PubMed/MEDLINE using the medical subject heading (MeSH) ‘prosthesis-related infections’ 1995 through 2010 without further restrictions. An analogous search was conducted for ISI Web of Knowledge. A total of 1409 reports were screened for original results, obtained by methods described in sufficient detail to make replication possible. We gave priority to methods for sample preparation, culture media, culture methods and incubation time. Clinical sensitivity and specificity were calculated where possible. We found evidence to support superiority of cultures obtained from the diluent after sonication of prosthetic implants in comparison with culturing tissue biopsies. Sonication parameters and accessory steps have been studied extensively, and thresholds for significant growth have been defined. Conversely, methods for processing of soft tissue biopsies have been studied to a limited extent. Culture of synovial fluid in blood culture vials has been shown to be more sensitive (90–92 %) than intraoperative swab cultures (68–76 %) and tissue cultures (77–82 %). Formal evaluation of agar media for culturing PJI specimens seemed to be lacking. The polymicrobial nature of PJIs supports the routine use of an assortment of media suitable for recovery of fastidious, slow-growing, anaerobic and sublethally damaged bacteria. A number of studies supported an incubation period for up to 14 days. Although we identified evidence-based improvements of culture methods, there is a need for more studies especially with regard to tissue biopsies. Culturing remains an important means to identify and characterize pathogenic micro-organisms and supplements the increasing number of culture-independent assays.