Improving Judicious Use of Heparin-Induced Thrombocytopenia Testing Through Electronic Health Record-Based Intervention.

BACKGROUND: Heparin-induced thrombocytopenia (HIT) is an immune-mediated drug reaction that can cause thromboembolism in the setting of thrombocytopenia. An enzyme-linked immunosorbent assay (ELISA)-based assay to screen for HIT antibodies (HAb) is available but has relatively low specificity and a correspondingly high false positive rate. The 4Ts score has been validated to determine the pretest probability of HIT. The authors hypothesized that an electronic health record (EHR)-based clinical decision support (CDS) tool incorporating the 4Ts score would reduce the volume of HAb orders. METHODS: After implementing a CDS tool into the EHR, the researchers retrospectively evaluated the impact from November 2019 to October 2021, compared to a preintervention period (January to October 2019). The primary outcome was average tests per month. Secondary outcomes included rates of tests ordered per total inpatient encounters and proportion of HAb sent despite low 4Ts score in the postintervention study period. RESULTS: Of 1,833 HAb sent during the study period, 1,217 occurred in the postintervention period. In the postintervention period compared with the preintervention period, the average orders per month was 50.5 (standard deviation [SD] 9.7) vs. 61.6 (SD 7.2) (p = 0.003), and the order incidence rate was 8.0 per 1,000 patient encounters postintervention vs. 9.2 per 1,000 patient encounters preintervention (rate ratio [RR] 0.87, 95% confidence interval [CI] 0.79-0.96, p = 0.002). Postintervention, 252 (20.7%) had a 4Ts score calculated as low probability, 759 (62.4%) as intermediate probability, and 131 (10.8%) as high probability, and 75 had no associated 4Ts score. CONCLUSION: Implementation of a simple CDS tool reduced the rate of HAb orders, reducing unnecessary HAb testing.

Recurrent connections facilitate symmetry perception in deep networks.

Symmetry is omnipresent in nature and perceived by the visual system of many species, as it facilitates detecting ecologically important classes of objects in our environment. Yet, the neural underpinnings of symmetry perception remain elusive, as they require abstraction of long-range spatial dependencies between image regions and are acquired with limited experience. In this paper, we evaluate Deep Neural Network (DNN) architectures on the task of learning symmetry perception from examples. We demonstrate that feed-forward DNNs that excel at modelling human performance on object recognition tasks, are unable to acquire a general notion of symmetry. This is the case even when the feed-forward DNNs are architected to capture long-range spatial dependencies, such as through 'dilated' convolutions and the 'transformers' design. By contrast, we find that recurrent architectures are capable of learning a general notion of symmetry by breaking down the symmetry's long-range spatial dependencies into a progression of local-range operations. These results suggest that recurrent connections likely play an important role in symmetry perception in artificial systems, and possibly, biological ones too.

Immediate and long-term effects of a team-based quality improvement training programme.

BACKGROUND: Although many studies of quality improvement (QI) education programmes report improvement in learners' knowledge and confidence, the impact on learners' future engagement in QI activities is largely unknown and few studies report project measures beyond completion of the programme. METHOD: We developed the Academy for Quality and Safety Improvement (AQSI) to prepare individuals, across multiple departments and professions, to lead QI. The 7-month programme consisted of class work and team-based project work. We assessed participants' knowledge using a multiple choice test and an adapted Quality Improvement Knowledge Assessment Test (QIKAT) before and after the programme. We evaluated participants' postprogramme QI activity and project status using surveys at 6 and 18 months. RESULTS: Over 5 years, 172 individuals and 32 teams participated. Participants had higher multiple choice test (71.9±12.7 vs 79.4±13.2; p<0.001) and adapted QIKAT scores (55.7±16.3 vs 61.8±14.7; p<0.001) after the programme. The majority of participants at 6 months indicated that they had applied knowledge and skills learnt to improve quality in their clinical area (129/148; 87.2%) and to implement QI interventions (92/148; 62.2%). At 18 months, nearly half (48/101; 47.5%) had led other QI projects and many (41/101; 40.6%) had provided QI mentorship to others. Overall, 14 (43.8%) teams had positive postintervention results at AQSI completion and 20 (62.5%) had positive results at some point (ie, completion, 6 months or 18 months after AQSI). CONCLUSIONS: A team-based QI training programme resulted in a high degree of participants' involvement in QI activities beyond completion of the programme. A majority of team projects showed improvement in project measures, often occurring after completion of the programme.

Triage documentation-based decision support to improve infectious disease risk screening and mitigate exposure.

Multidisciplinary focus group review of current triage practice identified gaps in identification of potentially infectious diseases. Modifications were made to triage and nursing assessment forms that were easy to maneuver, rapidly modifiable, and provided documentation-based decision support to expedite infection prevention measures. Development of a decision support infectious disease risk screening tool enhances outbreak preparedness, occupational safety, and response.