Real-world evaluation of AI driven COVID-19 triage for emergency admissions: External validation & operational assessment of lab-free and high-throughput screening solutions

Background Uncertainty in patients' COVID-19 status contributes to treatment delays, nosocomial transmission, and operational pressures in hospitals. However, typical turnaround times for batch-processed laboratory PCR tests remain 12-24h. Although rapid antigen lateral flow testing (LFD) has been widely adopted in UK emergency care settings, sensitivity is limited. We recently demonstrated that AI-driven triage (CURIAL-1.0) allows high-throughput COVID-19 screening using clinical data routinely available within 1h of arrival to hospital. Here we aimed to determine operational and safety improvements over standard-care, performing external/prospective evaluation across four NHS trusts with updated algorithms optimised for generalisability and speed, and deploying a novel lab-free screening pathway in a UK emergency department. Methods We rationalised predictors in CURIAL-1.0 to optimise separately for generalisability and speed, developing CURIAL-Lab with vital signs and routine laboratory blood predictors (FBC, U&E, LFT, CRP) and CURIAL-Rapide with vital signs and FBC alone. Models were calibrated during training to 90% sensitivity and validated externally for unscheduled admissions to Portsmouth University Hospitals, University Hospitals Birmingham and Bedfordshire Hospitals NHS trusts, and prospectively during the second-wave of the UK COVID-19 epidemic at Oxford University Hospitals (OUH). Predictions were generated using first-performed blood tests and vital signs and compared against confirmatory viral nucleic acid testing. Next, we retrospectively evaluated a novel clinical pathway triaging patients to COVID-19-suspected clinical areas where either model prediction or LFD results were positive, comparing sensitivity and NPV with LFD results alone. Lastly, we deployed CURIAL-Rapide alongside an approved point-of-care FBC analyser (OLO; SightDiagnostics, Israel) to provide lab-free COVID-19 screening in the John Radcliffe Hospital's Emergency Department (Oxford, UK), as trust-approved service improvement. Our primary improvement outcome was time-to-result availability; secondary outcomes were sensitivity, specificity, PPV, and NPV assessed against a PCR reference standard. We compared CURIAL-Rapide's performance with clinician triage and LFD results within standard-care. Results 72,223 patients met eligibility criteria across external and prospective validation sites. Model performance was consistent across trusts (CURIAL-Lab: AUROCs range 0.858-0.881; CURIAL-Rapide 0.836-0.854), with highest sensitivity achieved at Portsmouth University Hospitals (CURIAL-Lab:84.1% [95% Wilson's score CIs 82.5-85.7]; CURIAL-Rapide:83.5% [81.8 - 85.1]) at specificities of 71.3% (95% Wilson's score CIs: 70.9 - 71.8) and 63.6% (63.1 - 64.1). For 3,207 patients receiving LFD-triage within routine care for OUH admissions between December 23, 2021 and March 6, 2021, a combined clinical pathway increased sensitivity from 56.9% for LFDs alone (95% CI 51.7-62.0) to 88.2% with CURIAL-Rapide (84.4-91.1; AUROC 0.919) and 85.6% with CURIAL-Lab (81.6-88.9; AUROC 0.925). 520 patients were prospectively enrolled for point-of-care FBC analysis between February 18, 2021 and May 10, 2021, of whom 436 received confirmatory PCR testing within routine care and 10 (2.3%) tested positive. Median time from patient arrival to availability of CURIAL-Rapide result was 45:00 min (32-64), 16 minutes (26.3%) sooner than LFD results (61:00 min, 37-99; log-rank p<0.0001), and 6:52 h (90.2%) sooner than PCR results (7:37 h, 6:05-15:39; p<0.0001). Sensitivity and specificity of CURIAL-Rapide were 87.5% (52.9-97.8) and 85.4% (81.3-88.7), therefore achieving high NPV (99.7%, 98.2-99.9). CURIAL-Rapide correctly excluded COVID-19 for 58.5% of negative patients who were triaged by a clinician to COVID-19-suspected (amber) areas. Impact CURIAL-Lab & CURIAL-Rapide are generalisable, high-throughput screening tests for COVID-19, rapidly excluding the illness with higher NPV than LFDs. CURIAL-Rapide can be used in combination with near-patient FBC analysis for rapid, lab-free screening, and may reduce the number of COVID-19-negative patients triaged to enhanced precautions (amber) clinical areas.

Download Full-text

Paper-Based SERS Platform for One-Step Screening of Tetracycline in Milk

Scientific Reports ◽

10.1038/s41598-019-54380-y ◽

2019 ◽

Vol 9 (1) ◽

Cited By ~ 4

Author(s):

Ana Marques ◽

Bruno Veigas ◽

Andreia Araújo ◽

Beatriz Pagará ◽

Pedro Viana Baptista ◽

...

Keyword(s):

High Throughput ◽

High Throughput Screening ◽

Point Of Care ◽

Dairy Industry ◽

Principal Component ◽

Surface Enhanced Raman Spectroscopy ◽

Economic Losses ◽

Screening Methods ◽

Sers Substrate ◽

Molecular Signatures

AbstractThroughout the last decade, the expansion of food testing has been gradually moving towards ordinary high throughput screening methods performed on-site. The demand for point-of-care testing, able to distinguish molecular signatures with high accuracy, sensitivity and specificity has been significantly increasing. This new requirement relies on the on-site detection and monitorization of molecular signatures suitable for the surveillance of food production and processing. The widespread use of antibiotics has contributed to disease control of livestock but has also created problems for the dairy industry and consumers. Its therapeutic and subtherapeutic use has increased the risk of contamination in milk in enough concentrations to cause economic losses to the dairy industry and have a health impact in highly sensitive individuals. This study focuses on the development of a simple Surface-Enhanced Raman Spectroscopy (SERS) method for fast high throughput screening of tetracycline (TET) in milk. For this, we integrate a paper-based low-cost, fully recyclable and highly stable SERS platform, with a minimal sample preparation protocol. A two-microliter sample of milk solutions spiked with TET (from 0.01 to 1000 ppm) is dried on a silver nanoparticle coated cardboard substrate and measured via a Raman spectrophotometer. The SERS substrate showed to be extremely stable with a shelf life of several months. A global spectrum principal component analysis approach was used to test all the detected vibrational modes and their correlation with TET concentration. Peak intensity ratios (455 cm−1/1280 cm−1 and 874 cm−1/1397 cm−1) were found to be correlated with TET concentrations in milk, achieving a sensitivity as low as 0.1 ppm. Results indicate that this SERS method combined with portable Raman spectrometer is a potential tool that can be used on-site for the monitoring of TET residues and other antibiotics.

Download Full-text

High burden of cryptococcal antigenemia and meningitis among patients presenting at an emergency department in Maputo, Mozambique

PLoS ONE ◽

10.1371/journal.pone.0250195 ◽

2021 ◽

Vol 16 (4) ◽

pp. e0250195

Author(s):

Robert Deiss ◽

Carolina V. Loreti ◽

Ana G. Gutierrez ◽

Eudoxia Filipe ◽

Milton Tatia ◽

...

Keyword(s):

Emergency Department ◽

Cryptococcal Meningitis ◽

Screening Program ◽

Point Of Care ◽

Rapid Assessment ◽

Vital Signs ◽

Cd4 Count ◽

Sub Saharan Africa ◽

Induction Treatment ◽

Cryptococcal Antigen

Background Cryptococcal meningitis is a leading cause of HIV-related mortality in sub-Saharan Africa, however, screening for cryptococcal antigenemia has not been universally implemented. As a result, data concerning cryptococcal meningitis and antigenemia are sparse, and in Mozambique, the prevalence of both are unknown. Methods We performed a retrospective analysis of routinely collected data from a point-of-care cryptococcal antigen screening program at a public hospital in Maputo, Mozambique. HIV-positive patients admitted to the emergency department underwent CD4 count testing; those with pre-defined abnormal vital signs or CD4 count ≤ 200 cells/μL received cryptococcal antigen testing and lumbar punctures if indicated. Patients with CM were admitted to the hospital and treated with liposomal amphotericin B and flucytosine; their 12-week outcomes were ascertained through review of medical records or telephone contact by program staff made in the routine course of service delivery. Results Among 1,795 patients screened for cryptococcal antigenemia between March 2018—March 2019, 134 (7.5%) were positive. Of patients with cryptococcal antigenemia, 96 (71.6%) were diagnosed with CM, representing 5.4% of all screened patients. Treatment outcomes were available for 87 CM patients: 24 patients (27.6%) died during induction treatment and 63 (72.4%) survived until discharge; of these, 38 (60.3%) remained in care, 9 (14.3%) died, and 16 (25.3%) were lost-to follow-up at 12 weeks. Conclusions We found a high prevalence of cryptococcal antigenemia and meningitis among patients screened at an emergency department in Maputo, Mozambique. High mortality during and after induction therapy demonstrate missed opportunities for earlier detection of cryptococcal antigenemia, even as point-of-care screening and rapid assessment in an emergency room offer potential to improve outcomes.

Download Full-text

Abstract WP409: Impact of Nursing Expertise and Staffing Ratio on Compliance with Post-tPA/Endovascular Patient Monitoring

Stroke ◽

10.1161/str.48.suppl_1.wp409 ◽

2017 ◽

Vol 48 (suppl_1) ◽

Author(s):

Kelly Montgomery ◽

Danielle Sindelar ◽

Julie Fussner ◽

Erin Supan ◽

Cathy Sila

Keyword(s):

Emergency Department ◽

Medical Center ◽

Diagnostic Testing ◽

Vital Signs ◽

Quality Improvement Initiative ◽

Flow Sheet ◽

University Hospitals ◽

Patient Acuity ◽

Staffing Ratio ◽

Data Points

Introduction: In 2014, the Post-tPA/Endovascular nursing monitoring flow sheet was revised to harmonize with the NIHSS exam. The challenges to performance and documentation compliance posed by the intensity and frequency of assessments have been the focus of an ongoing quality improvement initiative. Methods: All Post-tPA/Endovascular monitoring flow sheets from June 2014- June 2016 at University Hospitals Case Medical Center were reviewed for presence of neurologic assessments, vital signs, and management of hypertension per protocol. Stroke staff conducted in-services on the enhanced assessments and modified NIHSS training and house staff mentored bedside RNs in performing the NIHSS. A tip sheet was developed for staff on the modified NIHSS and real time feedback was given on all outliers. Results: Of 459 patients, compliance with all of the 684 monitoring data points ranged from 67-100% in the Neuroscience ICU (Patient: RN ratio 2:1), 75-100% in the Neuro-Intermediate Unit (Patient: RN ratio 3:1) and 40-100% in the Emergency Department (Patient: RN ratio 4:1). Overall compliance to > 95% of data points was seen in all but 5 patients with missing flow sheets. Symptomatic hemorrhagic complications after IVtPA decreased from 6.5% to 2.7%. Root-cause analysis of missing data points revealed seven areas of opportunity: Interference by diagnostic testing (29%), during patient transportation (22%), and following endovascular treatment (15%) or due to travelling RNs (8%). Missing documentation was most frequent during the q15 minute phase due to the intensity of monitoring (11%)- with the Emergency Department the most vulnerable location- and less during the q30 minute (4%) or q1 hour (3%) assessments. Units with dedicated neuroscience nursing adjusted more rapidly to the revision compared to units that do not routinely perform such assessments. Conclusions: Optimum compliance with nursing assessments and monitoring occur when there is no interference with diagnostic testing or procedures, the patient needs were a high priority and the patient acuity was well matched to the RN staffing ratio. This data supports a care model where a neuroscience trained RN nurse transitions with the patient during the first 24 hours after tPA/Endovascular therapy.

Download Full-text

LO45: Does the use of point of care ultrasonography improve survival in emergency department patients with undifferentiated hypotension? The first Sonography in Hypotension and Cardiac Arrest in the Emergency Department (SHOC-ED1) Study; an international randomized controlled trial

CJEM ◽

10.1017/cem.2017.107 ◽

2017 ◽

Vol 19 (S1) ◽

pp. S43

Author(s):

P.R. Atkinson ◽

J. Milne ◽

L. Diegelmann ◽

H. Lamprecht ◽

M. Stander ◽

...

Keyword(s):

Emergency Department ◽

Randomized Controlled Trial ◽

Length Of Stay ◽

Diagnostic Accuracy ◽

Standard Care ◽

Primary Outcome ◽

Point Of Care ◽

Controlled Trial ◽

Randomized Controlled ◽

Secondary Outcomes

Introduction: Point of care ultrasound (PoCUS) is an established tool in the initial management of patients with undifferentiated hypotension in the emergency department (ED). While PoCUS protocols have been shown to improve early diagnostic accuracy, there is little published evidence for any mortality benefit. We report the findings from our international multicenter randomized controlled trial, assessing the impact of a PoCUS protocol on survival and key clinical outcomes. Methods: Recruitment occurred at 7 centres in North America (4) and South Africa (3). Scans were performed by PoCUS-trained physicians. Screening at triage identified patients (SBP<100 or shock index>1), randomized to PoCUS or control (standard care and no PoCUS) groups. Demographics, clinical details and study findings were collected prospectively. Initial and secondary diagnoses were recorded at 0 and 60 minutes, with ultrasound performed in the PoCUS group prior to secondary assessment. The primary outcome measure was 30-day/discharge mortality. Secondary outcome measures included diagnostic accuracy, changes in vital signs, acid-base status, and length of stay. Categorical data was analyzed using Fishers test, and continuous data by Student T test and multi-level log-regression testing. (GraphPad/SPSS) Final chart review was blinded to initial impressions and PoCUS findings. Results: 258 patients were enrolled with follow-up fully completed. Baseline comparisons confirmed effective randomization. There was no difference between groups for the primary outcome of mortality; PoCUS 32/129 (24.8%; 95% CI 14.3-35.3%) vs. Control 32/129 (24.8%; 95% CI 14.3-35.3%); RR 1.00 (95% CI 0.869 to 1.15; p=1.00). There were no differences in the secondary outcomes; ICU and total length of stay. Our sample size has a power of 0.80 (α:0.05) for a moderate effect size. Other secondary outcomes are reported separately. Conclusion: This is the first RCT to compare PoCUS to standard care for undifferentiated hypotensive ED patients. We did not find any mortality or length of stay benefits with the use of a PoCUS protocol, though a larger study is required to confirm these findings. While PoCUS may have diagnostic benefits, these may not translate into a survival benefit effect.

Download Full-text

Point of care lung ultrasound is useful when screening for CoVid-19 in Emergency Department patients.

10.1101/2020.06.09.20123836 ◽

2020 ◽

Cited By ~ 1

Author(s):

Andrea Hankins ◽

Heejung Bang ◽

Paul Walsh

Keyword(s):

Emergency Department ◽

Null Hypothesis ◽

Lung Ultrasound ◽

Point Of Care ◽

Vital Signs ◽

P Value ◽

Inclusion Criteria ◽

Upper Respiratory Infection ◽

Life Threatening ◽

Emergency Department Patients

Background CoVid-19 can be a life-threatening lung disease or a trivial upper respiratory infection depending on whether the alveoli are involved. Emergency department (ED) screening in symptomatic patients with normal vital signs is frequently limited to oro-nasopharyngeal swabs. We tested the null hypothesis that patients being screened for CoVid-19 in the ED with normal vital signs and without hypoxia would have a point-of-care lung ultrasound (LUS) consistent with CoVid-19 less than 2% of the time. Methods Subjects Subjects were identified from ED ultrasound logs. Inclusion criteria Age 14 years or older with symptoms prompting ED screening for CoVid-19. Exclusion criteria Known congestive heart failure or other chronic lung condition likely to cause excessive B lines on LUS. Intervention Structured blinded ultrasound review and chart review Analysis We used an exact hypothesis tests for binomial random variables. We also measured LUS diagnostic performance using computed tomography as the gold standard. Results We reviewed 77 charts; 62 met inclusion criteria. Vital signs were normal in 31 patients; 10 (32%) of these patients had LUS consistent with CoVid-19. We rejected the null hypothesis (p-value for bitest <0.001). The treating physicians' interpretation of their own point of care lung ultrasounds had a sensitivity of 100% (95% CI 75%, 100%) and specificity of 80% (95% CI 68%, 89%). Conclusion LUS has a meaningful detection rate for CoVid-19 in symptomatic emergency department patients with normal vital signs. We recommend at least LUS be used in addition to PCR testing when screening symptomatic ED patients for CoVid-19.

Download Full-text

A man with intolerant vomiting

Emergency Medicine Journal ◽

10.1136/emermed-2019-208863 ◽

2019 ◽

Vol 37 (1) ◽

pp. 41-49

Author(s):

Ya-Chin Liang ◽

Tou-Yuan Tsai

Keyword(s):

Emergency Department ◽

Point Of Care ◽

Vital Signs ◽

Point Of Care Ultrasound ◽

List Type ◽

Case Presentation ◽

Supplementary Video ◽

Healthy State ◽

Supplementary Material ◽

Large Meal

Case presentationA 57-year-old man was in a healthy state until 6 hours before presenting to the emergency department, when he suffered from more than 10 episodes of vomiting after a large meal at lunchtime. On physical examination, vital signs were unremarkable. Tenderness was noted over the epigastrium on palpation. Point-of-care ultrasound (POCUS) of the abdomen is shown in figure 1 and online supplementary video 1.Supplementary videoFigure 1Transverse (A) and longitudinal (B) ultrasonography of the epigastrium.QuestionWhat is the most likely diagnosis?Hypertrophic pylorus stenosisAortic dissectionSuperior mesentery artery (SMA) syndromeVolvulusFor answer see page 2

Download Full-text

Recent Advances in Microfluidic Paper-Based Analytical Devices toward High-Throughput Screening

Molecules ◽

10.3390/molecules25132970 ◽

2020 ◽

Vol 25 (13) ◽

pp. 2970

Author(s):

Siraprapa Boobphahom ◽

Mai Nguyet Ly ◽

Veasna Soum ◽

Nayoon Pyun ◽

Oh-Sun Kwon ◽

...

Keyword(s):

High Throughput ◽

High Throughput Screening ◽

Point Of Care ◽

Cost Effective ◽

Drug Analysis ◽

Detection Sensitivity ◽

Detection Methods ◽

Food And Beverage ◽

Rapid Result ◽

Recent Developments

Microfluidic paper-based analytical devices (µPADs) have become promising tools offering various analytical applications for chemical and biological assays at the point-of-care (POC). Compared to traditional microfluidic devices, µPADs offer notable advantages; they are cost-effective, easily fabricated, disposable, and portable. Because of our better understanding and advanced engineering of µPADs, multistep assays, high detection sensitivity, and rapid result readout have become possible, and recently developed µPADs have gained extensive interest in parallel analyses to detect biomarkers of interest. In this review, we focus on recent developments in order to achieve µPADs with high-throughput capability. We discuss existing fabrication techniques and designs, and we introduce and discuss current detection methods and their applications to multiplexed detection assays in relation to clinical diagnosis, drug analysis and screening, environmental monitoring, and food and beverage quality control. A summary with future perspectives for µPADs is also presented.

Download Full-text