Comparison of clinical characteristics and outcomes of patients with sepsis identified by the Sepsis-3 criteria by blood and urine culture results: A multicentre retrospective cohort study.

Background and Aims: Blood and urine are the most common culture testing for sepsis patients. This study aimed to compare clinical characteristics and outcomes of sepsis patients by blood and urine culture positivity and to identify factors associated with positive cultures. Methods: This retrospective study included patients aged ≥16 years with sepsis identified by the Sepsis-3 criteria presenting to the emergency department at four hospitals between 2017 and 2019 in Australia. Patient clinical outcomes were in-hospital mortality, intensive care unit (ICU) admission, hospital length of stay, and representation following discharge. Four culture groups were defined based on the positivity of blood cultures (BC) and urine cultures (UC) ordered within 24 h of triage. Results: Of 4109 patient encounters with sepsis, 2730 (66%) were nonbacteremic, urine culture-negative (BC-UC-); 767 (19%) nonbacteremic, urine culture-positive (BC-UC+); 359 (9%) bacteremic, urine culture-negative (BC+UC-); and 253 (6%) bacteremic, urine culture-positive (BC+UC+). Compared with BC-UC- patients, BC+UC- patients had the highest risk of ICU admission (adjusted odds ratio [AOR] 95% CI: 1.60 [1.18-2.18]) while BC-UC+ patients had lowest risk (adjusted odds ratio [AOR]: 0.56 [0.41-0.76]). BC+UC- patients had the highest risk of 3-day representation (AOR: 1.51 [1.02-2.25]) and second longest hospital stay (adjusted relative risk 1.17 [1.03-1.34]). Antibiotic administration before sample collection for culture was associated with lower odds of positive blood or urine culture results (AOR: 0.38, p < 0.0001). Conclusions: Enhanced clinical care should be beneficial for nongenitourinary sepsis patients (BC+UC-) who had the highest comparative risk of adverse clinical outcomes. Every effort needs to be made to collect relevant culture samples before antibiotic administration, to follow up on culture results, and tailor treatment accordingly.

Spiced RCT: Success and Pain Associated with Intravenous Cannulation in the Emergency Department Randomized Controlled Trial.

BACKGROUND: Peripheral venous cannulation is one of the most common procedures in medicine. A larger cannula allows higher rates of fluid to be provided if needed in a deteriorating patient; however, it is also perceived that larger-gauge cannula placement is associated with increased pain and procedural difficulty. OBJECTIVE: This study aimed to compare the pain and procedural difficulty experienced during insertion between 18-gauge (18G) and 20-gauge (20G) cannulas. METHODS: We conducted a single-blinded, randomized controlled trial on adult patients who required peripheral IV cannulation within a tertiary hospital emergency department between April and October 2018. Patients were randomized to either the 18G or 20G cannula group. The primary outcomes of the study-pain experienced by patients and procedural difficulties experienced by clinical staff-were recorded on two separate 10-cm visual analog scales. Other outcomes include first-attempt success rate, operator designation, complications, and the intent and actual use of the IV cannula were documented on preformatted questionnaires. RESULTS: Data from 178 patients were included in the analysis. Eighty-nine patients were allocated to each cannula group. There were no statistically or clinically significant differences between mean pain score (0.23; 95% CI 0.56-1.02; p = 0.5662) and mean procedural difficulty score (0.12; 95% CI 0.66-0.93; p = 0.7396). between the two groups. There was no difference in first-attempt success rate (73 of 89 vs. 75 of 89; p = 0.1288), complications (2 of 89 vs. 1 of 89) between the 20G group and 18G group, respectively. CONCLUSIONS: There was no significant difference between the 18G or 20G cannula for either pain experienced by patients or procedural difficulty experienced by clinicians.

Blood Culture Ordering After Sepsis Alerts and Subsequent Patient Outcomes: An Electronic Health Record-Based Study.

Sepsis is a global health priority associated with high mortality. Clinical decision support systems have been developed to support clinicians with sepsis management. Ordering blood cultures (BCs) for suspected sepsis patients are strongly recommended by clinical guidelines. However, limited evidence exists investigating BC ordering following sepsis alerts and subsequent patient outcomes. This study aimed to investigate this issue using electronic health record data from an acute care hospital in Australia. Of 4,092 patients, only 16.6% had a BC ordered following a sepsis alert. The median time from the first sepsis alert to a BC order was 15.3 hours. Patients had 5.89 times higher odds of being diagnosed with sepsis if a BC was ordered following a sepsis alert than those without BC ordered (p<0.0001). Further investigation is needed to understand reasons behind the delay or failure to order a BC despite receiving electronic sepsis alerts and how decision support can be optimized to improve patient outcomes.

Facilitators and barriers of appropriate and timely initiation of intravenous fluids in patients with sepsis in emergency departments: a consensus development Delphi study.

BACKGROUND: Sepsis is a life-threatening medical emergency in which appropriate and timely administration of intravenous fluids to patients with features of hypotension is critical to prevent multi-organ failure and subsequent death. However, compliance with recommended fluid administration is reported to be poor. There is a lack of consensus among emergency clinicians on some of the determinant factors influencing fluid administration in sepsis. Thus, the aim of this study was to identify the level of consensus among key stakeholders in emergency departments regarding the facilitators, barriers, and strategies to improve fluid administration. METHODS: The modified Delphi questionnaire with 23 statements exploring barriers, facilitators, and strategies to improve fluid administration was developed from the integration of findings from previous phases of the study involving emergency department clinicians. A two-round modified Delphi survey was conducted among key stakeholders with managerial, educational, supervision and leadership responsibilities using a "Reactive Delphi technique" from March 2023 to June 2023. The statements were rated for importance on a 9-point Likert scale. The RAND/UCLA Appropriateness Method (RAM) was used to identify the level of consensus (agreement/disagreement). RESULTS: Of the 21 panellists who completed Round 1 survey, 18 (86%) also completed Round 2. The panellists rated 9 out of 10 (90%) barriers, 3 out of 4 (75%) facilitators and all 9 (100%) improvement strategies as important. Out of the total 23 statements, 18 (78%) had agreement among the panellists. Incomplete vital signs at triage (Median = 9, IQR 7.25 to 9.00) as a barrier, awareness of importance of fluid administration in sepsis (Median = 9, IQR 8.00 to 9.00) as facilitator and provision of nurse-initiated intravenous fluids (Median = 9, IQR 8.00 to 9.00) as an improvement strategy were the highest rated statements. CONCLUSION: This is the first Delphi study identifying consensus on facilitators, barriers, and strategies to specifically improve intravenous fluid administration in sepsis in Australia. We identified 18 consensus-based factors associated with appropriate and timely administration of intravenous fluids in sepsis. This study offers empirical evidence to support the implementation of the identified strategies to improve patient outcomes.

The Emergency nurse Protocols Initiating Care-Sydney Triage to Admission Risk Tool (EPIC-START) trial: protocol for a stepped wedge implementation trial.

INTRODUCTION: Emergency department (ED) overcrowding is a global problem and a threat to the quality and safety of emergency care. Providing timely and safe emergency care therein is challenging. To address this in New South Wales (NSW), Australia, the Emergency nurse Protocol Initiating Care-Sydney Triage to Admission Risk Tool (EPIC-START) was developed. EPIC-START is a model of care incorporating EPIC protocols, the START patient admission prediction tool, and a clinical deterioration tool to support ED flow, timely care, and patient safety. The aim of this study is to evaluate the impact of EPIC-START implementation across 30 EDs on patient, implementation, and health service outcomes. METHODS AND ANALYSIS: This study protocol adopts an effectiveness-implementation hybrid design (Med Care 50: 217-226, 2012) and uses a stepped-wedge cluster randomised control trial of EPIC-START, including uptake and sustainability, within 30 EDs across four NSW local health districts spanning rural, regional, and metropolitan settings. Each cluster will be randomised independently of the research team to 1 of 4 dates until all EDs have been exposed to the intervention. Quantitative and qualitative evaluations will be conducted on data from medical records and routinely collected data, and patient, nursing, and medical staff pre- and post-surveys. ETHICS AND DISSEMINATION: Ethical approval for the research was received from the Sydney Local Health District Research Ethics Committee (Reference Number 2022/ETH01940) on 14 December 2022. TRIAL REGISTRATION: Australian and New Zealand Clinical trial, ACTRN12622001480774p. Registered on 27 October 2022.

Facilitators and barriers of appropriate and timely initial fluid administration in sepsis: A qualitative study.

BACKGROUND: Sepsis is a medical emergency requiring prompt recognition, and early administration of intravenous fluids and antibiotics. While compliance with appropriate and timely administration of intravenous fluids has been found to be poor, the reasons are not well understood. Therefore, we have explored the experiences and perceptions of emergency nurses and medical officers from four hospitals to identify the associated facilitators and barriers. METHODS: Qualitative design incorporating six focus group discussions and thematic analysis of data. A hybrid approach using both inductive and deductive reasoning was used. FINDINGS: Four key themes were developed: 1. Overcrowding and understaffing threaten appropriate fluid management in sepsis; 2. Variations in clinical practice results in suboptimal fluid management; 3. Challenges with clinical recognition of sepsis impedes timely fluid administration; 4. Top-down approach is necessary to improve fluid management. CONCLUSION: Themes highlighted the specific challenges associated with fluid administration in sepsis in the emergency department setting providing potential strategies to be implemented to improve practice and ultimately patient outcomes.

Association between timing and adequacy of antibiotics and adverse outcomes in patients with sepsis and septic shock: A multicentre retrospective cohort study.

OBJECTIVES: To investigate the association between the timing and adequacy of antibiotics administered to patients presenting with culture-positive sepsis and septic shock to the ED and in-hospital mortality and/or intensive care unit (ICU) admission. METHODS: Multicentre retrospective cohort study of ED presentations at four metropolitan hospitals in Sydney, Australia between January 2017 and November 2019. Encounters for patients aged ≥16 years meeting specified criteria for sepsis or septic shock with antibiotic administration within the first 6 h of presentation were included. RESULTS: Of 7611 encounters included in the study, 2328 (31%) were culture positive, and 2228 (29%) met the criteria for septic shock. In culture-positive sepsis encounters, partial or inadequate antibiotic coverage was associated with higher risk of death or ICU admission (adjusted odds ratio [AOR] 1.50, 95% confidence interval [CI] 1.04-2.06 and 1.95, 95% CI 1.28-2.99, respectively). This effect was not significant in septic shock encounters (AOR 1.10, 95% CI 0.64-1.88) with partial coverage and (AOR 1.63, 95% CI 0.81-3.3) inadequate coverage. Time to antibiotics was not significantly associated with the risk of mortality/ICU admission. This inference remained the same when analysis was restricted to cases with adequate antibiotic coverage. CONCLUSIONS: In a large multicentre sample of patients with culture-positive sepsis, inadequacy of antibiotics was associated with higher risk of in-hospital mortality or ICU admission.

Toxicology presentations to a tertiary unit in New South Wales during the COVID-19 pandemic first wave: A retrospective comparison study.

OBJECTIVE: To compare presentation numbers, class of exposure, poison severity score (PSS) and drugs ingested by patients in a tertiary toxicology service during the first wave of the COVID-19 pandemic to the corresponding time periods in 2018 and 2019. METHODS: A retrospective cohort observational study of patients admitted or consulted to the Western Sydney Toxicology Service (WSTS) from ED during February to May in 2018-2020. Patient age, sex, triage category, time and date of arrival, mode of arrival, type of poisoning, discharge location, length of stay and PSS were collected from WSTS database and electronic medical records. The total number of ED presentations, hospital admissions and toxicology admissions were gathered from hospital-based data services. RESULTS: There was an overall increase in toxicology presentations in February to May 2020 (n = 441) compared to 2019 (n = 333) and 2018 (n = 255). The daily rate of presentations increased in March to May 2020 with an overall rate ratio of 1.42, 95% confidence interval 1.23-1.63, P < 0.001. There was an increase in presentations across all drug types. From March to April 2020, there was significantly higher number of daily presentations for recreational drugs use compared to 2018. CONCLUSION: There was a relative increase in toxicology presentations during the COVID-19 pandemic compared to an overall decrease in presentations to ED. Recreational drug use increased significantly during the pandemic compared to 2018.

Identifying factors associated with intravenous fluid administration in patients with sepsis presenting to the emergency department: a retrospective cohort study.

BACKGROUND: Appropriate and timely administration of intravenous fluids to patients with sepsis-induced hypotension is one of the mainstays of sepsis management in the emergency department (ED), however, fluid resuscitation remains an ongoing challenge in ED. Our study has been undertaken with two specific aims: firstly, for patients with sepsis, to identify factors associated with receiving intravenous fluids while in the ED; and, secondly to identify determinants associated with the actual time to fluid administration. METHODS: We conducted a retrospective multicentre cohort study of adult ED presentations between October 2018 and May 2019 in four metropolitan hospitals in Western Sydney, Australia. Patients meeting pre-specified criteria for sepsis and septic shock and treated with antibiotics within the first 24 h of presentation were included. Multivariable models were used to identify factors associated with fluid administration in sepsis. RESULTS: Four thousand one hundred forty-six patients met the inclusion criteria, among these 2,300 (55.5%) patients with sepsis received intravenous fluids in ED. The median time to fluid administration from the time of diagnosis of sepsis was 1.6 h (Interquartile Range (IQR) 0.5 to 3.8), and the median volume of fluids administered was 1,100 mL (IQR 750 to 2058). Factors associated with patients receiving fluids were younger age (Odds Ratio (OR) 1.05, 95% Confidence Interval (CI (1.03 to 1.07), p < 0.001); lower systolic blood pressure (OR 1.11, 95% CI (1.08 to 1.13), p < 0.001); presenting to smaller hospital (OR 1.48, 95% CI (1.25 to 1.75, p < 0.001) and a Clinical Rapid Response alert activated (OR 1.64, 95% CI (1.28 to 2.11), p < 0.001). Patients with Triage Category 1 received fluids 101.22 min earlier (95% CI (59.3 to131.2), p < 0.001) and those with Category 2 received fluids 43.58 min earlier (95% CI (9.6 to 63.1), p < 0.001) compared to patients with Triage Category 3-5. Other factors associated with receiving fluids earlier included septic shock (-49.37 min (95% CI (-86.4 to -12.4), p < 0.001)); each mmol/L increase in serum lactate levels (-9.0 min, 95% CI (-15.7 to -2.3), p < 0.001) and presenting to smaller hospitals (-74.61 min, 95% CI (-94.0 to -55.3), p < 0.001). CONCLUSIONS: Younger age, greater severity of sepsis, and presenting to a smaller hospital increased the probability of receiving fluids and receiving it earlier. Recognition of these factors may assist in effective implementation of sepsis management guidelines which should translate into better patient outcomes. Future studies are needed to identify other associated factors that we have not explored.

Early fluid bolus in adults with sepsis in the emergency department: a systematic review, meta-analysis and narrative synthesis.

BACKGROUND: Early intravenous fluids for patients with sepsis presenting with hypoperfusion or shock in the emergency department remains one of the key recommendations of the Surviving Sepsis Campaign guidelines to reduce mortality. However, compliance with the recommendation remains poor. While several interventions have been implemented to improve early fluid administration as part of sepsis protocols, the extent to which they have improved compliance with fluid resuscitation is unknown. The factors associated with the lack of compliance are also poorly understood. METHODS: We conducted a systematic review, meta-analysis and narrative review to investigate the effectiveness of interventions in emergency departments in improving compliance with early fluid administration and examine the non-interventional facilitators and barriers that may influence appropriate fluid administration in adults with sepsis. We searched MEDLINE Ovid/PubMed, Ovid EMBASE, CINAHL, and SCOPUS databases for studies of any design to April 2021. We synthesised results from the studies reporting effectiveness of interventions in a meta-analysis and conducted a narrative synthesis of studies reporting non-interventional factors. RESULTS: We included 31 studies out of the 825 unique articles identified in the systematic review of which 21 were included in the meta-analysis and 11 in the narrative synthesis. In meta-analysis, interventions were associated with a 47% improvement in the rate of compliance [(Random Effects (RE) Relative Risk (RR) = 1.47, 95% Confidence Interval (CI), 1.25-1.74, p-value < 0.01)]; an average 24 min reduction in the time to fluids [RE mean difference = - 24.11(95% CI - 14.09 to - 34.14 min, p value < 0.01)], and patients receiving an additional 575 mL fluids [RE mean difference = 575.40 (95% CI 202.28-1353.08, p value < 0.01)]. The compliance rate of early fluid administration reported in the studies included in the narrative synthesis is 48% [RR = 0.48 (95% CI 0.24-0.72)]. CONCLUSION: Performance improvement interventions improve compliance and time and volume of fluids administered to patients with sepsis in the emergency department. While patient-related factors such as advanced age, co-morbidities, cryptic shock were associated with poor compliance, important organisational factors such as inexperience of clinicians, overcrowding and inter-hospital transfers were also identified. A comprehensive understanding of the facilitators and barriers to early fluid administration is essential to design quality improvement projects. PROSPERO REGISTRATION ID: CRD42021225417.

Can existing electronic medical records be used to quantify cardiovascular risk at point of care?

BACKGROUND: Using electronic data for cardiovascular risk stratification could help in prioritising healthcare access and optimise cardiovascular prevention. AIMS: To determine whether assessment of absolute cardiovascular risk (Australian absolute cardiovascular disease risk (ACVDR)) and short-term ischaemic risk (History, ECG, Age, Risk factors, and Troponin (HEART) score) is possible from available data in Electronic Medical Record (EMR) and My Health Record (MHR) of patients presenting with acute cardiac symptoms to a Rapid Access Cardiology Clinic (RACC). METHODS: Audit of EMR and MHR on 200 randomly selected adults who presented to RACC between 1 March 2017 and 4 February 2020. The main outcomes were the proportion of patients for which ACVDR score and HEART score could be calculated. RESULTS: Mean age was 55.2 ± 17.8 years and 43% were female. Most (85%) were referred from emergency for chest pain (52%). Forty-six percent had hypertension, 35% obesity, 20% diabetes mellitus, 17% ischaemic heart disease and 18% were current smokers. There was no significant difference in MHR accessibility with age, gender and number of comorbidities. An ACVDR score could be estimated for 17.5% (EMR) and 0% (MHR) of patients. None had complete data to estimate HEART score in either EMR or MHR. Most commonly missing variables for ACVDR score were blood pressure (MHR) and high-density lipoprotein cholesterol (EMR), and for HEART score the missing variables were body mass index and comorbidities (MHR and EMR). CONCLUSIONS: Significant gaps are apparent in electronic medical data capture of key variables to perform cardiovascular risk assessment. Medical data capture should prioritise the collection of clinically important data to help address gaps in cardiovascular management.

Association between intravenous fluid resuscitation and outcome among patients with suspected infection and sepsis: A retrospective cohort study.

OBJECTIVE: To investigate the association between timing and volume of intravenous fluids administered to ED patients with suspected infection and all-cause in-hospital mortality. METHODS: Retrospective cohort study of ED presentations at four metropolitan hospitals in Sydney, Australia, between October 2018 and May 2019. Patients over 16 years of age with suspected infection who received intravenous fluids within 24 h of presentation were included. RESULTS: During the study period, 7533 patients with suspected infection received intravenous fluids. Of these, 1996 (26.5%) and 231 (3.1%) had suspected sepsis and septic shock, respectively. Each 1000 mL increase in intravenous fluids administered was associated with a reduction in risk of in-hospital mortality (adjusted odds ratio [AOR] 0.87, 95% confidence interval [CI] 0.76-0.99). This association was stronger in patients with septic shock (AOR 0.66, 95% CI 0.49-0.89), and those admitted to intensive care unit (ICU) (AOR 0.74, 95% CI 0.56-0.96). Patients with suspected sepsis and septic shock who received a total volume of >3600 mL had lower in-hospital mortality (AOR 0.44, 95% CI 0.22-0.91; AOR 0.16, 95% CI 0.05-0.57) compared to those administered <3600 mL within the first 24 h of presenting to the ED. There was no association between the time of initiation of fluids and in-hospital mortality among survivors and non-survivors (2.3 vs 2.5 h, P = 0.50). CONCLUSION: We observed a reduction in risk of in-hospital mortality for each 1000 mL increase in intravenous fluids administered in patients with septic shock or admitted to ICU suggesting illness severity to be a likely effect modifier.

Closing the sepsis gap: an exploration of sepsis presentations at a remote north Australian emergency department.

INTRODUCTION: The objective of this study was to explore the demographic characteristics, disease specifics and outcomes of adult patients with suspected sepsis presenting to a remote Australian emergency department (ED). A retrospective, uninterrupted time series audit of ED patients presenting with suspected sepsis was conducted. A total of 189 remote presentations were reviewed based on the time of clinician identification of sepsis. METHODS: Retrospective cohort analysis was performed for all adult patients with suspected or confirmed sepsis. RESULTS: A majority of patients presenting with sepsis to a remote hospital were Indigenous (61.9%) with a large proportion (34.9%) presenting by ambulance. Median age was 50 years. Indigenous patients (44.7%, 95%CI 34.1-55.9) were more likely to meet the quick Sequential (sepsis-related) Organ Failure Assessment criteria compared to non-Indigenous patients (27.1% 95%CI 16.6-41.0) (p=0.05 95%CI -1.1-34.3) with higher rates of critical care admission (34.2% v 10.4%) (difference 23.8, p=0.003, 95%CI 7.7-37.5). Congruent with previous research, Indigenous status did not confer a difference in sepsis mortality (12.1% v 11.8%, p=0.91). CONCLUSION: Remote Indigenous patients have worse clinical sepsis scores, are more likely to present by ambulance and require skin and soft tissue source control. This cohort has higher lactate values and critical care requirements but similar mortality rates. Improving access to culturally safe medical care could address this disparity.

Utility of venous blood gases for the assessment of traumatic shock: a prospective observational study.

BACKGROUND: ABG samples are often obtained in trauma patients to assess shock severity. Venous blood gas (VBG) sampling, which is less invasive, has been widely used to assess other forms of shock. The study aim was to determine the agreement between VBG and ABG measurements in trauma. METHODS: Patients were enrolled at an Australian trauma centre between October 2016 and October 2018. Bland-Altman limits of agreement (LOA) between paired blood gas samples taken <30 min apart were used to quantify the extent of agreement. The impact of using only VBG measurements was considered using an a priori plan. Cases where venous sampling failed to detect 'concerning levels' were flagged using evidence-based cut-offs: pH ≤7.2, base deficit (BD) ≤-6, bicarbonate <21 and lactate ≥4. Case summaries of these patients were assessed by independent trauma clinicians as to whether an ABG would change expected management. RESULTS: During the study period 176 major trauma patients had valid paired blood gas samples available for analysis. The median time difference between paired measurements was 11 min (IQR 6-17). There was a predominance of men (81.8%) and blunt trauma (92.0%). Median Injury Severity Score was 13 (range 1-75) and inpatient mortality was 6.3%. Mean difference (ABG-VBG) and LOA between paired arterial and venous measurements were 0.036 (LOA -0.048 to 0.120) for pH, -1.27 mmol/L (LOA -4.35 to 1.81) for BD, -0.64 mmol/L (LOA -1.86 to 0.57) for lactate and -1.97 mmol/L (LOA -5.49 to 1.55) for bicarbonate. Independent assessment of the VBG 'false negative' cases (n=20) suggested an ABG would change circulatory management in two cases. CONCLUSIONS: In trauma patients VBG and ABG parameters displayed suboptimal agreement. However, in cases flagged as VBG 'false negative' independent review indicated that the availability of an ABG was unlikely to change management.

Quality use of diagnostic imaging in trauma, and the impact on emergency medical practice-a retrospective clinical audit.

PURPOSE: The current study aimed to evaluate the appropriateness of the decisions made to order diagnostic imaging for emergency trauma patients. METHODS: A retrospective audit of 642 clinical decisions for 370 non-consecutive trauma cases was conducted at a level 1 adult trauma centre. Radiographs and computed tomography investigations were compared for compliance with the American College of Radiology/European Society of Radiology (ACR/ESR) imaging guidelines. The non-compliant imaging orders were evaluated for the following medical officer ranks: Junior Medical Officer (JMO), Senior Resident Medical Officer (SRMO), Emergency Medicine Registrar (EMR) and Consultants/Career Medical Officer (CMO). The time of imaging order (day/evening or night shift), whether the imaging led to conservative or surgical patient management, and whether the imaging order decisions led to a change in the clinical management of the patient were also considered. RESULTS: Non-compliance with the ACR/ESR guidelines was low. At least half of non-compliant decisions made by each level of medical officer resulted in a change in patient management. In total, 11 (65%), 25 (53%), 12 (63%) and 11 (52%) of non-compliant imaging orders placed by JMOs, SRMOs, EMRs and Consultants, respectively, resulted in change of patient management. JMOs and SRMOs ordered a significantly higher proportion of non-compliant imaging studies for conservatively managed patients in comparison to surgically managed patients. CONCLUSION: This study highlighted that most non-compliant imaging orders benefited the patient.

Evaluation of an augmented emergency department electronic medical record-based sepsis alert.

OBJECTIVE: Electronic medical records-based alerts have shown mixed results in identifying ED sepsis. Augmenting clinical patient-flagging with automated alert systems may improve sepsis screening. We evaluate the performance of a hybrid alert to identify patients in ED with sepsis or in-hospital secondary outcomes from infection. METHODS: We extracted a dataset of all patients with sepsis during the study period at five participating Western Sydney EDs. We evaluated the hybrid alert's performance for identifying patients with a discharge diagnosis related to infection and modified sequential sepsis-related organ functional assessment (mSOFA) score ≥2 in ED and also compared the alert to rapid bedside screening tools to identify patients with infection for secondary outcomes of all-cause in-hospital death and/or intensive care unit admission. RESULTS: A total of 118 178 adult patients presented to participating EDs during study period with 1546 patients meeting ED sepsis criteria. The hybrid alert had a sensitivity - 71.2% (95% confidence interval 68.8-73.4), specificity - 96.4% (95% confidence interval 96.3-96.5) for identifying ED sepsis. Clinician flagging identified additional alert-negative 232 ED sepsis and 63 patients with secondary outcomes and 112 alert-positive patients with infection and ED mSOFA score <2 went on to die in hospital. CONCLUSION: The hybrid alert performed modestly in identifying ED sepsis and secondary outcomes from infection. Not all infected patients with a secondary outcome were identified by the alert or mSOFA score ≥2 threshold. Augmenting clinical practice with auto-alerts rather than pure automation should be considered as a potential for sepsis alerting until more reliable algorithms are available for safe use in clinical practice.