When the machine is wrong. Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machine-learning model.

BACKGROUND: A machine-learning model trained to recognize emergency calls regarding Out-of-Hospital Cardiac Arrest (OHCA) was tested in clinical practice at Copenhagen Emergency Medical Services (EMS) from September 2018 to December 2019. We aimed to investigate emergency call characteristics where the machine-learning model failed to recognize OHCA or misinterpreted a call as being OHCA. METHODS: All emergency calls were linked to the dispatch database and verified OHCAs were identified by linkage to the Danish Cardiac Arrest Registry. Calls with either false negative or false positive predictions of OHCA were evaluated by trained auditors. Descriptive analyses were performed with absolute numbers and percentages reported. RESULTS: The machine-learning model processed 169,236 calls to Copenhagen EMS and suspected 5,811 (3.4%) of the calls as OHCA, resulting in 84.5% sensitivity and 97.1% specificity. Among OHCAs not recognised by machine-learning model, a condition completely different from OHCA was presented by caller in 31% of the cases. In 28% of unrecognised calls, patient was reported breathing normally, and language barriers were identified in 23% of the cases. Among falsely suspected OHCA, the patient was reported unconscious in 28% of the cases, and in 13% of the false positive cases the machine-learning model interpreted calls regarding dead patients with irreversible signs of death as OHCA. CONCLUSION: Continuous optimization of the language model is needed to improve the prediction of OHCA and thereby improve sensitivity and specificity of the machine-learning model on recognising OHCA in emergency telephone calls.

Evaluation of a critical incident management system on mental health in lifeguard organisations: a retrospective study.

Background: Lifeguards may face many life-threatening situations during their careers and may be at increased risk of post-traumatic stress disorder (PTSD). Minimal evidence concerning critical incident management systems in lifeguard organisations exists. Objectives: To develop, implement and evaluate an operational system for critical incident management in lifeguard organisations. Methods: This retrospective study included data on occupational injury reports from 2013 to 2022 in TrygFonden Surf Lifesaving Denmark. All active lifeguards were invited to evaluate the system and the individual steps using an online questionnaire with three questions rated on a 5-point Likert scale. Primary outcome was a change in the frequency of psychological injury reports after system implementation in 2020. The secondary outcome was the lifeguards' satisfaction with the system. Results: After implementation, the average annual number of psychological injury reports increased 6.5-fold from 2 (2013-2019) to 13 (2020-2022), without changes to the number of critical incidents attended by the lifeguards. Sixty-six (33.8%) active lifeguards answered the questionnaire and agreed that follow-up after critical incidents was very important (mean score 4.7/5). Satisfaction with steps 1-2 and 3 of critical incident management among involved lifeguards was high (mean score 4.4/5 and 4.6/5, respectively). The system included an operational workflow diagram and incident report template presented in this study. Conclusions: The operational system for critical incident management may improve early recognition of symptoms for the prevention of PTSD. It may be used as a screening and decision tool for referral to a mental health professional.

Artificial intelligence in Emergency Medical Services dispatching: assessing the potential impact of an automatic speech recognition software on stroke detection taking the Capital Region of Denmark as case in point.

BACKGROUND AND PURPOSE: Stroke recognition at the Emergency Medical Services (EMS) impacts the stroke treatment and thus the related health outcome. At the EMS Copenhagen 66.2% of strokes are detected by the Emergency Medical Dispatcher (EMD) and in Denmark approximately 50% of stroke patients arrive at the hospital within the time-to-treatment. An automatic speech recognition software (ASR) can increase the recognition of Out-of-Hospital cardiac arrest (OHCA) at the EMS by 16%. This research aims to analyse the potential impact an ASR could have on stroke recognition at the EMS Copenhagen and the related treatment. METHODS: Stroke patient data (n = 9049) from the years 2016-2018 were analysed retrospectively, regarding correlations between stroke detection at the EMS and stroke specific, as well as personal characteristics such as stroke type, sex, age, weekday, time of day, year, EMS number contacted, and treatment. The possible increase in stroke detection through an ASR and the effect on stroke treatment was calculated based on the impact of an existing ASR to detect OHCA from CORTI AI. RESULTS: The Chi-Square test with the respective post-hoc test identified a negative correlation between stroke detection and females, the 1813-Medical Helpline, as well as weekends, and a positive correlation between stroke detection and treatment and thrombolysis. While the association analysis showed a moderate correlation between stroke detection and treatment the correlation to the other treatment options was weak or very weak. A potential increase in stroke detection to 61.19% with an ASR and hence an increase of thrombolysis by 5% in stroke patients calling within time-to-treatment was predicted. CONCLUSIONS: An ASR can potentially improve stroke recognition by EMDs and subsequent stroke treatment at the EMS Copenhagen. Based on the analysis results improvement of stroke recognition is particularly relevant for females, younger stroke patients, calls received through the 1813-Medical Helpline, and on weekends. TRIAL REGISTRATION: This study was registered at the Danish Data Protection Agency (PVH-2014-002) and the Danish Patient Safety Authority (R-21013122).

Behavioral factors associated with SARS-CoV-2 infection.

OBJECTIVE: To study the association between behavioural factors and incidence rates of SARS-CoV-2 infection. DESIGN: Case-control web-based questionnaire study. SETTING: Questionnaire data were collected in the Capital Region of Denmark in December 2020 when limited restrictions were in place, while the number of daily SARS-CoV-2 cases increased rapidly. PARTICIPANTS: 8913 cases of laboratory-confirmed SARS-CoV-2 infection were compared with two groups of controls: (1) 34 063 individuals with a negative SARS-CoV-2 test from the same date (negative controls, NCs) and 2) 25 989 individuals who had never been tested for a SARS-CoV-2 infection (untested controls, UC). Controls were matched on sex, age, test date and municipality. EXPOSURE: Activities during the 14 days prior to being tested positive for SARS-CoV-2 or during the same period for matched controls and precautions taken during the entire pandemic. MAIN OUTCOMES AND MEASURES: SARS-CoV-2 infection incidence rate ratios (IRR). RESULTS: Response rate was 41.4% (n=93 121). Using public transportation, grocery shopping (IRR: NC: 0.52; UC: 0.63) and outdoor sports activities (NC: 0.75; UC: 0.96) were not associated with increased rate of SARS-CoV-2 infection. Most precautions, for example, using hand sanitizer (NC: 0.79; UC: 0.98), physical distancing (NC: 0.79; UC: 0.82) and avoiding handshakes (NC: 0.74; UC: 0.77), were associated with a lower rate of infection. Activities associated with many close contacts, especially indoors, increased rate of infection. Except for working from home, all types of occupation were linked to increased rate of infection. CONCLUSIONS: In a community setting with moderate restrictions, activities such as using public transportation and grocery shopping with the relevant precautions were not associated with an increased rate of SARS-CoV-2 infection. Exposures and activities where safety measures are difficult to maintain might be important risk factors for infection. These findings may help public health authorities tailor their strategies for limiting the spread of SARS-CoV-2.

Symptom presentation of SARS-CoV-2-positive and negative patients: a nested case-control study among patients calling the emergency medical service and medical helpline.

OBJECTIVE: Currently effective symptom-based screening of patients suspected of COVID-19 is limited. We aimed to investigate age-related differences in symptom presentations of patients tested positive and negative for SARS-CoV-2. DESIGN: SETTING: Calls to the medical helpline (1-8-1-3) and emergency number (1-1-2) in Copenhagen, Denmark. At both medical services all calls are recorded. PARTICIPANTS: We included calls for patients who called for help/guidance at the medical helpline or emergency number prior to receiving a test for SARS-CoV-2 between April 1st and 20th 2020 (8423 patients). Among these calls, we randomly sampled recorded calls from 350 patients who later tested positive and 250 patients tested negative and registered symptoms described in the call. OUTCOME: RESULTS: After exclusions, 544 calls (312 SARS-CoV-2 positive and 232 negative) were included in the analysis. Fever and cough remained the two most common of COVID-19 symptoms across all age groups and approximately 42% of SARS-CoV-2 positive and 20% of negative presented with both fever and cough. Symptoms including nasal congestion, irritation/pain in throat, muscle/joint pain, loss of taste and smell, and headache were common symptoms of COVID-19 for patients younger than 60 years; whereas loss of appetite and feeling unwell were more commonly seen among patients over 60 years. Headache and loss of taste and smell were rare symptoms of COVID-19 among patients over 60 years. CONCLUSION: Our study identified age-related differences in symptom presentations of SARS-CoV-2-positive patients calling for help or medical advice. The specific symptoms of loss of smell or taste almost exclusively reported by patients younger than 60 years. Differences in symptom presentation across age groups must be considered when screening for COVID-19.

Does an emergency access button increase the patients' satisfaction and feeling of safety with the out-of-hours health services? A randomised controlled trial in Denmark.

OBJECTIVE: To investigate if the option to bypass the telephone queue can increase satisfaction and feeling of safety in callers. DESIGN: Randomised controlled parallel superiority trial. Data from questionnaire survey. SETTING: Two out-of-hours (OOH) services in Denmark. PARTICIPANTS: 217 510 citizens who called the OOH services between 4 September 2017 and 30 November 2017. RANDOMISATION: Two-faze study period: First half with randomisation of participants based on their date of birth; even date randomised to intervention, uneven date randomised to control group. Second half with all participants included in intervention group. INTERVENTION: Providing randomised callers (intervention group n=146 355) with the option to bypass the telephone waiting line through an emergency access button (EAB), while the rest got the normal service (control group n=71 155). All EAB users were invited to a questionnaire survey as well as random participants who did not use the EAB (of whom approximately 50% did not have the EAB option). MAIN OUTCOME MEASURES: Satisfaction and feeling of safety in callers. RESULTS: 2208 of 6704 (32.9%) invited callers answered the questionnaire (intervention group n=1415 (users n=621, non-users n=794); control group n=793). The OR for answering in the two categories with highest satisfaction when provided with the EAB option was 1.34 (95% CI 1.07 to 1.68) for satisfaction with the waiting time, 1.21 (95% CI 0.91 to 1.60) for overall satisfaction and 1.46 (95% CI 1.12 to 1.89) for feeling of safety. Approximately 72% (441/621) of EAB users reported that the EAB option increased their feeling of safety with the OOH services 'to a high degree' compared with 25% (197/794) of callers who had the EAB option without using it. CONCLUSIONS: The EAB can provide fast access to OOH telephone advice in case of severe illness. It favours citizens perceived in most need of urgent healthcare and significantly increases both feeling of safety and patient satisfaction. TRIAL REGISTRATION DETAILS: NCT02572115 (5 October 2015).

Smartphone Activation of Citizen Responders to Facilitate Defibrillation in Out-of-Hospital Cardiac Arrest.

BACKGROUND: Dispatching citizen responders through a smartphone application (app) holds the potential to increase bystander cardiopulmonary resuscitation (CPR) and defibrillation in out-of-hospital cardiac arrest (OHCA). OBJECTIVES: This study investigated arrival at the OHCA location of app-dispatched citizen responders before the Emergency Medical Services (EMS) and the association with bystander CPR and bystander defibrillation. METHODS: Suspected OHCAs with alerted citizen responders from September 1, 2017, to August 31, 2018, were included. Citizen responders located 1.8 km (1.1 miles) from the OHCA were dispatched to start CPR or retrieve an automated external defibrillator. OHCAs where at least 1 citizen responder arrived before EMS were compared with OHCAs where EMS arrived first. In both groups, random bystanders could be present before the arrival of citizen responders and the EMS. Primary outcomes were bystander CPR and bystander defibrillation, which included CPR and defibrillation by citizen responders and random bystanders. RESULTS: Citizen responders were alerted in 819 suspected OHCAs, of which 438 (53.5%) were confirmed cardiac arrests eligible for inclusion. At least 1 citizen responder arrived before EMS in 42.0% (n = 184) of all included OHCAs. When citizen responders arrived before EMS, the odds for bystander CPR increased (odds ratio: 1.76; 95% confidence interval: 1.07 to 2.91; p = 0.027) and the odds for bystander defibrillation more than tripled (odds ratio: 3.73; 95% confidence interval: 2.04 to 6.84; p < 0.001) compared with OHCAs in which citizen responders arrived after EMS. CONCLUSIONS: Arrival of app-dispatched citizen responders before EMS was associated with increased odds for bystander CPR and a more than 3-fold increase in odds for bystander defibrillation. (The HeartRunner Trial; NCT03835403).

Communication quality in telephone triage conducted by general practitioners, nurses or physicians: a quasi-experimental study using the AQTT to assess audio-recorded telephone calls to out-of-hours primary care in Denmark.

OBJECTIVES: To compare the quality of communication in out-of-hours (OOH) telephone triage conducted by general practitioners (GPs), nurses using a computerised decision support system and physicians with different medical specialities, and to explore the association between communication quality and efficiency, length of call and the accuracy of telephone triage. DESIGN: Natural quasi-experimental cross-sectional study. SETTING: Two Danish OOH services using different telephone triage models: a GP cooperative and the medical helpline 1813. PARTICIPANTS: 1294 audio-recorded randomly selected OOH telephone triage calls from 2016 conducted by GPs (n=423), nurses using CDSS (n=430) and physicians with different medical specialities (n=441). MAIN OUTCOME MEASURES: Twenty-four physicians assessed the calls. The panel used a validated assessment tool (Assessment of Quality in Telephone Triage, AQTT) to measure nine aspects of communication, overall perceived communication quality, efficiency and length of call. RESULTS: The risk of poor quality was significantly higher in calls triaged by GPs compared with calls triaged by nurses regarding 'allowing the caller to describe the situation' (GP: 13.5% nurse: 9.8%), 'mastering questioning techniques' (GP: 27.4% nurse: 21.1%), 'summarising' (GP: 33.0% nurse: 21.0%) and 'paying attention to caller's experience' (GP: 25.7% nurse: 17.0%). The risk of poor quality was significantly higher in calls triaged by physicians compared with calls triaged by GPs in five out of nine items. GP calls were significantly shorter (2 min 57 s) than nurse calls (4 min 44 s) and physician calls (4 min 1 s). Undertriaged calls were rated lower than optimally triaged calls for overall quality of communication (p<0.001) and all specific items. CONCLUSIONS: Compared with telephone triage by GPs, the communication quality was higher in calls triaged by nurses and lower in calls triaged by physicians with different medical specialities. However, calls triaged by nurses and physicians were longer and perceived less efficient. Quality of communication was associated with accurate triage.

Injury from electric scooters in Copenhagen: a retrospective cohort study.

OBJECTIVE: To analyse injuries related to manual and electric scooter use from January 2016 up to and including July 2019. SETTING: Electric scooter rental services were launched in Denmark in January 2019. The services were provided by private companies. Although rules for handling and riding scooters have been established, no reports either before or after introduction of electric scooters anticipated the full extent of use, and injuries to riders and pedestrians. PARTICIPANTS: All patient records mentioning manual or electric scooters. Records were reviewed, and data were stratified according to two groups: manual and electric scooters. INTERVENTIONS: A predefined survey was completed in all cases where 'scooter' was present. This contained variables such as type of scooter, type of participant, mechanism of injury, acuity, intoxication, referral to treatment facility. OUTCOME MEASURES: Among incidents involving scooters, summary statistics on continuous and categorical variables of interest were reported. RESULTS: 468 scooter-related injuries were recorded. We found that manual scooter riders were more likely to be children under the age of 15; fall alone-involving no other party; sustain contusions, sprains and lacerations; and bruise either their fingers or toes. Riders of electric scooters were likely to be 18-25 years, sustain facial bruising and lacerations requiring sutures, and be under the influence of alcohol or drugs. Non-riders of electric scooters were mostly elderly people who tripped over scooters, consequently sustaining moderate to severe injuries. CONCLUSION: There were two different types of population sustaining injuries from manual and electric scooters, respectively. The proportion of non-riders injured by electric scooters were surprisingly large (17%), and while electric scooters are here to stay, several apparently preventable injuries occur as a result of reckless driving and discarded electric scooters. Current rules for usage might not prevent unnecessary accidents and secure traffic safety and the lives of older individuals.

Satisfaction of 30 402 callers to a medical helpline of the Emergency Medical Services Copenhagen: a retrospective cohort study.

OBJECTIVES: To keep healthcare systems sustainable for future demands, many countries are developing a centralised telephone line for out-of-hours primary care services. To increase the quality of such services, more information is needed on factors that influence caller satisfaction. The aim of this study was to identify demographic and call-related characteristics that are associated with the patient satisfaction of callers to a medical helpline in Denmark. DESIGN: Retrospective cohort study on patient registry data and questionnaire results. SETTING: Non-emergency medical helpline in the Capital Region of Denmark. PARTICIPANTS: A random sample of 30 402 callers to the medical helpline between May 2016 and May 2018. PRIMARY AND SECONDARY OUTCOME MEASURES: Responses of a satisfaction questionnaire were linked to demographic and call-related dispatch data. Associations between the characteristics were analysed with multivariable logistic regression analysis with satisfaction as the dependent variable. A subgroup analysis was performed on callers for children aged between 0 and 4 years. RESULTS: Of the 30 402 analysed callers, 73.0% were satisfied with the medical helpline. Satisfaction was associated with calling for a somatic injury (OR: 1.96, 95% CI: 1.72 to 2.23), receiving a face-to-face consultation (OR: 2.27, 95% CI: 2.04 to 2.50) and a waiting time less than 10 min (OR: 1.82, 95% CI: 1.56 to 2.08). Callers for a 0-year to 4-year-old patient were more likely to be satisfied when they called for a somatic illness or received a telephone consultation, compared with the rest of the population (p<0.0001). CONCLUSION: Callers were in general satisfied with the medical helpline. Satisfaction was associated with reason for encounter, triage response and waiting time. People calling for 0-year to 4-year-old patients were, compared with the rest of the population, more frequently satisfied when they called for a somatic illness or received a telephone consultation.

Sociodemographic and health-related determinants for making repeated calls to a medical helpline: a prospective cohort study.

OBJECTIVES: To identify sociodemographic and health-related characteristics of callers' making repeated calls within 48 hours to a medical helpline, compared with those who only call once. SETTING: In the Capital Region of Denmark people with acute, non-life-threatening illnesses or injuries are triaged through a single-tier medical helpline for acute, healthcare services. PARTICIPANTS: People who called the medical helpline between 18 January and 9 February 2017 were invited to participate in the survey. During the period, 38 787 calls were handled and 12 902 agreed to participate. Calls were excluded because of the temporary civil registration number (n=78), the call was not made by the patient or a close relative (n=699), or survey responses were incomplete (n=19). Hence, the analysis included 12 106 calls, representing 11.131 callers' making single calls and 464 callers' making two or more calls within 48 hours. Callers' data (age, sex and caller identification) were collected from the medical helpline's electronic records. Data were enriched using the callers' self-rated health, self-evaluated degree of worry, and registry data on income, ethnicity and comorbidities. The OR for making repeated calls was calculated in a crude, sex-adjusted and age-adjusted analysis and in a mutually adjusted analysis. RESULTS: The crude logistic regression analysis showed that age, self-rated health, self-evaluated degree of worry, income, ethnicity and comorbidities were significantly associated with making repeated calls. In the mutually adjusted analysis associations decreased, however, odds ratios remained significantly decreased for callers with a household income in the middle (OR=0.71;95% CI 0.54 to 0.92) or highest (OR=0.68;95% CI 0.48 to 0.96) quartiles, whereas immigrants had borderline significantly increased OR (OR=1.34;95% CI 0.96 to 1.86) for making repeated calls. CONCLUSIONS: Findings suggest that income and ethnicity are potential determinants of callers' need to make additional calls within 48 hours to a medical helpline with triage function.

An international case-control study of interleukin-4Ralpha, interleukin-13, and cyclooxygenase-2 polymorphisms and glioblastoma risk.

Previous studies found that allergies are inversely related to risk of glioma. In an earlier publication, using data from a Swedish case-control study, Schwartzbaum et al. report an inverse relation between risk of glioblastoma and four single nucleotide polymorphisms (SNP) on two genes [interleukin (IL)-4Ralpha, IL-13] that are associated with allergies. In addition, recent studies suggest that IL-4 and IL-13 induce cyclooxygenase-2 (COX-2) to resolve brain inflammation. To see whether previous Swedish results (110 cases, 430 controls) would be replicated, we estimated the association between glioblastoma and two IL-4Ralpha (rs1805015, rs1801275) and two IL-13 (rs20541, rs1800925) SNPs and their haplotypes and one COX-2 SNP (-765GC) using additional English, Danish, and Finnish data (217 cases, 1,171 controls). Among general population controls, we evaluated associations between these haplotypes, the COX-2 SNP, and self-reported allergies. Our data did not support our original observations relating individual IL-4Ralpha, IL-13, or COX-2 SNPs to glioblastoma risk. However, the T-G IL-4Ralpha haplotype was associated with glioblastoma risk (odds ratio, 2.26; 95% confidence interval, 1.13-4.52) and there was a suggestion of an inverse relation between this haplotype and hayfever prevalence among controls (odds ratio, 0.38; 95% confidence interval, 0.14-1.03). The lack of support for a link between four IL-4Ralpha and IL-13 SNPs and glioblastoma may reflect the absence of associations or may result from uncontrolled confounding by haplotypes related both to those that we examined and glioblastoma. Nonetheless, the association between the T-G IL-4Ralpha haplotype and glioblastoma risk may indicate a role of immune factors in glioblastoma development.

Genetic variation in p53 and ATM haplotypes and risk of glioma and meningioma.

BACKGROUND: P53 and ATM are central checkpoint genes involved in the repair of DNA damage after ionising irradiation, which has been associated with risk of brain tumours. Therefore, we tested the hypothesis that polymorphisms and haplotypes in p53 and ATM could be associated with glioma and meningioma risk. MATERIAL AND METHODS: Six hundred and eighty glioma cases (298 glioblastoma (GBM)), 503 meningioma cases, and 1555 controls recruited in the Nordic-UK Interphone study, were analysed in association with three polymorphisms in p53 (rs2287499, rs1042533, rs1625895) and five polymorphisms in ATM ( rs228599, rs3092992, rs664143, rs170548, rs3092993). Haplotypes were constructed using the HAPLOSTAT program. RESULTS: The global statistical test of glioblastoma and p53 haplotypes was p = 0.02. The haplotype analysis on glioblastoma revealed the 1-2-2 haplotype (promotor-codon72-intron 6) had a frequency of 6.1% in cases compared with 9.8% in controls (p = 0.003). The 1-2-1 haplotype was significantly more frequent in GBM cases, 10.2%, than in controls, 7.3% (p = 0.02). The haplotype analysis in ATM revealed an increased frequency of the 1-1-1-2-1 haplotype in meningioma cases (33.8%) compared with controls (30.3%) (p = 0.03). The 2-1-2-1-1 haplotype had a lower frequency in meningioma cases (36.1%) than controls (40.7%) (p = 0.009). CONCLUSIONS: This study found both positive and negative associations of haplotypes in p53 for glioblastoma and ATM for meningioma. This study provides new data that could add to our understanding of brain tumour susceptibility.