Hasil untuk "Medical emergencies. Critical care. Intensive care. First aid"

Zhe Chen, Yusheng Liao, Shuyang Jiang et al.

Large language models hold promise for addressing medical challenges, such as medical diagnosis reasoning, research knowledge acquisition, clinical decision-making, and consumer health inquiry support. However, they often generate hallucinations due to limited medical knowledge. Incorporating external knowledge is therefore critical, which necessitates multi-source knowledge acquisition. We address this challenge by framing it as a source planning problem, which is to formulate context-appropriate queries tailored to the attributes of diverse sources. Existing approaches either overlook source planning or fail to achieve it effectively due to misalignment between the model's expectation of the sources and their actual content. To bridge this gap, we present MedOmniKB, a repository comprising multigenre and multi-structured medical knowledge sources. Leveraging these sources, we propose the Source Planning Optimisation method, which enhances multi-source utilisation. Our approach involves enabling an expert model to explore and evaluate potential plans while training a smaller model to learn source alignment. Experimental results demonstrate that our method substantially improves multi-source planning performance, enabling the optimised small model to achieve state-of-the-art results in leveraging diverse medical knowledge sources.

Behnam Rezaei Bezanjani, Seyyed Hamid Ghafouri, Reza Gholamrezaei

In recent years, the rapid integration of Internet of Things (IoT) devices into the healthcare sector has brought about revolutionary advancements in patient care and data management. While these technological innovations hold immense promise, they concurrently raise critical security concerns, particularly in safeguarding medical data against potential cyber threats. The sensitive nature of health-related information requires robust measures to ensure the confidentiality, integrity, and availability of patient data in IoT-enabled medical environments. Addressing the imperative need for enhanced security in IoT-based healthcare systems, we propose a comprehensive method encompassing three distinct phases. In the first phase, we implement Blockchain-Enabled Request and Transaction Encryption to strengthen data transaction security, providing an immutable and transparent framework. In the second phase, we introduce a Request Pattern Recognition Check that leverages diverse data sources to identify and block potential unauthorized access attempts. Finally, the third phase incorporates Feature Selection and a BiLSTM network to enhance the accuracy and efficiency of intrusion detection using advanced machine learning techniques. We compared the simulation results of the proposed method with three recent related methods: AIBPSF-IoMT, OMLIDS-PBIoT, and AIMMFIDS. The evaluation criteria include detection rate, false alarm rate, precision, recall, and accuracy - crucial benchmarks for assessing the overall performance of intrusion detection systems. Our findings show that the proposed method outperforms existing approaches across all evaluated criteria, demonstrating its effectiveness in improving the security of IoT-based healthcare systems.

Desheng Hu, Joachim Baumann, Aleksandra Urman et al.

Google Search increasingly surfaces AI-generated content through features like AI Overviews (AIO) and Featured Snippets (FS), which users frequently rely on despite having no control over their presentation. Through a systematic algorithm audit of 1,508 real baby care and pregnancy-related queries, we evaluate the quality and consistency of these information displays. Our robust evaluation framework assesses multiple quality dimensions, including answer consistency, relevance, presence of medical safeguards, source categories, and sentiment alignment. Our results reveal concerning gaps in information consistency, with information in AIO and FS displayed on the same search result page being inconsistent with each other in 33% of cases. Despite high relevance scores, both features critically lack medical safeguards (present in just 11% of AIO and 7% of FS responses). While health and wellness websites dominate source categories for both, AIO and FS, FS also often link to commercial sources. These findings have important implications for public health information access and demonstrate the need for stronger quality controls in AI-mediated health information. Our methodology provides a transferable framework for auditing AI systems across high-stakes domains where information quality directly impacts user well-being.

Joy Ming, Hawi H Tolera, Jiamin Tu et al.

This paper explores opportunities and challenges for data-driven advocacy to support home care workers, an often overlooked group of low-wage, frontline health workers. First, we investigate what data to collect and how to collect it in ways that preserve privacy and avoid burdening workers. Second, we examine how workers and advocates could use collected data to strengthen individual and collective advocacy efforts. Our qualitative study with 11 workers and 15 advocates highlights tensions between workers' desires for individual and immediate benefits and advocates' preferences to prioritize more collective and long-term benefits. We also uncover discrepancies between participants' expectations for how data might transform advocacy and their on-the-ground experiences collecting and using real data. Finally, we discuss future directions for data-driven worker advocacy, including combining different kinds of data to ameliorate challenges, leveraging advocates as data stewards, and accounting for workers' and organizations' heterogeneous goals.

Fred Mutisya, Shikoh Gitau, Christine Syovata et al.

Large Language Models(LLMs) hold promise for improving healthcare access in low-resource settings, but their effectiveness in African primary care remains underexplored. We present a methodology for creating a benchmark dataset and evaluation framework focused on Kenyan Level 2 and 3 clinical care. Our approach uses retrieval augmented generation (RAG) to ground clinical questions in Kenya's national guidelines, ensuring alignment with local standards. These guidelines were digitized, chunked, and indexed for semantic retrieval. Gemini Flash 2.0 Lite was then prompted with guideline excerpts to generate realistic clinical scenarios, multiple-choice questions, and rationale based answers in English and Swahili. Kenyan physicians co-created and refined the dataset, and a blinded expert review process ensured clinical accuracy, clarity, and cultural appropriateness. The resulting Alama Health QA dataset includes thousands of regulator-aligned question answer pairs across common outpatient conditions. Beyond accuracy, we introduce evaluation metrics that test clinical reasoning, safety, and adaptability such as rare case detection (Needle in the Haystack), stepwise logic (Decision Points), and contextual adaptability. Initial results reveal significant performance gaps when LLMs are applied to localized scenarios, consistent with findings that LLM accuracy is lower on African medical content than on US-based benchmarks. This work offers a replicable model for guideline-driven, dynamic benchmarking to support safe AI deployment in African health systems.

Donald Wright, Raphael Sherak, Lonnie Seo et al.

Background: Recent evidence suggests that intra-articular lidocaine (IAL) is an appropriate analgesic alternative to intravenous sedation (IV sedation) during shoulder dislocation reduction, however little is known about patient factors associated with IAL failure and need for subsequent IV sedation. Avoiding crossover is important, as repeated reduction attempts have been previously shown to increase the rate of procedural complications. Objectives: To identify patient level factors associated with crossover from IAL to IV sedation and associated complication rates and operational impacts. Methods: This retrospective observational cohort study evaluated the patient characteristics associated with crossover to unplanned IV sedation after IAL among adult patients undergoing ED reduction of an isolated, acute anterior shoulder dislocation from 2013 to 2021 in an urban, academic hospital system. Univariate analysis and multivariate logistic regression were used. Results: In total, 630 participants were identified who received IAL or procedural sedation. Of these, 182 (28.9 ​%) received IAL of whom 49 (26.9 ​%) subsequently required unplanned IV sedation. Participants had 1.205 (95 ​% CI 1.030, 1.418) fold increase in odds of unplanned sedation for every 10-year increase in age. Crossover to IV sedation was associated with greater rates of adverse events, ED length of stay, and quantity of opioids received compared to either IAL or IV sedation alone. Conclusions: Participants with unplanned IV sedation after IAL had more adverse events than those who received either method alone. Older age was associated with unplanned sedation after IAL. Prospective studies are needed to further define patient factors likely to contribute to failure of IAL.

Bjørn Ove Faldaas, Benjamin Stage Storm, Knut Tore Lappegård et al.

Abstract Background Identifying spontaneous circulation during cardiopulmonary resuscitation (CPR) is challenging. Current methods, which involve intermittent and time-consuming pulse checks, necessitate pauses in chest compressions. This issue is problematic in both in-hospital cardiac arrest and out-of-hospital cardiac arrest situations, where resources for identifying circulation during CPR may be limited. The fraction of chest compression plays a pivotal role in improving survival rates. To address this challenge, we evaluated a newly developed hands-free, continuous carotid Doppler system (RescueDoppler), designed to identify spontaneous circulation during chest compressions. In our study, we utilized a porcine model of cardiac arrest to investigate sequences of ventricular fibrillation, followed by defibrillation, and monitoring for the return of spontaneous circulation during chest compressions with the carotid Doppler system. We explored both manual compressions at 100 and 50 compressions per minute and mechanical compressions. To estimate the detection rate (i.e., sensitivity), we employed a logistic mixed model with animal identity as random effect. Results Offline analysis of Doppler color M-mode and spectral display successfully identified spontaneous circulation during chest compressions in all compression models. Spontaneous circulation was detected in 51 of 59 sequences, yielding an expected sensitivity of 98% with a 95% confidence interval of 59% to 99%. Conclusion The RescueDoppler, a continuous hands-free carotid Doppler system, demonstrates an expected sensitivity of 98% for identifying spontaneous circulation during both manual and mechanical chest compressions. Clinical studies are needed to further validate these findings.

Milad Naeimaei Aali, Felix Mannhardt, Pieter Jelle Toussaint

One of the challenges in healthcare processes, especially those related to multi-morbid patients who suffer from multiple disorders simultaneously, is not connecting the disorders in patients to process events and not linking events' activities to globally accepted terminology. Addressing this challenge introduces a new entity to the clinical process. On the other hand, it facilitates that the process is interpretable and analyzable across different healthcare systems. This paper aims to introduce a tool named CEKG that uses event logs, diagnosis data, ICD-10, SNOMED-CT, and mapping functions to satisfy these challenges by constructing event graphs for multi-morbid patients' care pathways automatically.

Langrui Zhou, Guang Li

The current mainstream multi-modal medical image-to-image translation methods face a contradiction. Supervised methods with outstanding performance rely on pixel-wise aligned training data to constrain the model optimization. However, obtaining pixel-wise aligned multi-modal medical image datasets is challenging. Unsupervised methods can be trained without paired data, but their reliability cannot be guaranteed. At present, there is no ideal multi-modal medical image-to-image translation method that can generate reliable translation results without the need for pixel-wise aligned data. This work aims to develop a novel medical image-to-image translation model that is independent of pixel-wise aligned data (MITIA), enabling reliable multi-modal medical image-to-image translation under the condition of misaligned training data. The proposed MITIA model utilizes a prior extraction network composed of a multi-modal medical image registration module and a multi-modal misalignment error detection module to extract pixel-level prior information from training data with misalignment errors to the largest extent. The extracted prior information is then used to construct a regularization term to constrain the optimization of the unsupervised cycle-consistent GAN model, restricting its solution space and thereby improving the performance and reliability of the generator. We trained the MITIA model using six datasets containing different misalignment errors and two well-aligned datasets. Subsequently, we compared the proposed method with six other state-of-the-art image-to-image translation methods. The results of both quantitative analysis and qualitative visual inspection indicate that MITIA achieves superior performance compared to the competing state-of-the-art methods, both on misaligned data and aligned data.

Mingjian Li, Mingyuan Meng, Michael Fulham et al.

Medical image representations can be learned through medical vision-language contrastive learning (mVLCL) where medical imaging reports are used as weak supervision through image-text alignment. These learned image representations can be transferred to and benefit various downstream medical vision tasks such as disease classification and segmentation. Recent mVLCL methods attempt to align image sub-regions and the report keywords as local-matchings. However, these methods aggregate all local-matchings via simple pooling operations while ignoring the inherent relations between them. These methods therefore fail to reason between local-matchings that are semantically related, e.g., local-matchings that correspond to the disease word and the location word (semantic-relations), and also fail to differentiate such clinically important local-matchings from others that correspond to less meaningful words, e.g., conjunction words (importance-relations). Hence, we propose a mVLCL method that models the inter-matching relations between local-matchings via a relation-enhanced contrastive learning framework (RECLF). In RECLF, we introduce a semantic-relation reasoning module (SRM) and an importance-relation reasoning module (IRM) to enable more fine-grained report supervision for image representation learning. We evaluated our method using six public benchmark datasets on four downstream tasks, including segmentation, zero-shot classification, linear classification, and cross-modal retrieval. Our results demonstrated the superiority of our RECLF over the state-of-the-art mVLCL methods with consistent improvements across single-modal and cross-modal tasks. These results suggest that our RECLF, by modelling the inter-matching relations, can learn improved medical image representations with better generalization capabilities.

Mehran Sotoodehnia, Maryam Khodayar, Alireza Jalali et al.

Abstract Introduction Difficult laryngoscopy and intubation are serious problems among critically ill patients in emergency department (ED) so utility of a rapid, accurate and noninvasive method for predicting of these patients are necessary. Ultrasonography has been recently used in this regard and this study was conducted to investigate the correlation of some introduced upper airway ultrasound parameters with difficult laryngoscopy / difficult intubation in patients referred to the ED. Method In this prospective observational study all patients ≥ 18-year-old who had an indication for rapid sequence intubation (RSI) were included. Ultrasound parameters including Hyoid Bone Visibility (HBV), Distance from Skin to Hyoid Bone (DSHB), Distance from Skin to Vocal Cords (DSVC), Distance from Skin to Thyroid Isthmus (DSTI), and Distance between Arytenoids Cartilages (DBAC) were measured in all cases. The patients underwent RSI and thereafter the patients’ baseline characteristics, Cormack-Lehane grade, number of attempted laryngoscopy were recorded in a pre-prepared check list and compared with measured ultrasound parameters. The “difficult laryngoscopy” was defined as Cormack-Lehane classification grades III/IV; and need for more than 3 intubation attempts was considered as “difficult intubation”. Results One hundred and twenty-three patients (52% male) were included of whom 10 patients (8.1%) were categorized as difficult laryngoscopy cases; and just 4 (3.3%) cases underwent more than 3 laryngoscopy attempts who considered as difficult intubation cases. The mean age of the patients in non-difficult and difficult intubation groups were 69.2 ± 15.16 and 68.77 ± 17.37 years, respectively (p > 0.05). There was no significant relationship between difficult laryngoscopy and HBV (p = 0.381) but has significant correlation with difficult intubation (p = 0.004). The DSHB had a significant correlation with difficult laryngoscopy (p = 0.002) but its correlation with difficult intubation was not significant (p = 0.629). The DSVC and DSTI had a significant relationship with both difficult laryngoscopy (p = 0.003 and p = 0.001), and difficult intubation (p = 0.025 and p = 0.001). The DBAC had not significant correlation neither with the difficult laryngoscopy (p = 0.142), nor with difficult intubation (p = 0.526). Conclusion The findings showed that ultrasound parameters including soft tissue DSHB, DSVC and DSTI could be proper predictors of difficult laryngoscopy. Also, HBV, DSVC and DSTI may be proper predictors for difficult intubation. But DBAC was not useful in this regard.

Allison Clark, Peyton Lampley, Vu Huy Tran

Introduction: Tension hydrothorax is an uncommon emergent condition in which hemodynamic instability and respiratory compromise may occur. Emergency physicians may diagnose tension hydrothorax by point-of-care ultrasound. Case Series: We discuss the key sonographic features assisting in identification. Four patients with history of malignancy who were found to have tension hydrothorax exhibited the following common ultrasound findings: massive, left-sided pleural effusion; complete, compressive atelectasis; and shift of cardiac structures into the right hemithorax, resulting in right-sided probe placement to obtain cardiac views. Conclusion: This is the first instance to our knowledge of point-of-care ultrasound findings in tension hydrothorax to be described in the literature.

Daniel Macis, Sara Perilli, Cristina Gena

Recently, it has been considering robotics to face world population aging. According to the WHO, in 2050 there will be about 2.1 billion people over 60 years old worldwide causing a persistent growing need of assistance and a shortage of manpower for delivering congruous assistance. Therefore, seniors' QoL is continuously threatened. Socially Assistive Robotics proposes itself as a solution. To improve SARs acceptability, it is necessary to tailor the system's characteristics with respect to the target needs and issues through the analysis of previous and current studies in the HRI field. Through the examination of the state of the art of social robotics in elderly care, past case studies and paper research about SARs' efficiency, it has been proposed two potential solution examples for two different scenarios, applying two different SARs: Pepper and Nao robots.

Saikat Roy, Gregor Koehler, Michael Baumgartner et al.

Owing to success in the data-rich domain of natural images, Transformers have recently become popular in medical image segmentation. However, the pairing of Transformers with convolutional blocks in varying architectural permutations leaves their relative effectiveness to open interpretation. We introduce Transformer Ablations that replace the Transformer blocks with plain linear operators to quantify this effectiveness. With experiments on 8 models on 2 medical image segmentation tasks, we explore -- 1) the replaceable nature of Transformer-learnt representations, 2) Transformer capacity alone cannot prevent representational replaceability and works in tandem with effective design, 3) The mere existence of explicit feature hierarchies in transformer blocks is more beneficial than accompanying self-attention modules, 4) Major spatial downsampling before Transformer modules should be used with caution.

Evi M. C. Huijben, Sina Amirrajab, Josien P. W. Pluim

Out-of-distribution (OOD) detection is crucial for the safety and reliability of artificial intelligence algorithms, especially in the medical domain. In the context of the Medical OOD (MOOD) detection challenge 2023, we propose a pipeline that combines a histogram-based method and a diffusion-based method. The histogram-based method is designed to accurately detect homogeneous anomalies in the toy examples of the challenge, such as blobs with constant intensity values. The diffusion-based method is based on one of the latest methods for unsupervised anomaly detection, called DDPM-OOD. We explore this method and propose extensive post-processing steps for pixel-level and sample-level anomaly detection on brain MRI and abdominal CT data provided by the challenge. Our results show that the proposed DDPM method is sensitive to blur and bias field samples, but faces challenges with anatomical deformation, black slice, and swapped patches. These findings suggest that further research is needed to improve the performance of DDPM for OOD detection in medical images.

Seyed Morsal Mosallami Aghili, Mehran Khoshfetrat, Ali Asgari et al.

Introduction: Evidence showed that cardiac complications may occur in coronavirus disease-19 (COVID-19) during the acute and post-infection phases. This study aimed to evaluate the association between the echocardiographic characteristics and in-hospital mortality of COVID-19 patients as well as the changes after one-month follow-up. Method: All adult (≥18 years old) hospitalized COVID-19 patients in need of echocardiography based on the guideline of the Iranian Society of Echocardiography for performing various types of echocardiography during the COVID-19 pandemic were included in this study. An expert cardiologist performed the echocardiography on all patients and also on all available patients one month after discharge. Results: 146 hospitalized cases of COVID-19 and 81 cases available for 1-month follow-up echocardiography were studied in this prospective study. Left ventricle wall hypokinesia, aorta valve stenosis, dilated Inferior Vena Cava (IVC), and Pulmonary Artery Systolic Pressure (PASP) of more than 35 were associated with 3.59 (95% CI: 1.19-10.79, p = 0.02), ‎11 (95% CI: 3.3 – 36.63, p = 0.001), ‎5.58 (95% CI: 1.04-29.41, p = 0.041)‎, and 2.91 (95% CI: 1.35 – 6.3, p = 0.001) times higher odds of mortality than healthy subjects. In 1-month follow-up of patients, deterioration in LVEF (p = 0.03) was detected in the not-fully vaccinated patients, and a significant decrease in PASP was observed in all cases (p = 0.04); but these changes were not clinically important. Conclusion: Left ventricle wall hypokinesia, aorta valve stenosis, dilated IVC, and PASP ≥ 35 were predictors of in-hospital mortality in our study. There were not any potential clinically significant differences in one-month echocardiographic follow-ups of the studied patients.

Thijs Kooi

The idea of using computers to read medical scans was introduced as early as 1966. However, limits to machine learning technology meant progress was slow initially. The Alexnet breakthrough in 2012 sparked new interest in the topic, which resulted in the release of 100s of medical AI solutions on the market. In spite of success for some diseases and modalities, many challenges remain. Research typically focuses on the development of specific applications or techniques, clinical evaluation, or meta analysis of clinical studies or techniques through surveys or challenges. However, limited attention has been given to the development process of improving real world performance. In this tutorial, we address the latter and discuss some techniques to conduct the development process in order to make this as efficient as possible.

Adam W. Gaffney

In March 2020, as coronavirus disease (COVID-19) began itsfirst deadly surge in the United States, policy-makers grasped that the nation’s wide gaps in health coverage would undercut our pandemic response. Containment of outbreaks requires testing and isolation, efforts that will be derailed if those with symptoms avoid care because of costs. Hence, a provision of the March 2020 Families First Coronavirus Response Act eliminated out-of-pocket costs for much severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) testing. However, the cost of treatment dwarfs that of testing. Hospitalization for COVID-19 pneumonia might impose costs of thousands or tensof thousandsofdollars, ormore, on the uninsuredorunderinsured.At thepandemic’s outset, an estimated 18 million Americans at increased risk of severe COVID-19 due to advanced age or comorbidities were inadequately insured (1). In April 2020, about 1 in 10 Americans said they would avoid seeking care because of costs if they believed they had COVID-19 (2). To address such concerns, the Coronavirus Aid, Relief, and Economic Security Act, passed that month, provided funds to coverCOVID-19 treatment for the uninsured. Meanwhile, some private insurers waived copays and deductibles for such care. Finally, the federal government fully subsidized the cost of COVID-19 vaccination for all. It has even warned providers not to charge out-of-pocket fees for any services related to vaccine administration. Such measures fell short of full protection.For instance, the complexityof our billing and coding system left some patients with inappropriatebills forcoronavirus testing (3). Someuninsuredpatientshospitalizedwith COVID-19 fell through the regulatory cracks andwere sent large bills (4).Aportionof those avoiding vaccination cited fear of costs as the reason, which was likely due to previous adverse experiences (5). Private insurers, meanwhile, have ended the copay/deductible waivers. With more than 600,000 dead from COVID-19, the United States’ overall response to the outbreak must be judged a failure. Still, it is probable that these COVID-19–specific coverage expansion measures made an important difference, even if they were insufficient. Although we lack data on their effects on outcomes, many previous studies have found that cost barriers deter all types of care, even for emergencies (6). Without an expansion of coverage for COVID-19–related care, it is probable thatmanymoremayhave avoided testing and potentially infected others, faced ruinous hospital bills during a period of widespread job loss, or even succumbed to COVID-19 because of delays in seeking medical attention or vaccination. Yet that raises an important, if uncomfortable, question. If such cost-related adverse outcomes seem unacceptable, indeed abhorrent, for patients with COVID-19, why are they considered acceptable for ourpatients with other lung diseases? A growing body of evidence has shed light on the inadequacy of coverage for patientswithrespiratoryillness. In2018,24.2% of those with asthma and 28.2% of those with chronic obstructive pulmonary disease (COPD) were inadequately insured (i.e., uninsured, or insured but unable to afford the costsofmedicalcareorprescriptiondrugs)(7). Stark racial and economic disparities in access to care magnify the greater burden of chronic respiratory disease experienced by disadvantaged populations (8), a pattern also seen with COVID-19 (1, 9). Inadequate access to care can lead to worse outcomes for such patients. Higher medication cost-sharing is associated with a greater likelihood of hospitalizations among older children with asthma (10). A province-wide increase in prescription drug cost-sharing in British Columbia led to more COPD and asthma admissions (11). A cross-sectional analysis of national survey data by my colleagues and I found greater foregone care and cost-related nonadherence to prescription drugs—and more hospitalizations—among individuals withCOPDandhigh deductibles relative to those with no or low deductibles (12). And for the critically ill, intensive care unit (ICU) treatment can translate into critically high medical bills. Estimated out-ofpocket costs over the last year of life for patients who use the ICU shortly before death amount to $26,993 for the uninsured and $10,022 for the privately insured (13). Somemay argue that enhanced coverage specifically for COVID-19-related care—but not other conditions—is justified given the contagious nature of the illness. When a

Antoni Torres, Anna Motos, Jordi Riera et al.

Camilla Metelmann, Bibiana Metelmann, Louisa Schuffert et al.

Abstract Background Bystander-initiated resuscitation is essential for surviving out-of-hospital cardiac arrest. Smartphone apps can provide real-time guidance for medical laypersons in these situations. Are these apps a beneficial addition to traditional resuscitation training? Methods In this controlled trial, we assessed the impact of app use on the quality of resuscitation (hands-off time, assessment of the patient’s condition, quality of chest compression, body and arm positioning). Pupils who have previously undergone a standardised resuscitation training, encountered a simulated cardiac arrest either (i) without an app (control group); (ii) with facultative app usage; or (iii) with mandatory app usage. Measurements were compared using generalised linear regression. Results 200 pupils attended this study with 74 pupils in control group, 65 in facultative group and 61 in mandatory group. Participants who had to use the app significantly delayed the check for breathing, call for help, and first compression, leading to longer total hands-off time. Hands-off time during chest compression did not differ significantly. The percentage of correct compression rate and correct compression depth was significantly higher when app use was mandatory. Assessment of the patient’s condition, and body and arm positioning did not differ. Conclusions Smartphone apps offering real-time guidance in resuscitation can improve the quality of chest compression but may also delay the start of resuscitation. Provided that the app gives easy-to-implement, guideline-compliant instructions and that the user is familiar with its operation, we recommend smartphone-guidance as an additional tool to hands-on CPR-training to increase the prevalence and quality of bystander-initiated CPR.