RESUMO
Chest X-rays (CXRs) play a pivotal role in cost-effective clinical assessment of various heart and lung related conditions. The urgency of COVID-19 diagnosis prompted their use in identifying conditions like lung opacity, pneumonia, and acute respiratory distress syndrome in pediatric patients. We propose an AI-driven solution for binary COVID-19 versus non-COVID-19 classification in pediatric CXRs. We present a Federated Self-Supervised Learning (FSSL) framework to enhance Vision Transformer (ViT) performance for COVID-19 detection in pediatric CXRs. ViT's prowess in vision-related binary classification tasks, combined with self-supervised pre-training on adult CXR data, forms the basis of the FSSL approach. We implement our strategy on the Rhino Health Federated Computing Platform (FCP), which ensures privacy and scalability for distributed data. The chest X-ray analysis using the federated SSL (CAFES) model, utilizes the FSSL-pre-trained ViT weights and demonstrated gains in accurately detecting COVID-19 when compared with a fully supervised model. Our FSSL-pre-trained ViT showed an area under the precision-recall curve (AUPR) of 0.952, which is 0.231 points higher than the fully supervised model for COVID-19 diagnosis using pediatric data. Our contributions include leveraging vision transformers for effective COVID-19 diagnosis from pediatric CXRs, employing distributed federated learning-based self-supervised pre-training on adult data, and improving pediatric COVID-19 diagnosis performance. This privacy-conscious approach aligns with HIPAA guidelines, paving the way for broader medical imaging applications.
RESUMO
Objectives: Procedural interoperability in health care requires information support and monitoring of a common work practice. Our aim was to devise an information model for a complete annotation of actions in clinical pathways that allow use of multiple plans concomitantly as several partial processes underlie any composite clinical process. Materials and Methods: The development of the information model was based on the integration of a defined protocol for clinical interoperability in the care of patients with chronic obstructive pulmonary disease and an observational study protocol for cohort characterization at the group level. In the clinical process patient reported outcome measures were included. Results: The clinical protocol and the observation study protocol were developed on the clinical level and a single plan definition was developed by merging of the protocols. The information model and a common data model that had been developed for care pathways was successfully implemented and data for the medical records and the observational study could be extracted independently. The interprofessional process support improved the communication between the stakeholders (health care professionals, clinical scientists and providers). Discussion: We successfully merged the processes and had a functionally successful pilot demonstrating a seamless appearance for the health care professionals, while at the same time it was possible to generate data that could serve quality registries and clinical research. The adopted data model was initially tested and hereby published to the public domain. Conclusion: The use of a patient centered information model and data annotation focused on the care pathway simplifies the annotation of data for different purposes and supports sharing of knowledge along the patient care path.
RESUMO
BACKGROUND AND PURPOSE: Intravenous thrombolysis is a well-established treatment for acute ischemic stroke. Our aim was to quantify the effect of each minute delay in door-to-needle time (DNT) on 90-day survival, intracerebral hemorrhagic complication <36 hours, and functional outcomes at 3 months, in routine clinical practice. METHODS: Our nationwide registry-based study included 14 132 adult patient admissions with ischemic stroke receiving intravenous thrombolysis from 2010 to 2017. Outcomes were analyzed using multivariable logistic regression, adjusting for potential confounders. RESULTS: Median DNT was 47 minutes, with an improvement from 65 to 38 minutes during the study. Median age was 74 years, and median National Institutes of Health Stroke Scale 8 points. We found a significant impact of each minute delay in DNT with reduced odds of survival by 0.6%, increased odds of intracerebral hemorrhagic and worse activities of daily living by 0.3%, and worse living conditions and mobility by 0.4%. CONCLUSIONS: Improving DNT is a key factor in achieving good outcomes after stroke. We estimate that in Sweden alone in 2017, compared with 2010, the shorter DNT achieved have saved 38 lives, avoided 8 intracerebral hemorrhagic transformations, and spared, respectively, 36, 51, and 52 patients from a worsening in activities of daily living, living conditions, and mobility. DNT is sensitive for interventions and should be targeted in quality improvement efforts.
Assuntos
Hospitalização/tendências , Acidente Vascular Cerebral/epidemiologia , Acidente Vascular Cerebral/terapia , Terapia Trombolítica/tendências , Tempo para o Tratamento/tendências , Idoso , Idoso de 80 Anos ou mais , Estudos de Coortes , Feminino , Humanos , Masculino , Estudos Retrospectivos , Acidente Vascular Cerebral/diagnóstico , Suécia/epidemiologia , Terapia Trombolítica/métodosRESUMO
Objectives: To update the sets of patient-centric outcomes measures ("standard-sets") developed by the not-for-profit organization ICHOM to become more readily applicable in patients with multimorbidity and to facilitate their implementation in health information systems. To that end we set out to (i) harmonize measures previously defined separately for different conditions, (ii) create clinical information models from the measures, and (iii) restructure the annotation to make the sets machine-readable. Materials and Methods: First, we harmonized the semantic meaning of individual measures across all the 28 standard-sets published to date, in a harmonized measure repository. Second, measures corresponding to four conditions (Breast cancer, Cataracts, Inflammatory bowel disease and Heart failure) were expressed as logical models and mapped to reference terminologies in a pilot study. Results: The harmonization of semantic meaning resulted in a consolidation of measures used across the standard-sets by 15%, from 3,178 to 2,712. These were all converted into a machine-readable format. 61% of the measures in the 4 pilot sets were bound to existing concepts in either SNOMED CT or LOINC. Discussion: The harmonization of ICHOM measures across conditions is expected to increase the applicability of ICHOM standard-sets to multi-morbid patients, as well as facilitate their implementation in health information systems. Conclusion: Harmonizing the ICHOM measures and making them machine-readable is expected to expedite the global adoption of systematic and interoperable outcomes measurement. In turn, we hope that the improved transparency on health outcomes that follows will let health systems across the globe learn from each other to the ultimate benefit of patients.
RESUMO
Background and Purpose: Studies of monthly variation in acute stroke care have led to conflicting results. Our objective was to study monthly variation and longitudinal trends in quality of care and patient survival following acute stroke. Methods: Our nationwide study included all adult patients (≥18 years) with acute stroke (ischemic or hemorrhagic), admitted to Swedish hospitals from 2011 to 2016, and that were registered in The Swedish Stroke Register (Riksstroke). We studied how month of admission and longitudinal trends affected acute stroke care and survival. We also studied resilience to this variation among hospitals with different levels of specialization. Results: We included 132,744 stroke admissions. The 90-day survival was highest in May and lowest in January (84.1 vs. 81.5%). Thrombolysis rates and door-to-needle time within 30 min increased from 2011 to 2016 (respectively, 7.3 vs. 12.8% and 7.7 vs. 28.7%). Admission to a stroke unit as first destination of hospital care was lowest in January and highest in June (78.3 vs. 80.5%). Stroke unit admission rates decreased in university hospitals from 2011 to 2016 (83.4 vs. 73.9%), while no such trend were observed in less specialized hospitals. All the differences above remained significant (p < 0.05) after adjustment for possible confounding factors. Conclusion: We found that month of admission and longitudinal trends both affect quality of care and survival of stroke patients in Sweden, and that the effects differ between hospital types. The observed variation suggests an opportunity to improve stroke care in Sweden. Future studies ought to focus on identifying the specific factors driving this variation, for subsequent targeting by quality improvement efforts.
RESUMO
OBJECTIVES: The aim of this work was to train machine learning models to identify patients at end of life with clinically meaningful diagnostic accuracy, using 30-day mortality in patients discharged from the emergency department (ED) as a proxy. DESIGN: Retrospective, population-based registry study. SETTING: Swedish health services. PRIMARY AND SECONDARY OUTCOME MEASURES: All cause 30-day mortality. METHODS: Electronic health records (EHRs) and administrative data were used to train six supervised machine learning models to predict all-cause mortality within 30 days in patients discharged from EDs in southern Sweden, Europe. PARTICIPANTS: The models were trained using 65 776 ED visits and validated on 55 164 visits from a separate ED to which the models were not exposed during training. RESULTS: The outcome occurred in 136 visits (0.21%) in the development set and in 83 visits (0.15%) in the validation set. The model with highest discrimination attained ROC-AUC 0.95 (95% CI 0.93 to 0.96), with sensitivity 0.87 (95% CI 0.80 to 0.93) and specificity 0.86 (0.86 to 0.86) on the validation set. CONCLUSIONS: Multiple models displayed excellent discrimination on the validation set and outperformed available indexes for short-term mortality prediction interms of ROC-AUC (by indirect comparison). The practical utility of the models increases as the data they were trained on did not require costly de novo collection but were real-world data generated as a by-product of routine care delivery.
Assuntos
Registros Eletrônicos de Saúde , Serviço Hospitalar de Emergência/estatística & dados numéricos , Aprendizado de Máquina , Mortalidade , Alta do Paciente/estatística & dados numéricos , Adulto , Idoso , Idoso de 80 Anos ou mais , Feminino , Humanos , Modelos Logísticos , Masculino , Pessoa de Meia-Idade , Sistema de Registros , Estudos Retrospectivos , Suécia/epidemiologia , Fatores de Tempo , Adulto JovemRESUMO
OBJECTIVES: A recent study of acute stroke patients in England and Wales revealed several patterns of temporal variation in quality of care. We hypothesized that similar patterns would be present in Sweden and aimed to describe these patterns. Additionally, we aimed to investigate whether hospital type conferred resilience against temporal variation. MATERIALS AND METHODS: We conducted this nationwide registry-based study using data from the Swedish Stroke Register (Riksstroke) including all adult patients registered with acute stroke between 2011 and 2015. Outcomes included process measures and survival. We modeled time of presentation as on/off-hours, shifts, day of week, 4-hour, and 12-hour time blocks. We studied hospital resilience by comparing outcomes across hospital types. RESULTS: A total of 113 862 stroke events in 72 hospitals were included. The process indicators and survival all showed significant temporal variation. Door-to-needle (DTN) time within 30 minutes was less likely during nighttime than daytime (OR 0.50; 95% CI 0.41-0.60). Patients admitted during off-hours had lower odds of direct stroke unit (SU) admission (OR 0.72; 95% CI 0.70-0.75). 30-day survival was lower in nighttime vs daytime presentations (OR 0.90, 95% CI 0.84-0.96). The effects of temporal variation differed significantly between hospital types for DTN time within 30 minutes and direct SU admission where university hospitals were more resilient than specialized non-university hospitals. CONCLUSIONS: Our study shows that variation in quality of care and survival is present throughout the whole week. We also found that university hospitals were more resilient to temporal variation than specialized non-university hospitals.
Assuntos
Fotoperíodo , Qualidade da Assistência à Saúde/estatística & dados numéricos , Sistema de Registros/estatística & dados numéricos , Acidente Vascular Cerebral/terapia , Adulto , Idoso , Feminino , Hospitalização/estatística & dados numéricos , Humanos , Masculino , Pessoa de Meia-Idade , Qualidade da Assistência à Saúde/normas , Acidente Vascular Cerebral/epidemiologia , SuéciaRESUMO
INTRODUCTION: It is well established that managing patients with acute stroke in dedicated stroke units is associated with improved functioning and survival. The objectives of this study are to investigate whether patients with acute stroke are less likely to be directly admitted to a stroke unit from the Emergency Department when hospital beds are scarce and to measure variation across hospitals in terms of this outcome. PATIENTS AND METHODS: This register study comprised data on patients with acute stroke admitted to 14 out of 72 Swedish hospitals in 2011-2014. Data from the Swedish stroke register were linked to administrative daily data on hospital bed occupancy (measured at 6 a.m.). Logistic regression analysis was used to analyse the association between bed occupancy and direct stroke unit admission. RESULTS: A total of 13,955 hospital admissions were included; 79.6% were directly admitted to a stroke unit from the Emergency Department. Each percentage increase in hospital bed occupancy was associated with a 1.5% decrease in odds of direct admission to a stroke unit (odds ratio = 0.985, 95% confidence interval = 0.978-0.992). The best-performing hospital exhibited an odds ratio of 3.8 (95% confidence interval = 2.6-5.5) for direct admission to a stroke unit versus the reference hospital. DISCUSSION AND CONCLUSION: We found an association between hospital crowding and reduced quality of care in acute stroke, portrayed by a lower likelihood of patients being directly admitted to a stroke unit from the Emergency Department. The magnitude of the effect varied considerably across hospitals.
RESUMO
BACKGROUND: Emergency department (ED) overcrowding is frequently described in terms of input- throughput and output. In order to reduce ED input, a concept called primary triage has been introduced in several Swedish EDs. In short, primary triage means that a nurse separately evaluates patients who present in the Emergency Department (ED) and either refers them to primary care or discharges them home, if their complaints are perceived as being of low acuity. The aim of the present study is to elucidate whether high levels of in-hospital bed occupancy are associated with decreased permeability in primary triage. The appropriateness of discharges from primary triage is assessed by 72-h revisits to the ED. METHODS: The study is a retrospective cohort study on administrative data from the ED at a 420-bed hospital in southern Sweden from 2011-2012. In addition to crude comparisons of proportions experiencing each outcome across strata of in-hospital bed occupancy, multivariate models are constructed in order to adjust for age, sex and other factors. RESULTS: A total of 37,129 visits to primary triage were included in the study. 53.4 % of these were admitted to the ED. Among the cases referred to another level of care, 8.8 % made an unplanned revisit to the ED within 72 h. The permeability of primary triage was not decreased at higher levels of in-hospital bed occupancy. Rather, the permeability was slightly higher at occupancy of 100-105 % compared to <95 % (OR 1.09 95 % CI 1.02-1.16). No significant association between in-hospital bed occupancy and the probability of 72-h revisits was observed. CONCLUSIONS: The absence of a decreased permeability of primary triage at times of high in-hospital bed occupancy is reassuring, as the opposite would have implied that patients might be denied entry not only to the hospital, but also to the ED, when in-hospital beds are scarce.
Assuntos
Ocupação de Leitos , Serviço Hospitalar de Emergência/organização & administração , Padrões de Prática em Enfermagem/estatística & dados numéricos , Triagem/métodos , Adolescente , Adulto , Idoso , Criança , Pré-Escolar , Serviço Hospitalar de Emergência/estatística & dados numéricos , Feminino , Humanos , Lactente , Masculino , Pessoa de Meia-Idade , Admissão do Paciente/estatística & dados numéricos , Alta do Paciente/estatística & dados numéricos , Estudos Retrospectivos , Suécia , Triagem/organização & administração , Triagem/estatística & dados numéricos , Adulto JovemRESUMO
Ageing populations and higher ambitions continuously drive healthcare costs in Sweden and worldwide. During the last two decades, downsizing hospital bed capacity has been the strategy for cutting expenditure in the Swedish healthcare system. However, the lack of implementation of new and viable outpatient alternatives has led to a widespread overcrowding problem in Swedish hospitals and emergency departments. The present study was conducted as a survey in hospital wards at two emergency hospitals in southwestern Sweden. Study aims were to assess the causes of hospitalization and indications for continuing in-hospital care in hospitalized geriatric patients (>80 years). The study shows that a very small number of patients are admitted barely because of social factors; however, there is a significant group where hospitalization is due to both social and medical factors. A large group of hospitalized patients over 80 year (37%) could receive their care outside the emergency hospital. About 30% of hospitalized patients are waiting for planning, and the majority of them waiting for social action and planning. Older patients with multiple diseases require healthcare but not hospitalization to the present extent. We should focus on developing additional forms of healthcare since avoidable hospitalization is a high cost for the society, but above all a risk for the individual.
Assuntos
Serviço Hospitalar de Emergência/estatística & dados numéricos , Hospitalização/estatística & dados numéricos , Idoso de 80 Anos ou mais , Humanos , Uso Excessivo dos Serviços de Saúde , Prontuários Médicos , Pesquisa Qualitativa , Inquéritos e Questionários , SuéciaRESUMO
BACKGROUND: Previous work has suggested that given a hospital's need to admit more patients from the emergency department (ED), high inpatient bed occupancy may encourage premature hospital discharges that favor the hospital's need for beds over patients' medical interests. We argue that the effects of such action would be measurable as a greater proportion of unplanned hospital readmissions among patients discharged when the hospital was full than when not. In response, the present study tested this hypothesis by investigating the association between inpatient bed occupancy at the time of hospital discharge and the 30-day readmission rate. METHODS: The sample included all inpatient admissions from the ED at a 420-bed emergency hospital in southern Sweden during 2011-2012 that resulted in discharge before 1 December 2012. The share of unplanned readmissions within 30 days was computed for levels of inpatient bed occupancy of <95%, 95-100%, 100-105% and >105% at the hour of discharge. A binary logistic regression model was constructed to adjust for age, time of discharge, and other factors that could affect the outcome. RESULTS: In all, 32,811 visits were included in the study, 9.9% of which resulted in an unplanned readmission within 30 days of discharge. The proportion of readmissions was 9.0% for occupancy levels of <95% at the patient's discharge, 10.2% for 95-100% occupancy, 10.8% for 100-105% occupancy, and 10.5% for >105% occupancy (p = 0.0001). Results from the multivariate models show that the OR (95% CI) of readmission was 1.11 (1.01-1.22) for patients discharged at 95-100% occupancy, 1.17 (1.06-1.29) at 100-105% occupancy, and 1.15 (0.99-1.34) at >105% occupancy. CONCLUSIONS: Results indicate that patients discharged from inpatient wards at times of high inpatient bed occupancy experience an increased risk of unplanned readmission within 30 days of discharge.
Assuntos
Ocupação de Leitos/estatística & dados numéricos , Alta do Paciente/estatística & dados numéricos , Readmissão do Paciente/estatística & dados numéricos , Serviço Hospitalar de Emergência/estatística & dados numéricos , Feminino , Humanos , Tempo de Internação/estatística & dados numéricos , Masculino , Estudos Retrospectivos , Fatores de Risco , Suécia , Fatores de TempoRESUMO
BACKGROUND: A possible downstream effect of high in-hospital bed occupancy is that patients in the emergency department (ED) who would benefit from in-hospital care are denied admission. The present study aimed at evaluating this hypothesis through investigating associations between in-hospital bed occupancy at the time of presentation in the ED and the probability for unplanned 72-hour (72-h) revisits to the ED among patients discharged at index. A second outcome was unplanned 72-h revisits resulting in admission. METHODS: All visits to the ED of a 420-bed emergency hospital in southern Sweden between 1 January 2011 and 31 December 2012, which did not result in admission, death, or transfer to another hospital were included. Revisiting fractions were computed for in-hospital occupancy intervals <85%, 85% to 90%, 90% to 95%, 95% to 100%, 100% to 105%, and ≥105%. Multivariate models were constructed in an attempt to take confounding factors from, e.g., presenting complaints, age, referral status, and triage priority into account. RESULTS: Included in the study are 81,878 visits. The fraction of unplanned 72-h revisits/unplanned 72-h revisits resulting in admission was 5.8%/1.4% overall, 6.2%/1.4% for occupancy <85%, 6.4%/1.5% for occupancy 85% to 90%, 5.8%/1.4% for occupancy 90% to 95%, 6.0%/1.6% for occupancy 95% to 100%, 5.4%/1.6% for occupancy 100% to 105%, and 4.9%/1.4% for occupancy ≥105%. In the multivariate models, a trend to lower probability of unplanned 72-h revisits was observed at occupancy ≥105% compared to occupancy <95% (OR 0.88, CI 0.76 to 1.01). No significant associations between in-hospital occupancy at index and the probability of making unplanned 72-h revisits resulting in admission were observed. CONCLUSIONS: The lack of associations between in-hospital occupancy and unplanned 72-h revisits does not support the hypothesis that ED patients are inappropriately discharged when in-hospital beds are scarce. The results are reassuring as they indicate that physicians are able to make good decisions, also while resources are constrained.
RESUMO
BACKGROUND: The association between emergency department (ED) overcrowding and poor patient outcomes is well described, with recent work suggesting that the phenomenon causes delays in time-sensitive interventions, such as resuscitation. Even though most researchers agree on the fact that admitted patients boarding in the ED is a major contributing factor to ED overcrowding, little work explicitly addresses whether in-hospital occupancy is associated to the probability of patients being admitted from the ED. The objective of the present study is to investigate whether such an association exists. METHODS: Retrospective analysis of data on all ED visits to Helsingborg General Hospital in southern Sweden between January 1, 2011, and December 31, 2012, was undertaken. The fraction of admitted patients was calculated separately for strata of in-hospital occupancy <95%, 95-100%, 100-105%, and >105%. Multivariate models were constructed in an attempt to take confounding factors, e.g., presenting complaints, age, referral status, triage priority, and sex into account. Subgroup analysis was performed for each specialty unit within the ED. RESULTS: Overall, 118,668 visits were included. The total admitted fraction was 30.9%. For levels of in-hospital occupancy <95%, 95-100%, 100-105%, and >105% the admitted fractions were 31.5%, 30.9%, 29.9%, and 28.7%, respectively. After taking confounding factors into account, the odds ratio for admission were 0.88 (CI 0.84-0.93, P >0.001) for occupancy level 95-100%, 0.82 (CI 0.78-0.87, P >0.001) for occupancy level 100-105%, and 0.74 (CI 0.67-0.81, P >0.001) for occupancy level >105%, relative to the odds ratio for admission at occupancy level <95%. A similar pattern was observed upon subgroup analysis. CONCLUSIONS: In-hospital occupancy was significantly associated with a decreased odds ratio for admission in the study population. One interpretation is that patients who would benefit from inpatient care instead received suboptimal care in outpatient settings at times of high in-hospital occupancy. A second interpretation is that physicians admit patients who could be managed safely in the outpatient setting, in times of good in-hospital bed availability. Physicians thereby expose patients to healthcare-associated infections and other hazards, in addition to consuming resources better needed by others.
RESUMO
BACKGROUND: In 2005, the Advanced Life Support (ALS) teams delivering pre-hospital care in RegionSkane in southern Sweden received additional support by physicians, who were part of "Pre-hospital acute teams" (PHAT). The study objective is to compare the incidence of pre-hospital medical interventions for trauma-patients cared for by conventional ALS teams and patients who received additional support by PHAT. METHODS: Trauma patients with Injury Severity Score (ISS) >9 were identified retrospectively in the national quality registry KVITTRA at three hospitals in RegionSkane, for the time period October 2005 to December 2008. Interventions include e.g. tracheal intubation, administration of i.v. fluids, neck immobilization and spine board usage. Confounding effects from trauma severity, trauma mechanism, vital parameters, age and sex were addressed in multivariate models. RESULTS: Data from 202 cases was included. 9 pre-hospital interventions were assessed. The incidence of endotracheal intubation and immobilisation of extremities was higher among patients in the PHAT-group compared to the ALS-only group (16.3% vs. 6.9%, p = 0.034) and (12.8% vs. 4.3%, p = 0.027) respectively. PHATs presence remained a significant predictor of these interventions also after taking confounding factors into account (OR 5.5, CL 1.5-19.7) and (OR 3.2 CI 1.0-9.8).PHAT was involved in a greater proportion of cases with <50.0% of survival (19.8% vs. 12.1%, p = 0.134). The average ISS was higher among cases receiving PHAT support in strata ISS 16-24 and ISS > 24 than cases in corresponding strata cared for by ALS teams alone (ISS 20.0 vs. 17.0, p = 0.048 and ISS 34.0 vs. 29.0, p = 0.019). CONCLUSIONS: The incidence of endotracheal intubation and immobilization of extremities was greater among patients supported by PHAT, compared to patients cared for by ALS teams alone. This finding has to be interpreted in the light of a selection-bias where PHAT support was directed to more severely injured patients.