Factors preventing nosocomial outbreak following a single case of COVID-19 diagnosed during hospitalization: A retrospective review

Objective: Our hospital experienced a hospital shutdown and quarantine for two weeks after one case of COVID-19 was diagnosed during hospitalization. We analyzed the reopening process following hospital closure and possible factors that prevented hospital spread. Methods: We retrospectively reviewed the confirmed patient’s medical records and results of epidemiological survey available from the infection control team of our hospital. Results: A total of 117 hospital staff members were tested, 26 of whom were self-isolated. Of the 54 inpatients tested, 28 on the same floor and two close contacts in the endoscopic room were quarantined in a single room. Finally, all quarantined hospital staff, inpatients and outpatients were tested for COVID-19 on the 14th day of close contact. The results were all negative, and the hospital work completely resumed. Conclusion: Although closing and isolating the hospital appeared to have played a useful role in preventing the spread of COVID-19 inside the hospital and to the local community, it is still debated whether or not the duration of hospital closure or quarantine was appropriate. The lessons from the two-week hospital closure suggest that wearing a mask, hand hygiene and the ward environment are important factors in preventing nosocomial outbreaks of COVID-19.

A scoping review of internal hospital crises and disasters in the Netherlands, 2000–2020

Background Internal hospital crises and disasters (IHCDs) are events that disrupt the routine functioning of a hospital while threatening the well-being of patients and staff. IHCDs may cause hospital closure, evacuations of patients and loss of healthcare capacity. The consequences may be ruinous for local communities. Although IHCDs occur with regularity, information on the frequency and types of these events is scarcely published in the medical literature. However, gray literature and popular media reports are widely available. We therefore conducted a scoping review of these literature sources to identify and characterize the IHCDs that occurred in Dutch hospitals from 2000 to 2020. The aim is to develop a systematic understanding of the frequency of the various types of IHCDs occurring in a prosperous nation such as the Netherlands. Methods A systematic scoping review of news articles retrieved from the LexisNexis database, Google, Google News, PubMed and EMBASE between 2000 and 2020. All articles mentioning the closure of a hospital department in the Netherlands were analyzed. Results A total of 134 IHCDs were identified in a 20-year time period. Of these IHCDs, there were 96 (71.6%) emergency department closures, 76 (56.7%) operation room closures, 56 (41.8%) evacuations, 26 (17.9%) reports of injured persons, and 2 (1.5%) reported casualties. Cascading events of multiple failures transpired in 39 (29.1%) IHCDs. The primary causes of IHCDs (as reported) were information and communication technology (ICT) failures, technical failures, fires, power failures, and hazardous material warnings. An average of 6.7 IHCDs occurred per year. From 2000–2009 there were 32 IHCDs, of which one concerned a primary ICT failure. Of the 102 IHCDs between 2010–2019, 32 were primary ICT failures. Conclusions IHCDs occur with some regularity in the Netherlands and have marked effects on hospital critical care departments, particularly emergency departments. Cascading events of multiple failures transpire nearly a third of the time, limiting the ability of a hospital to stave off closure due to failure. Emergency managers should therefore prioritize the risk of ICT failures and cascading events and train hospital staff accordingly.

On the Generation of Medical Dialogues for COVID-19

Under the pandemic of COVID-19, people experiencing COVID19-related symptoms or exposed to risk factors have a pressing need to consult doctors. Due to hospital closure, a lot of consulting services have been moved online. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialogue system that can provide COVID19-related consultations. We collected two dialogue datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. On these two datasets, we train several dialogue generation models based on Transformer, GPT, and BERT-GPT. Since the two COVID-19 dialogue datasets are small in size, which bear high risk of overfitting, we leverage transfer learning to mitigate data deficiency. Specifically, we take the pretrained models of Transformer, GPT, and BERT-GPT on dialog datasets and other large-scale texts, then finetune them on our CovidDialog datasets. Experiments demonstrate that these approaches are promising in generating meaningful medical dialogue about COVID-19. But more advanced approaches are needed to build a fully useful dialogue system that can offer accurate COVID-related consultations. The data and code are available at https://github.com/UCSD-AI4H/COVID-Dialogue

On the Generation of Medical Dialogues for COVID19

<div>Under the pandemic of COVID-19, people experiencing COVID19-related symptoms or exposed to risk factors have a pressing need to consult doctors. Due to hospital closure,</div><div>a lot of consulting services have been moved online. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialogue system that can provide COVID19-related consultations. We collected two dialogue datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. On these two datasets, we train several dialogue generation models based on Transformer, GPT, and BERT-GPT. Since the two COVID-19 dialogue datasets are small in size, which bear high risk of overftting, we leverage transfer learning to mitigate data deficiency. Specifically, we take the pretrained models of Transformer, GPT, and BERT-GPT on dialog datasets and other large-scale texts, then finetune them on our CovidDialog datasets. Experiments demonstrate that these approaches are promising in generating meaningful medical dialogues about COVID-19. But more advanced approaches are needed to build a fully useful dialogue system that can offer accurate COVID-related consultations. The data and code are available at https://github.com/UCSD-AI4H/COVID-Dialogue</div>

On the Generation of Medical Dialogues for COVID19

<div>Under the pandemic of COVID-19, people experiencing COVID19-related symptoms or exposed to risk factors have a pressing need to consult doctors. Due to hospital closure,</div><div>a lot of consulting services have been moved online. Because of the shortage of medical professionals, many people cannot receive online consultations timely. To address this problem, we aim to develop a medical dialogue system that can provide COVID19-related consultations. We collected two dialogue datasets - CovidDialog - (in English and Chinese respectively) containing conversations between doctors and patients about COVID-19. On these two datasets, we train several dialogue generation models based on Transformer, GPT, and BERT-GPT. Since the two COVID-19 dialogue datasets are small in size, which bear high risk of overftting, we leverage transfer learning to mitigate data deficiency. Specifically, we take the pretrained models of Transformer, GPT, and BERT-GPT on dialog datasets and other large-scale texts, then finetune them on our CovidDialog datasets. Experiments demonstrate that these approaches are promising in generating meaningful medical dialogues about COVID-19. But more advanced approaches are needed to build a fully useful dialogue system that can offer accurate COVID-related consultations. The data and code are available at https://github.com/UCSD-AI4H/COVID-Dialogue</div>