key: cord-0895962-tqy3i3ai authors: Gholamzadeh, Marsa; Abtahi, Hamidreza; Safdari, Reza title: Suggesting a framework for preparedness against the pandemic outbreak based on medical informatics solutions: a thematic analysis date: 2021-01-27 journal: Int J Health Plann Manage DOI: 10.1002/hpm.3106 sha: 299407653657457b13bbc7d958bd82a6d4f1437b doc_id: 895962 cord_uid: tqy3i3ai BACKGROUND: When an outbreak emerged, each country needs a coherent and preventive plan to deal with epidemics. In the era of technology, adopting informatics‐based solutions is essential. The main objective of this study is to propose a conceptual framework to provide a rapid and responsive surveillance system against pandemics. METHODS: A three‐step approach was employed in this research to develop a conceptual framework. These three steps comprise (1) literature review, (2) extracting and coding concepts, and determining main themes based on thematic analysis using ATLAS.ti® software, and (3) mapping concepts. Later, all of the results synthesized under expert consultation to design a conceptual framework based on the main themes and identified strategies related to medical informatics. RESULTS: In the literature review phase, 65 articles were identified as eligible studies for analysis. Through line by line coding in thematic analysis, more than 46 themes were extracted as potential foremost themes. Based on the key themes and strategies were employed by studies, the proposed framework designed in three main components. The most appropriate strategies that can be used in each section were identified based on the demands of each part and the available solutions. These solutions were employed in the final framework. CONCLUSION: The presented model in this study can be the first step for a better understanding of the potential of medical informatics solutions in promoting epidemic disease management. It can be applied as a reference model for designing intelligent surveillance systems to prepare for probable future pandemics. According to World Health Organization (WHO), the pandemic is commonly defined as the worldwide spread of a new communicable disease. 1, 2 The randomness and unpredictability of such diseases is a prominent feature of each new pandemic. This feature causes countries to face inevitable challenges. [3] [4] [5] When a disease outbreak began, most people lack natural immunity to fight it. It can cause a rapid transmission of new pandemic across countries over a specific amount of time. 6 The unfamiliar aspects and highly contagious nature of the current COVID-19 pandemic, have shown that every country requires a coherent and responsive plan to battle against the current pandemic. 7 By the 10 October 2020, there had been 1083140 COVID-19 deaths worldwide. A high mortality rate of this new virus has raised many concerns about the unpreparedness of countries to use appropriate methods to control, prevent and address unexpected epidemiological conditions. 8 Thus, during a pandemic outbreak, health care system preparedness is essential. In similar epidemics around the world, various countries have taken different approaches based on the last advancement of technologies and health-IT based solutions. 9 In former pandemics, various digital health strategies have been used with different approaches to control other epidemics such as the Middle East respiratory syndrome, severe acute respiratory syndrome and H1N1 flu. [10] [11] [12] For example, a team of scientists from Pakistan invented a smart tool called ID-Viewer as a decisionmaking system for predicting an infectious disease outbreak in 2016. 13 It was employed to detect the dengue epidemic for 20 weeks earlier by gathering and analysing all related information about dengue disease since 2011. 14 In a similar study conducted in 2012, Chinese researchers were able to predict abnormal outbreaks and warn the health system before its emergence by implementation a continuous and intelligent monitoring system to analyse real-time data of various diseases. 15, 16 The recent global epidemic has proved that e-health technologies can be used to control the spread of disease. In a recent report, Healthcare Information and Management Systems Society reported that digital tools such as telemedicine, remote patient monitoring, data analysis methods and even artificial intelligence (AI)-based solutions could play a significant role in restricting the prevalence of COVID- 19. 17 However, the sudden emergence of the epidemic has proven that just awareness of the latest technologies is not adequate. The most significant point to fight epidemic diseases by applying the latest technologies is to know how to use these tools most appropriately in outbreaks. 18 Consequently, to prepare for combat against the spread of diseases, an appropriate model based on the latest medical informatics solutions is needed. It seems that the time has come to use digital technologies at different levels of the health system based on the experiences of other countries to combat the epidemic of infectious diseases. 19 The main objective of this study is to propose a conceptual framework for designing an appropriate and comprehensive electronic surveillance system for preparedness against the pandemic. Other aims of this study include identifying the most proper solutions in medical informatics that can be used to predict, diagnose, control and manage the COVID-19 outbreak and other pandemics. In this qualitative research, a three-step approach was applied to outline a conceptual framework. These three steps included (1) literature review, (2) extracting concepts and main themes based on qualitative analysis, and (3) mapping concepts, and synthesizing the results with expert consultation. The qualitative synthesis was conducted based on the Standards for Reporting Qualitative Research checklist, which can help researchers to report their results correctly. 20 The systematic search was performed in four scientific databases, Web of Science, Scopus, PubMed and Science Direct from 2000 to August 2020. The search strategy and keywords are shown in Table A1 , Supplementary Appendix. All articles retrieved from database searches entered into Mendeley as the resource management software. The systematic review was conducted based on the preferred reporting items for systematic reviews and meta-analyses steps. Some inclusion and exclusion criteria were determined for reviewing articles. Articles included if they were original articles and if they used solutions or strategies to prepare for a sudden outbreak or control of an epidemic disease based on digital health. Solutions regarding social relations, military and cultural measures are not considered. Non-English papers, letter to editors, commentary papers, book chapters, short briefs, reports, technical reports, any reviews or meta-analysis were excluded. After duplication removal, the articles were omitted regarding the type of articles. Next, the remaining studies were reviewed based on titles and abstracts. All titles and abstracts of articles were examined to select eligible studies by reviewers. Marsa Gholamzadeh screened all titles and abstracts to find relevant articles. A second reviewer (Hamidreza Abtahi) reviewed a sample of studies randomly. Following, articles that met our inclusion criteria were selected for full-text review. After that, the full texts of relevant studies were screened thoroughly by all authors. If there was a disagreement between the authors, the final decision was made by Reza Safdari. Finally, the information of the included articles extracted based on characteristics such as author name, year of publication, title, purpose of the study, country, institution, proposed solution and type of disease and pandemic. Since thematic analysis known as one of the best methods in qualitative analysis, we applied it in our research. 21 Following a systematic investigation, the remained articles met our criteria were imported to ATLAS.ti® software to conduct an inquiry using inductive thematic analysis. It is a free famous software that is mostly used for content analysis by coding and analysing complex textual data. 22 All included articles were imported into the ATLAS.ti® software. All of the studies were screened line by line to code the preliminary idea. By connecting extracted codes, the fundamental themes were extended to achieve a thematic map. All potential themes were depicted in the form of a thematic tree to define themes and sub-themes. Coding and thematic analysis stages were conducted by one of the authors (Marsa Gholamzadeh) who had experience in analysing and reviewing studies under the supervision of health informatics experts. The information extracted by the researchers was re-examined to reach an agreement. The next reviewers (Hamidreza Abtahi and Reza Safdari) assessed and verified the extracted information. GHOLAMZADEH ET AL. Under expert consultation, all of the extracted themes were investigated and integrated. The initial model devised in this step is based on key themes and sub-themes. Then, in an iterative process, a conceptual framework was defined and redesigned to achieve the optimal model. In the following, the suggested conceptual model and the strategies for outbreak preparedness were described. The different parts of the proposed model are defined based on the solutions available concerning the various branches of medical informatics. Initial searches in scientific databases yielded 397 citations. After removing 71 duplicated articles, 326 citations were screened based on the type of studies. Next, 300 articles remained due to their relevancy in the abstract F I G U R E 1 Screening flow based on the preferred reporting items for systematic reviews and meta-analyses method screening phase. Then, the full-text of 249 studies were reviewed. Finally, 65 citations were identified as eligible studies to meet our criteria. The process of screening articles is shown in Figure 1 . A summary of the included articles based on predefined categories is described in Table 1 . The analysis of the most significant features of the reviewed articles is represented in Table 2 . Regarding the country, the United States has used medical informatics solutions more than any other country to control pandemics. On the other hand, the analysis revealed that most of the efforts devoted to controlling the spread of influenza outbreaks (41.54%) and employing medical informatics solutions. The trend of published articles regarding our objective had steady growth until the COVID-19 outbreak in 2020. Though, eight articles were published regarding outbreak preparedness from the beginning of 2020 up to 2 August 2020. Based on the central idea of our research, all studies were examined based on their tactics and strategies related to the medical informatics disciplines. Overall, all of the employed strategies can be devoted to 19 categories. These categories are represented based on their frequency and percentage in Table 3 . It is worth noting that most studies used a combination of different techniques and did not focus on just one specific solution but most strategies were related to developing AI-based models. After coding all the themes and sub-themes, more than 46 themes were extracted as potential main themes. By integrating all of the potential themes, the mapped network of themes is devised and illustrated in Figure 2 . Conclusively, all of the themes are summarized regarding recommended and applied strategies in three main categories. These categories were used to devise a conceptual model: The most appropriate solutions that can be used in each section are identified based on the integration of the requirements and the available solutions in the proposed model. In the following, the most suitable model for better epidemic management is designed in each section using the identified solutions. Analysis of articles showed that mathematical models have the potential to aid clinical decision-makers to forecast the next epidemic and prepare for a proper pandemic. Besides, the thematic analysis revealed that predictive modelling was the most common solution to develop an early warning system. It has the potential to predict outbreaks for providing emergency response. Analysis of studies showed that outbreak prediction models, AI-based algorithms and early warning systems, in combination with geographical positioning strategies can be adopted in this section. The schematic model of this subsystem is represented in Figure 3 . The first part of this model is an intelligent subsystem that investigates and interprets all recorded data in cross-sectional studies, death reports and case reports continuously to detect an abnormal pattern of a particular disease using built-in AI-based algorithms. In case of occurrence of an abnormal pattern, the system would be GHOLAMZADEH ET AL. placed in alert mode. Later, the subsystem of outbreak prediction will activate. Once epidemiologists confirm the outbreak, a survival system begins to run. Then the subsystems concerning clinical care and outbreak management will start to operate. The results of the epidemiological subsystem can also be used by clinical researchers for further studies. By integrating this subsystem with geographic systems, a model can be designed to visualize how the disease is transmitted, identifying infected areas and finding people at risk. Through analysing the obtained data and connecting the results to mobile applications, it is also possible to inform people who live in high-risk areas. During the disease outbreak, one of the challenges was considered by most articles is how to allocate health and human resources during the epidemic. Therefore, it is necessary to consider a subsystem to manage the proper allocation of resources to the health system. Reviewing literature showed that after a new disease outbreak, resource allocations could be handled by defining appropriate heuristic algorithms and analysing real-time data. The schematic model of the suggested subsystem is represented in Figure 4 . In an epidemic event, patient treatment and follow-up are the most significant issue to decrease the mortality rate. Hence, the clinical care subsystem is explained in the following based on the most useful strategies applied in reviewed articles. In a thematic analysis of the most influential ideas, clinical care planning was one of the most repetitive concepts extracted in the qualitative analysis stage of this research. The most common solutions that appeared in the analysis regarding this section include (1) implementing electronic health records for patient management, (2) using decision support systems and computerized physician order entry tool to make better decisions and prevent medical errors, and (3) model generation. Moreover, telemedicine can be utilized in clinical care planning. Due to the importance of the clinical system and its complexities, this system is better designed in two subsystems of monitoring outpatients and inpatients. The research and management subsystem model; NLP, natural language process GHOLAMZADEH ET AL. -23 The proposed system can be designed in such a way that the rapid alert system is activated when the disease is detected. By activating the alert status, this system will automatically give the necessary alerts to medical centres. Additionally, the latest treatment protocols will be provided to specialists in the form of embedded knowledge in such a framework. The inpatient subsystem model 24 - As it is apparent in Figure 5 , an appropriate strategy can be employed in this section to identify people at risk using knowledge obtained from evidence, and regular monitoring of epidemiological data. The early screening module is one of the main parts of the proposed framework. As soon as the imminent alert system is activated, the infected geographical areas could be recognized. The necessary warning messages can then be sent to residents of high-risk areas to prevent further outbreaks. Early screening of patients could be conducted in infected areas by applying different methods such as developing selfassessment websites and mobile-based applications, and telephone-based counselling. Accordingly, all of the people who were at the risk of exposure can be determined. Taking such an approach can also be effective in reducing the prevalence of COVID-19 in low-and middle-income countries. If the person is suspected of having the disease, he/she will be assessed based on standard checklists by health care providers. Then, the risk of disease in each person will be calculated. If the risk of disease is high, the patient will be monitored remotely through regular telephone consultation. For more investigation, the patient will be referred to medical centres if he needs it. Then, a medical record will be created for him. If he needs to be hospitalized, his information will be referred to the hospital subsystem. In a viral illness, the patient must be quarantined. To better communicate, secure social networks can be designed for patients to discuss their problems and share their experiences. These social networks should be managed under the supervision of health professionals to prevent the spread of untruths. Developing telemedicine programs is crucial to monitor non hospitalized patients in severe communicable diseases like COVID-19. So, monitoring patients through telemonitoring programs and virtual clinics could reduce disease transmission. It can be helpful in better controlling the spread of diseases such as COVID-19. Self-monitoring of people who have some symptoms is possible by employing such a system. Through telemonitoring, if the patient has the initial symptoms of the disease, it can first be followed remotely. After an initial investigation, the system referred the patient to medical centres if it is required. According to other studies, teleeducation can be used to better educate patients for the prevention and control of the disease using social networks and distance education. In the proposed model, once the patient's referred, the information is automatically entered into the system to create a file for the patient. During patient monitoring, control, and follow-up, all recorded information is continuously analysed to extract the disease pattern. At the same time, due to the connection of the system to the knowledge base, patients are followed according to the treatment protocols based on standard treatment steps. Physicians are also asked to enter their new findings concerning the disease in the relevant section. These new achievements could be added to the knowledge-base after expert approval. The details of the suggested strategy are shown in Figure 6 . At the time of hospitalization, the knowledge regarding the disease symptoms and its characteristics are extracted for better management and disease control using machine learning algorithms and real-time data analysis. The knowledge can be used for appropriate resource allocations in critically ill patients, prescribing the most effective drugs and prioritizing patients. Figure 7 , the proposed model is a four-layer model. These layers comprise the user interface layer, the logic layer, the application layer and the data layer. Despite each layer of the layered architecture pattern has a specific role and responsibility, all layers have interacted with each other to pass the information up to the presentation layer. Layered architecture pattern is the most common architecture pattern for developing web-based applications in healthcare settings. 31 Analysis of related articles showed that the most common strategy applied by researchers was designing early warning systems. In line with previous studies, it is apparent that the first step of 'preparation' is the continuous investigation and prevention of further outbreaks. 27 Accordingly, the popularity of mathematical models to forecast epidemic diseases has been increased to better help policy-makers in decision-making with the development of AIbased methods in other studies. 32 Consequently, outbreak investigation and prediction are considered as a prerequisite of preparedness to activate other subsystems in the form of an early detection system in the epidemiological section. In an epidemic, physicians should be aware of the best available evidence as soon as possible. Meanwhile, the work of the medical staff increases during the outbreak of the disease. Therefore, limited time is a common challenge in critical situations. As in the Coronavirus pandemic, the workload of healthcare providers has increased, and health care providers have to work in heavy workload conditions. 33, 34 To address this challenge, an AI-based technique such as text mining in the research and management section was considered to retrieve the required evidence as soon as possible. Recently, some studies were conducted to extract valuable knowledge from published literature and summarizing the most up-to-date research using natural language processing. 35 This solution can enable clinicians to access information from a huge amount of scientific evidence published recently. 36 Allocation of the appropriate resources and well-coordinated care during the time of the outbreak is one of the main themes and strategies extracted through our analysis. It accounted for another challenge that we usually face in an outbreak. 37 Thus, an intelligent procedure was considered in our proposed model using data mining mechanisms to solve this problem. Case findings and accurate diagnosis are crucial in controlling the outbreak of an epidemic. 38 Hence, a systematic process in patient management and clinical care planning is embedded to find infected people in the proposed model. The next important issue is related to the continuous care of infected patients. Thus, establishing a standard of care for triage and treatment could be remarkably efficient during a disease outbreak in disease control and enhance patient safety. 39 Patient management is also considered in two parts for an effective clinical care program. It comprises outpatient and inpatient modules using solutions such as applying decision support systems, implementing electronic health records, remote monitoring and utilizing data mining tools in the presented model. GHOLAMZADEH ET AL. -27 The proposed model tries to provide a framework to prepare for a sudden outbreak of disease by considering usable medical informatics methods. This model could provide deeper insights into the designing of a surveillance system for public health professionals from a medical informatics perspective. Such a framework can be optimized to be applied in different situations. Moreover, this study is the first stage of further research to validate the framework through focus group discussions, Delphi survey, or expert consensus. Once the model is validated, it can be considered as a reference framework in developing surveillance systems to prepare for the next epidemic. There are several limitations to this research. The proposed framework was designed from the author's point of view. However, this is only a proposed model for further studies and a new perspective on the control of communicable diseases regarding medical informatics solutions. Also, this is a non-validated model, and its validation will be examined in future studies. Not relevant. No funding to declare. The study involves only a review of literature without involving humans and/or animals. The authors have no ethical conflicts to disclose. Pandemic preparedness and response-lessons from the H1N1 influenza of The classical definition of a pandemic is not elusive Stationary distribution and extinction of stochastic coronavirus (COVID-19) epidemic model Pandemics: waves of disease, waves of hate from the Plague of Athens to A.I.D.S Randomness of dengue outbreaks on the equator SARS-CoV-2 and COVID-19 in older adults: what we may expect regarding pathogenesis, immune responses, and outcomes Getting ready for the next pandemic COVID-19: why we need to be more prepared and less scared We need compassionate leadership management based on evidence to defeat COVID-19 Digital health and the COVID-19 epidemic: an assessment framework for apps from an epidemiological and legal perspective Protecting residential care facilities from pandemic influenza US university response to H1N1: a study of access to online preparedness and response information Development of a resource modelling tool to support decision makers in pandemic influenza preparedness: The AsiaFluCap Simulator. BMC Publ Health A visual analytics architecture for infectious diseases surveillance and response management in Pakistan A systematic review of spatial decision support systems in public health informatics supporting the identification of high risk areas for zoonotic disease outbreaks Exploring the spatial and temporal relationships between mosquito population dynamics and dengue outbreaks based on climatic factors Surveillance and early warning systems of infectious disease in China: from 2012 to 2014 The Role of Health Technology and Informatics in a Global Public Health Emergency: Practices and Implications From the COVID-19 Pandemic Using Information Technology to Manage the COVID-19 Pandemic: Development of a Technical Framework Based on Practical Experience in China AI Techniques for COVID-19 Reporting qualitative research: standards, challenges, and implications for health design Thematic analysis: striving to meet the trustworthiness criteria Thematic content analysis using ATLAS. ti software: potentialities for researches in health Scientific and technological strategies proposal for the construction of digital public health emergency management system in China New technologies in predicting, preventing and controlling emerging infectious diseases Smart technologies for fighting pandemics: the techno-and human-driven approaches in controlling the virus transmission Creating a process for incorporating epidemiological modelling into outbreak management decisions Communicable Diseases and Outbreak Control Modelling the propagation of social response during a disease outbreak Outbreak Trends of Coronavirus Disease-2019 in India: a Prediction Assessing global preparedness for the next pandemic: development and application of an Epidemic Preparedness Index A layered architecture for uniform version management Modelling for disease preparedness and response How essential is to focus on physician's health and burnout in coronavirus (COVID-19) pandemic? Cureus Doctors' wellbeing: self-care during the covid-19 pandemic Information extraction from scientific articles: a survey Biomedical text mining for research rigor and integrity: tasks, challenges, directions. Briefings Bioinforma Which recommendations are considered essential for outbreak preparedness by first responders? Importance of diagnostics in epidemic and pandemic preparedness A systematic review on the causes of the transmission and control measures of outbreaks in long-term care facilities: back to basics of infection control Suggesting a framework for preparedness against the pandemic outbreak based on medical informatics solutions: a thematic analysis The author(s) declared no potential conflicts of interest concerning the research, authorship, and/or publication of this article. Reza Safdari: Conception and design of study, Acquisition of data, Analysis and/or interpretation of data, Revising the manuscript critically for important intellectual content. Hamidreza Abtahi: Conception and design of study, Acquisition of data, Draughting the manuscript. Marsa Gholamzadeh: Conception and design of study, Acquisition of data, Analysis and/or interpretation of data, Draughting the manuscript, Revising the manuscript critically for important intellectual content. Approval of the version of the manuscript to be published Data sharing not applicable to this article as no datasets were generated or analysed during the current study. https://orcid.org/0000-0001-6781-9342Hamidreza Abtahi https://orcid.org/0000-0002-1111-0497Reza Safdari https://orcid.org/0000-0002-4982-337X