Cost benefit analysis of alternative testing and quarantine policies for travelers for infection control: A case study of Singapore during the COVID-19 pandemic

Background Border control mitigates local infections but bears a heavy economic cost, especially for tourism-reliant countries. While studies have supported the efficacy of border control in suppressing cross-border transmission, the trade-off between costs from imported and secondary cases and from lost economic activities has not been studied. This case study of Singapore during the COVID-19 pandemic aims to understand the impacts of varying quarantine length and testing strategies on the economy and health system. Additionally, we explored the impact of permitting unvaccinated travelers to address emerging equity concerns. We assumed that community transmission is stable and vaccination rates are high enough that inbound travelers are not dissuaded from traveling. Methods The number of travelers was predicted considering that longer quarantine reduces willingness to travel. A micro-simulation model predicted the number of COVID-19 cases among travelers, the resultant secondary cases, and the probability of being symptomatic in each group. The incremental net monetary benefit (INB) of Singapore was quantified under each border-opening policy compared to pre-opening status, based on tourism receipts, cost/profit from testing and quarantine, and cost and health loss due to COVID-19 cases. Results Compared to polymerase chain reaction (PCR), rapid antigen test (ART) detects fewer imported cases but results in fewer secondary cases. Longer quarantine results in fewer cases but lower INB due to reduced tourism receipts. Assuming the proportion of unvaccinated travelers is small (8% locally and 24% globally), allowing unvaccinated travelers will accrue higher INB without exceeding the intensive care unit (ICU) capacity. The highest monthly INB from all travelers is $2,236.24 m, with 46.69 ICU cases per month, achieved with ARTs at pre-departure and on arrival without quarantine. The optimal policy in terms of highest INB is robust under changes to various model assumptions. Among all cost-benefit components, the top driver for INB is tourism receipts. Conclusions With high vaccination rates locally and globally alongside stable community transmission, opening borders to travelers regardless of vaccination status will increase economic growth in the destination country. The caseloads remain manageable without exceeding ICU capacity, and costs of cases are offset by the economic value generated from travelers.

Economic Analysis of Border Control Policies during COVID-19 Pandemic: A Modelling Study to Inform Cross-Border Travel Policy between Singapore and Thailand

With countries progressing towards high COVID-19 vaccination rates, strategies for border reopening are required. This study focuses on Thailand and Singapore, two countries that share significant tourism visitation, to illustrate a framework for optimizing COVID-19 testing and quarantine policies for bilateral travel with a focus on economic recovery. The timeframe is the month of October 2021, when Thailand and Singapore were preparing to reopen borders for bilateral travel. This study was conducted to provide evidence for the border reopening policy decisions. Incremental net benefit (INB) compared to the pre-opening period was quantified through a willingness-to-travel model, a micro-simulation COVID-19 transmission model and an economic model accounting for medical and non-medical costs/benefits. Multiple testing and quarantine policies were examined, and Pareto optimal (PO) policies and the most influential components were identified. The highest possible INB for Thailand is US $125.94 million, under a PO policy with no quarantine but with antigen rapid tests (ARTs) pre-departure and upon arrival to enter both countries. The highest possible INB for Singapore is US $29.78 million, under another PO policy with no quarantine on both sides, no testing to enter Thailand, and ARTs pre-departure and upon arrival to enter Singapore. Tourism receipts and costs/profits of testing and quarantine have greater economic impacts than that from COVID-19 transmission. Provided healthcare systems have sufficient capacity, great economic benefits can be gained for both countries by relaxing border control measures.

Economic Analysis of Border Control Policies during COVID-19 Pandemic: A Modelling Study to Inform Cross-Border Travel Policy between Singapore and Thailand.

With countries progressing towards high COVID-19 vaccination rates, strategies for border reopening are required. This study focuses on Thailand and Singapore, two countries that share significant tourism visitation, to illustrate a framework for optimizing COVID-19 testing and quarantine policies for bilateral travel with a focus on economic recovery. The timeframe is the month of October 2021, when Thailand and Singapore were preparing to reopen borders for bilateral travel. This study was conducted to provide evidence for the border reopening policy decisions. Incremental net benefit (INB) compared to the pre-opening period was quantified through a willingness-to-travel model, a micro-simulation COVID-19 transmission model and an economic model accounting for medical and non-medical costs/benefits. Multiple testing and quarantine policies were examined, and Pareto optimal (PO) policies and the most influential components were identified. The highest possible INB for Thailand is US $125.94 million, under a PO policy with no quarantine but with antigen rapid tests (ARTs) pre-departure and upon arrival to enter both countries. The highest possible INB for Singapore is US $29.78 million, under another PO policy with no quarantine on both sides, no testing to enter Thailand, and ARTs pre-departure and upon arrival to enter Singapore. Tourism receipts and costs/profits of testing and quarantine have greater economic impacts than that from COVID-19 transmission. Provided healthcare systems have sufficient capacity, great economic benefits can be gained for both countries by relaxing border control measures.

Cost benefit analysis of alternative testing and quarantine policies for travelers for infection control: A case study of Singapore during the COVID-19 pandemic.

Background: Border control mitigates local infections but bears a heavy economic cost, especially for tourism-reliant countries. While studies have supported the efficacy of border control in suppressing cross-border transmission, the trade-off between costs from imported and secondary cases and from lost economic activities has not been studied. This case study of Singapore during the COVID-19 pandemic aims to understand the impacts of varying quarantine length and testing strategies on the economy and health system. Additionally, we explored the impact of permitting unvaccinated travelers to address emerging equity concerns. We assumed that community transmission is stable and vaccination rates are high enough that inbound travelers are not dissuaded from traveling. Methods: The number of travelers was predicted considering that longer quarantine reduces willingness to travel. A micro-simulation model predicted the number of COVID-19 cases among travelers, the resultant secondary cases, and the probability of being symptomatic in each group. The incremental net monetary benefit (INB) of Singapore was quantified under each border-opening policy compared to pre-opening status, based on tourism receipts, cost/profit from testing and quarantine, and cost and health loss due to COVID-19 cases. Results: Compared to polymerase chain reaction (PCR), rapid antigen test (ART) detects fewer imported cases but results in fewer secondary cases. Longer quarantine results in fewer cases but lower INB due to reduced tourism receipts. Assuming the proportion of unvaccinated travelers is small (8% locally and 24% globally), allowing unvaccinated travelers will accrue higher INB without exceeding the intensive care unit (ICU) capacity. The highest monthly INB from all travelers is $2,236.24 m, with 46.69 ICU cases per month, achieved with ARTs at pre-departure and on arrival without quarantine. The optimal policy in terms of highest INB is robust under changes to various model assumptions. Among all cost-benefit components, the top driver for INB is tourism receipts. Conclusions: With high vaccination rates locally and globally alongside stable community transmission, opening borders to travelers regardless of vaccination status will increase economic growth in the destination country. The caseloads remain manageable without exceeding ICU capacity, and costs of cases are offset by the economic value generated from travelers.

Users' Beliefs Toward Physical Distancing in Facebook Pages of Public Health Authorities During COVID-19 Pandemic in Early 2020.

This study aims to describe Facebook users' beliefs toward physical distancing measures implemented during the Coronavirus disease (COVID-19) pandemic using the key constructs of the health belief model. A combination of rule-based filtering and manual classification methods was used to classify user comments on COVID-19 Facebook posts of three public health authorities: Centers for Disease Control and Prevention of the United States, Public Health England, and Ministry of Health, Singapore. A total of 104,304 comments were analyzed for posts published between 1 January, 2020, and 31 March, 2020, along with COVID-19 cases and deaths count data from the three countries. Findings indicate that the perceived benefits of physical distancing measures (n = 3,463; 3.3%) was three times higher than perceived barriers (n = 1,062; 1.0%). Perceived susceptibility to COVID-19 (n = 2,934; 2.8%) was higher compared with perceived severity (n = 2,081; 2.0%). Although susceptibility aspects of physical distancing were discussed more often at the start of the year, mentions on the benefits of intervention emerged stronger toward the end of the analysis period, highlighting the shift in beliefs. The health belief model is useful for understanding Facebook users' beliefs at a basic level, and it provides a scope for further improvement.

The Relative Importance of Vulnerability and Efficiency in COVID-19 Contact Tracing Programmes: A Discrete Choice Experiment.

Objectives: This study aims to assess the trade-offs between vulnerability and efficiency attributes of contact tracing programmes based on preferences of COVID-19 contact tracing practitioners, researchers and other relevant stakeholders at the global level. Methods: We conducted an online discrete choice experiment (DCE). Respondents were recruited globally to explore preferences according to country income level and the prevailing epidemiology of COVID-19 in the local setting. The DCE attributes represented efficiency (timeliness, completeness, number of contacts), vulnerability (vulnerable population), cooperation and privacy. A mixed-logit model and latent class analysis were used. Results: The number of respondents was 181. Timeliness was the most important attribute regardless of country income level and COVID-19 epidemiological condition. Vulnerability of contacts was the second most important attribute for low-to-lower-middle-income countries and third for upper-middle-to-high income countries. When normalised against conditional relative importance of timeliness, conditional relative importance of vulnerability ranged from 0.38 to 0.42. Conclusion: Vulnerability and efficiency criteria were both considered to be important attributes of contact tracing programmes. However, the relative values placed on these criteria varied significantly between epidemiological and economic context.

Physical or Virtual? - A Qualitative Inquiry Into Youth Perception and Preference for Safe Spaces (preprint)

Background Control measures from the COVID-19 pandemic had far-reaching and multifaceted consequences on the physical and mental health of youths. In particular, youths experienced a lack of safe space to turn to in times of need. Safe spaces, be it physical or virtual, provide social support and connection for youths, which are important for one’s mental health.Methods We conducted a qualitative study in Singapore to investigate how youths interpreted what safe spaces meant to them and how these spaces positively impacted their wellbeing. 48 youths were purposively recruited and took part in a one-on-one interview over Zoom that lasted between 45 minutes to 60 minutes each. Audio recordings of the interviews were transcribed, and data were thematically analysed.Results Findings from our study indicated that participants found three different interpretations for what made a space safe for them - namely as an absence of a negative experience, a neutral space where they could just be themselves, or an affirmative place to receive what they need. Youths preferred safe spaces that were physical in nature due to the relational aspects of the space which provided comfort and preferred safe spaces that were virtual in nature due to the logistical benefits of the space which allowed them to expand their social network.Conclusions Based on our findings, we propose to address youth mental health not only through schools and families, but adopting multi-sectoral public health approach, using physical and virtual safe spaces as both preventive and therapeutic strategies to address often hidden mental health challenges youth face in Singapore. When designing safe spaces for youths, it is important to consider the medium, content and audience involved to maximise youth engagement and in turn, improve the state of mental wellness for youths.

The Relative Importance of Vulnerability and Efficiency in COVID-19 Contact Tracing Programmes: A Discrete Choice Experiment

Objectives: This study aims to assess the trade-offs between vulnerability and efficiency attributes of contact tracing programmes based on preferences of COVID-19 contact tracing practitioners, researchers and other relevant stakeholders at the global level. Methods: We conducted an online discrete choice experiment (DCE). Respondents were recruited globally to explore preferences according to country income level and the prevailing epidemiology of COVID-19 in the local setting. The DCE attributes represented efficiency (timeliness, completeness, number of contacts), vulnerability (vulnerable population), cooperation and privacy. A mixed-logit model and latent class analysis were used. Results: The number of respondents was 181. Timeliness was the most important attribute regardless of country income level and COVID-19 epidemiological condition. Vulnerability of contacts was the second most important attribute for low-to-lower-middle-income countries and third for upper-middle-to-high income countries. When normalised against conditional relative importance of timeliness, conditional relative importance of vulnerability ranged from 0.38 to 0.42. Conclusion: Vulnerability and efficiency criteria were both considered to be important attributes of contact tracing programmes. However, the relative values placed on these criteria varied significantly between epidemiological and economic context.

A Data-Driven Analysis of the Economic Cost of Non-Pharmaceutical Interventions: A Cross-Country Comparison of Kenya, Singapore, and Thailand.

Objective: To estimate the economic impact of border closure and social distancing by estimating the decline of gross domestic product (GDP) in Kenya, Singapore and Thailand. Methods: We analysed secondary data retrospectively. To calculate impact of NPIs on GDP, the relationship between GDP and stock market index was examined using ordinary least squares (OLS). Then, autoregressive and moving averages (ARMA) model was used to examine the impact of NPI on stock market index. The change in GDP due to NPIs was derived by multiplying coefficients of OLS and ARMA models. Results: An increase in stock market index correlated with an increase in GDP, while both social distancing and border closure negatively correlated with stock market index. Implementation of NPIs correlated with the decline in GDP. Thai border closure had a greater decline in GDP than social distancing; Kenya exhibited the same trends; Singapore had the opposite trend. Conclusion: We quantified the magnitude of economic impact of NPIs in terms of GDP decline by linking stock market index and GDP. This approach may be applicable in other settings.

Professional Oversight of Emergency-Use Interventions and Monitoring Systems: Ethical Guidance From the Singapore Experience of COVID-19.

High degrees of uncertainty and a lack of effective therapeutic treatments have characterized the COVID-19 pandemic and the provision of drug products outside research settings has been controversial. International guidelines for providing patients with experimental interventions to treat infectious diseases outside of clinical trials exist but it is unclear if or how they should apply in settings where clinical trials and research are strongly regulated. We propose the Professional Oversight of Emergency-Use Interventions and Monitoring System (POEIMS) as an alternative pathway based on guidance developed for the ethical provision of experimental interventions to treat COVID-19 in Singapore. We support our proposal with justifications that establish moral duties for physicians to record outcomes data and for institutions to establish monitoring systems for reporting information on safety and effectiveness to the relevant authorities. Institutions also have a duty to support generation of evidence for what constitutes good clinical practice and so should ensure the unproven intervention is made the subject of research studies that can contribute to generalizable knowledge as soon as practical and that physicians remain committed to supporting learning health systems. We outline key differences between POEIMS and other pathways for the provision of experimental interventions in public health emergencies.

Determining quarantine length and testing frequency for international border opening during the COVID-19 pandemic.

BACKGROUND: The COVID-19 pandemic has resulted in the closure or partial closure of international borders in almost all countries. Here, we investigate the efficacy of imported case detection considering quarantine length and different testing measures for travellers on arrival. METHODS: We examine eight broad border control strategies from utilizing quarantine alone, pre-testing, entry and exit testing, and testing during quarantine. In comparing the efficacy of these strategies, we calculate the probability of detecting travellers who have been infected up to 2 weeks pre-departure according to their estimated incubation and infectious period. We estimate the number of undetected infected travellers permitted entry for these strategies across a prevalence range of 0.1-2% per million travellers. RESULTS: At 14-day quarantine, on average 2.2% (range: 0.5-8.2%) of imported infections are missed across the strategies, leading to 22 (5-82) imported cases at 0.1% prevalence per million travellers, increasing up to 430 (106-1641) at 2%. The strategy utilizing exit testing results in 3.9% (3.1-4.9%) of imported cases being missed at 7-day quarantine, down to 0.4% (0.3-0.7%) at 21-day quarantine, and the introduction of daily testing, as the most risk averse strategy, reduces the proportion further to 2.5-4.2% at day 7 and 0.1-0.2% at day 21 dependent on the tests used. Rapid antigen testing every 3 days in quarantine leads to 3% being missed at 7 days and 0.7% at 14 days, which is comparable to PCR testing with a 24-hour turnaround. CONCLUSIONS: Mandatory testing, at a minimal of pre-testing and on arrival, is strongly recommended where the length of quarantining should then be determined by the destination country's level of risk averseness, pandemic preparedness and origin of travellers. Repeated testing during quarantining should also be utilized to mitigate case importation risk and reduce the quarantining duration required.

Digital Support for Renal Patients Before and During the COVID-19 Pandemic: Examining the Efforts of Singapore Social Service Agencies in Facebook.

During the coronavirus disease 2019 (COVID-19) pandemic, social service agencies (SSAs) play a crucial role in supporting renal patients, who are particularly vulnerable to infections. Social media platforms such as Facebook, serves as an effective medium for these SSAs to disseminate information. Content analysis of the SSAs' Facebook posts can provide insights on whether Facebook has been adequately utilized during the COVID-19 pandemic and enable SSAs to improve their social media use in future pandemics. This study aimed to compare renal-related SSAs' Facebook post content before and during the COVID-19 pandemic. Facebook posts of three SSAs National Kidney Foundation (NKF), Kidney Dialysis Foundation (KDF), and Muslim Kidney Action Association (MKAC), posted during the pre-COVID-19 period (January 23, 2019 to June 2, 2019) and the peri-COVID-19 period (January 23, 2020 to June 1, 2020) were extracted. A classification scheme was developed by two coders with themes derived inductively and deductively. Each Facebook post was assigned with a theme. Quantitative analyses indicate that the number of Facebook posts increased from 115 in the pre-COVID-19 period to 293 in the peri-COVID-19 period. During peri-COVID-19, posts regarding lifestyle changes, donations and infectious disease surfaced. While the proportion of posts about encouraging kindness increased from one to 77 posts, the proportion of posts about community-based events and psychosocial support decreased from 44 to 15 posts and 17 to 10 posts respectively during the two periods. Facebook was found to be well-utilized by two of the three renal SSAs in engaging their beneficiaries during the pandemic. During future pandemics, renal SSAs should place emphasis on posts related to psychosocial support and encouraging kindness. Further studies are required to ascertain the impact of COVID-19 from the perspective of renal patients and also to validate the classification scheme which was developed in this study. The study's methodology and classification scheme can be used to guide future studies for evaluating the social media outreach performance of renal health support groups.

Avoiding Trouble Ahead: Lessons Learned and Suggestions for Economic Evaluations of COVID-19 Vaccines.

With vaccines for coronavirus disease 2019 (COVID-19) being introduced in countries across the world, policy makers are facing many practical considerations about how best to implement a vaccination programme. The supply of vaccines is insufficient for the global population, so decisions must be made as to which groups are prioritised for any vaccination and when. Furthermore, the aims of vaccination programmes will differ between countries, with some prioritising economic benefits that could stem from the relaxation of non-pharmaceutical interventions and others seeking simply to reduce the number of COVID-19 cases or deaths. This paper aims to share the experiences and lessons learned from conducting economic evaluations in Singapore and Thailand on hypothetical COVID-19 vaccines to provide a basis for other countries to develop their own contextualised economic evaluations, with particular focus on the key uncertainties, technical challenges, and characteristics that modellers should consider in partnership with key stakeholders. Which vaccines, vaccination strategies, and policy responses are most economically beneficial remains uncertain. It is therefore important for all governments to conduct their own analyses to inform local policy responses to COVID-19, including the implementation of COVID-19 vaccines in both the short and the long run. It is essential that such studies are designed, and ideally conducted, before vaccines are introduced so that policy decisions and implementation procedures are not delayed.

Impact of COVID-19 and partial lockdown on access to care, self-management and psychological well-being among people with diabetes: A cross-sectional study.

BACKGROUND: The impact of lockdown measures can be widespread, affecting both clinical and psychosocial aspects of health. This study aims to assess changes in health services access, self-care, behavioural, and psychological impact of COVID-19 and partial lockdown amongst diabetes patients in Singapore. METHODS: We conducted a cross-sectional online survey amongst people with diabetes with the Diabetes Health Profile-18 (DHP-18). Hierarchical regression analyses were performed for each DHP-18 subscale (Psychological Distress, Disinhibited Eating and Barriers to Activity) as dependent variables in separate models. RESULTS: Among 301 respondents, 45.2% were women, 67.1% of Chinese ethnicity, 24.2% were aged 40 to 49 years, 68.4% have Type 2 diabetes and 42.2% on oral medications alone. During the pandemic and the lockdown, nearly all respondents were able to receive care safely from the clinics they attend (94%) and obtain their medications and diabetes equipment and supplies (97%) when needed. Respondents reported less frequent engagement in physical activity (38%), checking of blood pressure (29%) and blood glucose (22%). Previous diagnosis of mental health conditions (ß = 9.33, P = .043), Type 1 diabetes (ß = 12.92, P = .023), number of diabetes-related comorbidities (ß = 3.16, P = .007) and Indian ethnicity (ß = 6.65, P = .034) were associated with higher psychological distress. Comorbidities were associated with higher disinhibited eating (ß = 2.49, P = .014) while ability to reach their doctor despite not going to the clinic is negatively associated with psychological distress (ß = -9.50 P = .002) and barriers to activity (ß = -7.53, P = .007). CONCLUSION: Health services access were minimally affected, but COVID-19 and lockdown had mixed impacts on self-care and management behaviours. Greater clinical care and attention should be provided to people with diabetes with multiple comorbidities and previous mental health disorders during the pandemic and lockdown.

Impact of COVID-19 on the outreach strategy of cancer social service agencies in Singapore: A pre-post analysis with Facebook data.

The Singapore government implemented multiple restrictive measures as the novel coronavirus infection (COVID-19) spread through the community, thereby affecting the support service of cancer-related social service agencies (cancer-SSAs). We are interested to understand how Singapore's cancer-SSAs utilized the social media platform Facebook to overcome the restrictions which were introduced due to COVID-19. Facebook posts from two prominent cancer-SSAs 365 Cancer Prevention Society (365CPS) and Singapore Cancer Society (SCS) between comparable periods in 2019 and 2020 were extracted. These posts were categorized using a classification scheme which was specifically developed by two coders with adequate intercoder reliability. Statistical analyses were performed to determine if there was a significant difference in the frequency of posts between 2019 and 2020, and across three specific periods in 2020. Results indicate that 365CPS appears to have adapted to the pandemic by increasing their posting frequency on Facebook in 2020, but the same was not evident for SCS. However, both SSAs tweaked their social media outreach strategy in line with social distancing measures, publishing posts detailing activities that beneficiaries can participate from home such as healthy recipes and virtual events. SSAs can scale up their efforts to achieve a higher level of health promotion and support for their beneficiaries. Further qualitative studies are required to ascertain the impact of COVID-19 from the perspective of beneficiaries and validate the classification scheme which was developed in this study. The classification scheme could be used in ascertaining the quality of social media communication between public/private support groups and patients.

Impact of the COVID-19 Pandemic on Door-to-Balloon Time for Primary Percutaneous Coronary Intervention　- Results From the Singapore Western STEMI Network.

BACKGROUND: Little is known about the effect of the coronavirus disease 2019 (COVID-19) pandemic and the outbreak response measures on door-to-balloon time (D2B). This study examined both D2B and clinical outcomes of patients with STEMI undergoing primary percutaneous coronary intervention (PPCI).Methods and Results:This was a retrospective study of 303 STEMI patients who presented directly or were transferred to a tertiary hospital in Singapore for PPCI from October 2019 to March 2020. We compared the clinical outcomes of patients admitted before (BOR) and during (DOR) the COVID-19 outbreak response. The study outcomes were in-hospital death, D2B, cardiogenic shock and 30-day readmission. For direct presentations, fewer patients in the DOR group achieved D2B time <90 min compared with the BOR group (71.4% vs. 80.9%, P=0.042). This was more apparent after exclusion of non-system delay cases (DOR 81.6% vs. BOR 95.9%, P=0.006). Prevalence of both out-of-hospital cardiac arrest (9.5% vs. 1.9%, P=0.003) and acute mitral regurgitation (31.6% vs. 17.5%, P=0.006) was higher in the DOR group. Mortality was similar between groups. Multivariable regression showed that longer D2B time was an independent predictor of death (odds ratio 1.005, 95% confidence interval 1.000-1.011, P=0.029). CONCLUSIONS: The COVID-19 pandemic and the outbreak response have had an adverse effect on PPCI service efficiency. The study reinforces the need to focus efforts on shortening D2B time, while maintaining infection control measures.

Use of Health Belief Model-Based Deep Learning Classifiers for COVID-19 Social Media Content to Examine Public Perceptions of Physical Distancing: Model Development and Case Study.

BACKGROUND: Public health authorities have been recommending interventions such as physical distancing and face masks, to curtail the transmission of coronavirus disease (COVID-19) within the community. Public perceptions toward such interventions should be identified to enable public health authorities to effectively address valid concerns. The Health Belief Model (HBM) has been used to characterize user-generated content from social media during previous outbreaks, with the aim of understanding the health behaviors of the public. OBJECTIVE: This study is aimed at developing and evaluating deep learning-based text classification models for classifying social media content posted during the COVID-19 outbreak, using the four key constructs of the HBM. We will specifically focus on content related to the physical distancing interventions put forth by public health authorities. We intend to test the model with a real-world case study. METHODS: The data set for this study was prepared by analyzing Facebook comments that were posted by the public in response to the COVID-19-related posts of three public health authorities: the Ministry of Health of Singapore (MOH), the Centers for Disease Control and Prevention, and Public Health England. The comments made in the context of physical distancing were manually classified with a Yes/No flag for each of the four HBM constructs: perceived severity, perceived susceptibility, perceived barriers, and perceived benefits. Using a curated data set of 16,752 comments, gated recurrent unit-based recurrent neural network models were trained and validated for text classification. Accuracy and binary cross-entropy loss were used to evaluate the model. Specificity, sensitivity, and balanced accuracy were used to evaluate the classification results in the MOH case study. RESULTS: The HBM text classification models achieved mean accuracy rates of 0.92, 0.95, 0.91, and 0.94 for the constructs of perceived susceptibility, perceived severity, perceived benefits, and perceived barriers, respectively. In the case study with MOH Facebook comments, specificity was above 96% for all HBM constructs. Sensitivity was 94.3% and 90.9% for perceived severity and perceived benefits, respectively. In addition, sensitivity was 79.6% and 81.5% for perceived susceptibility and perceived barriers, respectively. The classification models were able to accurately predict trends in the prevalence of the constructs for the time period examined in the case study. CONCLUSIONS: The deep learning-based text classifiers developed in this study help to determine public perceptions toward physical distancing, using the four key constructs of HBM. Health officials can make use of the classification model to characterize the health behaviors of the public through the lens of social media. In future studies, we intend to extend the model to study public perceptions of other important interventions by public health authorities.

Measuring the Outreach Efforts of Public Health Authorities and the Public Response on Facebook During the COVID-19 Pandemic in Early 2020: Cross-Country Comparison.

BACKGROUND: The coronavirus disease (COVID-19) pandemic presents one of the most challenging global crises at the dawn of a new decade. Public health authorities (PHAs) are increasingly adopting the use of social media such as Facebook to rapidly communicate and disseminate pandemic response measures to the public. Understanding of communication strategies across different PHAs and examining the public response on the social media landscapes can help improve practices for disseminating information to the public. OBJECTIVE: This study aims to examine COVID-19-related outreach efforts of PHAs in Singapore, the United States, and England, and the corresponding public response to these outreach efforts on Facebook. METHODS: Posts and comments from the Facebook pages of the Ministry of Health (MOH) in Singapore, the Centers for Disease Control and Prevention (CDC) in the United States, and Public Health England (PHE) in England were extracted from January 1, 2019, to March 18, 2020. Posts published before January 1, 2020, were categorized as pre-COVID-19, while the remaining posts were categorized as peri-COVID-19 posts. COVID-19-related posts were identified and classified into themes. Metrics used for measuring outreach and engagement were frequency, mean posts per day (PPD), mean reactions per post, mean shares per post, and mean comments per post. Responses to the COVID-19 posts were measured using frequency, mean sentiment polarity, positive to negative sentiments ratio (PNSR), and positive to negative emotions ratio (PNER). Toxicity in comments were identified and analyzed using frequency, mean likes per toxic comment, and mean replies per toxic comment. Trend analysis was performed to examine how the metrics varied with key events such as when COVID-19 was declared a pandemic. RESULTS: The MOH published more COVID-19 posts (n=271; mean PPD 5.0) compared to the CDC (n=94; mean PPD 2.2) and PHE (n=45; mean PPD 1.4). The mean number of comments per COVID-19 post was highest for the CDC (mean CPP 255.3) compared to the MOH (mean CPP 15.6) and PHE (mean CPP 12.5). Six major themes were identified, with posts about prevention and safety measures and situation updates being prevalent across the three PHAs. The themes of the MOH's posts were diverse, while the CDC and PHE posts focused on a few themes. Overall, response sentiments for the MOH posts (PNSR 0.94) were more favorable compared to response sentiments for the CDC (PNSR 0.57) and PHE (PNSR 0.55) posts. Toxic comments were rare (0.01%) across all PHAs. CONCLUSIONS: PHAs' extent of Facebook use for outreach purposes during the COVID-19 pandemic varied among the three PHAs, highlighting the strategies and approaches that other PHAs can potentially adopt. Our study showed that social media analysis was capable of providing insights about the communication strategies of PHAs during disease outbreaks.