A Bayesian longitudinal trend analysis of count data with Gaussian processes

Background The COVID-19 pandemic has had a profound global impact on governments, health care systems, economies, and populations around the world. Within the East Asia and Pacific region, some countries have mitigated the spread of the novel coronavirus effectively and largely avoided severe negative consequences, while others still struggle with containment. As the second wave reaches East Asia and the Pacific, it becomes more evident that additional SARS-CoV-2 surveillance is needed to track recent shifts, rates of increase, and persistence associated with the pandemic. Objective The goal of this study is to provide advanced surveillance metrics for COVID-19 transmission that account for speed, acceleration, jerk, persistence, and weekly shifts, to better understand country risk for explosive growth and those countries who are managing the pandemic successfully. Existing surveillance coupled with our dynamic metrics of transmission will inform health policy to control the COVID-19 pandemic until an effective vaccine is developed. We provide novel indicators to measure disease transmission. Methods Using a longitudinal trend analysis study design, we extracted 330 days of COVID-19 data from public health registries. We used an empirical difference equation to measure the daily number of cases in East Asia and the Pacific as a function of the prior number of cases, the level of testing, and weekly shift variables based on a dynamic panel model that was estimated using the generalized method of moments approach by implementing the Arellano-Bond estimator in R. Results The standard surveillance metrics for Indonesia, the Philippines, and Myanmar were concerning as they had the largest new caseloads at 4301, 2588, and 1387, respectively. When looking at the acceleration of new COVID-19 infections, we found that French Polynesia, Malaysia, and the Philippines had rates at 3.17, 0.22, and 0.06 per 100,000. These three countries also ranked highest in terms of jerk at 15.45, 0.10, and 0.04, respectively. Conclusions Two of the most populous countries in East Asia and the Pacific, Indonesia and the Philippines, have alarming surveillance metrics. These two countries rank highest in new infections in the region. The highest rates of speed, acceleration, and positive upwards jerk belong to French Polynesia, Malaysia, and the Philippines, and may result in explosive growth. While all countries in East Asia and the Pacific need to be cautious about reopening their countries since outbreaks are likely to occur in the second wave of COVID-19, the country of greatest concern is the Philippines. Based on standard and enhanced surveillance, the Philippines has not gained control of the COVID-19 epidemic, which is particularly troubling because the country ranks 4th in population in the region. Without extreme and rigid social distancing, quarantines, hygiene, and masking to reverse trends, the Philippines will remain on the global top 5 list of worst COVID-19 outbreaks resulting in high morbidity and mortality. The second wave will only exacerbate existing conditions and increase COVID-19 transmissions.

Download Full-text

Middle East and North Africa SARS-CoV-2 Surveillance: A Longitudinal Trend Analysis (Preprint)

10.2196/preprints.25830 ◽

2020 ◽

Author(s):

Lori Post ◽

Emily Marogi ◽

Charles B Moss ◽

Robert Leo Murphy ◽

Michael G Ison ◽

...

Keyword(s):

Public Health ◽

Saudi Arabia ◽

Middle East ◽

Trend Analysis ◽

North Africa ◽

Public Health Surveillance ◽

Generalized Method Of Moments ◽

Health Surveillance ◽

Infection Rates ◽

Longitudinal Trend

BACKGROUND The COVID-19 global pandemic has disrupted the lives of millions and forced countries to devise public health policies to reduce the pace of transmission. In the Middle East and North Africa, falling oil prices, disparities in wealth and public health infrastructure, and large refugee populations have significantly increased the COVID-19 disease burden. In light of these exacerbating factors, public health surveillance is particularly necessary to help leaders understand and implement effective disease control policies to reduce Sars-CoV-2 persistence and transmission. OBJECTIVE The goal of this study is to provide advanced surveillance metrics, in combination with traditional surveillance, for COVID-19 transmission that account for weekly shifts in the pandemic speed, acceleration, jerk and persistence, to better understand country risk for explosive growth and to better inform those who are managing the pandemic. Existing surveillance coupled with our dynamic metrics of transmission will inform health policy to control the COVID-19 pandemic until an effective vaccine is developed. METHODS Using a longitudinal trend analysis study design, we extracted 30 days of COVID-19 data from public health registries. We use an empirical difference equation to measure the daily number of cases in the Middle East and North Africa as a function of the prior number of cases, the level of testing, and weekly shift variables based on a dynamic panel data model that was estimated using the generalized method of moments (GMM) approach by implementing the Arellano-Bond estimator in R. RESULTS The regression Wald statistic is significant (χ^2 (5)=859.5, P<.001). The Sargan test is not significant, failing to reject the validity of over identifying restrictions (χ^2 (294)= 16 P=.99). Countries with the highest cumulative caseload of the novel coronavirus include Iran, Iraq, Saudi Arabia, and Israel with 530,380, 426,634, 342,202, and 303,109 cases respectively. Many of the smaller countries in MENA have higher infection rates than those countries with the highest caseloads. Oman has 33.3 new infections per 100,000 population while Bahrain has 12.1, Libya has 14, and Lebanon has 14.6. In order of most to least number of cumulative deaths since January 2020, Iran, Iraq, Egypt, and Saudi Arabia have 30,375, 10,254, 6,120, and 5,185 respectively. Israel, Bahrain, Lebanon, and Oman had the highest rates of COVID-19 persistence which are the number of new infections statistically related to new infections 7 days ago. Bahrain had positive speed, acceleration and jerk signaling the potential for explosive growth. CONCLUSIONS Static and dynamic public health surveillance metrics provide a more complete picture of pandemic progression across countries in MENA. Static measures capture data at a given point in time such as infection rates and death rates. By including speed, acceleration, jerk, and 7-day persistence, public health officials may design policy with an eye to the future. Iran, Iraq, Saudi Arabia, and Israel all demonstrated the highest rate of infections, acceleration, jerk, and 7-day persistence rates prompting public health leaders to increase prevention efforts. CLINICALTRIAL

Download Full-text

East Asia and the Pacific Surveillance of SARS-CoV-2 Wave Two: Longitudinal Trend Analysis (Preprint)

10.2196/preprints.25454 ◽

2020 ◽

Author(s):

Lori Ann Post ◽

Jasmine S Lin ◽

Charles B Moss ◽

Robert Leo Murphy ◽

Micahel G Ison ◽

...

Keyword(s):

East Asia ◽

Trend Analysis ◽

Generalized Method Of Moments ◽

French Polynesia ◽

The Philippines ◽

High Morbidity ◽

Negative Consequences ◽

Longitudinal Trend ◽

The Pacific ◽

Second Wave

BACKGROUND The COVID-19 pandemic has had a profound global impact on governments, healthcare systems, economies, and populations around the world. Within the East Asia and Pacific region, some countries have mitigated the spread of the novel coronavirus effectively and largely avoided severe negative consequences, while others still struggle with containment. As the second wave reaches East Asia and the Pacific, it becomes more obvious that additional SARS-CoV-2 surveillance is needed to track recent shifts in the pandemic, rates of increase, and persistence. OBJECTIVE The goal of this study is to provide advanced surveillance metrics for COVID-19 transmission that account for speed, acceleration, jerk,persistence, and weekly shifts in the pandemic, to better understand country risk for explosive growth and those who are managing the pandemic successfully. Existing surveillance coupled with our dynamic metrics of transmission will inform health policy to control the COVID-19 pandemic until an effective vaccine is developed. We provide novel indicators to measure the transmission of disease METHODS Using a longitudinal trend analysis study design, we extracted 330 days of COVID data from public health registries. We use an empirical difference equation to measure the daily number of cases in East Asia and the Pacific as a function of the prior number of cases, the level of testing, and weekly shift variables based on a dynamic panel model that was estimated using the generalized method of moments (GMM) approach by implementing the Arellano-Bond estimator in R. RESULTS Based on standard surveillance metrics, Indonesia, Philippines, and Myanmar are concerning because they have the largest new caseloads at 4,301, 2,588, and 1,387 respectively. However, when looking at the acceleration of new COVID-19 infections, we find that French Polynesia, Malaysia, and Philippines have the rates at 3.17, .22, and .06 per 100,000. These three countries also rank highest in jerk at 15.45, .10 and .04 respectively. CONCLUSIONS Two of the most populous countries in East Asia and the Pacific, Indonesia and the Philippines, have alarming surveillance metrics. These two countries rank highest in new infections in the region. The highest rates of speed, acceleration and positive upwards jerk belong to French Polynesia, Malaysia and the Philippines. Positive rates of speed, acceleration and upwards jerk are more likely to result in explosive growth. While all countries in East Asia and Pacific need to be cautious in regards to opening their countries because outbreaks are likely to occur in the second wave of COVID-19, the country of greatest concern remains the Philippines. Based on standard and enhanced surveillance, the Philippines has not gained control of the COVID-19 epidemic, which is particularly troubling because the country ranks 4th in population in the region. Without extreme and rigid social distancing, quarantines, hygiene, and masking to reverse trends, the Philippines will remain on the global top 5 list of worst COVID-19 outbreaks resulting in high morbidity and mortality. The second wave will only exacerbate existing conditions and increase COVID-19 transmissions. CLINICALTRIAL NA

Download Full-text