Epidemiological trends

We estimate the time-varying reproductive number, R_Ra transmission measure based on data from incident emergency department visits. Epidemiological trend It is selected By estimating the probability of this R_R Greater than 1 (map below). estimated R_R Values ​​above 1 indicate epidemic growth.

The second figure below shows the estimate R_Rand the uncertainty period is from September 25, 2024 through November 19, 2024 for the United States and each state reported. (Click on the map to view data for a specific state). while R_R It tells us whether the number of infections is likely to grow or decline, it does not reflect the burden of disease. View a summary of key data for COVID-19, influenza, and respiratory syncytial virus.

Covid-19

As of November 19, 2024, our estimates are that coronavirus (COVID-19) infections are increasing or likely to grow in 8 states, declining or likely to decrease in 15 states, and unchanged in 25 states.

flu

As of November 19, 2024, our estimates are that influenza infections are increasing or likely to grow in 28 states, declining or likely to decrease in one state, and unchanged in 12 states.

Simultaneous translation R_R

• R_R It is a data-driven measure of disease transmission. R_R It is appreciated in history R of the average number of new infections caused by each infectious person. R_R Accounts for current population susceptibility, public health interventions, and behavior.
• R_R >1 indicates that infections are increasing because each infected person, on average, causes more than one new infection R_R <1 indicates that infections are declining.
• R_R It could be Leading indicator From an increase or decrease in cases, hospitalizations, or deaths, because transmission of infection occurs before the case is confirmed, hospitalization, or death.
• The uncertainty range for each R_R The estimate determines the probability of increased infection. For example, if 75% of the uncertainty range lies above 1, there is a 75% chance that infection will grow in that location.
• When the data is sparse, the model is used for construction R_R Estimates will tend to generate estimates closer to 1 with broad, credible time periods, reflecting uncertainty in the true epidemiological trend over these time periods.

Caveats and limitations

• R_R Estimates are sensitive to related assumptions Intergenerational interval distribution.
• R_R Estimates may be overestimated or underestimated if the proportion of injuries resulting in emergency department visits changes suddenly. These estimates can be affected by shifts in clinical severity, increased or decreased use of clinical tests, or changes in reporting.

Knock

R_{R It is defined as the average number of new infections caused by each infected person at a given time, T. when RR > 1, the infection grows, and how long RR <1, infections are declining. The color classes in the above maps were determined by estimating the possible distribution RR Values ​​based on observed emergency department visit data and model assumptions (formally, a “credible interval”). Then we calculate the proportion of that reliable interval where RR > 1. Credible intervals are determined using the EpiNow2 package, which uses a Bayesian model to estimate RRWith delay adjustment and reporting effects.}

• If >90% of the distribution is the credible interval R_R > 1, the infection is growing
• If 76%-90% of the distribution is the credible interval R_R > 1, the infection is likely to grow
• If 26%-75% of the distribution is the reliable interval R_R > 1, the infection does not change (in this case, the reliable interval extends across 1, and contains a mixture of values ​​above and below 1.)
• If 10%-25% of the distribution is the reliable interval R_R > 1, infection rates are likely to decline; This equates to 75%-90% of the credible interval R_R ≥ 1
• If <10% of the distribution is the reliable interval R_R >1, infections are declining; This equates to >90% of the credible interval R_R ≥ 1
• Data used in estimation R_R It is updated frequently, and the numbers initially reported may be revised later. We manually review the data weekly and occasionally exclude implausible outliers, but we can still estimate them R_R.
• R_R Not estimated for states when: 1. fewer than 10 emergency department visits for COVID-19 were reported in each of the previous 2 weeks, 2. anomalies were detected in the reported values, and 3. The model did not pass the reliability checks.

R_R Estimates are derived from daily numbers of new emergency department visits for coronavirus (COVID-19) reported by National Syndrome Surveillance Program. this R_R: behind the model The article provides a more in-depth overview of the modeling approach used in estimation R_Rand strategies the CDC uses to verify the accuracy of estimates.

To estimate R_{R,We fit hypothetical models to the data using R packages EpiNow2, epinowcastOr using STAN models developed by the CDC for outbreak forecasting and analyses. the next Best practicesThese models adjust for lags from infection to surveillance, incomplete observation of recent infection events, and the effects of reporting over the course of a week, as well as the uncertainty of all these adjustments.}

Glossary of terms

• Generation interval: The time interval between infection times for the infected pair; That is, the difference in the time at which an individual (person j) became infected (person i) and the time during which this infectious agent (person i) became infected.
• Main indicator: A variable that provides an early indication of future trends in the outbreak, e.g. R_{RThis metric estimates the number of infections caused by an infected person in near real time.}
• Lagging indicator: A variable that provides a lagging indicator of future outbreak trends, for example, COVID-19 deaths, as this outcome occurs after cases occur.