With 1 in 40 Coloradans estimated to be infected, the impact of the Thanksgiving holiday won’t be seen for at least another week

REMOTE (Dec. 5, 2020): Colorado Department of Public Health and Environment (CDPHE) and the Colorado School of Public Health (ColoradoSPH) have released an updated modeling report.

Key findings from the modeling report:

• An estimated 1 of 40 Coloradans are infectious, the highest prevalence to date. 
• The effective reproductive number is approximately 1.2, lower than last week but indicating continued growth of the epidemic. Infections are continuing to rise, albeit at a slower rate than last week. Estimated transmission control has increased from 66% to 71% since last week.
• These estimates reflect infections occurring before Thanksgiving and do not yet reflect any impact of Thanksgiving travel or gatherings.
• Over the next two weeks, COVID-19 hospital demand will likely reach or exceed 2,120 beds, including 6,65 ICU beds. Because of increased contacts over Thanksgiving, demand could be greater.
• On the current trajectory, COVID-19 hospital and ICU demand may stay within surge capacity estimates over the next four weeks, even with modest increases in contacts over the Thanksgiving holiday. But further reductions in transmission control, associated with increases in contacts over the holidays could lead to hospital and ICU demand above surge capacity estimates.

ColoradoSPH assembled the expert modeling team, which relies on expertise from modeling scientists at ColoradoSPH and the University of Colorado School of Medicine at the CU Anschutz Medical Campus, as well as experts from the University of Colorado Boulder, the University of Colorado Denver, and Colorado State University. The models produced by the expert group are local, meaning models are based on Colorado data and assumptions apply specifically to Colorado, not national assumptions.

All previous modeling reports are available on the Colorado School of Public Health’s COVID-19 website. 

Continue to stay up to date by visiting covid19.colorado.gov.

The longer the state remains on current trajectory, the greater the change in transmission control needed 

REMOTE, (Oct. , 2020): The Colorado Department of Public Health and Environment (CDPHE) and the Colorado School of Public Health released an updated modeling report showing that hospitalizations from SARS-CoV-2 continue to increase rapidly across the state. 

On the current epidemic curve, Colorado will likely exceed the April peak in hospitalizations within two weeks. If the epidemic curve is not bent, Colorado could surpass intensive care unit (ICU) capacity in January. If contacts increase over the holidays (for example, due to gatherings between multiple households), ICU capacity could be exceeded in December. As announced by the state last week, gatherings should be limited to no more than 10 people from a maximum of two households. 

“There is a small window to improve transmission control over the next few weeks,” said Dr. Jonathan Samet, dean of the Colorado School of Public Health. “To limit increasing infections and avoid peaks that could strain healthcare capacity over the next three months, a substantial increase in transmission control is needed.”

The latest modeling provides projections based on COVID-19 hospital census data through October 26, 2020. The models are based on Colorado data and assumptions based on the current state of the science.

Key findings from the report:

• Hospitalizations continue to rapidly increase. On the current trajectory of the epidemic curve, Colorado will probably hit its highest number of hospitalizations due to COVID-19 by November 10, and the limits of ICU capacity may be reached in early to mid-January. If infections increase over the holidays because of gatherings and other reasons for increased contacts among people, ICU capacity could be exceeded in December. 
  
• The effective reproductive number is approximately 1.6 (with statistical uncertainty ranging from 1.47 to 1.70 at the 95% confidence level).
  
• Approximately 1 in 219 Coloradans are currently infectious (compared to 1 in 292 last week). This implies that the probability of encountering an infected person in the population is higher than it was at any point this summer. This estimate is generated from the model and assumes that not all infectious residents are captured by state surveillance systems.
  
• The estimated level of transmission control is currently 65% (for the period of September 28-October 13). Transmission control levels under about 79% will lead to increasing infections and an effective reproductive number greater than 1; and if contact rates are reduced and transmission control is above 79%, infections will decrease.
  
• Using an extended modeling approach based on case data, the modeling team estimates transmission control varies significantly by age group, with significant decreases in control levels among all ages over the last month. Individuals aged 20-39 have the highest infectious contact rates (transmission control = 60%), and contact rates have increased among individuals over 65 over recent weeks (transmission control = 76%).

The Colorado School of Public Health (ColoradoSPH) assembled the expert group that works with the state on modeling projections. The group includes modeling scientists at the ColoradoSPH and the University of Colorado School of Medicine at the CU Anschutz Medical Campus, as well as experts from the University of Colorado Boulder, University of Colorado Denver, and Colorado State University. 

All previous modeling reports are available on the Colorado School of Public Health’s COVID-19 website. 

The Colorado modeling team began using a new “transmission control” indicator in mid-October to describe the collective impact of all policies and behaviors on the spread of SARS-CoV-2. Transmission control captures ALL behavioral and policy changes in response to the SARS-CoV-2 pandemic including mask wearing, physical distancing, improved ventilation, working from home, contact tracing (including both isolation and quarantine), moving activities outside, and any seasonal impact. This approach has the advantage of requiring fewer assumptions and increasing accuracy for the Colorado model. In technical terms, the transmission control parameter describes the percent decrease in effective contacts between infected and susceptible individuals compared to pre-pandemic behavior. 

The state will continue to review data and model findings as the pandemic continues to inform policy decisions.  Continue to stay up to date by visiting covid19.colorado.gov.

REMOTE, Sept. 16: The Colorado Department of Public Health and Environment and Colorado School of Public Health released a new modeling report that indicates the spread of SARS-CoV-2 was reduced for much of July and August, leading to declines in hospitalizations and infections. However, in recent weeks, the estimated effective reproductive number has increased while hospitalizations have plateaued.

The latest modeling provides projections based on COVID-19 hospital census data to characterize the current status of the COVID-19 epidemic in Colorado, and the collective impact of efforts to reduce the spread of the SARS-CoV-2 virus. It also provides projections based on various policy scenarios around physical distancing, mask-wearing, and case detection, and containment. The models are based on Colorado data and assumptions based on the current state of the science.

Key findings from the report:

• COVID-19 hospitalizations have been decreasing since mid-July. This decline has slowed and hospitalizations appear to be at a plateau.
• The estimated effective reproductive number has increased following an upward trend over recent weeks. The current effective reproductive number is estimated to be between 0.93 and 1.14. Due to lags between infection and hospitalizations, this reflects transmission occurring through late-August (approximately August 25).
• The model-estimated number of infectious individuals in Colorado remains relatively low. The model-estimated rate of new infections as well as reported cases had been declining since mid-July, but the decline now appears to be leveling off.
• Our current estimate of physical distancing is between 65% and 72% (65% for the week of August 17-25 and 72% during the three-week period August 3-25). If physical distancing remains at 65%, we will begin to see gradual growth in cases and/or hospital demand.

The Colorado School of Public Health (ColoradoSPH) assembled the expert group that works with the state on modeling projections. The group includes modeling scientists at the ColoradoSPH and the University of Colorado School of Medicine at the CU Anschutz Medical Campus, as well as experts from the University of Colorado Boulder, University of Colorado Denver, and Colorado State University. 

All modeling reports are available on the Colorado School of Public Health’s COVID-19 website. 

The state will continue to review data and model findings as the pandemic continues to inform future policy decisions. 

DENVER (April 5, 2020): The state today released additional COVID-19 modeling data to the public. Governor Jared Polis first provided an in-depth analysis of the data during a press conference on March 27. The state will continue to review data as it evolves to inform future policy decisions.

The modeling data was produced by an expert team that the Colorado School of Public Health (ColoradoSPH) assembled to assist the Colorado Department of Public Health and Environment (CDPHE) in understanding the potential course of the pandemic in Colorado. 

Drawing on expertise from the ColoradoSPH and the University of Colorado School of Medicine at the CU Anschutz Medical Campus, and the University of Colorado in Denver and Boulder, a team of volunteer experts modeled the pandemic using approaches tailored to Colorado, updating the modeling as the disease continues to spread in the state.

The team uses a fundamental approach: the SEIR model. The basics of the models are intuitive: prior to infection, individuals are susceptible (S) and once exposed (E) and infected (I) they are contagious, whether symptomatic or asymptomatic; those infected may recover and become resistant (R) or become sufficiently ill to need hospitalization and possibly critical care and to die. This standard model is thus abbreviated as the SEIR model. 

Another important number in the model is the reproductive number (R0), that is, the average number of new cases generated per infected person at the beginning of the outbreak. (The Governor has referred to this number as the “R naught.”) If that number exceeds one then the infection will spread. Various figures for R0 have been reported at the start of the COVID-19 pandemic from different parts of the world, ranging from about 2 to 5. The reproductive number depends, in part, on the frequency of contacts between infected and uninfected individuals. The goal of social distancing, which we are all experiencing now, is to reduce these contacts and lower the reproductive number with a target value below one so that contagion ends. A critical question is: How much social distancing is needed to control the epidemic, and how long does it need to be in place?

The tables below provide two sets of numbers provided to CDPHE and the Governor’s Office. The numbers provided are for critical indicators and show the variation by the value of R0 at the beginning of the outbreak and the effectiveness of social distancing at reducing the contact rate, ranging from none to an 80% reduction. When these numbers were calculated, the team found that the R0 value for Colorado was likely above 3. To capture the uncertainty in R0 and the effectiveness of social distancing, estimates are provided for a range of values.

Model R0 3.5 4.0 with SD 0 to 80

Model R0 3.0 4.0 SD 0 to 60

The Colorado modeling team has continued to refine its models as the data accumulate on the course of the pandemic. There are other models that provide estimates for Colorado, but those models are not as closely linked to the state’s data.

Continue to stay up to date by visiting covid19.colorado.gov.