Tuesday, June 16, 2020

Incremental Risk of COVID-19 Exposure

My firm (AnalyticsIQ) is focused on assessing complex data to provide meaningful insights to our clients.  Hopefully, that gives us a little credibility in analyzing the voluminous COVID-19 transmission data.  We believe that there is still substantial risk of exposure (varying by location) for those who do not practice social distancing and/or wear some form of PPE.

One question that we have not seen addressed is:  “What is the risk from exposure to one incremental person?”.  We believe that the risk of exposure to one additional person is infinitesimal (< 0.001% per contact).

Here are the assumptions behind this assessment:

·       That the rate of infection (defined as COVID-19 positive and needing some level of medical care) over the course of a season (10 weeks) is around 3%
o   Assumes a lack of social distancing and/or wearing of PPE
o   For a frame of reference, this would result in about 10 million serious cases over a 10 week period in the USA
·       During ‘normal’ times an individual living in a large city would come in contact with about 300 individuals per week (some of whom would be actively infected)

Clearly, there is a difference in the type of contact – kissing someone on the lips has a higher risk of infection compared to passing within 5 feet of a stranger.  That said, the historical infection rate assumptions contain a mixture of all types of contacts.

It is safe to assume that adding social distancing and PPE will reduce the risk of infection, everything else being equal – meaning that assessing infection rates alone is an overly pessimistic view.  But…it is likely that there is a linear relationship between an individual’s ‘circle’ and their ultimate risk.  Reducing average contacts by 90% (to around 30 new contacts per week – or 300 during a season) will reduce the infection risk by an around 90%.  


While there is incremental risk to adding one seemingly healthy individual, it is minimal:  around 0.001% by the most pessimistic view.   While the degree of risk an individual is willing to take on is a personal choice, adding individual contacts has low risk. 

Wednesday, June 3, 2020

Take a Stand

We are all aware of the recent killings of unarmed black men at the hands of those who are paid to protect us:  police.  While it would be convenient to attribute these killings to the actions of rogue officers, the reality is that these events are reflective of a broken system.   The reaction to this injustice has been massive and is no doubt representative of all walks of life.  I am writing to tell you that AnalyticsIQ stands with those seeking peaceful change in the USA and we will no longer be silent in the face of perpetual racial injustice.

I have always said that I would have marched with Dr. Martin Luther King, Jr. had I been old enough (I was three when he was assassinated).  The people who marched in the deep southern US risked their lives to stand up to profound discrimination towards people of color.  They are heroes.  Some today are risking arrest and injury to peacefully demand change.

There are some things (actually many things) that are more important than money and financial success.  High up that list is taking a principled stand against discrimination of individuals based on their skin color, where the were born, their gender or gender identity and who they choose to love.  AnalyticsIQ demands that our employees live up to this standard, and we fully expect our vendors and customers to do the same.  I am happy to tell organizations that stand in the way of racial justice that we do not want their money.

AnalyticsIQ (along with me as an individual) will be making a financial commitment to support organizations that are fighting for racial justice.  The question for you:  do you want to be on the right side of history?

Thursday, March 26, 2020

COVID-19: The Numbers

It does appear that Italy has turned the corner in terms of new infections and related deaths:  both are still growing, but at last the rate of growth is declining.   The narrative that the USA is not an analog to Italy appears to be untrue: USA deaths are on track to surpass Italy’s experience by early next week.


ItalyUSA
 InfectionsDeaths InfectionsDeaths
  # Cases  NetNew  New%  Deaths  New%   # Cases  NetNew  New%  Deaths  New% 
3/5        3,858          769 25%           148 38%3/16    4,663         983 27%        86 26%
3/6        4,636          778 20%           197 33%3/17    6,411      1,748 37%      109 27%
3/7        5,883       1,247 27%           233 18%3/18    9,259      2,848 44%      150 38%
3/8        7,375       1,492 25%           366 57%3/19  13,789      4,530 49%      207 38%
3/9        9,172       1,797 24%           436 19%3/20  19,383      5,594 41%      256 24%
3/10      10,149          977 11%           631 45%3/21  24,207      4,824 25%      302 18%
3/11      12,462       2,313 23%           827 31%3/22  33,546      9,339 39%      419 39%
3/12      15,113       2,651 21%        1,016 23%3/23  43,734    10,188 30%      553 32%
3/13      17,660       2,547 17%        1,266 25%3/24  54,856    11,122 25%      780 41%
3/14      21,157       3,497 20%        1,441 14%3/25  68,211    13,355 24%   1,027 32%
3/15      24,747       3,590 17%        1,809 26%3/26    
3/16      27,980       3,233 13%        2,158 19%3/27    
3/17      31,506       3,526 13%        2,503 16%3/28    
3/18      35,713       4,207 13%        2,978 19%3/29    
3/19      41,035       5,322 15%        3,405 14%3/30    
3/20      47,021       5,986 15%        4,032 18%3/31    
3/21      53,578       6,557 14%        4,825 20%4/1    
3/22      59,138       5,560 10%        5,476 13%4/2    
3/23      63,927       4,789 8%        6,077 11%4/3    
3/24      69,176       5,249 8%        6,820 12%4/4    
3/25      74,386       5,210 8%        7,503 10%4/5     
Infections: # Cases = Total  cumulative infections through reported date.
Infections:  New% = Daily growth rate in reported new infections.
Deaths: # Cases = Total  cumulative deaths through reported date.
Deaths:  New% = Daily growth rate in reported new deaths.