I finally made headway on contact tracing percentages looking at 3 metrics:
1. What % of new cases is ADH able to link to a source?
2. How successful is contact tracing in breaking the chain of transmission?
3. What % of new cases respond to the survey of places they’ve visited?

As of June 18, about 75% of cases had been linked to a source. Some of these linkages are pretty easy to make. For example, if a new case works at a poultry plant, the assumption is made that the plant is the source.

We should note that linking cases back to a source does not mean that a contact tracer conducted an interview with that person. Contact tracing does not occur for those in correctional or nursing home facilities, and companies often conduct their own contact tracing.

Let’s look at the links as of June 18.

34.7% had contact with a known positive
22.8%, were residents or staff of a nursing home or correctional facility.
14.2% were part of a known or suspected cluster – this is usually going to be a workplace
-->>

1.5% had contact with someone who was under investigation for Covid 19
0.4% were linked through travel to an advisory location

If we take out the 7% under investigation, that leaves 26.4% of the cases unlinked, meaning that the state doesn’t know how that person was infected.

Metric 2: How successful is the state’s contact tracing efforts in breaking the chain of transmission?

The goal of contact tracing is to get ahead of the virus, reaching the contacts of the infected person before those contacts have a chance to spread the virus to others.

Dr. Jennifer Dillaha, epidemiologist with ADH, said the state tries to reach contacts of each new case within 1 to 2 days of testing positive, though she admitted it sometimes takes longer. And, as noted above, large corporations often conduct their own contact tracing.

I asked Dr. Dillaha if she felt ADH was getting ahead of the virus or if the state’s contact tracing was lagging too far behind to be effective.

She said ADH was not having the success it would like to see in getting people to isolate and quarantine in ways that would break that chain of transmission. Her reply was that it isn’t necessarily a timing issue, but a compliance issue.

ADH offers temporary housing and support services to people who live in large households that need to quarantine away from their families to avoid infecting others. Dr. Dillaha said ADH has encountered reluctance from Covid cases to accept the offer.

I can imagine that many new cases with Hispanic heritage might be reluctant to be “rehoused” by a government agency – images from detention camps and ICE officers fill my mind. She said ADH is working to build trust within minority communities and convey it doesn’t work with ICE.

The response seemed to be aligned with others from ADH, in which the focus is placed outward on lack of compliance or awareness by the patient – instead of looking inward at how the department can create systems and relationships that lead to compliance and active participation

Metric 3: What percentage of new cases respond to the survey of places they’ve visited in the 14 days prior to infection?

The governor and Dr. Smith use this survey to show that businesses related to phased re-openings aren’t at the root of Arkansas’ spike in cases. However, it’s important to know if the state is receiving a statistically significant number of responses when it makes that claim.

The long and short is that we don't know how many people are responding to the survey. Of the types of places listed in the survey, most people consistently visited a church, a doctor’s office, or a restaurant in the two weeks prior to their infection.

If only 100 people responded to the survey and all 100 of them had been to a church or if 2,000 people responded and only 100 of them had been to a church, that’s still 2% of active cases. But these present two very, very different scenarios. We just don't know the response rate.

The governor’s comments that he was told poultry workers were being infected outside of the work environment would seem to contradict the case he’s making with the survey response data.

If workers were being infected in the community at large, and the survey data the state is using is a relevant sample size, wouldn’t we see that transmission in the survey data?

To wrap up, I made some headway but still have many unanswered questions. I'm adding a new chart tonight that will track, a couple of times a week, the active cases and survey responses a bit better.

Full article below: https://arkansascovid.com/2020/06/deep-dive-into-3-contact-tracing-metrics/