Oregon

We noticed that this footnote also mentioned that results are deduplicated based on medical record, accession number, facility and result. Could you clarify the following questions, which will help us determine whether your count of tests belongs in our “specimens” our “encounters” field: Are accession numbers attached to individual specimens, or sets of specimens? Would an individual with two separate specimens collected for testing at one site on one day (e.g. a nasal and an oral sample that are not pooled) be counted once or twice?

Accession numbers are attached to individual specimens. We deduplicate on accession and patient ids such that two different specimens on the same person would result in 2 different ELRs.

We noticed a footnote on the county-level data noting that mentioned “non-antibody and non-‘info-only’” tests are de-duplicated to reach the final results. Our interpretation is that this means antibody and info-only tests are not included in the counts—not that they are included and not deduplicated. Could you confirm if this interpretation is correct?

We’re not sure what “info-only” testing means in this context—does this match the CDC definition of surveillance testing? (Testing meant to monitor population-level outbreaks and characterize incidence or prevalence, without using the information to diagnose individuals or screen individuals for COVID?)

Antibody and info-only tests are not included in the counts. “info only” ELRs provide case level information. For example, some hospitals report if a case is hospitalized or not through their ELRs. Since this ELR does not contain test result information, it would be classified as a “info only” ELR.

We noticed that we can access a daily history of the specimens on your dashboard—tank you so much for posting this! Are there plans underway to make this data downloadable in a machine-readable format ?

Yes, there are plans to make data downloadable in a machine-readable format. Summary table already exists (downloadable): https://public.tableau.com/profile/oregon.health.authority.covid.19#!/vizhome/OregonHealthAuthorityCOVID-19SummaryTable_15889676399110/OregonsCOVID-19ELRsSummaryTable

I’m writing from the COVID Tracking Project to offer our gratitude for the upcoming data release of Oregon’s COVID-19 testing figures using a metric that captures serial testing, and to make two high-priority (on our side) requests. The gratitude: We know these metric changeovers are difficult and time-consuming and are happening during a crisis, and we’re tremendously thankful to all of you who are working so hard. What we need to get Oregon’s numbers to update on the Johns Hopkins/New York Times/other trackers: Now that Oregon is releasing an “all tests” testing metric, we would love to get this to the trackers that calculate test positivity for US states so that Oregon’s figures are comparable with those of its neighbors. To do this, we need a complete, publicly available time series for the new testing metric. Once we have that, we can back-fill your testing data into the API field these organizations use to calculate test positivity and other testing measures. The other request: Some states have removed their unique people tested metric when they release a tests-based metric, which prevents state residents and the press from understanding testing utilization in those states. We would love to see the unique people metric carried forward as you release the new testing metric. (And as an Oregonian, I would personally love to see that as well.) Please let me know if you have any questions—our Data Quality team is scrambling to get ready to add this data, and again, many thanks from all 300+ of us at the COVID Tracking Project.

We anticipate that the Tableau dashboard will be fully updated by 12/21. This will include a **complete, publicly available time series. Unfortunately, we are unable to continue tracking and reporting our person-based metric. **It has placed considerable strain on our communicable disease database, which was not designed to house more than one million electronic laboratory results. Both antigen and PCR tests are included in our former and current metrics.

We are currently interpreting Total Tested as referring to the total number of individuals tested. Is this correct? If so, do you have plans to release the total number of tests conducted?

Do you have plans to release a full historical time-series of all testing and results? We have been capturing the total test results in Total Tested since 5/13/20. Do you plan to release a historical time series of that number? Will this historical data set be available consistently and in a machine readable format like (i.e. csv or json)?

Other states provide their COVID-19 testing and outcomes in machine readable format, but as far as we are aware Oregon Health Authority does not at this time. Do you have plans to release this data in a machine readable format (i.e. csv or json)?

Do you perform any deduplication to reach the Total Tested? If so, how? (i.e. instances swabbed per day/week; recording only one positive test per person, but multiple negative tests for the same person; etc.)? When reporting testing results, if the same person gets different test results on different days, would you report raw results as part of the daily positives or do you apply any kind of logic such as reporting only the first negative or positive result per individual?

We’re also interested in testing data in units of “testing encounters”, which can be defined as “the number of unique people who have been tested per day.” Under this metric, if a person is being tested once today and once again the following week, that counts as two tests, however, if the same person has two samples tested at a testing center visit, then that only counts as one testing encounter. Are you willing to or are you planning to release testing encounters? If so, will you be able to provide historical data consistently under a machine readable format (i.e. csv or json)?

OR emailed us timestamps for specific questions and answers in their recent Facebook video session.

Many thanks to everyone who joined us on Facebook for our live Q&A on COVID-19 data. If you missed it, you can still view the video here: http://ow.ly/rKza50AzuqW For your convenience, we've included the questions that were answered and their timestamps below. 13:24 – How many of the new cases reported each day are retests of people who already have COVID-19? 14:12 – Why don’t the case counts add up from one day to another? 16:04 – What happens if a presumptive case tests negative? 16:50 – Is it possible to know if someone dies of COVID-19 and not just with it? 18:05 – Why has there been a sharp increase in the percentage of cases with an unknown hospitalization status? 19:00 – What about HIPAA laws that protect our information? 20:20 – Does OHA have data on active and recovered cases? 21:30 – Is there data on the underlying conditions of people who die or have COVID-19? 22:10 – Will the Trump Administration’s new mandate that hospitals report COVID-19 numbers to HHS instead of CDC impact how OHA gets its numbers? 24:25 – What about false positives? Are those results skewing the data? 25:00 – How many of the presumptive cases turn out to be negative and are those numbers subtracted from the total? 26:10 – Is Oregon using antigen or pooled testing? 28:24 – Can you provide an overview of how Oregon is fairing in comparison to other states? 30:37 – Is there any data on where people are getting COVID-19?

Answers: 13:24 - OR is reporting people, not specimens 14:12 - May find out one person is from a different county, presumptive cases may be changed to confirmed 16:04 - Presumptive cases who test negative remain presumptive cases. If they test positive, they're moved to confirmed 16:50 - "dying of" COVID is classified as a "COVID related death" (people who die within 60 days of diagnosis or doctor lists COVID-19 as cause of death) ; people are tested posthumously ***no explicit statement on how these are categorized 18:05 - Needs to investigate each case and takes time to identify who's hospitalized or not 20:20 - Yes 21:30 - Mentions whether or not a a person who died had underlying conditions in daily press releases, but there's no specification of what the condition was to protect privacy -- there's no dataset 22:10 - HHS rule doesn't impact how OHA gets numbers, OHA gets numbers directly 24:25 - Not that worried about false positives skewing the data 25:00 - Presumptive cases are not subtracted 26:10 - Not currently using pooled testing... I think the doctor mixed up "antigen" with "antibody" ?? (but honestly, same) @28:11, he seemingly suggests that antigen tests are used for diagnoses

Is Oregon using pooled testing?

Pooled testing is not currently used by Oregon labs; may use in the future

Does the Oregon testing data include antigen diagnostic test results?

No, we don’t do antigen testing here in Oregon.

OR is currently only releasing aggregate congregate settings data (including long-term care facilities, group homes, prisons, shelters, etc.). Is it possible to get the case numbers in each of these types of settings separately? Or to pull out long-term care facility data specifically?

We are not breaking out the data by congregate setting type. (Note: follow up question asked about whether or not OR plans to break out this data in the future.)

Do the reported death counts include both lab confirmed cases and probable cases, or only lab confirmed cases?

Deaths include those among both lab-confirmed and presumed cases.