Charts are created automatically from data that I typically update around early afternoon of each day from official Oregon Health Authority daily updates.

Charts come from 3 cloud-based spreadsheets and may take some time to load.

I am not an authority - these charts should be treated as "FOR ENTERTAINMENT PURPOSES ONLY". Please assume I am an idiot and that I have cherry picked all data and everything you see here is either wrong or stupid..

These charts are created from OHA supplied "Daily Updates" data, and sometimes from OHA supplied "Data Dashboard".

The two official OHA sources do not always agree with one another and sometimes have significant discrepancies between them.
The OHA periodically announces they have removed a prior "new case" or "death" after new information indicated the patient did not have Covid-19. OHA periodically moves cases from one county to another, or even to out of state, based on new information. This has the effect of decrementing earlier counts. However, OHA never says on what day the decrement was implemented - consequently, I have to pick a random day to reduce by 1. This means my data may be slightly off from a later OHA data release or chart.
As of June 6th, OHA stopped releasing detailed data for Saturday and Sunday.

See many more notes at end of this file.

MAJOR CAUTION

THE CASE COUNTS ARE THE COUNTS AS RELEASED IN THE OHA DAILY UPDATES.

I HAVE DOWNLOADED THEIR "DATA DASHBOARD" WORKSHEET AND THE DATA VALUES THEY SHOW, EACH DAY, ARE ***SIGNIFICANTLY*** DIFFERENT THAN WHAT WAS REPORTED IN THEIR DAILY UPDATES. I HAVE NO WAY TO DETERMINE WHY THESE VALUES DIFFER SIGNIFICANTLY. DATES PRIOR TO MARCH 29 SHOW "DATA DASHBOARD" VALUES AND DATES FROM MARCH 29TH ONWARD SHOW THE DAILY UPDATE VALUES.

I will ask OHA about the data discrepancies but do not expect to receive a reply. Neither OHA nor the Governor's office respond to citizen inquiries.

TEST RATIO DATA

As of May 27th, during the prior week, state-wide positive tests were 1.7% of all those tested. The percent testing positive remained below 2% until (roughly) the Lincoln county fish processing plant outbreak and then the Island City church fiascos - when the weekly positive tests then crossed above 3% again.

The following reflects total test ratio and percentage since the very first test in March (which includes nearly two months of testing almost no one due to lack of tests such that positive test results were a greater percentage - you had to be almost dead to get tested then - unless you were the elite, like the Governor and her husband, who both were tested in March because they had head colds.)

The test ratio measures the total number tested for each confirmed case (e.g. 29:1). The goal is for this to be a high number such that 1/test ratio (e.g. 29) is a small percentage.

Total tests done for each confirmed case:

Percentage testing positive since first tests in March:

For the most recent weeks, the percent positive has been about 1.7%.

CHARTS

The "daily body count" reflects the day that OHA counted the death in their public information release. In reality, this is not the "day of death". Per May 10, 2020, the death count went up by 3 but all 3 deaths were from 3 to 10 days earlier. Therefore, use caution in interpreting a particular day's death count.

Daily body count

Cumulative Deaths Statewide, Log Scale

Cumulative Deaths, Linear Curve

This data was recorded by hand from OHA "Daily Updates", starting in late March. The OHA Data Dashboard does not provide a downloadable set of "death" data for comparison. Note that the "curve" is linear, and not the much discussed "exponential" curve.

Five Day Moving Average % Change in Daily Deaths

Case Fatality Rate for Oregon

New Cases Trends

CAUTION

Before May 4th: OHA reported *only* confirmed cases of Covid-19.
Starting on May 4th: OHA reported *both* confirmed cases and "presumptive" cases (no test). They initially kept reporting both confirmed and presumptive numbers separately.
Around May 26th: they began reporting only a sum of confirmed and presumptive cases and no longer provide separate numbers. (We could probably pull this out by comparing day by day "confirmed test" counts which are provided - but this means recording each day and comparing pairs of days.)
New case trend after May 26th can not be compared to pre-May 26th.
Starting June 6, OHA discontinued providing detailed data for Saturday and Sunday.

Daily new cases in Oregon (Line Chart)

NOTE - Protests and riots broke out in Portland and Eugene beginning May 28th and onwards. Protests were still occuring as of June 10, 2020. These activities did not include "social distancing" and it is possible that 1-2 weeks after the onset of these protests the state will see an increase in new Covid-19 cases. Or may be not. If "not" then it means the draconian closures of parks, state forests and US Forest Service lands was not necessary.

7-day moving average of new cases in Oregon

On the day I set up this chart, the average number of new cases per day was 56. Hence, the chart shows "above average" in red. This is based on the post March 29th "daily update" figures.

Daily new cases in Oregon (Daily Update) (Bar Chart)

Cumulative cases in Oregon (Daily Update), Linear curve

This chart is based on my logging the OHA "daily updates". However, the daily updates have significantly different daily values than OHA "Data Dashboard" data downloads. No explanation for that. Because OHA did release much information prior to the end of March, I did not record "daily update" values prior to March 29th. Note that the "curve" is linear and not the much discussed "exponential" curve.

Cumulative cases based on OHA Data Dashboard data, linear curve

Major discrepancies Between Daily Update and Data Dashboard

OHA releases a total of new cases for the prior day in a "Daily Update" release.

Separately, OHA has a "Data Dashboard" to display an OHA chart of daily new cases. From that page, the data may be downloaded.

When I compared the "Daily Update" values to the "Data Dashboard" I find significant discrepancies that are left unexplained by OHA. These discrepancies are in addition to the different "peak" days shown on various pages and "Data Dashboard" data values. This chart shows the "new cases" as reported by the Daily Update and separately by the Data Dashboard. Clearly, they are not the same.

Of interest:

Mar 8th, Governor declares "state of emergency" (but no specific actions)
Mar 13th, all school districts choose to close schools
Mar 23rd, Governor issues "lockdown order"
Mar 20, 23 and Apr 1: Peak "new hospitalizations" with 26, 27 and 28, respectively, based on OHA Data Dashboard data table. However, the OHA chart of hospitalizations shows the peak occuring on March 28th. No explanation for their data discrepancy.
May 15th, first rural counties end lockdown and advance to Phase 1 restrictions
Jun 24th, Governor mandates public mask wearing in Multnomah, Washington, Clackamas, Hood River, Marion, Polk and Lincoln counties. My hypothesis is this will have no impact on the new cases curve (social distancing, work place restrictions and handwashing are likely far larger impacts - and face masks will add very little to that in most situations). The State is now conducting a real world medical experiment, albeit, without informed consent of the participants, which is a violation of Federal law. But laws do not apply to public health.

Of interest, the "curve" was already flat between March 15th and March 23rd, when the lockdown order went into effect. Peak hospitalization day, per OHA, was March 23rd. This suggests that the "lockdowns" may not have had much impact on the shape of the curve. Social distancing, work place changes and hand washing may have been the important factors.

The curve was already linear for about 2 weeks before the lockdown went into effect. But note that "Flatten the curve" referred to the hospitalization curve, not new cases. The hospitalization curve was trending down in late March per the OHA's hospitalization chart.

Would love to have an "expert" provide input on this.

The "exponential" curve we heard much about appears to be true only at the very start of the epidemic. By mid-March, the curve was "flat". It may have flattened due to voluntary measures.

Note: In first half of June there were two large outbreaks. One at a fish processing plant in Newport, OR, and then several hundred at a Church in Island City, OR, next to La Grande. The latter was caused by the Church holding large events, tightly packed, with singing and dancing, including church services, a very large wedding, and a very large graduation ceremony. This resulted in the increasing slope of the curve in mid June. After some church goers tested positive, Covid-19 antigen tests were offered to everyone in attendance. There has been no report as to how many were symptomatic versus asymptomatic.

OHA Chart of Hospitalizations

Note their chart shows peak day as March 28th with 28 admissions. This chart is DIFFERENT than their Data Dashboard data which shows peak day of April 1 with 28 admissions. Again, the discrepancies throughout their data are never explained.

Flattening the Curve - and Hospital Use Increasing Metric

The media is tracking the increase in hospitalizations at 3 months after the original lockdowns as most states have done some degree of "re-opening". There are concerns that hospitalizations are increasing again (as of mid June). However increasing the sick, now, was the theory behind the original "flatten the curve". The goal was to lower the initial number of hospitalizations, to avoid overwhelming hospitals - BUT - and this is important - the intent was to shift those hospitalizations into the future, not eliminate them. We "flattened" the red curve and shifted those hospitalizations to the "blue curve". We are now on the increasing side of the blue curve. That is, if that hypothesis from February 27th, is correct.

OHA Chart of Peak Day of Covid-like Symptom Presentation to ERs

Peak Day is March 13th, with a slightly lower second peak on March 21.

Daily Increase as a Percent Change

Unofficial "Hospital Intensity" Metric

This chart graphs a ratio of the 5-day sum of hospitalizations to the 5-day sum of new cases. Initially, we had a lot of hospitalizations but daily new cases were not that high. As time went on, the number of daily new cases being diagnosed went up, but the number of hospitalized patients went down. This suggest that we are now diagnosing cases that are much milder, than before (likely due to increased testing), or that the disease is becoming weaker (questionable) or that out of hospital treatments have improved (unknown). UPDATE: Statement from OHA June 17 2020 -"The percentage of emergency department (ED) visits for COVID-19-like illness remain below 1%, suggesting that many of the cases now being reported are lower in severity or otherwise not coming through EDs"

Hospitalization and Hospital Resources In Use

Note: The number hospitalized for Covid 19 confirmed AND presumptive is always about twice as many as confirmed cases. We assume that a presumptive patient in the hospital is going to be tested, right? But since the confirmed case counts are always half the confirmed+presumptive total, this implies that most of the presumptive cases were not actually Covid-19? Or that the test do not work? Something makes no sense here.

For this reason, I believe the only valid metrics are the ICU confirmed cases and ventilators in use number.

As of Jun 6, OHA no longer releases hospital utilization data on Saturday and Sunday. Consequently, for every week after June 6, the last released figures of Thursday are filled in for Fri, Sat and Sun, and then resume updating on Monday when OHA again releases data - but only for Monday - FridayCAUTION: Because we have seen weekly peaks in numbers late in the week, this may artificially extend a high value (or low value) across Friday-Sunday, resulting in the visual trend being incorrect. For example, Thursday shows a high value - this is filled in for the missing Fri-Sat-Sun values. But unknown to us, the value dropped on Fri and remains low. But we have no way of knowing as OHA no longer releases any data for the weekends so the chart can be very misleading when days of data are missing each week. Corollary: Covid-19's threat to Oregon is now so diminished that weekend updates are not longer important.

Cumulative Hospitalized Cases Count from OHA Data Dashboard Data

Note that this is a linear chart and the curve is not exponential.

Hospital Confirmed Cases

Hospital Confirmed and Unconfirmed Cases

ICU Confirmed Cases

Ventilators in Use

Cases by County Trends

Top Counties Only

New Cases in Counties Having Large Numbers of Cases

At time of this writing, that is the Top Six counties

Percent of Daily New Cases in the Top Counties

Number of new cases in the top counties

Deschutes County

Daily new cases in Deschutes County

Deschutes County Cumulative Cases

Deschutes County Estimated Active Cases

In this chart, a case is defined as "Active" for 20 days after it was first reported. According to the OHA, the median time to recovery is 20 days. This is a sum of the total outstanding cases for the prior 20 days. This is unofficial and appears to underestimate the number of recoveries per OHA published estimates on recoveries.

Economic Data Section

Apple Mobility Data

Click here

Financial Data

These charts show the trillions of dollars printed by the Federal Reserve. Longer term, this leads to inflation - first in asset prices (hence stock market going up) and then later in consumer price inflation, although there are multiple factors at work.

TSA Passenger Data

Passenger Count as percent of prior year

Passenger count as percent of same day the prior week

OTHER DATA - ER Utilization, ER-Visits for Covid, Flu-like Symptoms

ER Utilitization and Percent of ER visits due to Covid-like symptoms

These charts of Oregon ER utilization and percent of ER visits that have Covid-like signs and symptoms are produced by the Oregon Health Authority. "Covid-like" symptoms are based on clinical observation and not a test. "Covid-like" is a subset of the CDC's "Influenza-like Illness" category.

ER visits began to fall in March due to publicity that ERs were overwhelmed, and fell further when the Governor, shortly thereafter, banned all non-emergency procedures, resulting in people not seeking care for bona fide reasons. This included a 50% drop in presentation of patients with heart attack symptoms and a 30% drop in presentation of patients with stroke symptoms. I also had a broken foot during this period and was unable to see a doctor. Since Oregon did not experience a spike due to Covid-19, Oregon hospitals have been running with -30 to -50% fewer patients. The loss of revenue has resulted in thousands of health care workers having hours cut, being furloughed or being laid off. Several rural hospitals report they could become bankrupt and close.

The goal is that the % of Covid-like presentations to Oregon ERs is below 1.5%. This is the threshold value for off season ILI per the OHA. The blue line in the 2nd chart is a measure of Covid-like ER presentations, as a percent. As of June 11, per OHA, the statewide Covid-like presentation to ERs is now 0.6%.

OHA ONLY UPDATING THESE CHARTS ONCE PER WEEK - FACE PLANT - GO HERE for OHA official charts (new features from OHA).

This chart includes:

Updated Covid-19-like illness at ERs chart
Percent tests as positive
Percent cases not traced to known source
New Covid-19 cases
New Covid-19 cases with/without known source

Official Criteria for Re-Opening Includes that Influenza-like symptoms are at a low level

Oregon's Official criteria for re-opening counties includes that the CDC report of Influenza-Like Illness (ILI) is at a very low level for each county (preferably below 1.5% of ER visits). Covid-19 is a subset of "ILI".

Here is a link to CDC ILINet State Activity Report

Here is a link to the Oregon government ILI pages.

Here is the chart I captured - Oregon discontinued updating this chart in mid-April because the flu season is over, or something. Considering we are in the midst of a pandemic it seems that this data and chart ought to continue being updated but they are not.

The State of Oregon says that they are taking into consideration a comparison of today's illness outbreak to the Influenza-Like Illness tracking and that a low value is part of the criteria for re-opening the economy. For Oregon this will be compared on a per County basis but "per County" data is not made publicly available so I can only show the state level data.

"Covid-like" is a subset of ILI and is determined solely by clinical observations of symptoms. Neither it nor influenza diagnosis is based on test per the CDC.

CDC produced chart of ILI for the nation

CDC produced chart of ILI for OREGON

The next charts show the CDC influenza + pneuomonia mortality chart for the entire state of Oregon. This chart is updated on this page, infrequently. THIS IS NOT THE CHART OF ILI.

For comparison, here is New York City

For comparison, here is Michigan

This page and all content on this page is FOR ENTERTAINMENT PURPOSES ONLY.

-------------------------------------------------------------------------------------

IF YOU JUST WANT THE CHARTS WITHOUT EXPLANATIONS, DATA CHANGE HISTORY, ETC, THEN JUST PAGE DOWN AND WAIT FOR THE CHARTS TO LOAD.

OHA CHARTS

OHA hides data from public. (One of several reports on this topic by The Oregonian.)

Link: Access to CLI presentation at ERs, % tests that are positive, New Cases, Cases traced to sources, hospitalization data by state and county, active monitoring capacity, and county indicators.
Link: Demographic data, disease severity data
Tests and outcomes by county
Covid-19 press releases

Use caution when interpreting any trend lines. I created this page solely for my own quick reference from any computer. Assume all data is incorrect or cherry-picked to present a biased picture. Any interpretation you make is at your sole discretion and responsbility.

Charts here may look slightly different than charts provided by others. This usually occurs because OHA sometimes changes past data. Some times I catch the changes, sometimes I do not. The same occurs for official charts you may see online elsewhere - they often miss the "past data" changes. On June 15th, I discovered that the Daily News Update value seem to be slightly different than the "News release" page values. I have gone back and changed some of the recent daily totals to those in the "News release".

Link to Oregonian's "by zipcode" case tracker after OHA released zip code data on May 5th: Link to Oregon Zip Code Tracker

OHA REPORTING CHANGES

Up until well into April, you had to be hospitalized with pneumonia, or have all "official" symptoms and live in a skilled nursing center, or have all "official" symptoms and have been in close contact with a confirmed case. No one else could be tested. Unless you were the Governor or her husband, both of whom received Covid-19 tests, contrary to the rules, during the 3rd week of March because they had "head colds".
Beginning in May, testing became much more widely available in Oregon. But if sick before then, there was no testing available to anyone. I developed a head cold on the evening of March 3rd, which then progressed to head ache, muscle aches and pains, slight sore throat, low grade fever and then an unusual dry cough starting on March 8th. This dry cough continued until March 23rd, went away for several days, and then returned for about a week but a much lower level. Two doctors I spoke to said it was unfortunate that I could not be tested for Covid-19 back then. Unless I had been the Governor, of course. In late May, my own doctor suggested getting an antibody test; however, its provided by a private lab and we know nothing about its sensitivity or specificity/accuracy. I declined to get the test as it did not seem the results would be "actionable" besides satisfying a curiosity.
Until early April, OHA data was incomplete and sometimes missing. OHA released only a statewide count - no county data - during most of March. And provided no information on hospital utilization. By May, in my area, there are now at least two drive up Covid-19 test centers available, both operated by urgent care clinics. No doctor referal is required - they only ask that you call them in advance. During approximately the first week of May, the total number of tests run daily has doubled since the week before (and quadrupled by June). According to an epidemiologist, "new case counts" are expected to rise nationally, in part due to testing eligibility now being relaxed and the wider availability of testing. This means that higher confirmed cases in May is probaby an artifact of far more tests being run now and not an indicator of a greater prevelance of the disease.
Example: Recently, some workers in a fish processing plant tested positive so all workers were tested. 124 of them tested positive - but incredibly, 95% were asymptomatic per the OHA. That is a lot of confirmed cases that would never have been counted just weeks earlier. This, in turn, coincided with the first weekend of "missing data" from OHA such that on Monday, the updated total had skyrocketed by about 350 cases (three days with an unusual event). This, in turn, led to news media nationwide reporting that Oregon was experiencing a "surge" of "Second wave" cases, even though by Monday and Tuesday, the count was back in the normal daily range.
Before May 4th, OHA released a count only of "test confirmed" cases.
Starting on May 4th, OHA began releasing a headline number that includes both "presumptive" and "test confirmed" combined. This resulted in an upwards inflection point in the official charts. For the state, their daily report still split these numbers out so that my charts continued to use only the "Test confirmed" value - until OHA stopped releasing the individual numbers on May 26th. NOTE - OHA DOES NOT split out these numbers in the "by County" report so that after May 4th, there is a slightly higher number of reported cases causing an inflection point in County data. Additionally, presumptive cases are sometimes later thrown out and this may or may not be reflected in my accumulated data tables for the County.
Before May 23rd, OHA released data "current as of 8:00 am" each day. On May 23rd, they changed the reporting period to "12:01 am" of each day. This added an additional 8 hours of data to May 23rd.
Beginning on May 4th, OHA began including "presumptive" + "test confirmed". A presumptive case may include those with a negative test result - OHA --> "Includes cases confirmed by diagnostic testing and presumptive cases. Presumptive cases are those without a positive diagnostic test who present COVID-19-like symptoms and had close contact with a confirmed case."
About May 26th, OHA stopped providing separate "confirmed" and "presumptive" numbers so that the public total combines both confirmed and presumptive. This appears to inflate the number of daily Covid-19 cases by up to 10% - having gone from confirmed-only (and hard to get tested) to confirmed+presumptive (no test) while the number of daily tests has nearly tripled over the month (quadrupled by early June). In other words, the trend data is largely garbage.
As of June 6, the OHA discontinued giving a full daily update on Saturday and Sunday. Which means it may no longer be possible to accurately track anything. OHA has come under intense media criticism for their lack of transparency, their reluctance throughout this pandemic to release information to the public, and their continued reluctance to release important information.

PEAK DAYS TIMELINE

March 8: Institute for Disease Modeling (in a late May report) says Oregon's curve began to bend on March 8 owing to voluntary measures (IDM)
March 15: Peak day of Covid-like illness presenting at ERs in Oregon (OHA chart)
March 23: Governor announces state-wide "lockdown" program. Charts show this lockdown had no impact on the already decreasing trend line.
March 23: Earlier OHA charts and CLI data implied "peak" hospitalization was around March 23d. Their "data dashboard" data, through June 15th, puts the peak days at 26 on 3/20, 27 on 3/23 and 28 on 4/1 (OHA)
April 5: Peak day of hospitalizations in Deschutes County (St Charles Hospital, Bend)
April 5, 9 and May 20 [*]: Peak days of "new cases" in Deschutes County
April 9: Peak day of state-wide ICU bed usage by confirmed patients (OHA)
April 6: Peak day of state-wide ventilator usage by confirmed patients (OHA)
April 27: Peak day for state-wide deaths (OHA)

[*] Note on May 20th. According to KATU, and then the Oregonian, a large family social gathering took place in Deschutes County prior to the May 15th "Phase 1" re-opening plan with an ill family member. As part of newly expanded contact tracing, after this case was confirmed other attendees were then tested and found to be positive. All of the contacts took place BEFORE THE MAY 15th "RE-OPENING" - Deschutes County Public Health said very specifically this occurred before May 15th AND was not due to community spread. About 75% of all new cases - what the media called a "spike" - that week were due to this get together. Much of the news media reported this as "Breaking news: Central Oregon sees surge of cases DUE TO RE-OPENING" which was not true but made for a scary headline. A related issue is the use of perjorative wording by the media: an "increase" is a "surge" or "spike" - because the media knows that fear sells. But this increases anxiety and health problems associated with anxiety

Comments from OHA on the Effect of More Testing

From the Oregon Health Authority:

Readers might notice that, although Oregon is moving towards reopening some businesses and relaxing some of the restrictions on daily activity, more COVID-19 cases and deaths were reported during the week of May 4 through May 11 than during the preceding week (394 cases and 14 deaths). However, the actual number of new infections (also called disease "incidence") probably did not increase; rather the increase in reported cases is most likely due to increased testing and increasing numbers of presumptive infections. The number of COVID-19 tests reported in Oregon rose from 12,684 in the week ending May 1 to 14,481 in the week ending May 8, a 14% increase, while the proportion of tests that were positive fell to 3.1%, its lowest level since OHA began tracking COVID-19 testing

Relative Risk Observations

To help keep personal panic under control, it may be helpful to put the risks in perspective by comparing that risk to other common risks.

As of the end of May 2020:

You are 2x to 5x more likely to be in a car crash in Oregon TODAY than being diagnosed with Covid-19 TODAY.
You are about 5x to 12x more likely to be hospitalized for a crash TODAY than to be hospitalized for Covid-19 TODAY.
You are slightly more likely to die in a car crash in Oregon TODAY than to die of Covid-19 TODAY.

The multiplier varies, per day, due to differing numbers of diagnosed new cases and deaths reported. The normal range is "2x to 5x" due to the variation in daily new cases identified.

An OSU random sample survey of over 600 people in Corvallis, and over 600 people in Bend, found zero cases of Covid-19. (Each person was tested for Covid-19). Since they could not find anyone in their sample who had Covid-19, they instead used a model and estimated that 1 in 1,000 people have Covid-19 at any time. But they honestly do not know - a model is not data. If you meet a random selection of people in Bend, the probability that you contact someone with Covid-19 is p=0.001. It is said that if you maintain six foot social distancing, your odds of contracting Covid-19 fall by an additional one to two orders of magnitude (this appears to be another model output, not based on data). Thus, we then have p=0.0001 or may be p=0.00001. Furthermore, the CDC and WHO say you need to have extended close contact (various estimates are 5 to 15 minutes) which cuts the p value even more. It seems that the liklihood you pick up Covid-19 from a random, momentary near contact in a grocery store is close to zero. One does not need to be an expert to see this. NOTE - this only applies in my community. The prevelance may be very different in your community. It would be extremely helpful if public health would provide honest assessments so that the public could gauge relative risk, rather than causing severe anxiety, angry mask shaming, and other societal dysfunction. The State of Utah has published data on where contacts occur - per Utah, of the fully traced contacts (about 85% of all cases) about 60% are extended close family contacts, about 25% are extended close social contacts, about 5% are at work, and the remaining 10% are "travel related" or "other/unknown". This is useful data that can help to control anxiety. BUT you should not extrapolate Utah's data to your community - different lockdown rules, different family dynamics, etc. This is why local public health should release similar information.