Ontario Regional Dashboard 1030am Aug 2
York, SW, and Chatham among regions with higher active case today. Windsor, Ottawa, Toronto slightly.
(More incl some notes on "active" at end )
#Covid19 #Covid19Ontario #CovidOntario #Onhealth #CovidCanada
 https://public.tableau.com/profile/bill.comeau#!/vizhome/OntarioCovid-19CaseMap/Dashboard1

Full rankings today based on active counts.

Per 100K - this is where smaller regions come into the spotlight using case incidence rates.

Full regional numbers.

On active cases: the province uses a cutoff rule for "resolved" which is essentially set as 14 days after first symptom onset if not hospitalized (or dead).

This rule does not mean a case is fully recovered. They could even still have symptoms but not considered serious enough for hospitalization.
The regional websites will differ for two reasons:
1. They report at different times, often before the provincial process which cuts off

... usually at 4pm the day before then reports at 1030am next day.

2. They use different definitions for resolved. Eg, York uses this definition which is based on presence of symptoms. That means York will have fewer resolved (and therefore more "active") as seen in next tweet.

I took a look at York's numbers last night. You can see a large gap in resolved (-150) as expected. That plus the fact it reports *after* the provincial 4pm data cutoff means it has more total cases and more resolved. The net result is 181 more active.

Which is more accurate or better? It depends.
Checking York's website after 4pm the day before is more timely in new and total cases.
It's actives include those past 14 days that *reportedly* still have symptoms.
IF York monitors symptom status every day and keeps accurate logs

then its "active" is a good indicator that is more realistic on a case's status.
The issue for the province is simple: each region can have different rules, bookkeeping and monitoring to check resolved vs active. Some may be slow or inconsistent, others very sharp.

The only way to have a consistent measurable definition is to have a clear and measurable rule. That's the 14 day rule. In many ways it approximates a 14 day new case trend. It's also undercounting the total cases that remain in limbo between 14 days and no symptoms

We can't say for sure which is "more accurate". If you want the latest results, check the regions. If you want to see a broader definition of who is truly unresolved, check regions that choose a rule like York's (if they record it accurately).

...If you want consistent comparisons on a definition of active, use the province's number, or if you want recent case development status (~ 14 days).
Until the province and regions synchronize, this will be a confusion.

I would recommend they maintain "narrow" and "wide" definitions:
1. The current arbitrary Ontario rule, call it "Not Hospitalized 14 days after symptom onset"
2. A "Resolved based on no symptoms at 60 day checkup"

So if someone asks why the province or the region is "wrong", point them to this thread and hope people like @janephilpott or @celliottability can help sort this out because tracking the severity of cases consistently and accurately is critical. /end
tl;dr ? it's complicated

Addendum: why do I say active cases at the provincial level approximate 14 day new cases? This is why: