By 4/11: Please input your thoughts on Diagnosis Date here to get us started on our next data element discussion. Archived

·         How do you determine the date of diagnosis? Do you use only the provider diagnosis from provider interview or chart review?

SNHD: From Provider report/chart review or in certain circumstances, date of confirmatory test.

·         Could date of diagnosis instead be the date public health confirmed a case status, or when the lab confirmed the test result?

SNHD: This does not seem like a good idea because sometimes and investigation will go on for a long period of time and this would be a huge delay from onset or from doctor diagnosis or from lab date. For example, if we need to send a sample to CDC for confirmation testing this could take weeks and the case status would not be updated until they were received. 

Lab confirmation is a preferred method, when available. but we should use smaple collection date instead of lab confirmed date (test_result_date).

·         If the you do not have the source or information specified above, do you use other information to populate this field or leave it blank? 

I think if we do not have a clinical diagnosis date, then the lab date is used? Right? 

SNHD: We leave it blank

Attached you will find a document summarizing the feedback/thoughts of the various disease programs here in California. 

CSTE Data Std Workgrp - Diagnosis Dates.docx 14.3 KB • Download

For Arizona:
In practice, we commonly use two different things:  date of provider diagnosis, if available (from provider report, chart review, provider interview) or date of lab results / lab confirmation.  Using date public health confirmed a case status would be rare, if ever.  

Date of diagnosis is filled in for about 85% of the confirmed/probable cases in our system, though we would not be able to distinguish the source (provider dx vs. lab results) and some of the dates may not make sense in relation to the original report of the case.  In general, we have not put a lot of weight on obtaining the diagnosis date.  I believe our practice of filling in lab dates may be a carry-over from older systems that did not have as many fields for entering distinct dates, and we have been re-thinking that recently.  

From our Vector group:  date of lab results.  

From our STD group:  PRISM does not have a date of diagnosis field. For chlamydia/gonorrhea they use the date of positive laboratory test. Since most cases are asymptomatic, cases really should not be diagnosed without a lab test.  Syphilis is a bit more complicated. About half of our cases are identified in the latent stage and symptoms of syphilis can mimic a lot of other diseases, so again providers should not be making a diagnosis until they have done the appropriate testing. For syphilis we usually use the date of initial test (if the case ends up being a false positive, it’s set as a non-morbidity and the date of initial test is not counted as a diagnosis date). 

In NYC in the Bureau of Communicable Disease, we calculate the date of diagnosis.  It is an absolutely key field for us and we classify all of our cases by date of diagnosis and use it to calculate the MMWR week (I believe and will try to verify!).  Here is what we do.
We use an algorithm - which is the earlier date of two dates:
EITHER
1) Date of diagnosis provided by a report from a medical provider - this can come through an electronic report from a provider or can be manually entered by an investigator based on a medical record review or interview with a provider, 
OR
2) A calculated value based on a hierarchy - If there is a positive lab test for the condition we use the earliest date of collection for all the positive lab tests.  If there is no collection date for a specimen with a positive lab test, we use the report date - for this we use the date the report was originated by the reporter if available, or the date the report was received by the health department if the originating date is not available.  Finally, if there is no report date, we use the create date in our own Maven system.

Oregon
How do you determine the date of diagnosis? Do you use only the provider diagnosis from provider interview or chart review?

We would use patient interview, if not able to be interviewed medical diagnosis. Not so important for many of our diseases - more important are onset, report and specimen. Key field for HIV - capture both HIV and AIDs onset date. Make note of where the date comes from

Could date of diagnosis instead be the date public health confirmed a case status, or when the lab confirmed the test result?

No we have a different fields for that. Diagnosis date is not a required field and is not populated with any calculations

If the you do not have the source or information specified above, do you use other information to populate this field or leave it blank? 
Leave it blank