This repository is intended for jurisdictions to easily share resources related to the COVID-19 outbreak response. Please refer to the CDC website for the most up-to-date official guidance for public health departments. _______________________________________________________________________________________________________________________
The audience for this Basecamp is CSTE members working at STLT health departments participating in COVID-19 response activities. Appropriate resources to upload include health department-created guidance documents, resources, or other useful tools that may be helpful to other jurisdictions during their response. Forums may be used to pose questions for discussion among STLT jurisdictions. ______________________________________________________________________________________________________________________ Prior to using this Basecamp, please ensure you review the **INSTRUCTIONS FOR BASECAMP USAGE** folder, which contains "Basecamp Repository Instructions and Rules" and "How to Change Basecamp Notification Settings." Before posting in the Forums, please review the first posting “How to Use Forums” for rules and instructions. _______________________________________________________________________________________________________________________ IMPORTANT: This Basecamp is not housed behind the CSTE firewall and users are not monitored by the CSTE National Office, therefore CSTE does not recommend uploading any sensitive resources.
_______________________________________________________________________________________________________________________
Note: This Basecamp is not monitored by CSTE National Office staff on a daily basis. If you need to REPORT a Basecamp user, forum posting, or uploaded resource, please contact the National Office at preparedness@cste.org or call 770-458-3811 during normal business hours.
I am working with all of our COVID ELR data (positives and non-positives are reportable) to answer several questions, e.g., testing volume, people tested, etc. as well as set up more long-term surveillance using these data. I've made a lot of progress with respect to data cleaning and deduplication but would be interested in connecting with others doing similar work to share code or coding ideas, data visualization ideas, and other strategies for working with these data. I know I still have work to do on my deduplication algorithm, and would be very interested in others' approaches to this issue. If anyone out there is doing something similar, please let me know!
Comments & Events
jing
Hi Erin,
We perform data deduplication routinely in our program area (of CDPH). Both LinkPlus and MatchPro (https://surveillance.cancer.gov/matchpro/download) are good options. I believe both accommodate probabilistic matching. We also have SAS codes for more speedy matching supplemented with review of unexact matches divided by certainty levels. Names and DoB are matching input requisite (SSN optional). The SAS compged function is the matching workhorse but seems not as fast as indexing and key look up. We can chat offline if interested.
Jing
Notified 138 people
Ken Komatsu,EB Member-At-Large
HI Erin, A related question, wondering how many states use only ELR data for their testing metrics and how many are entering all their paper/faxed reports for more complete data? We are only using ELR data.
Notified 219 people
Erin Murray
Hi Ken,
We are only using ELR data for molecular/PCR tests in our testing metrics as well. The one caveat to this is that we are receiving some lab results in CSV files from labs, and those are being incorporated into our surveillance system and included as well, as if they had come in via more traditional ELR mechanisms.
Notified 220 people
Sherri Davidson
We are including most* ELR and non-ELR diagnostic tests in our testing counts, but only ELR diagnostic tests in our percent positives. Same as Erin, our *.csv files appear as traditional ELR in our surveillance system. (*I exclude ELR that indicate the testing was referenced out or non-ELR that appear to be a duplicate manual entry of an ELR.)
Notified 220 people
Jim Collins
Testing Metrics Michigan Cleaned ELR Yes; Cleaned CSV (that we convert to ELR to be consumed by our surveillance sytem) Yes; manually entered positive results from fax/phone referrals (extreme minority of cases here) NO
Notified 220 people
Katherine McCombs
We use ELR and hand-entered lab results. Curious how many hand-entered lab results other states have? Are you planning to enter them eventually?
Notified 221 people
Sarah Park,State Epidemiologist
Yes, ELR data. For now, either uploading .CSV or having national guardsmen data enter any faxed (ugh) reports from these pop-up labs, but if/when that volume goes up? Surge testing planned by HHS here in the coming couple of weeks, and we're hearing we'll likely receive the results as flat files. ??!!!
We perform data deduplication routinely in our program area (of CDPH). Both LinkPlus and MatchPro (https://surveillance.cancer.gov/matchpro/download) are good options. I believe both accommodate probabilistic matching. We also have SAS codes for more speedy matching supplemented with review of unexact matches divided by certainty levels. Names and DoB are matching input requisite (SSN optional). The SAS compged function is the matching workhorse but seems not as fast as indexing and key look up. We can chat offline if interested.
Jing
A related question, wondering how many states use only ELR data for their testing metrics and how many are entering all their paper/faxed reports for more complete data? We are only using ELR data.
We are only using ELR data for molecular/PCR tests in our testing metrics as well. The one caveat to this is that we are receiving some lab results in CSV files from labs, and those are being incorporated into our surveillance system and included as well, as if they had come in via more traditional ELR mechanisms.
Cleaned ELR Yes;
Cleaned CSV (that we convert to ELR to be consumed by our surveillance sytem)
Yes; manually entered positive results from fax/phone referrals (extreme minority of cases here) NO