Attachment 2 - Vendor Technical Questions and Answers.pdf

Attachment 2 - Vendor Technical Questions and Answers

EMS Data Access Updated: 8/4/2020

Q1. Can you specify the dataset / data fields that you are requiring for this project?

A1. Access to and ability to query near real-time and historical data at the county and state level are needed to identify cases meeting various clinical syndromes defined by CDC to produce near real-time count data by day for counties and states. Clinical syndrome definition building by CDC should allow for text analysis of narrative data as well as searching for medical diagnosis/billing codes. Data should be able to be queried from a web interface and downloaded as csv or MS Excel files. No limits on the number of queries or downloads should exist as data may need to be queried daily or weekly. The proposal should state availability of variables by which data can be stratified including age, gender, race, and other available fields.

Q2. What date ranges are you seeking data for? ImageTrend has been collecting EMS data since 2002. How far back are you seeking data? In addition, if this is an ongoing data set, would you like access to the dataset in real time and at what frequency would you require dataset updates?

A2. We understand that the time period of historical data is likely to differ by state, please indicate historical data availability by state. Ongoing real-time access to state/county data is necessary; data that is updated in near real-time (please specify approximate time lag by state from when clinical encounter occurs to when data is available to CDC, i.e., 1 hour, 1 day, 1 week, etc) should be available to CDC analysts to query and download in a web-based application at any time (application should function completely in a browser; due to CDC security, software cannot be downloaded to CDC machines).

Q3. In terms of geographical identifiable data, would you require the data be identifiable by state, county, zip code, census tract, specific geo-address locations, or some other variation?

A3. Data should be identifiable by state and county, at minimum. Note if smaller levels of geography are available.

Q4. When you refer to “data,” are you requiring the ability to download the transactional/row level data for the entire dataset, or do you want to view the aggregated/rolled up data calculations using a visualization platform?

A4. Access to individual patient-level record data would only be needed to develop the clinical syndrome and case definitions that would comprise various queries. Syndrome/case definitions will need to allow for querying the narrative text of the record for various words and phrases as well as medical diagnosis/billing codes. For each case definition that is built, CDC analysts would query the near real-time data which would need to return aggregated counts by day by county and state. This information can be stratified by other variables such as age, gender, race, etc.

Q5. What format would you like to receive / download the data in (i.e. secure FTP export of CSV or SQL DB; remote SQL access; PowerBi; NEMSIS v3 XML; etc.)?

A5. Data can be downloaded as CSV. As data access is needed in real-time and for queries that may change based on public health conditions, data download should be on-demand from a web-based application that allows for specifying various queries, rather than scheduled, static, or infrequent downloads. No limits on the number of queries or downloads should exist as data may need to be queried daily or weekly.