The desired outcomes for the meeting are:
A. To reach the greatest mutual understanding of how each partner plans to implement the CDS tool
B. Resolve technical questions that could become blockers to implementation
C. Get a rough ballpark date for when the implementation itself would be started for each partner
| Info |
|---|
Meeting Resources |
Attendees
|
| CDC | NACHC INFORMATICS | HealthFlow | athenahealth |
|---|---|---|---|---|---|
Andrew Hamilton +Shelly Sital JR xJR Segovia Jena xJena Wallader Gemkow xNicole Padula | Michael xMichael Lieberman Seren xSeren Karasu | Karen xKaren Hoover Aileen xAileen Ya-Lin Dejene Parrish Patrick Schoen (HRSA) Mary Tanner Anne Kimball WeiMing Kevin Delaney Pamela xPamela Grudua xNinad Mishra | Julia Ray Pedro Crystal Jayson | +Ken Allgood Thai Lam xhJames James | xNele Jessel |
Agenda/Notes
Topic | Speaker | Notes | Action Items | ||
|---|---|---|---|---|---|
Welcome | Julia | ||||
12:00 - 12:30 | All Partners
| All | We reviewed SOW, Task 1 is almost complete, need data upload to NACHC. |
| |
12:30 - 12:45 | AllianceChicago
| AllianceChicago |
Implemented first site last year.
Karen - Does the CDS tool include the value sets for the high-risk indicators that were in the IG, that NACHC, CDS. Nicole - Yes, we incorporated the value sets. Ninad - In the EMR system, the value sets are built in. Nicole - Yes, the risk indicators are built in, and kicks off a questionnaire. Julia clarified - you can implement the IG/tool or you can integrate into the EMR. Ninad - Great beginning - how installation-specific is it? If you install in one Clinic at AC, will the algorithm work in another clinic? John - The IG provides a framework, and every application is the same. Nicole, Shelly, Aileen, Karen - The tool does not automatically order a test; the test they order is a rapid test, not a lab test. The changes that AC made to the process were made with feedback from the clinic site about their workflow and what is possible. Pedro contributed that all risk factor implementation were in different forms and the patient had to see different providers and what AC is showing is one, integrated form. Also, they may have to provide consent before they get tested. Ninad - Is this mapped to.. what is pulled into decision support. Julia - It is mapped, this clinic might have consent requirements rather than automatically do the order. Karen - CDC has recommended since 2006 that no consent is required. Julia - sometimes clinics keep it anyway just to be sure. Pedro - this helps them document refusing, opting out, was this offered to everyone, so they use this to show who refused. Karen - Where does our CDS tool fit into this template? Aileen - if they find risk, it flags the check-box. Will the provider also ask the flagged check-box questions? Nicole - Yes the provider would also ask these questions. Karen - So this form comes up as a result of using the CDS tool. Nicole - This will not show up automatically, it has to be called up. Karen - The point of this was to skip the provider so that ppl can get tests. This is antithetical to what we are trying to accomplish. Julia - Could AC speak with the site about automating it so it bypasses the care team? Karen - It requires human action - the point of this was to bypass the care team. Building in human contact is antithetical to the vision, which was to make HIV testing a routine part of health care. Julia - If is a rapid test, the patient does it themselves. Karen - The point is to get the blood drawn and get the HIV lab version ordered by the system. We don’t want to put it on the provider to have to click the boxes and take steps. Julia - AC is moving it from the provider to the care team.. Karen - If we have care team intervention, the rate of testing will not reach 100%; it might be more like 20% based on experience. Mike - A clinician is always responsible for an order. There has to be a system-level decision. Karen - This is a proof of concept, and there needs to be an automated option with the system-level decisions made at the start so that we can push the boundaries to get more testing - the rate is lower than it should be. Mike: Epic shows what is “due” - we have an HIV screening health maintenance topic. All adults screened once, and at-risk patients get screened yearly. We don’t have a really good definition of high risk. Improve upon the EHR to find the information in the record to show which patients should be screened. We envision that the MA will order everthing that is due for the patient. The MA can order or get practioner to sign off. There is human intervention. Aileen - Can Mike confirm - they are not using the CDS tool? Mike - It is implemented in different ways - a smart set bpa, to present the HIV test order for signature. Aileen - How is this different from your current model? Mike - Improving ID of at-risk patients; we will implement an alert for provider to make it more visible that they can order the test with one click. CDS will identify the at-risk patients (will not be using FHIR). Karen - You can make this go away by ordering the test. She thinks it is a good beginning that CDS will identify the at-risk patients. Mike - It is difficult to get info on multiple partners, getting paid for sex. Karen - CDS is based on codes, these things not part of CDS tool as it was built. It is based on objective criteria, codes. Julia - We are not asking partners to implement something that is not in the EHR; working with HealthFlow and Athena to help with that. Ken - I agree with the CDC on their vision - right now, CDS is being designed to identify the risk and get pro-active engagement, reduce the number of clicks from clinical team. The CDS does the back-end work. there still needs to be a human in the loop, but eventually get the human to do less. Nele - Agree with Karen’s goal; but agree with Mike that is difficult to find those kinds of specific data elements - it is usually done in a questionnaire - it does not usually come into the EHR from a lab result, it is usually something captured in the history. Once it is in the chart, I can find the data, but it has to get into the chart. Karen - sometimes there are lab results that are indicative of ppl at risk who are not getting tests. Example, a person has three rectal infections with gonorrhea, the CDS tool can pick that up. Nele - Yes, understood - it could be based on the objective data ideally. Karen - example a 14-year old with history of STDs who tested positive for HIV - she should have been given an HIV test and education based on the STD results. Ideally - something that pulls EHR data, patient could fill out questionnaire int eh waiting room… some way to communicated to the patient that they have reason for getting a test and learn to protect themselves. Pedro - This is the first-year that clinics have had to report the number of 33% (offers or tests?). We are changing the scenario. We can put this out there, but hcs have to look at their workflow and what they are competing with other priorities… They have the workflows set up and now have to report on quality improvement every year. There will be a higher uptake. Karen: We are trying to take the exceptionalism out of HIV testing. The advantage of a CDS tool is that it could pick up ppl at risk so they can get PrEP rather than be on a lifetime of anti-retrovirals. | ||
12:45 - 13:00 | OCHIN
| OCHIN |
| See above. | |
13:00 - 13:10 | Connect-A-Thon Part II | John GreshAdmin | Challenges: John showed the demo…
2. coming to consensus on the workflow takes time up front, how do we come together as a team to create a single application that uses all of this, and agree that we can move forward. Nele - Is this ready for use? John - This is a proof of concept. How to you make the choice of what is going to be used from all the data? Nele - How much of this is autopopulated. Athenahealth can turn on this app. Who are you looking at to build the SMART on FHIR app? Karen - Would be good to meet again as a group. |
| |
13:10 - 13:30 | Questions and Discussion | CDC, All | |||
13:30 | Close | Julia |