February 8, 2011 CBCC Conference Call

Community-Based Collaborative Care Working Group Meeting

Back to CBCC Main Page

Agenda

1. (05 min) Roll call, approve minutes January 18th, call for additional agenda items & accept agenda
2. (20 min) Review action items January 18th,
3. (35 min) Review Domain Analysis Model for SHIPPS

Minutes

1. Action Items

Review action items from 20010118

Action item 1: Follow up with Ken with respect to PQRI (Mary Ann)

• Are they developing additional measures or are they are working (adoption of existing measures) with the same measures that NQF is managing?
• Are there other measures other than these 5 we are working with

Action item 2: To follow up on what it means to have a registry reporting process in place. (Serafina)

Action item 3: To review 2011 PQRI Implementation Guide may be useful in terms of a modeling perspective

2. Resolutions

Motion to approve minutes from 20110201 by Ioana and seconded by Richard

3. Updates/Discussion

Back to CBCC Main Page

Physician Quality Reporting Initiative (PQRI)

PQRI is a voluntary physician quality reporting initiative sponsored by CMS, it enables providers to report on their quality measure for an incentive payment. Eligible providers are specified and can voluntary report on, as of 2010, 216 measures. These measures are basically claims based reporting process or registry process. They are trying to evolve into an EHR. There is some relevance to us through their claims based process. Their individual measures identified can be combined with the NCQA certification, which are the measures that NQF uses. The measures they report on have to be NQF endorsed. There is an overlap 2011 implementation guide how to approach the claims based reporting may be useful to look at from a modeling perspective with what we are doing. Defines categories codes QCodes.

Question: How much more analysis are we going to do with respect to PQRI?

Ioana: PHq9 is a PQRI measure

Richard: Are there other PQRI measures we should be looking at

Ioana: PQRI is more in the business of operationalizing measures that in creating new measures.

Serafina: They also maintain measures and do have a call for new measure 2012

Ioana: Are they developing additional measures or are they are working with the same measures that NQF is managing?

Richard: If people are developing new BH measures we ought to know who they are.

Action: Follow up with Ken.

• Are they developing additional measures or are they are working with the same measures that NQF is managing? Are they developing additional measures or just adoption of existing measures?
• Are there other measures other than these 5 we are working with (Mary Ann)

Action item: To follow up on what it means to be a registry reporting process. (Serafina)

Action item: To review 2011 PQRI Implementation Guide may be useful in terms of a modeling perspective

Domain Analysis Model for SHIPPS The organization of the document is set out to accomplish what we have done in our project scope statement.

Information Maturity Levels

• How information evolves from unstructured to structured
• Local codes and then eventuality standard meta data with standards based encoding
• Identify drawbacks from various types of information
• Why it is desirable to move to the highest level of maturity to support secondary uses of clinical data
• This will include data segmentation and real time quality measures

Functional Analysis

• Discussions on infrastructure functions that are currently available in EHR Functional Model v2
• Gaps identified in the Supporting Functions
• Currently no ability to specify how the consent directive makes it into the EHR e.g. no provision for either EHR or PHR based entries of privacy preferences
• Main focus missing functions in supporting privacy
• As we go through our analysis of quality measures we may find missing functions that enable real time quality measures

Quality Measures Analysis

• Analyzing quality measure, the 8 measures we identify in Behavioral Health
• We looked at the data required to compute them
• How to represent the constraints that are part of the eMeasure e.g. negative 60 day period to determine that we have a new episode of diagnosis
• Our analysis will identify the state of affairs today e.g. measures are computed from claims data there is an assumption and necessity to tie the substance abuse procedure and diagnosis to encounter because people get paid for services. In the future it may be possible, if problem lists are maintained by EHR to look directly when a diagnosis was entered or date of onset by looking directly at the problem list. This would make the computation a lot simpler so it you were looking for a new episode you would look at the date of onset that occurs during the measure period versus inferring that this is the first episode by looking at the negative diagnosis

Richard: When services are more rationally organized in terms of networks instead of individual providers treating you somehow ‘de nova’ every patient by themselves. We could have a more sophisticated measure here for date of onset.

Date of onset

• What we have seen is that it’s not really the date of onset necessarily but the date it was recorded in the context of a certain encounter.
• When it appears on a problem list who reported it and when it really started. There will be multiple dates.
• Depending on how date of onset is defined and how the EHR is capturing that data it is debatable how the information is going to be used.
• When it comes to new substance abuse episode the date of onset could have been prior to a date when it was observed in a new measure period. The date of onset may not be available at all you just have the information when it was recorded.
• May not be able to come to agreement to what date should be used; it needs to be applicable to the measure.
• You could have e.g. 4 or 5 dates of onset depending on the number of providers you go to

Self-reported data versus assessment made in an organization

• In this measure we are talking about date of onset that would have been recorded / attested to by a physician. We will capture the information for this measure as it relates to an encounter.
• We will capture the ‘as is’ and ‘to be’ and we will create 2 separate views of the model to illustrate the point of claims based associations versus problem list based. Where possible we will include the constraints using Object Constraint Language (OCL). Ongoing discussions within Structured Documents:

Structure Documents discussion

• Regarding GELLO and Java Script that may be used to describe the conditional logical and numerator and denominator in a measure.
• We know these can be described these precisely and we will demonstrate these data fields can be used create Structured Query Language clauses and segments that could be processed against a compatible data base.

That’s the information analysis. We are also going to include:

Meta Data for Privacy

• We will include examples of protective information that has necessary Meta data to enable it for processing. Our top priority should be CDA base data as this is a candidate for many projects. We will demonstrate how the information we’ve already completed for composite privacy would apply to CDA R2 and identify the Meta data required to process and this will provide us an opportunity to do gap analysis of what additional Meta data may be necessary.
• We will leverage the extensive analysis what we’ve done on what is required to describe privacy policy and we will look at the corresponding Meta data that will have to be carried by the data itself to enable the automatic access control and data segmentation operations that people may need to support.

Mike announcements

HIMMS conference Feb 20 demo (VA, SAMHSA and Jerico) use of tagged Meta data and Health EHR to enforce patient Privacy Consent in two ways:

• Compliant with the Presidents Council that advices on Science and Technology, Health Information Technology Report that was produced in December last year.
• The Presidents National Strategy for Trusted identities in cyber space, with a draft national strategy that this is also compliant with and can be found at www@NIST.gov/NSTIC. Recently commerce was directed to lead this activity. So there is significant activity here and in the US Health Information technology standards and policy committees in regarding using tagged Meta data forcing patient privacy.

High priority

• Functional analysis as it ties into the EHR WG
• Identify protected information as it relates to Structured Documents
• Ioana to sit on a panel discussion at HIMMS. Tuesday 1 p.m. Meaningful Privacy Choices: The challenges of granular patient consent.

Glossary of terms Include links to:

• NQF QDS
• Harmonized Security and Policy DAM
• New terms e.g. data segmentation, NCVHS introduced it in their letter to the Secretary of Health and Human Services
  • Security people have been using term segmentation
  • SAMHSA emphasis on more granular consent
  • PHR vendors have shown a great deal of interest in segmentation

Publication Under Domain Analysis Model and it will be similar to what we have done in the past e.g. use the Security Domain Analysis Model as an example.

Meeting adjourned 2:56 EST