PH+Reporting+User+Story+-+Syndromic+Measure+Reporting

include component="page" wikiName="siframework" page="PHRI Header" =User Story: Syndromic Measure Reporting=

Contact Info:
John Page (john.page@agilex.com, phone: 703.889.3977) -- Agilex Aleena Dhar (aleena.dhar@agilex.com, phone: 703.889.3187) -- Agilex

Date Received:
11/18/2011

We are submitting the Syndromic Measure Reporting User Story for consideration by the ONC S&I Framework Public Health Reporting Initiative (PHRI). This user story addresses the ability for a provider/facility to report syndromic measures to a public health agency, in order to provide a near-real time Situation Awareness capability for the most commonly anticipated infectious health events. It is anticipated that it would operate in parallel with the Syndromic Surveillance using UC and ER Data user story which is being submitted by the International Society of Disease Surveillance (ISDS), which defines surveillance reporting of patient level data using the Minimum Data Set (MDS).. The complementary approach we are suggesting in this user story is based on the Distribute Model, and supports a less data-intensive surveillance capability than can be highly automated, and can scale up to support non-urgent ambulatory and inpatient settings.
 * 1.1 Introduction**

The aim of this User Story is to supplement the Syndromic Surveillance using UC and ER Data with a lightweight, aggregate reporting of syndromic surveillance measures directly from the EHR. The set of measures will be limited to a small set of common syndromes of interest, such Influenza-Like Illness (ILI) and gastrointestinal (GI). The advantages to reporting these measures directly from the EHRs are:
 * 1.2 User Story Narrative**
 * 1.2.1 Goal**
 * EHRs will have access to more historical data and elements outside the ISDS Minimum Data set in constructing measures. PHAs will need to collect raw data for a significant period of time to obtain historical thresholds with the
 * PHAs may share understanding and situational awareness based on shared measures, in addition to defining their own analytics on the ISDS MDS.
 * Only aggregate data is transmitted outside of the source’s firewall minimizing concerns about HIPAA, PHI, and DURSAs/BAs that must address the requirements for a limited data set. This allows a minimal set of surveillance measures that can represent the “default” reporting obligation for sources/PHAs that may resist sharing more detailed data under other user stories..
 * Cost of implementing these measures is absorbed by the EHR Vendors community, not the PH community. PHAs that are limited in their ability to invest in analytics for the ISDS MDS will be able to receive these summary measures in human readable form.
 * Near real time availability of data- no time delays or consistency issues by different PHAs applying different Syndromic definitions. (The latter was a concern with Distribute)
 * Much smaller bandwidth and data storage requirements, which may greatly assist in scaling up to addressing non-ER and non-urgent care settings.

However, it is equally important to stress what this user story will NOT be able to do.


 * 1) It is NOT a replacement for the analysis of patient-level data, such as in the ISDS MDS.
 * 2) Any detection based on free text fields will be severely limited.
 * 3) These measures will only be effective at identifying pre-defined syndromes. They will NOT. be effective at identifying new health threats that do not conform to the syndrome definition.
 * 4) These measures should NOT be considered of a high enough confidence level to trigger the broadcast automated alerts or notifications (unlike the original Biosense.). They should be viewed as status board indicators to invite more investigation by epidemiologists and public health officials.

Together with the Syndromic Surveillance using UC and ER Data user story, these twin approaches can provider both an immediate, minimal surveillance capability and a deep, flexible surveillance capability.

The abstract model depicts the Syndromic Measure user story where the actors are the participants directly involved in the user story. **Fig 1: Syndromic Measures Abstract Model**
 * 1.2.2 Abstract Model**

Currently, we would envision this data would be collected at ER and Urgent Care facilities, and focus on initial/discharge diagnosis codes and chief complaints, and would not require any data that would not already be in a typical EHR. It would be collected as part of the normal admission/treatment/discharge process- there should be no need to add any collection requirements, or to rely on data that is on paper forms.
 * 1.2.3 Description of Data Reporting Events, Actors and Triggers**

//Pre-Conditions// The Reporting agreement parameters are addressed in the Reporting Registration User Story that is submitted separately.

//Assumptions// A data partnership agreement needs to be in place between the source and the PHA which includes the raw and derived data that PHA is able to share with other PHAs.

//Primary Actors:// Since the reports will be sent on a chronologically triggered process, the actor will be the EHR or a reporting system integrated with it. || //Triggers//: The report is to be triggered chronologically, every 12-24 hours. It may also be desirable to trigger an immediate report of an ICD-9 code corresponding to a grave threat (Hemorrhagic Fever, Smallpox, etc) is entered, but the latter may be framed as a separate User Story.
 * **Actors** || **Description** ||
 * Admissions Agent || Captures most of admissions/demographic data as part of the admissions/discharge process ||
 * Care Provider || The source (provider, practice or facility sending the information) providing initial/modified diagnosis codes, notes, etc.
 * PHA Official or Analyst || Recipient of data ||

An elaboration of this Story could also address summary reports for longer time periods, or to also capture historical norms.

//Use by Public Health:// Ideally, this data would be simultaneously made available to local, state and national PHAs to support situation awareness indicators at all three levels. This data could easily be output in a format that would be easily supported by analytical tools, spreadsheets, geospatial displays and dashboards. It is not intended to replace investigation by public health officials or epidemiologists, but rather to identify regions which may be at risk. In addition, since these measures will use a lot less bandwidth than a surveillance data feed, this approach can be applied to

The discussion of data needs to address core level elements, defintions of syndromes, and the definition of measures. In addition, some suggestions will be offered for the message/document used to transport the measures.
 * 1.2.4 Data (and Standards)**

//Data Elements:// Chief Complaints and Diagnosis Codes are defined in the C83 specification under HITSP/S&I Framework, and can be mapped to IHE conformance profiles and to HL7.

We strongly suggest the that data elements used to construct measures be drawn, as much as possible, from the MDS data elements listed on pages 56-65 in the ISDS recommendations.@http://www.syndromic.org/uploads/files/ISDSRecommendation_FINAL.pdf

//Syndrome Definitions:// Syndrome definitions will be built out of these elements. While we understand that many PHAs will wish to create their own syndrome definitions for their own use, a common definition is important to compare data in different regions without forging a myriad of agreements. The definitions of the syndromes themselves should be based on existing definitions, such as those provided by the CDC at:@http://www.bt.cdc.gov/surveillance/syndromedef/index.asp

It may also be helpful to crosswalk these definitions with those used by the U.S. Outpatient Influenza-like Illness Surveillance Network (ILIN).

Another valuable source for a crosswalk would the ESSENCE study on code-based ILI identification at:@http://wwwnc.cdc.gov/eid/article/13/2/06-0557_article.htm

//Measure Definitions:// A “Measure” brings together all of the elements needed to calculate the how prevalent a given syndrome is in encounters. This includes the inclusion/exclusion criteria. In addition to defining whether or not an encounter fits the syndrome definition, these criteria may also wish to limit the encounters included to those in ER or Urgent care facilities. The measure will also need to reflect data used for the denominator, and possibly support sub grouping by fields such as hour of the day.

Currently, we are unaware of any explicit measure definitions that provide this additional information for the syndrome definitions previously identified. We do suggest that the Health Quality Measure Format (HQMF) be evaluated for defining these measures. HQMF is currently being used by the CMS and National Quality Forum (NQF), and is an XML-based format to specify quality measures precisely enough for automation. The HL7 CDA Quality Reporting Documentation Architecture (QRDA) may be the appropriate document for transmitting these measures, although the QRDA category for reporting measures is being revised for balloting by HL7. The advantages of using HQMF and QRDA are that the CMS is already proposing them for Stage 2 Meaningful Use, and for quality measure reporting. This will allow the PH Reporting Initiative to take advantage of different standards that are being supporting through Meaningful Use going forward.

This user story is based to a large degree on the success of the Distribute program, which was stood up in a few months to support H1N1 surveillance, and was able to provide a limited surveillance capability over a larger patient base then Biosense. The greatest drawback to Distribute program was that each source provided their own syndromic definitions, which made them less comparable. By applying a common set of syndrome definitions, we have an approach that it is lightweight, supports a high degree of federation, and can serve as an example of “low hanging fruit” and an early success for the PH Reporting initiative.
 * 1.2.4 Other Information**

The use of existing standards, where applicable, has been documented in Sections 1.2.3 and 1.2.4.

Due to time constraints, we have not yet approached a stakeholder group to support this initiative going forward. If there is interest in further refining this user story, we would reach out to the Distribute Community of Practice (CoP) and the Primary Care Information Project (PCIP) operated by the New York City Department of Health and Mental Hygiene.
 * 1.3 Stakeholder Commitment**

Any comments/questions may be directed to:
 * 1.4 Contact Information**

John Page Agilex 703-889-3977 John.page@agilex.com

Aleena Dhar Agilex 703-889-3187 Aleena.dhar@agilex.com

QRDA -- Quality Reporting Document Architecture NQF -- National Quality Forum MU -- Meaningful Use CDC -- Centers for Disease Control and Prevention CMS -- Centers for Medicare & Medicaid Services ONC -- Office of the National Coordinator S&I -- Standards and Implementation PH -- Public Health PHA -- Public Health Agencies PHRI -- Public Health Reporting Initiative EHR -- Electronic Health Record HQMF -- Health Quality Measure Format STST -- Surveillance Technical Steering Team HAIR -- Health Acquired Infection Report MDS -- Minimum Data Set ISDS -- International Society of Disease Surveillance NwHIN -- Nationwide Health Information Network POC -- Point of Contact OID -- Object Identifier DURSA -- Data Use and Reciprocal Support Agreement BA -- Business Agreement IRB -- Institutional Review Board PKI -- Public Key Infrastructure HL7 -- Health Level 7 CDA -- Clinical Document Architecture HIPAA -- Health Insurance Portability and Accountability Act HIT -- Health Information Technology SMHP -- State Medicaid HIT Plan XML -- eXtensible Markup Language ER -- Emergency Room IHE -- Integrating the Health Enterprise XDR -- Cross-Enterprise Document Reliable Exchange XDM -- Cross-Enterprise Document Media Exchange PHIN -- Public Health Information Network PHIN MS -- Public Health Information Network Messaging System
 * 1.5 Acronyms**

Supporting Files:

 * **Description** || **File** ||
 * This document contains the initial draft user story submission. || [[file:Initial Draft Submission - Syndromic Measure Reporting - November 18 2011.docx]] ||  ||

Comments:
Please comment on this User Story using the "Discussion" tab at the top of the screen.

include component="page" wikiName="siframework" page="space.template.inc_contentleft_end"