1
2
3
4
Best example is PH response to question are IZ standards ready. Show you the PH value of several specific components. Need for us to be ready with the RCKMS for extension of interoperability gains similar to IZ 5
6
7
8
9
10
11
12
13
14
15
16
1380 Mumps reports in INPC, equals 36 cases per 100,000 when the MMWR rate is <1/100,000 CDC 2014 Mumps Cases and Outbreaks From January 1 to April 18, 2014, 332 people in the United States have been reported to have mumps. Outbreaks in at least two U.S. universities have contributed to these cases: 17
18
19
20
21
22
RCKMS Overview (5 min) History Scope Timeline Architecture (5 min) Context Use Case Overview flow Component diagram Application architecture Pilot (10 min) Features included Pilot jurisdictions & conditions Criteria and value sets Production implementation (15 min) Components to be included Partner engagement Governance External interfaces PHIN VADS, other terminology providers Ontology Vocabulary Summary (5) 23
24
25
26
Design Objectives Format for specifying computable CDS knowledge Knowledge can be importedinto existing CDS systems Not creating a new execution format Format must be flexible Support different CDS 27
intervention types More than alerts and reminders Knowledge must be portable 27
Need Standard CDS data model that is simple and intuitive for a typical CDS knowledge engineer to understand and use Relevant Prior Work Evaluated HL7 Consolidated Clinical Document Architecture (C-CDA) HL7 Quality Reporting Document Architecture (QRDA) 28
HL7 Fast Healthcare Interoperability Resources (FHIR) HL7 Virtual Medical Record (vmr) IHC Clinical Element Models, OpenEHR templates, others Decision HL7 vmr with templates derived from C- CDA and QRDA 28
29
Note that these are implementers of CDS only the first one is DSS vmr compliant 30
31
32
33
34
Reporting of conditions is largely manual Jurisdictions create their own rules for reporting levels of specificity differ typically only human-readable Reporters have great difficulty finding, interpreting and implementing the correct rules Rule changes are not communicated timely or effectively to reporters Difficult to automate detection and electronic reporting Process is labor-intensive, inefficient and often results in incomplete, slow reporting 35
36
37
Who, What, When, Where and How of Reporting Who is required to report (e.g., Hospital, Healthcare Provider, Lab) What-information should be used to decide if a report needs to be made When should the report be sent (e.g., 2 hr, 24 hr, 10 days) Where should the report be sent (e.g., local HD, state HD, and where within the HD) How should the report be sent (e.g., ELR, phone, fax, mail) What link to specification for information that should be included in the report 38
39
40
For the translation, we used the HeDSchema Framework Artifact Utility to perform the translation to the AllscriptsCREF specification. We developed the translation extension for CREF as part of the pilot. As far as lessons learned, I think one thing that became clear is just how important the Value Sets and Terminologies work is. Where there is an established value set that both sides can reference, semantics can be correctly established and preserved through the translation, so that's an indispensable aspect of the HeD effort thus far. Also, the discussion about "canonical" versus "covering" representations of concepts is something that I think is worth having. Even with the terminology mapping, that is something that needs to be considered as part of authoring content, as well as consuming content. This may have been more of an issue on the NQF-0068 pilot, but I think it did come up on the SDI pilot as well. 41
42
43
44
45
46
47
For the NQF-0068 measure, we found during the translation that different source systems represented the same concepts using different classes. For example, a substance administration may be represented with a SubstanceAdministrationEvent, or with Procedure, depending on the source system. So the choice we had was to either change the artifact so that it looked for both types, or require the source systems to produce consistent data. Obviously the first option is easier, but it does raise the question of how best to handle this, and whether the artifacts should be written "canonically" or changed to accommodate these variances as a practical solution to the problem. 48
49
Ken 50
Notes: We want to be careful about the piloting of UC2 the wording should be that we piloted different parts of HeD specifically vmr and an earlier version of the standard 51
Ken 52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
XPath, the XML Path Language, is a query languagefor selecting nodesfrom an XML document. In addition, XPathmay be used to compute values (e.g., strings, numbers, or Booleanvalues) from the content of an XML document. XPathwas defined by the World Wide Web Consortium(W3C). [1] 71
Expected Outcome Document and Disseminate Lesson Learned Define Challenges and Opportunities Develop a Reference Implementation Community-empowerment through reusable components Technical Assistance Inform ONC s S&I Initiative Meaningful Use Stage III National Electronic Disease Surveillance System cooperative agreement support* Evaluate recipient activities Develop meaningful measures 72
73
74
75
76
77
78
79
80
SEE IF YOU CAN BLOW THESE BULLET POINTS (MAJOR) BY EVEN ONE FONT SIZE! First, lets take a look at the New York City pilot, which is overseen by the PCIP, a bureau within the New York City Department of Health and Mental Hygiene (DOHMH). The PCIP was created in 2005 in partnership with the Health IT vendor, eclinicalworks, with the mission to develop and implement a public health-enabled EHR in ambulatory care practices, serving medically underserved communities. Queries are sent through a network data partner, New York City Hub Health System. The data sources for this query health system are 751 total practices, employing 4207 providers and serving an estimated 2.5 million patients. The goal of this pilot is to use query health to expand population health monitoring in New York City to improve understanding of population health quality/performance measures, chronic disease trends (diabetes, hypertension), infectious disease outbreaks, and to incorporate the essential technical and operational elements from the Query Health pilot project into the statewide health information exchange architecture, known as Statewide Health Information Network for New York, orshin-ny. 81
Theinformation requestor in the NYC query health pilot project is the PCIP Project at the NYC Department of Health. The queries are submitted through a Network Data Partner, the Hub System-which acts as an intermediary between the NYC Department of Health and the network of independent ambulatory care practices. In this infrastructure, the Hub System builds the queries, distributes them to the network of practices, collects the data from the practices, and returns it in a de-identified, aggregate form to the NYC Dept. of Health. 82
Aprimary driver/leader of the query health project in NYC is Dr. Michael Buck. This slide presents a mapgenerated using aggregated Query Health data that shows obesity prevalence in 13 areas of NYC that includes multiple boroughs. In this study, obesity was defined as individuals whose body mass index (BMI) was over 29.9. When the data were mapped, it was clear that the prevalence of obesity was not evenly distributed across the areas studied. Populations living west of the East River had a higher prevalence of obesity than populations living east of the river. This is a real-life example of how query health data can be used by local public officials to design interventions to target populations at highest risk. 83
Thisdiagram displays the infrastructure of the MDPHnetinitiative. Therequester is the Massachusetts Department of Public Heath, and it issues a query through the PopMedNet system. PopMedNetis analogous to the NYC PCIP Hub Health System, in that it distributes the query to each responder participating in the network. Thequery then goes into a software called ESP.ESP was developed at The Harvard Center for Excellence in Public Health Informatics, and it is the ESPnetdata architecture that allows the network to work across various EHR systems. The ESP software, which is installed behind each responder s firewall, automatically queries data at the individual practice level and receives a standardized nightly data upload from the responder s EHR. The de-identified and aggregated data are then returned to MDPH through PopMedNet. 84
Mr.Josh Vogel, Epidemiologist at The Massachusetts Department of Public Health, has kindly provided us with two examplesof the preliminary results from the Massachusetts Query Health Pilot. This slide shows ILI related clinic visits from September 2009 to March 2012, including the number of flu vaccinations were administered during this time period. Data such as these can be used by public health officials in the State of Massachusetts to view trends in these two flu-related public health measures over time. 85
The second example provided by Mr. Vogel was a gestational diabetes pilot. In this case, MDPH used query health data to assess the effectiveness of PSA announcements designed to encourage pregnant women to get tested for gestational diabetes mellitus (GDM). MDPH sponsored a media campaign from the 1 st -25 th of June in 2011 and then collected aggregate data from responding providers at MLCHC and AtriusHealth before and after the campaign. The goal was to use the number of HbA1C tests ordered before and after the PSA campaign as a means of assessing the media campaign s effectiveness. 86
87