Session 10 Standardizing Health Care Data Across an Enterprise

  Standardizing Health-Care Data Across an Enterprise PERTEMUAN Ke - 9 Taufik Rendi Anggara., MT Manajemen Informasi Kesehatan Fakultas Ilmu Ilmu Kesehatan

LEARNING OBJECTIVES

  • Describe the utility of standard reference terminologies in health information exchange.
  • Identify factors to consider when selecting context-appropriate terminology standards.
  • Describe the process of mapping between

    terminologies, as well as identify types of tools

    available to assist in mapping.
  • • Explain why a Terminology Service is important in

    health information exchange, and describe the

    process for its implementation.

  ROLE OF TERMINOLOGIES IN HEALTH

CARE

  Terminologies in health-care support the documentation of observations, treatments, and outcomes that clinicians put in the patient chart—which is increasingly performed using electronic health record (EHR) systems. Although there exist a fair number of “major” or standard terminologies like the NIC and NOC, there are many more terminologies developed for specifc purposes. A local terminology is one that is created for a specifc purpose by a single organization, such as a laboratory, hospital, clinic, or pharmacy. For example, a local terminology may be used by the laboratory supporting a large health system to provide “user friendly” terms to physicians who order the tests.

  

STANDARD TERMINOLOGIES IN HEALTH

CARE

  The use of reference terminologies is the foundation for facilitating semantic interoperability— maintaining data’s original meaning during HIE. Standardized reference terminologies accomplish this by providing a structured, comparable technical language that enables the data to have meaning outside of the originating system. As a result, comparability of data can allow for accurate and consistent measurement, aggregation, analysis, and reporting of information. Terminology standards also allow data reusability, allowing standardized terms to be stored as a single concept and used multiple ways in a variety of applications. Examples of standard terminologies in health care are presented in Table

  9.1.

TABLE 9.1 Commonly Used Standard Terminologies in Health Care SELECTION OF STANDARD TERMINOLOGIES FOR HIE

  Given there are a variety of standard terminologies, HIE leaders must determine which terminology standard(s) should be selected. The process of selecting terminology standard(s) for a given HIE service or context of use requires consideration of several factors, including:

  • underlying health needs and priorities of the health system;
  • overall goals and objectives of the HIE or specifc HIE service;
  • the types or categories of data to be exchanged among a network of systems;
  • the workfow which will generate the data to be exchanged or in which data will be queried.

CURRENT USE OF STANDARD

  

As previously described, a wide variety of reference terminologies are in a

TERMINOLOGIES

mature state of development and are readily available for use in health

information systems [9]. However, despite availability, actual use in the

real-world is reported to be limited, and local terms remain highly

prevalent. For example, Dixon et al. [10] examined the use of LOINC and

SNOMED CT coding in more than 7 million electronic laboratory messages

exchanged in two states between 2010 and 2011. The researchers found

that fewer than 17% of all messages reported by laboratories to public

health authorities utilized a LOINC term to identify the test performed or a

SNOMED CT term to identify the test result. A separate but similar study,

conducted by Dhakal et al. [11], investigated 63 non-federal hospitals’

use of LOINC and SNOMED CT terms in electronic laboratory messages

sent to a national surveillance system managed by the CDC. Analysis of

over 14 million messages from various states and jurisdictions found that

23% of test orders and 33% of test results had missing or non-informative

codes. Lack of standardized terms limited CDC’s ability to classify results

and analyze patterns across jurisdictions at a national level.

  

MAPPING

  Linking terms from one terminology to another is a process called mapping. Mapping can be defned as a correspondence between the concepts or terms in two diferent terminologies [3]. In mapping, the source code is the origin of the map (ie, the dataset from which one is mapping). The target code is that in which the relationship, or equivalence, is being defned [12]. In other words, the target code is that into which the source code is being translated. For example, mapping can be done to link terms from an interface terminology to a reference terminology. In this situation, the interface terminology would be considered the source, and the reference terminology the target.

  

Cont.

  Equivalence

  It should be noted that mapping does not always result in a 1- to-1 linkage from the source code to the target. Instead, many types of equivalence, or correspondence, can be observed, including:

  • 1-to-1 exact, where a single source code is mapped to a single target code with identical clinical meaning;
  • 1-to-1 approximate, where a single source code is mapped to a single target code with similar but not identical clinical meaning;
  • 1-to-many, where a single source code is mapped to multiple target codes with similar clinical meaning;
  • Many-to-1, where multiple source codes are mapped to a

  

Cont.

  

Variations and combinations of these may also occur. An illustration of

the various types of mapping equivalence is presented in Fig. 9.1. At

times, it may occur that a source code may not be translatable or

“mapped” to a target code, resulting in an orphan code. Such an

occurrence may be due to various causes, which have been

elucidated by Lin and colleagues , including:

  • no equivalent concept for the source code exists in the target code;
  • •the source code has an ambiguous meaning, and this lack of meaning

    precludes any linkage into a meaningful target code;
  • •the source code is overly specifc, preventing linkage to a less-

    specifc target code;
  • •the data are narrative “free” text with no meaning outside a specifc

    clinical case;
  • •the source code is institution-specifc, such as one related to internal

    processing procedures, and therefore may have no meaning outside

TABLE 9.1 Types of mapping equivalence.

  

Cont.

  Mapping Process

  The process of mapping between two terminologies is iterative, complex, and can be labor-intensive. Much like selection of terminology standards, it is context-specifc and highly dependent on the use case necessitating HIE as well as the characteristics of the two terminologies. Furthermore, variations in interpretation of data standards are possible, requiring signifcant efort to harmonize approaches prior to implementation. For this reason, it is essential to establish a set of mapping rules to guide the entire process prior to undertaking subsequent steps [17,18] in order to ensure an efcient and efective process. The American Health Information Management Association published a set of best practices that ofer guidance on mapping that might be of use.

TERMINOLOGY SERVICES

  Implementation of Terminology Services

The implementation of Terminology Services within an HIE is a multistep process

that begins with an assessment of available resources and the development of a

project plan. To do so, the following types of activities must be addressed : •Choosing, acquiring, installing, and confguring software and hardware.

  • •Populating the Terminology Service with the required dictionaries and

    terminologies.
  • •Enhancing the point-of-care systems, insurance systems, and/or national

    registry systems that will interact with the Terminology Services so that they can

    communicate with the Terminology Service through the Interoperability Layer of

    the architecture.
  • Documenting the technical information required to support the system.
  • Testing the system to ensure that it is operating as planned.
  • •Developing and implementing policies and procedures required to support the

    system and business processes, and training users on these topics. These

    include processes for system maintenance and backup as well as processes,

    such as loading dictionary updates

TERMINOLOGY SERVICES

  Terminology Service Support

Implementation of a Terminology Service is just the beginning. Once in

place, ongoing, iterative support is necessary to meet the needs of the

evolving feld. Common types of support include:

  • •Operations support: This includes the development and maintenance of

    policies and procedures that are necessary to support business and

    technical processes for the Terminology Service. It may also comprise

    tasks such as evaluating adherence to the selected terminology

    standards.
  • •Terminology support: Standards are modifed and improved over time.

    Newer versions of standards may impact previous mappings due to new

    concepts, retired concepts, or structural changes [21]. Because of this,

    the Terminology Services must have an ongoing analysis of impact of any

    terminology or mapping changes. In addition, the Terminology Service

    should maintain a history (versions, efective dates, etc.) of all the

    terminologies supported so that references to historical data can be

TERMINOLOGY SERVICES

  Terminology Service Support (Cont)

  • •Help desk support: As more and more users and organizations utilize

    the Terminology Services, it may become important to identify,

    document, and triage responses to types of requests from the various

    users. For each request, it may be necessary to track who requested

    it, why it was requested, and when it was requested. Detailed

    information of how the issue was resolved should also be

    documented for reference purposes.
  • •Training of support personnel: Successful implementation of the

    support plan is grounded in efective training of all personnel charged

    with overseeing it

TERMINOLOGY SERVICES

  Evaluation of Terminology Service

Once implemented with necessary support services in place, it

becomes essential to conduct ongoing evaluation of the HIE’s

Terminology Services to ensure efciency and efectiveness of the

operation. Appropriate timing of these evaluations will be determined

by the HIE, depending on need as well as data and resource

availability. For example, they may be conducted annually, quarterly,

or on a real-time basis. Varying aspects of the service can be

assessed, including but not limited to: frequency of codes’ use;

correctness of methods of code validation; frequency, types, and

causes of errors; and prevalence of redundancy/overcoding [3]. Any

concerns or areas of weakness regarding these features or others

should be referred to administrators for correction.

  

SUMMARY

  Data standardization is an ongoing challenge but is key to facilitating semantic interoperability in HIE. Through the mapping of local terminologies to reference standards, efective exchange of clinical data between various entities in the health care ecosystem can occur, allowing for accurate and consistent measurement, analysis, and communication of information. The implementation of standards and Terminology Services is a complex undertaking that requires consideration of context-specifc economic, political, social, and technological factors. Despite challenges, the feld of data standardization is currently at a stage of rapid advancement due to increasing cooperation and international guidance. While much work remains, the future is promising for semantic interoperability.