LISMI Division Issues for Comment

Since completion of the Institute of Medicine study on the Computer-based Patient Record in 1991, the momentum toward the Electronic Health Record (EHR) has been increasing. The 1993 federal effort for "Healthcare Reform" only accelerated an effort that had been already underway for over twenty years. The passage in September 1996 of the Health Insurance Portability and Accountability Act (HIPAA) provided an explicit step to this end by addressing the administrative simplification issues. Use of information technology in the clinical laboratory at the very earliest efforts had already placed it at the core of this major pathway of change. In 1991 the first standard, by ASTM Technical Committee E-31, E-1384 “Standard Practice for the Structure and Content of the HER” was also issued. It had been preceded in 1988 by a messaging standard for the exchange of laboratory data in concert with message structure by HL7. Likewise, an ASTM Standard Practice on the core demographic data (Registration, Admitting, Discharge and Transfer - RADT) that underpins all patient care, both care record and ancillary services including the clinical laboratory, data management functions also was issued by ASTM E-31. During this time there was also much activity at the international level, particularly in IUPAC and the IFCC. In 1991, when the European (CEN) Commission European Normalization effort got fully underway, a significant part of the laboratory-related informatics work shifted to the CEN TC 251 on Medical Informatics, but not entirely. The late 1995 publishing of the IUPAC Silver Book drew on the concurrent CEN and IUPAC/IFCC focus within Europe.

Events in the United States have now matured. After a four year period as a temporary Healthcare Informatics Standards Planning Panel (HISPP), a permanent Healthcare Informatics Standards Board (ANSI HISB) was established by ANSI. It first met in February 1996. The the Association for Diagnostics & Laboratory Medicine (formerly AACC) Laboratory Information Systems Division has actively represented the clinical laboratory up to this point and the College of American Pathologists is also voting member of the new HISB. The the Association for Diagnostics & Laboratory Medicine (ADLM) intends to be a full voting member, which must include a membership in ANSI as a precondition for HISB full membership. The HISB has catalyzed the formation of an ISO Technical Committee (TC-215) in Health Informatics. In this context the ADLM has a clear collaborative role, in concert with its Academy, the NACB, to help lead the formation of the informatics standards that will shape the integration of the clinical laboratory with the EHR domain of the practitioner.

In 1996 the following questions were posted on the NACB Web Page and comments were solicited. They are even more relevant now due to the greatly increased interest in the benefits of information technology due to the passage of the Health Insurance Portability and Accountability Act of September 1996 and the federal agency efforts to encourage health informatics standards uses for the Electronic Health Record (EHR). These questions are reposted here in order to solicit comment on them as a basis for planning educational and informational programs by the ADLM LISMI Division that could be offered jointly with other ADLM Divisions and the NACB.

The Central Questions

1. What essential functions can be identified that are critical to integrating
laboratory services with the delivery of care to patients?

2. What information that is used either in the laboratory or in the patient
encounter supports these functions?

3. What styles of dialog and flows of data can be identified that provide the needed
data to the practitioner when and where it is needed?

4. What supporting knowledge structures can be identified that are part of the
overall information domain common to the laboratory and the practitioner?

5. What data representations and structures that characterize the laboratory and
EHR information domains, support these functions?

6. What informatics skills in the comprehensive skills profile of the laboratorian
enable the provision of the essential functions?

7. How does the laboratory use the information within its domain to manage the
resources needed to provide the essential functions?

Beginning Discussion on Each Question

Respond to LISMI Division contact at email: [email protected]


The key functions so far identified that are critical to functional integration are the Request for Laboratory Services and Interpretation of Results. They have been identified, but not characterized, in the NCCLS 8A (former ASTM Standard E-1639 Requirements for Clinical Laboratory Information Management Systems.. The need is for documentation of these two functions as they appear in different forms within the full range of care settings that will now exist in the EHR Era. The issue here is what capabilities can be provided due to the systematics of the software and which functions need to be facilitated by the software but require active involvement of the laboratorian in a dialog with the practitioner. The assumption is that an effective solution to the request/report of observations functions can also provide for the characterization of the resource management attributes that may be part of the optional selection of clinical laboratory services. Laboratorians will have to explain to resource managers how the associated attributes for this purpose meet the resource management requirements. In the "Capitated" situation there will be different understandings needed to avoid the "Game-Playing" that has attended fee-for-service. Different incentives and rewards in these different financial structures are at work and these incentives must be identified during the characterization of the functions themselves.

Question: What are the key “business” functions for both the EHR and the Clinical Laboratory?

Question 1 Responses and Comments


There is a range of information, both in the laboratory and the practice setting, that attend a request for laboratory services and contribute to the Interpretation of Results. It is organized in the form of a Clinical Order - the formal form of the request. While the general form of a clinical order is recognized in ASTM E-1384 Standard Practice for the Structure and Content of the Electronic Health Record (EHR) and in both ASTM and HL7 standards on messages, the specific attributes needed by the laboratory for clinical decision support are still indistinct in the informatics standards and almost unrecognized consciously by either the laboratorian or practitioner community. Each part of the Clinical Order has a purpose related to effective execution of the request. The concept of Clinical Orders has been generally part of the inpatient (hospital) environment historically but recognition of the function, and its attendant structure, in the non-hospital (that range of non-fully-resident) settings that now characterizes the present healthcare environment has not yet been widely recognized. In addition to the data contained in the Clinical Order structure, there are Knowledge Representation data structures related to the context-insensitive attributes of either clinical laboratory services or other healthcare enterprise attributes that are associated with a clinical order.

Question: What are the key Information Functions and Data Structures needed for the EHR and the Clinical Laboratory?

Question 2 Responses and Comments


The closely related question to the nature of the information is that of the style of communication. The objective is to get from the practitioner to the laboratorian a clear, accurate and timely request relating to the clinical situation of the individual patient and then to return to the practitioner a full interpretation of the meaning of the observation(s) made by the laboratory. In the current and future telecommunication environment, the style of interaction will be a major factor in the success of this communication. For clinical laboratorians to also be able to communicate with the system designers and software engineers who will build such environments, the first priority is to clearly and completely document the required style of interaction. NCCLS has a standard, GP-19-A “Laboratory Instruments and Data Management Systems: Design of Software User Interfaces and End-User Software Systems Validation, Operation, and Monitoring” that was initially stated for primarily for instrumentation. It is being referred to by the NCCLS LIS8A standard (prior ASTM E-1639) “Functional Requirements for CLIMS” as a source of general principles for user interaction. Much more specific design guidance must be given by laboratorians to implementers that are involved relating to specific care settings, such as Point-Of-Care-Testing. NCCLS will be addressing documenting implementation-independent dialog "Views" as part of the work program of the NCCLS Area Committee on Automation and Informtics. Working Laboratorians MUST be involved if the needed concepts are to be effectively documented that describe the appropriate dialog style.

Question: What are the key dialog styles that make the system valuable to users?

Question 3 Responses and Comments


The knowledge structures that support the essential functions must be understood to be those context-insensitive (to the individual patient) constellations of attributes that represent the published research and experience that has, until now, been gathered via various print and electronic compendia. One example is the current CPT4 lexicon of laboratory procedures by which we identify what is requested and reported, what methodologies we use at what expense and effectiveness, what is charged to the eventual payer and what interventions lead to what effects that are documented in the research literature. Another example is the "Disease" lexicon, typified by the ubiquitous ICD-9-CM by which all health conditions are now officially categorized but which is being replaced by the ICD-10-CM. The diagnosis vocabulary has had a major impact on the integration of the laboratory with the practice setting. We can no longer leave these knowledge structures solely to the health policy statisticians. Without laboratorian's input, the statisticians can do little more to correct its impacts on the interpretation of laboratory observations. These are but two familiar examples. Such vocabularies and knowledge structures can not just be left to "industry" or to "government". They must be broadly based. The ADLM will be pushing for the creation of new software tools to expedite the development of much improved terminologies.

Question: What are the associated Knowledge Structures needed for integration of the EHR with The Clinical Laboratory?

Question 4 Responses and Comments


The data representations and structures in both the practice environment (EHR and associated records) and the laboratory ( Clinical Laboratory Information Management Systems - CLIMS) need to be clearly identified. All of the concepts that we manipulate can be dealt with either as "data elements" - a generic classification of concepts which have value sets such as a numeric or lexical value that represents a concept. Some of the data that we record is done in the practice setting, some in the clinical laboratory and some in other settings such as the home or an environmental laboratory. We must understand the nature of the data that are captured and its representation. One problem area unresolved by the IFCC/IUPAC work is the nature of ordinal scales, decision points, the measurement process, uncertainty of values and its representation. Work of the ISO TC-212, for which NCCLS is Secretariat, is addressing this and the related issues of reference materials. Explicitly and formally, by means of modeling, we must describe the data in the information domains that are common to the integration of lab and EHR. Already, early models for the RADT of the EHR and for CLIMS exist in ASTM E-1715, NCCLS LIS-8A (prior ASTM E-1639), LIS 9A (prior ASTM E-2118) “Standard Guide for Coordination of Clinical Laboratory Services Within the Electronic Health Record Environment and Networked Architectures” standards resp. The HL7 message standards developer is also creating a formal model to match with the CEN messaging model. Laboratorians will need to understand what these formalisms say and what the mean in terms applicable to the laboratory and the EHR. Laboratorians will need to know how these will be used by implementers, even if they do not involve themselves in implementation. This means that laboratorians must understand the implementation process if they are to master the direction of acquisition of information systems products and services or are involved in the creation of software tools having the capabilities supporting the key functions identified in Question 1. Since the key part of this process is Requirements, the critical role of the laboratorian is understanding content and its conventions as represented by data representations and data structure. Mastery of data Structure and Representation will enable mastery of Requirements.

Question: What are the key data structures (Data Objects) and Data Representations (Vocabularies) needed for integration of the HER and the Clinical Laboratory?

Question 5 Responses and Comments


One key aspect of the practice of the profession of the clinical laboratorian will be understanding how the essential functions in Question 1 are conducted using the skills of the performers. The laboratorian will need to know who has the skills and knowledge to perform each step in the requested services and so will need to know how to characterize not only his/her own skills and knowledge profile but also the profiles of others in the affected domain. This managerial skill will include the knowledge as well as psychomotor skills and integrative behavioral skills that are needed in management. The laboratorian will also need to know how to characterize the work as well as the people. Currently for scientific and technical occupations, no structured knowledge base is organized to do this in the way that will be needed in the clinical laboratory. Having characterized the skills profiles for the essential functions to be provided to the practitioner community, the laboratorian will need to know, and defend to the resource mangers, the proposed staffing pattern. Informatics skills will be key to the needed management skill profile. Thus, the laboratorian must "First, Know Thyself".

Question: What are the key knowledge and psychomotor skills in informatics need by the clinical laboratory specialties?

Question 6 Responses and Comments


Management of the clinical laboratory will draw on data first recorded either in the EHR or in the laboratory. Current resource management emphasis on "outcomes" and "practice guidelines" must be more strictly founded in understanding what management criteria are used for what decisions. Clinical Laboratory management must be founded in how the management data originate in the initial observations recorded in the process of providing the essential functions. Resource Management will build on the answers developed in Questions 5 and 6 and it will develop the management data needed to support those criteria that are derived from the original observations. It will help rephrase management questions (such as policy and epidemiology) in terms of those original observations. If original data do not exist, management will act to introduce original observations that not only provide the support for management decisions but also have a driving force in support of the essential patient care functions.

Question: What key resource management information is needed by the clinical laboratory?

Question 7 Responses and Comments