Report of Consensus

General Statements

Preamble

Primary care represents a complex clinical activity which does not presently have a commonly accepted clinical vocabulary and classification scheme. In order to adequately support the multiple dimensions of primary care (including patient care support, relevant research, and teaching activities), we endorse the concept, development, and evaluation of a standardized nomenclature or vocabulary for primary care.

Framing Issue

A serious problem exists with the capture of structured information in primary care for optimal health care delivery (primarily for patient care and secondarily for epidemiologic and research purposes). There is a need to develop consensus international standards for clinical vocabulary and classification in primary care.

Vision

Our vision is for primary health care providers to record data about patients easily and accurately at the point of care, in such a manner that clinically relevant information is available for the primary purpose of supporting the care of individual patients. Additionally, this information should be comparable with information collected elsewhere in the world.

Standards Statements and Critical Elements

The following characteristics are viewed as essential for a primary care clinical vocabulary. (No rank order is implied.) Must support and enhance the best patient-centered care.

• Is clinically relevant to primary care (and primary health care practitioners).
  • Supports the uncertainty inherent in primary care.
  • Deals with the temporal evolution of clinical problems.
  • Is able to deal with clinically relevant nuances.
  • Is able to deal with pertinent negative terms.
• Must have a rich vocabulary.
  • Is unambiguous, nonredundant, and appropriate.
  • Supports synonyms and homonyms.
  • Supports qualifiers and modifiers of clinical terms.
  • Supports mapping to other code sets to facilitate all aspects of clinical practice and administration.
  • Supports the documentation needed for education and training.
• Must allow easy data collection.
  • Permits seamless and effortless data collection.
  • Permits reliable and valid data input and output.
• Must have a sound architectural infrastructure.
  • Is broadly available.
  • Is sufficiently flexible to deal with multiple levels of granularity (detail).
  • Can deal with hierarchical and multiaxial structures.
  • Can be used internationally.
  • Has adequate structure for maintenance and updates, including version control.
  • Can be extended in a disciplined way.
  • Is multilingual.
  • Is multicultural.
• Must allow consistent use in all possible primary care locations.
  • Can be used in all primary health care delivery settings.
  • Supports work flow in primary care settings.
• Must support aggregation and analysis of data.
  • Supports clinical protocol and guideline implementation.
  • Is suitable for research.
  • Is suitable for decision support.
  • Supports patient-focused outcome measures, including functional status and quality.

Strengths and Weaknesses of Examined Vocabularies

The overwhelming consensus of conference participants was that the development of clinical vocabularies is a process in progress. At this time, it would be premature, misleading, and destructive to recommend any one vocabulary over another. In time, the best qualities of each system may merge to form one system that best meets the needs of primary care. The eventual standard will be either a suite of vocabularies or a blending of the uniquenesses and strengths of existing vocabularies into a single system.

The participants recommended that primary care practitioners map and aggressively use all of the examined vocabularies at the point of service, while researchers use standardized methods to examine and evaluate the vocabularies. Such studies could identify gaps and strengths, which could be communicated among those developing vocabularies. The outcome of such efforts could be a cooperative development, using the strengths of all existing systems. It is in this spirit that the following identified strengths and weaknesses are listed.

ICPC (International Classification of Primary Care)

Strengths

• International.
• Multilingual.
• Widely available.
• Reasonably good for epidemiologic research.
• Developed specifically for primary care.
• Has been used for "on the fly" coding.
• Supports concepts of comorbidity well.
• Supports concept of episode of care.

Weaknesses

• Does not deal well with evolution of disease or disease diagnosis.
• Moderate synonym list; limits level of granularity.
• Does not map broadly to other codes.
• Does not fully support the coding necessary for clinical patient care or administrative needs.

READ Codes

Strengths

• Initially written by a general practitioner for general practitioners.
• Used extensively in practice within the United Kingdom.
• Endorsed by U.K. Clinical Professions in 1994.
• Broad clinical coverage across disciplines.
• Mapping and tool facilities available.
• Implemented by multiple vendors.

Weaknesses

• Uncertainty about version 3.1 integrity and version control.
• Evaluation not complete.
• Available only in English.

SNOMED (Systematized Nomenclature of Human and Veterinary Medicine)

Strengths

• Multiaxial and hierarchical.
• Has both breadth and depth.
• Underlying knowledge representation.
• Significant momentum in the United States.
• Multilingual.

Weaknesses

• Too broad.
• Too granular for routine primary care.
• No syntax for combining terms.
• Efficient use in primary care setting not yet demonstrated on a large scale.
• Not originally developed with primary care in mind.

UMLS (Unified Medical Language System)

UMLS is not a coding system. It is too large and complex for direct use for primary care coding. It does, however, allow health professionals and researchers to integrate data from different sources. The question is: Can UMLS be used as the link to connect the coding systems proposed for primary care?

Strengths

• Huge resources behind development of the system.
• Metathesaurus with substantial breadth and potential to represent the broad range of concepts in primary care.
• Nonmedical words in the specialist lexicon that may be useful in primary care.
• Includes substantial portions of SNOMED.

Weaknesses

• Not specific to primary care.
• No specific primary care sources in current vocabularies, although this is to be remedied.
• Not truly international at present.
• Many exceptions identified in current version.

Glossary

Early in the deliberations, it became evident that participants varied in their understanding and use of terms such as nomenclature, vocabulary, classification, and coding. A glossary of terms is necessary.

Primary Care Focus

We agreed that the main purpose of a clinical vocabulary for primary care is to support patient care at the point of service. Everything else—administration, epidemiology, and aggregation of data—while essential, remains secondary.

Developmental Process

To facilitate the development of vocabularies, an overall process should be designed and implemented. It should include but not be limited to:

• Clearly defining goals.
• Developing and adopting a standardized evaluation methodology.
• Studies of vocabularies already in use in primary care.

Recommendations for Future Actions and Direction

The following recommendations for action and direction are far from complete. However, carrying them out would result in a good beginning in moving toward international standards in primary care.

• Link the three coding systems discussed (ICPC, Read, and SNOMED) in the National Library of Medicine's UMLS.
• Encourage and help developers and users of the four schemes to work with each other toward strategies that focus on cooperative development of coding and classification systems.
• An adequate infrastructure for primary care informatics, including centers for primary care coding, should be established and funded in each country.
• Develop and adopt a glossary of primary care informatics, including a definition and scope of use of the term "clinical vocabulary."
• Organize and conduct a followup conference to continue the work begun at this conference. Request further funding to support a future and expanded conference from all sources, especially the Agency for Health Care Policy Research (AHCPR).
• Identify or develop and adopt standard processes and methodologies to evaluate clinical vocabularies.
• Develop a problem list for each vocabulary that can be used in working toward an international standard.
• Explore the concept and development of a minimum data set for ambulatory primary care.
• Investigate methods and processes for developing the new terms to be added to a vocabulary for primary care.
• Establish a process to update current vocabularies and standards.
• Seek national and international support.
  • Develop a list of strategic national and international organizations to educate about the need for standards for clinical vocabulary in primary care informatics.
  • Request review and endorsement of the conference proceedings by these key organizations.
  • Seek endorsement and agreement from key organizations and agencies.
  • Request funding from key organizations for centers for primary care informatics, with special emphasis on developing clinical vocabularies.
• Share conference outcomes.
  • Publish and disseminate the conference proceedings as soon as possible.
  • Submit proceedings to the AMIA Board of Directors for information, approval, and endorsement.
  • Make presentations and publish articles about the process and content of the conference. Specifically, make presentations at upcoming AMIA conferences.
  • Identify other interested agencies and organizations for possible distribution and/or inclusion in existing databases.
• Build support, linkages, and liaison with other groups.
  • Work with the American Academy of Family Physicians (AAFP) technical panel for endorsement and action.
  • Propose, develop, and/or conduct coding seminars or courses for AAFP, the Society of Teachers of Family Medicine (STFM), private practitioners, and other primary care groups to increase their interest in or knowledge of coding.
  • Develop a strategy for liaison with private practitioners and user groups for recommendations that would match their needs and systems.
• Continue support for the participants and conference process.
  • Continue to frame the issue of moving closer to international standards by continuously clarifying the purpose statement, goal definition, etc.
  • Explore multidisciplinary involvement and consensus, especially an effort to include the entire scope of primary care.
  • Continue regular communication among conference participants so members can keep working as a group, thus having greater influence with established organizations.
  • Share the conference process methodology and encourage use of the process to gain consensus at the local, regional, and national levels.
  • Develop and/or define a research and education agenda in primary care informatics.

Return to Contents
Proceed to Next Section