Frequency-based LOINC collections

Empirically-derived collections and and other data about LOINC usage in information systems.

Value set of universal laboratory order codes from LOINC

The purpose of this value set is to define a set of universal laboratory order codes that would be recommended and adopted by developers of order entry systems for delivery in HL7 messages to laboratories, where they could be understood and fulfilled. This value is designed to cover greater than 95% of the test ordering volume in the US, and was developed with both empirical and consensus-driven approaches. This list is referenced by the HITSP C80 Clinical Document and Messaging Terminology Construct in Table 2-96 "Laboratory Order Value Set". C80 states that it "should be considered a minimum 'starter' set" and "does not attempt to include all possible LOINC codes" or all possible lab orders.

Comments and suggestions are welcome at commontestandorders@mail.nih.gov

Value Set

• Common Lab Orders Value Set (Version 1)
  File Version: Common Lab Orders 1 | File type: application/pdf | Release Date: 2010-01-14
  OID: 1.3.6.1.4.1.12009.10.2.2

HITSP C80

• HITSP C80 - Clinical Document and Message Terminology Component
  Table 2-97 "Laboratory Order Value Set" reference to the LOINC Common Lab Orders value set
  

Common LOINC Result Codes

Prioritizing the mapping effort on the modest subset of most common laboratory result codes is a convincing way to lower the barriers to interoperability in a way that reduces the effort without dramatically sacrificing the rewards. One empiric analysis has shown that approximately 20% of laboratory observations (typically less than 800 codes) accounts for 99% of the result volume, and that this same set of common tests accounts for all the results of more than 99% of patients.

Common Result Codes in LOINC and RELMA

Based on an analysis of 3 years (2006-2008) of laboratory result data from the Indiana Network for Patient Care (INPC), one of the oldest and largest health information exchanges in the US, we calculated the relative percent volume by test result. The values of percent volume for LOINC codes mapped to tests accounting for 99.5% of the INPC volume are stored in the INPC_PERCENTAGE field of the main LOINC table. This includes about 1090 LOINC terms. Within the RELMA mapping program, the "common lab tests only" filter limits searches to only return codes in this common test list.

Future Work
Regenstrief and the LOINC Committee plan to synthesize data from a more representative sample of result volume for the purpose of creating an enhanced "common result code" value set. This expanded analysis would be included in the LOINC and RELMA distribution similarly to how the INPC data currently is. 

Recommended Reading

• Vreeman DJ, Finnell JT, Overhage JM. A Rationale for Parsimonious Laboratory Term Mapping by Frequency. Proc AMIA Symp. 2007;771-775.