Strategies of Clinical Data Entry

From Clinfowiki
Jump to: navigation, search

Phenotypic Rich Data Can Be Used for Medical Research During Routine Use of the EHR

Medical research advances our understanding of disease and the effect or harm with treatment. Detailed and rigid study protocols are created to test a hypothesis using a highly selective group of patients.

The studies are extensive and require outside funding, use detailed and unique documentation forms cared for by specially trained medical and research personnel. It is frequently difficult for physicians to apply findings from these studies when caring for typical and unselective group of patients.

Advancements in computer systems and electronic health record (EHR) as well as the federal government’s through the Office of National Coordinator (ONC) push towards all physicians using EHR over the next several years has increased the percentage of physician practices with these systems. [1]. Clinical data stored in the EHR is now available for use in ways not possible with paper. There is currently no standard clinical documentation structure or method of data entry such as free text typing, dictating, or with commercial templates.

Methods have been devised to gather and interpret unstructured data, using natural language processing (NLP)(2), or programs such as MedLEE(3), rbChart(4) converting it to structured languages, or standard vocabularies and terminologies (UMLS, ICD, SNOMED-CT).

As the number of EHR systems in physician practices increases the methods for data entry and coding will become standard or diverge depending on how the systems are certified and meaningful use is defined. The potential for dramatic increase of EHR data used in research will benefit from better coded data provided by the EHR itself or improved methods of post processing listed above.

We have created a method for data entry using PowerForms in a Cerner commercial EHR that creates data elements stored in a business objects database that are pulled in to create a document for the visit and also are available for later statistical analysis.(5)


References

  1. Centers for M, Medicaid Services HHS. Medicare and Medicaid programs; electronic health record incentive program. Final rule. Fed Regist. 2010/08/04 ed. p. 44313-588.
  1. 1
  2. de Lusignan S, van Weel C. The use of routinely collected computer data for research in primary care: opportunities and challenges. Fam Pract. 2006 April 2006;23(2):253-63.
  3. Chiang JH, Lin JW, Yang CW. Automated evaluation of electronic discharge notes to assess quality of care for cardiovascular diseases using Medical Language Extraction and Encoding System (MedLEE). J Am Med Inform Assoc. May 1;17(3):245-52.
  4. Lawson Jr W, Lobach DF, Hales JW, Russell JH, Guilarte WD, Haisty CK, et al., editors. Simultaneously Collecting Coded Data and Free Text Clinical Narrative Using a Content-Independent Data Model: The rbChart System1999: American Medical Informatics Association.
  5. Stone CD. Unpublished data. 2011.

Submitted by Chris D. Stone