Original Article
Diabetics can be identified in an electronic medical record using laboratory tests and prescriptions

https://doi.org/10.1016/j.jclinepi.2010.04.007Get rights and content

Abstract

Objective

With the increasing use of electronic medical records (EMRs) comes the potential to efficiently evaluate and improve quality of care. We set out to determine if diabetics could be accurately identified using structured data contained within an EMR.

Study Design and Setting

We used a 5% random sample of adult patients (969 patients) within a convenience sample of 17 primary care physicians using Practices Solutions EMR in Ontario. A reference standard of diabetes status was manually confirmed by reviewing each patient's record. Accuracy for identifying people with diabetes was assessed using various combinations of laboratory tests and prescriptions. EMR data was also compared with administrative data.

Results

A rule of one elevated blood sugar or a prescription for an antidiabetic medication had a 83.1% sensitivity, 98.2% specificity, 80.0% positive predictive value (PPV) and 98.5% negative predictive value (NPV) compared with the reference standard of diabetes status.

Conclusion

We found that the use of structured data within an EMR could be used to identify patients with diabetes. Our results have positive implications for policy makers, researchers, and clinicians as they develop registries of diabetic patients to examine quality of care using EMR data.

Introduction

What is new?

  • Identification of diabetics in Canada at a population level has historically relied on physician billing and hospitalization administrative data. With the uptake of electronic medical records (EMRs), more detailed clinical information is available, but methods to identify patients with particular disease conditions are needed.

  • Using a combination of prescriptions for antidiabetic medications and laboratory tests results, diabetics can be identified within an EMR with accuracy similar to administrative data.

  • Use of these structured variables within an EMR will allow for easy identification of diabetics within a large population with an EMR record.

In Canada, the identification of diabetics within a population has traditionally relied on administrative data, namely, a combination of physician billing data and hospital discharge abstracts [1], [2]. Although administrative data is useful for monitoring prevalence and incidence within large populations, the lack of availability of comprehensive clinical information limits its utility in measuring more detailed aspects of diabetes management and care. With the increasing use of electronic medical records (EMRs), identification of diabetics within an EMR could greatly aid physicians in monitoring the management of this complex chronic disease as prescriptions and laboratory results are readily available.

The use of EMRs in Canada, although growing rapidly, is only in its infancy compared with many other countries. With approximately 26% uptake of EMRs in Canada [3], dozens of vendors and no physician incentives to enter or code data in a standardized format, the use of EMR data for research purposes may be challenging [4]. Thus methods to identify patients with particular disease conditions within an EMR need to be developed.

Searching the cumulative patient profile of a primary care EMR to identify diabetics theoretically should be a straightforward process. However, physicians vary in the degree and manner in which they populate their EMRs, and they often use different acronyms. This inconsistency can increase the complexity of identifying diabetics through simple text searching. Diabetes is a condition with discrete laboratory value tests that can make a definitive diagnosis, and medications for treatment are fairly specific. Thus, the nature of this condition may lend itself well to identification of affected patients through analysis of structured data within an EMR.

Using EMR data that is linked to administrative data, we set out to determine if diabetics could be identified through structured EMR data and/or more accurately than administrative data.

Section snippets

Methods

Through the Canadian Cardiovascular Outcomes Research Team (CCORT), we collected data from 17 physicians using Practice Solutions EMR (Ottawa, Ontario) for a minimum of 2 years. Data from the EMR was extracted, encrypted, and securely transferred to the Institute for Clinical Evaluative Sciences (ICES) from June to December of 2007. ICES is a prescribed entity under the Ontario Privacy and Health Information Protection Act, and the data was handled as per standard operating procedures to

Results

All but one of our 17 convenience sample physicians distributed throughout Southern Ontario were in group practice. The average length of time using Practice Solutions EMR was 7.3 years, and the average years in practice were 20.5. The average age of the 5% random sample of our active adult cohort that was chart abstracted was 49.6 years, and the gender split was 58.1% female. All of the algorithms had high specificity and NPV and varied mainly in sensitivity and PPV.

The criteria of only one

Discussion

The results of our study show that the presence of diabetes within a general patient population on an EMR can be identified with a reasonable degree of accuracy through structured data contained within the EMR. Interestingly, we found that the use of administrative data performed slightly better than the structured data.

We did not find as high a sensitivity or PPV found in other studies using EMR data in the United Kingdom [10], [11], where general practitioners code in disease conditions or in

Acknowledgments

Thank you to Adenieki Mornan for her assistance in the preparation of this manuscript. This work was funded by a Canadian Institutes of Health Research Team Grant in Cardiovascular Outcomes Research to the Canadian Cardiovascular Outcomes Research Team (CCORT; www.ccort.ca).

This study was supported by the ICES, which is funded by an annual grant from the Ontario Ministry of Health and Long-Term Care (MOHLTC). The opinions, results and conclusions reported in this article are those of the

References (13)

  • M. Wilchesky et al.

    Validation of diagnostic codes within medical services claims

    J Clin Epidemiol

    (2004)
  • Public Health Agency of Canada. Report from the national diabetes surveillance system: diabetes in Canada, 2008...
  • J.E. Hux et al.

    Diabetes in Ontario: determination of prevalence and incidence using a validated administrative data algorithm

    Diabetes Care

    (2002)
  • National Physician Survey. National results by FP/GP or other specialist, sex, age and ll physicians. The College of...
  • T. Mitiku et al.

    Using data from electronic medical records: theory versus practice

    Healthc Q

    (2008)
  • Institute for Clinical Evaluative Sciences. Privacy code: protecting personal health information at ICES....
There are more references available in the full text version of this article.

Cited by (0)

Dr Manuel holds a CIHR/PHAC Chair in applied public health.

View full text