Short communication
Coded entry versus free-text and alert overrides: What you get depends on how you ask

https://doi.org/10.1016/j.ijmedinf.2010.08.003Get rights and content

Abstract

Purpose

A key trade-off in computerized clinical documentation exists between collecting coded data versus free-text. Coded data are more readily computer-readable and easier to reuse in different contexts. However, clinical information often exceeds the scope of commonly available terminologies, and coding may be resisted by providers. Alert override reasons are one domain for which agreed-upon terminologies are rarely used. Few data are available on how the collection of information affects the responses of providers.

Methods

We took advantage of a natural experiment and compared coded and uncoded reasons for drug–drug interaction (DDI) alert overrides entered in two inpatient prescribing systems with an identical DDI database but with one system offering coded reasons and the other free-text entry. We only included alerts which were issued in both sites and which physicians had to acknowledge.

Results

Over a one-year study period, 15,636 alerts were issued. The reasons for override entered in the coded approach matched the free-text site in only 46%. When using free-text, physicians provided many reasons not among the coded options, and often reported that they considered the alert inappropriate, including their rationale regarding this. However, the information entered as free-text included many typing and spelling errors, and the same concept was often represented in different ways, e.g. 209 different ways in which “will monitor as recommended” was noted.

Conclusions

The reasons for alert override vary substantially according to the data entry type, which implies that data entry choice may lead to substantial distortion of the underlying data.

Introduction

In healthcare-related documentation, the increasing amount of clinical data combined with the need to reuse and exchange information underscores the importance of how data are collected. Large parts of clinical workflow, including diagnosis, drug treatment, and patient monitoring have been successfully transferred from paper-based to computerized documentation [1]. Apart from the potential for saving time [2], computerized documentation facilitates standardization of documentation, which should in turn enhance the data comparability, facilitate its reuse, and potentially improve the capture of key details [3], [4]. Standardization can be achieved by inclusion of structured vocabularies which refer to distinct terms or codes as well as prephrased sentences which are accessible via drop-down menus or activation of radio buttons [5].

However, clinical conditions are complex, and good vocabularies do not exist for all domains [6], [7]. Thus, computerized systems often allow the (additional) capture of information into free-text fields, which enables collection of a richness of detail not possible with coded entry. However, if not controlled for plausibility, such coexistent use can result in contradicting information [8], and indeed, Singh et al. have documented hazardous discrepancies between coded and free-text entries in a computerized physician order entry platform [9]. Hence, coded entry can be particularly problematic when the fields do not sufficiently cover the domain the user seeks to describe.

In order to assess how information entered by using a coded approach compared to free-text differs when only one data entry format is available, we took advantage of a natural experiment, and assessed the acknowledgement of drug–drug interaction (DDI) alerts in two different inpatient settings, both containing the same DDI alerts but one with coded and one with free-text entry only.

Section snippets

Description of study sites and data collection

We retrospectively compared the textual information by which physicians acknowledged a DDI alert in two electronic prescription systems. Both systems were implemented and in long-term use for inpatient care in large tertiary care university hospitals. The underlying DDI knowledge base was identical for both systems.

For DDI alerts of major severity, the physician could in response to the alert in both systems either cancel the prescribed drug or keep the prescription and specify a reason.

Results

In system A, 6526 DDI alerts were issued, and physicians cancelled the current or discontinued the pre-existing drug in 586 and 812 cases, respectively (in total 21.4%). Thus, the physician prescribed both drugs concurrently in 5128 cases, for each of those selecting one or several prephrased reasons. In system B, physicians cancelled the prescription of the drug triggering the DDI alert or discontinued the previously prescribed drug comparably frequent in 971 and 876 of 9106 DDI alerts (in

Discussion

We analyzed a large dataset of drug interaction alert overrides to assess how physicians acknowledge electronic warnings when they were offered coded reasons versus free-text to override an alert. The underlying DDI database was identical and the two systems were implemented in comparable inpatient care settings. While physicians decided to keep both or cancel either one of the interacting drug with comparable frequency, the reasons they used to justify keeping an interacting medication order

Conclusion

These results illustrate some of the pros and cons of coded versus free-text entry. Both approaches have advantages and disadvantages. Coded entry—while desirable in many ways—can create serious distortions if the reasons that users are looking for are not among the available options. If coded entry is to be widely used, it must be monitored regularly, and this study points out some of the measurements which should be carried out in an ongoing way.

Acknowledgements

This work was supported in part by a fellowship within the Postdoc-program of the German Academic Exchange Service (DAAD) Bonn, Germany and by the Health Information Technology Center for Education and Research in Therapeutics at Brigham and Women's Hospital, Boston, MA, supported by the Agency for Healthcare Research and Quality, Rockville, MD.

Summary points

What was already known on the topic?

  • Coded entry in computerized physician order entry systems is time-saving and can enhance system

Cited by (0)

View full text