View source for Minimizing Electronic Health Record Patient-Note Mismatches

This is a review of ''Wilcox, Chen, and Hripcsak's'' 2011 article, ''Minimizing Electronic Health Record Patient-Note Mismatches''.<ref name="Wilcox 2011"> Wilcox, A. B., Chen, Y. H., & Hripcsak, G. Minimizing electronic health record patient-note mismatches. doi: 10.1136/amiajnl-2010-000068. http://jamia.oxfordjournals.org/content/18/4/511</ref>

== Background ==

The authors of the article wanted to reduce the occurrence of clinicians documenting notes on the wrong patient’s electronic chart, which they referred to as “patient-note mismatch.” They learned about these mismatches from physician self-reports. 

In a patient note-mismatch, health and diagnosis information from a patient is placed into another patient’s health record. The authors were interested in the study because of limited information on Adverse Events (AE) <ref name="Adverse Events"> - Adverse Events (AE). He, Y., Sarntivijai, S., Lin, Y., Xiang, Z., Guo, A., Zhang, S., ... & Smith, B. (2014). OAE: the ontology of adverse events. J Biomed Semant, 5(1), 29.</ref>  caused by patient note-mismatch.  

== Methods ==

The authors noted the difficulty and time-intensiveness of manually reviewing the electronic records to obtain an accurate measure of the note mismatches.  They decided to estimate the occurrence of these mismatches by limiting their sample to patient admission notes: they would consider a note to be a mismatch if the gender of the patient mentioned in the note is different from the patient’s actual listed gender.  The occurrence of the mismatch was noted to be 0.5%.

== Solution ==

User Interface Change:
To decrease the occurrence of patient-note mismatches, they designed a dialogue box that would pop up and ask the physician to verify that they are saving the note on the correct patient.  It prompts them to reenter their password as a method of confirmation.

The use of  [http://en.wikipedia.org/wiki/Natural_language_processing Natural Language Processing (NLP)] was proposed as a means to decrease the rate of patient note-mismatch.  This tool can be used as a warning flag to alert the user.  

== Results ==

The change in the user interface decreased the amount of mismatches from 0.5% to 0.3%.  The number of physician self-reported mismatches decreased by 43%. 

Results from the study did indicate a low incidence of patient-note mismatch. However, the study emphasized that further reduction from the low incidence rate could be achieved through the use of Natural Language Processing (NLP) serving as a warning flag to alert the user.

== Comments ==

The introduction of a confirmation dialogue box was an effective [[Importance of Workflow Analysis During Physician Office EMR Implementation|workflow]] change that resulted in the reduction of the amount of patient-note mismatches. Asking for the user’s password as a means of confirmation is more than what is necessary.  As an analogy, online banking does not ask for your password every time you make an online transaction. Clicking a button to confirm would have been an adequate hard stop.  

The authors mentioned another feature of the EHR that should have been changed to further reduce documentation errors: the screen where clinicians write their notes in the electronic medical record is separate from the interface for viewing patient data, therefore there is no visual as to which patient they are documenting on.  A better interface would be one where the patient’s picture and demographics is visible on all pages, i.e., name and medical record number can be seen on the top left of each page.

== References==
<references/>

[[Category:Interface, Usability and Accessibility]]
[[Category:Reviews]]
[[Category: EHR]]