[Skip to Content]
[Skip to Content Landing]
Research Letter
April 2017

Assessing Frequency and Risk of Weight Entry Errors in Pediatrics

Author Affiliations
  • 1Department of Information Services, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
  • 2Division of Hospital Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
  • 3Division of Biomedical Informatics, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
  • 4James M. Anderson Center for Health Systems Excellence, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
  • 5Division of Emergency Medicine, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
  • 6Division of Pharmacy, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio
 

Copyright 2017 American Medical Association. All Rights Reserved.

JAMA Pediatr. 2017;171(4):392-393. doi:10.1001/jamapediatrics.2016.3865

Pediatric patient weights are commonly used by electronic health records to calculate weight-based medication doses. Patient weights, which are frequently user entered, are susceptible to data entry errors, such as digit omission, addition, or transposition or entry of the right data into the wrong patient’s medical record.1 This means that errant weight entries can lead to errors in medication calculation, which has been previously reported.2,3 Outside of case reports, to our knowledge, there is little published regarding the frequency of or the risk associated with these errors.

Methods

The Institutional Review Board of Cincinnati Children’s Hospital Medical Center judged this work to be nonhuman subjects research and provided a waiver of informed consent.

We set out to characterize the frequency of and the risk associated with weight entry errors at our institution. To accomplish this, we developed an electronic trigger tool that was applied to all new weights entered into our electronic health record during the study period to notify members of the study team if a newly entered patient weight varied more than 50% from that patient’s last recorded weight, provided the previous entry was within 30 days of the new weight entry. This temporal restriction served 2 purposes. First, it reduced false-positive alerts by eliminating patients with long gaps between weight measurements who we could reasonably expect would grow more than 50% between weights. Second, this time limit targeted the population we hypothesized to be at highest risk: hospitalized patients who were more likely to have frequent weights entered into our system.

Receipt of this alert then prompted a medical record review of the flagged patient, including past weights, growth chart, medical diagnoses, clinical notes, and other context of the new weight entry in question, followed by determination of the hypothetical cause of the deviant weight entry. Standard causes investigated included common errors, such as digit or unit transposition. Medical records in which more common errors appeared implausible to the reviewer were investigated for other possible sources of the weight entry, such as vital signs or patient equipment noted in proximity to the weight entry. If these values matched or closely matched the entered weight, then the entry was determined to have been errantly transposed in the weight field. In cases in which weight entry errors were identified, follow-up medical record review was performed to determine whether a weight had been corrected and, if so, when this correction occurred in regards to the triggered entry. Other data collected during review included the clinical setting of the weight entry. The χ2 test was used to determine the variance of weight entry errors between settings.

Results

Between October 24, 2014, and July 18, 2015, 135 alerts fired from 401 029 weight entries (Table 1). After medical record review, 135 alerts were determined to be true positives. The 5 false-positive alerts were normally growing infants whose prior weight entry was near the end of our 30-day cutoff. The emergency department had a significantly higher rate of entry errors per weight entry (1.063 errors per 1000 weight entries; P < .001) than other departments, with ambulatory areas having the lowest rate. Table 2 illustrates the frequency of error types noted during medical record review. Most weight entry errors had no clear attribution. Among those with identifiable attribution, pounds/kilograms transposition was most common (68.3%).

Table 1.  
Error Rates by Environmenta
Error Rates by Environmenta
Table 2.  
Weight Entry Error Frequency by Type
Weight Entry Error Frequency by Type

Eighteen of 130 patients (13.8%) with weight entry errors were hospitalized. Of these 18 patients, 16 had weights that were corrected prior to discharge, with a median (range) time to correction of 2.96 (0-380) hours. Notably, 3 of 16 patients (18.8%) had weight corrections that took longer than 24 hours.

Four of 18 hospitalized patients (22.2%) with errant weights were prescribed 20 medication orders prior to correction of their weights. Of these orders, none were based on weight.

Discussion

Our study demonstrates that weight entry errors are relatively uncommon at our institution, and the period of vulnerability for patients with errant weights is small. However, weight entry errors still pose a significant risk to efforts aimed at reducing medication errors, especially in pediatrics, where medications are frequently written based on weight. Urgent and emergent settings appear to be at highest risk for weight entry errors, although the inpatient population is at highest risk for medication errors secondary to errant weight entries owing to the frequency of medication ordering in the inpatient setting.

Conclusions

This study suggests that relatively simple tools can be used to identify weight entry errors before they present safety risks to patients. More sophisticated methods should be investigated to mitigate entry errors that could be harder to detect, such as patients presenting after long absence or lower-magnitude errors.

Back to top
Article Information

Corresponding Author: Philip A. Hagedorn, MD, Division of Hospital Medicine, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Ave, MLC 3024, Cincinnati, OH 45229 (philip.hagedorn@cchmc.org).

Published Online: February 6, 2017. doi:10.1001/jamapediatrics.2016.3865

Author Contributions: Dr Spooner had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Kirkendall, Kouril, Dexheimer, Courter, Minich, Spooner.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Hagedorn, Kouril, Dexheimer, Minich.

Critical revision of the manuscript for important intellectual content: Hagedorn, Kirkendall, Kouril, Dexheimer, Courter, Spooner.

Statistical analysis: Hagedorn, Kirkendall, Kouril, Spooner.

Administrative, technical, or material support: Kirkendall, Kouril, Dexheimer, Courter, Minich, Spooner.

Study supervision: Kirkendall, Spooner.

Conflict of Interest Disclosures: None reported.

Additional Contributions: We thank Monifa Mahdi, MBA (Department of Information Services, Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio) and Rahul Damania, MD (Northeast Ohio Medical University College of Medicine, Rootstown, Ohio), for their contributions to data collection and medical record reviews. They were not compensated for their work.

References
1.
Gupta  A, Raja  AS, Khorasani  R.  Examining clinical decision support integrity: is clinician self-reported data entry accurate?  J Am Med Inform Assoc. 2014;21(1):23-26.PubMedGoogle ScholarCrossref
2.
Bokser SJ. A weighty mistake. https://psnet.ahrq.gov/webmm/case/293/a-weighty-mistake. Accessed March 16, 2016.
3.
Goldberg  SI, Shubina  M, Niemierko  A, Turchin  A.  A weighty problem: identification, characteristics and risk factors for errors in EMR data.  AMIA Annu Symp Proc. 2010;2010:251-255.PubMedGoogle Scholar
×