ELIMINATION OF HUMAN ERRORS IN THE LABELING OF
URINE ANALYSIS SAMPLES
IE 548
Antar GutierrezPaul StelsonMylie Tong
Diane Van Scoter
Outline Labeling Errors – Background Magnitude of the Problem Process Background Urine Analysis Testing UA Sample Labeling UA Sample Labeling for Lab Literature Review Model Development Example Process HMSEM/FMEA Workbook Summary & Recommendations
Labeling Errors - Background ED identified the labeling error problem If it is not fixed, it could jeopardize the
ED’s ability to perform these tests If tests are not performed in ED, they
must be sent to the hospital laboratory, causing:
Delayed Results Longer Stay in ED for Patients Higher Cost for ED Patients
Magnitude of the Problem
Emergency Room Dip test per year Number of containers sent to central lab From 10 to 15 % have issues
Issues 9 out of 10 have 2 labels from different
patients 1 out of 10 had no label on the specimen
container
11,559 6,159 923
831 90
Process Background Walk-in Patients to ED
Check in at front desk where their information is put into the EPIC system
Receive a wrist band to identify them Are seen by a Triage nurse to get their general
information and have it entered into the EPIC system
Patients are assessed based on the ESI 5 level system to determine the order of being seen by a physician
Patients are taken into the ED for examination Generic labels are printed at the front desk
Some
times 4
hours of
waiting time
Process Background (Continued)
Ambulance Patients to ED Generic information is radioed in while the
ambulance is en-route Patient goes directly into ED trauma room Patient receives wristband in trauma room Patient data is entered into EPIC from bedside Generic labels printed at Unit Secretary’s desk
Urine Analysis Testing Testing Decision: Made by Nurse or
Doctor Triage Nurse decides to do Urine Analysis test if
patient has one of the threshold conditions: Age over 60 years Pregnant Abdominal pain Fever Altered Mental Status Psychiatric Patient
Doctor makes decision based on examination in ED or based on results of Nurse authorized urine dip test. A bar-coded label is created for the specific urine test. These samples get sent to the laboratory for full analysis.
Not consistent within the nurses
UA Sample Labeling Specimen container must have a patient
label affixed, it can be a generic or bar-coded label
Generic labels must be brought to the patients room for walk-in patients and from the Unit Nurse’s desk for the ambulance patients
All bar coded, Dr. requisition labels come from the Unit Nurse desk
No documented procedure for movement of labels from the printer to the patient’s room
UA Sample Labeling for Lab
Laboratory will accept UA specimens if: Generic label is on the specimen jar and
bar-coded label is on the double bagging
Generic label is on the specimen jar and copy of doctors order is placed in the double bags
Bar-coded label is placed on the specimen jar and it is double bagged
Literature Review Battles, 1999
Identified 6 sources (root causes) of errors for ED blood transfusions:
Patient Assessment Care Planning Laboratory Procedures Staff Related Factors Equipment Related Factors Information Related Factors
Literature Review (Continued)
Croskerry and Sinclair (2001) Identified unique characteristics of ED leading to
errors: High levels of diagnostic uncertainty High decision density High cognitive load High levels of activity Inexperience of some physicians and nurses Interruptions and distractions Uneven and abbreviated care Narrow time windows Shift work / Shift Changes Compromised teamwork Poor feedback
Literature Review (Continued)
Welch (2006) Identified 4 types of cognitive sources leading to
errors: Availability Heuristic
<the more easily the heuristic can be brought to mind, the more available and the higher the probability associated with a particular event> (Wickens and Hollands, 2000)
Confirmation Bias <the tendency for individuals to seek information that
supports their conclusion and not seek information that conflicts with it> (Wickens and Hollands, 2000)
Coning of Attention <stress effect reduces the breadth of focus, which can
result in a flawed decision making process> (Wickens and Hollands, 2000)
Reversion <due to stress, participant keeps trying same response
even though it has been unsuccessful> (Wickens and Hollands, 2000)
Model Development IDEF0
Modeling software that focuses on information and resource flow and requirements.
User friendly design facilitates model development and relationship awareness.
Our model highlights aspects of the process relevant to the problem, mislabeled UA specimens.
Model progression was an iterative process Our SME Carol Bonnono was crucial to
understanding and modeling the process
Model Development Model Iterations
Weekly inquiries to Carol More information more model detail more
information required Through many inquiries we found that labeling
process had very little standardization or individual task ownership.
Too many different paths/options, none of which have a standard operating routine.
Example ProcessModel Decomposition A-0
Example ProcessModel Decomposition A0: Diagnose patient
Example ProcessModel Decomposition A1: Admit Patient
Example ProcessModel Decomposition A15: Deliver label to patients room or chart
Notes:Destination OptionsPersonnel Options
HMSEM/FMEA Workbook Human-Machine Systems Engineering
Methodology Workbook Tool that helps categorize and define individual
tasks for analysis. Integrates Human Fallibilities Identification and
Remediation Data Base to expose possible modes of failure for important tasks.
Failure Modes & Effects Analysis summarizes all the possible human fallibilities relating to cognition for a specific task. User then extrapolates how the fallibility could manifest for the specific task.
HMSEM/FMEA Workbook Human-Machine Systems Engineering
Methodology Workbook The results from the workbook analysis
highlights potential failure modes. With the potential errors we can postulate
recommendations to limit the manifestations of human errors.
Summary & Discussion Human Error Human Fallibilities Recommendations
Human ErrorPotential Errors Found in UA analysis
project Failure to notice label is needed Failure to collect generic label in chart or
patient’s room Forget to attach label Test specimen may be incorrectly labeled Deposit specimen in test room even when
it’s not labeled Failure to collect specific label Generic label does not match specific
label
Human FallibilityHuman Fallibilities found from HFIRDB
and FMEA process Workload effect Working memory capacity and duration Stress performance influence Bottleneck effect Intramodality performance decrement Retroactive memory interference Stage-defined resource effect Implementation cost bias
Recommendations Establish a threshold for the patients who need
the UA Generic labels are to be accordion folded and
lopped through the wristband and stapled through the margins
Define the responsibilities of delivering the label and collecting the label
Generic labels should be printed on a paper with color
After patient comes from the restroom with the sample then the nurse can get one of the labels from the wristband and label the container
Recommendations Test Room
Big sign in patient’s room which says: IF TEST IS POSITIVE TAKE IT BACK TO PATIENT’S ROOM, IF NEGATIVE DUMP IT!
Shelf with post-it note where the attending can write down patient’s information as a place holder for label
At the Unit’s secretary Desk Print a blank bar-coded label in between
different patient’s labels