Kaiser Permanente Experience with Automating the IHI Global Trigger Tool


Nov 5, 2013 (4 years and 8 months ago)


Kaiser Permanente Experience

Automating the IHI Global Trigger Tool

presented at AHRQ 2010 Annual Meeting

September 29, 2010

Dot Snow, MPH

National Patient

Permanente Foundation Health Plan

Fast Facts About Kaiser Permanente

Founded in 1945, Kaiser Permanente is the nation's largest not
profit health plan, serving 8.7 million members

Among the innovations it has brought to U.S. health care are:

prepaid insurance which spreads the cost to make it more

physician group practice with a focus on preventing illness
as much as on caring for the sick

an organized delivery system, putting as many services as
possible under one roof

implementation of the largest civilian electronic health
record in the world

Fast Facts About Kaiser Permanente

Medical Centers:

Medical Offices:



Operate with a Labor Management Partnership

Local markets in Northern California, Southern
California, Hawaii, Colorado, Ohio, Georgia,
Oregon/Washington and the Mid
Atlantic States

Trigger Tool/TIDS Experiences

Initial Pilot of IHI Global Trigger

IHI Global Trigger Tools Pilot (2006)

Data came from 2 medical centers representing 2,363
patient days across 400 patients

37% of patients experienced an adverse event, with 20
25% having more than one adverse event

47% of the adverse events led to increased use of care

Focal Trigger Tool Pilot of IV Heparin Therapy

Automated Adverse Event Monitoring Program (AAEMP)

Other Trigger Tool Study Results

Adverse Drug Events: ADEs/1000 medication doses = 2.67%



RK The adverse drug event trigger tool: A practical methodology for measuring medication
related harm. Journal Quality and Safety in Health Care June 2003)

ICU Trigger Tool Study: 1450 adverse events found in 55% of patients; with 28% of patients
having > 1 event




D Methodology and rationale for the
measurement of harm with trigger tools. Quality and Safety in Health Care.

December 2003)

IHI Collaborative Surgical Trigger Tool Data: adverse events found in 14.6% of patients

(Griffin FA,

DC. Detection of adverse events in surgical patients using

Tool approach. Qual.
. Health Care 2008 17: 253

The Office of Inspector General pilot study on national incidence of adverse events in
hospitals showed ~30% of

patients experienced some type of adverse event
during hospitalization. Results from the National Incidence Study to be released later in
(Lee Adler, MD; 2010 NPSF Congress presentation)

Application of Initial Pilot Results

Initial pilot identified IV heparin therapy as an area for focal
trigger tool review

Four improvement opportunities identified:

Clarified the logic in the regional algorithm

Provided additional teaching of protocol to improve

Promoted alternative therapies with less risk for
dropping platelets

Adjusted tools in ED to allow for documentation of
double checks

Automated Adverse Event Monitoring Program


funded research project

Represents a collaborative effort between
Kaiser Permanente and Computer Sciences

Seeks to automate the IHI GTT methodology
utilizing data from KP

Will be used to detect and evaluate triggers
that allows for the measurement of adverse
events in a health care system

Process Steps in Building the AAEMP Application

Identify and define triggers

Map triggers to locations in data tables from EMR

Build application

Validate triggers and data from EMR locations

Define operational workflows

Prepare clinical site for pilot test

pilot: refine trigger definition and application
based on sensitivity / specificity analyses

Identify and Define Triggers

Started with triggers from IHI GTT

Triggers must be located in discrete fields within
the EMR; info located in text or notes will not be
available for trigger detection

Worked with AAEMP Steering Committee to
carefully define triggers including threshold
values for positive triggers and specific
inclusion or exclusion criteria

Added 7 infection specific triggers to test in
AAEMP pilot, including CAUTI, VAP, and sepsis

Map Triggers to Data Table Locations

Used Clarity data tables rather than Chronicle
server to avoid interference with operations

Time intensive step

requires knowledge of
regional configurations and info location in

Want to identify all possible locations of trigger
information within EMR to maximize reliability of
the application to identify triggers and values

Build AAEMP Application

Original proposal called for a Java application and
Oracle database

Requires collaboration with IT to determine
technical specifications, feasibility assessment,
system architecture and test plan

Front end of application completed

Back end not built at that time; ongoing roll
out of
EMR to medical centers was top IT priority

Adapted AAEMP for Access Database

Identified a medical center that used Access
to build a surveillance system similar in
concept to AAEMP

Vendor modified application to run in
Access and incorporate GTT triggers

Application loaded onto desktop computers
rather than a system server

Validate Trigger and Data from EMR Locations

The application is tested using real patient
data to determine that the information
coming back from Clarity tables is indeed
the trigger information expected to be

Define Operational Workflows

Application designed to assess medical records of
all hospitalized patients and produce a list of
patients with positive trigger firings

Nurse reviewer from Quality Department reviews
medical record to determine whether an adverse
event is present

If adverse event is present, categorize severity
level according to NCC
MERP levels of harm

Physician validates nurse’s review and information
is added to database

Prepare Clinical Site for Pilot Test

All members of pilot site team complete required
IRB trainings and certifications

Pilot Site Training:

Overview of IHI GTT and methodology

Practice trigger and adverse event evaluation using
sample charts

Demonstrate AAEMP application and reports

Clarify operational workflow

Schedule series of check
in meetings to discuss
emergent issues

May want to review initial cases early in pilot as a check
for inter
rater reliability

Planned Post
Pilot Analyses

Sensitivity / specificity analyses of triggers with
subsequent refinement of trigger definitions and

Return on investment (lives saved; shorter LOS;
cost reductions)

How orders in EMR could be changed to re
/ re
build for reliability in order to reduce harm to
our patients (e.g., what feedback, directed
up, and/or actions need to be hardwired
into orders; what best practice alerts need to

Considerations / Challenges

Regions have made unique configuration
choices that impact the location of data in
Clarity tables after the daily ETL process

Trigger information may be documented in
multiple places within EMR

Implementation demands greatest with initial
sites; further spread would benefit from
pilot site experiences

Considerations / Challenges

Articulate agreements for how information
will be shared and acted upon

Sequencing trigger tool activity within a
system’s EMR rollout and/or upgrades

validation and testing activities may compete
with production schedules and resource


AAEMP allows detection of triggers in near real time

based methodology is a retrospective review and trigger
identification/detection is well after the fact


allows Kaiser Permanente to augment
the classic IHI global trigger tool methodology in two
important and innovative ways:

searches the medical records of

hospitalized patients
instead of a sample, generating larger samples and revealing
consistent patterns

allows for close surveillance of patients for the rapid detection
and prevention or amelioration of harm


Can replace existing surveillance activities;
sampling of records can allow fit with
existing resources

An automated trigger tool can:

shine light on areas of improvement

Assess impact of specific performance
improvement initiatives

Concluding Observations and Thoughts

A strong multidisciplinary steering
committee is essential

Numerous IT issues need to be addressed:

System architecture

Security considerations

Vendor relationships

IT standards and specifications

Will be a powerful application of a validated
adverse event detection tool