automating safety detection and surveillance: the ...automating safety detection and surveillance:...

Automating Safety Detection and

Surveillance: The Automated Adverse Event

Detection Project (AAEDP)

Eric S. Kirkendall, M.D.

Objectives

• Describe the AAEDP and it’s role in our safety systems at Cincinnati Children’s

• Discuss the data types and information attained via the AAEDP

– Automated vs. Manual Collection

• Review current triggers and safety events detected by the AAEDP

• Highlight collaborative efforts involving the AAEDP

When It Began…

• Started July 2006 from CCHMC Center for Health

Care Quality and Health Policy & Clinical

Effectiveness

• Started as an AHRQ CERTs grant

– Only pediatric CERTs site at this time

• Automated system meant to detect harm and inform

improvement teams

– Meant to overcome shortcomings of voluntary reporting

systems

4

CCHMC

Detection

of Harm

Safety

Reporting

System &

SSE’s Never

Events

SSI

VAP

BSI

Codes

Outside

ICU

CHCA

ADE

Tool

AAEDP

Pressure

Ulcers

GTT

PIV’s

AAEDP Workflow Process…

• Clinical Data Repository queried

• Reports generated and sent to application

Report Generation

• Clinical Investigator retrieves reports

• Manually reviews reports and EHR data

• Makes manual exclusions

• Captures event characteristics

• Determines if adverse event occurred (if possible)

• Enters information/data into adverse event database

Manual Investigation

• Data Analyst compiles data, analyzes statistically

• Produces run charts, graphs, etc

Data Processing

• Output is taken to improvement and governance teams

• Create new improvement teams if necessary

Presentation to Stakeholders

Technical Aspects

• Clinical Data Repository queried

• Daily reports reside in Business Objects Infoview

application

• Investigator logs into inbox, reviews and downloads

report in Excel (or PDF, Word, text, etc)

• Results loaded into separate MS Access relational

database (more advanced database in the works)

Objectives






Granularity of Data

• Automated periodic reports contain basic

demographic information

– Can be very granular, depending on the

complexity of the query and needs of the

investigator

• More advanced data (such as medical information

that requires clinical judgment) is gathered manually

• Bridging the gap with more advanced queries

– Nephrotoxin list with Acute Kidney Injury trigger

Data (continued)

• Mainly medication-related detection

– Can get very specific drug information including dose, route

of admin, formulation, concomitant meds

– Can use graphing/synopsis functions to tie timing of

medications to vital signs, other events

• Only medical device-related triggers are IV

infiltrates/phlebitis and oversedation (PCA-related)

– No specific information is gathered initially, but is available

through EHR

Objectives






1st Triggers Followed

• Narcan, Flumazenil, Digibind, Vit K-Warfarin,

started July 2006

• Glucose/Insulin added Dec 2006

• Hyaluronidase added Sept 2007

• Phentolamine added Feb 2008

• Glucagon, Infiltrates/Phlebitis added Feb

2008

Adverse Drug Events

Naloxone

Trigger Started

Glucose

Trigger Started

Hyaluronidase

Trigger Started

IV Chart Review

Started

20

30

40

50

60

70

80

Ju

l

Au

g

Sep

Oct

No

v

Dec

Jan

Fe

b

Ma

r

Ap

r

Ma

y

Ju

n

Ju

l

Au

g

Sep

Oct

No

v

Dec

Jan

Fe

b

Ma

r

Ap

r

Ma

y

Ju

n

Ju

l

Au

g

Sep

Oct

No

v

Dec

Jan

Fe

b

Ma

r

Ap

r

Ma

y

Ju

n

Ju

l

Au

g

2006 2007 2008 2009

# A

dv

ers

e E

ve

nts

Other

IV Infiltration (Chart Review) (03-25-08)

IV Infitration (Hyaluronidase) (09-01-07)

Hypoglycemia (Glucose Bolus) (12-08-06)

Opiate Over-sedation (Naloxone) (07-08-06)

Max ADEs Detected (3-month Average)

Current / Reduced Monthly ADEs vs. Detection Capability Base Level (3-month Average)

Max ADEs Detected (3-Month Avg) Desired Direction

ACurrent/Reduced Monthly ADEs Desired Direction___

(Relative to Level of Max ADEs Detected)____

Current Triggers Followed

• Active Improvement – Narcan

– Flumazenil

– Digibind

– Glucose/Insulin

– Glucagon

– Hyaluronidase

– Phentolamine

– Vit K after warfarin

– Infiltrates/Phlebitis

• Surveillance Only • Benztropine

• Kayaxelate

• PACU to ICU

• Any Transfers to Higher Level of Care

CCHMC Central Venous Catheter (CVC) Associated

Laboratory Confirmed Bloodstream Infections (LCBIs)

0.2

1.8

1.10.9 1.0

0.5

1.2

0.70.9

1.2 1.31.5

0.9

0.8

0.90.9

1.9

2.7

1.11.41.4

0.8

2.2

2.12.4

2.5

1.5

1.8

1.2

1.6

1.31.5

2.6

1.81.6

1.1

2.5

2.0

1.4

2.4

1.9

3.7

4.0

2.5

3.0

2.9

1.2 0.9

0

1

2

3

4

5

6

7

8

Q3 '04

Q4 '04

Q1 '05

Q2 '05

Q3 '05

Q4 '05

Q1 '06

Q2 '06

Q3 '06

Q4 '06

Q1 '07

Q2 '07

Ju

l '0

7

Au

g '07

Sep

'07

Oct

'07

No

v '07

Dec '07

Jan

'08

Feb

'08

Mar

'08

Ap

r '0

8

May '08

Ju

n '08

Ju

l '0

8

Au

g '08

Sep

'08

Oct

'08

No

v '08

Dec '08

Jan

'09

Feb

'09

Mar

'09

Ap

r '0

9

May '09

Ju

n '09

Ju

l '0

9

Au

g '09

Sep

'09

Oct

'09

No

v '09

Dec '09

Jan

'10

Feb

'10

Mar

'10

Ap

r '1

0

May '10

Ju

n '10

FY2005 FY2006 FY2007 FY2008 FY2009 FY2010

Infe

cti

on

s p

er

10

00

De

vic

e D

ay

s

CVC-LCBIs Control Limits Goals [0.8 (Jul06-Jun07) / 1.0 (Jul07-Jun08) ] Baselines

Q3/04 - Revised Care Practices and CHG

Scrub for Line Care

Q3/04 - CHG Scrub for CVC Insertions

Q4/04 - Maximal sterile barriers and CHG

Q4/04 - Interventional Radiology

New CCHMC

Collaborative

Began, Q1/06

CA-BSI NACHRI

derived CVC bundle

rolled out to hospital

3/15/07

Q2/05 - MaxPlus Cap in PICU B4 & A6

Q2/05 - MaxPlus Cap on A5N2

Q3/05 - MaxPlus Cap cancelled on

A5N2

Q3/05 - MaxPlus Caps cancelled.

Updated Thru 30Jun10 by Kate Rich, Division of Health Policy & Clinical

Effectiveness

Source: Infection Control Dept.

Desired

Direction of

Change

Chart Type: u-chart

3/17/09 - Microclave

Cap Use

Housewide

Jan09 - Microclave

Cap Use ICUs

Feb09 - Microclave Cap

Use HemOnc/BMT

Spiros device

evaluated 8/1/09-

9/2/09

24 26 21 36 34 24 18 24 15 21 27 12 6 6 7 9 9 9 5 8 5 3 8 6 5 8 8 6 6 5 5 4 4 1 4 4 4 9 5 4 4 2 5 3 4 5 6

8340

8741

8546

9034

9080

8890

9324

10054

10450

10667

11007

10491

3861

4094

3827

3625

3706

3434

3439

3790

3757

3651

3705

3840

4180

4484

4177

4159

4353

4306

4545

4326

4834

4476

4595

4483

4620

4874

4393

4463

4130

4309

4003

4015

4496

4113

4680Device

Days

Infections

Common Demographic and Event Characteristics of

Acute Kidney Injury (AKI) Automated Trigger

Common Demographic Information Collected from Patient's Chart

Regarding Each Active Trigger Event:

Medical Record Number

Encounter Number

DOB

Date of Admission

Primary Diagnosis

Secondary Diagnosis

Weight (Kg)

Gender

Ethnicity

Date of Trigger

Time of Trigger

Location of Trigger occurrence

Location of Event occurrence (i.e.Unit and whether in-hospital or not)

Date of Event occurrence

Time of Event occurrence

Hospital Service Involved

Trigger Dose if Applicable

Exclusion Criteria for Creatinine Trigger

Any patient who has had cardiac bypass within 72 hours prior to

maximum creatinine value

Patient’s LOS (length of stay) is < 3 days

Patient with previous history of Dialysis

Time to Creatinine recovery (i.e. return to baseline value) is ≤ 2 days

Minimum value of 0.5mg/dL (Leave off for now; will evaluate later

after charts are reviewed…)

Additional Event Characteristics for Creatinine Trigger

Was the patient on IV fluids prior to the increased creatinine? (yes/no)

Was the Patient on Dialysis at any time during this admission?

Has Patient had a BMT? This admission or a previous admission?

Is the Patient receiving a Nephrotoxic medication? (record drug and

dose and frequency)

Are medication monitoring levels (peaks and troughs) ordered

appropriately for Gentamicin, Amikacin and/or Tobramycin and

Vancomycin?

Are dosage adjustments made appropriately if needed?

Are monitoring levels ordered again after re-adjustments to dosage?

Post-event monitoring: Renal panels and IVFs ordered appropriately

after increase in creatinine noted?

Team Members

• Clinical Investigator, Elizabeth Kloppenborg, RN

• Director, AAEDP, Eric Kirkendall, MD

• ADE Team members

– Patient Safety Officer, Stephen Muething, MD

– Lead Pharmacists,

– Many others: QI specialists, Data Analysts, Project

Leaders, Legal

• Higher Governance, HPCE

Objectives






Collaboration

• Local – Improvement Teams ~ Opiate Task

Force

• Regional – SPS/Ohio Collaborative (more on

that in a minute)

• National – AAEDC (David Stockwell at

CNMC), Publications

SPS

• Collaborative

An Ohio Pediatric Hospital

Collaborative for Patient Safety

8 Pediatric Hospitals Working Together

• Shared aspiration of making Ohio the safest

place in the nation for Children’s health care

• Innovative & strategic approach to building a

State-Wide Pediatric Collaborative

Improvement Model

• Transparency of Data: each hospital’s

willingness to share successes and failures in

order to improve healthcare in Ohio for

children

Ohio Children’s Hospitals

Quality Improvement Initiatives

• CABSI (NACHRI)

• MRT Code (OCHA)

• ADE (OCHSPS)

• SSI (OCHSPS)

• Bedside Care (CHCA)

• Pressure Ulcers (CHCA)

• Procedural Never Events

(CHCA, 2008)

• ADE – Sustain & Spread

(CHCA, 2008)

• MDRO (CHCA, 2008)

• Inpatient Throughput (CHCA,

2007)

• Eliminating Codes (CHCA)

• ED Wait Times (CHCA,

2006)

• SSI (CHCA, 2006)

• CABSI (CHCA, 2005)

• ADE-Narcotic (CHCA, 2005)

July 2010 December 2015

Eliminating Serious Harm in Ohio’s Children’s Hospitals

Goals:-Eliminate serious harm in Ohio’s pediatric hospitals

-Reduce the overall cost of health care in the state

-Develop ongoing learning network

-Build a sustainable state-wide infrastructure.

X

Jul-10 - Dec-10

Jul-10 to Mar- 11

PLANNING PHASE

Activities:-will building with stakeholders,-fundraising-formation of PSO-action plan development-Diagnostic evaluations at 6 remaining sites-training of risk managers

-construction of data systems

Jan-11

Milestones:

-funding identified

-PSO formed

-action plans resourced/launched

-evaluation results completed

-data systems and DBs complete

-training complete

Dec-10 - Dec-12IMPLEMENTATION PHASE

Activities:-focus on specific SSE reduction with specific aim-Develop state-wide error prevention behaviors and associated tools-error prevention training for all front-line-leadership consulting and leadership training-Safety Coach program at all sites-simulation training for high risk areas-root cause analysis training and continuous learning-action plan and "lessons learned" network-6 month improvement cycles on common cause areas-Improvement Advisor Training-Periodic site visits by SPS leadership

-Continued expansion of shared data for specific sources of harm.

Dec-12

Milestones:

-SSE reduction goals met

-Completion Trainings:

IHI QI Training

Error Preventioin

Root Cause Analysis

-Safety Coach Programs implemented

-Conducted multiple Learning Sessions

-Implementation of RCA Plans

Jan-13 - Dec-15EXPANSION PHASE

Activities:-Establishing stable QI infrastructure and learning network-Further development of QI capability at all levels-Continued development of HRO practices at each site-Continue 6 month improvement cycles for common causes of harm as revealed by data.-Further develop state-wide resources for simulation training

-cementing permanent learning networks.

Dec-15

Milestones:

-Advanced QI Trainings Completed

-Funding ID’d for Simulation Training

-Launch a new state-wide

Improvement Team every 6 months

-Reliable QI infrastructure and

Learning Network implemented

Questions…

automating safety detection and surveillance: the ...automating safety detection and surveillance:...

Documents