Anticipating and Preventing Adverse Drug Events (ADEs):

Decreasing the Need to Rescue Hospitalized Patients from Opioid-related Complications

Ashley Meyers, BSN, RN-BC, PCCN-K

Craig Havican, BSN, RN

About Sparrow Health System» Sparrow Hospital - Lansing

» 733 beds

» 30,000 inpatient discharges

» Surgery: 8,162 IP, 12,776 OP

» 4,200+ births, Level 3 RNICU

» 117,000+ annual ED visits

» 960+ Providers, 6500+ Caregivers, 2300+ Volunteers

» Sparrow Specialty Hospital (LTACH)

» Sparrow Clinton, Ionia and Carson Hospitals

» Ambulatory clinics and services

2

Sparrow Offices» Ambulatory clinics – 60 locations, 400+ Providers » Outpatient visits – 590,000 visits/year» Variety of specialties and services

3

• Oncology• OB/Gyn• Orthopedics• Pain Management• Pediatrics• Perinatal• Surgery• Urgent Care• Weight Management• Wound and Hyperbaric

• Behavioral Health• Cardiology, CVT Surgery• Diabetes/Endocrinology• Family Medicine• FastCare Retail Clinics• Gastroenterology• Geriatrics / Senior Health• Infusion Centers• Internal Medicine• Nephrology• Neurology

Local Problem

» We are in an opioid public health emergency

» Opioid administration in hospitalized patients too often harms those it is intended to help (ADE)

» Inpatient administration of the opioid antagonist naloxone for overmedication is evidence of overuse or misuse

» Our data suggested that we could do better

» Timely identification and intervention for patients at risk or with early evidence of respiratory depression should help (effectiveness, safety, cost)

4

Local Problem

» Validated tools exist to decrease the risk of iatrogenic overdose

» Assessment of level of sedation

» Determination of opioid safety score

» Triggering interventions before naloxone rescue is required

» Well- designed, pervasively used EMR tools should help improve clinical outcomes and decrease costs

» No EMR-integrated tools and workflows to improve

5

Drivers to Take Action: 2012-2014

» Escalating national opioid crisis; Michigan’s ranking

» 10th in opioid prescribing, 18th in opioid deaths)

» Joint Commission Sentinel Alerts

» American Society for Pain Management Nursing Guidelines

» MHA Keystone Center Opioid ADE Prevention Initiative

6

Front-Line Nurses Leading the Way

» Sparrow Pain Resource Nurses (SPRNs) began looking at quality data in 2014

» Nearly 1% of patients we were treating with opioids in the hospital experienced opioid-induced respiratory depression (OIRD) requiring naloxone rescue

» To improve this, we set a hospital goal to decrease the rate of OIRD requiring naloxone rescue using:

» People: Governance, leadership, clinicians, IT

» Processes: Policies, workflows, Lean methods, PDCA

» Technology: IT (EMR); devices (ETC02 - capnography)

7

8

Sparrow Baseline Data

• Sparrow rate of opioid-induced respiratory depression (OIRD) as measured by percent of inpatients on opioids requiring naloxone rescue administration

• 2014: 0.72%

• 2015: 0.73% and increasing

9

Narcotics Accounting for Naloxone Use

Fentanyl Hydrocodone/APAP

Hydromorphone Morphine

Oxycodone

2015 Naloxone Rate Trend Before MOSS Implementation

» MOSS Documentation = 0%

10

Project Description and Goal

In acutely ill adult inpatients, does an EMR-integrated risk assessment tool aimed at preventing OIRD, implemented using

accepted nursing workflows, decrease the incidence of OIRD compared to no risk assessment, as measured by naloxone use?

Goal:

» Decrease IP naloxone rescue rate to ≤0.65%*

» Goal is not 0%; some patients will require naloxone rescue even with use of best practices

» Focus on appropriate use and preventive monitoring

11

* Benchmark for the 7 Caymich Michigan member hospitals, all 550+ beds

Benchmarks

» Naloxone Target Rate* = 0.65%» Numerator = # of patients receiving opioid & naloxone

» Denominator = # of patients receiving opioid (any route)

» Aligned with

» MHA Keystone Pain Management Collaborative

» Hospital Improvement Innovation Network (HIIN)

*Khelemsky et al., 2015, Caymich, 2017

12

Design and Implementation

» The knowledge tools we decided to use

» Pasero Opioid-Induced Sedation Scale (POSS)1

» Michigan Opioid Safety Score (MOSS)2

» The IT tool we decided to use: Epic, because it is…

» Key to our Sparrow Way and care transformation goals

» Where clinical care gets done…and documented

» How we deploy CDS: Risk scores, BPAs, nursing care plans

» If this works, we can share it with other Epic organizations

13

1. Pasero C. J Perianesth Nurs. 2009;23:186

2. Soto R, Yaldou B. J Perianesth Nurs. 2015;30:196

Design and Implementation

•Review literature•Identify best practices

•People, process, technology•Gather a guiding coalition•Communicate for buy-in

Decide what good looks like

•Assessment documentation•Scales & scores•Nursing care plans•CDS tools, displays•Policy-supported workflows

Build the solution in EMR •Application testing

•Integrated testing•MOSS education•FMEA, address findings•Policy implications

Test, Talk, Teach

•Nursing leadership sign-offs•EMR workflow training•Put into nursing practice•Measure, monitor, adjust

Go-live & PDCA

14

More about MOSS» Combines

» Health risk assessment

» Respiratory rate

» Modified POSS (mPOSS)

15

Soto R, Yaldou B. J Perianesth Nurs. 2015;30:196

A SWOT Analysis of Paper MOSS Deployment to Inform Our Conversion to EMR

16

Strengths• Supporting Literature• Pain Champions• Clear Assessment Times

Multimodal Pain Management Order Sets

Weaknesses• Double Documentation

(Paper & Electronic)• Turnover of RN/Nurse

Leadership

Opportunities• Improve Patient Safety• Increase RN autonomy • Decrease RRT• Decrease Narcan Use

Threats• Other competing pilots• Engagement

Sisco, Cooper, & Rayburn, 2014

From Paper to EMR

17

Preoccupation with Failure to Promote Success Every Step of the Way: FMEA*

Plan EMR Build Evaluate

Documentation Committee

Across applications

View for other disciplines

Method

Audience

Other Disciplines

Content

Go-Live Support

Timeline

Just in time education

Educate Implement

1 2 3 4 5

Key Stakeholders

Align with other initiatives

Timeline

Ongoing PI

Risk Mgmt reporting

PDCA follow-up plan

(Harpel & Giannini, 2014)

18

*FMEA = Failure Mode and Effects Analysis

How Health IT Was Used

» Use standard processes for nursing documentation (EMR flowsheets) to capture data for MOSS value

» Health risks: e.g., obstructive sleep apnea, age, other sedatives

» Respiratory rate: document once, use many times

» mPOSS sedation assessment

19

How Health IT Was Used

» Program the EMR to use nursing documentation to calculate the MOSS value

» Display the MOSS value where nurses can see and interpret it

» Provide usable, actionable clinical decision support (CDS) to drive best practices

20

MOSS Scoring and Action

21

How Health IT Was UsedNursing Documentation Display of MOSS Value

22

How Health IT Was Used

Clinical decision support

» Alerts, advises to add nursing care plan or exclusion

» Adds and opens care plan

Unit level reports

» Department managers

» Rapid Response Team (RRT) nurses

23

How Health IT Was Used: Care Plans

24

Value Derived: Improved ProcessesAdherence to Best Practice

25

Trai

ning

2

Policy

Req. Doc.

Trai

ning

1

Value Derived: Patient OutcomesSurpassed Naloxone Rate Benchmark (≤0.65%)

MeanNaloxone

Rate

Relative Change in Naloxone

Rate

# of Patients

Receiving Naloxone

Per Year

Mean IncidencePer Month

2014 0.72% NA 179 14.9

2015 0.74% 2.6% 174 14.5

2016 0.69% -3.2% 157 13

2017 0.51% -23.6% 113 9.4

2018 YTD* 0.60% -15.8% 68 9.7

26

*Data through July 2018

Value Derived: Patient OutcomesDeclining Naloxone Rescue Rate

27

Trai

ning

1

Trai

ning

2

Polic

y

Req.

Doc

.

Value Derived: Improved Processes & Outcomes

MOSS documentation Naloxone rate (Goal ≤0.65%)

28

• Lower morbidity• Fewer opioid ADEs

Value Derived: Cost Avoidance as a Result of Clinical Improvements

» Estimated cost per ADE*

» Non-ICU = $13,994; ICU = $19,685

29

* Sultana J, Cutroneo P, Trifirò G. J Pharmacol Pharmacother. 2013; 4:S73-7.

Mean Naloxone Rate

Raw IncidencePer Year

Mean IncidencePer Month

Estimated Cost Avoidance

(Raw Incidence vs. Baseline Year)2014 0.72 179 14.9

2015 0.73 174 14.5 $70K - $98K

2016 0.69 157 13 $308K - $431K

2017 0.51 113 9.4 $924K - $1,293K

2018 YTD 0.6 58 9.6 $510K - 714K

2018 Projected $874K - 1,224K

• Fewer patients harmed = 142• Cost avoidance = $1.8M - $2.5M

External Recognition for Our IT-enabled Best Practice: Epic Clinical Program

30

External Recognition for Our IT-enabled Best Practice: ECRI Institute

» ECRI Institute - a nonprofit organization, dedicated to bringing applied scientific research in healthcare to uncover the best approaches to improving patient care, marrying experience and independence with the objectivity of evidence-based research

» MOSS program recognition» Better processes» Naloxone reductions» Greater staff comfort in

administering opioids safely» Criteria-based prescribing

restrictions (fentanyl)» Safer order sets (PCA)» Pushing nonpharmacologic pain

management modalities (e.g., heat and cold, aromatherapy, pet therapy)

31

Capital and Operational Expenses

» Capital expenses = $ 0

» Operational expenses = $57,375

32

Activity CostPI/Project Planning; 120 total hours $ 3,988 Committee work; 20 total hours $ 665 Live and online Nursing Education; 1.5 hours each $ 49,845 Go-Live Support; 40 total hours $ 1,329 EMR analyst time; 60 total hours $ 1,549 TOTAL $ 57,375

Lessons Learned

33

» Involve your frontline staff at the beginning

» Sense of ownership of the problem and solution

» Partner with IT team at the beginning ofimplementation planning

» One time education & training isn’t enough

» Include at-the-elbow support

» Technology and training do not ensure sustainability

» Need clear expectations and accountability

Thank you!

34