High Reliability and Microsystem Stress - Brady & Hausfield · 4/12/2016 1 High Reliability and Microsystem Stress Helping leaders identify and mitigate unit level stress: Next steps

4/12/2016

1

High Reliability and Microsystem Stress

Helping leaders identify and mitigate unit level stress: Next steps towards the journey of high reliability

Whittney Brady RN, DNPJackie Hausfeld, RN, MSN, NEA-BC

Objectives

Quantitative metrics and qualitative measures indicative ofmicrosystem stress

Describe mitigation and escalation strategies at the unit, microsystem and organizational levels to prevent serious harm and other types of poor outcomes in stressed systems.

Discuss a systematic approach to predict stressed microsystems.

Mitigate

Predict

We have no financial, professional or personal conflict of interest to disclose

4/12/2016

2

Develop a system to identify, mitigate and predict microsystem

stress in order to prevent serious harm and other undesirable

outcomes.

Global Aim

Identify Mitigate Predict

Gray box = completed interventionKEY Green box = what we’re working on right now

White box = future work

Definitions

Organizational Level = Macrosystem

Unit Level = Microsystem

Inpatient System Level = Mesosystem

CANCER & BLOOD DISEASES INSTITUTE (CBDI)

The First Stressed Microsystem

4/12/2016

3

• 56 beds in CBDI 6/13



• 360 new oncology patients per year

• 100-110 bone marrow transplants per year

Identify Volume

CBDI: Quantitative Measures


Identify Volume

Identify Staffing


4/12/2016

4

Less Experienced NursesStaffingIdentify


Identify Acuity

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ju

ly_1

1 (n

=1

247)

Au

g_1

1 (n

=1

094)

Se

pt_

11 (

n=

1122

)

Oct

_11

(n=

123

8)

No

v_

11 (

n=

129

5)

De

c_11

(n

=1

380

)

Jan

_1

2 (n

=15

26)

Fe

b_

12 (

n=

1362

)

Mar

_12

(n

=14

34)

Ap

r_12

(n

=15

50)

May

_12

(n=

135

2)

Ju

n_

12 (

n=

1410

)

Ju

l_12

(n

=1

501

)

Au

g_1

2 (n

=1

415)

Se

p_

12 (

n=

12

40)

Oct

_12

(n=

128

0)

No

v_

12 (

n=

105

8)

De

c_12

(n

=1

136

)

Jan

_1

3 (n

=12

28)

Fe

b_

13 (

n=

1081

)

Mar

_13

(n

=12

34)

Ap

r_13

(n

=13

14)

May

_13

(n=

136

8)

Ju

ne

_13

(n=

124

6)

Ju

l_13

(n

=1

695

)

Au

g_1

3 (n

=1

652)

Se

p_

13 (

n=

14

56)

Oct

_13

(n=

160

6)

No

v_

13 (

n=

147

3)

De

c_13

(n

=1

414

)

Jan

_1

4 (n

=15

53)

Fe

b_

14 (

n=

1426

)

Mar

_14

(n

=17

74)

Ap

r_14

(n

=21

57)

May

_14

(n=

222

2)

Pri

mar

y B

SI R

ate

per

100

0 lin

e d

ays

Month

Primary BSI Rate in CCHMC CBDI (July 2011-May 2014)

Monthly Primary BSI Rate Median BSI rate Control Limits


Serious Harm: BSI

• Stabilization of current processes

• 2 person dressing changes

• Daily prevention standard rounding with real time feedback

Stressed Microsystem: CBDI

Mitigate

UnitUnit

• Increased education to float staff and review of CVC care by all staff• Physician engagement in BSI prevention work• Pre assignment of float staff

Interventions

Implementation of a system to improve allocation of resources and support to deescalate system stress• Implementation of a experienced based knowledge bonus

Inpatient System

Inpatient System

OrganizationOrganization

4/12/2016

5

0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

4.0

Ju

l-1

1 (n

=1

247)

Au

g-1

1 (

n=

1094

)S

ep

-11

(n=

112

2)O

ct-1

1 (

n=

12

38)

No

v-1

1 (n

=12

95)

De

c-1

1 (n

=1

380)

Jan

-12

(n

=1

526)

Fe

b-1

2 (n

=1

362)

Mar

-12

(n=

143

4)A

pr-

12

(n

=1

550)

May

-12

(n

=13

52)

Ju

n-1

2 (n

=1

410)

Ju

l-1

2 (n

=1

501)

Au

g-1

2 (

n=

1415

)S

ep

-12

(n=

124

0)O

ct-1

2 (

n=

12

80)

No

v-1

2 (n

=10

58)

De

c-1

2 (n

=1

136)

Jan

-13

(n

=1

228)

Fe

b-1

3 (n

=1

081)

Mar

-13

(n=

123

4)A

pr-

13

(n

=1

314)

May

-13

(n

=13

68)

Ju

n-1

3 (n

=1

246)

Ju

l-1

3 (n

=1

695)

Au

g-1

3 (

n=

1652

)S

ep

-13

(n=

145

6)O

ct-1

3 (

n=

16

06)

No

v-1

3 (n

=14

73)

De

c-1

3 (n

=1

414)

Jan

-14

(n

=1

553)

Fe

b-1

4 (n

=1

426)

Mar

-14

(n=

177

4)A

pr-

14

(n

=2

157)

May

-14

(n

=22

22)

Ju

n-1

4 (n

=2

143)

Ju

l-1

4 (n

=1

437)

Au

g-1

4 (

n=

1680

)S

ep

-14

(n=

156

0)O

ct-1

4 (

n=

16

78)

No

v-1

4 (n

=08

00)

Blo

od

Str

eam

In

fect

ion

s p

er 1

000

Lin

e D

ays

Month (number of line days)

Primary Blood Stream Infection Rate in the Cancer and Blood Disease Institute (Infections / 1000 line days)

Monthly Blood Stream Infection Rate Average Rate of Blood Stream Infections Control Limits

Acute increase in census, phase 1 patients, relapsed refractory patients, national and

Increased percentage of floating and inexperienced nursing

resources

Implementation of:• Identification of high

risk patients• Improved daily CHG

bathing/oral care compliance

• Increased awareness of high BSI-risk patients

• Assistance for nurses performing high BSI-risk procedures

• System to improve allocation of resources

Stressed Microsystem: CBDI Outcome

• Decrease in primary BSI rate from 1.8 primary BSIs per 1000 line days to 0.21 BSIs per 1000 line days.

• Prolonged stress in complex systems with high-risk patients can contribute to increased BSI rates.

• Identifying key processes and executing mitigation strategies at the unit, microsystem and organizational levels can stabilize outcomes when under stress.

• Building on continued learnings from CBDI helped to identify the next stressed microsystem: NICU.

Stressed Microsystem: CBDI Key Findings

NEWBORN INTENSIVE CARE UNIT (NICU) – CY2014

The Second Stressed Microsystem

4/12/2016

6

• Record High Census

• High Acuity

• Major Construction

• New Staff

NICU

30

35

40

45

50

55

60

01/

01/

140

1/0

6/14

01/

11/

140

1/1

6/14

01/

21/

140

1/2

6/14

01/

31/

140

2/0

5/14

02/

10/

140

2/1

5/14

02/

20/

140

2/2

5/14

03/

02/

140

3/0

7/14

03/

12/

140

3/1

7/14

03/

22/

140

3/2

7/14

04/

01/

140

4/0

6/14

04/

11/

140

4/1

6/14

04/

21/

140

4/2

6/14

05/

01/

140

5/0

6/14

05/

11/

140

5/1

6/14

05/

21/

140

5/2

6/14

05/

31/

140

6/0

5/14

06/

10/

140

6/1

5/14

06/

20/

140

6/2

5/14

06/

30/

140

7/0

5/14

07/

10/

140

7/1

5/14

07/

20/

140

7/2

5/14

07/

30/

140

8/0

4/14

08/

09/

140

8/1

4/14

08/

19/

140

8/2

4/14

08/

29/

140

9/0

3/14

09/

08/

140

9/1

3/14

09/

18/

140

9/2

3/14

09/

28/

141

0/0

3/14

10/

08/

141

0/1

3/14

10/

18/

14

Dai

ly C

ensu

s

Date

NICU Daily CensusCY 2014

Average Daily Census Goals Linear (Average Daily Census)

NICU ADC 47

Licensed Bed #59

Adjusted ADC 49.5

Identify Volume

NICU: Record High Census

NICU: High Acuity

Identify Acuity

• Record number of CDH Patients with ALOS of 72 days

• “Managing census” utilizing level II and III NICUs in our region

• Landscape of the NICU has changed: Cincinnati Fetal Care Center

Fetal Care Patients

4/12/2016

7

NICU: Major Construction

NICU patients located on 5 different units.

October, 2014 –We had just over 200 RN’s…

…94 had been hired since 1/1/2013

NICU: New Staff

40

42

44

46

48

50

52

54

FY12 FY13 FY14 FY15 FY16

ADC

NICU ADC by Fiscal Year

NICU: Quantitative Measures

Identify Volume

FY 13: ADC 45FY 16: ADC 53

4/12/2016

8

NICU FTE’s by Fiscal Year

NICU: Quantitative Measures

Identify Staffing

0

50

100

150

200

250

300

FY12 FY13 FY14 FY15 FY16

FTE's

Increase in over 48 FTE’s

Snap Shot: Quantitative Metrics

10/19/14 – 10/25/14– 108% occupancy to

budgeted ADC

– (4.31) variance to budgeted HPPD

– 13% operational vacancy (before we added more FTEs)

– 1185 hours of float staff

PICC Team

Stressed Microsystem: NICU

Mitigate

UnitUnit

Leadership Prevention Standard Rounds: all patients on all units.Weekly report out on all serious harm in leadership meeting. Pre-assignment of float staff.

Interventions

Implementation of a system to improve allocation of resources Organizational support to deescalate system stress• Implementation of a experienced based knowledge bonus• Added FTE’s

Inpatient System

Inpatient System

OrganizationOrganization

• Targeted rounding • Prediction (Watchers) Multi disciplinary Huddles 4 times per day

4/12/2016

9

Stressed Microsystem: NICU Outcome

0.0

2.0

4.0

6.0

8.0

10.0

12.0

14.0

16.0

18.0

20.0

01/

01/

09 (

n=

14

85)

03/

01/

09 (

n=

13

26)

05/

01/

09 (

n=

14

66)

07/

01/

09 (

n=

12

70)

09/

01/

09 (

n=

13

90)

11/

01/

09 (

n=

13

90)

01/

01/

10 (

n=

12

80)

03/

01/

10 (

n=

13

93)

05/

01/

10 (

n=

15

60)

07/

01/

10 (

n=

12

61)

09/

01/

10 (

n=

12

70)

11/

01/

10 (

n=

12

97)

01/

01/

11 (

n=

12

61)

03/

01/

11 (

n=

14

59)

05/

01/

11 (

n=

13

55)

07/

01/

11 (

n=

14

83)

09/

01/

11 (

n=

14

10)

11/

01/

11 (

n=

14

29)

01/

01/

12 (

n=

14

48)

03/

01/

12 (

n=

12

80)

05/

01/

12 (

n=

14

04)

07/

01/

12 (

n=

14

23)

09/

01/

12 (

n=

14

74)

11/

01/

12 (

n=

13

56)

01/

01/

13 (

n=

13

95)

03/

01/

13 (

n=

14

63)

05/

01/

13 (

n=

12

40)

07/

01/

13 (

n=

14

50)

09/

01/

13 (

n=

13

57)

11/

01/

13 (

n=

15

94)

01/

01/

14 (

n=

13

29)

03/

01/

14 (

n=

16

41)

05/

01/

14 (

n=

16

44)

07/

01/

14 (

n=

16

57)

09/

01/

14 (

n=

16

06)

11/

01/

14 (

n=

15

66)

01/

01/

15 (

n=

13

17)

03/

01/

15 (

n=

12

89)

05/

01/

15 (

n=

16

20)

07/

01/

15 (

n=

14

76)

09/

01/

15 (

n=

14

76)

11/

01/

15 (

n=

15

60)

Infe

ctio

n p

er 1

000

pat

ien

t d

ays

B4 NICU Nosocomial Infection Data

Infection rate Average infection rate Control Limits

Mean = 3.32/1000 pt

Stressed Microsystem: NICU

QUALITATIVE FACTORSListening to families and staff…

4/12/2016

10

NICU: Qualitative Measures

Family Stress

Identify

Issue 143Aug 22 – Aug 28

NICU Notes

Week 2 Results – 4th FloorVery Supported – 50Somewhat Supported – 0Minimally Supported - 0Week 1 Results – 6th FloorVery Supported – 12Somewhat Supported – 0Minimally Supported – 0

(this one started 8/21)

Parents Feedback

4/12/2016

11

Staff Stress

NICU: Qualitative Measures

Issue 143Aug 22 – Aug 28

NICU Notes

Staff Qualitative Stress Measure

0

5

10

15

20

25

30

35

40

45

Yellow

Orange

RedMove to B Pod

Move to 500 Pod

Busy night – could have used more MTs

Move to A Pod

4/12/2016

12

NICU: Staff Definitions

• Good day, went well• Appropriate assignments• Not feeling stressed• Not feeling like you couldn’t

get things done in a timely manner

• Well supported• People there to help• Peers with good attitudes• Able to take a break and lunch• Able to teach families• Received the resources form

the house you requested

• Busy assignment but received the help needed

• Overall busy and unorganized• Some support but needed

more • Assignment busy and you

don’t have the supplies you need

• Chaotic and cannot catch up • Road trip, assignment changes

but received the help• Had to change assignments

during the shift• Changes in patient condition

Yellow Orange

• Super busy with no help• Inappropriate assignments• Leave work feeling over whelmed• Entire unit busy, you know but

there is nothing that can be done• No support from the people you

work with• Assignment unsafe • Staff not flexible• House takes your resources away

and creates less than ideal assignments/admit plan

• “Stupid busy” – phones ringing off the hook

• Staff with bad attitudes

NICU: Staff Definitions

Red

Qualitative Scoring

IS YOUR UNIT . . . . .

• GREEN: Routine risk/stress level within normal variability met by daily operations

• YELLOW: Minimal risk/stress level with some variability met by minor operational adjustments

• ORANGE: Moderate risk/stress level with high level of variability, predicted or unanticipated that require considerable number of interventions and support

• RED: High risk/stress level with a high amount of variability predicted or unanticipated, that require a large amount of intervention and support but very challenging to meet.

Predict

4/12/2016

13

• Can see the entire day in 4 hour blocks

System Level Qualitative Data Predict

Capturing Impact of Prolonged Stress on Staff

Initial Testing

• Unit staff used colors denote their stress level for the day

• 4 level color scale utilized

• Staff defined what each color represented

• Expanded separate rating process by charge nurses

• Correlation found between charge nurse and aggregate staff

ratings

Microsystem Stress: Qualitative Predict

Capturing Impact of Prolonged Stress on Staff

Current Process• Charge nurses determine overall color rating each shift with

input from staff and key roles on their unit

• Rating is entered into automated system every 4 hours and

comment entered if rated orange or red

• Comments provide information for resource allocation

• Comments also give insight into why the unit “feels” stressed

• Shift and aggregate data is utilized for shift decisions and

trending

Microsystem Stress: Qualitative Predict

4/12/2016

14

BUILDING A SYSTEMWhat did we learn?

Develop a system to identify, mitigate and predict microsystem

stress in order to prevent serious harm and other undesirable

outcomes.

Global Aim

Identify Mitigate Predict



4/12/2016

15

Team Name: Stressed Microsystems TeamDate: September 14, 2015 Revision: 11

System Level Key Diagram

Develop a system to identify, mitigate

and predict microsystem

stress in order to prevent serious harm (and other

undesirable outcomes).

Global Aim

Right factors (quantitative* and qualitative) are identified, validated, then utilized

Primary Key Drivers

Timely access to the right data representing right factors

Effective data analysis, review and data driven decisions

Roles and processes for management and decision-making are clear

Appropriate oversight and support by leadership

Identification and validation of quantitative factors• Duration Stressed System*

Sub‐Projects

Identification and validation of qualitative factors• Assessment of stress level

by nursing



Identification and validation of quantitative factors• Volume• Staffing• Patient Acuity

Mitigation and PredictionStrategies

QUANTITATIVE FACTORSShow me the numbers!

Microsystem Stress: Quantitative

• Microsystem Quantitative Daily Indicator Data- Reviewed published evidence- Validated relationship between indicators and harm - Indicators

• Actual and budgeted Average Daily Census (ADC) and percent occupancy

• Average actual Nursing Hours Per Patient Day (NHPPD) to budget

• Operational vacancy rate- Medical Leave of Absence (MLOA), orientation, hired

and waiting for boards, posted and not yet hired• Percent of float staff used• Multiple sites of care

Predict

4/12/2016

16

Updated List of Predictor Variables

Variable Periodicity CBDI Variable NICU Variable

SRU Hours (RNs) Monthly X X

Location Daily X X

Percent Occupancy Daily X X

NHPPD Hours X X

Number of International Patient - Estimated Monthly X

BMT New US Referral Monthly X

BMT Num Chemo Doses Monthly X

ONC Active Phase I Monthly X

*Note: data sources, periodicity, and assumptions subject to change based on final model. Current data sources used as they contained historical data

Microsystem Stress: Staffing

Pick correct

shift

Add requests

for needed

staff

Predict

NHPPD

Microsystem Stress: Staffing Predict

4/12/2016

17

MITIGATION ESCALATION AND PREDICTION

So now what?

B4 NICU 59 89.5% 52.8 52.2 98.9% 88.5% 18.2 17.0 ‐1.2 ‐6.6% 9.0% 685.1 12.8% 740 13.8% 7 1.4% 4.8% 0.0% 4.8%

Microsystem Stress Report Predict

Inpatient Unit Level I InterventionsGreen Yellow Orange Red

Attend bed huddle and Safety meeting. Match

clinical resources to patient acuity and care needs. Offer any additional staff to the house.

Assess available clinical resources and

ability to care for patients based on acuity and care needs.

Unit level clinical and medical operational

leaders to work on screening admissions and patient placement in collaboration with flow coordinators/MPS lead.


leaders discuss/determine need to reschedule pre-admissions and/or defer pts.

Continue with standard unit practices. Predict

& plan for admissions, discharges, and other flow factors today & looking forward.

Ask staff to work extra for defined shift with

resource need.

Make AVP aware of staffing and unit

operations.


leaders discuss ability to stop admissions and/or transferring patients to another facility.

Smooth resources & post shifts not at core and

also ask clinical staff and standby to pick up extra shifts based on volume.

Ask available current staff to work an

additional 4 hours.

Increase leadership rounding. Evaluate the need for the Director to take

charge/support role or continue with unit leadership activities to address unit operations.

D/C patients that meet criteria in a timely

manner.

Request appropriate SRU/float staff for

support such as RN, PCA, HUC, and Sitter.

Evaluate the need to cancel OPT/Education if

resource needs are not satisfactorily met.

All hands on deck and attending meetings and

other activities based on patient care needs and safety being met.

Predict operational vacancy and staffing

impact short term and long term.

Evaluate the need to move support roles into

charge or the direct care role.

Evaluate the need to move a manager into

charge or direct care role.

Strategize for increased RN hiring and

orientation for large numbers of open positions. Utilize creative methodologies that expand beyond the unit.

Evaluate the ability to adjust Assignment with

Preceptor/Orientee for Phase IV orientees close to completing orientation.

Temporarily increase staff FTE as open

positions filled

Evaluate the need to cancel unit meetings or

cancel staff attendance to department and division meetings.

Evaluate the need for additional support from

Pastoral services or other resources

Consider purchasing food for staff.

Evaluate the need for Organizational Support if

Ongoing Orange.

Mitigate

Ask available current staff to

work additional 4 hours

Evaluate need to move manager into

charge or direct care roll

All meetings and other non clinical

activities cancelled and resources

reassigned

4/12/2016

18

Inpatient Unit Level II InterventionsGreen Yellow Orange Red

Maintain current processes with distribution of SRU/Float Resources.

Consider microsystems that have been stressed for over a week in distribution

of resources.

Include AVP/VP in discussion around support for unit microsystem.

Implement all applicable interventions denoted at Orange level.

Evaluate the ability to partner with another unit with similar competency

and has a lower volume or more positive operational vacancy.

Evaluate the need to pre-assign some SRU resources to promote consistency

in support and decrease the staffing gap. Increase Month’s Team support.

Evaluate the need to increase RN and Allied Health resources permanently

related to new trends in ADC. Implement if appropriate.

Evaluate the need for a special pay program based on prediction of

operational vacancy and longer term staffing gaps.

Evaluate the ability to cancel or hold off on accepting Destination and

Tertiary Patients depending on clinical need, impact on program, etc.

Support manager and educators working extra clinical shifts.

Evaluate the need for the use of Supplemental staff. Post positions if

needed.

Provide support to providers to assist with rounding and other clinical work.

Mitigate

Evaluate the need for a special pay

program based on prediction of operational

vacancy and longer term staffing

gaps.

Dashboard Analysis

Average Weekly Occupancy

• 13 units: Average ADC over budget ADC

• 7 units: >85% Occupancy

• 5 units: >90% Occupancy

Average Nursing Hours per Patient Day (NHPPD)

• 7 units: Overstaffed by >5%

• 7 units: Understaffed by >5%

• 4 units: Within target range = GOAL

Operational Vacancy Rate

• 5 units: >10% vacancy rate

Float Use

• 6 units: >10%

– Decreasing over time as new hires leave orientation

– 2 units <12 beds

Qualitative

• 3 units: >10% of shifts rated orange or red

Predict

How are we using this information?

• Guides drill-downs into the data, why are the number

high or low and do we have opportunity?

• Initiative around sitter use

• Supports responding to trended data:

• Increase and/or reissuing RN FTEs

• Increase SRU RNs preassigned to an area

• Implement a knowledge bonus

• Utilize in decision making around distribution of resources

from SRU

• Helps to predict intervention needs and explain current state

• Trended data helps to show duration

Identify

Mitigate

Predict

4/12/2016

19

Summary of Data/Analytics

1. Performed statistical analysis to inform what measures might lead to harm (tested with CBDI/NICU)

2. Operationalized a microsystem stress measure that could be collected and sustained in the inpatient setting (nursing) (informed by PDAS cycles in CBDI/NICU)

3. Built a patient services operations system to collect and feed back the data (used for various nursing processes to ensure use). Incorporate PMRS dashboard reporting into PS system to help inpatient units mitigate.

4. Using SPC and empirical analysis to see if correlation exists between harm+concerns+803-SAFE calls (composite measure of “not good care”)

5. Future: determine if statistical analysis would show relationships with stress duration and outcome to help us be able to predict.

Microsystem Dashboard CONCEPTMicrosystem Outcomes

Microsystem “Key” Processes

(Nursing) Capacity Demand

DRAFT Some measures are not completely operationalized. Measures are owned by various groups.

Composite Measure

Harm & Stress - TCC

Time Period of Stress Data

Special Cause

Last Updated 3/9/2015 by A. Anneken, James M. Anderson Center for Health Systems Excellence

4/12/2016

20

Summary of Learnings

• Both quantitative and qualitative metrics are helpful in identifying unit and system level stress

• Standardized mitigation and escalation strategies expedite decision making and execution of interventions

• Examination of trending data supports prediction and early detection of stressed systems

Next Steps

• Incorporate year to date data into the report

• Consistently review and understand weekly trended data

• Quantitative and qualitative data utilized in decision making and resource allocation

• Spread to other mesosystems beyond inpatient

• Transition to utilizing new Daily Microsystem Report

Questions?

Documents

High Reliability and Microsystem Stress - Brady & Hausfield · 4/12/2016 1 High Reliability and Microsystem Stress Helping leaders identify and mitigate unit level stress: Next steps