innovation spotlight - himss20...1 innovation spotlight session #ni2, february 19, 2017 tanna...
TRANSCRIPT
1
Innovation Spotlight Session #NI2, February 19, 2017
Tanna Nelson, MSN, RN-BC, CPHIMS, Texas Health Resources
Sally Okun, VP Advocacy, Policy & Patient Safety, PatientsLikeMe
Kimberly Ellis Krakowski, MSN, RN, CENP, CAHIMS, Associate Chief Nursing Information Officer, Inova Health System
2
Conflict of Interest
Tanna Nelson, MSN, RN-BC
Kimberly Ellis Krakowski, MSN, RN, CENP, CAHIMS
Have no real or apparent conflicts of interest to report.
3
Conflict of Interest
Sally Okun
Salary: PatientsLikeMe
Consulting Fees (e.g., advisory boards): Commonwealth Fund, PCORI
Ownership Interest (stocks, stock options or other ownership interest excluding diversified mutual
funds): PatientsLikeMe
4
Learning Objectives
• Outline the reasons why applying innovation will bring value to an
organization
• Describe lessons learned when developing a new approach or application
• Discuss how to overcome challenges and realize benefits when developing
and implementing an innovative solution
5
Predictive Analytics for Reducing Readmissions within 30 days of Discharge
Session #, February 19, 2017
Tanna Nelson, MSN, RN-BC, CPHIMS, Texas Health Resources
6
Texas Health Resources - Organizational Background• Texas health resources is one of the largest faith-
based, nonprofit health care delivery systems in the united states and the largest in north Texas in terms of patients served.
• The system's primary service area consists of 16 counties in north central Texas, home to more than 6.8 million people.
6
7
Texas Health Resources• 25 hospitals in North Texas
• 14 wholly owned hospitals
• 133,903 Inpatient Visits
• 1,238,392 Outpatient Encounters
• 469,309 ED Visits
• 89,452 Surgeries
• 27,200 Deliveries
• 5,500 Active Physicians
• 7,500 RN’s
• 22,000 Employees
8
Business Model
9
Texas Health Resources & Readmission Risks
• Used multiple ‘brand’ name readmission risk indicators for 3 years
– Not effective/efficient enough in targeted outreach
– Some tools proprietary and risk factors were unknown
• Gap in managing and reducing readmissions
– Limited resources
– Can’t reach every patient but need to reach the right patients
• Requested for more data that defined our unique population
• Formation of a Readmission Taskforce
10
Latest Readmission Risk Indicator Tool: LACE+
11
Challenges: Data• All Variables aren’t in EHR:
– Case-mix group (CMG) score reduces c-statistic (0.753 vs. 0.743) (van Walraven, Wong, & Forster, 2012)
– Alternate Level of Care (ALC) Status
• Disease Conditions vague (mild, moderate, severe):
– Difficult to interpret and maintain (Quan et al., 2005)
– Literature guidance out of date
• Documentation inconsistency (Problem List vs. Patient History)
• Source of truth: Registration vs. Clinical documentation
12
Challenges: Clinical Interventions• Resource constraints
– Unable to address all high risk patients (~4,500/month)
• Risk stratification:
– Too many high risk patients identified who did not readmit (84.6%)
– Too many high utilization patients with low or medium risk scores
13
Formation of Innovation Group • Membership
– Two physician champions
– Nursing/Nursing Informatics
– Clinical Analytics/Biostatistician
– Care Transition Managers
• Focus
– Concept development
– Version review and approval, ensuring tool fits into provider workflows
– Development of interventions
– EHR feasibility, maintainability, replicability
14
Strategic Goal
• Create a predictive scoring tool:– Tailored to THR’s specific readmission populations
– All attributes must be available in EHR prior to discharge
– Improve the trustworthiness of the risk designation
• Focus intervention efforts on High Risk and Medium-High Risk patients
– Remain budget and resource neutral while providing complex case management to 100%
of high risk cases
– Providing manageable workloads
• C-stat goal of 0.78 to 0.80+– Elevate from a fair tool to a good tool
15
Clinical/Nursing Informatics Responsibilities• Analytics applications: SAS EG and SPSS Modeler, Statistics
• Created test environment for algorithm changes and statistical analysis
• Identification of source of truth and documentation reliability
• Data mining and validation
• Dataset preparation
• Determining and prioritizing indicators for evaluation
• Evaluation of indicators and readmission risk
• Variable weighting/scoring
• Build and testing in EHR
• Training and implementation
16
Texas Health Readmission Indicator List (THRIL)
• Systematic analysis
• Incremental change
• Careful evaluation of impact
17
From LACE+ to THRILv1• Addressed source of truth issues
• Reweighted disease conditions based on readmission rates
• Utilized patient history documentation in addition to Problem List
• Added new conditions (sepsis, antepartum complications, pneumonia)
• Re-stratified risk categories
• Adjusted age ranges, admission counts, point assignments
• Added raw counts of ED utilization and hospital admissions to target high utilizers
18
Case Management InterventionsLow risk = ≤ 28
– DC Education begins on day of admission; meds reconciled; follow-up appointment made by the CNL
Medium risk = 29-58– DC Education begins on day of admission; find a PCP if necessary; CTM makes follow-up
appointment; Meds reconciled; community resources as indicated
Medium-High risk = 59-80 – DC education begins on day of admission; CTM arranges home health, rehab, skilled care
based on criteria and patient acuity. Refer to Transition Housecalls if possible.
High risk = ≥ 81– Complex case management; assessment for advance directives, end of life planning,
palliative care / hospice appropriateness
19
Readmission Rate for ‘High Risk’ Patients
15.4% 15.4%16.6% 14.6% 14.7%
29.2% 28.4% 24.7% 26.0% 23.5% 26.0%
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16 Dec-16
Coun
t of
Patients
Cate
gori
ze
d
as H
igh R
isk
Readmitted
Not Readmitted
The height of each bar represents the total number of patients categorized as ‘High Risk’ for readmission.
The percentage displayed above each bar is the readmission rate for the ‘High Risk’ patient population. Higher percentages are better,
meaning we are identifying more readmitters in the High Risk bucket.
LACE+ THRIL v1
20
Comparison of LACE+ to THRILv1
0.66
0.68
0.7
0.72
0.74
0.76
0.78
0.8
C-s
tat S
core
Oct-15 Nov-15 Dec-15 Jan-16 Feb-16 Mar-16 Apr-16 May-16 Jun-16 Jul-16 Aug-16 Sep-16 Oct-16 Nov-16
LACE+ 0.712 0.721 0.73 0.744 0.754 0.75 0.748 0.744
THRIL v1 0.784 0.771 0.77 0.777 0.761 0.76
21
Where Are We Going?
LACE + Updates THRIL v1
22
From THRILv1 to THRIL v2
– Medical History list count
– Surgical History list count
– Allergy list count
– Schedule I & II allergy count
– Braden Score <19 at discharge
– Existence of a Pressure Ulcer
– Count of pain score of 10 is reported
– Isolation status
– Marital status
– Payer
– Substance abuse
– Behavioral Health diagnosis
• Expand Palliative Care programs
• Widen focus to include Medium-High risk patients
• Incorporate new attributes:
23
Lessons Learned– The making of a predictive tool is not a short-term project
– Requires dedicated resources and project management
– Allow for ample time to test and adjust scores and weights
– Avoid scope-creep
– Study the marketplace for attributes
– Be patient
24
ReferencesQuan, H., Sundararajan, V., Halfon, P., Fong, A., Burnand, B., Luthi, J., . . . Ghali, W. (2005). Coding
algorithms for defining comorbidities in ICD-9-CM and ICD-10 administrative data. Medical Care, 43(11), 1130-1139.
van Walraven, C., Wong, J., & Forster, A. (2012). LACE+ index: extension of a validated index to predict early death or urgent readmission after hospital discharge using administrative data. Open Med, 6(3), e80-e90. Retrieved September 2016, from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3659212/
26
Patient Generated Data Driving InnovationSession ID: NI2 February 19, 2017
Sally Okun, VP Advocacy, Policy & Patient Safety
PatientsLikeMe
27
28
About PatientsLikeMe
Our mission is to improve the lives of patients through new knowledge derived from
shared real-world experiences and outcomes
• Founded in 2004 as a direct response to one
family’s experience with chronic disease
• Online, open, patient-facing community for patients
living with and managing illness
• Started in ALS, expanded to any condition in 2011
• Deep patient data and experience in many life-
changing conditions
• 30+ million structured data points
• 3+ million free-text posts
• 15+ PRO measures
• 460,000+ patients
• 2,500+ conditions
• 90+ peer-reviewed publications
• Patient-generated taxonomy
• Patient-informed principles
Patients Data Insights
Basic Information (age, sex, etc.)
Diseases, Conditions(early signs, diagnosis status, etc.)
Treatments & Side Effects(Rx, OTC, Supp., non-drug, etc.)
General & Specific Symptoms(onset, severity status, etc.)
Quality of Life & Behavior Status(all patients, some disease specific)
Outcome Measures of Disease(disease dependent)
Patient-generated narrative data in forum
discussions, journals and feeds
Emerging data source experiments(wearable/sensors, EHRs, claims, 'omics, specimens)
Engagement
Knowledge
Evidence
Standards
Data Integrity
Empowerment
Patient Data Conventions
Patient voice
translated into
computable
clinically relevant
data elements
Data codified using:
• ICD10
• SNOMED
• MedDRA
• ICF
31
32
Patient-informed principles
33
• Sufficient rigor for peer-review publication in leading journals
• Publish open access wherever possible
• Every patient has opportunity to take part in research if they want
• Use of IRB to ensure ethical conduct & oversight
• Describe patient reported and generated data objectively
• Provide “givebacks” to show patients the value of taking part in research
• Uphold the core values of PatientsLikeMe established in 2004:
o Patients first – always
o Honor patients’ trust – always be principled stewards of patients' data
o Transparency – always be clear about who we are working with
o Openness – always empower bidirectional sharing and communication
Research science principles
34
Understanding unmet needs
Uncontrollable yawning in ALS • Some ALS patients reported yawning
dozens of times per day, sometimes
painfully dislocating their jaws
• We added “Excessive yawning” to
symptom list and gathered data from
254 ALS patients in just 4 weeks
• Found association between yawning
severity and patients whose first
symptoms were in their mouths and
throat vs limbs
• Impact: Identified possible drug
target, unmet need for symptom
relief, and contributed to medical
hypothesis generation
mild926 patients (36%)
none1058 patients (41%)
moderate503 patients (19%)
severe94 patients (4%)
35
Quality of care in epilepsy
• Partnered with American Academy of
Neurology to develop a self-report
instrument to illuminate patient experience
with current state of care
• Found significant differences between
physician types; patients treated by non-
specialists receiving less quality care
• Identified care gaps around side effect
management, surgery referral, reproductive
issues in women
• Impact: Lead to changes in neurology
training and informed quality measures in
epilepsy for National Quality Forum
Illuminating care differences
36
Sleep issues more prevalent in chronic illness
• Insights from patient-reported data to inform future clinical development
• Cross condition study of 51,000 with insomnia and survey of 5256 patients across 11 comorbid conditions
• Most patients not diagnosed with sleep disorder and consequently not being treated for sleep problems
• Impact: Insights shaped methods of education; identified strong link between chronic conditions and sleep problems; targeting for clinical trialsnyt
Insights about daily life
37
Empowering action
Patient-Led Trial of Lithium to Slow ALS • A small Italian study suggested lithium carbonate significantly slowed ALS (N=16 treated)
• Over 160 PLM patients sought lithium from providers and tracked their outcomes
• We developed a matching algorithm using historical controls instead of a placebo group
• Refuted findings of original study within 6 months
• Impact: Hypothesis to result 3-5 years faster than multiple phase III RCT’s. Patients stopped using ineffective treatment
38
...in our increasingly connected and networked world
data reported and generated by our real lived experiences
is essential if we are to achieve the promise of a
continuously learning health system
To learn, listen well to impressions voiced by patients first~sally okun~
39
Questions
Sally Okun
Twitter: @SallyOkun
LinkedIn: https://www.linkedin.com/in/sally-okun-3139a02
Please remember to complete your online session evaluation
40
Dancing with Disruption February 19, 2017
Kimberly Ellis Krakowski, MSN, RN, CENP, CAHIMS
Associate Chief Nursing Information Officer
Inova Health System
Organization logo(s) may be placed on this slide
41
Speaker Introduction
Kimberly Ellis Krakowski, MSN, RN, CENP, CAHIMS
Assoc. Chief Nursing Information Officer,
Sr. Director of Informatics and Clinical Applications
Inova Health System
42
SBAR
Situation
• Inova Health System was approached by Booz Allen Hamilton to partner and
conduct research using wearable devices and proprietary algorithms.
43
SBAR
Background
• While the first wearable device was created in 1961 by two mathematicians
to cheat at roulette, it was when the original FitBit electronic activity tracker
entered the market in 2009 that it became a popular household item rather a
newsworthy technology.
44
SBAR
Assessment
• The organization’s EHR offers functionality to import a patient’s wearable
health device data into the patient portal and to become part of the patient’s
health record where it will be reviewed by a health provider.
• The organization made the decision to configure this functionality.
1. Congestive Heart Failure patient population for monitoring activity, HR
and weight.
2. Post operative orthopedic patients for monitoring activity levels.
45
SBAR
Recommendation:• Research Option #1
– Provide Orthopedic patients wearable activity devices at discharge for
outpatient monitoring. The bedside nurse would be responsible for
provide education and documenting proficiency to the
patient/caregivers.
• Research option #2
– Provide nurses a wearable activity devices to measure activity and
sleep patterns and compare within rotating and non-rotating shifts.
46
IRBObjectives• The primary objective was to remotely identify differential sleep
patterns between rotating/non-rotating nurses for sleep length and quality.
• The secondary objective was to identify optimum engagement methods with nurses by sharing their sleep patterns with encouragement to enhance sleep quality.
• An additional goal of this study are to expose nurses on the utility of wearable sensors for ultimate application towards future patients.
47
https://www.youtube.com/watch?v=GA8z7f7a2Pk&sns=em
48
Questions?
49
Thank you!