TE TECUP UPP P Vi Virtual ual Meeting eeting Lordn via Adobe - - PDF document

6/12/2020 1

TE TECUP UPP P Vi Virtual ual Meeting eeting

Lordn via Adobe Stock

Agenda

9:00am PT/12:00pm ET Welcome & Introductions 9:15am PT/12:15pm ET Project Re-orientation 9:30am PT/12:30pm ET Beta Test Preliminary Findings / Q & A 10:00am PT/1:00pm ET Discussion:

• COVID impact on practices
• Stakeholder support
• Measure performance and quality improvement

10:50am PT/ 1:50pm ET Next Steps & Closing Comments 11:00am PT/ 2:00pm ET Close

2

1 2

6/12/2020 2

Katherine Ast, MSW, LCSW AAHPM Project Director Kelly McKenna, MA AAHPM Project Manager Joe Rotella, MD, MBA, FAAHPM (CMO) Wendy-Jo Toyama, MBA CAE (CEO) Rodney Tucker, MD MMM (President) AAHPM Team Leadership Sangeeta Ahluwalia, PhD Carrie Farmer, PhD RAND Project Directors Jessica Phillips, MS RAND Project Manager Amy Melnick, MPA Coalition Project Director Gwynn Sullivan, MSN Coalition Project Manager Cozzie King Coalition Manager Sydney Dy, MD, MS, FAAHPM Mary Ersek, PhD, RN, FPCN TECUPP Co-Chairs

Welcome (from the project team)!

3

Potential Conflicts of Interest Disclosed

• Name,

me, Creden edential tials – List conflict(s) noted here.

3 4

6/12/2020 3

Does anyone have any new conflicts nflicts of interest rest to report?

5 6

PROJECT RE-ORIENTATION

5 6

6/12/2020 4

7

TECUPP roles and responsibilities

• Provide input on key decisions
• Engage in group discussions
• Share informed opinions freely
• Remember the importance of a “by us, for us” process

8

Project goal:

Develop two wo patien ient-rep eported ed qu quali ality ty measu sures es of outpatien atient t palliati iative e car are e exper erience ience for CMS’s Merit-Based Incentive Payment System (MIPS) under the Quality Payment Program (QPP) created by MACRA

7 8

6/12/2020 5

Measures under development

9

Palliative care outpatients’ experience

• f feeling heard and understood

Measure Name Measure Name Measure Description Measure Description

Percentage of patients aged 18 years and

• lder per year who are fielded a patient

experience survey within 3 months of an

• utpatient palliative care visit, who report

feeling heard and understood by their palliative care provider and team over the last six months Palliative care outpatients’ experience

• f receiving desired help for pain

Measure Name Measure Name Measure Description Measure Description

Percentage of patients aged 18 years and

• lder per year who are fielded a patient

experience survey within 3 months of an

• utpatient palliative care visit, who report

having pain and wanting help for their pain, and who report getting the help they wanted for their pain by their palliative care provider and team over the last six months

Project timeline

10

2019 2020 2021 2022

Information Gathering

Sept 2018 – Jun 2019

Cognitive Testing

Mar 2019 – Feb 2020

Alpha Field Test

Aug – Oct 2019

Beta Field Test

Nov 2019 – Nov 2020

Public Comment

Dec 2020 – Feb 2021

Final Measure Specifications

Jan – Jun 2021

NQF Intent to Submit

Aug 2021 (Fall 2021 review cycle)

Submit Measures to MUC List Jun 2021

PREPARATION TESTING

NQF Submission

Nov 2021 [AAHPM]

TECUPP Meeting

Jun 2020

Final Business Case Oct 2020

FINALIZATION/ENDORSEMENT

9 10

6/12/2020 6

TESTING

12

Goal: l: Establish data collection procedures and prepare for national field test (beta)

• What response rates could we expect?
• Would eligibility criteria yield an adequate sample?
• Could a web-based survey be feasible for beta?
• What could preliminary data tell us about the items and measures?

Parti ticip ipants ants: : 300 patients across 5 outpatient palliative care programs Methods hods: : Mixed mode survey administration

Alpha field test

11 12

6/12/2020 7

Alpha field test: key findings

Respons nse rates were e as antic icipat ated ed

• 40% (120 completed surveys, out of 300 fielded), with a range of 26% to 53%

across 5 programs

Eligib igibilit ility criteria eria (1 outpati atien ent t visit it in 3 months hs) are feasib ible le and optim imal al

• Of 996 patients with 1 visit in 6 months, 662 (66%) had the eligible visit in 3

months

• Of 662 patients (48%), 318 had only one visit in 3 months

Web-base ased d survey is feasib ible le to test t in beta ta Potential for topping out with “heard and understood” (H/U) data elements

• Revised response option from “very true” to “completely true”

The convenien nience e sample le limit ited ed program gram and patient ient repres esen entat tation ion

• Beta testing will use a nationally representative sample of programs, distributed

by geographic region and program type

13 14

Goal: l: Examine reliability and validity of proposed quality measures and explore measure implementation for the QPP Sample le size e goal: : 6,000-7,500 sampled patients for 2,400-3,000 completed surveys (assuming 40% response rate) Methods hods: : “Enhanced” mixed mode administration (web to mail to phone) in both English and Spanish (planned, pending completed translation)

Beta field test

Schedule edule: : November 2019 to November 2020, but currently paused due to COVID-19; last round of fielding was 3/30/20 to 5/25/20

13 14

6/12/2020 8

Beta program recruitment

15

Hospi spice Midwes est Northea heast st South West st TOTAL AL Targeted Number of Programs 2 1 3 1 7 Programs Recruited (with executed DUAs) 2 2 5 1 10 10 Percent of Target 100% 200% 167% 100% 143% Hospi spital Midwes est Northea heast st South West st TOTAL AL Targeted Number of Programs 5 9 7 7 28 28 Programs Recruited (with executed DUAs) 5 6 7 7 25 25 Percent of Target 100% 67% 100% 100% 89% 89% Other her Midwes est Northea heast st South West st TOTAL AL Targeted Number of Programs 3 2 5 5 15 15 Programs Recruited (with executed DUAs) 2 3 2 1 8 Percent of Target 67% 150% 40% 20% 53% 53% TOTAL AL PRO ROGRAMS AMS 9 11 11 14 14 9 43 43

As of April 2020, we have received 914 completed beta test surveys

Survey ey adminis inistrat ation ion Number er Number of surveys fielded 2030 Number of patients eligible for inclusion 1811 Number of completed surveys 914 Mail surveys 424 (46%) Phone surveys 384 (42%) Web surveys 106 (12%) Response rate (914/1811) 51%

Patients removed from denominator 219 Deceased 165 Other 54

16

15 16

6/12/2020 9

Patient/respondent characteristics

Chara racter cterist stics cs (n=914) 4) % (n) or M (SD) Age 64.2

(sd = 13.8, N = 913)

Male 44.6%

(n = 407, N = 913)

Proxy Assistance 20.5%

(n = 187, N = 914) Reason for proxy assistance… Phone Mail Web Count of Proxy Response by Mode 72 102 13 Patient Helped Answer Some Questions 8.3% (n = 6)

• Read the questions to me
• 48%

(n = 49) 53.8% (n = 7) Wrote down the answers I gave

• 47.1%

(n = 48) 7.7% (n = 1) Answered for me

• 38.2%

(n = 39) 46.2% (n = 6) Translated into my language

• 2.9%

(n = 3) 0% (n = 0) Helped in some other way

• 5.9%

(n = 6) 0% (n = 0) 17

Patient/respondent characteristics (n=914)

18

17 18

6/12/2020 10

Most patients reported feeling heard and understood by their outpatient palliative care provider and team

DE# Questi tion

• n

Completely y True Very True Somewhat t True A Little Bit True Not at All True N

Q12 I felt heard and understood by this provider and team. 71.3% 21.3% 4.2% 2.1% 1.2% 908 Q13 I trusted this provider and team. 75.5% 18.4% 3.9% 1.7% 0.7% 909 Q14 I felt comfortable asking this provider and team questions. 79.4% 15.1% 3.3% 1.3% 0.9% 908 Q15 I could tell this provider and team anything, even things I might not tell anyone else. 60.5% 23.5% 10.8% 2.6% 2.7% 899 I felt this provider and team … Q16 Put my best interests first when making recommendations about my care. 73.6% 18.8% 4.7% 1.8% 1.1% 909 Q17 Always told me the truth about my health, even if there was bad news. 76.9% 17.0% 4.2% 1.2% 0.7% 904 Q18 Saw me as a person, not just someone with a medical problem. 78.4% 15.3% 4.0% 1.4% 0.9% 908 Q19 Knew what worried me most about my health. 64.0% 24.5% 8.9% 0.9% 1.8% 903 Q20 Understood what is important to me in my life. 64.7% 22.4% 9.9% 1.3% 1.8% 903 Q21 Would know what I would want done if I was unconscious or in a coma. 55.7% 22.3% 14.1% 2.6% 5.4% 875 19

Most patients who wanted help for their pain received the help they wanted

20

19 20

6/12/2020 11

Similarly, most who wanted emotional support from their provider and team reported receiving it

21

We evaluated the heard and understood (H/U) multi-item scale

22

• Inter-item correlations
• Item-total correlations
• Reliability
• Factor analysis

21 22

6/12/2020 12

Feeling “heard and understood” could be measured with a multi-item scale

I felt heard and understood by this provider and team I trusted this provider and team I felt comfortable asking this provider and team questions I could tell this provider and team anything, even things I might not tell anyone else I felt this provider and team put my best interests first when making recommendations about my care I felt this provider and team always told the truth about my health, even if there was bad news I felt this provider and team saw me as a person, not just someone with a medical problem I felt this provider and team knew what worried me most about my health I felt this provider and team understood what is important to me in my life I felt this provider and team would know what I would want done if I was unconscious or in a coma

23

Item total correlations and Cronbach’s alpha reliability of H/U items support unidimensional scale

Coefficient alpha = 0.94 DE # Data ta Elements ts Ite tem-tota

• tal correlatio

tion Q16 I felt this provider and team put my best interests first when making recommendations about my care 0.84 Q20 I felt this provider and team understood what is important to me in my life 0.82 Q19 19 I felt this provider and team knew what worried me most about my health 0.81 Q13 13 I trusted this provider and team 0.81 Q18 I felt this provider and team saw me as a person, not just someone with a medical problem 0.80 Q14 I felt comfortable asking this provider and team questions 0.78 Q12 I felt heard and understood by this provider and team 0.76 Q17 I felt this provider and team always told the truth about my health, even if there was bad news 0.73 Q15 15 I could tell this provider and team anything, even things I might not tell anyone else 0.66 Q21 I felt this provider and team would know what I would want done if I was unconscious or in a coma 0.52

24

23 24

6/12/2020 13

EFA and CFA of H/U items support multi-item unidimensional scale

Model fit from CFA: CFI=0.96, SRMR=0.03, RMSEA = 0.09 DE # Data ta Elements ts CFA loadings Q16 I felt this provider and team put my best interests first when making recommendations about my care 0.89 Q13 13 I trusted this provider and team 0.85 Q18 I felt this provider and team saw me as a person, not just someone with a medical problem 0.83 Q20 I felt this provider and team understood what is important to me in my life 0.83 Q19 19 I felt this provider and team knew what worried me most about my health 0.82 Q14 I felt comfortable asking this provider and team questions 0.82 Q12 I felt heard and understood by this provider and team 0.80 Q17 I felt this provider and team always told the truth about my health, even if there was bad news 0.76 Q15 15 I could tell this provider and team anything, even things I might not tell anyone else 0.68 Q21 I felt this provider and team would know what I would want done if I was unconscious or in a coma 0.53

25

H/U composite (a few things to consider)

Item content ent redunda dancy y (e.g., g., doublet ets) s)

• Do we need both items?
• If not, which do we retain?

Imp mpor

• rta

tant content ent comp mponen

• nents

ts to inclu lude e in comp mposit

• site

e

• Broad (e.g., Q12) vs. specific item content (e.g., Q21)

Final al number of items (e.g., g., all items ms, , 6 i items ms, , 4 i items ms) to be used for comp mpos

• sit

ite e

• What is the smallest set of items that would be acceptable?
• How does the performance of different possible composites (e.g., 4

item vs 6 items) compare?

26

25 26

6/12/2020 14

Once options for a composite are developed, we can proceed with validity analyses

Converge ergent nt va validi idity ty Association between H/U composite and:

• CAHPS Communication

composite

• Receiving desired help for pain
• Receiving desired emotional

support

27

Know

• wn

n group ups va validi idity ty Differences in H/U composite, receiving desired help for pain, and receiving desired emotional support by:

• PHQ-2 positive screen
• Cognitive function
• Overall general health
• Gender

Next steps for testing

This is preli limi minar ary y data only—in interpret with h caution

• n

Further er develop lopmen ment of the measure: e:

• Analyses are planned to be repeated with more data
• Factor loadings were exploratory and additional analyses will be

performed

• Risk adjustment and reliability analyses will require more data to

inform

COVID ID-19 9 may y affect t respons

• nse

e pattern erns, s, and analy lyses es may need d to be adjusted ed

28

27 28

6/12/2020 15

DISCUSSION COVID impact on practices

• What is the impact on palliative care practices,

particularly changes to care delivery?

• What is the impact on patient experience?
• What is the impact on palliative care quality

measurement?

30

29 30

6/12/2020 16

Stakeholder support

• What does broad (and successful) support for

these measures look like?

• How best can we reach the key audiences for these

measures?

• What do you perceive as the value of these

measures?

31

Measure performance and quality improvement

• How might programs use measure information

to drive quality improvements?

• What would quality improvement around “heard

and understood” and “receiving help for pain” look like?

32

31 32

6/12/2020 17

33

NEXT STEPS & CLOSING COMMENTS

Th Thank nk you! u!

Lordn via Adobe Stock

33 34