jonas&de&souza,&md,&mba assistant&professor&&& · 2018-09-24 ·...

Post on 09-Jul-2020

0 Views

Category:

Documents

0 Downloads

Preview:

Click to see full reader

TRANSCRIPT

   

Jonas  de  Souza,  MD,  MBA  Assistant  Professor      

The  University  of  Chicago  

•  I  have  no  conflicts  of  interests  

Disclosures

•  I  have  no  conflicts  of  interests  today  – Moving  to  Humana  in  July  2017  –  I  will  have  conflicts  of  interests  

Disclosures

•  Describe  how  we  developed  and  validated  the  COmprehensive  Score  for  financial  Toxicity  PRO  

•  Illustrate  applicaIons  of  the  COST-­‐PRO  in  pracIce  

•  Discuss  potenIal  future  direcIons  

Financial Toxicity

Costs to the payer

•   I  have  no  conflicts  of  interest  to  disclose  

 

Presented  by:  Jonas  de  Souza,  MD  

MEDICARE 2004 PPPY: $37799

MEDICARE 2014 PPPY: $51566

COMMERCIAL 2004 PPPY: $55789

COMMERCIAL 2014 PPPY: $90656

Costs to the payer

•   I  have  no  conflicts  of  interest  to  disclose  

 

Presented  by:  Jonas  de  Souza,  MD  

MEDICARE 2004

MEDICARE 2014

COMMERCIAL 2004

COMMERCIAL 2014

62%

36%

Cost-sharing is increasing

Presented  by:  Jonas  de  Souza,  MD  

VISIBLE TO THE PATIENTS

Patients only see the tip of the iceberg

Presented  by:  Jonas  de  Souza,  MD  

•  COST-­‐SHARING,  PREMIUMS  •  INDIRECT  COSTS  (LOSS  OF  INCOME)  •  DIRECT  NON-­‐MEDICAL  (CHILDCARE,  ETC)    

The tip of the iceberg sinks the ship

VISIBLE TO THE PATIENTS •  COST-­‐SHARING,  PREMIUMS  •  INDIRECT  COSTS  (LOSS  OF  INCOME)  •  DIRECT  NON-­‐MEDICAL  (CHILDCARE,  ETC)    

BANKRUPTCY IN CANCER PATIENTS LINKED TO EARLY

MORTALITY (RAMSEY ET AL)

FINANCIAL BURDEN LINKED TO HRQOL

(ZAFAR ET AL) (FENN ET AL)

COST-SHARING RELATED TO ADHERENCE

(DUSETZINA ET AL)

Presented  by:  Jonas  de  Souza,  MD  

“A cancer patient killed himself after he was told he had been “A cancer patient killed himself after he was told he had been refused a wonder drug by his local trust.” a wonder drug by his local trust.”

Ratain MJ: Biomarkers and Clinical Care. AAAS/FDLI Colloquium, Personalized Medicine in an Era of Health Care Reform, 2009

Financial  Toxicity  Grade  5?  

Ratain MJ: Biomarkers and Clinical Care. AAAS/FDLI Colloquium, Personalized Medicine in an Era of Health Care Reform, 2009  

A  PaIent-­‐Centered  View  

A  lot,  some,    a  liYle,  not  at  all  

From  qualitaFve  to  quanFtaFve  

 COST  =  COmprehensive  Score  for  financial  Toxicity  

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  Development of the COST-PROM

de Souza JA, Yap B, Hlubocky FJ, Cella D, Ratain MJ, Daugherty CK. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

•   PaIents  with  advanced  cancers  

•   Who  received  chemotherapy  for  at  least  3  months  

•   Last  treatment  not  longer  than  6  months  prior  to  the  assessment      

A  PaIent-­‐Centered  View  Development of the COST-PROM – 1/4 steps

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

•   The  qualitaIve  interviews  with  20  paIents,  literature  review  and  expert  opinion  resulted  in  147  items  

•   A_er  reducIon  due  to  overlapping  content,  a  total  of  58  items  were  retained:  31  items  from  paIents,  25  items  from  the  literature,  2  from  experts    • 58  items  were  classified  into  8  financial  items,  13  resource  items,  17  affect  items,  10  coping  items,  and  10  family  items    

58 items retained

A  PaIent-­‐Centered  View  Development of the COST-PROM – 1/4 steps

A  PaIent-­‐Centered  View  Development of the COST-PROM – 2/4 steps

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

•   Importance  score  =  frequency  x  importance    •   Retain  all  items  unIl  at  least  3  items  per  domain  had  been  retained  –  28  items  excluded  

•   30  items  going  to  step  3    

58 items retained

30 items retained

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

58 items retained

30 items retained

Inter-item correlation Item-total correlation

Preliminary factor analysis

35 patients

46 patients

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

 •   Following  FACT  formadng  

•   CollecIng  sociodemographics      

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

•  Pairs of items with an IIC >0.7 were identified and, within each correlated pair, the item with the highest importance score in step 2 was retained

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

•  Pairs of items with an IIC >0.7 were identified and, within each correlated pair, the item with the highest importance score in step 2 was retained

•  Items with nonsignificant item-total correlations (ITCs) (ie, not significantly related to the instrument total score; P  >  .05) were excluded

35 patients

46 patients

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

• Interim  factor  analysis  revealed  2  factors,  an  items/factor  raIo  of  6  and  wide  communaliIes  

16 items retained

13 items retained

A  PaIent-­‐Centered  View  Development of the COST-PROM – 3/4 steps

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

16 items retained

13 items retained

58 items retained

30 items retained

•   Based  on  preliminary  analysis,  simulaIon  work1  esImated  a  minimum  of  95  paIents  would  be  considered  appropriate  to  factor  analysis  1 - Mundfrom DJ, Shaw DG, Tian Lu K. Minimum sample size recommendations for conducting factor analyses. Int J Testing. 2005;5: 159-168.

A  PaIent-­‐Centered  View  Development of the COST-PROM – 4/4 steps

20 patients

35 patients

46 patients +

54 patients

Factor Analysis

16 items retained

13 items retained

58 items retained

30 items retained

A  PaIent-­‐Centered  View  Development of the COST-PROM – 4/4 steps

20 patients

35 patients

46 patients

100 patients 46 from step 3 plus 54 only for factor analysis

•  Items  with  loadings  less  than  0.5  to  be  removed  

•   Two  items  removed:      “My  insurance  does  not  provide  adequate  treatment  coverage  for  my  illness  and  care-­‐related  expenses”  (0.44)    “I  rely  on  friends  or  family  to  help  with  the  costs  of  health  care”  (0.43)  

Final: 11 items

 COST  =  COmprehensive  Score  for  financial  Toxicity  

 •   11  items  •   1-­‐  factor  soluIon  explained  93%  of  variance  in  the  data  •   Cronbach-­‐alpha  0.9  •   Mean  IIC:  0.47    •   Mean  ITC:  0.71  •   1  financial  item,  2  resources  and  8  affect  items    

   de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  Development of the COST-PROM

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

    Clinical  Relevance?

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  COST-PROM - HRQOL

•   Financial  Toxicity  (Hypotheses)  •   Correlates  with  HRQOL  •   Independent  of  mood  states  •   Independent  of  personal  adtudes  or  traits  •   Independent  of  willingness  to  discuss  costs  •   Controlled  for  clinical  trial  parIcipaIon    

Approached:    375  paIents  who  have  been  on  chemotherapy  for  at  least  3  months  

Agreed  to  ParFcipate:  236  paIents  (62.9%)  

Analyzed:  233  paIents  (62.1%)  

Withdrew:  3  paIents  due  to  the  financial  nature  of  the  study  

(0.8%)  Declined  to  ParFcipate:  139  paIents  (37.1%)  

233 patients with advanced stage cancers were recruited at UCM-affiliated centers

de Souza, J. A., Yap, B. J., Wroblewski, K., Blinder, V., Araújo, F. S., Hlubocky, F. J., Nicholas, L. H., O'Connor, J. M., Brockstein, B., Ratain, M. J., Daugherty, C. K. and Cella, D. (2017), Measuring financial toxicity as a clinically relevant patient-reported outcome: The validation of the COmprehensive Score for financial Toxicity (COST). Cancer, 123: 476–484.

Characteristics of respondents and non-respondents in the study

Respondents n = 233

Non-respondents n = 142

p-value Age (mean±SD) 58.4±11.5 63.6±12.8 <0.001 Gender

Female 136 (58.4%) 69 (48.9%) 0.08 Male 97 (41.6%) 72 (51.1%)

Marital Status Married 170 (73.3%) 87 (61.3%) 0.02 Divorced/separated/widowed 38 (16.4%) 26 (18.3%) Never married 24 (10.3%) 29 (20.4%)

Race/ethnicity Caucasian, non-Hispanic 154 (66.4%) 80 (56.3%) 0.07 Non-Caucasian 78 (33.6%) 62 (44.6%)

Insurance type Private or employer-based 144 (61.8%) 72 (50.7%) 0.07 Medicare (+ or - supplement) 73 (31.3%) 59 (41.6%) Medicaid 13 (5.6%) 11 (7.8%) COBRA continuation coverage 3 (1.3%) 0 (0.0%)

Household income (mean±SD)1 $68,107±28,232 $62,936±25,718 0.07 P-values are from t-tests or chi-square tests between participants and non-participants in the study.1 Income was based on Census Data related to a patient’s zipcode. !

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  COST-PROM - HRQOL

COST-PROM

HRQOL

Mood States

Social Desirability

OOP

Admissions

Compliance

Willingness to Discuss Costs

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  COST-PROM - HRQOL

COST-PROM

HRQOL

Mood States

Social Desirability

OOP

Admissions

Compliance

Willingness to Discuss Costs

•   Brief  POMS:  r  =  -­‐0.26,  p  <0.01  

•   Marlowe-­‐Crowne  Social  Desirability:  r=  0.16,  p  =  0.28  

•   FACT-­‐G:  r  =  0.42,  p  <  0.001  •   PWB:  r  =  0.35,  p  <  0.05  •   SWB:  r  =  0.33,  p  <  0.05  •   EWB:  r=  0.32,  p  <  0.05  •   FWB:  r  =  0.41,  p  <  0.01  

de Souza JA, Yap B, Hlubocky FJ, et al. The Development of a Financial Toxicity Patient Reported Outcome in Cancer: The COST Measure. Cancer. 2014. http://onlinelibrary.wiley.com/doi/10.1002/cncr.28814/abstract

A  PaIent-­‐Centered  View  COST-PROM - HRQOL

OOP

Admissions

Compliance

Willingness to Discuss Costs

•   ParIal  correlaIons    •   Age,  ECOG  PS  •   variables  found  to  be  significantly  associated  with  COST  on  mulIvariate  analyses:    

• Income,  employment,  ethnicity,  inpaIent  admission,  psychological  distress  

•   FACT-­‐G:  0.31,  p<0.01  

•   EORTC:  0.20,  p<0.001  

A  PaIent-­‐Centered  View  COST-PROM - HRQOL

COST-PROM Income (FPL)

OOP

Admissions

Compliance •   r=0.28,  p<0.01  

ER visits

Admissions

A  PaIent-­‐Centered  View  COST-PROM - Predictors

No statistically significant interactions were found, including the interaction between income and employment.

•  PaIent-­‐centered  •  ScienIfically-­‐derived  •  Correlated  with  a  meaningful  outcome  (HRQoL)  

   

A  PaIent-­‐Centered  View  COST-PROM

•  PaIent-­‐centered  •  ScienIfically-­‐derived  •  Correlated  with  a  meaningful  outcome  (HRQoL)  

   Useful  in  clinic            or  research?  

 

A  PaIent-­‐Centered  View  COST-PROM

ApplicaIon  Cross-­‐secIonal  Studies  in  Single  Diseases

ApplicaIon  Understanding  Cost  Discussion  Preferences

ApplicaIon  Understanding  Cost  Discussion  Preferences

Preferred  Cost  Communicator:  Who  do  you  want  to  talk  to  about  costs?    

Note: 3% of patients selected “other” as their preferred cost communicator. “Other” responses included: “expert in the hospital”, “hospital administrator”, “I don’t know”, “Medicare/Medicaid”, “whoever can negotiate with insurance” and 2 responses of “insurance company employee”.

Preferred  Cost  Communicator:  Who  do  you  want  to  talk  to  about  costs?    

•  Physician:    –  PaIents  with  less  financial  toxicity  are  more  likely  to  want  to  discuss  

costs  with  their  physician  (OR=  1.02,  p=0.05)  –  Those  on  oral  drugs  are  3.4  Fmes  more  likely  than  those  on  IV  therapy  to  

want  to  talk  to  their  physicians  about  costs  (p=0.005)  

•  Social  Worker:  –  PaIents  with  increased  financial  toxicity  are  more  likely  to  want  to  speak  

to  a  social  worker  (OR=  0.97,  p=0.04)  –  Non-­‐Caucasians  are  2.26  Fmes  more  likely  to  want  to  speak  to  a  Social  

Worker  (p=0.01)  

•  Financial  Counselor:  –  A  trend  is  seen;  paIents  who  want  to  speak  to  a  financial  counselor  have  

increased  financial  toxicity  (OR=0.60,  p=0.16)  

Preferred  Cost  Communicator:  Who  do  you  want  to  talk  to  about  costs?    

•  Physician:    –  PaIents  with  less  financial  toxicity  are  more  likely  to  want  to  discuss  

costs  with  their  physician  (OR=  1.02,  p=0.05)  –  Those  on  oral  drugs  are  3.4  Fmes  more  likely  than  those  on  IV  therapy  to  

want  to  talk  to  their  physicians  about  costs  (p=0.005)  

•  Social  Worker:  –  PaIents  with  increased  financial  toxicity  are  more  likely  to  want  to  speak  

to  a  social  worker  (OR=  0.97,  p=0.04)  –  Non-­‐Caucasians  are  2.26  Fmes  more  likely  to  want  to  speak  to  a  Social  

Worker  (p=0.01)  

•  Financial  Counselor:  –  A  trend  is  seen;  paIents  who  want  to  speak  to  a  financial  counselor  have  

increased  financial  toxicity  (OR=0.60,  p=0.16)  

Preferred  Cost  Communicator:  Who  do  you  want  to  talk  to  about  costs?    

•  Physician:    –  PaIents  with  less  financial  toxicity  are  more  likely  to  want  to  discuss  

costs  with  their  physician  (OR=  1.02,  p=0.05)  –  Those  on  oral  drugs  are  3.4  Fmes  more  likely  than  those  on  IV  therapy  to  

want  to  talk  to  their  physicians  about  costs  (p=0.005)  

•  Social  Worker:  –  PaIents  with  increased  financial  toxicity  are  more  likely  to  want  to  speak  

to  a  social  worker  (OR=  0.97,  p=0.04)  –  Non-­‐Caucasians  are  2.26  Fmes  more  likely  to  want  to  speak  to  a  Social  

Worker  (p=0.01)  

•  Financial  Counselor:  –  A  trend  is  seen;  paIents  who  want  to  speak  to  a  financial  counselor  have  

increased  financial  toxicity  (OR=0.60,  p=0.16)  

•  CollaboraIon  with  the  North  American  Thyroid  Cancer  Survivorship  Study  

•  591  paIents  with  thyroid  cancer  within  the  past  6  years  were  surveyed  in  2  countries:    

–  553  (93.6%)  in  the  U.S.,  and  38  (6.4%)  in  Canada  

•  In  mulIvariate  analyses,  the  independent  predictors  of  worse  Financial  Toxicity  were  lower  income  (p  <  0.001),  female  gender  (p  =  0.01),  lower  educaIonal  level  (p=0.002),  treatment  in  the  U.S.,  (p=0.002),  and  worse  HRQoL  (p<0.001)  

ApplicaIon  Comparison  of  Distress  Between  Countries

www.costofcancercare.org  

www.costofcancercare.org  (as  of  Feb  2017)  

�  Agreement with Patient Access Network Foundation �  All patients received co-pay assistance �  Patients assessed at baseline, 1-month, 3-month and

6-month �  308 patients with baseline, 1-month and 3-month

data

www.costofcancercare.org  (as  of  Feb  2017)  

�  COST changes over time!

�  COST (FT) improved over 3 months

�  HRQoL did not improve

Validation sample: mean COST  ±  SD, 22.23  ±  11.89)

www.costofcancercare.org  (as  of  Feb  2017)  

�  Score changes over time! �  Can find out why their

financial toxicity improved? �  Self-report use of

navigators, financial counselors, transportation vouchers, support groups

�  Or only the copay assistance?

Validation sample: mean COST  ±  SD, 22.23  ±  11.89)

www.costofcancercare.org  (as  of  Feb  2017)  

�  179 pts (42%) had a navigator, 106 (34%), a social worker, 107 (35%), a financial counselor, 94 (31%) a support group, and 50 (16%), received a transportation voucher

� None of these “interventions” were associated with improved financial toxicity

www.costofcancercare.org  (as  of  Feb  2017)  

�  179 pts (42%) had a navigator, 106 (34%), a social worker, 107 (35%), a financial counselor, 94 (31%) a support group, and 50 (16%), received a transportation voucher

� None of these “interventions” were associated with improved financial toxicity

� Annals of obvious research: in a selected population of patients seeking help to pay for their drugs, nothing much helps but paying for their drugs!

   COST

FACT-G

Grade 0 Financial Toxicity

Grade 1

Grade 2

Grade 3

Hig

her b

ette

r qua

lity

of li

fe

Sig

nific

ant a

nd M

eani

ngfu

l

CTCAE (Common Terminology Criteria for Adverse Events) •  Grade 0, none •  Grade 1 Mild; asymptomatic or mild symptoms; clinical or diagnostic

observations only; intervention not indicated. •  Grade 2 Moderate; minimal, local or noninvasive intervention indicated •  Grade 3 Severe or medically significant but not immediately life-threatening

ApplicaIon  Grading  Financial  Toxicity  

Currently  about  865  paIents

Future  direcIons  

�  ICD10 code for financial toxicity

�  Post-marketing measurement �  A PRO to be required in the post-approval drug

process?

Future  direcIons  My  Dream  Approach  to  Financial  Toxicity  

(aka  “my  interrupted  project”)  

�  We are making progress in survival and outcomes

�  We do expensive things to our patients �  Let’s not ever forget how we also impact

their finances �  Let’s use science and evidence to develop

solutions

In  Summary  

 Table 1: Demographics, Part 1 Survey Participants (n=233)

Age Mean, median [Range]

  58.4, 59 [27-88]

Sex Female Male

  58.4% (136) 41.6% (97)

Race Caucasian, Non-Hispanic Non-Caucasian

66.4% (154) 33.6% (78)

Employment Status Working (Full or Part-Time) Unemployed Retired Disability Other

  33.6% (78) 9.1% (21) 32.8% (76) 19.4% (45) 5.1% (12)

Education Level Less than college Some college Completed college or higher

  22.3% (52) 30.5% (71) 47.2% (110)

Marital Status Married/Cohabitating Divorced/Widowed/Separated Single, Never Married

  73.3% (170) 16.4% (38) 10.3% (24)

Insurance Type Private/Employer Insurance Medicare Medicaid COBRA

  61.8% (144) 31.3% (73) 5.6% (13) 1.3% (3)

Clinical Trial Status Yes No

  27.9% (65) 72.1% (168)

 Table 1: Demographics, Part 2 Survey Participants (n=233)

Length of Disease < 1 year > 1 year

  38.6% (90) 61.4% (143)

Line of Therapy Adjuvant/Neoadjuvant 1st Line Metastatic 2nd Line Metastatic 3rd Line Metastatic 4th Line Metastatic or Later

  20.2% (47) 35.6% (83) 18.9% (44) 13.7% (32) 11.6% (27)

COST Score Mean, median [Range]

  22.2, 23 [0-44]

Financial Well Being Score Mean, median [Range]

  5.8, 5.5 [1-10]

FACT-G Quality of Life Score Mean, median [Range]

  77.0, 79 [23.8-108]

EORTC Quality of Life Score Mean, median [Range]

  61.5, 66.7 [0-100]

Income- Median Zip Mean, median [Range]

  $68,048.9, $60,325 [$21,155-$201,354]

Income- Final Mean, median [Range]

  $90,003.9, $67,282 [$0-$1,600,000]

Federal Poverty Level Mean 25th Percentile 50th Percentile 75th Percentile

  514.6% 276.3% 376.7% 555.5%

Treatment Center University of Chicago Hospitals* NorthShore University Health System

  85.4% (199) 14.6% (34)

Treatment Mode of Administration Oral Drug Only IV Drug (IV Drug Only or in Combination with Oral)

  12.0% (28) 88.0% (205)

CommunicaFon  Preferences:  My  doctor  should  consider  my  out-­‐of-­‐pocket  costs  as  

he/she  makes  medical  decisions  

Knowledge/Experience: My doctor knows how much I am spending on

out-of-pocket medical costs.

Scores Survey Participants (n=233)

COST Score Median [Range]

  23 [0-44]

POMS Median [Range]

  14 [1-50]

FACT-G Quality of Life Score Median [Range]

  79 [23.8-108]

EORTC Quality of Life Score Median [Range]

  66.7 [0-100]

Median Scores

top related