enhancement of meld

Enhancement of MELD

W. Ray Kim, MDMayo Clinic College of

Medicine Rochester, MN

Plummer Building, Rochester, MN

Agenda

• How to make MELD better

%As an indicator of pretransplant mortality• Add more variable(s): MELDNa• Optimize model: Refit MELD

%As an indicator of posttransplant outcome• Limitations

115 120 125 130 135 140 145

0.5

1.0

5.0

10.0R

isk

of

3-M

onth

M

ort

alit

y

(aft

er

ad

just

ing f

or

MELD

)

Na (mEq/L)

Serum Sodium and Mortality

Kim NEJM 2008;1018

UNOS Registrants, 2005N=6769

MELDNa: Incorporating Na to MELD

6 10 15 20 25 30 35 40

125

130

135

140

MELD

Na (

mEq

/L)

MELDNa = MELD - Na - 0.025*MELD*(140-Na) + 140

Kim NEJM 2008;1018

Validation: MELDNa vs MELD

<20 20-29 30+ <20 20-29 30+ MELD

Ob

serv

ed

/Exp

ecte

d

Based on MELD Based on MELDNa

Kim NEJM 2008;1018

Refitting CoefficientsMethods• Waitlist data obtained from OPTN from 2005-2008:

%Model derivation set (05-06): n=14,214

%Model validation set (07-08): n=13,945 • All adult, primary LTx candidates with end stage

liver disease included (HCC and status 1 excluded) • The proportional hazards regression analysis

predicting mortality within 90 days of listing

%Define lower and upper bounds

%Optimize coefficients

%Bilirubin, Creatinine, INR, and Na

Bilirubin

Bilirubin

Rela

tive R

isk

0.1 0.5 1.0 5.0 50.0

0.0

50.5

05.0

0

Bilirubin and Relative Risk of Death at 90 Days Adjusted for creatinine and INR

05-06 data

0.1

1.0

1

0.0

Bilirubin

Bilirubin

Rela

tive R

isk

0.1 0.5 1.0 5.0 50.0

0.0

50.5

05.0

0

splineMELD


05-06 data

LB=1.0

Current MELD

0.1

1.0

1

0.0

Bilirubin

Bilirubin

Rela

tive R

isk

0.1 0.5 1.0 5.0 50.0

0.0

50.5

05.0

0

splineRefit


05-06 data

LB=1.0

Refit

0.1

1.0

1

0.0

Creatinine

Creatinine

Rela

tive

Ris

k

0.1 0.5 1.0 5.0 10.0

0.5

1.0

5.0

10.

0

Creatinine and Relative Risk of Death at 90 Days Adjusted for bilirubin and INR

05-06 data

Creatinine

Creatinine

Rel

ativ

e R

isk

0.1 0.5 1.0 5.0 10.0

0.5

1.0

5.0

10.0

splineMELD


05-06 data

UB=4.0LB=1.0

Current MELD

Creatinine

Creatinine

Rel

ativ

e R

isk

0.1 0.5 1.0 5.0 10.0

0.5

1.0

5.0

10.0

splineRefit


05-06 data

LB=0.8 UB=3.0

Refit

INR

INR

Rela

tive R

isk

1 5 10

0.0

50.5

05.0

0

INR and Relative Risk of Death at 90 Days Adjusted for bilirubin and creatinine

05-06 data

0.1

1

.0

10

.0

INR

INR

Rel

ativ

e R

isk

1 5 10

0.05

0.50

5.00

splineMELD


05-06 data

LB=1.0

Current MELD

0.1

1

.0

10

.0

INR

INR

Rel

ativ

e R

isk

1 5 10

0.05

0.50

5.00

splineRefit


05-06 data

Refit

UB=3.0LB=1.0

0.1

1

.0

10

.0

Sodium

Na

Rela

tive R

isk

120 130 140 150 160

0.5

1.0

5.0

10.

0

Na and Relative Risk of Death at 90 Days Adjusted for bilirubin, creatinine and INR

05-06 data

Sodium

Na

Rela

tive R

isk

120 130 140 150 160

0.5

1.0

5.0

10.

0

splineRefit

Na and Relative Risk of Death at 90 Days Adjusted for bilirubin, creatinine and INR

05-06 data

Refit

LB=125 UB=140

Validation: Bilirubin

Bilirubin

Rela

tive R

isk

0.05 0.50 5.00 50.00

0.0

10.1

01.0

010.0

0

splineMELDSRTRRefit


07-08 data

*

*Sharma et al. Gastro. 2008;1575

Validation: Creatinine

Creatinine

Rela

tive R

isk

0.1 0.5 1.0 5.0 10.0

0.5

1.0

5.0

10.0

splineMELDSRTRRefit


07-08 data


*

Validation: INR

INR

Rela

tive R

isk

1 5 10

0.1

0.5

5.0

splineMELDSRTRRefit


07-08 data


*1.0

10.0

Validation: MELD

MELD

Rela

tive R

isk

0 20 40 60

0.0

10.1

01.0

010.0

0

Survival at 90 Days

2007-2008 UNOS Data

07-08 data

LB (6) and UB (40) ignored for comparison purpose

Validation: SRTR MELD

SRTR

Rela

tive R

isk

2 4 6 8 10

0.1

1.0

10.0

100.0

Survival at 90 Days

2007-2008 UNOS Data

07-08 data

Validation: Refit MELD(Na)

Refit MELD Na

Rela

tive R

isk

0 2 4 6 8

0.1

1.0

10.0

100.0

Survival at 90 Days

2007-2008 UNOS Data

07-08 data

The Bottom Line

Validation Set

(2007-2008)Model Chi-Square Concordance (SE)

MELD 2263 0.8653 (0.0062)

SRTR MELD 2055 (-8) 0.8581 (0.0063)

MELDNa 2370 (+7) 0.8758 (0.0059)

Refit MELDNa 2374 (+11) 0.8768 (0.0058)

‘Kitchen Sink’ ≠ Crystal Ball‘Unsupervised’ incorporation of all potentially relevant

variables

0.8768 0.8823 0.8879

0.7

0.8

0.9

1

Refit MELDNa + Benefit Variables* Kitchen Sink**

* includes: refit MELDNa, age, BMI, albumin, HCV, diabetes, prior malignancies** includes refit MELDNa, age, sex, race, albumin, HCV, HBV, diabetes, prior malignancies, life support, encephalopathy, ascites, previous abdominal surgery, portal vein thrombosis

C-s

tati

sti

cs

MELD and Resource Utilization

1.5

2.53.0

4.0

6.1

0

1

2

3

4

5

6

7

<10 10-19 20-29 30-39 >40

0

5

10

15

20

25

30

35

<10 10-19 20-29 30-39 >40

FloorICU

pRBC LOS

1518

26

29

25

Kim ATC. 2004

Transplant Benefit

0%

20%

40%

60%

80%

100%

1 2 3 4 5 6 7 8 9 10

Survivalw/o LTx

Survival after LTx

10.3 14.7 16.8 18.5 20.5 22.5 25.1 28.4 32.7 38.7

90 D

ays S

urv

ival

C-statistics 0.57 0.88

Refit MELDNa Deciles

Mean MELD

Conclusions• Ways to improve prediction of waitlist mortality

%Add more variable: MELD Na• Represents a meaningful improvement • Big change in mortality in a subgroup of

patients%Optimize model coeffcients: Refit MELDNa

• New upper and lower bounds:• Creatinine: 0.8-3.0• INR: 1.0-3.0

%Limit to predictability• MELD and post-transplant outcome

%Statistically significant impact on survival and resource utilization

%Limited accuracy

enhancement of meld

Documents