32
rmando Teixeira-Pinto cademyHealth, Orlando ‘07 Analysis of Non-commensurate Outcomes

Analysis of Non-commensurate Outcomes

  • Upload
    seamus

  • View
    41

  • Download
    0

Embed Size (px)

DESCRIPTION

Analysis of Non-commensurate Outcomes. Armando Teixeira-Pinto AcademyHealth, Orlando ‘07. Agenda. Introduction Example: HRQOL after intensive care Common approach to multiple outcomes The latent variable model HRQOL results Discussion and summary. The city of PORTO. The city of PORTO. - PowerPoint PPT Presentation

Citation preview

Page 1: Analysis of  Non-commensurate Outcomes

Armando Teixeira-PintoAcademyHealth, Orlando ‘07

Analysis of Non-commensurate Outcomes

Page 2: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Agenda

IntroductionExample: HRQOL after intensive careCommon approach to multiple outcomesThe latent variable modelHRQOL resultsDiscussion and summary

Page 3: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

The city of PORTO

Page 4: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

The city of PORTO

Page 5: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

The city of PORTO

Page 6: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Introduction

Multiple outcomes are often collected in health studies Longitudinal data, repeated measurements,

multiple informants, multi-dimension outcome (health related quality of life), multiple surrogates for an outcome of interest

Typically these outcomes are correlated. For outcomes measured in the same scale

there are several multivariate methods implemented in commercial software Generalized linear mixed model, GEE, GLM,

MANOVA…

Page 7: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Introduction

Often the outcomes are non-commensurate (mixed type) as for example a binary and a continuous outcome

Common approach: Analyze each outcome separately (univariate

framework) ignoring the correlation A multivariate approach will:

Use the additional information contained in the correlation between outcomes

Permit better control over Type I error rates Answer intrinsically multivariate questions Be helpful in some situations of missing data

Page 8: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Motivation example

Quality of life after Intensive Care Objective: evaluate health related quality of life

(HRQOL) of patients 6 months after ICU discharge.

Study the association with: Age Previous health state

Non-chronic disease Chronic disease with no disability Chronic disease with disability

Apache II score Severity score at ICU admission

Page 9: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Instrument EQ-5D

Measuring HRQOLEQ-5D is a standardized instrument for

use as a measure of health outcome. Applicable to a wide range of health

conditions and treatments, it provides a simple descriptive profile and a single index value for health status based on 5 health related dimensions.

Includes a question about patient’s perception of his/hers HRQOL

Page 10: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Page 11: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Instrument EQ-5D

We’ll consider two outcomesEQ-5D index

Summarizes the 5 dimensions of the EQ5D

Continuous outcomeD-VAS (visual analogue scale)

VAS Dichotomized <=50 and >50Binary outcome

And the three covariates:Age ; Previous health state; Apache

II

Page 12: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Common approach

Data for the HRQOL after ICU stay: 4 years of data collection One intensive care unit from a tertiary hospital in

Portugal 485 patients participated in the study The EQ-5D index was available for all the patients Only 366 patients answered the question

associated with the D-VAS

Common approach: Linear model for the EQ-5D index Logistic or probit regression for D-VAS

Page 13: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Multiple outcomes

EQ-5D indexEQ-5D index

D-VASD-VAS

age

previous health state

Apache II

age

previous health state

Apache II

n=485

n=366

Page 14: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Multiple outcomes

EQ-5D indexEQ-5D index

D-VASD-VAS

age

previous health state

Apache II

age

previous health state

Apache II

n=485

n=366

Page 15: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Instrument EQ-5D

Page 16: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Instrument EQ-5D

Page 17: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Why should we use a multivariate method?

Missing values of D-VAS are associated with lower HRQOL

For a separate model for D-VAS we have missing not a random (MNAR) and the regression estimates might be biased

Because the two outcomes are correlated, in a joint model, we can ‘borrow’ information from the EQ-5d index and reduce the bias for the estimates associated with D-VAS

Page 18: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Multiple outcomes

If the outcomes are of the same type, we could assume a multivariate distribution for the outcomes

For example, two continuous outcomes

2221

2121

2

1 ,

MVN

Page 19: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Binary and continuous outcomes

For mixed type of outcomes there is no obvious multivariate distribution Strategy: Avoid direct specification of the joint

distribution

Latent variable model for yb, yc Introduce a latent variable, u, and assume that

conditional on u the outcomes are independent

f(yb, yc)= f(yb, yc ,u) du =

= f(yb, yc |u) f(u) du

= f(yb |u) f(yc| u) f(u) du

Page 20: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Binary and continuous outcomes

Latent variable model

f(yb |u) f(yc| u) f(u) du

We can specify separate equations for the outcomes conditional on u.

The latent variable is modeling the correlation between the outcomes

Page 21: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Latent model

Mathematically speaking:

b and c are scale factors “adjusting” the latent variable to the different scales of the outcomes

),0(~ ),,0(~

)1(

22ccu

cccTcc

bbTbb

NNu

uXy

uXyPprobit

Page 22: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Latent model

However this models has parameters that are non-identifiable and we have to fix some of them

It can be shown that the correct way to fix some of the parameters is:

),0(~ ),,0(~

)1(

22ccui

cccTcc

bTbb

NNu

uXy

uXyPprobit

Page 23: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Latent model

The interpretation of b ’s referring to the effect of the covariates on the outcome yb is conditional on u, i.e., yb |u

The ‘marginal’ effect can be obtained:

),0(~ ),,0(~

)1(

22ccui

cccTcc

bTbb

NNu

uXy

uXyPprobit

21 u

b

IMPORTANT NOTE: The models are for yb |u and yc |u . I omit the conditional from the equations for simplification.

Page 24: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Latent model

A nice feature of this model is that it can be easily implemented in commercial stats software With SAS, use PROC NLMIXED

),0(~ ),,0(~

)1(

22ccui

cccTcc

bTbb

NNu

uXy

uXyPprobit

The same is true for c ’s, but because of the linear link the interpretation is the same for yc |u and yc

Page 25: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

SAS code to fit the Latent Model

#SAS code to maximize the likelihood resulting from the latent variable model for the HRQOL example;

proc nlmixed data=Icu.Euroqolreduced technique=newrap;#initial values;

parms a1=-0.9 b1=.02 c1=-1 d1=0 a2=104 b2=-.2 c2=-9 d2=-4 sigmau=1 sigma2=15 ; bounds sigma2>0, sigmau>0;

#likelihood;part1=a1 + b1*age + c1*apache +d1*pstate+ u;part2=eq5d - (a2 + b2*age +c2*apache + d2*pstate) - u*sigma2;if missing(dvas) then loglik=-log(sigma2)-.5*1/(sigma2**2)*(part2)**2;else loglik =dvas*log(PROBNORM (part1))+(1-dvas)*log(PROBNORM (-part1))-log(sigma2) -5*1/(sigma2**2)*(part2)**2;

#model (actually you can put any variable other than eq5d with complete observations;model eq5d ~ general(loglik) ;

random u ~ normal(0,sigmau**2) subject=idnumb;

#computes the ‘marginalized’ parameters for the probit model;estimate ‘intercept' a1/sqrt(1+sigmau**2);estimate 'age_marg' b1/sqrt(1+sigmau**2);estimate 'apache_marg' c1/sqrt(1+sigmau**2);estimate ‘pstate_marg’ d1/sqrt(1+sigmau**2);run;

Page 26: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Results of the HRQOL study

Univariate Latent model

Coefficient P-value Coefficient P-value

EQ-5D Index (n=485)

Age -0.24(0.06)

<0.01 -0.24(0.06)

<0.01

Previous state -8.12(1.53)

<0.01 -8.12(1.53)

<0.01

Apache II ~0(0.15)

~1 ~0(0.16)

~1

D-VAS (n=366)

Age -0.01(0.005)

0.01 -0.01(0.005)

0.03

Previous state -0.46(0.11)

<0.01 -0.49(0.11)

<0.01

Apache II -0.018(0.011)

0.09 -0.027(0.010)

<0.01

Page 27: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Results of the HRQOL study

Univariate Latent model

Coefficient P-value Coefficient P-value

EQ-5D Index (n=485)

Age -0.24(0.06)

<0.01 -0.24(0.06)

<0.01

Previous state -8.12(1.53)

<0.01 -8.12(1.53)

<0.01

Apache II ~0(0.15)

~1 ~0(0.16)

~1

D-VAS (n=366)

Age -0.01(0.005)

0.01 -0.01(0.005)

0.03

Previous state -0.46(0.11)

<0.01 -0.49(0.11)

<0.01

Apache II -0.018(0.011)

0.09 -0.027(0.010)

<0.01

Page 28: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Results of the HRQOL study

The analysis suggests that the severity of the episode leading to the ICU admission is associated with the patients perception of his/hers HRQOL but not with the EQ-5D index

This effect would not be noticed with univariate analysis

Taking into account the correlation between the two outcomes (crude = 0.42) helped to reduce the bias of the effects estimates

Page 29: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Other approaches

Other strategies presented in the literature:Factorization method:

f(yb, yc) = f(yb)f(yc| yb) or f(yb, yc) = f(yc)f(yb| yc)

Extension of weighted GEEs to non-commensurate outcomes

Other strategies for the missing data can also be used, e.g., multiple imputation

Page 30: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Extention to more than two outcomes

),0(~

)(

)(

)(

)(

2

33333

22222

11111

u

kkTkkk

T

T

T

Nu

uXyEg

uXyEg

uXyEg

uXyEg

For k outcomes:

Page 31: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

“Take home” message

Complete cases

+

Same covariates for all the outcomes

Univariate approach

Multivariate approach

Complete cases

+

Different covariates for the the outcomes

Univariate approach less efficient (larger std. errors)

Multivariate approach more efficient (smaller std. errors)

Missing data on the outcomes

Univariate approach may lead to biased estimates

Multivariate approach may reduce the bias

Page 32: Analysis of  Non-commensurate Outcomes

A. Teixeira-Pinto

AcademyHealth, Orlando 2007

Thank you for your attention!

Slides available at:

http://users.med.up.pt/tpinto/ahealth.ppt