Skip to main content

Table 6 Phase IV (Step 13–15) Value, model, and evaluate health state utilities

From: A scoping review to create a framework for the steps in developing condition-specific preference-based instruments de novo or from an existing non-preference-based instrument: use of item response theory or Rasch analysis

 

ABC-UI [31, 32]

AQL-5D [33,34,35,36,37,38,39]

IUI [40, 41]

OAB-5D [9, 36, 42,43,44]

EORTC-8D [7, 45,46,47,48]

QLU-C10D [49,50,51,52,53,54,55,56,57,58]

HAQ-PBM, QLQ-PBM, MSIS-PBM [61]

FACT-8D [62, 63]

DEMQOL-U & -Proxy-U [8, 71,72,73,74]

AD-5D [75,76,77]

CARIES-QC-U [112, 113]

DUI [78, 79]

DHP-3D & 5D [80]

HASMID-8 & 10 [81, 82]

DMD-QoL-8D [83]

CP-6D [68]

ECHOHIS-4D [115]

13. Elicit health state utility values

 Whose utilities

  Patients

         

X

 

X

 

X

   

  General public

X

X

X

X

X

X

X

 

X

X

X

 

X

X

X

X

X

  Carers

        

X+++

X

       

 Valuation method

  TTOa

X

X

X

X

X

 

X

 

X

   

X

    

  DCEb or DCE-TTOc

     

X++

 

X

 

X

X

  

X

X

X

X

  VASd

  

X

        

X

     

  EQ-VAS

                 

  BWSe

         

X

X

      

  RSf

                 

  SGg

           

X

     

 Method of selecting health states

  Orthogonal or balanced design

X

X

 

X

X

X+

X

 

X

X

  

X

    

  Single and multi-attribute health states to enable modeling

  

X

        

X

     

  D-efficiency

         

X

X

  

X

 

X

X

  C-efficiency

       

X

      

X

  

  Corner states

  

X

        

X

     

  Rasch vignette

                 

  Own health state

 

X@

               

  Intermediate and anchor states

 

X

              

X

  Naming health state

                 

14. Model utility values

  Individual level data

X

X

 

X

X

X

X

 

X

X

X

 

X

X

X

X

X

  Aggregate (mean) data

X

X

X

X

X

X

 

X

   

X

X

    

 Functional form

  Additive function

 

X

   

X

X

X

     

X

 

X

X

  Multiplicative function

 

X^

X

        

X

     

 Model type

  Conditional logit

     

X+

 

X

 

X

X

  

X

X

X

X

  Mixed logit

     

X+

 

X

       

X

 

  Multinomial logit

         

X

       

  Ordinary least squares

X

X

 

X

X

   

X

   

X

    

  Tobit

                 

  Random effects

      

X

          

  Multiattribute utility function

  

X

        

X

     

 Estimation method

  Maximum likelihood estimation

X

X

  

X

            

  Expected a posteriori

   

X

             

15. Evaluate utility function

  Regression model coefficients*

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

X

  Consistency of coefficients with descriptive system**

X

X

  

X

X

X

X

X

X

 

X

X

X

X

X

X

 Fit statistics

  RMSE§

X

       

X

   

X

    

  MAE§§

X

X

X

X

X

 

X

 

X

   

X

    

  AIC§§§

X

    

X+

 

X

 

X

     

X

X

  BIC§§§§

X

    

X+

 

X

 

X

     

X

X

  R2 / Adjusted R2

 

X

X

X

    

X

 

X

X

 

X^^

X^^

  
 

WAITe [110]

NEWQoL-6D [84,85,86]

CORE-6D [88,89,90]

ReQoL-UI [91,92,93,94]

MSIS-8D & -P [27, 97,98,99,100]

Neuro-QoL derived NQU [101, 102]

P-PBMSI [103,104,105,106]

MF-8D [107]

MHOM RA [117]

Vis-QoL [118,119,120,121,122,123,124]

VFQ-UI [125,126,127,128,129,130]

PBI-WRQL [111]

PB-WRQL [111]

PB-HIV [131]

CFQ-R-8D [69]

MobQoL-7D [96]

13. Elicit health state utility values

 Whose utilities

  Patients

    

X

X

X

 

X

X

X

X

X

X

 

X

  General public

X

X

X

X

X

X

 

X

  

X

   

X

X

  Carers

                

Valuation method

    

  TTOa

X

X

X

X

X

  

X

X

X

X

   

X

 

  DCEb or DCE-TTOc

X

               

  VASd

           

X

   

X

  EQ-VAS

            

X

X

  

  BWSe

  

X

             

  RSf

      

X

  

X

      

  SGg

     

X

X

         

 Method of selecting health states

  Orthogonal or balanced design

 

X

     

X

        

  Single and multi-attribute health states to enable modeling

     

X

X

  

X

      

  D-efficiency

                

  C-efficiency

                

  Corner states

     

X

X

  

X

      

  Rasch vignette

  

X

X

            

  Own health state

            

X

X

 

X

  Intermediate and anchor states

X

              

X

  Naming health state

    

X

           

14. Model utility values

  Individual level data

X

X

 

X

       

X

  

X

X

  Aggregate (mean) data

 

X

X

  

X

X

  

X

      

 Functional form

  Additive function

 

X

X

 

X

X

  

X

 

X

X

  

X

 

  Multiplicative function

  

X

  

X

X

  

X

      

  Logistic function

            

X

X

  

  Conditional logit

                

  Mixed logit

                

  Multinomial logit

                

  Ordinary least squares

 

X

 

X

   

X

  

X

     

  Tobit

              

X

 

  Random effects

    

X

         

X

 

  Multiattribute utility function

         

X

      

 Estimation method

  Maximum likelihood estimation

   

X

   

X

        

  Expected a posteriori

                

15. Evaluate utility function

  Regression model coefficients*

X

X

X

X

X

X

X

X

X

 

X

X

X

X

X

X

  Consistency of coefficients with descriptive system**

 

X

X

X

X

X

X

X

  

X

X

X

 

X

 

 Fit statistics

  RMSE§

 

X

X

X

X

     

X

     

  MAE§§

 

X

 

X

X

  

X

      

X

 

  AIC§§§

   

X

          

X

 

  BIC§§§§

   

X

          

X

 

  R2 / Adjusted R2

  

X

 

X

     

X

X

    
  1.  + EORTC-derived QLQ-C10 utility weights were elicited in different countries (Australia, UK, Canada, France, Germany, Netherlands, US) using similar methods. + + EORTC-QLU C10D DCE-TTO valuation task included 11 attributes – the 10 dimensions of the QLU C10D and 1 attribute of time. + + + Carers valued the DEMQOL Proxy-U
  2. a: time trade off (TTO), b: discrete choice experiment (DCE), c: discrete choice experiment – time trade off (DCE-TTO), d: visual analogue scale (VAS), e: best worst scaling (BWS), f: rating scale (RS), g: standard gamble (SG)
  3. @Respondents valued their own health state if they had asthma
  4. *Regression model coefficient significance. **Consistency of coefficients with descriptive system parameters i.e., worse health should produce lower utilities. §root mean square error (RMSE), §§: mean absolute error (MAE), §§§: Akaike’s information criterion (AIC), §§§§: Bayesian Information Criterion (BIC)
  5. ^Linear parametric models were fitted by Yang, and a multiplicative Bayesian models were fitted by Kharroubi.^^Pseudo R2