Systematic literature review of utility values associated with type 2 diabetes-related complications Beaudet A 1 , Clegg J 1 , Lloyd A 2 Reference Valuation (Tariff if EQ-5D) Statistical methods n Mean age Respondent recruitment Country Bagust et al (5) EQ-5D (3) LMRM 4,641 67 Clinical trial: CODE-2 study BL, IT, ES, NL, UK Clarke et al. (6) EQ-5D (3) Tobit model 3,192 62 Clinical trial: UKPDS study UK Currie et al. (7) EQ-5D (3) Multivariate analysis 1,305 62 Postal survey UK Fenwick et al. (8) EQ-5D (3) Multivariate quantile regression model 577 66 Specialized eye clinics AU Glasziou et al. (9) EQ-5D (3) Mean values 978 67 Clinical trial: ADVANCE study AU Kiberd and Jindal (10)* TTO Mean values 17 NR Health care workers CA Kontodimopoulos et al.(11) EQ-5D (3) LMRM 319 65 Diabetology outpatient department GR Langelaan et al. (12) EQ-5D (26) Mean values 128 42 Rehabilitation centre for visually impaired adults NL Laupacis et al. (13)* TTO Mean values 168 42 Transplant waiting list CA Lee et al. (14) EQ-5D (27) Univariate model 858 58 Outpatient clinic of university hospital KR Lloyd et al. (15) EQ-5D (3) Univariate model 122 62 Five clinical sites UK Marrett et al. (16) EQ-5D (28) Mix linear regression model 1,984 58 An Internet-based survey US Matza et al. (17) EQ-5D (3) Least square means 129 56 Advertisement in newspapers UK O'Reilly DJ et al. (18) EQ-5D (28) OLS mean regression 1,147 64 Community setting CA Quah et al. (19) EQ-5D (3) Multiple regression model 699 63 Polyclinic laboratory SG Redekop et al. (20) EQ-5D (3) OLS linear regression 1,136 65 General practitioners NL Smith et al. (21) EQ-5D (28) LMRM 2,074 66 Diabetes registry population USA Solli et al. (22) EQ-5D (3) LMRM 356 64 Diabetes Association NO Sullivan et al. (23) EQ-5D (28) CLAD model 2,039 45 Nationally representative survey US Vexiau et al. (24) EQ-5D (3) LMRM 400 62 Primary care office FR Wasserfallen et al. (25) EQ-5D (3) Mean values 455 64 19 dialysis centres CH Methods • At the scoping stage, the list of health state utility values to be captured was identified by reviewing the list of microvascular and macrovascular complications associated with T2DM and commonly used within existing models. • The search strings combined 1) T2DM, 2) utility and 3) individual complication search terms. The literature searches were conducted on May 24th, 2012 in Medline & Medline In - Process, Embase, EconLIT and NHS Economic Evaluation Database. Results are presented in Figure 1. • Articles were included if they reported a cohort study performed in adults reporting the effect of T2DM complications on utility values. Exclusion criteria included non - English publications, paediatric population, instruments without conversion to utility measure or the effect of a specific therapies on utility values. • When articles presenting EQ - 5D index data were available for a given complication, only these articles were included in the relevance assessment. • A set of utility values was selected following the relevance and quality assessment. The use of values generated using a multiple regression was preferred over use of the unadjusted data. Given the high number of T2DM complications homogeneity of the estimates was considered important. Figure 1. PRISMA Flow diagram * Except for the complication “renal transplant” for which no EQ - 5D values could be identified. CLAD: Censored Least Absolute Deviations Estimator, LMRM: Linear Multivariate Regression Model, NR: not reported, OLS: Ordinary Least Square PDB54 Objective • Type 2 diabetes mellitus (T2DM) represents a major public health burden. Consequently, several T2DM treatments will be submitted to health technology assessment agencies for reimbursement over the next years and assessed according to their relative cost - effectiveness. • The NICE reference case for the measurement and valuation of health for use in cost - effectiveness models emphases the QALY as the preferred measure of the benefit (1, 4). EQ - 5D has been shown to be valid, reliable and responsive in T2DM (2). Other NICE preferences include: patient reported outcomes, tariffs elicited with the general population, UK setting. • The reference case states that the quality and the relevance of the studies should be assessed following the same principles as that for clinical effectiveness reviews, however there are currently no agreed reporting standards to justify the selection of a set of relevant utility values. • The aim of this systematic literature review was to provide a set of utility values associated with T2DM - related complications in line with NICE reference case to be used in modelling studies. Table 1. Description of included studies ↓ ↓ ↓ ↓ ↓ T2DM without complication (6) 0.785 ( 0.681, 0.889) Excess BMI, per kg/m² above 25 (5) - 0.006 (- 0.008,- 0.004) Minor hypoglycaemia (7) - 0.014 (- 0.031, 0.003) Cataract (14) - 0.016 (- 0.031, - 0.001) Renal transplant (10) 0.762 ( 0.658, 0.866) Macular oedema (8) - 0.040 (- 0.066,- 0.014) Non proliferative background DR (8) - 0.040 (- 0.066,- 0.014) Major hypoglycaemia event (7) - 0.047 (- 0.074, - 0.020) Proteinuria (5) - 0.048 (- 0.091, - 0.005) Myocardial infarction (6) - 0.055 (- 0.067, - 0.042) Peripheral vascular disease (5) - 0.061 (- 0.090,- 0.032) Vision threatening DR (8) - 0.070 (- 0.099,- 0.041) Severe vision loss (6) - 0.074 (- 0.025,- 0.124) Neuropathy (5) - 0.084 (- 0.111, - 0.057) Ischemic heart disease (6) - 0.090 (- 0.126, - 0.054) Heart failure (6) - 0.108 (- 0.169, - 0.048) Haemodialysis (25) - 0.164 (- 0.222,- 0.105) Stroke (6) - 0.164 (- 0.274, - 0.054) Active ulcer (5) - 0.170 (- 0.207, - 0.133) Peritoneal dialysis (25) - 0.204 (- 0.342,- 0.066) Amputation event (6) - 0.280 (- 0.389,- 0.170) 15. Lloyd A. et al. Diabet Med 2008 May;25(5):618-24. 16. Marrett E et al. Diabetes Obes Metab 2009 Dec;11(12):1138-44. 17. Matza LS et al. Curr Med Res Opin 2007 Sep;23(9):2051-62. 18. O'Reilly DJ et al. Quality of Life Research 2011 Aug;20(6):939-43. 19. Quah JHM et al. 2011 Jun 1;40(6):June. 20. Redekop WK et al. Diabetes Care 2002;25(3):458-63. 21. Smith DH et al. Qual Life Res 2008 Dec;17(10):1277-84. 22. Solli O et al. Health Qual Life Outcomes 2010;8(18). 23. Sullivan PW et al. Qual Life Res 2008 Oct;17(8):1063-71. 24. Vexiau P et al. Diabetes Obes Metab 2008 Jun;10 Suppl 1:16-24. 25. Wasserfallen JB et al. Nephrol Dial Transplant 2004 Jun;19(6):1594-9 26. Lamers LM et al. Ned Tijdschr Geneeskd 2005 Jul 9;149(28):1574-8. 27. Jo MW et al. Value Health 2008 Dec;11(7):1186-9. 28. Shaw JW et al. Med Care 2005 Mar;43(3):203-20. References 1. Brazier J, Longworth L. TSD 8 An introduction to the measurement and valuation of health for NICE submissions. Decision Support Unit 2011 2. Janssen MF et al. Diabet Med 2011 Apr;28(4):395-413. 3. Dolan P. Med Care 1997 Nov;35(11):1095-108. 4. National Institute for Health and Clinical Excellence. Briefing paper for methods review workshop on key issues in utility measurement. 2007 5. Bagust A, Beale S. Health Econ 2005 Mar;14(3):217-30. 6. Clarke P et al. Med Decis Making 2002 Jul;22(4):340-9. 7. Currie CJ et al. Curr.Med.Res.Opin. 22[8], 1523-1534. 2006. 8. Fenwick EK et al. Invest Ophthalmol Vis Sci 2012 Feb 13;53(2):677-84. 9. Glasziou P et al. Health Qual Life Outcomes 2007;5:21. 10. Kiberd BA, Jindal KK BMJ 1995 Dec 16;311(7020):1595-9. 11. Kontodimopoulos N et al. Eur J Health Econ 2012 Feb;13(1):111-20. 12. Langelaan M et al. Ophthalmic Epidemiol 2007 May;14(3):119-26 13. Laupacis A et al. Kidney Int 1996 Jul;50(1):235-42. 14. Lee WJ et al. J Korean Med Sci 2012 Mar;27(3):255-60. Conclusions • This study generates utility inputs suitable for use in cost-effectiveness modelling elicited with a robust methodology. To our knowledge, it is the first review to specifically assess the appropriateness of the studies from the perspective of the NICE reference case. • This study presents several limitations. Statistical methods used as well the reporting of variability and statistically significance measures were inconsistent across studies. The populations compared were also heterogenous. • An area that would benefit from further research is the calculation of utility value for patients experiencing several complications, an important consideration for T2DM patients who typically develop several complications over time. • This set of values should improve the robustness of T2DM modelling outcomes in line with NICE requirements. Clarke et al (2002) reports values for a number of complications and therefore, provides consistency across a number of complications. • The major limitation in terms of interpreting the values is that they have been sourced from different studies, using different methodologies and populations. • Future research could focus on eliciting a coherent set of values for T2DM - related complications in line with the NICE reference case and to define the variance around the utility value point estimates. CI: confidence interval, DR: diabetic retinopathy. * When the minor and major hypoglycaemia disutilities were converted to an annual instead of quarterly impact, the values reported were - 0.004 and - 0.012. 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 Utility 15th Annual European International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Congress Berlin, Germany November 3 − 7 2012 Sponsored by IMS Health Results • The articles were systematically assessed for relevance with the NICE reference case. • The quality of the studies is reviewed in Table 1. It was possible to identify relevant values elicited with the EQ-5D for all pre-specified T2DM complications except for renal transplant following diabetic nephropathy. • Figure 2 presents the suggested utility values for T2DM complications with the estimated 95% confidence intervals. Figure 2. Suggested utility set for T2DM modelling (95% confidence interval) 1 IMS Health, Basel, Switzerland. 2 IMS Health, London, United Kingdom. Records identified through database searching (n = 19,195) Additional records identified through other sources (n = 0) Records without duplicates (n = 16,578) Records screened (n = 16,574) Records excluded (n = 16,235) Full-text articles assessed for eligibility (n = 339) Reasons for exclusion: No utility values presented • Utility values associated with a specific intervention • No utility values associated diabetes complications Studies included in qualitative assessment (n = 61) Studies presenting EQ-5D index values included for relevance assessment (n = 21)*
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Systematic literature review of utility values associated with type 2 diabetes-related complicationsBeaudet A 1, Clegg J 1, Lloyd A 2
Reference Valuation (Tariff if EQ-5D)
Statistical methods
n Mean age
Respondent recruitment
Country
Bagust et al (5) EQ-5D (3) LMRM 4,641 67 Clinical trial: CODE-2 study
BL, IT, ES, NL, UK
Clarke et al. (6) EQ-5D (3) Tobit model 3,192 62 Clinical trial: UKPDS study UK
Currie et al. (7) EQ-5D (3) Multivariate analysis 1,305 62 Postal survey UK
Fenwick et al. (8) EQ-5D (3) Multivariate quantile regression model
577 66 Specialized eye clinics
AU
Glasziou et al. (9) EQ-5D (3) Mean values 978 67 Clinical trial: ADVANCE study
AU
Kiberd and Jindal (10)* TTO Mean values 17 NR Health care workers CA
Kontodimopoulos et al.(11)
EQ-5D (3) LMRM 319 65 Diabetology outpatient department
GR
Langelaan et al. (12) EQ-5D (26) Mean values 128 42 Rehabilitation centre for visually impaired adults
NL
Laupacis et al. (13)* TTO Mean values 168 42 Transplant waiting list CA
Lee et al. (14) EQ-5D (27) Univariate model 858 58 Outpatient clinic of university hospital
KR
Lloyd et al. (15) EQ-5D (3) Univariate model 122 62 Five clinical sites UK
Marrett et al. (16) EQ-5D (28) Mix linear regression model
1,984 58 An Internet-based survey
US
Matza et al. (17) EQ-5D (3) Least square means 129 56 Advertisement in newspapers
UK
O'Reilly DJ et al. (18) EQ-5D (28) OLS mean regression 1,147 64 Community setting CA
Quah et al. (19) EQ-5D (3) Multiple regression model
699 63 Polyclinic laboratory SG
Redekop et al. (20) EQ-5D (3) OLS linear regression 1,136 65 General practitioners NL
Smith et al. (21) EQ-5D (28) LMRM 2,074 66 Diabetes registry population USA
Solli et al. (22) EQ-5D (3) LMRM 356 64 Diabetes Association NO
Sullivan et al. (23) EQ-5D (28) CLAD model 2,039 45 Nationally representative survey
US
Vexiau et al. (24) EQ-5D (3) LMRM 400 62 Primary care office FR
Wasserfallen et al. (25) EQ-5D (3) Mean values 455 64 19 dialysis centres CH
Methods •At the scoping stage, the list of health state utility values to be captured was identified
by reviewing the list of microvascular and macrovascular complications associated with T2DM and commonly used within existing models.
• The search strings combined 1) T2DM, 2) utility and 3) individual complication search terms. The literature searches were conducted on May 24th, 2012 in Medline & Medline In - Process, Embase, EconLIT and NHS Economic Evaluation Database. Results are presented in Figure 1.
•Articles were included if they reported a cohort study performed in adults reporting the effect of T2DM complications on utility values. Exclusion criteria included non - English publications, paediatric population, instruments without conversion to utility measure or the effect of a specific therapies on utility values.
•When articles presenting EQ - 5D index data were available for a given complication, only these articles were included in the relevance assessment.
•A set of utility values was selected following the relevance and quality assessment. The use of values generated using a multiple regression was preferred over use of the unadjusted data. Given the high number of T2DM complications homogeneity of the estimates was considered important.
Figure 1. PRISMA Flow diagram
* Except for the complication “renal transplant” for which no EQ - 5D values could be identified.
CLAD: Censored Least Absolute Deviations Estimator, LMRM: Linear Multivariate Regression Model, NR: not reported, OLS: Ordinary Least Square
PDB54
Objective• Type 2 diabetes mellitus (T2DM) represents a major public health burden.
Consequently, several T2DM treatments will be submitted to health technology assessment agencies for reimbursement over the next years and assessed according to their relative cost - effectiveness.
• The NICE reference case for the measurement and valuation of health for use in cost - effectiveness models emphases the QALY as the preferred measure of the benefit (1, 4). EQ - 5D has been shown to be valid, reliable and responsive in T2DM (2). Other NICE preferences include: patient reported outcomes, tariffs elicited with the general population, UK setting.
• The reference case states that the quality and the relevance of the studies should be assessed following the same principles as that for clinical effectiveness reviews, however there are currently no agreed reporting standards to justify the selection of a set of relevant utility values.
• The aim of this systematic literature review was to provide a set of utility values associated with T2DM - related complications in line with NICE reference case to be used in modelling studies.
Table 1. Description of included studies
↓
↓
↓
↓
↓
T2DM without complication (6) 0.785 ( 0.681, 0.889)Excess BMI, per kg/m² above 25 (5) - 0.006 (- 0.008,- 0.004)Minor hypoglycaemia (7) - 0.014 (- 0.031, 0.003) Cataract (14) - 0.016 (- 0.031, - 0.001)Renal transplant (10) 0.762 ( 0.658, 0.866)Macular oedema (8) - 0.040 (- 0.066,- 0.014)Non proliferative background DR (8) - 0.040 (- 0.066,- 0.014)Major hypoglycaemia event (7) - 0.047 (- 0.074, - 0.020)Proteinuria (5) - 0.048 (- 0.091, - 0.005)Myocardial infarction (6) - 0.055 (- 0.067, - 0.042)Peripheral vascular disease (5) - 0.061 (- 0.090,- 0.032)Vision threatening DR (8) - 0.070 (- 0.099,- 0.041)Severe vision loss (6) - 0.074 (- 0.025,- 0.124)Neuropathy (5) - 0.084 (- 0.111, - 0.057)Ischemic heart disease (6) - 0.090 (- 0.126, - 0.054)Heart failure (6) - 0.108 (- 0.169, - 0.048)Haemodialysis (25) - 0.164 (- 0.222,- 0.105)Stroke (6) - 0.164 (- 0.274, - 0.054)Active ulcer (5) - 0.170 (- 0.207, - 0.133)Peritoneal dialysis (25) - 0.204 (- 0.342,- 0.066)Amputation event (6) - 0.280 (- 0.389,- 0.170)
15. Lloyd A. et al. Diabet Med 2008 May;25(5):618-24.16. Marrett E et al. Diabetes Obes Metab 2009 Dec;11(12):1138-44.17. Matza LS et al. Curr Med Res Opin 2007 Sep;23(9):2051-62.18. O'Reilly DJ et al. Quality of Life Research 2011 Aug;20(6):939-43. 19. Quah JHM et al. 2011 Jun 1;40(6):June. 20. Redekop WK et al. Diabetes Care 2002;25(3):458-63.21. Smith DH et al. Qual Life Res 2008 Dec;17(10):1277-84.22. Solli O et al. Health Qual Life Outcomes 2010;8(18).
23. Sullivan PW et al. Qual Life Res 2008 Oct;17(8):1063-71.24. Vexiau P et al. Diabetes Obes Metab 2008 Jun;10 Suppl 1:16-24.25. Wasserfallen JB et al. Nephrol Dial Transplant 2004
Jun;19(6):1594-926. Lamers LM et al. Ned Tijdschr Geneeskd 2005 Jul 9;149(28):1574-8.27. Jo MW et al. Value Health 2008 Dec;11(7):1186-9.28. Shaw JW et al. Med Care 2005 Mar;43(3):203-20.
References1. Brazier J, Longworth L. TSD 8 An introduction to the measurement
and valuation of health for NICE submissions. Decision Support Unit 20112. Janssen MF et al. Diabet Med 2011 Apr;28(4):395-413.3. Dolan P. Med Care 1997 Nov;35(11):1095-108.4. National Institute for Health and Clinical Excellence. Briefing paper for
methods review workshop on key issues in utility measurement. 20075. Bagust A, Beale S. Health Econ 2005 Mar;14(3):217-30. 6. Clarke P et al. Med Decis Making 2002 Jul;22(4):340-9.
7. Currie CJ et al. Curr.Med.Res.Opin. 22[8], 1523-1534. 2006.8. Fenwick EK et al. Invest Ophthalmol Vis Sci 2012 Feb 13;53(2):677-84.9. Glasziou P et al. Health Qual Life Outcomes 2007;5:21.10. Kiberd BA, Jindal KK BMJ 1995 Dec 16;311(7020):1595-9.11. Kontodimopoulos N et al. Eur J Health Econ 2012 Feb;13(1):111-20.12. Langelaan M et al. Ophthalmic Epidemiol 2007 May;14(3):119-2613. Laupacis A et al. Kidney Int 1996 Jul;50(1):235-42.14. Lee WJ et al. J Korean Med Sci 2012 Mar;27(3):255-60.
Conclusions • This study generates utility inputs suitable for use in cost-effectiveness modelling
elicited with a robust methodology. To our knowledge, it is the first review to specifically assess the appropriateness of the studies from the perspective of the NICE reference case.
• This study presents several limitations. Statistical methods used as well the reporting of variability and statistically significance measures were inconsistent across studies. The populations compared were also heterogenous.
•An area that would benefit from further research is the calculation of utility value for patients experiencing several complications, an important consideration for T2DM patients who typically develop several complications over time.
• This set of values should improve the robustness of T2DM modelling outcomes in line with NICE requirements. Clarke et al (2002) reports values for a number of complications and therefore, provides consistency across a number of complications.
• The major limitation in terms of interpreting the values is that they have been sourced from different studies, using different methodologies and populations.
•Future research could focus on eliciting a coherent set of values for T2DM - related complications in line with the NICE reference case and to define the variance around the utility value point estimates.
CI: confidence interval, DR: diabetic retinopathy. * When the minor and major hypoglycaemia disutilities were converted to an annual instead of quarterly impact, the values reported were - 0.004 and - 0.012.
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0Utility
15th Annual European International Society for Pharmacoeconomics and Outcomes Research (ISPOR) Congress Berlin, Germany November 3 − 7 2012 Sponsored by IMS Health
Results • The articles were systematically assessed for relevance with the NICE reference case.• The quality of the studies is reviewed in Table 1. It was possible to identify relevant
values elicited with the EQ-5D for all pre-specified T2DM complications except for renal transplant following diabetic nephropathy.
•Figure 2 presents the suggested utility values for T2DM complications with the estimated 95% confidence intervals.
Figure 2. Suggested utility set for T2DM modelling (95% confidence interval)
1 IMS Health, Basel, Switzerland. 2 IMS Health, London, United Kingdom.
Records identified through database searching (n = 19,195)
Additional records identified through other sources (n = 0)
Records without duplicates (n = 16,578)
Records screened (n = 16,574) Records excluded (n = 16,235)
Full-text articles assessed for eligibility (n = 339) Reasons for exclusion: No utility values presented•Utility values associated with a specific
intervention•No utility values associated diabetes
complications
Studies included in qualitative assessment (n = 61)
Studies presenting EQ-5D index values included for relevance assessment (n = 21)*
Related Documents
![1.6 Business internal: Activity 01 - IB ECONOMICS€¦ · 12. Complete the following table by calculating the marginal utility values. [4 marks] Table 1: Total utility and marginal](https://static.cupdf.com/doc/110x72/611da4f0b6807c6a7c3eef71/16-business-internal-activity-01-ib-economics-12-complete-the-following-table.jpg)