Modelling Partially & Completely Missing Preference-Based Outcome Measures (PBOMs) Keith Abrams Department of Health Sciences, University of Leicester,

Slides:



Advertisements
Similar presentations
USE OF EVIDENCE IN DECISION MODELS: An appraisal of health technology assessments in the UK Nicola Cooper Centre for Biostatistics & Genetic Epidemiology,
Advertisements

USE OF EVIDENCE IN DECISION MODELS: An appraisal of health technology assessments in the UK Nicola Cooper Centre for Biostatistics & Genetic Epidemiology,
Estimating the cost-effectiveness of an intervention in a clinical trial when partial cost information is available: A Bayesian approach Nicola Cooper.
Evidence synthesis of competing interventions when there is inconsistency in how effectiveness outcomes are measured across studies Nicola Cooper Centre.
Grandparenting and health in Europe: a longitudinal analysis Di Gessa G, Glaser K and Tinker A Institute of Gerontology, Department of Social Science,
The Incorporation of Meta-Analysis Results into Evidence-Based Decision Modelling Nicola Cooper, Alex Sutton, Keith Abrams, Paul Lambert, David Jones Department.
Exploring uncertainty in cost effectiveness analysis NICE International and HITAP copyright © 2013 Francis Ruiz NICE International (acknowledgements to:
MPS Research UnitCHEBS Workshop - April Anne Whitehead Medical and Pharmaceutical Statistics Research Unit The University of Reading Sample size.
Design & Interpretation of Randomized Trials: A Clinician’s Perspective Francis KL Chan Department of Medicine & Therapeutics CUHK.
“Personality, Socioeconomic Status, and All-Cause Mortality in the United States” - Chapman BP et al. Journal Club 02/24/11.
Estimation and Reporting of Heterogeneity of Treatment Effects in Observational Comparative Effectiveness Research Prepared for: Agency for Healthcare.
Opportunities for Bayesian analysis in evaluation of health-care interventions David Spiegelhalter MRC Biostatistics Unit Cambridge
Use of Bayesian Methods for Markov Modelling in Cost Effectiveness Analysis: An application to taxane use in advanced breast cancer Nicola Cooper, Keith.
The Importance of Decision Analytic Modelling in Evaluating Health Care Interventions Mark Sculpher Professor of Health Economics Centre for Health Economics.
1 A Bayesian Non-Inferiority Approach to Evaluation of Bridging Studies Chin-Fu Hsiao, Jen-Pei Liu Division of Biostatistics and Bioinformatics National.
Modelling Cost Effectiveness of Behaviour Modification Programmes and Effects on Medication Case Study of Education Programmes in Diabetes Professor Alan.
USE OF LAPLACE APPROXIMATIONS TO SIGNIFICANTLY IMPROVE THE EFFICIENCY
Structural uncertainty from an economists’ perspective
The Cost-Effectiveness and Value of Information Associated with Biologic Drugs for the Treatment of Psoriatic Arthritis Y Bravo Vergel, N Hawkins, C Asseburg,
Results Doubling the sample size from n=50 to 100, increases the value of a 6 month study by 64% but the value of a 3 years study by only 34%. Similarly.
1 Equivalence and Bioequivalence: Frequentist and Bayesian views on sample size Mike Campbell ScHARR CHEBS FOCUS fortnight 1/04/03.
Clinical trials methodology group Simon Gates 9 February 2006.
Results 2 (cont’d) c) Long term observational data on the duration of effective response Observational data on n=50 has EVSI = £867 d) Collect data on.
Valuing the SF-6D: a nonparametric approach using individual level preference data Part 1): The SF-6D and its valuation Samer A Kharroubi, Tony O’Hagan,
A Two Level Monte Carlo Approach To Calculating
1 Transforming the efficiency of Partial EVSI computation Alan Brennan Health Economics and Decision Science (HEDS) Samer Kharroubi Centre for Bayesian.
17/12/2002 CHEBS Launch Seminar CHEBS Activities and Plans Tony O’Hagan Director.
Generalised Evidence Synthesis Keith Abrams, Cosetta Minelli, Nicola Cooper & Alex Sutton Medical Statistics Group Department of Health Sciences, University.
Decision Analysis as a Basis for Estimating Cost- Effectiveness: The Experience of the National Institute for Health and Clinical Excellence in the UK.
USE OF EVIDENCE IN DECISION MODELS: An appraisal of health technology assessments in the UK Nicola Cooper Centre for Biostatistics & Genetic Epidemiology,
PEAS wprkshop 2 Non-response and what to do about it Gillian Raab Professor of Applied Statistics Napier University.
DISCUSSION Alex Sutton Centre for Biostatistics & Genetic Epidemiology, University of Leicester.
Copyright restrictions may apply A Randomized Trial of Nebulized 3% Hypertonic Saline With Epinephrine in the Treatment of Acute Bronchiolitis in the Emergency.
MAPPING THE DIABETES HEALTH PROFILE (DHP-18) ONTO THE EQ-5D AND SF-6D GENERIC PREFERENCE BASED MEASURES OF HEALTH Brendan Mulhern 1, Keith Meadows 2, Donna.
Economic evaluation of health programmes Department of Epidemiology, Biostatistics and Occupational Health Class no. 16: Economic Evaluation using Decision.
Advanced Statistics for Interventional Cardiologists.
Modeling Menstrual Cycle Length in Pre- and Peri-Menopausal Women Michael Elliott Xiaobi Huang Sioban Harlow University of Michigan School of Public Health.
Decision Models Based on Individual Patient and Summary Data Mark Sculpher Neil Hawkins Centre for Health Economics, University of York Workshop: Towards.
Real-world effectiveness of nicotine replacement therapy in pregnancy Leonie S. Brose, PhD Andy McEwen, PhD & Robert West, PhD University College London.
Guide to Handling Missing Information Contacting researchers Algebraic recalculations, conversions and approximations Imputation method (substituting missing.
Linkage between SSCAS data and mortality data. Patients’ outcome Determined by: Prior health and personal characteristics Severity of illness Effectiveness.
Journal Club Alcohol, Other Drugs, and Health: Current Evidence May–June 2012.
Combining prevalence estimates from multiple sources Julian Flowers.
Incorporating heterogeneity in meta-analyses: A case study Liz Stojanovski University of Newcastle Presentation at IBS Taupo, New Zealand, 2009.
Department of SOCIAL MEDICINE Producing Small Area Estimates of the Need for Hip and Knee Replacement Surgery ANDY JUDGE Nicky Welton Mary Shaw Yoav Ben-Shlomo.
Data Analysis – Statistical Issues Bernd Genser, PhD Instituto de Saúde Coletiva, Universidade Federal da Bahia, Salvador
Meta-analysis and “statistical aggregation” Dave Thompson Dept. of Biostatistics and Epidemiology College of Public Health, OUHSC Learning to Practice.
1 THE ROLE OF COVARIATES IN CLINICAL TRIALS ANALYSES Ralph B. D’Agostino, Sr., PhD Boston University FDA ODAC March 13, 2006.
Patricia Guyot1,2, Nicky J Welton1, AE Ades1
Lecture 9: Analysis of intervention studies Randomized trial - categorical outcome Measures of risk: –incidence rate of an adverse event (death, etc) It.
Introduction ESTIMATING THE MINIMALLY IMPORTANT DIFFERENCE (MID) OF THE DIABETES HEALTH PROFILE (DHP-18) FOR TYPE 1 AND TYPE 2 DIABETES MELLITIS Brendan.
The Diabetic Retinopathy Clinical Research Network Effect of Diabetes Education During Retinal Ophthalmology Visits on Diabetes Control (Protocol M) 11.
Bayesian Statistics & Innovative Trial Design April 3, 2006 Jane Perlmutter
Simulation Study for Longitudinal Data with Nonignorable Missing Data Rong Liu, PhD Candidate Dr. Ramakrishnan, Advisor Department of Biostatistics Virginia.
Multilevel and multifrailty models. Overview  Multifrailty versus multilevel Only one cluster, two frailties in cluster e.g., prognostic index (PI) analysis,
A shared random effects transition model for longitudinal count data with informative missingness Jinhui Li Joint work with Yingnian Wu, Xiaowei Yang.
Armando Teixeira-Pinto AcademyHealth, Orlando ‘07 Analysis of Non-commensurate Outcomes.
Course: Research in Biomedicine and Health III Seminar 5: Critical assessment of evidence.
Capture-recapture Models for Open Populations “Single-age Models” 6.13 UF-2015.
Is a meta-analysis right for me? Jaime Peters June 2014.
Meta-analysis of observational studies Nicole Vogelzangs Department of Psychiatry & EMGO + institute.
INFERENCE FOR BIG DATA Mike Daniels The University of Texas at Austin Department of Statistics & Data Sciences Department of Integrative Biology.
Prediction and Missing Data. Summarising Distributions ● Models are often large and complex ● Often only interested in some parameters – e.g. not so interested.
Global burden of diseases
Multiple Imputation using SOLAS for Missing Data Analysis
Simon Thompson University of Cambridge
Mark Rothmann U.S. Food and Drug Administration September 14, 2018
20 times 80 is enough Ben van Hout 10/4/19
Nadia Howlader, PhD National Cancer Institute
Yu Du, PhD Research Scientist Eli Lilly and Company
Presentation transcript:

Modelling Partially & Completely Missing Preference-Based Outcome Measures (PBOMs) Keith Abrams Department of Health Sciences, University of Leicester, UK John Brazier, Tony O’Hagan, Samer Kharroubi Centre for Bayesian Statistics in Health Economics (CHEBS), University of Sheffield, UK Aki Tsuchiya Sheffield Health Economics Group (SHEG), University of Sheffield, UK

Outline Missing PBOM Data Modelling Missing PBOM Data Example - IBS Study Other Scenarios & Links with Other Work

Why is data missing? Partially Missing –Missing PBOM on some individuals –Missing dimensions of PBOM Completely Missing Data –Only collected on a sub-sample –Not collected at all

Why is missing data important? Efficiency Precision Missing at Random (MAR) conditional on covariates & other outcomes, e.g. random sub-sample Informative missing data, e.g. death, produce biased results

Solutions to missing data? Collect PBOMs in RCTs evaluating health- care interventions Minimise missing data Complete case analysis – problem if substantial ‘missing data’ Model missing data –Mean value imputation – limited applicability –Multiple Imputation – accounts for uncertainty –Regression imputation – uses relationship between covariates & PBOMs –Bayesian approach …

Multivariate Model - 1 Missing data are treated as parameters & imputed/predicted using the posterior predictive distribution

Multivariate Model - 2 Product Normal Formulation:

Example – IBS Study N=161 patients with Irritable Bowel Syndrome (IBS) PBOMs: SF-6D & EQ-5D NPBOMs: IBS-QoL [34 items] (Overall & 8 sub-scales) Demographic data: age & sex Assessed at Baseline & 3 months follow-up Missing Data –Age & Sex complete –IBS n=13 –EQ-5D n=4 –IBS & EQ-5D n=1

Example – IBS Study Using Multivariate Model for EQ-5D & IBS (+ age & sex) implemented in WinBUGS using ‘vague’ priors Overall (n=161) –EQ-5D: (2.039) & 95% CrI: to Assume 81 patients do not have EQ-5D Complete Case Analysis (n=80) –EQ-5D: (3.117) & 95% CrI: to Predictive Analysis (n=80+81) –EQ-5D: (2.789) & 95% CrI: to 65.88

IBS Study – Extension & Further Work 1 Distributional assumptions (both model & prior distributions), e.g. IBS study EQ-5D: -7.7 to +100 Sensitivity to prior distributions, especially on variance parameters Use of other studies which may have considered both EQ-5D & IBS-QoL – prior distributions, possibly down weighted according to patient population considered Modelling of 8 sub-scales of IBS-QoL & relationship with EQ-5D

IBS Study – Extension & Further Work 2 Additional baseline demographics, e.g. employment Consideration of whether certain individuals (defined by demographics & IBS-QoL scores) are ‘poorly’ predicted – Conditional Predictive Ordinates (CPOs) Assessment of predictive performance – cross-validation

Applications to Other Scenarios Studies which have not used a PBOM at all BUT where there are other studies which have, i.e. ‘Borrowing Strength’ Assumptions – Exchangeable, i.e. the relationship between PBOM & NPBOM is the same across studies ‘Bank of Reference Studies’ for common conditions/diseases, BUT … –Should not be seen as a replacement for well designed studies which use PBOMs –Limited use when there is a treatment-baseline interaction, which might be different for PBOMs & NPBOMs

Links with Other Work Regression-based approaches (‘mapping’) (Tsuchiya et al, 2003) Cross-Calibration (Parmigiani et al, 2003) – categorical data Modelling missing cost data (Lambert et al, 2003) Missing data due to death (informative missing data) –`Quality Adjusted Survival’ techniques (including multi-state modelling) –Joint modelling of both PBOM/NPBOMs & Missing/Death Process (Billingham, 2002)

References Akehurst RL et al. Health-Related Quality of Life and Cost Impact of Irritable Bowel Syndrome in a UK Primary Care Setting. Pharmacoeconomics 2002;20(7): Billingham LJ, Abrams KR. A Bayesian method for synthesizing evidence. The Confidence Profile Method. SMMR 1990;6(1): Lambert PC et al. Glasziou P & Irwig L. An evidence-based approach to individualising treatment. BMJ 1995;311: Prevost TC et al. Hierarchical models in generalised synthesis of evidence: an example based on studies of breast cancer. Stat Med 2000;19: Sutton AJ & Abrams KR. Bayesian methods in meta-analysis and evidence synthesis. SMMR 2001;10(4): Spiegelhalter DJ, Abrams KR, Myles JP. Bayesian Approaches to Clinical Trials & Health-care Evaluation. London: Wiley, 2003.