USE OF LAPLACE APPROXIMATIONS TO SIGNIFICANTLY IMPROVE THE EFFICIENCY OF EXPECTED VALUE OF SAMPLE INFORMATION COMPUTATIONS. Purpose: to describe a novel process for transforming the efficiency of partial EVSI computations in health economic decision models. Background: Brennan et al. (1,2) Claxton et al (3) have promoted EVSI • as a measure of the societal value of research designs Bayes • to identify optimal sample sizes for primary data collection. Current Mathematical Formulation and 2 level algorithm: Partial EVSI for Parameters Current methods involve: • a two level Monte Carlo simulation algorithm • a large number of calculations. = the parameters for the model (uncertain currently). d = set of possible decisions or strategies. NB(d, ) = the net benefit for decision d, and parameters d Step 2: Evaluate Expected Value of the Proposed Research (Perform the [loop] using Monte Carlo simulation a large number of times) Expected net benefit | proposed data: = E Xi max E NBd , | X i d Partial Expected Value of the Proposed Sample Information (3) – (1) EVSI E Xi max E NBd , | X i max E NB(d, ) d d ^ 2 Level Algorithm 1000 x 1000 iterations inner expectation of net benefit over | Xi (1,000 outer x 1,000 inner simulations) £1,400 (5) ^ b ^ E NB | X i NB d , i- d , i NB d , i i i NB d , i NB d , i 1 E Xi max NB d , max E NB(d, )(6) d d = 15 minutes £1,000 1st order Laplace Approximation 1000 iterations (2) (3) (4) 2nd order term = the posterior mode of the probability distribution for uncertain parameters + and - are the solutions to a series of non-linear equations incorporating the posterior ^ ^ mode , and J the observed information (J= -l’’( ) ) i.e.the 2nd derivative of the prior likelihood function. J-1 is the posterior variance covariance matrix. α+ and α-are analytic expressions in terms of the prior, the first derivative of the likelihood function and the function v( ) itself. a b c d Parameters e g Uncertainty in Means Patient Level Variability Standard Deviations Standard Deviations T0 T1 (T1 - T0) T0 T1 T0 T1 £ 1,000 £ 1,500 £ 500 £ 1 £ 1 £ 500 £ 500 10% 8% -2% 2% 2% 25% 25% 5.20 6.10 0.90 1.00 1.00 4.00 4.00 £ 400 £ 400 £ £ 200 £ 200 £ 200 £ 200 Cost of drug % admissions Days in Hospital Cost per Day % Responding Utility change if respond Duration of response (years) 70% 0.3000 3.0 80% 0.3000 3.0 10% - 10% 0.1000 0.5 10% 0.0500 1.0 20% 0.2000 1.0 20% 0.2000 2.0 % Side effects 25% 20% -5% Change in utility if side effect -0.10 -0.10 0.00 Duration of side effect (years) 0.50 0.50 Total Cost £ 1,208 £ 1,695 £ 487 Total QALY 0.6175 0.7100 0.0925 Cost per QALY £ 1,956 £ 2,388 £ 5,267 Net Benefit (threshold = £10,000 per QALY) £ 4,967 £ 5,405 £ 438 10% 0.02 0.20 5% 0.02 0.20 20% 0.10 0.80 10% 0.10 0.80 Bayesian Updating: the normal case The 2 level algorithm 0 = prior mean 0 = prior standard deviation I0 = 1/σ02 pop = patient level standard deviation n. = sample size Bayesian Updating: the normal case 1 st order Laplace approximation prior precision = sample mean from N~(μ0sample, σ2pop/n) = 2pop/n = sample variance IS = 1/2X = precision of the sample mean I 0 0 I S X I0 I S = posterior mean /n 2 2 2 / n 0 = posterior standard deviation pop 0 1 2 pop ^ Posterior mode = Posterior mean μ1 2X 1 £400 utility observational study £200 % response trial 0 50 100 150 200 250 Sample Size (n) 1st order Laplace marginally below 2 level algorithm 2 level algorithm using 1000 x 1000 is a slight over-estimate (150,000 simulations) £1,400 £1,200 % response/ utility trial + duration obsn'l study £1,000 % response and utility trial £800 £600 duration of response observational study £400 utility observational study £200 % response trial £0 0 50 100 150 200 250 Sample Size (n) % Difference: (2 level Algorithm) - (1st order Laplace Approximation) duration of utility response % response observational observational % response n trial study study and utility trial 0 0% 0% 0% 0% 10 22% 8% 0% 13% 25 19% 2% 5% 8% 50 11% 1% 8% 5% 100 17% 1% 2% 2% 200 15% 4% 2% 3% % response/ utility trial + duration obsn'l study 0% 5% 4% 4% 3% 3% Conclusions μ0sample = random sample from N~(0, 0) X duration of response observational study = 18 seconds Results are very similar order of magnitude f £600 EVSI :- 1st Order Laplace Approximations Illustrative Model with Normally Distributed Uncertain Parameters • Two treatments T1 and T0 • normally distributed cost and benefit parameters Parameter Mean Values given Existing Evidence % response and utility trial £800 £0 Comparison Illustrative Model % response/ utility trial + duration obsn'l study £1,200 Only One Expectation in the formula is recalculated for each simulated dataset collected Xi (1) ^ b ^ Laplace E v | X v i- i v i i i v i v (5) __________________________________________ i 1 ______ EVSI :- 2 Level Algorithm Results Computation Times We use Laplace approximation to evaluate the EVSI inner expectation , hence ^ Methodology: Laplace Approximation Sweeting and Kharroubi4 have developed a 2nd order approximation to evaluate the an expectation of function v() given available data X. 1st order term d outer expectation of net benefit over Xi Posterior mode This is a 2 level simulation due to 2 expectations (e.g. 1000 x 1000 model simulations) E Xi max E NBd , | X i For EVSI the first term in the formula is 1st order approximate EVSI = Decide on research design i.e. parameters to collect data on (i), sample size, etc. •[Start loop] • Sample the data collection: • a) sample the true underlying value for parameter of interest (i) from its prior uncertainty • b) sample simulated data ( Xi ) given the sampled true underlying value of i •Synthesise existing evidence with simulated data result is a simulated posterior probability distribution for value of parameter of interest. •Evaluate the net benefit for each strategy given the new data and make a ‘revised decision’ (Re-run probabilistic analysis using Monte Carlo simulation on the decision model ) Results ^ Step 1: Define a Data Collection Exercise, Simulate the Variety of Possible Results i = the parameters of interest - we propose to collect data on these -i = the other parameters (those not of interest, i.e. remaining uncertainty) Net benefit of ‘revised decision’ | simulated data[i]: = max E NBd , | X i d •[Loop back] Application of Laplace Approximation to EVSI formula EVSI (£) Step 0: Analysis based on Current information •Set up the decision model •Characterise the uncertainty in each parameter with prior probability distributions •Calculate the baseline decision and its expected net benefit Given current information chose decision giving maximum expected net benefit. Expected net benefit | current information = max E NB(d, ) . EVSI (£) Alan Brennan, Samer Kharroubi University of Sheffield, England. a.brennan@sheffield.ac.uk CHEBS (1). This novel application of Laplace approximations short-cuts the calculation of EVSI (2). A simple illustrative model shows that EVSI calculations using the new approach are in line with those produced by the longer 2 level Monte-Carlo sampling method. (3). Computation time reductions depend on the number of Monte-Carlo samples used to evaluate EVSI, but can be seen to be up to 100 times shorter using the approximation method. (4). Application to more complex models and assessment of the value of the 2 nd order term are needed 1 Brennan, A., Chilcott, J. B, Kharroubi, S, O'Hagan, A. Calculating Expected Value of Perfect Information:Resolution of Conflicting Methods via a Two Level Monte Carlo Approach, presented at the 24th Annual Meeting of SMDM, October 23rd, 2002, Washington. 2002. Submitted - Journal of Medical Decision Making 2 Brennan, A. B., Chilcott, J. B., Kharroubi, S., O'Hagan, A. A Two Level Monte Carlo Approach to Calculation Expected Value of Sample Information: How To Value a Research Design. Presented at the 24th Annual Meeting of SMDM, October 23rd, 2002, Washington. 2002. 3 Claxton, K, Ades, T. Efficient Research Design: An Application of Value of Information Analysis to an Economic Model of Zanamivir. Presented at the 24th Annual Meeting of the Society for Medical Decision Making, October 21st, 2002, Washington. 2002. 4 Sweeting, J, Kharroubi, S. Some New Formulae for Posterior Expectations and Bartlett Corrections. Sociedad de Estadistica e Investigacion Operative Test, (Accepted) 2003 Acknowledgements: Particular thanks to Professor Tony O’Hagan for encouraging our ongoing work.