1 / 45

Methods for Cost Estimation in CEA: the GLM Approach Henry Glick University of Pennsylvania

Methods for Cost Estimation in CEA: the GLM Approach Henry Glick University of Pennsylvania AcademyHealth Issues in Cost-Effectiveness Analysis Washington, DC 06/10/2008. Outline. Policy-relevant parameter for cost-effectiveness Problems posed by nonnormality of cost data

Download Presentation

Methods for Cost Estimation in CEA: the GLM Approach Henry Glick University of Pennsylvania

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Methods for Cost Estimation in CEA: the GLM Approach Henry Glick University of Pennsylvania AcademyHealth Issues in Cost-Effectiveness Analysis Washington, DC 06/10/2008

  2. Outline • Policy-relevant parameter for cost-effectiveness • Problems posed by nonnormality of cost data • Generalized linear models as a response to the problems • Identifying links and families (gets a little technical) • General comments My objective is to provide practical advice for ways implement GLM models. Slides available at: http://www.uphs.upenn.edu/dgimhsr/presentations.htm

  3. Policy Relevant Parameter for CEA • Policy relevant parameter: differences in the arithmetic, or sample, mean • In welfare economics, a project is cost-beneficial if the winners from any policy gain enough to be able to compensate the losers and still be better off themselves • Thus, we need a parameter that allows us to determine how much the losers lose, or cost, and how much the winners win, or benefit • From a budgetary perspective, decision makers can use the arithmetic mean to determine how much they will spend on a program

  4. Policy Relevant Parameter for CEA (2) • In both cases, substitution of some other parameter for the sample mean can be justified only if it provides a better estimate of gains and losses or spending

  5. Are Sample Means Always the Best Estimator? • In simulation, when cost data are sufficiently nonnormal, the relative bias (truth - observed)2 for other parameters such as the median or adjusted geometric mean can sometimes be lower than the relative bias observed for the arithmetic mean • HOWEVER, • Distribution required to be sufficiently nonnormality that ln(cost) is also substantially nonnormally distributed • In actual data, since we never know truth, it is difficult to determine whether other parameters will have lower relative bias than sample mean

  6. The Problem • Common feature of cost data is right-skewness (i.e., long, heavy, right tails)

  7. The Problem (cont.) • Distributions with long, heavy, right tails will have a mean that differs from the median, independent of “outliers” • Cost data also can’t be negative, and can have large fractions of observations with 0 cost • Nonnormality of cost data can pose problems for common parametric tests such as t-test, ANOVA, and OLS regression

  8. Common (Relatively Bad) Responses To Violation Of Normality • Adopt nonparametric tests of other characteristics of the distribution that are not as affected by the nonnormality of the distribution (“biostatistical” approach) • Transform the data so they approximate a normal distribution (“classic econometric” approach)

  9. Recommended Response: Adopt More Flexible Models • Generalized Linear Models (GLM) • Have the advantages of the log models, but • don’t require normality or homoscedasticity • and evaluate a direct of the difference in cost and don’t raise problems related to retransformation from the scale of estimation to the raw scale • To build them, one must identify a "link function" and a "family“ (based on the data)

  10. Stata and SAS Code • STATA code: glm y x, link(linkname) family (familyname) • General SAS code (not appropriate for gamma family / log link): proc genmod; model y=x/ link=linkname dist=familyname; run;

  11. SAS Code for a Gamma Family / Log Link • When running gamma/log models, the general SAS code drops observations with an outcome of 0 • If you want to maintain these observations and are predicting y as a function of x (M Buntin): proc genmod; a = _mean_; b = _resp_; d = b/a + log(a) variance var = a2 deviance dev =d; model y = x / link = log; run;

  12. The Link Function • Link function directly characterizes how the linear combination of the predictors is related to the prediction on the original scale • e.g., predictions from the identity link -- which is used in OLS -- equal:

  13. The Log Link • Log link is most commonly used in literature • When we adopt the log link, we are assuming: ln(E(y/x))=Xβ • GLM with a log link differs from log OLS in part because in log OLS, one is assuming: E(ln(y)/x)=Xβ • ln(E(y/x) ≠ E(ln(y)/x) i.e. log of the mean  mean of the log costs

  14. ln(E(y/x) ≠ E(ln(y)/x) * Difference = 0; † Difference = 0.179047

  15. Comparison of Results of GLM Gamma/Log and log OLS Regression

  16. The Power Link Function • Stata’s power link provides a flexible link function • It allows generation of a wide variety of named and unnamed links, e.g., • power 1 = Identity link; = BiXi • power .5 = Square root link; = (BiXi)2 • power .25: = (BiXi)4 • power 0 = log link; = exp(BiXi) • power -1 = reciprocal link; = 1/(BiXi) • power -2 = inverse quadratic; = 1/(BiXi)0.5

  17. Negative Power Links • Retranslation of negative power links to the raw scale: • When using a negative power link, negative coefficients yield larger estimates on the raw scale; positive coefficients yield smaller estimates

  18. Selecting a Link Function • There is no single test that identifies the appropriate link • Instead can employ multiple tests of fit • :Pregibon link test checks linearity of response on scale of estimation • Modified Hosmer and Lemeshow test checks for systematic bias in fit on raw scale • Pearson’s correlation test checks for systematic bias in fit on raw scale • Ideally, all 3 tests will yield nonsignificant p-values • Others (e.g., Hardin and Hilbe) have proposed use of (larger) log likelihood, (smaller) deviance, AIC and BIC statistics

  19. The Family • Specifies the distribution that reflects the mean-variance relationship • Gaussian: Constant variance • Poisson: Variance is proportional to mean • Gamma: Variance is proportional to square of mean • Inverse Gaussian or Wald: Variance is proportional to cube of mean • Use of the poisson, gamma, and inverse Gausian families relax the assumption of homoscedasticity

  20. Modified Park Test • A “constructive” test that recommends a family given a particular link function • Implemented after GLM regression that uses the particular link • The test predicts the square of the residuals (res2) as a function of the log of the predictions (lnyhat) by use of a GLM with a log link and gamma family to • Stata code glm res2 lnyhat,link(log) family(gamma), robust • If weights or clustering are used in the original GLM, same weights and clustering should be used for modified Park test

  21. Recommended Family, Modified Park Test • Recommended family derived from the coefficient for lnyhat: • If coefficient ~=0, Gaussian • If coefficient ~=1, Poisson • If coefficient ~=2, Gamma • If coefficient ~=3, Inverse Gaussian or Wald • Given the absence of families for negative coefficients: • If coefficient < -0.5, consider subtracting all observations from maximum-valued observation and rerunning analysis

  22. Example, GLM gamma/log glm cost treat dis* bl*,link(log) family(gamma) miscand1.dta

  23. GLM Diagnostics, log/gamma miscand1.dta

  24. Change Family to Poisson and Rerun Model miscand1.dta

  25. GLM Diagnostics, log/poisson miscand1.dta

  26. Passes Tests, But Can We Improve the Link? • Iteratively evaluate power links (in 0.1 intervals) between -2 and 2 • Use the modified Park test to select a family • Evaluate the fit statistics • Don’t show you the results here, but I then fine tune the power link in 0.01 intervals within candidate regions of the power link

  27. Results for Selected Power Links miscand1.dta

  28. Individual Criteria Do Not Uniquely Identify Link, But… miscand1.dta

  29. Can We Develop a Summary Measure? miscand1.dta

  30. LL, AIC, BIC, and Deviance miscand1.dta

  31. Why Not Simply Use AIC and BIC? • In the current example: • AIC, BIC, log likelihood, and deviance all agreed • They yielded an answer that was similar answer to that from the Pearson correlation test, Pregibon link test, and modified Hosmer and Lemeshow tests • Power link 0.3 • AIC, BIC, log likelihood, (and deviance?) already commonly used for decisions about model fit • Why do we need the new tests?

  32. AIC / BIC • There are at least 3 reasons why in the long run log likelihood, AIC, BIC, and deviance are unlikely to be the recommended tests for identifying the appropriate link function • First, when there are a large fraction of observations with zero cost: • The recommendations from log likelihood / AIC agree with each other • The recommendations from BIC / deviance agree with each other • But the log likelihood/AIC recommendations differ from the BIC/deviation recommendations

  33. AIC / BIC (2) • Second, the 4 statistics aren’t stable across families, and the shifts in their magnitude across families do not provide information about which family/link is best • For example, in a dataset where the modified Park test recommends a gamma family for power links < 0.4, but recommends a poisson family for power links > 0.5, the magnitude of the AIC statistic shifts from ~18 for the gamma family to ~454 for the poisson family • Although the smaller AIC values associated with power links < 0.4 suggest that these links have the better fit, the Pearson, Pregibon, and H&M tests all suggest that the power links > 0.5 are actually superior

  34. AIC / BIC (3) • Third, while this instability across families is less of a problem when our statistical packages offer 4 continuous families only, it will eliminate comparability across links when statistical packages begin to offer more flexible GLM power families • i.e., when we don’t have to choose between Gaussian (0) and poisson (1) families, but instead can use a family of 0.7 • In this case, given that each power will be associated with a slightly different family, it will be impossible to compare the resulting AIC/BIC statistics

  35. Extended Estimating Equations • Basu and Rathouz (2005) have proposed use of extended estimating equations (EEE) which estimate the link function and family along with the coefficients and standard errors • Tends to need a large number of observations (thousands not hundreds) to converge • Currently can’t take the results and use them with a simple GLM command (makes bootstrapping of resulting models cumbersome)

  36. GLM Disadvantages • Disadvantages • Can suffer substantial precision losses • If heavy-tailed (log) error term, i.e., log-scale residuals have high kurtosis (>3) • If family is misspecified

  37. Review • The distribution of cost can pose problems for common parametric tests of cost • Responses in the literature that suggest that we should evaluate something other than the difference in the sample mean (or a direct transformation of this difference) – e.g., nonparametric tests of other characteristics of the distribution or transformations of cost – generally create more problems than they solve • Use of more flexible models that evaluate a direct transformation of the difference in cost generally pose fewer problems • Does require we identify functional forms for the relationship between the predictors and the mean and for the variance structure

  38. EXTRA SLIDES

  39. Fine Tuning (1) miscand1.dta

  40. Fine Tuning (2) miscand1.dta

  41. AIC / BIC • (When there are a large fraction of 0s) log likelihood / AIC recommendations differ from BIC / deviance recommendations • 4 statistics aren’t stable across families, and shifts in their magnitude across families do not provide information about which family/link is best

  42. Different Recommendations heroin2.dta

  43. 4 Statistics Aren’t Stable Across Families rchapter5.dta

  44. 4 Statistics Aren’t Stable Across Families rchapter5.dta

  45. Limitations If Power Family Becomes Available • Generally not a big problem when we are limited to the 4 named continuous families • Because, as in the example, within families we can look for the power at which the statistics reach a maximum (ll) or maximum (AIC, BIC, deviance) • When a more flexible GLM family is added to our statistical packages that allows a family of 0.7 or 1.3, rather than forcing us to round to a poisson family (1.0), the change in the scale of the AIC, etc., will make these statistics difficult, if not impossible, to use

More Related