Abstract
We develop a new tool for assessing the sensitivity of findings on treatment effectiveness to differential follow-up rates in the two treatment conditions being compared. The method censors the group with the higher response rate to create a synthetic respondent group that is then compared with the observed cases in the other condition to estimate a treatment effect. Censoring is done under various assumptions about the strength of the relationship between follow-up and outcomes to determine how informative differential dropout can alter inferences relative to estimates from models that assume the data are Missing at Random. The method provides an intuitive measure for understanding the strength of the association between outcomes and dropout that would be required to alter inferences about treatment effects. Our approach is motivated by translational research in which treatments found to be effective under experimental conditions are tested in standard treatment settings. In such applications, follow-up rates in the experimental setting are likely to be substantially higher than in the standard setting, especially when observational data are used in the evaluation. We test the method on a case study evaluation of the effectiveness of an evidence-supported adolescent substance abuse treatment program (Motivational Enhancement Therapy/Cognitive Behavioral Therapy-5) delivered by community-based treatment providers relative to its performance in a controlled research trial. In this case study, follow-up rates in the community-based settings were extremely low (54 %) compared to the experimental setting (95 %) giving raise to concerns about non-ignorable drop-out.
Similar content being viewed by others
Abbreviations
- ASMD:
-
Absolute Standardized Mean Difference
- CSAT:
-
Center for Substance Abuse Treatment
- CYT:
-
Cannabis Youth Trial
- EAT:
-
Effective Adolescent Treatment
- EPS:
-
Emotional Problems Scale
- GAIN:
-
Global Appraisal of Individual Needs
- GBM:
-
Generalized Boosted Model
- IAS:
-
Illegal Activities Scale
- ISNI:
-
Index of Sensitivity to NonIgnorability
- MAR:
-
Missing at Random
- MCAR:
-
Missing Completely at Random
- MET/CBT–5:
-
Motivational Enhancement Therapy/Cognitive Behavioral Therapy-5
- SAMHSA:
-
Substance Abuse and Mental Health Services Administration
- SFS:
-
Substance Frequency Scale
- SPS:
-
Substance Problem Scale
- SSI:
-
Standardized Selection Index
References
Alford, G.S., Koehler, R.A., Leonard, J.: Alcoholics anonymous-narcotics anonymous model inpatient treatment of chemically dependent adolescents: a 2-year outcome study. J. Stud. Alcohol 52, 118–126 (1991)
American Society of Addiction Medicine: ASAM Patient Placement Criteria for the Treatment of Substance-Related Disorders, 2nd revised edn. Lippincott Williams & Wilkins, Chevy Chase (2001)
Barkham, M., Stiles, W.B., Connell, J., Twigg, E., Leach, C., Lucock, M., Mellor-Clark, J., Bower, P., King, M., Shapiro, D.A., Hardy, G.E., Greenberg, L., Angus, L.: Effects of psychological therapies in randomized trials and practice-based studies. Br. J. Clin. Psychol. 47, 397–415 (2008)
Bertelsen, M., Jeppesen, P., Petersen, L., Thorup, A., Øhlenschlaeger, J., le Quach, P., Christensen, T.Ø., Krarup, G., Jørgensen, P., Nordentoft, M.: Five-year follow-up of a randomized multicenter trial of intensive early intervention vs standard treatment for patients with a first episode of psychotic illness: the OPUS trial. Arch. Gen. Psychiatry 65(7), 762–771 (2008)
Blanco, C., Olfson, M., Goodwin, R.D., Ogburn, E., Liebowitz, M.R., Nunes, E.V., Hasin, D.S.: Generalizability of clinical trial results for major depression to community samples: results from the National epidemiologic survey on alcohol and related conditions. J. Clin. Psychiatry 69, 1276–1280 (2008a)
Blanco, C., Olfson, M., Okuda, M., Nunes, E.V., Liu, S.M., Hasin, D.S.: Generalizability of clinical trials for alcohol dependence to community samples. Drug Alcohol Depend. 98, 123–128 (2008b)
Brown, S.A., Gleghorn, A., Schuckit, M.A., Myers, M.G., Mott, M.A.: Conduct disorder among adolescent alcohol and drug abusers. J. Stud. Alcohol 57, 314–324 (1996)
Brown, S.A., Tapert, S.F., Tate, S.R., Abrantes, A.M.: The role of alcohol in adolescent relapse and outcome. J. Psychoact. Drugs 32, 107–115 (2000)
Cady, M.E., Winters, K.C., Jordan, D.A., Solberg, K.B., Stinchfield, R.D.: Motivation to change as a predictor of treatment outcome for adolescent substance abusers. J. Child Adolesc. Subst. Abuse 5, 73–91 (1996). doi:10.1300/J029v05n01_04
Christensen, H., Griffiths, K.M., Farrer, L.: Adherence in internet interventions for anxiety and depression. J. Med. Internet Res. 11(2), e13 (2009)
Cochran, W.G.: The effectiveness of adjustment by subclassification in removing bias in observational studies. Biometrics 24(2), 295–313 (1968)
Cohen, J.: A power primer. Psychol. Bull. 112, 155–159 (1992)
Copas, J.B., Eguchi, S.: Local sensitivity approximations for selectivity bias. J. R. Stat. Soc. Ser. B 63, 871–895 (2001)
Copas, J.B., Li, H.G.: Inference for non-random samples (with discussion). J. R. Stat. Soc. Ser. B 59, 55–95 (1997)
Daniels, M.J., Hogan, J.W.: Reparameterizing the pattern mixture model for sensitivity analyses under informative dropout. Biometrics 65, 1241–1248 (2000)
Dennis, M.L.: Global Appraisal of Individual Needs (GAIN): administration guide for the GAIN and related measures, Version: 1299. Chestnut Health Systems, Bloomington (1999)
Dennis, M.L., Godley, S.H., Diamond, G., Tims, F.M., Babor, T., Donaldson, J., Liddle, H., Titus, J.C., Kaminer, Y., Webb, C., Hamilton, N., Funk, R.: The Cannabis Youth Treatment, CYT Study: Main findings from two randomized trials. J. Subst. Abuse Treat. 27(3), 197–213 (2004)
Dennis, M.L., Titus, J.C., Diamond, G., Donaldson, J., Godley, S.H., Tims, F.M., Webb, C., Kaminer, Y., Babor, T., Roebuck, M.C., Godley, M.D., Hamilton, N., Liddle, H., Scott, C.K., C. Y. T. Steering Committee: The Cannabis Youth Treatment (CYT) experiment: rationale, study design and analysis plans. Addiction 97(Suppl 1), 16–34 (2002)
Desmond, D.P., Maddux, J.F., Johnson, T.H., Confer, B.A.: Obtaining follow-up interviews for treatment evaluation. J. Subst. Abuse Treat. 12(2), 95–102 (1995)
Diggle, P., Kenward, M.G.: Informative drop-out in longitudinal data analysis (with discussion). Appl. Stat. 43, 49–73 (1994)
Dodge, R., Sindelar, J., Sinha, R.: The role of depression symptoms in predicting drug abstinence in outpatient substance abuse treatment. J. Subst. Abuse Treat. 28, 189–196 (2005)
Frangakis, C.E., Rubin, D.B.: Principal stratification in causal inference. Biometrics 58, 21–29 (2002)
Friedman, A.S., Glickman, N.W.: Program characteristics for successful treatment of adolescent drug abuse. J. Nerv. Ment. Dis. 174(11), 669–679 (1986)
Friedman, A.S., Glickman, N.W., Morrissey, M.R.: Prediction to successful treatment outcome by client characteristics and retention in treatment in adolescent drug treatment programs: a large-scale cross validation study. J. Drug Educ. 16, 149–165 (1986)
Fuller, W.A., McLoughlin, M.M., Baker, H.D.: Regression weighting in the presence of nonresponse with application to the 1987–1988 Nationwide Food Consumption Survey. Surv. Methodol. 20, 75–85 (1994)
Glynn, R.J., Laird, N., Rubin, D.: Selection modeling vs. mixture modeling with nonignorable response. In: Wainer, H. (ed.) Drawing Inferences from Self-Selected Samples. Springer, New York (1986)
Graubard, B.I., Korn, E.L.: Predictive margins with survey data. Biometrics 55(2), 652–659 (1999)
Greenlees, J.S., Reece, W.S., Zieschang, K.D.: Imputation of missing values when the probability of response depends on the variable being imputed. J. Am. Stat. Assoc. 77(378), 251–261 (1982)
Gruber, V.A., Delucchi, K.L., Kielstein, A., Batki, S.L.: A randomized trial of 6-month methadone maintenance with standard or minimal counseling versus 21-day methadone detoxification. Drug Alcohol Depend. 94(1–3), 199–206 (2008)
Harder, V.S., Stuart, E.A., Anthony, J.C.: Propensity score techniques and the assessment of measured covariate balance to test causal associations in psychological research. Psychol. Methods 15(3), 234–249 (2010)
Hedeker, D., Gibbons, R.D.: Longitudinal Data Analysis. Wiley, Hoboken (2006)
Hedeker, D., Mermelstein, R.J., Demirtas, H.: Analysis of binary outcomes with missing data: missing = smoking, last observation carried forward, and a little multiple imputation. Addiction 102, 1564–1573 (2007)
Hirano, K., Imbens, G.W., Ridder, G.: Efficient estimation of average treatment effects using the estimated propensity score. Econometrica 71, 1161–1189 (2003)
Holland, P.W.: Statistics and causal inference (with discussion). J. Am. Stat. Assoc. 81, 945–960 (1986)
Hunter, S., Ramchand, R., Griffin, B.A., Suttorp, M.J., McCaffrey, D.F., Morral, A.R.: The effectiveness of community-based delivery of an evidence-based treatment for adolescent substance use. J. Subst. Abuse Treat. (in press)
Jenson, J.M., Wells, E.A., Plotnick, R.D., Hawkins, J.D., Catalano, R.F.: The effects of skills and intentions to use drugs on post-treatment drug use of adolescents. Am. J. Drug Alcohol Abuse 19, 1–18 (1993)
Kang, J.D.Y., Schafer, J.L.: Demystifying double robustness: a comparison of alternative strategies for estimating a population mean from incomplete data. Stat. Sci. 22(4), 523–539 (2007)
Kennedy, B.P., Minami, M.: The Beech Hill Hospital/outward bound adolescent chemical dependency treatment program. J. Subst. Abuse Treat. 10, 395–406 (1993)
Lee, D.S.: Training, wages, and sample selection: Estimating sharp bounds on treatment effects. NBER Working Paper 11721. National Bureau of Economic Research (2005)
Lee, D.S.: Training, wages, and sample selection: estimating sharp bounds on treatment effects. Rev. Econ. Stud. 76(3), 1071–1102 (2009)
Little, R.J.A., Rubin, D.B.: Statistical Analysis with Missing Data, 2nd edn. Wiley, New York (2002)
Little, R.J.A., Wang, Y.: Pattern-mixture models for multivariate incomplete data with covariates. Biometrics 52, 98–111 (1996)
MacKinnon, A., Griffiths, K.M., Christensen, H.: Comparative randomised trial of online cognitive-behavioural therapy and an information website for depression: 12-month outcomes. BJP 192, 130–134 (2008). doi:10.1192/bjp.bp.106.032078
McCaffrey, D.F., Ridgeway, G., Morral, A.R.: Propensity score estimation with boosted regression for evaluating causal effects in observational studies. Psychol. Methods 9, 403–425 (2004)
McGuigan, K.A., Ellickson, P.L., Hays, R.D., Bell, R.M.: Adjusting for attrition in school-based samples: bias, precision, and cost trade-offs of three methods. Eval. Rev. 21, 554–567 (1997)
Miller, W.R., Sorensen, J., Selzer, J., Brigham, G.: Disseminating evidence-based practices in substance abuse treatment: a review with suggestions. J. Subst. Abuse Treat. 31(1), 25–39 (2006)
Miller, W.R., Zweben, J., Johnson, W.R.: Evidence-based treatment: why, what, where, when, and how? J. Subst. Abuse Treat. 29, 267–276 (2005)
Morral, A.R., McCaffrey, D.F., Ridgeway, G.: Effectiveness of community-based treatment for substance-abusing adolescents: 12-month outcomes of youths entering Phoenix Academy or alternative probation dispositions. Psychol. Addict. Behav. 18, 257–268 (2004)
Myers, M.G., Brown, S.A., Mott, M.A.: Preadolescent conduct disorder behaviors predict relapse and progression of addiction for adolescent alcohol and drug abusers. Alcohol Clin. Exp. Res. 19, 1528–1536 (1995)
Neugebauer, R., Van der Laan, M.J.: Why prefer double robust estimates? J. Stat. Plan. Inference 129(1–2), 405–426 (2005)
Okuda, M., Hasin, D.S., Olfson, M., Khan, S.S., Nunes, E.V., Montoya, I., Liu, S.M., Grant, B.F., Blanco, C.: Generalizability of clinical trials for cannabis dependence to community samples. Drug Alcohol Depend. 111(1–2), 177–181 (2010)
Ramchand, R., Griffin, B.A., Harris, K.M., Morral, A.R.: Using a cross-study design to assess the efficacy of MET/CBT-5 in treating adolescents with cannabis-related disorders. J. Stud. Alcohol Drugs 72(3), 380–389 (2011)
Richter, S.S., Brown, S.A., Mott, M.A.: The impact of social support and self-esteem on adolescent substance abuse treatment outcome. J. Subst. Abuse 3, 371–385 (1991)
Ridgeway, G., McCaffrey, D., Morral, A., Burgette, L., Griffin, B.A.: Toolkit for weighting and analysis of nonequivalent groups: a tutorial for the twang package. R package (2012)
Robins, J.M., Rotnitzky, A., Zhao, L.P.: Analysis of semiparametric regression models for repeated outcomes in the presence of missing data. J. Am. Stat. Assoc. 90, 106–121 (1995)
Rosenbaum, P.R., Rubin, D.B.: The central role of the propensity score in observational studies for causal effects. Biometrika 70, 41–55 (1983)
Rosenbaum, P.R., Rubin, D.B.: Estimating the effects caused by treatments—comment. J. Am. Stat. Assoc. 79, 26–28 (1984)
Rosenbaum, P.R., Rubin, D.B.: The bias due to incomplete matching. Biometrics 41, 103–116 (1985)
Rubin, D.B.: Estimating causal effects of treatments in randomized and nonrandomized studies. J. Educ. Psychol. 66, 688–701 (1976a)
Rubin, D.B.: Inference and missing data. Biometrika 63, 581–592 (1976b)
Rush, T.V.: Predicting treatment outcome for juvenile and young adult clients in the Pennsylvania substance abuse system. In: Beschner, G.M., Friedman, A.S. (eds.) Youth drug abuse: problems, issues, and treatment, pp. 629–656. Lexington Books, Lexington (1979)
Sampl, S., Kadden, R.: Motivational enhancement therapy and cognitive behavioral therapy for adolescent cannabis users: 5 sessions. In: CYT Cannabis Youth Treatment Series, vol. 1. Center for Substance Abuse Treatment, Substance Abuse and Mental Health Services Administration, Rockville, MD (2001)
Scharfstein, D.O., Rotnitzky, A., Robins, J.M.: Adjusting for nonignorable dropouts using semiparametric nonresponse models. J. Am. Stat. Assoc. 94, 1096–1120 (1999)
Schluchter, M.D.: Methods for the analysis of informatively censored longitudinal data. Stat. Med. 11, 1861–1870 (1992)
Scott, C.K.: A replicable model for achieving over 90% follow-up rates in longitudinal studies of substance abusers. Drug Alcohol Depend. 7, 21–36 (2004)
Shoemaker, R.H., Sherry, P.: Posttreatment factors influencing outcome of adolescent chemical dependency treatment. J. Child Adolesc. Subst. Abuse 2, 89–106 (1991)
Troxel, A.B.: A comparative analysis of quality of life data from a Southwest Oncology Group randomized trial of advanced colorectal cancer. Stat. Med. 17, 767–779 (1998)
Troxel, A.B., Harrington, D.P., Lipsitz, S.R.: Analysis of longitudinal measurements with non-ignorable non-monotone missing values. Appl. Stat. 47, 425–438 (1998)
Troxel, A.B., Ma, G., Heitjan, D.F.: An index of local sensitivity to nonignorability. Stat. Sin. 14, 1221–1237 (2004)
Verbeke, G., Molenberghs, G., Thijs, H., Lesaffre, E., Kenward, M.G.: Sensitivity analysis for nonrandom dropout: a local influence approach. Biometrics 57, 7–14 (2001)
Walton, M.A., Ramanathan, C.S., Reischil, T.M.: Tracking substance abusers in longitudinal research: understanding follow-up contact difficulty. Am. J. Community Psychol. 26, 233–253 (1998)
Weisz, J., Donenberg, G., Han, S., Weiss, B.: Bridging the gap between laboratory and clinic in child and adolescent psychotherapy. J. Consult. Clin. Psychol. 63(5), 688–701 (1995)
Witbrodt, J., Bond, J., Kaskutas, L.A., Weisner, C., Jaeger, G., Pating, D., Moore, C.: Day hospital and residential addiction treatment: randomized and nonrandomized managed care clients. J. Consult. Clin. Psychol. 75(6), 947–959 (2007)
Wooldridge, J.M.: Econometric Analysis of Cross Section and Panel Data, 1st edn. The MIT Press, Cambridge (2001)
Wooldridge, J.M.: Econometric Analysis of Cross Section and Panel Data. MIT Press, Cambridge (2002)
Wu, M.C., Carroll, R.J.: Estimation and comparison of changes in the presence of informative right censoring by modeling the censoring process. Biometrics 44, 175–188 (1988)
Xie, H., Qian, Y.: Measuring the impact of nonignorability in panel data with non-monotone nonresponse. J. Appl. Econ. 27(1), 129–159 (2012)
Acknowledgments
This research was supported by the National Institute on Drug Abuse (5R01DA017507-05; PI: Morral and 1R01DA015697; PI: McCaffrey). The authors would like to thank Andrew Morral for comments provided on earlier drafts of this research.
Author information
Authors and Affiliations
Corresponding author
Appendix: Details on propensity score weighting
Appendix: Details on propensity score weighting
The propensity score weights were estimated using a Generalized Boosted Model (GBM) which is a non-parametric estimation technique that can account for a large number of pre-treatment covariates. GBM adaptively captures the functional form of the relationship between the covariates and treatment selection with less bias than traditional approaches (McCaffrey et al. 2004; Harder et al. 2010; Lee 2009). The “twang” package in R was used to conduct the weight estimation (Ridgeway et al. 2012).
Following the weighting procedure, we calculated the standardized mean difference of each of the pretreatment variables to measure the similarity of the CYT and propensity-score weighted comparison EAT samples. The standardized mean difference is defined using the following formula:
where \( \hat{\mu }_{j} \) denotes the estimated mean value for the treatment and comparison conditions (j = t and c, respectively); \( \hat{\sigma }_{t} \) denotes the estimated standard deviation in the treatment condition (the CYT/experimental youth) for a given pretreatment variable. Values of 0 for a standardized mean difference thus represent no difference in means while values of + or −1 represent one standard deviation difference between the two groups. Absolute standardized differences greater than 0.25 are considered to be ‘moderate effect size differences’ (Cochran 1968). To control for the possible confounding effects of pretreatment group differences that remain after weighting (Neugebauer and Van der Laan 2005), any variables for which the Absolute Standardized Mean Difference (ASMD) was greater than 0.25 were included in our regression models.
We began this analysis with 60 pretreatment variables related to substance abuse patient placement criteria established by the American Society for Addiction Medicine (American Society of Addiction Medicine 2001); these 60 variables have been used in previous investigations examining the effectiveness of adolescent drug treatment (McCaffrey et al. 2004). Many of these characteristics have been shown in prior work to influence substance abuse treatment outcomes, including pretreatment levels of substance use (Alford et al. 1991; Shoemaker and Sherry 1991; Jenson et al. 1993; Kennedy and Minami 1993), symptoms associated with emotional well-being and criminality (Alford et al. 1991; Brown et al. 1996, 2000; Myers et al. 1995), academic/scholastic attendance/performance (Rush 1979; Shoemaker and Sherry 1991), employment (Rush 1979), sociodemographics (Alford et al. 1991; Cady et al. 1996; Friedman and Glickman 1986; Friedman et al. 1986; Shoemaker and Sherry 1991), and social/familial substance use (Richter et al. 1991; Shoemaker and Sherry 1991).
Since controlling for variables unrelated to the outcomes may not improve estimates of treatment effects (Wooldridge 2001), we selected from this initial set of 60 pretreatment covariates only those pretreatment covariates that explained at least one percent of the variance in our outcomes for inclusion our propensity score model. This resulted in a total of 22 pretreatment variables which are listed in Table 1. Missing values on the 22 pretreatment variables were low (mean = 0.60 % and max = 4.87 %) and were controlled for in the propensity score model by balancing the two groups on missing value indicators for each variable in addition to balancing on the variables themselves.
Rights and permissions
About this article
Cite this article
Griffin, B.A., McCaffrey, D.F., Ramchand, R. et al. Assessing the sensitivity of treatment effect estimates to differential follow-up rates: implications for translational research. Health Serv Outcomes Res Method 12, 84–103 (2012). https://doi.org/10.1007/s10742-012-0089-7
Received:
Revised:
Accepted:
Published:
Issue Date:
DOI: https://doi.org/10.1007/s10742-012-0089-7