Abstract
Prospective observational studies, which provide information on the effectiveness of interventions in natural settings, may complement results from randomised clinical trials in the evaluation of health technologies. However, observational studies are subject to a number of potential methodological weaknesses, mainly selection and observer bias. This paper reviews and applies various methods to control for selection bias in the estimation of treatment effects and proposes novel ways to assess the presence of observer bias. We also address the issues of estimation and inference in a multilevel setting. We describe and compare the use of regression methods, propensity score matching, fixed-effects models incorporating investigator characteristics, and a multilevel, hierarchical model using Bayesian estimation techniques in the control of selection bias. We also propose to assess the existence of observer bias in observational studies by comparing patient- and investigator-reported outcomes. To illustrate these methods, we have used data from the SOHO (Schizophrenia Outpatient Health Outcomes) study, a large, prospective, observational study of health outcomes associated with the treatment of schizophrenia.
The methods used to adjust for differences between treatment groups that could cause selection bias yielded comparable results, reinforcing the validity of the findings. Also, the assessment of observer bias did not show that it existed in the SOHO study. Observational studies, when properly conducted and when using adequate statistical methods, can provide valid information on the evaluation of health technologies.
Similar content being viewed by others
References
Hofer A, Hummer M, Huber R, et al. Selection bias in clinical trials with antipsychotics. J Clin Psychopharmacol 2000; 20: 699–702
Wells KB. Treatment research at the crossroads: the scientific interface of clinical trials and effectiveness research. Am J Psychiatry 1999; 156: 5–10
Rubin DB. Estimating causal effects from large data sets using propensity scores. Ann Intern Med 1997; 127(8 Pt 2): 757–63
Concato J, Shah N, Horwitz RI. Randomised, controlled trials, observational studies, and the hierarchy of research designs. N Engl J Med 2000 Jun 22; 342(25): 1887–92
Benson K, Hartz AJ. A comparison of observational studies and randomised, controlled trials. N Engl J Med 2000 Jun 22; 342(25): 1878–86
Black N. Why we need observational studies to evaluate the effectiveness of healthcare. BMJ 1996 May 11; 312(7040): 1215–8
Black N. What observational studies can offer decision makers. Horm Res. 1999; 51 Suppl. 1: 44–9
Britton A, McPherson K, McKee M, et al. Choosing between randomised and non-randomised studies: a systematic review. Health Technol Assess 1998; 2(13): I–iv, 1-124
Pocock SJ, Elbourne DR. Randomised trials or observational tribulations? N Engl J Med 2000 Jun 22; 342(25): 1907–9
Friedman HS. Observational studies and randomised trials [letter]. N Engl J Med 2000 Oct 19; 343(16): 1195–6
Kunz R, Khan KS, Neumayer HH. Observational studies and randomised trials [letter]. N Engl J Med 2000 Oct 19; 343(16): 1194–5
Sacks HS. Observational studies and randomised trials [letter]. N Engl J Med 2000 Oct 19; 343(16): 1195
Liu PY, Anderson G, Crowley JJ. Observational studies and randomised trials [letter]. N Engl J Med 2000 Oct 19; 343(16): 1195
Smith RP, Meier P. Observational studies and randomised trials [letter]. N Engl J Med 2000 Oct 19; 343(16): 1196
Hlatky MA, Califf RM, Harrell Jr FE, et al. Comparison of predictions based on observational data with the results of randomised controlled clinical trials of coronary artery bypass surgery. J Am Coll Cardiol 1988; 11: 237–45
Kunz R, Oxman AD. The unpredictability paradox: review of empirical comparisons of randomised and non-randomised clinical trials. BMJ 1998; 317: 1185–90
Wolfe RA. Observational studies are just as effective as randomised clinical trials. Blood Purif 2000; 18(4): 323–6
Greene T. Are observational studies ‘just as effective’ as randomised clinical trials? Blood Purif 2000; 18(4): 317–22
Vineis P. Proof in observational medicine. J Epidemiol Commun Health 1997; 51: 9–13
Holmberg L, Baum M, Adami HO. On the scientific inference from clinical trials. J Eval Clin Pract 1999 May; 5(2): 157–62
Wilson GT. The clinical utility of randomised controlled trials. Int J Eat Disord 1998 Jul; 24(1): 13–29
Ellenberg JH. Selection bias in observational and experimental studies. Stat Med 1994 Mar 15-Apr 15; 13(5–7): 557–67
Kemler MA, de Vet HC. Does randomisation introduce bias in unblinded trials? Epidemiology 2000 Mar; 11(2): 228
Rabeneck L, Viscoli CM, Horwitz RI. Problems in the conduct and analysis of randomised clinical trials. Are we getting the right answers to the wrong questions? Arch Intern Med 1992 Mar; 152(3): 507–12
Caplan LR. Is the promise of randomised control trials (“evidence-based medicine”) overstated? Curr Neurol Neurosci Rep 2002; 2: 1–8
Haro JM, Edgell ET, Jones PB, et al. The European Schizophrenia Outpatient Health Outcome (SOHO) study: rationale, methods and recruitment. Acta Psychiatr Scand 2003; 107: 222–32
Haro JM, Edgell ET, Frewer P, et al. The European Schizophrenia Outpatient Health Outcomes (SOHO) study: baseline findings across country and treatment. Acta Psychiatr Scand Suppl 2003; 416: 7–15
Haro JM, Edgell ET, Novick D, et al. Effectiveness of antip-sychotic treatment for schizophrenia: 6-month results of the pan-European Schizophrenia Outpatient Health Outcomes (SOHO) study. Acta Psychiatr Scand 2005; 111: 220–31
Haro JM, Kamath SA, Ochoa S, et al. The clinical global impression-schizophrenia (CGI-SCH) scale: a simple instrument to measure the diversity of symptoms present in schizophrenia. Acta Psychiatr Scand Suppl 2003; 416: 16–23
Guy W. Clinical global impression. In: Guy W, editor. ECDEU assessment manual for psychopharmacology, revised. Rock-ville (MD): National Institute of Mental Health, 1976: 217–22
Williams A. EuroQol: a new facility for the measurement of health-related quality of life. Health Policy 1990; 16: 199–208
Kind P, Hardman G, Macran S. UK population norms for EQ-5D. York Centre for Health Economics Discussion Paper, 172. York: Centre for Health Economics, University of York, 1999
Lambert M, Haro JM, Novick D, et al. Olanzapine vs. other antipsychotics in actual out-patient settings: six months tolera-bility results from the European Schizophrenia Out-patient Health Outcomes study. Acta Psychiatr Scand 2005; 111: 232–43
Rosenbaum PR. Observational studies. 2nd ed. New York: Springer-Verlag, 2002
Starr TB, Dalcorso RD, Levine RJ. Fertility of workers: a comparison of logistic regression and indirect standardization. Am J Epidemiol 1986; 123: 490–8
Altman DG. Comparability of randomised groups. Statistician 1985; 34: 125–36
Rubin DB, Thomas N. Matching using estimated propensity scores: relating theory to practice. Biometrics 1996; 52: 249–64
Cochrane WG. The effectiveness of adjustment by subclassification in removing bias in observational studies. Biometrics 1968; 24: 295–313
Drake C. Effects of misspecifications of the propensity score on estimators of treatment effects. Biometrics 1993; 49: 1231–6
Hylan TR, Crown WH, Meneades L, et al. Tricyclic antidepres-sant and selective serotonin reuptake inhibitors antidepressant selection and health care costs in the naturalistic setting: a multivariate analysis. J Affect Disord 1998; 47: 71–9
Lu M. The productivity of mental health care: an instrumental variable approach. J Ment Health Policy Econ 1999; 2: 59–71
Salkever DS, Slade EP, Karakus M, et al. Estimation of antip-sychotic effects of hospitalisation risk in a naturalistic study with selection of unobservables. J Nerv Ment Dis 2004; 192: 119–28
Hadley J, Polsky D, Mandelblatt JS, et al. An exploratory instrumental variable analysis of the outcomes of localized breast cancer treatments in a Medicare population. Health Econ 2003; 12: 171–86
Bound J, Jaeger DA, Baker RM. Problems with instrumental variables estimation when the correlation between the instruments and the endogenous explanatory variable is weak. J Am Stat Assoc 1995; 90: 443–50
Heckman JJ. Sample specification bias as a specification error. Econometrica 1979; 47: 153–61
Wooldridge JM. Econometric analysis of cross section and panel data. Cambridge (MA): MIT Press, 2002
Lee M-J. Panel data econometrics: method-of-moments and limited dependent variables. San Diego (CA): Academic Press, 2002
Spiegelhalter DJ, Feedman LS, Parmar MKB. Bayesian approaches to randomized trials (with discussion). J R Stat Soc Ser A Stat Soc 1994; 157: 357–416
Jones DA. A Bayesian approach to the economic evaluation of health care technologies. In: Spiker B, editor. Quality of life and pharmaeconomics in clinical trials. 2nd ed. Philadelphia (PA): Lippincott-Raven, 1996: 1189–96
Sculpher MJ, Pang FS, Manca A, et al. Generalisability in economic evaluation studies in health care: a review and case-studies. Health Technol Assess 2004; 8: 1–206
Rice N, Leyland A. Multilevel models: applications to health data. J Health Serv Res Policy 1996; 1: 154–64
Leyland AH, Goldstein H. Multilevel modelling of health statistics. Chichester: Willey, 2001
Goldstein H, Browne W, Rasbash J. Multilevel modelling of medical data. Stat Med 2002; 21: 3291–315
Manca A, Rice N, Sculpher MJ, et al. Assessing generalisability by location in trial-based cost-effectiveness analysis: the use of multilevel models [published erratum appears in Health Econ 2005; 14 (5): 486]. Health Econ 2005; 14: 47–85
Grieve R, Nixon R, Thompson SG, et al. Using multilevel models for assessing the variability of multinational resource use and cost data. Health Econ 2005; 14: 185–96
Chien CF, Steinwachs DM, Lehman A, et al. Provider continuity and outcomes of care for persons with schizophrenia. Ment Health Serv Res 2000; 2: 201–11
Haro JM, Salvador-Carulla L, Cabases J, et al. Utilisation of mental health services and costs of patients with schizophrenia in three areas of Spain. Br J Psychiatry 1998; 173: 334–40
Acknowledgements
Frank Windmeijer received monetary compensation from Eli Lily and Company for econometric advice. Josep Maria Haro received monetary compensation from Eli Lily and Company for his participation in the SOHO Advisory Board. David Suarez is providing statistical consultancy work for Lilly. Stathis Kontodimas and Mark Ratcliffe are Eli Lilly and Company employees.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
About this article
Cite this article
Haro, J.M., Kontodimas, S., Negrin, M.A. et al. Methodological aspects in the assessment of treatment effects in observational health outcomes studies. Appl Health Econ Health Policy 5, 11–25 (2006). https://doi.org/10.2165/00148365-200605010-00003
Published:
Issue Date:
DOI: https://doi.org/10.2165/00148365-200605010-00003