Pharmacological Treatment of Adult Attention-Deficit/Hyperactivity Disorder (ADHD) in a Longitudinal Observational Study: Estimated Treatment Effect Strengthened by Improved Covariate Balance

An improved method for estimation of causal effects from observationaldata is demonstrated. Applications in medicine have been few, and the purposeof the present study is to contribute new clinical insight by means of this newand more sophisticated analysis. Long term effect of medication for adult ADHDpatients is not resolved. A model with causal parameters to represent effect ofmedication was formulated, which accounts for time-varying confounding andselection-bias from loss to follow-up. The popular marginal structural model(MSM) for causal inference, of Robins etal., adjusts for time-varying confounding, but suffers from lack of robustnessfor misspecification in the weights. Recent work by Imai and Ratkovic [1][2] achieves robustness in the MSM, through improved covariate balance(CBMSM). The CBMSM (freely available software) was compared with a standard fitof a MSM and a naive regression model, to give a robust estimate of the truetreatment effect in 250 previously non-medicated adults, treated for one year,in a specialized ADHD outpatient clinic in Norway. Covariate balance wasgreatly improved, resulting in a stronger treatment effect than without thisimprovement. In terms of treatment effect per week, early stages seemed to havethe strongest influence. An estimated average reduction of 4 units on thesymptom scale assessed at 12 weeks, for hypothetical medication in the 9 - 12weeks period compared to no medication in this period, was found. Thetreatment effect persisted throughout the whole year, with an estimated averagereduction of 0.7 units per week on symptoms assessed at one year, forhypothetical medication in the last 13 weeks of the year, compared to nomedication in this period. The present findings support a strong and causaldirect and indirect effect of pharmacological treatment of adults with ADHD onimprovement in symptoms, and with a stronger treatment effect than has beenreported.