OBJECTIVES:   Application of a segmented times series model to measure the effect of guidelines on outcomes measures.  METHODS: To isolate the effect of guidelines, we need to control for three different factors: 1) Baseline differences between the two groups, 2) Step-wise differences at the intervention point, and 3) Trend differences after the intervention. The segmented times series model was combined with the propensity score matching technique. The segmented time series model contained two predictor variables: the binary intervention variable and an interval coding for time. The kitchen sink approach was used for propensity score matching and the segmented time series model controlled for the confounding influence of any underlying trend. The final model ensured that any estimated change in the mean level of the series after intervention was not simply due to the series’ trend. RESULTS: Using U.S. claims data, we analyzed the effect of the American Psychiatric Association’s consensus statement on glucose monitoring for patients on atypical antipsychotic drugs. Glucose screening rose 1% per quarter among antipsychotic-treated patients before release of the guidelines, compared to 0.5% per quarter after (P=0.005 for trend).  Monitoring rates were 16.07% before release of the guidelines and 18.76% after (P<0.001). CONCLUSIONS:   The segmented time series model can provide a clear picture about both trend and intervention effect when analyzing the effects of guidelines.