Skip to main content
Figure 3 | BMC Research Notes

Figure 3

From: Benefits of applying a proxy eligibility period when using electronic health records for outcomes research: a simulation study

Figure 3

Bias in parameter estimates. Box plots of the bias in selected parameter estimates by percentage of individuals with missing data for the restricted analysis and the unrestricted analysis. a Percentage of patients with cardiovascular disease, b percentage of patients experiencing shortness of breath, c mean number of shortness of breath occurrences per patient, d percentage of patients prescribed ICS/LABA, e mean number of ICS/LABA prescriptions per patient, f percentage of patients with an exacerbation, g mean number of exacerbations per patient, h percentage of patients with a hospitalization, and i mean number of hospitalizations per patient.

Back to article page