Polypharmacy in a hospitalized psychiatric population: risk estimation and damage quantification

Figures

• 

  Fig. 1. Linear regression of “Hospital stay length (days)” and “Number of psychodrugs”. The residual distribution is represented by the black histogram. In the superior central white area it is presented the linear regression function, the correlation (r), p value for the significant test of the correlation and the determination coefficient (R2)

• 

  Fig. 2. a. “Hospital stay length (days)” versus “Number of psychodrugs” for subgroups patients: “1 diagnostic” and “> 1 diagnostic”. The regression equations, correlations, p values and determination coefficient are presented in the correspondent colour for each subgroup. b, c, and d. Means and standard errors of hospital stay length, psychotropic drugs, and illicit drugs by subgroups. ns means no statistical significance by test Mann Whitney U

• 

  Fig. 3. Five subgroups of patients with different treatments. Comparison for the 4 groups with Kruskal Wallis test. Post Hoc test by Nemenyi. Each symbol (cross, square, diamond, triangle or circle) represent one patient. Horizontal black tick lines show the mean group. Superior left bar plot shows mean and standard error for hospital stay length (days) by subgroups: Comb versus Mono (total patients of Atypical + Typical); Null hypothesis significant test was Mann Whitney U. *** represents p value < 0.001, ** p < 0.01 y * p < 0.05