Inappropriate Prescribing in the Elderly

The analysis of inappropriate prescriptions at admission "Day 0" will be evaluated as a single outcome measure by aggregating data using the STOPP/START criteria. Specific assessments will be reported separately, as they involve different types of information and units of measure:

• Prevalence of inappropriate prescriptions: The proportion of patients with prescriptions deviating from STOPP/START criteria (%).
• Characteristics of patient history: Data such as active and passive diseases (count), place of origin (categorical variable), reason for admission (categorical variable), and presence of ADL support (binary variable: Yes/No).
• Clinical indicators: Charlson Comorbidity Index (numerical score), number of medications (count), presence of geriatric syndromes (binary variable: Yes/No).
• Mobility status: Whether the patient is bedridden or ambulatory (categorical variable).
• Hospitalization history: Number of admissions in the past year (count).

This analysis focuses on the prevalence of inappropriate prescriptions at discharge "Day 1", defined according to adherence to or deviation from the recommendations specified in the STOPP/START criteria.

-Prevalence of Inappropriate Prescriptions (%): The proportion of patients with prescriptions deviating from the STOPP/START criteria.

-Destination After Discharge: Categorical variable (e.g., home alone, home with assistance, nursing home, rehabilitation unit).

• Length of Stay (Days)
• Presence of Geriatric Syndromes (delirium, cognitive decline, urinary ) during Hospitalization Unit of Measurement: Binary variable (Yes/No).
• Barthel Index at Discharge Unit of Measurement: Numerical score (in points). Description: The Barthel Index score at the time of discharge. Information for each outcome measure will be obtained from discharge letters and from the physician who cared for the patient for the majority of the hospitalization.

Percentage of cases in which inappropriate prescriptions are identified at the time of admission "Day 0" and discharge "Day 1" for each of the individual STOPP/START criteria, i.e., for each of its recommendations.

The association between the prevalence of inappropriate prescriptions at the time of admission "Day 0" and discharge "Day 1" and clinical variables of interest (such as place of origin, level of care, comorbidities, polypharmacy, and multidimensional geriatric assessment) was analyzed using univariate and multivariate regression. The univariate regression explored the association of each variable with the prevalence of inappropriate prescriptions, while the multivariate regression was used to estimate the simultaneous effect of multiple variables, controlling for potential confounding factors.