While working as the lead data scientist at MetaStar I helped to analyze the impact of a community pharmacy based intervention to reduce the rate of hospital admissions and readmissions. Patients enrolled in the intervention had the community pharamcy deliver medications to the homes of patients and educate them as well. We found that enrolling patients in the program reduced their rate of admissions.
Abstract
Background
Direct pharmacist care has been associated with substantial reduction in hospital admission and readmission rates and other positive outcomes, as compared with the absence of such care.
Objective
To decrease readmissions for community pharmacy patients through a program of improved medication packaging, delivery and patient education.
Design
Comparison of the number of admissions and readmissions for each patient enrolled in the program, comparing the time elapsed since enrollment with the equivalent period prior to enrollment.
Setting
A community pharmacy in Kenosha, Wisconsin.
Patients
Medicare beneficiaries served by the community pharmacy conducting the intervention. This includes 263 patients, 167 of which are Medicare beneficiaries, who have been placed in the intervention group as of June 2016.
Intervention
A voluntary program to package medications according to patient-specific characteristics and physician orders, to deliver medication to patients’ homes, and to educate and follow up with patients regarding problems with adherence.
Measurements
Hospital admissions and readmissions post-enrollment as compared with the equivalent pre-enrollment period.
Results
An analysis that limits the study period to a year centered on the patient’s enrollment date in the PACT intervention found a highly statistically significant (p < 0.01) reduction in admissions. An analysis that included the entire duration of the patient’s enrollment in PACT also found a statistically significant (p < 0.001) reduction in admissions. However, neither analytic technique found a statistically significant reduction in readmissions (p=0.2 and 0.1 respectively).
Limitations
That the study was unable to show a decrease in readmissions to accompany the decrease in admissions may be due to the success of the intervention in decreasing the denominator as well as the numerator of the readmissions measure. In addition, the study has not stratified for changes in the intervention over time, and for differences in patient characteristics or outcomes other than admissions and readmissions.
