KELA AND KANTA RELIABILITY

OBJECTIVE

the project's objective is to asses if kela purchases, kanta prescription and kanta delivery datasets are reliable, for doing so we want to:

1. check quality of datasets
2. understand problems contained in the data
3. join the datasets together

DATA

the data used in the project have been filtered for a specific group of medicines: statins (ATC code starting with C10AA) Moreover, years have been restricted from 2014 to 2020 for quality and necessity reason

SCRIPTS

PRE-PROCESSING

• filter_kela_purchase.R, filter_kanta_prescription.R and filter_kanta_delivery.R are used to filter for statins the original yearly datasets
• data_joiner.R is responsible for joining single datasets from year 2014 to 2020, as output it produce the 3 completely filtered datasets: joined_statins_purchases.rds, joined_statins_prescriptions.rds and joined_statins_deliveries.rds
• data_joiner.R is also responsible for joinig the 3 filtered datasets in a unique one: final_full_df.rds (FULL JOIN) and final_inner_df.rds (INNER JOIN)

ANALYSIS

• Quality assessment analysis is performed with the MAIN_quality_check.R file
• Analysis over the joined data is instead performed with the MAIN_analysis.R
• Analysis over the single filtered datasets (not joined) is performed in order to create:

1. distribution plot of the lag in days between purchases/prescriptions/deliveris, this is created with preprocess_plot_distribution.R+PLOT_distributions.R
2. trajectory plots of a group of patients in each dataset to asses the quality of information. for this plot use the file PLOT_trajectories.R