The primary objective of this study is to analyse medication data from the BioCog Study with the ApoMining-Database and to determine the positive and negative predictive value from the ApoMining-Database for prediction of postoperative delirium (POD).

From 2014-2017 the BioCog Study (ClinicalTrials.gov Identifier: NCT02265263) collected data from 439 perioperative elderly patients in Campus Virchow - Klinikum, Universitätsmedizin Berlin. According to the study protocol, a delirium assessment was performed each day after operation until the 7th postoperative day. At this time, it is known which patients developed postoperative delirium (POD) and which did not. Additionally, from each patient data on long-term medication before operation and applied medication during operation and in the recovery room were collected. In this study, the investiagtors analyse the long-term medication and the perioperative medication from BioCog-Study patients with the ApoMining-Database (http://www.apothesen.de/index.php?id=878; ApoThesenGmbH; Bad Münstereifel; Germany). The ApoMining-Database is a medication database, which analyse tolerability and risks of medications for the elderly. The database generates a hit once a medication reveal a risk for delirium. Additionally, the database can calculate the anticholinergic burden of the medication according to the prescribing information. Whereas the investigators already know, which patient developed a POD, they will determine positive and negative predictive value from the ApoMining-Database for prediction of postoperative delirium.

Inclusion Criteria:

• From BioCog study (NCT02265263)

Exclusion Criteria:

• From BioCog study (NCT02265263)

Additionally for this analysis:

Inclusion criteria:

• Enrollment at Campus Virchow - Klinikum, Charité - Universitätsmedizin Berlin

Exclusion criteria:

• No delirium assessment
• No long-term medication before operation

Elderly patients undergoing elective surgery