A.Landschaft (1,2), G.Hirt (2), A.Ihlenfeld (2), K.Gastmeier (3), S.Wirz (4)

1. Information Services Department, Boston Children’s Hospital, Boston, MA 02115;

2. Copeia GmbH, Bergisch Gladbach, Germany D-51469,

3. Dr. med. Knud Gastmeier,Praxis für Spezielle ambulante Palliativmedizin, Potsdam,Germany D-14482,

4. Department for Anesthesiology, Intensive Medicine, Pain and Palliative Medicine - Center for Pain Medicine; Weaning Center CURA Hospital Bad Honnef/GFO Clinics Bonn, Bad Honnef, Germany, D-53604

Introduction: Since March 2017, the number of German patients prescribed cannabis medical products grew. Aims: To assess German patients' free text narratives about medical cannabis therapy.

Methods: We conducted a nationwide, anonymous survey on medically prescribed cannabis products. Patients answered standardized questions online in a text dialog (“chatbot”) and had the option to write free text. We used Python to process the text using standard natural language processing (NLP) techniques to remove common words. Consequently, we calculated word frequencies in the full cohort and in two sub-cohorts.

Results: We distributed brochures with a hyperlink to the online questionnaire through a network of 50 cannabis pharmacies and patient organizations. Of the 1580 participants who interacted with the questionnaire, 1035 provided consent to use their answers for research purposes. A sub-cohort of 292 participants (193 male, 99 female) wrote free text of more than 32 characters.

Conclusions: Almost one-third of the participants used the optional free text field, which suggests important content to express in their own words. Word frequency analysis indicates focus on the clinical outcomes of the treatment and also a concern about its cost and reimbursement from statutory health insurance. In table 1, four terms refer to cost (‘kostenübernahme’,‘krankenkasse’, ‘kosten’, ‘krankenkassen’), with a higher proportion reported by males for all of them. In female narratives there is a higher rate of clinical terms.

This initial result requires further advanced NLP analysis.

Keywords: cannabis, natural language processing, patients’ survey, Germany.

Acknowledgements: We thank the Verband der Cannabis versorgenden Apotheken (VCA) for helping us to distribute flyers with a link to the online survey to patients of 50 German cannabis pharmacies.

This abstract is listed in the journal: «Medical Cannabis and Cannabinoids» (Karger Publishers, Basel, Switzerland) is the official journal of SSCM Swiss Society of Cannabis in Medicin and the SSCM-IACM IACM - International Association for Cannabinoid Medicines Cannabinoid Conference 2022. Cannabis Medicinalis