Prescription Errors in a Mental Healthcare Facility: A Content Analysis of Brief Interviews

Background. The role of prescription errors (PE) in mental healthcare environments is not well understood. There are few studies that have examined PE in psychiatric hospitals. Although the occurrence of PE in mental health facilities is less frequent than in general hospitals, the specifics of PE remain unclear.

Objective. The study aims to shed light on both the circumstances of each PE in a regional psychiatric hospital in Germany and the factors that shape the nature of PE.

Methods. This qualitative narrative investigation was part of a large-scale one-year observational study (January 2023 – 2024) conducted at St. Josef psychiatric hospital AMEOS Oberhausen in Germany. The therapeutic protocols of the patients were anonymized, and PE were identified according to a modified EQUIP method. For each PE, a double-blind interview was conducted and recorded. Through the interviews, we extracted the keywords based on their frequency and by using NVivo 13 software we build up a codebook followed by a Delphi consensus to clarify the nature of the PE. The observers also recorded the prescriber's behavioral patterns. An analysis of the thematic patterns, semantic groups, and principal factors was executed.

Results. In total, 6020 therapeutic plans were compiled, and 150 PE were detected, leading to an error rate of 2.5%. The predominant PE were associated with medications for musculoskeletal and joint disorders, followed by gastrointestinal drugs and central nervous system (CNS) medications, including psychotropics. The mean duration of the interviews was 2.6 ± 0.8 minutes. The average age of the prescribers was 35.5 ± 4.9 years, with 52.1% being male. Through the interviews, we extracted the following keywords: minimum - shortest time, time - underpressure, unrecognizable - unidentifiable script, vision loss - lack of lighting, lack of knowledge, stress - work stress. There were few cases reported where prescribers failed to pay attention, experienced communication breakdowns, suffered from lack of sleep, had concentration issues, issued unnecessary prescriptions, responded to patient requests, and made misspellings. The average meeting duration was 20 ± 4.2 minutes. The Delphi consensus indicated that there are four primary semantic groups: features of time, reading - spelling challenges, poor lighting, and communication difficulties. These are identified as key factors contributing to PE in mental health contexts, and a hypothesis regarding the nature of writing PE is discussed.

Conclusions. Although nearly all PE were classified as minor in severity, the qualitative research indicated variability in the responses from prescribers regarding each PE. Nonetheless, there are prominent semantic categories, such as temporal aspects, the act of writing, and environmental conditions, that could clarify the  characteristics of PE. We suggest that the process of documenting PE is a complex undertaking with checkpoints that should be adhered to. More research is needed to clarify the risks and factors associated with the nature of PE, in a way to develop interventions to address them.