Search

Background Half of older people in hospital have a pre-admission medicine prescribed that is potentially inappropriate. Deprescribing research has historically focused on the primary care setting. The aim of this thesis was to develop a practitioner behaviour change intervention for enhancing deprescribing in the hospital setting. Methods Underpinned by behavioural science, the research programme comprised four empirical studies: evaluation of existing hospital deprescribing activity; survey of patients’ and carers’ attitudes towards deprescribing; focus groups with geriatricians and pharmacists to identify key barriers and enablers to address in an intervention; expert panel consensus study to select Behaviour Change techniques (BCTs) for the intervention. Results Deprescribing in hospital occurred for 0.6% of pre-admission medicines, of which 84.1% was reactive in response to harm and 15.9% proactive to prevent harm. Deprescribing in hospital was acceptable to patients and carers: 97.4% and 76.3% respectively were willing to accept a doctor’s deprescribing proposition. Geriatricians and pharmacists described several existing deprescribing enablers in hospital including alignment with their generalist role/knowledge and routine patient monitoring. Key barriers to deprescribing were a misconception of patients’ and carers’ resistance to deprescribing, pharmacists’ perception that deprescribing is riskier than continuing to prescribe, pharmacists’ working patterns limiting capacity to support deprescribing and it being a low hospital priority. Introduction of incentives to deprescribe was an enabler. Six BCTs were selected and characterised to address the key barriers and enabler: social comparison (two distinct characterisations); salience of consequences; pros and cons; restructure the physical environment; action planning. Conclusion There is significant scope to increase deprescribing in hospital and this is acceptable to patients and carers. The behavioural intervention to enhance geriatrician and pharmacist led deprescribing requires modelling to determine the optimal configuration of BCTs. Subsequent testing of the intervention is necessary to determine efficacy at enhancing deprescribing and impact on patient outcomes.

Table S2. Caregiver binary logistic regression model. Full binary logistic regression model data. (DOCX 16 kb)

Table S1. Patient binary logistic regression model. Full binary logistic regression model data. (DOCX 16 kb)

Deprescribing is the planned and supervised process of stopping a medication when that medication may no longer be necessary, is causing more harm than good or does not align with a person's healthcare goals and treatment preferences. There has been an explosion of deprescribing research published over the last 10 years. Refereed/Peer-reviewed