Polypharmacy—commonly defined as concurrent use of five or more medications—raises the risk of adverse outcomes, drug–drug interactions and inappropriate prescribing in older adults and people with multimorbidity. The planned process of tapering, discontinuing or withdrawing medicines, known as deprescribing, aims to reduce pill burden, lower interaction and adverse event risks, and optimise therapy. Certain drug classes (for example, benzodiazepines, antipsychotics, long-acting sulfonylureas, proton pump inhibitors and others) are commonly considered for deprescribing when risks exceed benefits. The implementation of deprescribing requires healthcare professionals to have appropriate knowledge, attitudes and access to guidelines, but practice and familiarity with deprescribing tools have been inconsistent across settings. This study assessed healthcare professionals’ knowledge, attitude and practice (KAP) towards deprescribing in Gondar City, Ethiopia.
A facility-based cross-sectional study was conducted from 9 May to June 2025 at two public hospitals in Gondar City: University of Gondar Comprehensive Specialized Hospital (UOGCSH) and Ayira General Hospital. The source population included physicians, nurses and pharmacists working at these hospitals. Eligible participants were healthcare professionals directly involved in deprescribing activity, on duty during the study period and willing to participate. Midwives, laboratory personnel, non-healthcare staff and those not directly engaged in deprescribing activities were excluded.
A self-administered structured questionnaire—adapted from prior literature and implemented via Kobo Toolbox—collected data on sociodemographic characteristics and KAP towards deprescribing. Knowledge was measured with 30 items (score range 0–30) and classified using Bloom’s cut-offs into good (>80%), moderate (60–80%) and poor (<60%); for regression, moderate was grouped with good. Attitude was assessed with 14 Likert-scale items (0–4 per item) yielding a maximum score of 56. Practice was measured by reported engagement in deprescribing actions and related behaviours; overall attitude and practice scores were categorised as negative (<50%), moderate (50%–70%) and positive (>70%). The questionnaire underwent a 5% pretest and internal consistency was reported with Cronbach’s alpha of 0.84. Ethical clearance was obtained from the University of Gondar School of Pharmacy Ethical Review Board.
Data were checked for completeness and analysed using SPSS v.25. Descriptive statistics included frequencies, means and standard deviations. Bivariate logistic regression was used to screen candidate variables (p<0.25), which were then entered into multivariable logistic regression. Associations are presented as crude and adjusted odds ratios with 95% confidence intervals; significance was set at p<0.05. Model diagnostics included goodness-of-fit testing, variance inflation factor checks for multicollinearity, and addressing outliers and missing data as described in the source.
A total of 389 healthcare professionals completed the survey (92.2% response rate). The majority were female (54.2%) and the largest age group was 21–28 years (46.0%), followed by 29–35 years (30.1%). Most participants were single (66.1%) and worked at UOGCSH (91.5%). By profession, nurses composed 65.6% of respondents, with senior doctors representing 18.0%. Nearly half (48.3%) had 0–4 years of work experience, and 88.7% graduated from government institutions.
Most respondents (90.5%) had heard about deprescribing of potentially inappropriate medications. Participants commonly perceived deprescribing as appropriate for patients on polypharmacy (71.0%), those on long-term medications for chronic conditions (66.1%) and older adults aged 65 years and above (61.7%). Regarding drug classes suitable for deprescribing, benzodiazepines were most frequently identified (54.2%), followed by antihypertensives (43.4%). Approximately 78.7% reported awareness of existing deprescribing guidelines and 55.5% felt they had sufficient knowledge to perform deprescribing. Overall, 32% of participants were classified as having good knowledge of deprescribing (using the study’s scoring criteria), while 49% demonstrated poor knowledge.
Attitudinal items indicated that a majority supported educational and structural measures to strengthen deprescribing. For example, 65.8% strongly agreed that providing training to healthcare providers is essential, and 58.6% strongly agreed on the importance of establishing standardised multidisciplinary teams. Over half (57.8%) agreed that deprescribing improves patients’ quality of life. Using the study’s categorisation, 51.7% of respondents had a positive attitude towards deprescribing.
Reported clinical behaviours showed that 76.9% of respondents consciously practised deprescribing in their work, and 74.3% most often used a tapering-down approach. Documentation of deprescribing outcomes was reported by 65.0% of participants. Long-acting sulfonylureas were the most frequently deprescribed medication class reported by 53.2% of respondents. Despite this, only 24.4% had received training on deprescribing during their work experience. Overall practice categorisation found that 27.2% of healthcare professionals demonstrated good practice, while over half exhibited lower practice levels.
Multivariable logistic regression identified several factors positively associated with good knowledge: male gender, graduation from a governmental institution, and being a pharmacist. The source reports adjusted odds ratios and 95% confidence intervals for these associations. Specific values and statistical measures are presented in the original article.
Strengths of the study include inclusion of multiple healthcare professional categories involved in medication management and a validated questionnaire with internal consistency assessment. Limitations are the cross-sectional design based on self-reported responses, exclusion of patient-related factors and perspectives, limited prior local studies for comparison, and restriction to a specific region that may limit direct generalisability to other contexts.
The study found limited awareness and practice related to deprescribing among healthcare professionals in Gondar City: only about one-third had good knowledge and good practice, while roughly half had positive attitudes. The authors recommend incorporating deprescribing into undergraduate curricula and providing targeted training for healthcare professionals to improve knowledge and practice. All data relevant to the study were reported in the article or uploaded as supplementary information according to the source.