A myriad of landmark breakthroughs from 2025 demonstrate that the clinical impact of modern therapeutics – including artificial intelligence – depends as much on disciplined hospital pharmacy implementation as it does on scientific innovation. Here, Dr João Gonçalves PhD shares the final part of his top five highlights from 2025 and his concluding thoughts on the past 12 months in hospital pharmacy.
The fifth theme of 2025 does not involve a single molecule or platform, but a maturation in how hospital pharmacy considers safety and digital innovation.
Guntschnig et al argued for tackling medication errors through a systems approach that improves patient safety, reinforcing the idea that mistakes emerge from design, interfaces and governance rather than from individual shortcomings.1
In parallel, another study argued that hospital pharmacy should seize opportunities from artificial intelligence (AI) and innovation, explicitly connecting AI to workflow modifications and improvements in medication-use processes.2
These contributions matter because they move the conversation away from vague aspirations and towards the uncomfortable but necessary work of operational engineering encompassing measurement, accountability and governance.
AI and medication safety
In many hospitals, medication safety is still pursued through incremental additions: more training, more checklists, more warnings. This approach often produces diminishing returns and can even increase risk through complexity, alert fatigue and inconsistent adoption.
The safety framework clarifies what makes safety improvements credible: identifying failure modes, designing controls proportionate to risk, reducing variability in high-alert pathways and measuring outcomes in a way that facilitates learning.
This approach naturally aligns with the pharmacist’s role as a system guardian. Pharmacy oversees the entire medication journey, including points where errors are most likely to occur in transitions of care, dose adjustments, off-label protocols and complex preparations.
Responsible AI in hospital pharmacy
In this context, AI should be regarded neither as a threat nor as a marketing label. It is a tool that can reduce risk when used within robust governance structures but can increase risk if deployed without clear accountability.
The discourse implicitly sets a standard for what ‘responsible AI’ in hospital pharmacy should entail: rigorous model validation, explicit performance metrics, integration into workflows in ways that reduce rather than increase cognitive burden and ongoing post-deployment monitoring for drift or bias.
Crucially, AI should be applied where it can deliver measurable benefit, such as triaging medication reconciliation, forecasting demand to mitigate shortages, prioritising verification queues, and performing dose-range checks for high-alert medicines. These are areas where pharmacy already has baseline data and where performance can be objectively measured.
Taken together, this positions hospital pharmacy towards a more engineering-focused role, not only managing medicines but actively designing safer medication systems and overseeing AI interventions as clinical risk tools.
Disciplined implementation as a new competitive edge
Taken individually, the five developments discussed in this series may appear disparate:
- An advanced therapy medicinal product proof-of-concept in autoimmune disease
- Radioligand therapy in oncology
- Incretins in heart failure with preserved ejection fraction
- A novel antimicrobial for gonorrhoea
- Systems-based medication safety with AI.
In reality, they converge on a single operational truth: the modern hospital’s ability to generate patient benefit increasingly relies on how effectively it implements and governs therapy pathways. Scientific progress is now closely linked to process excellence.
For hospital pharmacists, the strategic stance for 2026 is therefore not just to follow the literature – it is to convert the findings into operational capability.
That involves developing governance structures that can handle complex therapies without fragility; creating pathways that reduce variability; establishing metrics that focus on what matters (dose delivery reliability, adverse events, admissions, and total care costs); and embedding digital tools into workflows in a way that reduces risk rather than introducing new ones.
In a health system facing financial and capacity pressures, this is where hospital pharmacy can deliver disproportionate value – not by controlling every decision, but by designing the system to make decisions more consistent, auditable and aligned with evidence.
Author
João Gonçalves PharmD PhD
Faculty of Pharmacy, University of Lisbon and Imed Research Institute for Medicines, Lisbon, Portugal
References
1 Guntschnig S et al. Tackling medication errors: how a systems approach improves patient safety. Eur J Hosp Pharm 2025;0:1:doi:10.1136/ejhpharm-2025-004533.
2 Zavaleta-Monestel E, Martinez Sesmero JM. Artificial intelligence and innovation in hospital pharmacy: embracing opportunities. Eur J Hosp Pharm 2025:doi:10.1136/ejhpharm-2025-004755.
