Assessment of Automation Models in Hospital Pharmacy: Systematic Review of Technologies, Practices, and Clinical Impacts

Abstract

Medication management in hospitals is a complex process that encompasses every step from prescription to administration, involving multiple healthcare professionals. This process is prone to various errors that can compromise patient safety and generate significant human and financial costs. Automation in hospital pharmacies represents a major advancement, enhancing patient safety, optimizing professional practices, and reducing hospital expenses. This study aims to analyze the different types of automation systems used in hospital pharmacies, assess the impact of automation, and explore its benefits as well as the challenges and limitations associated with its implementation. A literature search was conducted using the ScienceDirect, PubMed, and Scopus databases, covering the period from 1992 to 2024. A total of 129 relevant articles related to the automation of medication preparation and distribution, as well as its challenges and perspectives were included in this study. Automated technologies significantly contribute to reducing medication errors, strengthening traceability, optimizing inventory management, and alleviating the workload of healthcare professionals. However, challenges persist, particularly in terms of costs, integration with existing processes, and staff training. The use of artificial intelligence offers promising prospects for improving the accuracy and operational efficiency of automation systems.

Introduction

Automation, defined as the use of technologies to perform tasks autonomously while minimizing human intervention, has become established in various sectors to improve the efficiency, accuracy and safety of processes. ¹ In hospitals, automation is increasingly applied to pharmacy practices, where it plays a key role in optimizing medication management and reducing medication errors.^2,3 These errors can occur at different stages of the drug circuit, including during prescription affecting the dosage, form or route of administration,^4-6 as well as during the preparation, dispensing and administration phases.^7-11 Figure 1 illustrates the stages of medication management and potential errors at each step. They are responsible for many adverse events, sometimes serious, which compromise patient safety and increase the workload of healthcare staff as well as hospital expenses.^12-16

Figure 1.

Ishikawa diagram, showing the different types of errors.

Critical steps in the drug circuit have been standardized and secured through the use of technologies such as automated prescribing, ¹⁷ unit dose systems,^18,19 barcode tracking systems,^10,20 and more recently radio frequency identification (RFID). ²¹ These technologies have significantly reduced the risks associated with errors in prescribing, drug selection, and patient identification. Once these steps are secured, automation also extends to the drug preparation and distribution phases. At this level, it takes various forms, including preparation robots, dispensing robots and automated distribution cabinets (ADC). The combination of these technologies has made it possible to reduce the burden of low value-added logistics tasks, ensure operational efficiency, guarantee the traceability of operations and optimize inventory management. ²²

The objective of this literature review is to analyze the impact and interest of automation in hospital pharmacy. It aims to provide an overview of the benefits it brings, while addressing the challenges and limitations associated with its implementation, and identifying avenues for improvement to optimize the safety of hospital pharmaceutical practices.

Method

As an initial step, a preliminary search was conducted on Google and YouTube to identify innovative systems in automation. The information gathered was then validated through a bibliographic search conducted using the ScienceDirect, PubMed and Scopus databases, covering the period from 1992 and 2024. The following keywords were used: Medication Errors, Dispensing Automation, Preparation Automation, Costs, Drug Safety, Electronic Prescription, Machine Learning, and AI. The search was restricted to articles written in French and English. The selection was made on the basis of the titles and abstracts of the articles. The reference lists of relevant articles were also manually reviewed to identify additional studies that could be included. A total of 129 relevant articles were included in this analysis.

History of Automation in Hospital Pharmacy

Over the decades, automation in hospital pharmacies has gradually developed, driven by the need to improve patient safety and optimize the efficiency of medication management. In the late 1960s, the first automation initiatives emerged with the introduction of unit dose dispensing programs, ²³ then in the 1970s, the first computerized management systems were introduced, facilitating inventory management and medication traceability. From the 1990s, technological advances allowed the development of robots specialized in the preparation and distribution of unit doses, these devices have considerably improved accuracy and reduced dispensing errors, while optimizing workflows.^24,25 In October 1992, one of the first automated dispensing cabinets, was installed in the emergency department at the University of California San Francisco Medical Center. A few months later, in April 1993, the technology was deployed in other units, including a post-anesthesia care unit, two 36-bed surgical units, and an 8-bed intensive care unit. ²⁶ The rise of automated dispensing cabinets and barcode tracking systems in the 2000s has enhanced process security, while enabling accurate tracking of inventory and medication administration. More recently, the integration of advanced technologies, such as artificial intelligence and data analytics, has revolutionized hospital pharmacy automation, providing more robust tools to meet the increasing demands for safety, efficiency, and management in modern hospital environments, while paving the way for increased personalization of care. ²⁷

Analysis of Different Automation Technologies

Hospital pharmacy automation systems fall into 2 broad categories, Figure 2 summarizes the different systems that will be developed

Figure 2.

Classification of different automated systems in hospital pharmacy.

• 1. Centralized systems: Include robots designed for the preparation and dispensing of medications from a central pharmacy. These systems use a combination of technologies such as barcodes, visual identification (cameras), RFID, and other advanced methods. Dispensing robots are used to dispense medications in single doses, they are able to handle individually packaged solid medications, such as tablets and capsules. The latter are stored on internal rods, which the robot uses to pick up the doses required for a patient, before placing them in a dedicated envelope for distribution. ²⁸ There is also non-unit medication dispensing systems. ²⁹

Recently, centralized pharmacy automation systems have expanded to include the use of robots dedicated to sterile drug preparation, particularly for intravenous anticancer treatments.^23,30,31 These include robotic devices for the preparation of intravenous (IV) medications placed in clean rooms, ³² automated systems for total parenteral nutrition, ³³ and for systems for the preparation non sterile product.

Intravenous Workflow Management Systems (IVWMS) ensure the safe preparation of intravenous medications using technologies such as barcode scanning, standardized processes, and volumetric or gravimetric verification. These systems enhance safety in sterile compounding. ³⁴

Pneumatic Tube Systems (PTSs) are another commonly used technology to rapidly transport drugs, samples, or other medical supplies between the central pharmacy and hospital wards. These systems operate using compressed air or vacuum to move cylindrical containers through a network of tubes. ³⁵ PTS significantly reduces drug transport time, reduces staff travel, and preserves the properties of drugs, even sensitive ones, during transport. This technology not only improves operational efficiency, but also improves responsiveness in emergencies, especially for critical drugs.^36,37 However, this system has some drawbacks such as high cost and difficulty of installation, impact on certain drugs (proteins).^38,39

• 2. Decentralized systems: are implemented near the point of care, such as in nursing units or hospital wards, they include satellite pharmacies and ADCs, ⁴⁰ the latter, located near the care units, allow rapid and secure access to medications. Automated medicine cabinets, also called decentralized automatons or computerized secure cabinets, represent an evolution of traditional cabinets in care units. They are controlled by a central unit and access is secured by an identifier, biometric recognition (fingerprints) or a password. The caregiver selects the patient, and the search for medications is done either manually (by trade name or International Nonproprietary Name) or via computerized prescription. Access to certain non-prescription products can be restricted, but ADCs can be programmed to allow access without specific authorization when appropriate. The medications are stored in secure drawers, the opening of which is computerized. Each drawer contains a single specialty and can be controlled by different systems (flaps, diodes, intrusion detection).

According to a 2005 survey in the United States, approximately 75% of hospitals used a centralized system for drug distribution, while 25% had adopted decentralized systems. By 2011, the trend toward decentralization had accelerated, with 40% of hospitals having adopted a decentralized model and 58% planning to do so. Among decentralized equipment, 89% of hospitals used ADCs, while 11% used centralized dispensing robots, 18% used carousels, which are drug storage and retrieval systems used to automate drug dispensing in the central pharmacy, and 34% incorporated a machine-readable coding system to verify doses before dispensing.^41,42

The automation of the bedside medication administration process is advancing rapidly, with an increasing integration of robotic technologies. Lim et al. ⁴³ describe the implementation of a robotic nursing assistant (RAN) in a hospital ward, designed to support vital sign monitoring and medication administration. Additionally Nanobots, which are small electronic devices, operating autonomously and remotely controlled to perform certain specific functions including drug administration or diagnosis. ⁴⁴

Smart pumps represent an advancement in the parenteral administration of medications in hospital settings. They are equipped with dose management software and a database that establishes allowable concentrations for each medication, with minimum and maximum limits, as well as alerts in case of exceeding these limits. ⁴⁵

Impact of Automation on the Patient Care

Reduction of Medication Errors (Patient Safety)

The use of robots has shown a significant reduction in dispensing errors. ⁴² A study conducted in a tertiary hospital in Brazil showed that the implementation of automated dispensing robots within the central pharmacy reduced the frequency of errors in the prescription (from 26% in 2013 to 15% in 2017) and distribution phases (from 36% in 2013 to 33% in 2017). ²⁹

ADCs provide highly secure access through biometric technologies, including user codes or fingerprint recognition. These systems limit access to medications to authorized personnel only, after verification of their identity and precise identification of the patient via automated medical prescription. This device makes it possible in particular to restrict access to certain sensitive medications. The installation of ADCs in hospital departments has proven effective in reducing medication errors, such as omission of doses or errors in the timing of administration. ⁴⁶ Several studies have shown these benefits: Fanning et al. ⁴⁷ reported a 64.7% reduction in medication selection and preparation errors in an emergency department. Similarly, Tu et al. ⁴⁸ found that in intensive care, prescribing errors decreased from 3.03 to 1.75 per 100,000 prescriptions, while dispensing errors decreased to zero per 100,000 dispensations. Chapuis et al., ⁴⁹ found a decrease in the overall error rate in an intervention unit from 20.4% before the implementation of ADCs to 13.5% after. Another study found a 53% reduction in administration errors, with a 79.1% decrease in dosing errors and a 93.7% decrease in selection errors following the combined use of a unit-dose dispensing robot and an ADC. ⁵⁰ Dib et al. ⁵¹ found a 27% reduction in medication errors after the introduction of s in 5 care units. Martin et al. ⁵² evaluated an ADC system in 4 departments and observed a decrease in missed doses from 29% to 24%. Cottney et al., ⁵³ reported a reduction in administration errors, from 8.9% to 7.2%, however, this decrease, primarily due to a reduction in low-severity errors, was not clinically significant. These results highlight the positive impact of automation on patient safety, by significantly reducing errors in the different stages of the medication circuit. ⁵⁴

Reduction in Drug Administration Times

Automation has significantly improved medication administration time, including reducing delays and missed doses. This is primarily due to the proximity of medication cabinets, which facilitate faster retrieval and administration of needed medications.^55,56 Martin et al. ⁵² reported a 324% increase in the availability of pharmaceuticals in care areas, with over 95% of needed medications directly accessible at the ward level, and this improvement was accompanied by an 88% reduction in prescription processing time. In addition, the secure storage provided by these cabinets has made it possible to include specially managed medicines (eg, table products) previously subject to strict centralized storage protocols.

Ray et al. ⁵⁷ found that the average waiting time for the first dose of a new medication decreased from 45 minutes with the traditional system to just 1 minute with ADCs. Similarly, Douglas et al. ⁵⁶ showed that the installation of ADCs reduced medication administration time by 40%, with time savings observed in several departments, such as oncology and orthopedics. Shirley et al. ⁵⁸ also noted an increase in the proportion of medications administered on time, from 59% to 77% after the introduction of automated systems.

In critical departments, rapid administration of medications is crucial, especially when life-threatening conditions are at stake, for example, early antibiotic therapy is associated with improved outcomes in conditions such as sepsis, ⁵⁹ meningitis, ⁶⁰ and open fractures. ⁶¹ Lo et al. ⁶² observed that the time to first dose of piperacillin/tazobactam was significantly reduced in intensive care units and medical wards. Hitti et al. ⁶³ confirmed that stocking broad-spectrum antibiotics in emergency department s significantly decreased the time to antibiotic administration in patients with severe sepsis, in line with the Surviving Care recommendations Sepsis Campaign, which advocates administration within 3 hours of hospital admission. ⁴⁷

Finally, the optimization of preparation processes through the use of robots has allowed a significant reduction in the time needed to prepare and deliver drugs, thus contributing to more efficient management of the drug circuit. Table 1 summarizes the main studies on the impact of automation on medication administration time.

Table 1.

Results of Studies on the Impact of Automation on Administration Times.

Type of automated system studied	Year place	Type of study	Sample size	Criteria studied	Results	Author and reference
ADC decentralized	2014, 377-bed teaching hospital	Retrospective analysis	Data from 121 patients	Total time from prescription to administration	1.7-h reduction in mean from prescription to administration time	Lo et al 62
ADC decentralized	2012	Observational retrospective	110 patients	Average antibiotic prescription time administration	Time from prescription to administration reduced from 55 to 26 min. Average time from arrival to antibiotic intake was reduced by 70 min	Hitti et al. 63
ADC decentralized	2017, university hospital	Retrospective/prospective study		Impact on administration time	Reduction of: 40% in administration time	Douglas et al. 56
ADC decentralized	2014, East London NHS Foundation Trust (ELFT)	Prospective study		Evaluate the impact of automation on reducing administration time	Reduction in administration time from 2.94 to 2.37 min per dose	Cottny et al. 53
ADC decentralized	1997, four departments within a major academic hospital	Prospective study		Prescription processing time	88% decrease compared to pharmacy distribution	Martin et al. 52
ADC decentralized	Mercy hospital in Scranton, Pennsylvania, a 270-bed tertiary care	Observational befor, after		To examine the effects of an automated dispensing system on medication administration	The mean time difference between actual and scheduled drug administration decreased from 129.84to 101 min	Shirely et al. 58

Impact of Automation on Healthcare Professionals

Impact on Security (Health and Legal)

Automated systems minimize the need for frequent drug handling, thereby reducing the risk of exposure of healthcare professionals to potentially toxic substances, particularly during the preparation and distribution of cytotoxic or narcotic drugs.^64-67 The Failure Modes Effects and Criticality Analysis (FMECA) method was used to assess the risks of exposure of personnel in a central unit for the preparation of cytotoxic drugs, the result showing that the risk of exposure of personnel to cytotoxic drugs remains significant. ⁶⁸ The introduction of a preparation robot within this unit has considerably improved operator safety. This technology reduced musculoskeletal disorders by 89%, decreased errors in the selection of drugs and diluents by 60%, and improved traceability to 97%. ⁶⁹ In addition, automated systems record each step of drug management, from prescription to administration. This comprehensive documentation strengthens the legal protection of professionals in the event of disputes or errors, by offering impeccable traceability.

Impact on Drug Recovery Time

Medication retrieval time, defined as the interval between staff entering the medication room and picking up medications, plays a crucial role in nursing workflow. A rapid retrieval process limits interruptions and improves the responsiveness of professionals to patient needs, while a prolonged time can cause delays and increase their workload. According to Roman et al., ⁷⁰ the retrieval of controlled medications, typically stored in a locked safe, was significantly faster with ADCs, saving 36.1 seconds. Another study conducted in North America confirmed these results, reporting a decrease in the average retrieval time from 107 to 48 seconds. ²⁶ These time savings are particularly critical in emergency situations, where rapid access to specific, often restricted medications is a key performance indicator (KPI). ⁷¹

Impact on Staff Work Efficiency and Time Management

Drug dispensing in hospitals is a complex process, involving multiple steps and human resources. In manual dispensing systems, pharmacists verify and validate prescriptions, technicians prepare orders and manage inventory, and nurses, collect medications for administration. Automated systems have transformed this workflow. Technicians now focus on dispensing, while pharmacists spend more time verifying prescriptions, delegating repetitive tasks. This transition has significantly transformed the time allocation and roles of professionals involved in medication management.^72-74

The consensus of studies indicates that automation reduces the time pharmacists spend on manual tasks, while increasing the workload of technicians due to the increased complexity of inventory management.^52,75-77 Nurses, on the other hand, spend less time on medication-related tasks, allowing them to devote more time to direct patient care. This reduction in workload is explained by fewer stockouts and fewer trips to the pharmacy for refills.^26,49,76-79 However, some studies, such as that of Noparatayaporn et al., ⁷² report that pharmacists have seen their responsibilities increase, particularly with regard to monitoring and checking of automated systems. Table 2 presents studies highlighting the impact of automation on healthcare professionals’ time management.

Table 2.

Studies Showing the Impact of Automatons on Professionals’ Time.

Type of automated system studied	Type of study	Year and place	Nurses’ time	Pharmacists’ time	Technicians’ time	Authors
ADC decentralized	Descriptive and comparative study	1995, University Medical Center	Inventory time reduced from 5.04 to 0.36/hr/week	Reduction to 0.5% full-time equivalent total	Elimination of billing time	Schwarz and Brodowy 26
ADC decentralized	Financial analysis	1992, 1000-bed referral hospital	Reduction of time spent on drugs from 10.2% to 5.6%. Reduced time spent on records/documentation from 28.0% to 16.9%. Time spent on patients increased from 20.0% to 28.6%		Time spent on stock increased from 7.17 to 48.96 min	Lee et al. 76
ADC decentralized	Comparative study	1996, 400-bed university hospital	Decrease in percentage of drug-related activity from 20.7% to 18.4%	Medicine unit: clinical activity increased from 36.5% to 49%. Surgical intensive care unit: clinical activities increased from 27.97% to 35%		Guerrero et al. 77
ADC decentralized		1995, University of California, San Diego Medical Center		Reduction in dispensing time from 25% to 5% 3 years after installation of ADMs		Ray et al. 57
ADC decentralized	Comparative study	2017, University hospital, 2100 beds Bangkok, Thailand		Addition of ADM selection and verification functions. Increase from 46.84 to 117.61 FTE	Reduction from 132.66 to 55.38% FTE	Noparatayaporn et al. 72
ADC decentralized	Comparative study	1997		Time reduction of 46%. With 16% increase in dispensing interventions	36% increase in time	Martin et al. 52
ADC decentralized		2014, UK hospital	66 min/service/day saved			Cottney 53
ADC decentralized	Comparative study	2015, University Hospital Surgical intensive care unit	Time saved with an average of 14.7 h/day/33 beds		Increase of 3.5 h/day devoted to storage activities	Chapuis et al. 75
ADC decentralized	Cost-benefit type medical economic study	2008 an intensive care unit at Grenoble University Hospital, a 2000-bed establishment	1.9 h/day reduction in time spent on storage. Reduction of 0.4 h/day for orderlies		Additional time of 0.7 h/day	Kheniene et al. 80
Centralized drug retrieval cabinet and robotic XR 2		2024		Decrease of 0.77 FTE	Decrease of 1.76 FTE for pharmacy technicians	Lin et al. 74
Centralized automated drug distribution		Mono-centric retrospective 6 months before 6 months after		Time gain 0.67 h/day	Gain of 1.32 h/day. Storage managers had an increase of 0.95 h/day	Baraka et al. 81
Centralized robot, medication carousel	Comparative study	2010, Mercy Hospital and Medical Center			Reduction from 1.25 to 0.4 FTE (72%)	Steven et al. 73

FTE = full-time equivalent.

Impact on Working Conditions

The installation of automated dispensing systems has also optimized the management of drug requests and returns. De-Carvalho et al., ⁸² reported a 71% decrease in urgent requests and a 30% reduction in products returned to the pharmacy. Almalki et al. ⁸³ also found a 72% decrease in the number of items returned each month, ⁸³ similarly a study of the Saudi medical service conducted in 5 nursing units showed that overall drug dispensing decreased by an average of 43%. ⁴⁶ These advances have reduced interruptions due to urgent orders and drug returns.

ADCs have also helped reduce difficulties in accessing medication, with Ardern et al. ⁸⁴ finding that 66.1% of staff had experienced problems with storage keys prior to their implementation, problems that have now been resolved with the system.

Nursing staff satisfaction with these systems has been examined by several studies, the majority of which report that working conditions have become more comfortable and secure, and that their productivity has increased due to the time saved.^49,81,83-86

Economic Impact of Automation

Implementing automated systems in medication management and dispensing offers significant economic benefits, including reduced drug costs, losses due to expired products, and personnel expenses. The total cost of an automated system includes equipment, renovations, and the associated computer system. The economic impact varies from 1 automation system to another:

Centralized Preparation Machines

Lin et al showed that adoption of the IVWMS significantly reduced wasted and missed IV doses by 14,176 and 2268 doses, respectively. Consequently, the overall savings from using the system amounted to $144,019 over 3 months. ⁸⁷

A retrospective cost-benefit analysis conducted at the National Institute of Oncology in Morocco, evaluating the automation of cytotoxic drug preparation using an automated system compared to manual methods, showed a reduction in annual drug consumption costs of 19.74%, a saving on drug-related expenses of $41,228.27. It also showed a decrease in personnel costs of $48,073.35. ⁸⁸ Although the system incurred a high initial investment of $2,934,098.74, the cumulative savings demonstrate its potential to break even within 2 years, particularly due to its efficiency in reducing medication waste through vial-sharing capabilities.

Carignani et al. ⁸⁹ conducted a study to evaluate the economic sustainability of automation by comparing the costs of manual and automated preparation models in onco-hematology. The analysis revealed an annual savings of $91,275, with a marginal savings of $5.85 per preparation and a return on investment estimated at 3.3 years for a robotic system with a lifespan of 8 years. These findings demonstrate the economic viability of the investment. ⁸⁹

Centralized Distribution Machines

A cost-benefit analysis by Baraka et al. ⁸¹ showed that after implementing an automated supply system, cytotoxic drug consumption decreased by 9%, expired drugs decreased by 98.3%, stockouts also decreased by 41.1%. This resulted in a total saving of $134,874 in just 1 year. According to Temple, after implementing a dispensing carousel in the pharmacy, the filling of ADCs reduced the amount of technical labor required to support drug distribution, on the other hand the storage cost was reduced by $25,059. ⁹⁰ Other studies have shown similar results, related to the savings following the reduction in the value of the drug stock.^78,91 A recent study found that after 7 years of using the machines, the actual return on investment reached +$163,331, confirming that, despite their high initial cost, ADCs represent a profitable investment. ⁹²

Decentralized Distribution Machines

Another study made on ADC, showed that 1 year after implementation, personnel costs were reduced by $14,444, mainly through reduced working hours and reassignment of tasks. ⁷⁹ Regarding the cost of expired medications, Chapuis et al. ⁷⁵ found that the use of automated dispensing cabinets (s) in 3 intensive care units reduced losses by $15,742 per year. Similarly, Almalki et al. ⁸³ reported estimated annual savings of approximately $750,000 from a 57% reduction in medications expired, contributing to a total annual saving of $4.1 million.

Schwarz and Brodowy ²⁶ concluded that replacing single-use cassettes with s resulted in total cost savings of $2.08 million over 5 years, largely due to reduced staff time. In a similar study, Wise et al. ⁷⁹ estimated annual cost savings of $80,910, due to time saved by healthcare professionals using ADCs. Studies by Batson et al. ⁹³ also found additional cost reductions through labor savings, reduced storage costs, and fewer expired medications. Finally, Khenien et al. ⁷⁸ observed a 56% reduction in the value of medication stock in wards, representing a saving of $15,701 in addition to $24,326 saved through reduced staff hours.

Automation in hospital pharmacy, whether applied to preparation or dispensing, contributes significantly to economic optimization by reducing costs related to medications, losses and labor, these savings strengthen the argument for the widespread adoption of these technologies.

Challenges and Limitations of Automation in Hospital Pharmacy

Automation in hospital pharmacy, although beneficial, is accompanied by significant challenges. These can be financial, organizational or technical, requiring strategic planning for successful implementation. ⁹⁴ Table 3 outlines the advantages and disadvantages of various automation systems, offering insight into the complexities of automation adoption.

Table 3.

Advantages and Disadvantages of Different Automation Systems.

System	Advantages	Disadvantage
Automated dispensing cabinets (ADC)	- Reduced medication errors - Reduced medication administration and retrieval time - Improved inventory traceability - Reduced staff workload - Reduced costs	- High initial cost - Requires maintenance - Staff training required - Risks in the event of technical failure - Requires continuous inventory control - Risk of selection error
Robotic preparation systems	- High precision in preparation - Reduced risk of contamination - Protects staff and the environment - Optimized time	- Significant investment - Costly regular maintenance - Complexity of integration with existing hospital IT systems
Pneumatic tubes	- Reduced drug transport time - Reduces staff travel	- Impact on the stability of certain pharmaceutical products (proteins) - Difficult installation - High energy consumption
Electronic prescription	- Limiting incorrect prescriptions - Improved communication between professionals	- Variable adoption by doctors - Interoperability problems with other systems
Traceability systems (RFID)	- Precise location of drugs - Detailed tracking of the product route	- High installation cost - Compatibility with other technologies required

Financial Challenges and Cost Management

Installing automated systems represents a significant initial investment, often prohibitive for hospitals with limited budgets, ⁹⁵ and maintenance costs are also a major constraint. ⁹⁶ Hospitals must allocate resources for maintenance, software updates, hardware repairs and staff training, and effective management of these costs is crucial to maximizing the return on investment ⁹⁷ of automation systems and ensuring their sustainability. ¹⁴

Training and User Adaptation

The transition to automated systems can be accompanied by sociotechnical challenges, including resistance from healthcare personnel. ⁹⁸ This makes extensive training essential to familiarize staff with the new devices. Craswell et al. ⁹⁹ observed that users experienced anxiety during the first uses of the system due to their lack of familiarity with computerized medication retrieval processes, thus, it is important to include comprehensive and ongoing training of the relevant staff when introducing an ADC.^100,101 This significantly reduced the rate of user errors, highlighting the importance of involving them in the design, installation, and evaluation of new technological systems.^99,100

Training programs aimed at supporting the adoption of automation in healthcare have proven effective in improving professionals’ skills and reducing errors. Straight et al., ¹⁰² demonstrated that training nurses using a process improvement strategy, including an online self-learning module, led to a reduction in the error rate associated with automated systems.

Similarly, training programs for pharmacists have focused on enhancing digital skills, such as using automated dispensing systems and telepharmacy technologies. These programs have successfully improved pharmacists’ efficiency and confidence in handling automation. ¹⁰³ Practical sessions, where pharmacists interact directly with automated systems, have further reduced resistance and anxiety, facilitating smoother integration into daily workflows. ¹⁰⁴

Ongoing workshops and refresher courses help pharmacists stay updated on new technologies and best practices, ensuring continuous adaptation. ¹⁰³

Technical Issues and Downtime

Automated system failures can disrupt workflows, requiring technical adjustments and process reorganization to ensure continuity of delivery. A study by Hanuscak et al., ¹⁰⁵ showed that uncontrollable downtime can be attributed to interface, software, or hardware failures or malfunctions, which could jeopardize patient safety. Among the challenges encountered, users have reported operational issues such as sunlight interference on machine screens, which can affect their functionality. ⁹⁹ Disruptions, such as power outages, brownouts, and surges, are a major concern because they can interrupt medication availability and disrupt critical processes. To address these disruptions, hospitals can install backup generators and implement manual workflow procedures to ensure continuity of medication distribution.^106,107

The challenges associated with automation in hospital pharmacy are significant but surmountable with proactive management. Strategic planning, proper training, and technical mitigation measures can reduce these barriers and enable hospitals to fully leverage the benefits of automation.

Limitations of Automation Applications

For unit-dose dispensing machines, a deblistering or primary deconditioning phase is essential to obtain single doses; this step is a weak link which compromises the quality of the medication, with a risk of instability, ¹⁰⁸ contamination or interaction with other medications. ¹⁰⁹

Regarding ADCs, issues of practical use and accessibility have been raised:

• Complex user authentication through the requirement of passwords, badges or biometric verification adds security but can delay access, particularly in emergency situations where immediate access to medications is essential.

• The risk of selection errors encountered with drawers containing several drugs, where users can access all the drugs present in the same drawer. ⁴⁷ On the other hand, certain systems with compartmentalized pockets allow access to a single line of an article depending on the user.

• Incorrect stocking or filling errors (eg, placing medications in the wrong drawer) can lead to significant picking and dispensing errors. ⁵ Inventory discrepancy and expired medications can sometimes contain expired or missing medications due to inaccurate inventory tracking or poor replenishment practices; monitoring stock levels manually or through inventory systems is essential.

• ADC are often large and can require significant space, making them difficult to install in smaller or crowded healthcare facilities, and their capacity for a variety of medications is limited. Depending on the model, some ADCs may have limited space for specific requirements, such as refrigeration, which may limit their usefulness in facilities with a wide variety of pharmaceutical needs.

• Among the most common complaints from nurses using the decentralized dispensing systems is the need to queue during peak hours, often compounded by interruptions from other staff.^46,110

Integration of AI in Automated Drug Dispensing and Preparation Systems

Artificial intelligence (AI) is a technological system that uses advanced algorithms and networks to mimic human intelligence, including learning, reasoning, and decision-making, ¹¹¹ its application has extended to the pharmaceutical sector, where it improves productivity by reducing the workload of professionals. ¹¹² The success of AI relies on the availability of large volumes of reliable and high-quality data to support decisions that impact patient safety and ensure therapeutic efficacy, which may involve potential drug interactions or dosage recommendations based on integrated patient data. ¹¹³ The integration of AI into automated dispensing systems can improve patient safety, prevent stockouts by monitoring inventory levels and average drug consumption per hospital department, and optimize drug use. ¹¹⁴ Finally, AI allows clinical pharmacists to focus more on patient care rather than administrative tasks. AI-based systems can analyze distribution patterns, predict periods of high demand, and adjust staffing or machine usage to improve efficiency. ¹¹⁵ Praveen et al. ¹¹⁶ examined the integration of AI technologies in optimizing inventory management and forecasting customer demand. Their study highlights substantial enhancements in forecast accuracy and inventory turnover rates resulting from AI implementation, while also exploring its potential future impact on supply chain management.

AI enables robots to adjust quantities in real time and detect any deviation from predefined quality standards, reducing the risk of human errors and irregularities. This not only improves the reliability of pharmaceutical preparations, but also patient safety, by minimizing the risks of incorrect administrations. ¹¹⁷

Future and Patents of Automation in Hospital Pharmacy

The future of automation is experiencing rapid technological advancements and continued growth in patented innovations, with the rise of artificial intelligence and machine learning, automation is becoming increasingly sophisticated and versatile, opening the way to expanded applications

Recent patents highlight significant innovations across all stages of preparation, dispensing, and distribution.

Drug Preparation and Compounding

Several innovations focus on automating the preparation and compounding of medications, ensuring efficiency and precision. These advancement include drug dosage preparation systems, ¹¹⁸ automated pharmaceutical compounding systems, ¹¹⁹ automated sterile compounding stations that prepare intravenous medications, manage inventory, and print storage labels with expiration dates, ¹²⁰ and systems that automate hazardous drug compounding using robotic arms and real-time monitoring. ¹²¹

Drug Dispensing and Delivery

Other patents focus on automating the dispensing and delivery of drugs, improving accuracy and speed. Innovation include methods for managing drug preparation, dispensing, and recycling, ¹²² systems designed to improve dispensing efficiency,^123,124 medication retrieval system featuring a robot designed to automate the process of retrieving and delivering medications in a hospital setting, ¹²⁵ perfusion system designed to directly administer an antineoplastic drug dose to a patient, calculated based on the patient’s parameters, ¹²⁶ Infusion pump systems that process therapy data to generate and output tailored protocols for optimized drug delivery, ¹²⁷ and system utilizing Simultaneous Localization and Mapping (SLAM), allowing robots to autonomously manage inventory and deliver medications to patients. ¹²⁸

Patient-Related Systems

Innovations also extend to improve patient care, with systems that monitor and improve patient-specific therapies, one such innovation is parenteral nutritional diagnostic systems that analyze muscle quantity and quality to assess nutritional status and offer precise recommendations. ¹²⁶ These innovations are presented in Table 4.

Table 4.

Some Patents on Automation in Hospital Pharmacy.

Patent number	Patent title	Author(s) name	Publication date	Reference
1-WO US US11250957B2	Medication preparation system	Dennis Tribble	2022-02-15	118
2-US US11182728B2	Medication workflow management	Nivaldo Diaz	2021-11-23	122
3-WO US JP AU CA ZA US10064579B2	System and method for dynamically adjusting patient therapy	Joseph Condurso	2018-09-04	124
4-US10315851B2	Apparatuses, systems, and methods for improving the efficiency of medication distribution	Shawn T. Greyshock	2019-06-11	123
5-WO EP US EP2457549B1	Automated pharmacy admixture system (apas)	Alex H. Reinhardt	2016-06-08	119
6-CN113146576B	Robot based on medicine taking system and control method	Université de technologie de Hefei	2024-08-20	125
7-ES2968415T3	Perfusion system	Subhas Balaram Bhowmick	2024-05-09	126
8-S11200979B2	Pharmacy automation using autonomous robot	Neil S et al.	2021-12-14	128
9-US20240013150A1	Automated preparation of medications in anticipation of use	Dennis Trible et al.	2024-01-11	120
10-US12083317B2	Parenteral nutrition diagnostic system, apparatus, and method	Stephen A et al.	2024-09-10	129
11-US11160921B2	Pump infusion system	Achilleas Tsoukalis	2021-11-02	127
12-ES2856339T3	Robotic system for compounding drugs	Marino Kriheli et al.	2021-09-27	121

Conclusion

This work highlights the significant evolution of automation technologies in hospitals, which play a crucial role in the transformation of care practices. These technologies, whether automated distribution systems, robotic preparation medication or traceability solutions, have demonstrated their ability to improve patient safety, reduce medication errors and optimize workflows for healthcare professionals.

However, their implementation is not without challenges, with financial constraints, organizational complexities and technical limitations remaining major obstacles to widespread adoption.

The future of automation in hospitals is bright, particularly with the increasing integration of artificial intelligence and machine learning systems, which open up new possibilities for further personalizing care and optimizing resources.