The application of ABC-VED with multi-criteria analysis for drug inventory management

Abstract

Effective drug inventory management is critical for ensuring the availability of essential medicines while minimizing costs and reducing wastages in healthcare institutions. This study presents a comprehensive approach that integrates Always Better Control (ABC) and Vital Essential Desirable (VED) analysis with multi criteria analysis to optimize pharmaceutical inventory control. By combining these two methods, inventory items are stratified into more meaningful categories for prioritization. Research is warranted in light of the case Hospital’s ongoing inventory management problem. This project uses both qualitative and quantitative data collection approaches in order to design a drug inventory control system specifically for the pharmaceutical department at the case Hospital. A normalization technique is applied to avoid data quality issues, reduce data redundancy, improve data analysis, and enhance data security. The Economic Order Quantity (EOQ) model and forecasting methods are also applied. In the end, the inventory cost is decreased from 207,600 to 170,998B ETB, a noteworthy 17.63% improvement that can be applied in real-world situations. The proposed ABC-VED with multi criteria analysis framework enables healthcare facilities to allocate resources more efficiently, improve stock management, and ensure the uninterrupted supply of critical medications. This integrated approach supports evidence based policy making in pharmacy management and offers a scalable model adaptable to varied healthcare settings.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-32068-w.

Introduction

The concept of inventory relates to the stock of materials stored in a warehouse, serving purposes such as managing uncertainty, facilitating wholesale purchasing, reducing waiting times, maintaining safety stock for seasonal variations, and optimizing the manufacturing process to meet customer demands¹. Inventory management involves a systematic approach to identify and control items requiring special attention using various tools. Always Better Control (ABC) is a crucial tool for assessing the class of items based on their annual consumption value, particularly focusing on highly important drugs that impact both patient health and organizational expenses. However, some researchers have criticized ABC for overlooking the importance of items beyond their annual consumption². To address this limitation, the Vital Essential Desirable (VED) analysis has emerged as a complementary practice. Integrating ABC, VED, and lead-time considerations is essential to fully capture the significance of these analyses³. Economic Order Quantity (EOQ) is another inventory control technique used to balance an organization’s annual holding and ordering costs, which was overlooked in previous research⁴. Hospitals, as providers of vital healthcare services, require continuous and optimal drug supplies to their pharmacies. Numerous inventory control challenges have been identified in hospitals⁵. Effective inventory management is not only crucial for patient care but also for controlling drug stock^6,7. Reducing the number of items in storage decreases holding costs but increases setup or ordering costs due to a higher frequency of orders. However, reducing the number of drug items also increases the risk of stock outs. The challenges of inventory management in healthcare organizations have prompted researchers to focus on further investigations, particularly in the pharmaceutical sector^8,9. The increasing annual expenses resulting from improper drug inventory management in several developing countries further highlight the need for such research. Consequently, this study takes Ayder Referral Hospital in Ethiopia as a case study to explore the combined use of ABC-VED analysis, forecasting techniques, and the EOQ model. Therefore, this study develops an integrated drug classification model combining ABC–VED analysis with multi-criteria analysis, balancing inventory cost minimization and availability of essential medicines in hospital systems.

The objective of this research is to develop an ABC-VED classification system with multi-criteria analysis for drug inventory management at Ayder Referral Hospital to ensure a balance between minimizing costs and ensuring the continuous availability of vital medicines, which is of highest importance in the healthcare sector. To meet this research objective the following research questions are addressed: How can the ABC-VED classification system be modified and tailored to meet Ayder Referral Hospital’s demands for medicine inventory management? What important parameters and elements need to be taken into account in the multi-criteria analysis for Ayder Referral Hospital’s medication inventory management? What are the possible advantages and difficulties of applying the suggested strategy to Ayder Referral Hospital’s medication inventory management? How does Ayder Referral Hospital’s medicine inventory management function better overall, in terms of cost-effectiveness and efficiency, with the use of the created ABC-VED and multi-criteria analysis approach? How to minimizing inventory cost and ensuring the continuous availability of critical drugs? What suggestions are there to improve the efficiency of the ABC-VED drug inventory management approach in healthcare settings using a multi-criteria analysis approach?

Related literatures

Inventory management is crucial in both manufacturing and service industries, with a particular focus on hospitals. Researchers have extensively studied inventory control in service industries, including hospitals^8,10. The objective of inventory control is to optimize the balance between ordering and carrying costs while meeting demand and service requirements¹¹. However, in developing countries like Ethiopia, inventory management in hospitals still requires individual effort due to inadequate demand and supply proportions, as indicated by historical data. The calculation of buffer or safety stock as an essential aspect of inventory management¹¹. Maintaining the appropriate stock level of items is a fundamental responsibility of inventory management^12,13.

In the context of hospital services, the management of drug inventory is critical for ensuring the availability and safety of essential drugs. Inventory management plays a vital role in maintaining the desired stock of drugs. Drug management activities in hospitals involve drug selection, purchasing, storage, and inventory control¹². In drug selection, hospital pharmacists are responsible for making informed decisions regarding the type, quantity, specifications, and sources of drugs, considering factors such as price, lead time, service quality, demand fulfillment, and other necessary criteria¹⁴.

Inventory management practice: pharmaceutical category

The pharmaceutical sector holds significant importance in the healthcare industry, primarily due to the high costs associated with drug storage and control parameters^8,15. Consequently, it becomes crucial to establish efficient drug inventory management practices in the pharmaceutical field to ensure the timely delivery of desired drugs to patients at an affordable cost and in optimal condition⁸.

In the pharmaceutical sector, dealing with uncertainties often leads to challenges such as drug shortages or excessive quantities, which significantly impact the industry^8,16–18. Implementing effective inventory management practices in the pharmaceutical sector can enhance the key performance indicators (KPIs) of drug supply chains and reduce liquidity levels in China’s public hospitals^7,16. In contrast, inadequate drug inventory management can have detrimental effects on the financial stability of medical organizations, resulting in unnecessary expenses^19,20.

In their fiscal investigation of drug expenses in a government medical college hospital, highlighted the need for a robust control mechanism for drugs¹⁹. Their findings revealed that the annual drug usage accounted for only 11.59% of the total hospital capital. Analyzing the percentage cost of each drug assisted in determining the appropriate order quantity and reorder point for high-value but less desirable drugs. By considering the cost inflation index, the study showed that annual drug expenditure increased by approximately 2.84% when factoring in inflation. Ultimately, the study concluded that the combination of ABC-VED analysis in drug classification helped streamline and reduce the number of drugs in inventory.

Measuring and enhancing the availability of drugs and the effectiveness of stock management systems in medical organizations is crucial²⁰. Predicting potential drug shortages is essential to prevent the costs associated with excessive stock levels. By implementing a suitable inventory control system backed by sufficient data, pharmaceutical companies can adjust their inventory policies accordingly²⁰.

As described in the comparison table of previous studies in appendix 2, various scholars have conducted research on inventory management and its control systems. Despite the wide spread use of ABC and VED analysis in healthcare inventory management, the literature reveals a lack of integrated frameworks that holistically consider both consumption value and clinical criticality with additional multi criterial analysis. Moreover, existing models often overlook the contextual complexities of hospital pharmacies, such as the need for prioritizing lifesaving drugs under budget constraints, lead time, and integrated it with forecasting technique. In response to these gaps, the aim of this study is to develop ABC-VED with multi criteria analysis for drug inventory classification and prioritization to ensure a balance between minimizing costs and ensuring the continuous availability of vital medicines, which is of highest importance in the healthcare sector.

Research methods, tools and techniques

Research design

This study have used a mixed-methodologies strategy, including quantitative and qualitative data gathering and analysis methods. Multiple stages of the study also carried out, encompassing data collection or gathering, research tools and techniques used for the study, data analysis, model development, and assessment.

Data collection

A mix of primary and secondary data collection techniques were used in this investigation. Owing to the delicate nature of the workplace, the quantitative features of the data were given more weight. Information on a number of variables, such as the number of orders placed in a given year, ordering costs, drug unit costs, annual demand, the number of drugs in stock, current drug classification techniques, drug criticality, stock status, lead time, patient service, and drug availability, were gathered as part of the primary data collection process. The hospital’s capacity was maintained during the data collection period, which ran from June 2016 to May 2020. To ensure the representativeness of the drug sample used in this study, we conducted a comparative analysis between the sample and the overall hospital drug inventory across key attributes, including therapeutic class, dosage form, and price range. The research sample size consisted of 412 frequently over stocked and stocked out drugs selected from the 3000 medicines available in the case hospital, which is acceptable since this represents approximately 13% of the total drugs²¹.

Data Analysis, materials and methods

There are numerous widely used inventory classification methods, each with a distinct strategy and objective. The following are a few of the most popular methods:

ABC Analysis is a method for classifying inventory items according to their worth or significance to the company. Three categories are used to classify the items: A, B, and C. Items in category “A” usually make up a small portion of the entire inventory but are highly valuable or significant. Items in category ‘’B’’ are moderate value and usage. Items in Category C are often more plentiful and have a lower value than those in category B. “Items were classed according to annual consumption value. A-class items are the top 15% of items, accounting for 75% of the overall value. B-class items account for the next 25% of things, contributing 20% of the value, while C-class items account for the remaining 60%, contributing only 5% of the value.” Effective resource allocation and prioritization of inventory management activities are facilitated by this classification. ABC can be used in the subsequent contexts:

Hospitals can investigate the factors that influence costs in relation to various activities thanks to ABC. This data aids in locating areas that can be optimized or cost-cut, which improves financial management and the distribution of resources more effectively. This analysis assists in identifying high-cost areas or services that might need to be adjusted or subject to additional scrutiny.

VED Analysis: Inventory items are categorized using VED Analysis according to how important they are to the organization’s operations. It can be used to efficiently manage medical equipment and supplies:

Inventory Prioritization: Using VED, resources can be ranked according to how important an item is. Items deemed vital, like life-saving medications or emergency gear, are given the utmost care and supervision. Items that are desirable but less critical are controlled less strictly than essential items that are required for regular care.

Stock Management: VED helps hospitals keep the right amount of essential items in stock so they are always available when needed. Additionally, it aids in the identification of non-moving or slow-moving items, enabling improved inventory control and preventing needless stock accumulation or waste.

Risk Mitigation: VED helps reduce the likelihood of stock outs or shortages of essential supplies by classifying items according to their criticality. By keeping an adequate stock of essential supplies, hospitals can lower the risk of delays in patient care or operational effectiveness.

Inventory items are classified using FSN Analysis, which takes into account how they are typically consumed. Three categories are used to classify the items: F, S, and N. F represents fast-moving goods with high rates of consumption and regular turnover. S stands for items that move slowly and have a moderate rate of turnover and consumption. N stands for non-moving items, which frequently signal possible obsolescence or overstocking because of their low turnover and consumption rates. This classification aids in locating products that could require special attention when regards to stock rotation and inventory management.

Inventory items are categorized using the High-Medium-Low (HML) Analysis system according to their unit prices or costs. Three categories are used to classify items: High, Medium, and Low. Based on the item value, this classification aids in selecting appropriate inventory control measures. Low-value items might have less strict control requirements, but high-value items could require more frequent monitoring and more strict control. By providing organizations insights into their inventory profiles, these inventory classification techniques help them make well-informed decisions about inventory management, resource allocation, risk mitigation, and optimization.

As you see from these analysis all of them are used single criteria to classify items. However, it does not give the correct position or class of items. To alleviate these limitation this study applied the combination of ABC-VED analysis with multi criteria factors. It has different advantage especially for pharmacy hospitals. The ABC-VED method is unique among available approaches due to its comprehensive classification, ease of use, resource optimization, and benefits for risk mitigation.

Holistic Classification: The ABC-VED approach combines the value-based (ABC) and criticality/control requirement-based (VED) classification of items. Taking into account the importance of inventory items to the organization and control requirements, this through approach provides a broader view of inventory items. ABC-VED provides a more accurate understanding of the significance of the inventory by integrating criticality and value, which promotes decision-making.

Simplistic and Ease of Implementation: The ABC-VED approach is comparatively simple to implement. Organizations can easily access and understand the categorization because it is predicated on readily quantifiable criteria like value and criticality. Because of its simplicity, the classification can be applied and communicated to pertinent stakeholders in an efficient manner, which supports efficient inventory management procedures.

Resource Allocation Optimization: ABC-VED helps with resource allocation optimization by taking inventory items’ value and criticality into account. Important things, which are highly valued and need extremely strict controls, get the time and funding they need. On the other hand, less important things are handled proportionately, which saves needless expenses and labor. This methodology guarantees the optimal allocation of resources, thereby improving the overall efficacy of inventory management.

Risk Mitigation: ABC-VED’s capacity to recognize and control risks related to inventory items is one of its main advantages. Organizations can actively reduce the risks of stock outs, production delays, and disruptions by concentrating on essential items and putting strict control measures in place. This proactive approach to risk management improves customer satisfaction and guarantees seamless operations.

Currently, the hospital applies VED sometimes called VEN analysis for classifying drugs which is based on their criticality or function. Vital drugs are critical for patient care and have a direct impact on health outcomes. Essential drugs are necessary for routine treatment but may have alternatives. Desirable drugs are non-essential and have a lower priority. However, VED/VEN analysis alone does not give the correct position or class of items. So to alleviate this limitation combination of ABC and VED analysis is somehow good due to the listed advantages. By combining the ABC and VED analyses, drugs are classified into different categories based on both their consumption value and criticality. This classification helps in prioritizing inventory management efforts. For example, drugs categorized as “A” and “V” items (high consumption value and vital) require special attention and close monitoring to ensure their availability and prevent stock outs. On the other hand, drugs categorized as “C” and “D” items (low consumption value and non-essential) may have lower priority in terms of inventory control.

But, still it has some limitation, because it doesn’t consider other factors except the cost or consumption value and criticality of the items while combining ABC and VED analysis. So this study incorporated this factors and other advantages when combining ABC-VED analysis to make the class of the drug is accurate. The ABC-VED with multi criteria analysis provides a framework for identifying drugs that require stringent inventory management practices, such as frequent monitoring, adequate stock levels, and efficient procurement strategies. It helps healthcare organizations optimize their inventory by focusing resources on critical drugs while reducing costs associated with less essential items.

The data analysis tool utilized in this study was MS-Excel. For the ABC analysis, the data was inputted into Microsoft Excel, and the drugs were arranged in descending order based on their Annual Dollar Usage (ADU). The drugs were then classified into categories A, B, and C. This classification method demonstrates the limited influence of a single criterion on the drug classification. Additionally, the study examined the existing drug classification system employed by the hospital. After analyzing the two common drug classification techniques, the study focused on the specific approach of combining ABC-VED with multi-criteria analysis. To develop this method, the study followed the following patterns or procedures.

An inventory with “i” drugs items and the items is categorized based on “j” criteria. The measurement of the i^th item under the j^th criteria is denoted as “y_ij”. All measurements are formulated to develop an improved inventory model.

The action using transformation (y_ij- min_i=1, 2,.,n {y_ij})/(max_i=1,2, …,n {y_ij} - min_i=1,2,.,n {y_ij}) can be fundamental to change all measurement in a 0–1 scale for all drugs. To simplify the inventory classification under multiple criteria, a positive weight w_ij which is the weight of influence of performance of the i^th drug item under the j^th criteria is denoted. The score of the i^th item (denoted as S_i) is expressed as a weighted sum of performance measures under multiple criteria. A weighted linear optimization model is used to calculate the score of each drug item accordingly².

1

2

3

Where: S_i stands for item score for i^th drug (shown in Table 3), W stands for the weight of each parameter from 0 to 1. The weight of each parameter which are, (average unit cost, annual dollar usage, critical factor and lead-time of the drug) is decided by the committee formed in the case study and the sum must be 1 (Appendix 3). And “Y” stands for the multiple measures of the i^th item under the j^th criteria, and s.t stands for subjected to the constraint. This means the weights in the multi-criteria analysis were determined using expert judgement and a structured deliberative procedure. A panel of five subject matter experts, each with appropriate experience in environmental science, policy analysis, and socioeconomic evaluation, was assembled to determine the relative importance of each criterion. Initially, each expert individually ranked the criteria according to their considered relevance to the study objectives. These individual rankings were then debated in a mediated session with the goal of reaching a consensus. A weighted scoring system was used, in which rankings were assigned numerical values and the aggregated scores from all participants were normalized to provide the final weights. This approach was chosen over formal procedures like equal weighting because of the manageable amount of criteria, the need of adding domain-specific insights, and practical factors such as time and budget limitations. The resulting weights represent expert opinion and aim to strike a compromise between methodological rigor and contextual relevance.

Table 3.

Comparisons between the existing & improved drug classification.

Items Name	ABC classification based on:
	ADU (Most familiar method)	Criticality (Existing System)	ADU, Criticality & lead time (Improved System)
Atrovastatin - Tablet - 10 mg	B	A	A
Clonazepam - Tablet - 0.5 mg	C	B	C
Codeine Phosphate - Tablet - 30 mg	C	A	A
Diazepam - Tablet - 5 mg	C	B	C
Lidocaine Hydrochloride - Injection - 2% in 20 ml	A	C	C
Metoclopramide - Injection - 5 mg/ml in 2 ml ampoule	A	A	C
Nifedipine - Tablet - 20 mg	B	A	C
Omeperazole – Capsule (Enclosing Enteric Coated Granules) - 20 mg	A	A	C
Pantoprazole - Tablet - 40 mg	B	A	C
Cimetidine- Tablet- 400 mg	C	C	B
…	…	…	…
Verapamil- Tablet-400 mg	C	B	C

All scores S_i (i = 1, 2. n) constantly within a scale of 0–1 as all measures y_ij are in 0–1 scale and all limited scores are in an equivalent base when both weights and measurement are normalized. Following the principle of ABC – VED with multi-criteria analysis, the drugs were classified into classes ‘A’, ‘B’, and ‘C’ as depicted in Table 3. Once the 72 drugs were categorized using this analysis, the subsequent step entailed employing quantitative time series forecasting techniques to predict the demand for the upcoming year²².

Assumptions of the weighted linear optimization model

• I.The model assumes that the relationship between decision variables and the objective function is linear. This simplifies the problem but may not capture real-world nonlinear behaviors.
• II.All input data (demand, cost, lead time, and criticality scores) are assumed to be known and constant. The model does not account for uncertainty or variability.
• III.It presumes that the decision-maker can assign appropriate weights to each criterion, reflecting their relative importance accurately.
• IV.The model assumes fixed thresholds for classifying items into A, B, or C categories, which may not always reflect dynamic business needs.

Limitations of the weighted linear optimization model

• A.Choosing weights for each criterion is often subjective and may not reflect actual business priorities unless validated by domain experts or through historical data.
• B.It does not incorporate stochastic elements like demand variability, lead time uncertainty, or supply disruptions, which are common in real world inventory systems.
• C.The accuracy of results heavily depends on the quality and completeness of input data. Inaccurate or outdated data can lead to sub-optimal or misleading classifications.

The reason why we are using Weighted Linear Optimization Deterministic Model is:

1. In many practical settings, especially small to medium enterprises, probabilistic data like demand distributions may not be available or reliable.

2. It avoids the complexity of modeling uncertainty, making it suitable for environments where data is relatively stable or well-estimated.

3. Businesses often rely on historical averages or expert estimates, which naturally fit into deterministic models.

4. It avoids the complexity of modeling uncertainty, making it suitable for environments where data is relatively stable or well-estimated.

In order to enhance the precision of drug demand forecasting across various categories (A, B, and C), healthcare organizations can integrate forecasting techniques with the ABC-VED approach and multi-criteria analysis. This combination enables more informed decision-making in terms of inventory levels, replenishment strategies, and resource allocation. The ultimate objective is to attain optimal drug availability while mitigating the risks associated with excessive inventory or stock shortages.

Moreover, taking into account the distinctive characteristics of each drug, a customized forecasting technique is devised for the 29 drugs categorized under class ‘A’. This tailored approach guarantees precise predictions of the demand for each drug, considering its individual attributes and consumption patterns. By employing this analysis, healthcare organizations can efficiently strategize for the future, ensuring the availability of crucial drugs while optimizing inventory management practices.

Using the forecasted demand value as a constant demand Economic Order Quantity (EOQ) model are also utilized to develop an optimal number of drug items and minimizing the total inventory cost developed by^23,24.

4

Where: Q stands for quantity (EOQ), D is demand, S stands for ordering cost, C is unit cost and H stands for holding cost per unit. The total annual inventory cost is the sum of annual ordering costs and annual carrying cost such that:

5

In addition to the Economic Order Quantity (EOQ) technique, two other important methods employed in inventory management are the Reorder Point (ROP) and Safety Stock. The Reorder Point indicates the threshold at which an item needs to be re-ordered when its quantity on hand falls below a certain level. On the other hand, Safety Stock is the additional inventory maintained to prevent back orders and accommodate uncertain demand. When there is uncertainty in demand, it is necessary to maintain an adequate level of safety stock to fulfill the desired demand. To determine the timing of the next order, three factors need to be considered: demand, lead time, and safety stock. The reorder point without safety stock can be computed as:

6

Where, R = reorder point in units, d = daily demand in unit and L = lead time in days. Based on this approach the results of drug inventory model investigated and discussed accordingly.

Results and discussion

Classification of drugs using ABC analysis

ABC analysis is a method used to categorize items based on a single criterion, specifically the cost or annual consumption value of items. In the case of Ayder Referral Hospital, there are over 3000 types of drugs. Among these, approximately 412 items experience fluctuations in terms of stock outs and overstocking. Out of these 412 items, 72 drugs exhibit frequent fluctuations, and therefore, the study focuses on these specific items for classification.

The data for these drugs are recorded in an MS-Excel spreadsheet and arranged in descending order based on their Annual Dollar Usage (ADU). Subsequently, cumulative cost, cumulative cost percentage, and the percentage of the number of items are calculated. Based on the capital assets, the items are then categorized into groups A, B, and C. The results of the classification of these 72 drugs based on annual dollar usage, as collected from Ayder Referral Hospital, are presented in Appendix 1.

Which displays the classification of drugs solely based on their annual dollar usage, without taking into consideration any additional criteria. This means that the classification of drugs in the appendix is solely determined by their consumption value or cost, without considering factors such as criticality, consumption patterns, lead time or other relevant aspects. The focus is specifically on the monetary aspect of the drugs’ usage, allowing for a simplified classification based on their economic impact.

Classification of drugs in the existing system

As stated in the problem statement of Ayder Referral Hospital, there have been instances of irregular inventory levels specifically in certain categories of drugs. At Ayder Referral Hospital, VED/VEN analysis is the current drug classification method used based on the information available from the hospital document. Using this classification method, drugs are grouped according to how important or critical they are. Table 1 displays the current drug classifications used in the hospital.

Table 1.

Classification of drugs in the existing systems.

(source: reports from ayder referral Hospital).

S/No	Generic Name of The Medicine	Dosage Form (S)	VEN Category
1.	Omeperazole	Capsule - 20 mg	V
2.	Metoclopramide	Injection - 5 mg/ml in 2 ml ampoule	V
3.	Lidocaine Hydrochloride	Injection - 2% in 20 ml	N
4.	Nifedipine	Tablet - 20 mg	V
5.	Pantoprazole	Tablet - 40 mg	V
6.	Atrovastatin	Tablet - 10 mg	V
7.	Codeine Phosphate	Tablet - 30 mg	V
8.	Clonazepam	Tablet - 0.5 mg	E
9.	Diazepam	Tablet - 5 mg	E
10.	Cimetidine	Tablet- 400 mg	V
	….	….	….
72.	Verapamil	Tablet-400 mg	E

The table mentioned above draws attention to a problem with the way medications are currently classified in hospital settings. It implies that other significant factors may go unnoticed by the current classification approach, which might not take criteria into account beyond criticality. It highlights that while classifying medications as vital, essential, or desirable, consideration must be given to the healthcare environment, particular patient populations, and the frequency of diseases or conditions.

It is implied that there is a lack of a systematic and thorough evaluation of the drugs in the process by saying that the hospital’s classification is dependent on subjective assessments based on personal knowledge. This subjective method may introduce biases and reduce the drug classification system’s efficacy. It is essential to add more criteria to the classification procedure in order to address this problem. These standards might include things like how well the medication works, how affordable it is, and how long it takes to follow treatment plans or guidelines. The classification can become more patient-centered, evidence-based, and in line with industry best practices by taking into account a wider range of factors.

In addition, the way that medications are classified ought to be flexible enough to accommodate the unique circumstances of the hospital. The importance and prioritization of drugs can be affected by the patient populations, disease prevalence, and resource availability that differ between hospitals. As a result, not every situation calls for a standardized approach, and the classification scheme should be adaptable enough to take into account unique requirements and situations.

Classification of drugs in the improved system

One popular technique for classifying medications according to their cost and criticality is the ABC-VED analysis. The cost component entails evaluating the financial impact of pharmaceuticals, taking into account things like total expenditure and procurement costs. The significance of medications for patient care and medical necessity is the main emphasis of the criticality component.

Nevertheless, focusing just on price and importance could leave out other crucial elements that have an impact on medication management and supply chain effectiveness. Lead time is one of these variables; it’s the amount of time needed to obtain or restock a medication after an order is placed. Lead time has a direct impact on patient care by influencing drug availability and timely delivery.

Healthcare organizations can obtain a more comprehensive understanding of drug grouping by incorporating lead time as a criterion in the ABC-VED analysis. Drugs that might have longer procurement cycles, possible delays, or supply chain vulnerabilities can be identified by taking lead time into account. Better planning, inventory control, and reducing the possibility of stock outs or delays in patient care are all made possible by this information.

Healthcare organizations can improve the efficacy and precision of their drug classification by extending the analysis to include lead time. This method guarantees that medications are categorized taking into account availability and related procurement time, in addition to cost and criticality. Accordingly, the classification is summarized as shown in Table 2 below.

Table 2.

Classification of drug by combining ABC-VED with multi criteria analysis (Improved system).

Item Name	Received Quantity	Average Unit Cost (Birr)	Annual Usage (Birr)	Critical Factor (0–1)	Lead Time (Day)	Item Score (0–1)	Group
Atrovastatin - Tablet - 10 mg	6,000	6.90275	41416.5	1	5	1	A
Clonazepam - Tablet - 0.5 mg	20,000	0.94535	18,907	0.5	7	0.49	C
Codeine Phosphate - Tablet - 30 mg	20,100	1.080438	21716.8	1	4	1	A
Diazepam - Tablet - 5 mg	26,300	0.893717	12758.18	0.5	4	0.49	C
Lidocaine Hydrochloride - Injection - 1% In 50 ml Vial	4	9.2	36.8	1	12	0.04	C
Metoclopramide - Injection - 5 mg/Ml In 2 ml Ampoule	38,630	5.858028	152289.33	1	4	0.02	C
Nifedipine - Tablet - 20 mg	263,920	0.458792	87222.2	1	6	0	C
Omeperazole Capsule(Enclosing Enteric Coated Granules) - 20 mg	684,000	0.322777	166088.22	1	9	0	C
Pantoprazole - Tablet - 40 mg	15,420	2.84502	43278.6	1	14	0.02	C
Cimetidine- Tablet- 400 mg	16,400	0.46662	3947.45	0.01	14	0.52	B
…	…	…	…	…	…	…	…
Verapamil- Tablet-40 mg	60	0.32925	19.755	0.5	9	0.002	C

The multiple measures of the i^th item under the j^th criteria is denoted as y_ij and converted into a single score of drug item are utilized in the research work². The table presented demonstrates how drugs can be categorized differently based on the criteria applied. It specifically provides an example using the drug omeprazole to illustrate this point. According to the VEN analysis, which likely considers factors such as the importance of the drug for critical medical conditions, omeprazole is classified as a vital drug. This categorization suggests that omeprazole is considered crucial for immediate patient care, possibly for conditions such as severe acid reflux or stomach ulcers.

On the other hand, omeprazole is rated as desirable when using a multi-criteria measurement approach like ABC-VED with lead time. This suggests that other variables, such as demand volume, cost, and time needed to obtain the medication, are considered under this particular methodology in addition to criticality. Omeprazole is classified as a non-critical medication and belongs in the desirable category based on these extra factors.

This illustration demonstrates how different drugs are classified based on the measurement standards used. It highlights the fact that various analytical frameworks can result in disparate classifications and that, when assessing the significance and priority of drugs, it is crucial to take into account a variety of viewpoints and criteria. Different measurement approaches may take into account different goals, priorities, and contextual factors, which can lead to discrepancies in classification.

Healthcare companies can make better decisions about medication management, resource allocation, and patient care by recognizing these variations and comprehending the effects of different measurement criteria. It emphasizes how crucial it is to choose suitable measurement frameworks that complement the goals of the company and offer a thorough understanding of drug prioritization. To illustrate the analysis, few drug classes are taken as shown in Table 3.

The comparison table revealed discrepancies in drug classification when using ABC analysis, VEN analysis, and ABC-VED with multi-criteria analysis. This suggests that a single factor alone cannot accurately determine the position or class of a drug. When considering only ABC analysis based on cost, 14 out of 72 drugs were grouped under class ‘A’, accounting for approximately 69% of the annual drug budget. Similarly, when considering VED analysis alone, 40 out of 72 drugs were classified as vital or class ‘A’, representing 76% of the annual drug budget. However, the VED classification varied among hospitals based on patient healthcare needs. It was observed that drugs of desirable nature were classified as vital or class ‘A’ due to the use of a single criterion. For instance, in this study, drugs like Omeprazole - Capsule (Enclosing Enteric Coated Granules) − 20 mg and Metoclopramide - Injection − 5 mg/ml in 2 ml ampoule were classified as vital or class ‘A’, consuming a significant portion of the budget. Therefore, the desirable group of drugs cannot be ignored entirely. Consequently, it is crucial to consider both the cost factor and criticality of drugs, and utilize multi-criteria analysis when grouping drugs. The integration of ABC-VED with multi-criteria analysis facilitates prioritization. The ABC-VED matrix clearly indicates that all stocks are not equally valuable and do not require the same management focus.

The integration of ABC-VED analysis with Multi-Criteria Analysis (MCA) in this study provides a more comprehensive framework for drug inventory management by addressing both cost-based and criticality-based prioritization in a structured, decision-oriented manner. While traditional ABC (Always, Better, Control) analysis focuses solely on the financial value of inventory items and VED (Vital, Essential, Desirable) categorization emphasizes clinical importance, their combination when enhanced with MCA allows for the simultaneous consideration of multiple, often conflicting, criteria such as cost, criticality, and lead time. Previous research has applied ABC-VED matrices by considering the cost and criticality of the drug by ignoring other factors and used fixed weighting schemes, potentially oversimplifying decision-making processes. In contrast, this study applied ABC-VED analysis by considering different criteria like, demand, criticality, cost, and lead time and propose a transparent weighting approach using expert-based MCA, allowing for context-sensitive prioritization that takes into account both operational restrictions and healthcare priorities. This integrated approach improves the resilience and adaptability of medicine inventory management, resulting in more efficient resource allocation and better patient outcomes.

Categorizing drugs using the ABC-VED with multi-criteria analysis allows us to focus on vital and costly drugs, which require high management priority and strict control as they consume the largest portion of the budget. This study found that vital and high consumption value drugs accounted for approximately 67.4% of the budget. Implementing rational drug use by eliminating nonessential drugs and imposing a fixed budget for this category can result in substantial savings without compromising patient care. The ABC-VED with multi-criteria matrix helps improve drug availability, reduce emergency purchases, and ensure adequate inventory control, thereby reducing financial burdens.

The results indicate that in the existing drug classification technique, 76% of the total drugs were categorized as class ‘A’. However, after implementing ABC-VED with multi-criteria drug classification, the number of class ‘A’ drugs could be reduced to 67.4%, indicating improved cost efficiency and prioritization in procurement. This reduction corresponds to a decrease in annual drug consumption from 2,736,077.5 ETB to 2,421,965.75 ETB. These findings highlight the importance of employing ABC-VED with multi-criteria analysis to accurately classify drugs. When we see the.

Demand forecasting for class ‘A’ drugs

From the ABC classification, only 29 out of 72 drugs are grouped under class ‘A’ since these items are vital-few and costly it needs serious attention. So this study is focused on these drugs while forecasting the demand. In doing so, the behavior of each drug is identified first. To select the proper forecasting technique, a quantitative (time series) methods are applied. The measure of forecast accuracy the so-called mean absolute deviation (MAD) is also used in the research output²⁵. For example the best appropriate forecasting technique for drug number 1 and drug number 4 is shown in the Fig. 1 below.

Fig. 1.

Y-trend equation graph of drug number 1.

In Table 4 below represents the linear regression forecasting technique is the best for predicting drug 1 based on the minimum result of measure of forecasting error with regard to the drug behavior as compared to other forecasting techniques.

Table 4.

The values of forecasting error for each technique (for drug number 1).

Comparison of the forecasting error
Forecasting Techniques	Measure of error
	MAD	MSE	MAPE
Simple moving	1475	2,451,250	12.32
Weighted moving	1980	3,920,400	15.84
Exponential smoothing	896	1,182,300	7.59
Linear regression	400	164,500	3.73

As showed in Table 4 above linear regression forecasting technique is the best technique for forecasting drug number 1. Since, Regression analysis examines the relationship between drug demand and other variables, such as patient demographics, marketing efforts, or economic indicators. The table also highlight, compared to the other forecasting technique linear regression has Mean Absolute Deviation (MAD) value which is 400. It helps to identify the factors that significantly impact drug demand and enables the creation of predictive models. Once the appropriate forecasting technique has determined then the forecast value of each year is calculated using trend line as showed in Table 5 below.

Table 5.

Forecast value for drug number1 using linear regression technique.

Drug1
Year (X)	Demand (Y)	Predicted (Y)	Deviation	Sq. of dev.	Sum of sqr.	Abs dev.	Sum of abs	APE	Sum of APE	MAD	MSE	MAPE
2016–2017	10,100	9670	430	184,900	658,000	430	1600	4.26	14.9	400	164,500	3.7
2017–2018	10,000	10,490	−490	240,100		490		4.9
2018–2019	11,000	11,310	−310	96,100		310		2.82
2019–2020	12,500	12,130	370	136,900		370		2.96
2020–2021	-	12,950

According to the above table (Table 5), the forecast value for drug number 1 is stated to be 12,950 units. This suggests that based on the forecasting technique employed and the available data, the estimated demand or consumption for drug number 1 is projected to be 12,950 units within the specified time frame. Similarly, for drug number 4, the appropriate forecasting technique and its corresponding result are discussed in Table 6; Fig. 2 below.

Table 6.

Forecast value for drug number 4 using exponential smoothing.

Drug 4					Forecast
Year	2016–2017	2017–2018	2018–2019	2019–2020	2020–2021	MAD	MSE	MAPE
Demand	1405	1600	750	2170	-
alpha = 0.1		1405	1424.5	1357.05	1438.345	186.9	128,207	15.509
alpha = 0.2		1405	1444	1305.2	1478.16	194.9	140,838	16.064
alpha = 0.3		1405	1463.5	1249.45	1525.615	203.2	154,947	16.638
alpha = 0.4		1405	1483	1189.8	1581.88	212	170,678	17.232
alpha = 0.5		1405	1502.5	1126.25	1648.125	221.3	188,188	17.847
alpha = 0.6		1405	1522	1058.8	1725.52	230.9	207,642	18.481
alpha = 0.7		1405	1541.5	987.45	1815.235	241	229,214	19.135
alpha = 0.8		1405	1561	912.2	1918.44	262.5	253,090	19.809
alpha = 0.9		1405	1580.5	833.05	2036.305	262.5	279,466	20.503

Fig. 2.

Exponential smoothing graph of drug number 4.

The table serves as an example of drug forecasting using the simple exponential smoothing technique. It showcases the variation between the forecasted values and the measure of error for different alpha values. Alpha represents the smoothing constant, which determines the weight given to past observations in the forecasting process. The table emphasizes the predicted value of drug number 4 with the least mean absolute deviation error. Therefore the forecasted value for drug number 4 is stated to be 1438 units, with a smoothing constant (alpha) of 0.1. This means that when alpha is set to 0.1, the forecasted value of 1438 units for drug number 4 yields the minimum forecasting error.

The forecasted value of drug number 4 in the last four years and the next year with different alpha value is shown in Fig. 2 below.

The predicted values for drug number 4 are shown graphically in the figure. It illustrates how these values change depending on which alpha value is used in the forecasting process. Understanding the relationship between alpha and the resulting forecasts is made easier with the aid of the graph, which illustrates how the smoothing constant affects the predicted values.

Thus, a study is conducted using a similar methodology to ascertain the most suitable forecasting technique for the demand of 29 class ‘A’ drugs. This inquiry most likely entails examining various forecasting models or methodologies and rating their efficacy according to the least amount of error. Finding the most precise and trustworthy forecasting method to estimate the demand for these medications is the goal.

The above table (Table 7) shows the predicted values for all class ‘A’ medications for the next year. The anticipated demand for each medication during that time is shown by these values. The forecasts most likely stem from past data, patterns, the makeup of drugs, and the amount of error factors taken into account during the forecasting process. These estimated values are regarded as constants in the demand. This presumption makes it possible to determine how much medication should be ordered or restocked.

Table 7.

The forecasted demand for the next possible year.

S/N	Drug Name	Units	Appropriate Forecasting Technique	Forecasted Demand
1.	Acetylsalicylic Acid - Tablet (Microfined) − 300 mg	Tablet	linear regression	12,950
2.	Adrenaline (Epinephrine) - Injection − 0.1% in 1 ml ampoule	Ampoule	linear regression	23,228
3.	Aluminum Hydroxide + Magnesium Hydroxide + Simethicon - Oral Suspension − (225 mg + 200 mg + 50 mg)/5 ml	Bottle	exponential smoothing	1495
4.	Aminophylline - Injection − 250 mg/10 ml in 10 ml vial	Vial/ampoule	exponential smoothing	1438
5.	Ampicillin - Capsule − 500 mg	Box	linear regression	150
6.	Ampicillin Sodium - Injection − 500 mg	Ampoule	linear regression	21,748
7.	Atrovastatin - Tablet − 10 mg	Pack	exponential smoothing	5999
8.	Atrovastatin - Tablet − 40 mg	Pack	exponential smoothing	8681
9.	Beclometasone Dipropionate - Inhalation − 100mcg/dose	Bottle	exponential smoothing	5837
10.	Bisacodyl - Tablet − 5 mg	Pack	exponential smoothing	5112
11.	Castor Oil - Oral Liquid − 30 ml	Bottle	exponential smoothing	415
12.	Chlorpromazine Hydrochloride - Injection − 50 mg/ml in 2 ml Ampoule	Ampoule	exponential smoothing	412
13.	Cimetidine - Injection − 200 mg/ml in 2 ml ampoule	Ampoule	linear regression	61,075
14.	Codeine Phosphate - Linctus − 15 mg/5 ml	Bottle	exponential smoothing	19
15.	Codeine Phosphate - Tablet − 30 mg	Pack	linear regression	18,200
16.	Dextromethorphan Hydrobromide - Syrup − 15 mg/5 ml	Bottle	linear regression	3150
17.	Diazepam - Injection − 5 mg/ml in 2 ml ampoule	Ampoule	simple moving average	5447
18.	Diclofenac Sodium - Injection − 25 mg/ml in 3 ml ampoule	Ampoule	linear regression	27,320
19.	Diclofenac Sodium - Tablet − 50 mg	Box	linear regression	101,100
20.	Digoxin - Injection − 0.5 mg/ml in 2 ml Ampoule	Ampoule	simple moving average	1011
21.	Enalapril Maleate - Tablet − 10 mg	Pack	linear regression	97,100
22.	Enalapril Maleate - Tablet − 5 mg	Pack	linear regression	170,900
23.	Fentanyl - Injection − 0.05 mg/ml in 2 ml Ampoule	Ampoule	linear regression	385
24.	Frusemide - Injection − 10 mg/ml in 2 ml ampoule	Ampoule	linear regression	27,600
25.	Frusemide - Tablet − 40 mg	Tin(of 500 tab)	linear regression	1,177,550
26.	Haloperidol - Injection − 5 mg/ml in 1 ml Ampoule	Ampoule	linear regression	10,455
27.	Haloperidol - Tablet − 1.5 mg	Box(of 100)	linear regression	176,450
28.	Hydralazine - Injection − 20 mg/ml in 1 ml ampoule	Ampoule	simple moving average	7468
29.	Lidocaine Hydrochloride - Spray − 10% in 80 g	Ampoule	linear regression	65

An effective method for determining the ideal order quantity that keeps costs to a minimum while keeping sufficient stock on hand is the economic order quantity (EOQ) model. The hospital can determine the most cost-effective order quantities for each drug by taking into account variables like ordering costs, carrying costs, and demand variability and using the forecasted values as inputs.

For the hospital to have an efficient supply chain, the predicted values and the ensuing computations are essential. The hospital can minimize inventory costs and potential stock outs while guaranteeing a sufficient supply of drugs by precisely estimating the demand and determining appropriate order quantities.

Economic order quantity (EOQ) and improved inventory cost

The expense of maintaining inventory is an important factor in supply that shouldn’t be disregarded^26,27. Based on the yearly ordering cost, the pharmaceutical department has set standard values for holding (carrying) costs. 20% to 25% of the unit cost is the usual range for the average holding cost²⁸.

The classical EOQ model, while foundational in inventory theory, relies on assumptions such as constant demand, fixed costs, and the absence of shortages conditions that rarely align with the dynamic and high risk environment of hospital pharmacies. The EOQ formula applied in this study appears unmodified. The applicability of the traditional EOQ model for the selected drugs, demand and lead time were relatively stable. Even if used as a baseline, it does not account for health care specific constraints. We mentioned different approaches as a future study in conclusion section. The first interesting approach is to incorporate dynamic patient demand patterns. Another essential area is to address supply chain concerns, such as delivery delays, expiration date of drugs, disruptions caused by geopolitical events, or variations in supplier performance. Future research should look into stochastic inventory models or simulation-based approaches that incorporate these uncertainties in order to better plan for interruptions, and use sensitivity analysis to show how changes in weights affect classification.

Equation 5 illustrates the application of the Economic Order Quantity (EOQ) model in calculating the overall cost of the current inventory management system as well as the suggested enhanced system. The total cost for 29 chosen drug items categorized as class “A” is summarized in the Table 8 below.

Table 8.

Annual cost of the drug item for the existed & improved value.

S/N	Drugs Items	Cost Category
		Existed Value			Improved Value
		Carrying cost (birr)	Ordering cost (birr)	Total Cost (birr)	Holding cost (birr)	Ordering cost (birr)		Total Cost (birr)
1.	Acetylsalicylic Acid - Tablet (Microfined) − 300 mg	155.71	8820	8975.71	1277.38	1277.38		2554.76
2.	Adrenaline (Epinephrine) - Injection − 0.1% in 1 ml ampoule	6758.59	4620	11378.59	6865.82	6865.81		13731.6
3.	Aluminum Hydroxide + Magnesium Hydroxide + Simethicon - Oral Suspension − (225 mg + 200 mg + 50 mg)/5 ml	994.90	7550	8544.90	2569.39	2569.39		5138.79
4.	Aminophylline - Injection − 250 mg/10 ml in 10 ml vial	870.28	5840	6710.28	2221.45	2221.45		4442.91
5.	Ampicillin - Capsule − 500 mg	1491.72	560	2051.72	143.62	143.62		287.23
6.	Ampicillin Sodium - Injection − 500 mg	1022.57	5440	6462.57	2832.43	2832.43		5664.87
7.	Atrovastatintablet10mg	585.87	5640	6225.87	2701.56	2701.56		5403.12
8.	Atrovastatintablet40mg	1986.97	5440	7426.97	2922.27	2922.27		5844.54
9.	Beclometasone Dipropionate - Inhalation − 100mcg/dose	2347.17	8050	10397.17	6377.06	6377.06		12754.1
10.	Bisacodyl - Tablet − 5 mg	141.38	1160	1301.38	293.54	293.54		587.08
11.	Castor Oil - Oral Liquid − 30 ml	179.51	6120	6299.51	1544.99	1544.99		3089.99
12.	Chlorpromazine Hydrochloride - Injection − 50 mg/ml in 2 ml Ampoule	120.85	1120	1240.85	285.33	285.33		570.66
13.	Cimetidine - Injection − 200 mg/ml in 2 ml ampoule	4385.58	5840	10225.58	6802.89	6802.89		13605.7
14.	Codeine Phosphate - Linctus − 15 mg/5 ml	25.32	480	505.32	160.20	160.20		320.40
15.	Codeine Phosphate - Tablet − 30 mg	344.39	3930	4274.39	1794.43	1794.43		3588.85
16.	Dextromethorphan Hydrobromide - Syrup − 15 mg/5 ml	2664.62	4740	7404.62	3020.08	3020.08		6040.17
17.	Diazepam - Injection − 5 mg/ml in 2 ml ampoule	185.94	10,880	11065.94	1589.80	1589.80		3179.60
18.	Diclofenac Sodium - Injection − 25 mg/ml in 3 ml ampoule	832.72	3930	4762.72	2799.88	2799.88		5599.76
19.	Diclofenac Sodium - Tablet − 50 mg	593.27	3780	4373.27	1289.38	1289.38		2578.76
20.	Digoxin - Injection − 0.5 mg/ml in 2 ml Ampoule	423.77	9660	10083.77	3056.55	3056.55		6113.10
21.	Enalapril Maleate - Tablet − 10 mg	738.31	5040	5778.31	2618.42	2618.42		5236.84
22.	Enalapril Maleate - Tablet − 5 mg	2062.74	5040	7102.74	3103.83	3103.83		6207.66
23.	Fentanyl - Injection − 0.05 mg/ml in 2 ml Ampoule	110.10	2490	2600.10	836.07	836.07		1672.14
24.	Frusemide - Injection − 10 mg/ml in 2 ml ampoule	10320.88	5240	15560.88	3576.76	3576.76		7153.52
25.	Frusemide - Tablet − 40 mg	2313.62	9170	11483.62	7422.18	7422.18		14844.4
26.	Haloperidol - Injection − 5 mg/ml in 1 ml Ampoule	606.97	4380	4986.97	2542.43	2542.43		5084.86
27.	Haloperidol - Tablet − 1.5 mg	2950.58	5240	8190.58	4717.21	4717.21		9434.42
28.	Hydralazine - Injection − 20 mg/ml in 1 ml ampoule	11454.09	4380	15834.09	8380.94	8380.94		16761.9
29.	Lidocaine Hydrochloride - Spray − 10% in 80 g	191.91	6160	6351.91	1753.16	1753.16		3506.32
Total Cost				207,600				170,998
Saved Value					36,602

As the above table (Table 8) shows, the hospital successfully lowers its overall cost from 207,600 ETB to 170,998 ETB by putting the suggested approach into practice. This corresponds to a notable 17.63% cost reduction, resulting in significant annual savings of roughly 36,602 ETB. The study makes a number of suggestions to enhance the Ayder Referral Hospital’s medication inventory management. First, it recommends using quantitative forecasting methods in conjunction with the implementation of a reorder point and an economic order quantity (EOQ) model. By taking these steps, drug stock-outs and overstocks should be reduced, resulting in a more balanced and effective inventory level.

The analysis demonstrated significant improvements in the efficiency of the drug procurement and inventory management system following the intervention. The 17.63% cost reduction is particularly noteworthy and aligns with findings from previous studies. As⁶ reported a similar trend in cost optimization post ABC-VED implementation, with a cost reduction of approximately 18% and improved availability of essential medications. In⁶, the implementation of ABC-VED analysis in a tertiary hospital led to a 9% drop in class ‘A’ drugs, which is a comparable outcome to our 8.6% decrease (from 76% to 67.4%). Aligning inventory control with clinical priorities is crucial⁷. Their study observed a 20% increase in essential drug availability, echoing our findings of improved essential drug availability after the intervention.

These comparisons reinforce the effectiveness of integrating the ABC-VED with multi criteria analysis into hospital procurement systems. Not only does it facilitate cost containment, but it also ensure that critical and lifesaving medications are prioritized.

The study also emphasizes how crucial it is to apply multi-criteria analysis in conjunction with the ABC-VED method for the hospital pharmacy to ensure optimal resource utilization and eliminate out-of-stock situations. This all-encompassing strategy improves the EOQ model’s forecasting accuracy, which helps with efficient inventory management.

The necessity of enhancing Ayder Referral Hospital’s medication inventory management system is also emphasized in the paper. The suggested improvements are centered on maximizing medication availability and lowering inventory expenses. This entails using the EOQ model, inventory models, classification strategies, and forecasting techniques. The hospital can improve its inventory management procedures and guarantee that medications are available when needed while cutting costs by putting these strategies into practice. The limitations of this study include its conduct at a single institution, which may affect generalizability, and the lack of control over external factors such as market fluctuations and supplier changes.

Conclusion

This study focused on developing an improved drug inventory management system at Ayder Referral Hospital through the integration of ABC-VED analysis with multi-criteria analysis. The study followed a mixed-methods approach, combining quantitative and qualitative data analysis to gather comprehensive insights into the hospital’s drug inventory management practices.

The findings of the research show that the integration of ABC-VED analysis and multi-criteria analysis can significantly improve the efficiency and effectiveness of drug inventory management. The ABC-VED analysis permitted for the classification of drugs based on their consumption value and criticality for patient care. This classification provides a foundation for ranking inventory control efforts and allocating resources accordingly.

The multi-criteria analysis further improved the inventory management system by incorporating multiple factors such as cost, lead time, demand availability, and criticality. By assigning weights to each criterion and using a multi-criteria decision-making technique, drugs were ranked and prioritized based on their overall performance in the well-defined criteria. This approach enables more informed decision-making regarding drug procurement, stock levels, and replenishment policies.

The developed drug inventory management model, which integrates ABC-VED and multi-criteria analysis, offers a comprehensive framework for Ayder Referral Hospital to improve drug inventory control. The model provides procedures for setting appropriate reorder points, order quantities, and safety stock levels, taking into account both the consumption value and criticality of drugs, as well as other dynamic factors such as demand forecasting and lead time variation. The achieved solution like optimal cost, allocation, and quantity represents the best possible outcome given the constraints and assumptions through the model. The Linear Model efficiently formulates the problem and delivers the best solution within the constraints of its assumptions. While it may not account for all real-world complications, it is an effective and useful tool for preliminary analysis and decision-making. Future research could improve the model by integrating uncertainty, nonlinear dynamics, or integer limitations, depending on the situation and goals.

In general, this study were identified, several crucial issues affecting the drug inventory management system at Ayder Referral Hospital. The combination of ABC-VED with Multi-Criteria Decision Analysis creates a comprehensive, evidence based paradigm for efficient drug inventory management in healthcare settings. This hybrid model not only ensures the availability of vital and effective pharmaceuticals, but it also tackles key operational issues such as lead time, storage limits, stock out risk, and supplier reliability. Using this technique, healthcare institutions can shift from reactive inventory procedures to proactive and strategic inventory planning, resulting in better patient outcomes, fewer operational costs, and greater decision-making capabilities. The results demonstrated a reduction of 17.63% in the total inventory cost, and these improvements were successfully implemented in practice. Therefore, the application of inventory management systems in the pharmaceutical field is deemed essential for developing an optimal drug inventory system in stock, and the findings are also relevant for similar service industries.

However, it is recommended that from a management standpoint, hospitals and pharmacies should use category-based control systems. High-priority medications necessitate rigorous inventory management, regular stock assessments, and buffer stock arrangements. These items should be controlled by senior chemists or purchasing managers. To ensure consistency and accuracy, a multidisciplinary inventory committee made up of pharmacists, clinicians, procurement officers, and data analysts should be formed to evaluate inventory classifications on a regular basis and adjust procurement strategies in response to both clinical needs and operational realities.

Policymakers must incorporate the ABC-VED with multi criteria analysis framework into hospital standard operating procedures and national drug procurement rules. Aligning this methodology with critical drug lists and healthcare quality standards can promote rational drug use while also ensuring optimal resource allocation.

Finally, regular audits and performance monitoring utilizing key indicators, such as stock-out rates, inventory turnover, and drug waste, should be carried out to assess the inventory strategy’s success and inform ongoing development. Adopting the ABC-VED wit multi criteria analysis model allows healthcare organizations to better align medication procurement and management with clinical priorities, financial restrictions, and operational risks, resulting in more robust, efficient, and patient-centered pharmaceutical services. The limitation of this study is its deterministic assumptions and single-hospital data. While the combination of ABC-VED with multi criteria analysis has considerably improved the prioritization and efficiency of drug inventory management, various areas for future research remain open to improve its applicability and resilience. The first interesting approach is to incorporate dynamic patient demand patterns. Another essential area is to address supply chain concerns, such as delivery delays, expiration date of drugs, disruptions caused by geopolitical events, or variations in supplier performance. Future research should look into stochastic inventory models or simulation-based approaches that incorporate these uncertainties in order to better plan for interruptions, and use sensitivity analysis to show how changes in weights affect classification.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Assefa Beyene Ayalew: Since he is the corresponding and first author, he critically drafted and revised the article for significant intellectual content and made a significant contribution to the concept of the article.Aschale Getnet Alemu, and Alebachew Mengistu Worku: Significant contribution on data collection and interpretation of the data.

Data availability

The data that supports the findings of this study are available within the article.

Declarations

Competing interests

The authors declare no competing interests.

Ethical statement

Establishing a framework for the development and application of ABC-VED with Multi-Criteria Analysis for drug inventory management is the aim of this ethical statement. Our decision-making process has been aided by this framework, which also makes sure that our activities are consistent with moral standards and ideals.

Authors have developed and implemented ABC-VED with Multi-Criteria Analysis for drug inventory management in a way that upholds ethical considerations, environmental responsibility, patient welfare, fairness, efficiency, transparency, and accountability. We have done this by adhering to these ethical principles. While upholding the highest ethical standards, this ethical framework will direct our decision-making and help us deliver high-quality healthcare services.