Inventory management is understood here as the coordinated set of activities involved in procuring, storing, distributing, and accounting for medicines and medical supplies so that patients and clinical staff can obtain what they need, when they need it, while costs and losses are kept within acceptable bounds ([3]). Three constructs anchor the empirical analysis. Inventory shrinkage refers to the loss of inventory through causes such as expiry, damage, theft, or obsolescence, and is typically measured as the gap between recorded and physically verified stock ([1]; [16]; [18]). Inventory records accuracy refers to the degree to which recorded stock data (quantities, locations, descriptions) correspond to physical stock on hand ([1]). Staff competence level refers to the knowledge, skills, and proficiency of personnel responsible for inventory-related tasks. Organizational performance, the outcome of interest, is operationalized through indicators of service delivery that are meaningful in a hospital pharmacy context: drug availability, planning and scheduling effectiveness, and perceived contribution to patient/customer service.

In public health supply systems, the costs of poor inventory practice include shortage costs (emergency procurement, clinical delay, patient risk) and holding costs (storage, handling, obsolescence, and loss) ([12]). Where forecasts diverge from actual demand, facilities tend either toward stock-outs or toward costly excess holding, both of which erode performance ([24]). In the East African public-hospital context, [18] found that inventory shrinkage was negatively associated with organizational performance while inventory-records accuracy was positively associated with it, a pattern broadly consistent with the theoretical expectation that loss control and record integrity pull performance in opposite directions when one is weak.

Within the African public-health sector specifically, [13] documented that weak inventory-control practices for anti-malarial medicines in Ethiopian public facilities produced avoidable stock-outs and wastage, and [2] similarly found significant gaps in inventory management for maternal and child health commodities in Ethiopian public hospitals. In Zambia specifically, the World Health Organization’s assessment of the country’s multi-tiered medicine distribution system found persistent gaps between the intended Outside the health sector, [9] reported that flour milling firms in Nigeria that adopted best-practice, model-based inventory management achieved better capacity utilization, higher service levels and shorter lead times than firms relying on ad hoc, demand-based approaches. Similarly, [21] found that effective inventory control was associated with improved sales effectiveness and reduced operational costs at a Nigerian bottling company, and [22] found that economic order quantity and inventory control practices were positively and significantly correlated with the performance of steel firms in Kenya. These findings, while drawn from adjacent sectors, support the general proposition that the quality of inventory practice, not merely its existence, is associated with organizational outcomes and actual performance of the public supply chain, particularly at the point where central-level stock is translated into facility-level availability ([28]), while national audit reports have repeatedly identified stock discrepancies and losses from expired or unaccounted-for medicines in public hospitals as a recurring weakness ([20]). Taken together, the regional and Zambian evidence suggests that inventory shrinkage, records accuracy, and staff competence are plausible, complementary levers for improving hospital pharmacy performance, which is the premise this study tests empirically at Ndola Teaching Hospital.

The study draws on the Theory of Constraints, which holds that the performance of any system is limited by a small number of binding constraints at any given time, and that performance improves when those constraints are identified and systematically addressed rather than treated as fixed ([7]). [14] apply this reasoning to a public-sector pharmaceutical supply chain, showing that constraints arise at each link between the medical stores, the hospital and its clinics, and that identifying and addressing them improves the availability of supplies. Applied to hospital pharmacy inventory, the theory frames shrinkage, records inaccuracy, and staff incompetence as candidate constraints on performance: each represents a point at which the flow of usable stock to patients can be interrupted or degraded, and each is, in principle, manageable through targeted intervention. A comparable approach is taken by [10], who draw on the Theory of Constraints to examine procurement processes, ICT infrastructure, distribution channels and staff competency as factors affecting the efficiency of the medical supply chain to public health centres in Kenya.

Following [8] account of a conceptual framework as an interlinked network of constructs that together explain a phenomenon, this study conceptualizes inventory shrinkage, inventory records accuracy, and staff competence level as independent variables acting on organizational performance, the dependent variable (Figure 1). The framework guided both the design of the survey instrument and the specification of the regression model reported in Section 4.

Figure 1. Conceptual framework linking inventory management practices to organizational performance.

The study adopted a descriptive case-study design combined with a mixed-methods approach, integrating quantitative survey data with open-ended qualitative responses. A case-study design was selected because it allows an in-depth, contextually grounded account of inventory management practice at a single, purposively selected institution, rather than a shallower comparison across many facilities ([4]; [15]), and descriptive research of this kind is well suited to capturing the present state of a phenomenon.

Ndola Teaching Hospital employs approximately 500 healthcare workers across clinical and non-clinical functions. The target population for this study, however, was the subset of staff directly involved in pharmaceutical inventory management: pharmacists, pharmacy technologists, and supply-chain/logistics personnel, comprising a total of 96 eligible staff. Given the manageable size of this target group relative to the wider hospital workforce, the study adopted a census approach, inviting all 96 eligible staff to participate rather than drawing a probability sample from the larger population of 500. All 96 questionnaires were completed and returned, yielding a 100% response rate within the target population. This census strategy, rather than a probability sample, is the primary reason the achieved sample is smaller than the hospital’s total workforce: it reflects the size of the eligible population, not a sampling shortfall, though single-site, cross-sectional data of this kind still limit generalization beyond Ndola Teaching Hospital (see Section 6).

Primary data were collected using a structured questionnaire administered via Google Forms over a two-week window, combining closed five-point Likert-scale items (1 = strongly disagree to 5 = strongly agree) with open-ended questions.

Quantitative data were analyzed in SPSS (version 20). Descriptive statistics (frequencies and percentages) were used to summarize demographic characteristics and Likert-scale responses. The Shapiro-Wilk test and visual inspection of Q-Q plots indicated that the outcome variable departed from normality; logarithmic, square-root, and reciprocal transformations did not materially improve normality. Given this, and consistent with common practice for moderately sized survey samples, Pearson correlation and ordinary least squares (OLS) regression were retained as the primary inferential tools, with results interpreted cautiously in light of the non-normality. Pearson’s correlation coefficient was used to assess the strength and direction of association between organizational performance and each of the three independent variables, and a multiple linear regression was estimated with organizational performance as the dependent variable. Open-ended responses were analyzed thematically.

The study received ethical clearance from the Ethics Review Committee of the University of Lusaka. Participation was voluntary and based on informed consent; respondent identities were not recorded, and all data were held confidentially and reported only in aggregate form.

Table 1 summarizes the demographic profile of the 96 respondents. The sample was slightly female-majority (54.0%), concentrated in the 30 - 39 age band (59.4%) and the pharmacist and pharmacy-technologist professions (91.3% combined), and reasonably experienced, with 58.0% reporting five or more years of service at the hospital.

Table 1. Demographic characteristics of respondents (N = 96).

Table 2summarizes respondents’ agreement with statements on inventory shrinkage, records accuracy, competence, and organizational performance. The strongest agreement, at or above 90%, concerned the effect of long procurement procedures on performance (97.1%), up-to-date records on accuracy (93.2%), and proper inventory management on customer service (92.75%). Among the drivers of shrinkage, damage (82.6%) and obsolescence (81.2%) were seen as considerably more consequential than theft (59.4%). Competence in pharmaceutical inventory management was rated moderately high or high by all respondents (100%), and on-the-job training was the most strongly preferred development modality (92.75%).

Table 2. Summary of perceived inventory management practices, by construct.

Open-ended responses converged on two dominant themes: about two-fifths of respondents (40.6%) called for more skilled, adequately staffed inventory teams and continued on-site training, and close to half (49.3%) called for faster, more capable inventory software. Smaller numbers raised regular stock assessments, closer ward-level monitoring, and openness to innovation. These themes reinforce the quantitative pattern: staff capability and systems, more than shrinkage control alone, dominated respondents’ own framing of what would improve inventory management at the hospital.

Table 3 presents the Pearson correlation matrix. Organizational performance was weakly and non-significantly, negatively correlated with inventory shrinkage (r = −0.0264). It was positively correlated with competence level (r = 0.2921) and, more strongly, with inventory records accuracy (r = 0.3728). Among the independent variables themselves, correlations were weak throughout (r = 0.0015 to 0.1577), consistent with shrinkage, records accuracy, and competence being related but conceptually distinct practices.

Table 3.Pearson correlation matrix.

A multiple linear regression of organizational performance on inventory shrinkage, competence level, and inventory records accuracy was statistically significant overall, F(3, 92) = 5.31, p = 0.0025, explaining approximately 19.7% of the variance in organizational performance (R² = 0.1968; adjusted R² = 0.1597; RMSE = 10.184) (Table 4). Inventory shrinkage was not a statistically significant predictor (b = −0.0356, p = 0.689), whereas both competence level (b = 0.5632, p = 0.038) and inventory records accuracy (b = 0.5002, p = 0.004) were statistically significant, positive predictors of performance (Table 5): each one-unit improvement in competence level and in records accuracy was associated with roughly half a unit of predicted improvement in organizational performance.

Table 4.Regression model summary.

Table 5.Regression coefficients (dependent variable: organizational performance).