The path to performance of public versus private healthcare supply chains: case study of Uganda

Improving performance of healthcare supply chains (HCSCs) in low- and middle-income countries (LMICs) remains a priority for national governments and international aid organizations (Ministry of Health, Uganda, 2015; The Global Fund, 2017). The World Health Organization (WHO) estimated that by 2018, 2 billion people [would] have no access to basic medicines, causing a cascade of preventable misery (Chan, 2018:14). The limited availability of medicines disproportionately affects especially LMICs (Olalekan, et al., 2024; Wirtz, et al., 2016).

Despite the considerable investments made to improve HCSCs in LMICs, improvement interventions were not effective resulting in continued low performance (The Global Fund, 2017). Extant studies on LMICs highlight that lack of accurate data on delivery, distribution and storage affects performance of HCSCs (Abdulkadir et al., 2024; Miko and Abbas, 2024) worsened especially by pilferage (Abdulkadir et al., 2024). Mandal and Jha (2018) and later Bak et al. (2023), underlined that HCSCs performance were impacted by the level of collaboration existent in HCSCs. In the case of Uganda, collaboration across HCSCs actors still poses its challenges (Mutebi et al., 2020) including the generic collaboration framework (Namagembe, 2020) and absence of mechanisms to ensure collaboration (Mutebi et al., 2020).

The long-time debate on improving performance of HCSCs in LMICs has been the role of public versus private sector with a view to benchmark across the sectors (Cameron et al., 2009). This approach might have succeeded with the collaborative understanding of performance improvement that Kraiselburd and Yadav (2013) and Stulens et al. (2021) identified as critical for HCSCs in LMIC. On the other hand, Tukamuhabwa et al. (2017) emphasized small-scale, chronic disruptive events, rather than discrete, large-scale catastrophic events as typically a threat to Uganda HCSCs’ performance. Hence, unpacking performance will allow supply chain (SC) collaboration, fine-tuning interventions and hence increasing their effectiveness (Mutebi et al., 2020).

Therefore, this paper aims to develop a framework for SC collaboration in the context of public versus private HCSCs by exploring the following research questions:

Answers to these questions will help to navigate through fragmented collaboration practices, that stand in the way of improving performance of HCSCs in LMICs (Friday et al., 2021).

One of the main healthcare challenges faced by LMICs is hemorrhage, contributing upto 28.4% of maternal deaths in sub-Saharan Africa (Cresswell, et al., 2025). The main medicines used for prevention and treatment of hemorrhage is oxytocin and misoprostol (WHO, 2012), which this study focused on, yet they are often out of stock and expired in Uganda (Ministry of Health, Uganda, 2017a).

Following this introduction, this paper presents a literature review on collaboration and HCSC performance, a methodology outlining an embedded case study in Uganda, followed by findings, analysis, discussion and conclusions.

The definition of performance measures can be a starting point for SC collaboration (Bak et al., 2023; Ralston et al., 2017; Stulens et al., 2021), however, identification of performance measures in the healthcare sector remains yet a challenge (Ramanathan and Ramanathan, 2011) as performance measures may differ across different levels and different contexts of HCSCs (Bak et al., 2023). Therefore, performance measures need to be specified at different levels in the SC (Simchi-Levi et al., 2009).

Stockout and stock expiration are commonly used performance indicators for HCSCs in LMICs (The Global Fund, 2017; Kumar and Das, 2023). Whereby, stockout occurs when there is no medicine left in the health facility (HF) store for a minimum period of one day (Ministry of Health, Uganda, 2017b). Stock expiration occurs beyond the expiry date, that is:

Thus, decrease in stock expiration is an important element of SC performance and can motivate enhanced collaboration (Kwon and Kim, 2018).

In comparison, stock accumulation is greatly influenced by ordering practices (Mohamed and Huynh, 2023) where more frequent ordering allows to keep stocks within set maximum levels above which stock is more likely to expire before use (Simchi-Levi et al., 2009). Hence, stock accumulation provides an alert, signaling its risk for future stock expiration in HCSCs. On the other hand, health service providers seek […] reliable uninterrupted supplies (Kwon and Kim, 2018:143), whereby LMICs rely highly on global HCSCs for medicines (Onyango, 2024; Stulens et al., 2021). Disruption in flow of medicines can be a major challenge for the LMICs (Simchi-Levi et al., 2009; Stulens et al., 2021), for example during COVID-19 pandemic (Arji et al., 2023; Olalekan, et al., 2024; and Onyango, 2024). Disruptions may be due to insufficient financing (Kraiselburd and Yadav, 2013; Privett and Gonsalvez, 2014; Yadav, 2015), and complexity of multi-level HCSCs within the country particularly the challenge of last mile (Jahre, et al., 2012; Yadav, 2015). Hence, supply chain (SC) flow – especially of information, funds and products can provide an alert, to signal the risk of future stockouts or stock expiration in HCSCs.

Collaboration in HCSCs lags behind other sectors due to the investments required and the challenges associated with assessing benefits for its SC members (Friday, et al., 2021), yet improved collaboration would overcome uncertainties inherent in complex HCSCs (Stulens et al., 2021). To assess SC collaboration; strategy, structure, systems and processes have been utilized (Rodrigues et al., 2004; Romule et al., 2020). Strategy is defined as the alternate ways in which organizations define their product-market domains(Miles et al., 1978:546). In the context of collaboration, this involves the alignment of each player’s business goals with those of the collaboration (Angerhofer and Angelides, 2006). Structure refers to “patterned regularity that is reproduced as its members interact and communicate about their interactions to coordinate actions and sustain the state of being organized” (Wagner, 2025:1854). Systems are defined as complex entities with clearly delineated boundaries, comprising interconnected components that interact with one another (Alvargonzález, 2019). Process on the other hand is defined as “a set of logically related tasks performed to achieve a defined business outcome” (Vollmann, 1996, p. 60).

Collaboration in supply chains involve inter-organizational cooperation and coordination (Gulati et al., 2012) where two or more autonomous SC members jointly plan and execute operations (Cao and Zhang, 2011). However, for HCSCs in LMICs cooperation is underdeveloped as evidenced by the absence of or non-commitment to strategic plans and/or goals (The Global Fund, 2017; Yadav, 2015). According to Mandal and Jha (2018:2668) the benefits of SC collaboration can be realized fully when every key SC entity is ready to cooperate. Cooperation happens when all SC members agree on shared goals and benefits, as well as their responsibilities (Ralston et al., 2017). Therefore, shared goals (Liu et al., 2022) and benefits (Mandal and Jha, 2018) are strategies for SC collaboration impacting resource allocation in terms of finances and budgeting and SC performance (Romule et al., 2020). In a resource constrained setting, cost sharing is a key responsibility of collaborators and according to the Ministry of Finance, Planning, and Economic Development, Uganda (2024), a key government strategy is to allocate a dedicated budget to the central medical stores for all public HFs. This practice is called “contribution of resources.”

Therefore, this paper explores the three extant strategies for collaboration applicable to HCSCs in LMICs – shared goals, contribution of resources and shared benefits.

The overall performance of HCSCs is enabled through the coordination of [several] business functions (Vijayasarathy, 2010: 492). However, a key weakness of HCSCs in LMICs is absence and inefficiency of coordination structures yet stakeholders hold divergent objectives (Yadav, 2015). This is exemplified by The Global Fund (2017) report that in most countries, governance structures are ineffective at resolving supply chain interventions. Lack of coordination leads to incomplete demand information and lack of shipment visibility, resulting in excessive inventory and waste, due to stock expiration (Privett and Gonsalvez, 2014).

This gap highlights the necessity for formal mechanisms to foster SC collaboration. Extant literature suggests four foundational structures: Supply chain relationships – the broad variety of contractual arrangements that form the basis for collaboration between independent organizations (Kraiselburd and Yadav, 2013); Coordinating institutions – dedicated groups or individuals tasked with extemporaneously troubleshooting and solving emerging challenges (Gulati et al., 2012); Coordinating committees – formal committees vital for engaging multiple stakeholders in long-term agendas (Kraiselburd and Yadav, 2013); and government guidelines – the formal and informal rules and policies set by government to govern supply chain operations, identified as necessary to provide a framework for collaboration (Gulati et al., 2012; Mandal and Jha, 2018; Chen et al., 2020).

The utilization of information systems and technology to facilitate data capture, analysis and information sharing has been demonstrated to enhance SC performance (Angerhofer and Angelides, 2006; Simchi-Levi et al., 2009). Extant studies have highlighted that there is paucity of accurate data on delivery, distribution and storage (Abdulkadir et al., 2024). Furthermore, deficiencies in SC performance measurement compromise the performance of HCSCs in LMICs (Kraiselburd and Yadav, 2013; Miko and Abbas, 2024; Privett and Gonsalvez, 2014; and Yadav, 2015). The creation of an integrated performance measurement system to monitor SC performance has been proven vital in the healthcare context in Morocco, a low-middle income country (Chorfi et al., 2018).

Based on the foregoing, this paper explores information systems including information technology and information sharing, as well as SC performance measurement as the systems for healthcare SC collaboration in LMICs.

Organizations are integrated into a SC through inter-organizational processes (Lambert et al., 2005), a key ingredient for SC integration (Flynn et al., 2010). The first process framework proposed for SC collaboration is the Supply Chain Operations Reference (SCOR) model (Simchi-Levi et al., 2009). The SCOR model provides a good benchmark for industries to compare and assess their SC processes including Plan, Source, Deliver and Return (Lockamy and McCormack, 2004).

In the context of HCSCs, collaboration processes between key stakeholders have been shown to mitigate stockouts (Friday et al., 2021). In relation to planning, Abdulkadir et al. (2024) contended that demand-driven decision-making informed by real data would enhance the provision of medicines to patients in Nigeria. Regarding sourcing, Kessy et al. (2024) argued that fostering close interactions between suppliers and HFs would engender trust in Tanzania. Regarding delivery, the study by Kessy et al. (2024) revealed that the main difference between public and private medical commodities supply chains is in the extent to which the different steps in the process are standardized. Finally, regarding return processes, Ribeiro et al. (2021) highlighted the absence of standardized processes for reverse flows of hospitals in Brazil and the associated negative financial and environmental consequences. The interdependent nature of the four processes is identified as the determining factor for SC performance (Angerhofer and Angelides, 2006).

Therefore, given its demonstrated applicability in HCSCs in LMIC, this study adopted the dimensions of the SCOR model to deepen the process dimension of collaboration.

Overall, this review of literature reveals that a holistic assessment of all collaboration (sub-) dimensions is underdeveloped in the extant literature, and even more so in the context of HCSCs in LMICs. Our empirical study is grounded upon such holistic approach as set out in Table 1.

Themes from literature review were fine-tuned through theory-data iterations to enable consideration of all potential explanations of low performance of HCSCs (Eisenhardt, 1989). Figure 1 shows the research framework that guided data gathering, analyses and presentation of results.

This research is grounded upon an embedded multi-case study design to explore a contemporary phenomenon within its real-life context; and answering “why” and “how” questions (Yin, 2014), in this case the functioning of HCSCs. The study’s unit of analysis is the supply chain. Interviewees, acting as organizational keepers, provided information focused on their organizations’ roles within the supply chain, with data collected from diverse SC members. The multi-case study design enables theoretical replication and fosters in-depth understanding of the HCSCs, by generating findings specific to sub-cases, permitting within- and across-case comparisons across different HCSC layers and types. This approach empirically grounds findings and maximizes construct validity through diverse inquiry and multiple sources of evidence (Eisenhardt, 1989).

According to Ministry of Finance, Planning, and Economic Development, Uganda (2024) and Uganda Bureau of Statistics (2024), Uganda is a low-income country with a government per capita medical supply budget of USD4.01 – far below the USD13-25 estimated as the minimum requirement for LMICs (Wirtz, et al., 2016). Expiration of medicines albeit the constrained budget and recurring stockouts indicates systemic medicine SC challenges. Therefore, Uganda serves as a typical case (Yin, 2014) due to its pronounced budgetary constraints and ineffective HCSCs, challenges experienced by many LMICs (The Global Fund, 2017). By examining Uganda’s context, this study provides insights into systemic challenges in HCSCs that are transferable to other LMICs. While contextual differences may limit direct generalizability, the findings offer a framework applicable across LMICs.

In Uganda, the MOH is responsible for policy direction, while National Drug Authority is responsible for regulation. National Medical Stores (NMS) is the lead supplier to public HFs (Ministry of Health, Uganda, 1993), whereas Joint Medical Store (JMS) and several other private firms are engaged in the supply, manufacture, warehousing, distribution use and dispensing of pharmaceuticals. Uganda imports over 90% of its medicines (Rugumambaju and Kutyabami, 2010). Health Facilities owned by the Government (public) or privately can be classified into low level (HCI to HCIII) and high level (HCIV, General Hospital, Regional Referral Hospital and National Referral Hospital) (Ministry of Health, Uganda, 2016). Figure 2 provides an illustration of the HCSCs studied.

Cases were selected following replication logic to provide examples of polar types (Eisenhardt, 1989, p. 537). To appreciate the variation of context along the SC, this study compared public and private HCSCs as well as high-level vs low-level HFs.

Data collected from documentation, archival records and 39 interviews informed the study. Data on stockout and expiration of medicines were extracted from bimonthly MOH stock status reports and annual Auditor General’s reports to rate supply chain performance. Interviewees suggested informative documents (reports and strategic plans) which were added. Face-to-face interviews in English followed an interview guide (see  Appendix 1) and were audio recorded and transcribed verbatim. Interviewees included officers at lead supply agencies (NMS and JMS), MOH, development agencies, districts and HFs (see  Appendix 2).

Documents and interview transcripts were uploaded into NVivo and coded thematically. Following theory-data iterations (Eisenhardt, 1989), themes included strategies for SC collaboration, structures for SC collaboration, systems for SC collaboration and SC collaboration processes as well as the performance measures of SC flow, stockout, stock accumulation and stock expiration (see  Appendix 3).

Drawing on Yin (2014), this study adopted several strategies to enhance validity and reliability. Construct validity was ensured using multiple data sources. Internal and external validity were enhanced by utilizing an embedded multiple-case study design, employing replication logic and limiting generalization to theory and context (Yin, 2014). Finally, a case study protocol ensured reliability, per Yin (2014)’s recommendations.

Reliability of content analysis was assessed for stability and reproducibility using intercoder reliability (ICR), the extent to which two or more coders agree on the coding of a given text (Campbell et al., 2013). Cohen’s kappa (Cohen, 1960) Cohen (1960) was used to examine ICR (MacPhail et al., 2016). Consensus was built on the coding approach through two rounds of independent coding of same text by the three investigators and joint assessment of ICR by the team (MacPhail et al., 2016).

Antecedent variables, assessed for presence on a 4-point Likert scale (1=absent, 4=very high), intentionally excluded a neutral option to force determination (Chyung et al., 2017). Where applicable, performance was similarly rated (1=very poor, 4=very good). The first author assigned Likert scale scores which were reviewed by the coauthors. Potential discrepancies were debated by the author team and led to updating of some scores. Association was then derived from absolute score difference between presence of dimensions and performance rating (0–3): (0 strongly associated), (1 moderately), (2 weakly), (3 not associated). Antecedent variables exhibiting an absolute score difference of two or less were deemed to be impactful on SC performance.

This section details performance dimensions. Stockout and stock expiration were widely recognized SC performance measures in the cases. Stockout was a commonly used SC performance measure and a major challenge for public HFs. [W]e have had stockouts of especially oxytocin and it really affected us very much. […] It was from November of last year to around February of this year (KII Public HCIV). In addition, national stock status reports indicated frequent stockouts in public and low-level HFs compared to private and high-level ones, respectively. Structured data on stock expiration (location, quantity, and value) were scarce, despite self-reported incidences. Unlike stockout, expiration data were disaggregated only by HF level, yet KIIs revealed that expiration was rare in the private sector; we have not had any expirations (KII Private Hospital), another interviewee noting [n]o, those ones do not expire (KII Private HCIII).

On the other hand, stock accumulation and SC flow were less recognized SC performance measures. On SC flow, the Auditor General’s report of February 2021 identified disruptions in flow of funds intended to finance medicines for public HFs. This was corroborated by a national officer who claimed that [w]e don’t have visibility into how much money NMS was allocated to procure a specific product. Yet interviews could not pinpoint where and how disruptions occurred. Stock accumulation was only noted in extreme cases for example, misoprostol we always have much even that last week I gave some to the municipal council. (KII Public Hospital) and not always a challenge under the public HCSC because it mitigated against stockouts, as an interviewee suggested [i]t (stock accumulation) cannot be because it is always shared with other facilities. (KII District Officer). In the private HFs stock accumulation was sometimes deliberate to prevent stockouts. We order a large quantity of oxytocin, so it is always there (KII Private Hospital). Table 2 shows the applicability of SC performance dimensions in public versus private HCSCs.

Therefore, the performance of HCSCs in Uganda was measured by two final SC performance outcomes 1) stockout and 2) stock expiration. In contrast, the other SC performance indicators, namely 1) SC flow and 2) stock accumulation were not yet widely applied.

In this section, we provide empirical evidence of each dimension of SC collaboration. Table 3 presents the degree to which dimensions of SC collaboration were present in public versus private HCSCs, whereas Table 4 shows the resultant framework.

Despite presence of a national pharmaceutical sector strategic plan (Ministry of Health, Uganda, 2015), it did not result in shared goals among SC members as noted by the district official [w]e just show them the dangers and we show them how to utilize that outcome, but it is not in the district plan (KII District Official). Similarly, the private sector was not keen on strategic plans noting that [i]t (strategic plan) is there with the administration (KII Private Hosp).

Regarding the contribution of resources, public HCSC funds were paid directly to the public lead supply agency, NMS (Ministry of Finance, Planning, and Economic Development, Uganda, 2024). In contrast, private HFs mainly relied on facility generated user fees without first contributing the funds to a pool; on the side of what (they) sell there is no challenge, on the side of the free commodities there is not any indication of source of funding– (KII National, JMS).

On shared benefits, none accrued from appropriate performance in public HFs because (t)he incentive system is not given according to performance like avoiding stockouts. […] the incentives are about you reporting to the work for the 30 days (KII National). In contrast, there was a strong shared financial incentive for good performance in the private HCSCs which relied on user fees. Consequently, the private lead SC agency, JMS ensured that medicines were always available; [f]or the one we sell, we do not face any challenge because for us we do cost recovery. (KII National, JMS).

Based on the findings of cross-case analyses shared benefits were strongly associated with SC performance in both public and private healthcare settings. On the other hand, shared goals were only strongly associated with performance in public HCSCs, whereas the contribution of resources was not associated with performance in either HCSC (see  Appendix 4).

Supply chain relationships in public HCSCs were not favorable but driven by mandate of the lead supply agency (Ministry of Health, Uganda, 1993), and communication was only out of necessity. [W]ith medical stores […], there is nothing much apart from the deliveries. […] the only time they communicate is when the order has delayed (KII Public HCIII). In contrast, SC relationships were more favorable in the private SC which relied on formal contracts and long-term supply agreements. [W]e have a very good relationship with them because they can give this institution a credit […] we always tell them what we need, and their management collaborates (KII Private Hospital).

Coordinating institutions and committees in both public and private HCSCs were underdeveloped, lacking dedicated technical staff for medicines. Supply chain management was an additional, on-the-job role. Consequently, an interviewee noted that [c]oordination is a challenge when you go down to the districts; they don’t have someone whose main obligation is logistics (KII National, MOH). Coordinating committees were undermined by duplication and nonattendance. Besides, the official Technical Working Group (Ministry of Health, Uganda, 2015) was another Commodity Security Group which was also undermined by nonattendance. [I]n the meeting we invite over 70 people but 30 regularly attend (KII National, QPPU). The private lead supply agency, JMS concurred that [t]he RH meetings are so many […], so some of those might not yield much because some stakeholders may not be involved. (KII National, JMS).

At HF level, many public high-level HFs lacked functional Medicines and Therapeutics Committees (MTCs) to ensure that allocated budgets are used to procure medicines, with an interviewee stating that there was no functional MTCs (KII Public Hospital). Low level public HFs received a predetermined kit and thus had no MTCs. Similarly, private HFs either lacked functional MTCs or they were on and off. [W]e have a therapeutic committee, but it has spells of activity and spells of non-activity especially during the Covid period (KII Private Hospital).

Public HFs showed lower adherence to government guidelines, such as those for dispensing misoprostol. This led to inconsistent use of alternative medicines (misoprostol versus oxytocin) and fueled misconceptions about misoprostol’s use in illegal abortions. There is an abuse of misoprostol by schoolgirls which affects the availability of the drug (KII District Official). This undermined demand forecasting in the public HFs. Similarly, the logistics management guidelines (Ministry of Health, Uganda, 2017c) were not adhered to under the public HCSC. For instance, the guidelines on accepting donations were disregarded, which resulted in their expiration:

Further, the range of services provided by public lower-level facilities were not in tandem with medicines supplied They just tend to refer even simple cases which can be managed at those health centers, so we end up serving an overwhelming number of patients– KII Public Hospital.

Cross-case analyses showed that formal SC relationships and adherence to government guidelines were strongly associated with performance in both public and private HCSCs. However, coordinating institutions and committees were only strongly associated with performance in public HCSCs (see  Appendix 5).

The paucity of information sharing between HFs and the lead SC agencies was attributable to technological limitations in both public and private HCSCs. Information technology initiatives in the public HCSCs were dependent on stand-alone donor-funded electronic Logistics Management Information Systems (eLMIS) tools. These initiatives encountered difficulties with integration due to the presence of varying systems, inadequate infrastructure and high internet costs:

Consequently, stock levels were not monitored at the central level. [I] want to know what exactly is happening at facility level, (KII National, QPPU).

On the other hand, several private hospitals used Enterprise Resource Planning (ERP) Systems enabling visibility. [T]hey can be seeing the patients who are coming to the pharmacy in the system (KII Private Hospital). Surprisingly, private high-level HFs which did not have ERP systems still used paper records effectively:

In addition to technological limitations, information sharing was not seen as a common practice in both public and private HCSCs. For instance, members of staff at one private high-level HF were uncertain as to why they should share information with suppliers; for the medicine they do not have any hand in contributing for us, why should we give them a report?. Similarly, lead agency stock status reports were irregular with one interviewee claiming that [n]ational level data is not readily available. […] You have to wait for somebody to send you a mail once in a while (KII National Officer). Further, the reports lacked information on allocation of funds, status of procurement and inventory pipeline information as stated by one interviewee that [w]e don’t have visibility into how much money NMS has allocated to procure a specific product. Neither do we have information on how much is going to be delivered and when. (KII National Officer).

There was minimal alignment behind joint KPIs for public and private HFs. The National Pharmaceutical Sector Strategic Plan (NPSSP) included KPIs agreed among SC members (Ministry of Health, Uganda, 2015). However, the Supervision Performance Assessment and Recognition Strategy (SPARS) was popular among public and private HFs for SC performance measurement but used a different set of KPIs (Ladwar et al., 2021). Further, the MOH annual health sector performance report used three KPIs different from those in SPARS and the NPSSP (Ministry of Health, Uganda, 2017b) and perceived as externally imposed by the lead SC agency; I don’t see much (value) because, for instance when you calculate their order fulfillment rate, it’s like for the wrong consumption but it means much to them because it can maybe help them improve their performance. But for us here, I don’t think so (KII Public Hospital). Consequently, joint SC performance review (Ministry of Health, Uganda, 2020) was ineffective because the indicators used were not the same as those in the NPSSP. An interviewee summarized the situation as … we spend time developing a plan and we don’t have enough time to see if we are on track (KII National Officer).

Cross-case analyses showed that Information Technology was strongly associated with performance in both public and private HCSCs. Conversely, information sharing, joint KPIs and joint performance monitoring were strongly associated with SC performance only in public but not private healthcare settings (see  Appendix 6).

Guided by the SCOR model, this section shows how the SC collaboration processes influenced performance of public versus private HCSCs in Uganda.

Demand forecasting was inadequate both in public and private HCSCs. Particularly, public HFs faced fixed budget allocation and lacked an evidence-based approach; [t]hey really didn’t use a scientific approach to do their forecasting, so it was more based on estimation and haggling here my department needs this (KII Public Hospital). Whereas low-level public HFs with predetermined medicines kits were more affected [u]sually, the cheapest means is to endure with whatever it is, if patients can buy those who are severely sick, they buy until the next supply (KII Public HCIII). In parallel, private HFs suffered bullwhip effect; […] you never know when they have money to come and procure. I have no guarantee that they will come (KII Private Warehouse JMS).

Interestingly, the internal procurement processes were effective in both public and private HCSCs under the lead agencies (NMS and JMS, respectively) an interviewee noted:

In this context the warehousing was found adequate under both public and private HCSCs, except for private cold chain capacity being limited. Hence, when JMS experienced oxytocin shortages, alternative suppliers often lacked stock due to limited cold chain storage capacity. An interviewee stated that:

Nevertheless, private HFs still obtained oxytocin from alternative sources. There is also a pharmacy in town which has a fridge. We can always get the medicine from there in case of any shortages (KII Private Hospital).

On inventory management, private HCSCs were relatively better; [w]e check on the stock levels monthly […] we do physical counting and also do weekly checkups depending on the consumption.– KII private HCIII. In contrast, the public HCSC delivered short shelf medicines suggesting suboptimal inventory management; It (oxytocin) was delivered by then it had three months shelf life, and the quantity was even more than what we needed. It was pushed (KII Public Hospital). Further, public HCSCs suffered pilferage:

Regarding order processing, public low-level HFs received predetermined kits decided at the beginning of each planning cycle but perceived by staff as a push logistics system; [s]ometimes we get more than what we need and sometimes we even have to give to some other health units (KII Public HCIII). Similarly, public high-level HFs received unsolicited donated medicines; [i]t was a donation; it came in a lot […] they just decided to divide this among the seventeen regional referrals. We didn’t order, they just received them, so they were pushed (KII Public Hosp). In contrast, deliveries to the private HFs were in line with orders placed for both high- and low-level private HFs.

Considering, last mile distribution, it was well conducted in both cases, public and private HCSCs using Third-Party Logistics (3PL) without noticeable delays; […] the one that they gave the contract is working well, they are ok, so far there is no challenge (KII Public District).

Redistribution being common in public HCSCs enabled low-level HFs (HCII) to provide maternal care whereas higher level HFs were inaccessible, yet human resource capacity existed at low-level HFs (HCII); [to] avoid any shortages, HCIIIs should factor in supplies of HCIIs […] so that when they receive, they can also supply enough supplies to the HCIIs (KII Public DHO). On the other hand, redistribution was rare in the private HFs:

Cross-case analyses on the SCOR processes revealed varied associations. Source (warehousing & inventory management) and product delivery (order processing) were strongly associated with performance in both public and private HCSCs. Planning (demand forecast and supply plan) was strongly associated with performance of high-level public and private HFs. Conversely, source (procurement) and product delivery (last mile distribution) were strongly associated with performance of only the private HCSC, whereas product return (redistribution) was not strongly associated with the performance of any of the HCSCs (see  Appendix 7).

The above findings, along the different dimensions of collaboration are summarized in the proposed framework for SC collaboration in LMIC public and private HCSCs (see Table 4). The integrative and holistic nature of our framework, associating precise facets of collaboration with SC performance allows to counter fragmented collaboration practices.

The study findings indicated that monitoring SC flow and stock accumulation may result in alerts regarding the more tracked SC performance – indicators stockout and stock expiration. However, as the study findings highlighted, this practice remains rather underdeveloped in both public and private HCSCs. Similarly, recent academic discussions highlighted the importance of monitoring external environment data to facilitate adaptation to unanticipated circumstances, such as damaged transportation infrastructures or custom-related challenges (Bak et al., 2023: Polater, 2021). Whilst extant literature has previously indicated that stockouts and stock expiration are more prevalent in public compared to the private HFs in LMICs (Olaniran et al., 2022; Mohan et al., 2024), the current study indicates that these occur more frequently at the low-level HFs which are typically the first point of contact. The study also reveals that an increased presence of shared benefits, SC relationships, adherence to government guidelines, information technology, inventory management and order processing is associated with higher performance in both public and private HCSCs in Uganda, a LMIC.

Tracking stockout and expiration in both public and private HCSCs is consistent with Rodrigues et al. (2004) who recommended tracking stockout at the final consumption point as well as Rodrigues et al. (2004) and Supply-Chain Council (2012) who recommended tracking costs of the SC (including expiration). More frequent stockout and expiration of medicines in the public HCSCs is consistent with Cameron et al (2009) as cited by Kraiselburd and Yadav (2013) who found superior supply chain performance in private compared to the public HCSCs. In the case of Uganda, Namyalo et al. (2025) iterated the challenge of stockouts in public HFs that provided free services. Yet, this study shows that there are differences in the degree of associations between dimensions of collaboration and performance across public and private HCSCs, which is consistent with contingency theory, i.e. the best way to organize depends on the environment to which the organization relates (Scott and Davis, 2015p 127), and Yadav (2015) that management of HCSCs in LMICs is context specific.

According to the study by Polater (2021), the most important capability for identifying solutions to problems in humanitarian SCs is “being collaborative”. The present study has enabled more precise determination of the relationships between different dimensions of collaboration and SC performance. The findings indicated that the presence of shared goals was a catalyst for collaboration, thereby corroborating the findings of Angerhofer and Angelides (2006) and Ralston et al. (2017). The present findings concur with the conclusions of The Global Fund (2017), which emphasized that ineffectiveness of country ownership and governance in LMICs is frequently attributable to an absence of strategic frameworks. This phenomenon results in interventions that are driven by ad-hoc donor requirements rather than coherent national strategies, and provides much needed evidence of the impact of shared goals impact on healthcare improvement in LMICs (Kalaris, 2024).

The study also reveals that public HCSCs, in contrast to the private ones, exhibit an absence of shared benefits for high performance and demonstrate a weak correlation between employee benefits and performance. Nevertheless, the efficacy of last-mile distribution in both public and private HCSCs can be attributed to the utilization of outsourcing. This finding is consistent with the conclusions of Yadav (2015), who determined that a paucity of incentives results in substandard performance in public HCSCs. Finally, the current study reveals that the strategy of contribution of resources exhibits no association with SC performance. Thus, for public HCSCs in LMICs, shared goals and benefits are probable more efficacious strategies to enhance SC performance in comparison to the present strategy of contribution of resources.

Regarding structures for SC collaboration, the current study reveals that overreliance on mandates instead of contracts in public HCSCs weakens SC relationships. This corroborates extant literature that overlapping mandates create ambiguity in responsibility for tasks, which fractures relationships and undermines performance (Yadav, 2015) – a problem exacerbated by the ineffective governance structures that hinder accountability (The Global Fund, 2017). Replacing this model with defined contractual obligations would clarify roles, reduce the need for complex oversight and ultimately strengthen supply chain relationships.

Consistent with Gulati et al. (2012), the current study suggests that SC coordination may be achieved through formal contracts and informal operational rules. The former is consistent with Holt and Ghobadian (2009) who recognized that clear guidelines could lead to common goals within the SC, ultimately boosting performance. The latter is consistent with Friday et al. (2021) who emphasized the importance of alignment of uncertainty and risk perceptions of the different decision-takers on inventory levels along HCSCs.

Regarding systems for SC collaboration, strong associations have been found between the adoption of information technology and performance in both public and private HCSCs, which is consistent with the findings of Angerhofer and Angelides (2006), Miko and Abbas (2024) and Arji et al. (2023). The adoption of emerging technologies such as artificial intelligence, blockchain, drone technology and the Internet of Things has revolutionized many industries and has the potential to enhance the efficiency and responsiveness of complex health supply chains (Singh, 2025). The relatively better performance in private HFs that did not have high levels of information technology and information sharing is consistent with claims by Abdulkadir et al. (2024) and Simchi-Levi et al. (2009) that the benefits which may accrue from adoption of information technology are contingent on existence of appropriate SC practices.

In the context of SC collaboration processes, the absence of fundamental warehousing and inventory management practices such as the maintenance of updated stock records, regular physical counts and medicine tracking for accountability in public HFs is consistent with the findings of Privett and Gonsalvez (2014). These authors identified warehouse management and inventory management among the challenges affecting HCSCs. Surprisingly, private low- and some high-level HFs, even without ERPs, maintained better records and accountability than public counterparts, presumably due to financial incentives.

Reliance on a push system due to a lack of capacity to order in public low-level HFs is consistent with the findings of Privett and Gonsalvez (2014) on order management challenges and Yadav (2015) on the mismatch between skills and system design in HCSCs in LMICs where HF in-charges who are best positioned to order lack the necessary skills. It is therefore vital to strengthening order processing and management if the performance of public HCSCs is to be improved.

Finally, operationalization of healthcare SC processes in LMICs along the established SCOR model is consistent with Lockamy and McCormack (2004); Lambert et al. (2005); Simchi-Levi et al. (2009) and Asrol et al. (2021) in demonstrating the SCOR model as a process framework which helps in measuring and comparing SC performance across sectors in this case public versus private HCSCs.

A framework for SC collaboration has been developed to counter fragmented collaboration practices and enable more effective interventions within the context of public versus private HCSCs in LMICs. Consequently, this paper makes valuable contribution to theory, practice and policy development in this area.

The paper contributes to theory by addressing the call for middle-range theorizing as outlined by Pellathy et al. (2018) and the call for research on collaboration in HCSCs made by Friday et al. (2021), focusing on mechanisms of HCSCs in LMICs. The operationalization of the dimensions of SC collaboration in HCSCs along the established themes of strategy, structure, systems and processes extends the work of Rodrigues et al. (2004). Similarly, operationalization of the SC processes in HCSCs in LMICs along the established SCOR model demonstrates it as a process framework which helps to measure and compare SC performance across sectors, in this case public versus private HCSCs.

Second, this paper contributes to practice by providing a framework of antecedents to improve performance of HCSCs in LMICs. The performance of the public HCSCs in LMICs could be enhanced by benchmarking against the private HCSCs to introduce shared benefits and appropriate SC relationships; promote adherence to government guidelines; adopt and improve existing information technologies as well as improving processes such as warehousing, inventory management and order processing. Recommendations regarding public versus private and high-level versus low-level approaches enable practitioners to deploy the most appropriate set of strategies, structures, systems and processes identified as strongly associated with performance for each situation thereby boosting SC performance with limited resources.

Finally, this paper provides policymakers in the public and private healthcare with a comprehensive framework for SC collaboration, complementing previous fragmented approaches. This framework facilitates development of policies for public HCSCs to benchmark against and leverage best practices of private HCSCs.

This study provides substantial data in the context of Uganda, with a particular focus on two specific medicines: misoprostol and oxytocin. This provides a foundation for future research. To validate generalizability across LMICs, it is recommended that similar investigations be conducted in middle-income countries. Future research directions include the analysis of other medicine types (e.g. those requiring cold chain), the quantitative testing of the proposed framework using models such as Jbaily et al. (2020) and the exploration of how collaboration impacts specific HCSC performance indicators. It is hoped that this study will stimulate further research and promote effective interventions to improve medicine access in LMICs.

1. How has been the supply of medicines to your facility. Probe for information on Misoprostol and Oxytocin.

2. Refer to the graphs (focusing devices) on stock levels of Misoprostol and Oxytocin. Are the graphs accurate? What happened during that period?

3. Can you recall a major incident when the flow of products, information, money, or decisions in the selected supply chain was disrupted? If yes,

   • What products, information, money, or decisions were involved?

   • When did the disruption occur?

   • Where did the disruption occur?

   • Who was responsible for the disruption?

   • What incidents preceded the disruption?

4. Can you recall a major stock-out of essential medicines? If yes,

   • What essential medicines had stock-out?

   • When did the stock-out occur?

   • Where did they stock-out occur?

   • Who was responsible for the stock-out?

   • What incidents preceded the stock-out?

5. Can you recall a major expiration of essential medicines? If yes,

   • What essential medicines expired?

   • When did the expiration occur?

   • Where did the expiration occur?

   • Who was responsible for the expiration?

   • What incidents preceded the expiration?

6. What are the strategic supply chain related objectives of your organization?

7. What are the supply chain related Key Performance Indicators (KPIs) of your organization?

8. How do you coordinate with other members of your supply chain?

9. What rules govern your supply chain operations?

Source(s): Authors’ own work