4

Challenge: Ineffective Supply Chains

Supply chains are a critical component of a systematic treatment program for disease, said Prashant Yadav, director of the Healthcare Research Initiative at the William Davidson Institute at the University of Michigan. Although many people envision cartons and warehouses, Yadav explained that a supply chain is a complete ecosystem of organizations, people, technology, activities, information, and resources that together ensure that a product travels from where it is manufactured to the patient.

THE EFFECT OF DEMAND ON THE SUPPLY CHAINS

Supply chain management is a well-developed scientific discipline, said Yadav. Supply chains deliver medicines, but also return critical information to planners regarding need, demand, and consumption. In some cases, he said, functioning supply chains can also play a role in demand creation. Yadav expanded on lack of demand as it relates to supply chains, noting that demand is different from need. He described a “low-demand-low-supply trap,” where the delivery system remains stuck in a suboptimal state of low use, low availability, and high cost, unless there is some intervention. If demand is low because of lack of provider or patient awareness, or some other reason such as affordability or availability, then that poor demand implies a small market size. A small perceived market size offers little incentive to invest in supply systems from a business perspective. If supply chains are weak, the margins and mark-ups (i.e., per-unit product costs) become larger, the product availability decreases, and that in turn further lowers demand. According to Yadav, investment in supply systems reflecting true market potential, and not current perceived demand, is needed to get the supply chain out of this trap.

Yadav highlighted some of the differences between supply chains in developed and emerging market pharmaceutical systems, such as those found in developing countries. In developed countries there is a strong presence of public and private insurance and limited out-of-pocket

expenditure; strong, well-defined laws and enforcement of regulations; and distribution by large organizations with nationwide coverage and relatively low mark-ups. By contrast, payment in developing countries is through either out-of-pocket or direct government purchasing of medicines for a government-run system; regulatory structures that can be fragmented and weak with ill-defined and poorly enforced laws; the wholesaling and retail pharmacy system is not strong enough to act as a major mechanism for the supply of medicines, and mark-ups are high. In developing countries, the physical supply of medicines is dominated by a government-run and -owned system with a limited, fragmented private distribution market with little or no nationwide coverage. Patients obtain medicines from private-sector pharmacies, second-tier pharmacies, chemical sellers, or public-sector community health workers, health centers, district hospitals, and medical stores (Smith and Yadav, 2012; WHO, 2011d; Yadav, 2010).

Public-Sector Medicine Supply Chains

In a government, or public-sector, supply chain system, there is a cycle of uncertainty, Yadav said. Manufacturers sell directly to the country’s Ministry of Health (MOH), which receives financing from the Ministry of Finance or other sources. MOH purchases feed into a central medical store, after which medicines are distributed to health care facilities. In most cases, there are no processes by which information can be fed back into the system to provide data on what medications were financed, supplied, and distributed. In addition, data on use and demand are not supplied. Improving access requires a concerted effort to understand demand and develop a well-managed financing and procurement process, Yadav said.

Yadav highlighted several factors that lead to poor availability of medicines at health clinics in government-run systems. The timing of funds disbursement from the Ministry of Finance, or other external source, for the purchase of medicines from manufacturers can be variable and uncertain, impacting procurement cycles and supply. Procurement processes can be archaic, he said, and lead times from manufacturers long. Yadav suggested that the government’s control on distribution can lead to weak incentives and poor information flow along the length of the supply chain. He added that poor tracking of consumption and staff

capacity to manage inventory, stock, and dispense can be inadequate in some countries.

Another problem is what Yadav described as an unnecessary level of complexity. The structure of the supply chain is mapped exactly to the administrative structure of the country, resulting in multiple tiers and stopping points along the supply chain. Complexity can negatively influence distribution, resulting in a phenomenon called “the bullwhip effect” (Lee et al., 1997). Yadav explained that this effect is the result of small variations in patient demand at the clinic that are amplified as information is processed upstream through stopping points in a multi-tiered distribution system—from health facility to district and provincial stores, central medical store, procurement, and finally, to the manufacturer. Fewer tiers in the distribution system might help it remain in sync with actual demand, Yadav said.

Forecasting demand can also be challenging. If replenishment intervals are frequent (e.g., monthly orders), forecasting of demand can be more accurate. However, procurement departments may place orders once every 1 to 2 years, leading to wide uncertainty that manifests as either stock-outs or excess stock, Yadav said. Currently, information about use is collected through surveys. Although surveys are relevant for evaluation, effective supply chains are based on high frequency or continuous information feedback. A few participants stressed the importance of regular monitoring of systems to check the availability of products and the number of patients in need every month. A participant noted that at the clinic level, managers do not have the tools to assess quantification and determine future medication needs. Within a hospital there is often little or no communication between those prescribing and those ordering the medications. Another participant noted that there are also concerns about medicines expiring, which leads to ordering of limited quantities, increasing the potential for stock-outs. Yadav agreed that the capacity for quantification of procurement might best exist at a district level and not necessarily at a primary health center level. Particularly for unstable demand environments, purchasing should not be based on previous consumption data if the intent is to accurately scale-up the effort, said Yadav.

There are also infrastructure or “last-mile transport” challenges in getting products to primary health centers, which are often in rural areas. However, Yadav said that in his personal experience, even the health center closest to the central medical store or hospital might be out of stock of at least 30 percent of the core medicines. This is not necessarily

an infrastructure challenge, he said, but could be an issue of poor information flow and weak incentives. Providing supervision and training within health facilities on inventory management, forecasting, procurement, and requisitioning, Yadav said, may help to alleviate this issue of stock-outs.

Yadav suggested shifting tasks to where there is greater capacity. For example, in systems with weak clinic-level capacity for ordering/requisitioning, perhaps the district pharmacist or provincial pharmacist can be the locus of decision making. In systems with challenges delivering medicines to rural areas, it may be effective to combine information collection and product distribution. In Zimbabwe’s Delivery Team Topping Up¹ system, for example, staff from the district visit every health center to deliver essential products, conduct inventory, and resupply in the same visit. The district staff member is able to capture local knowledge about demand and make decisions about requisitioning—removing the responsibility from the clinic staff. Similar “moving warehouse” pilot programs are under way in Mozambique, Nigeria, and Senegal.

Private-Sector Medicine Supply Chains

Within the private sector, Yadav explained that most pharmacies rely on a “cash-and-carry model,” in which products are bought directly from the wholesaler rather than through a distributor. While large retail pharmacies may receive some form of credit from the wholesaler, small-town rural pharmacies are typically not extended credit and must purchase medicines on a cash basis. Yadav noted that this might result in retail pharmacy owners, with limited working capital, only stocking medications that sell quickly. In addition, pharmacy staff often travel significant distances, with fewer trips for longer distances, resulting in less frequent opportunities to restock supplies. In comparison, Yadav mentioned that private pharmacies and private wholesalers function well, delivering products to even the most remote areas, if incentives are structured appropriately (Yadav et al., 2012). For example, products in high demand in remote areas offer financial incentive to wholesalers to travel long distances given the potential profit. Supply chains that are patient-centric are critical, said Yadav. He suggested that supply chains

______________

¹Part of the USAID/Deliver Project. See http://deliver.jsi.com/dhome/countries/countrynews?p_persp=PERSP_DLVR_CNTRY_ZW.

be designed as a mechanism to deliver the product to where the patient seeks care, rather than asking the patient to come to where the product is available.

Barriers to Getting Product into the Supply Chain

Before an MNS medicine can enter any supply chain, public or private, it must be registered in the country, said Yadav. If the market size is small and the cost to serve that market is high, the business case to enter the market is weak, Yadav said. Another factor for a manufacturer to consider is high costs associated with time, effort, and resources needed to register a product. During the discussion, Samji noted that the level of complexity of regulations can be a significant barrier. For example, each country requires its own registration number on the product packaging, however, product volumes in these countries are low. Producing country-specific packaging adds complexity and cost for the manufacturer, which can lead to increased prices. Samji suggested a potential solution might be a mutual recognition system among country regulators to reduce the complexity of developing packaging for each individual market.

Although some manufacturers register their products out of corporate social responsibility, given the high costs and relatively small market size, Yadav noted that many do not. When manufacturers choose not to register products in a country, this leaves few or no supply sources, which leads to higher prices, a lack of availability, or both. The question, Yadav said, is whether this perception of low market size or high cost to serve is accurate, or whether it is due to lack of data about the market. Developing strategies to reduce the time, effort, and transaction costs for registration is important, said Yadav. For example, rather than raising registration fees, creating an appropriate financing model for the developing country’s regulatory authority may be beneficial.

Several participants discussed leveraging regional block structures²—regional areas that share common institutional practices, goals, or currency—to harmonize regulations across local procurement. Yadav noted that there are some initiatives addressing this through harmonization, such as the African Medicines Regulatory Harmonization (AMRH) program.³ The goal of this program is to improve public

______________

²See http://www.usitc.gov/publications/332/3650/pub3650_ch3.pdf.

³See http://www.amrh.org.

health by increasing access to good quality, safe and effective medicines through the harmonization of medicines regulations, including the reduction of the time taken to register essential medicines for the treatment of diseases. A participant pointed out that the WHO Regional Office for Africa could explain to countries the advantages of pooling orders into larger volumes. However, one participant noted that it may be difficult to harmonize procurement within regional blocks that do not share a common language. Another participant noted the additional regulatory challenges of procurement of controlled medicines for acute management of MNS disorders, such as injectable phenobarbital.⁴ Alem added that most medicines are purchased by SSA countries from abroad, and foreign currency is a challenge, as the capacity to generate foreign currency can be low. Although some medicines are now being manufactured locally, companies still struggle with importing raw materials, he added.

Yadav highlighted several additional challenges facing pharmaceutical companies in serving markets in SSA. The lack of good distribution practice (GDP)-compliant distributors, particularly in the private sector, may lead to pharmaceutical companies not wanting to enter the market, due to potential reputational risks associated with loss of integrity (e.g., inappropriate handling or storage) and high distribution fees. In general, Yadav added, the wholesale market in SSA is fragmented. Given the larger quantity of wholesalers and the fixed costs associated with each, patients are subjected to high medicine prices compared to markets with fewer wholesalers. High retail prices are also associated with mark-ups added by intermediaries at multiple levels in the distribution system.

Breaking the Low-Demand-Low-Supply Cycle

In summary, many complex factors contribute to poor availability of medicines. These problems are complex and require action on multiple fronts. However, the complexity and multi-dimensionality of the problem can not be an excuse for inaction, Yadav said. Lessons can be learned from successful supply chains for other consumer products in SSA (Yadav et al., 2013). Coca-Cola, for example, is distributed broadly

______________

⁴Phenobarbital is a barbiturate used to control seizures, alleviate anxiety, and prevent withdrawal symptoms from barbiturate dependency. See http://www.nlm.nih.gov/medlineplus/druginfo/meds/a682007.html.

throughout SSA through independent wholesalers and retailers to reach a larger number of consumers. The company leverages local knowledge about the integrity and pricing of their product once it reaches the consumer to ensure that the distributor is complying with contractual agreements (Yadav et al., 2013).

Yadav concluded by noting that breaking the low-demand-low supply cycle will require discussion and action on:

• Better forecasting and needs assessment, to include inventory management training for staff;
• The redesign and simplification of distribution structures;
• Better information collection and flow;
• The creation of agile procurement structures;
• Higher frequency of deliveries;
• Incentives and accountability in the supply chain;
• Private-sector transport and distribution; and
• Working capital credit for private pharmacies.

LESSONS LEARNED FOR ADDRESSING INEFFECTIVE SUPPLY CHAINS

As previously mentioned, five example programs addressing access to medicines were presented during the workshop to facilitate exploration of best practices and lessons learned from other programs. The examples were selected by planning committee members and included two country-level programs, an infectious disease project, and two noncommunicable disease programs. Highlights from the presentations of the lessons learned for addressing ineffective supply chains are provided in Box 4-1. A full description of the examples as presented can be found in Appendix A.

CHALLENGES AND OPPORTUNITIES FOR ADDRESSING INEFFECTIVE SUPPLY CHAINS

In preparation for the focused discussion on supply chains, Giorgis summarized the issues for supply chains that were discussed in the presentations and example programs. The biggest procurer of MNS medicines in SSA are governments, however, several participants noted that MNS medicines are not always procured in quantities capable of addressing the need. Several participants noted that to ensure medications are available on a timely basis, there is a need for continuous, reliable data monitoring of demand and use. According to a few participants, the capacity of dispensary and pharmacy staff to appropriately requisition and purchase medicines based on demand can impact supply.

Following the focused discussion, Tarun Dua, medical officer in the Department of Mental Health and Substance Abuse at WHO, reported that five priority barriers were identified by various participants relative to selection of MNS medicines: (1) absence of quality, timely information, including data collection and analysis; (2) deficiencies in the allocation and training of human resources for supply chains; (3) lack of coordination at all levels of the procurement chain; (4) inefficiencies across different tiers of the supply chain; and (5) long procurement lead

times with little transparency of the process. All constraints and/or barriers and potential opportunities noted by individual participants are included in Table 4-1.

Dua noted that many participants stressed the need for country-specific solutions because no one-size-fits-all solution can be applied globally. Several participants agreed that each country faces unique challenges related to supply chains.

A few participants noted that development of an information network systems approach for improved communication among tiers might increase the quality and timeliness of information flowing along the supply chain system. In addition, a few participants stressed that training on data collection and analysis might improve forecasting of need and reduce stock-outs and overstocking of medicines. Beyond training of supply chain staff, many participants noted that there are deficiencies in allocation of human resources across supply chains. Dua said several participants suggested that policies to increase the number of workers to supply chains might lead to increased knowledge of the system, again improving forecasting of need.

A few suggestions by different participants were offered as opportunities to improve coordination among tiers of the procurement chain, including the use of mobile technology to improve information flow and the establishment of therapeutic committees at health institutions. One participant noted that these committees could consider information on needs, stocks, and supply chain logistics and conduct coordination efforts across the supply chain.

Dua added that several participants noted inefficiencies across different tiers of the supply chain, including the many layers and steps needed to move medicines. A few participants suggested that reduction in the number of tiers between central warehouses and patient distribution points might result in faster procurement and delivery of essential medicines. Another mechanism suggested by a participant for addressing this challenge was improved transportation and potentially the outsourcing of transportation to private distributors. Finally, Dua noted that one participant emphasized that opportunities could be found through the use of complementary and/or multiple procurement agencies.

Wrapping up the overview of the focused discussion, Dua indicated that early, frequent, and transparent consultations with manufacturers might reduce lead time for procurement, stock-outs, and overstocking. Through reliable and diversified manufacturer sources, Dua noted that

prepositioning of medicines and raw materials might improve, reducing the time for delivery of essential medicines.

TABLE 4-1 Opportunities to Address Ineffective Supply Chains for Essential Medicines as Identified by Individual Workshop Participants¹

______________

¹This table presents challenges and opportunities discussed by one or more workshop participants. During the workshop, individual participants engaged in active discussions. In some cases, participants expressed unique ideas and/or differing opinions. However, because this is a summary of workshop comments and does not provide consensus recommendations, workshop rapporteurs endeavored to include all workshop participant comments. This table and its content should be attributed to the rapporteurs of this summary as informed by the workshop.

______________

^aAtalay Alem

^bTarun Dua

^cMapoko Ilondo

^dJafary Liana

^eMamuye Mussie

^fIsmet Smaji

^gPrashant Yadav

^hPaul Zintl

This page intentionally left blank.