A recent paper by FDA officials explored why and how manufacturing-quality problems could combine with other economic and technological factors to result in shortages of generic sterile injectable drugs.
The paper, written by Janet Woodcock, Director of FDA’s Center for Drug Evaluation and Research (CDER), and Marta Wosinka, director of economics staff at CDER, noted that the fundamental problem identified is “the inability of the market to observe and reward quality.” An important, but largely unrecognized aspect of quality is the ability to reliably meet customer demand, they wrote.
“This lack of reward for quality can reinforce price competition and encourage manufacturers to keep costs down by minimizing quality investments.” In the case of sterile injectables, “there is very little margin for error, so a lack of sustained investment in infrastructure and vigilant quality focus can produce what economists term a “bad market equilibrium,” particularly the problem of reliability, in which quality problems are rife.”
As a result, the authors proposed that FDA reward quality as a way to avoid manufacturing problems that are causing shortages. Woodcock and Wosinka noted in their piece, published in Clinical Pharmacology & Therapeutics, that FDA must use its regulatory flexibility “on behalf of patients” to avoid shortages and further strengthen the incentive to “push the envelope” on quality.
While they express FDA’s commitment to new tools and approaches to prevent and resolve drug shortages, some of which have “already borne fruit,” the authors recognize that the shortage crisis has reached a point “where FDA needs to engage the marketplace to help address the manufacturing problems, which unilateral FDA actions have not been able to prevent.”
Rewarding quality, however, may not be simple because there are various factors that prevent drugmakers from sufficiently investing in their operations – aging facilities, new opportunities to produce other medicines and contract manufacturing. Nevertheless, Woodcock and Wosinka propose that FDA “support the buyers and payers in their purchase and reimbursement decisions by providing them with meaningful manufacturing quality metrics.”
They note that “this general approach has been successfully used in many other settings where quality is difficult to observe or quality signals are difficult to interpret. Restaurant grades, HMO scorecards or even a US Pharmacopeia stamp on vitamins are just a few among many tools that utilize this concept. It would then be up to the marketplace to answer the ultimate question: how much are we willing to pay for quality?
The authors noted that their explanation for drug shortages is specific to generic sterile injectable products, so caution should be applied when extrapolating the conclusions of this article to other drug shortages, particularly oral products and brand name injectables.
By early 2010, two companies decided to halt production of propofol after inspections revealed sterility and quality issues, leaving only one company to meet the market demand. With the remaining firm unable to sufficiently ramp up production and inventories running out, a drug shortage occurred.
In 2011, drug shortages rose to unprecedented levels, with 251 medically necessary drugs affected. Seventy-three percent of these drugs were sterile-injectable products such as the anesthetic propofol, injectable morphine, epinephrine used in cardiac arrest and anaphylactic shock, intravenous nutritional supplements, and chemotherapy agents. The resulting health crisis has been particularly pronounced because these drugs are medically necessary, meaning that they are used to treat or prevent a serious disease or medical condition and that there is no appropriate substitute.
Although today’s buyers may now be more aware of quality problems with sterile injectables, “buyers appear to generally discount quality concerns and to focus principally on obtaining the lowest prices possible.” This behavior is likely based on a belief that all marketed products are of equivalent quality. “The resulting lack of reward for quality may encourage manufacturers to keep costs down by, for example, minimizing quality-related investments in areas such as maintenance of production facilities and equipment, quality control testing and oversight, and timely response to early indicators of quality problems.”
Not every production disruption turns into a shortage, but virtually all shortages are preceded by disruptions in production. While some supply disruptions are unanticipated by the manufacturer, as a result of natural disasters or unforeseeable disruptions in the supply of the active pharmaceutical ingredient, they are few and far between.
“Instead, drug shortages are first and foremost driven by the inability of various firms to maintain production because of the failure of quality management in facilities that produce the finished dosage form of the drug (rather than the active ingredient). These failures were directly responsible for 56% of sterile injectable drug shortages in 2011.”
Other shortages caused by delays and capacity issues are driven by quality-related shutdowns. For example, shortages resulting from manufacturing failures for one drug increased demand for another drug by so much that companies producing it were not able to catch up with demand. In this case, quality-based shortages of one drug lead to shortages of another drug (up to 7% of sterile-injectable shortages).
Some firms have discontinued older, medically necessary products. Product discontinuations, which accounted for 9% of sterile-injectable shortages in 2011, are often triggered by capacity constraints, many of them brought about by quality-related production line disruptions. Economic theory suggests that companies would generally consider a product to be worth producing if the revenue it generates can cover the costs of production. “This production rule no longer holds true when the firm reaches production capacity and managers are forced to make trade-offs among products, exposing products with low–average revenue for potential cuts as companies try to optimize profitability of their production lineup. Such pressures could be contributing to the discontinuations we observe.”
Woodcock and Wosinka noted that President Obama and Congress ensured that companies must properly notify FDA of potential shortages, given the time required to increase production at an existing line, initiate a new site, or (the most challenging) bring on a new manufacturer, shortages may occur even with advance notice of discontinuation.
The current shortage crisis with generic sterile injectable drugs follows directly from contamination problems at multiple facilities—problems that have required correction through upgrades of systems and/or production processes. The problems leading to shortages have included detection of shards of glass and metal in vials as well as of bacteria, endotoxin, or mold in what are supposed to be sterile products. These problems have differing etiologies some can be linked to insufficient maintenance of production facilities and equipment; some are a result of antiquated/inadequate aseptic operation design; and others stem from suboptimal quality control testing and oversight and lack of timely responses to indicators of quality problems.
The scope of these problems and the resulting remediation needs have varied across firms. In some cases, the widespread and systemic nature of the manufacturing failures has led firms to shut down all production while they address the problems. In most cases, firms have been able to continue production while improvements were being made, although often at a reduced rate.
Economic Incentive to Minimize Investment in Manufacturing Quality
The drivers behind the quality problems we describe above are multifaceted, but we postulate that at their core is the failure of the market to sufficiently reward quality.
Market does not reward quality: At the “heart of the quality problem is the fact that generic manufacturers compete on price.” Buyers—in this case, hospitals and clinics—consider any given generic products as perfect substitutes, giving manufacturers little room for differentiation. “Therefore, buyers have not been attuned to differences in the quality of production. In their minds, the products are of sufficient quality if they are on the market,” thus “the lack of reliability of supply stemming from quality failures has not seemed to be a factor of concern.”
“Another reason for lack of awareness of quality differences results from providers’ inability to accurately assess production quality for a given drug. Providers are more likely to suspect a causal link and therefore report an adverse event if a drug is used in a relatively healthy patient population. But, generally, injection and infusion drugs are administered to patients who have compromised immune systems and little ability to fight infections and/or other comorbidities.”
“Few health-care providers consider whether substandard manufacturing could be the source of an adverse event, both because of the general high quality of the US drug supply, in which product defects are quite rare, or because certain adverse events, such as infection, are common in the treated population. As such, the propofol example is more an exception than the rule. In that case, providers were able to establish a potential causal link because the drug is often administered to relatively healthy patients.”
“Despite buyers’ unwillingness to pay a premium for quality, many firms strive to exceed minimum manufacturing standards. Nonetheless, economic incentives create pressures on firms to minimize expenditures and thus generate a conflict between patient safety and profit.”
Quality is not fully transparent: A drug manufacturer is responsible for implementing dependable daily operations that assure consistent drug quality. Management’s daily decisions on myriad issues involving equipment, materials, maintenance, staff qualifications, supervision, process control, and investigations can significantly impact sterility assurance and will ultimately determine the quality of the drugs that are shipped from a given facility. The primary mode of sterile injectable production, aseptic processing, is not a forgiving process so any such variable, if not properly attended to, can lead to contamination.
The FDA assesses whether a facility is in a state of control through periodic inspections that provide an evaluation of a firm’s manufacturing operations, including their system for quality management. Almost all major sterile-injectable manufacturing facilities producing for the US market are inspected at least every 2 years, because they are based domestically. Between inspections, the FDA relies, in part, on firms to be forthcoming about their quality problems, which they report by issuing defect reports.
Defects in sterile injectable products can be difficult to detect because microbial contamination may be non-uniform and episodic. This lack of uniformity can confound conventional sampling plans. In addition, microbial contamination can increase after production, which means that more sensitive testing must be done at production time.
For these reasons, “FDA has made attempts over the last few years to enhance its audit of a site’s state of control and quality problems by developing a cadre of highly trained and experienced inspectors to conduct inspections of complex facilities such as those that produce sterile injectables.”
Exercising regulatory flexibility to avert shortages may have unintended consequences: If FDA uncovers signals of emerging problems, it seeks the firm’s cooperation in voluntarily correcting the potentially major issue before there is an impact on product quality. The agency will weigh the need to prevent adverse health outcomes but also the long-term incentives for firms to keep all of their facilities and processing lines up to quality standards.
FDA’s need to use regulatory flexibility on behalf of patients to avert and mitigate shortages could have unintended long-term consequences when coupled with the market’s lack of reward for quality. “Economic models predict that, in the face of the seeming intertemporal inconsistency created by dual FDA objectives, quality investments would be lower than if the FDA could use preemptive enforcement without regard for disruptions in medically necessary products. This dynamic may further reinforce the economic incentives to minimize quality investments given the nature of competition (based on price, not quality).”
Factors Reinforcing Insufficient Quality Investments
“The market’s lack of recognition of quality differences in generic sterile injectable products can influence firms to adopt a reactive approach to quality management. In such cases, a firm might wait until problems are identified before taking action, in contrast to taking a proactive and vigilant approach that is necessary to sustain a dependable operation.”
When dealing with such a reactive quality-management approach, the medically necessary status of certain sterile injectable drugs impels the FDA to consider the medical ramifications of lack of access. This in turn may give short-sighted firms an incentive to manufacture under a minimum level of control. In addition to the key drivers mentioned above, several other factors may have further exacerbated insufficient quality investments.
Aging facilities: In contrast to the production of tablets, facilities producing generic sterile injectables are primarily based in the United States because of high transportation costs associated with liquids that often require climate control. New manufacturing lines have been added by various manufacturers over the past decade, but some primary production lines and facilities are quite old.
“A review of inspection reports for key facilities suggests that some facilities at the heart of current drug shortages have been in operation continually since the 1960,” and “certain manufacturing lines have undergone only limited upgrades during that time while running 24 hours a day, 7 days per week, a finding consistent with a 2011 report.”
Such aging facilities require upgrades, but low margins may make the economics of such investments unattractive. Although contemporary facilities are highly automated and use isolators and other separation technologies to protect the processing line from contamination risks, older facilities typically include processing lines and facility layouts that are less effective at mitigating the various operational variables that pose risks to product sterility.
If the equipment is not well designed or is poorly maintained, repeated or extensive manual interventions often occur due to mechanical problems. When production line operators perform manual activities near an insufficiently protected product, they raise the risk of microbial contamination. Inadequately maintained or designed equipment has also resulted in particles being shed into the product, including metal or glass shavings.
New production opportunities: Magnifying the problem of aging facilities, firms have been adding many new products to their portfolios. Even before the drug-shortage crisis, generic sterile injectable lines were generally operating at full capacity. This means that without further expansion in capacity, new marketing opportunities would force firms to make trade-offs between which products to produce and at what volume.
In addition, line operators need to turn over lines between products more frequently if those new products are added rather than replacing what is in the current production lineup. This line turnover can additionally introduce opportunities for sterility problems that can compromise product quality, especially if it involves production lines that are not fully automated.
Contracting practices: Not only are buyers unable to observe manufacturing quality, but firms that contract out manufacturing of their product often do not have the same level of insight into or oversight of the contract manufacturer’s quality systems as they would have into their own. Over-commitment on manufacturing capacity by a contract manufacturer can lead to an unsustainably high number of products on each line and substandard oversight of the process.
Contracted products are primarily branded sterile injectable products marketed by emerging companies without their own production facilities. Their manufacturing is then contracted not to large branded manufacturers but to firms that primarily produce and market many generic products. These relationships are proprietary, so buyers and their representatives generally have no knowledge that a drug they are buying may be made by a contract manufacturer.
Failure-to-supply clauses, which are commonly used in contracts between manufacturers and their buyers or buyers’ representatives, are also problematic. While such clauses are supposed to impose a penalty for failure to supply a product, their incentive for firms to avoid supply problems has not been realized “because there often is no alternative source” when a drug shortage occurs.
Economic downturn: The recession may have additionally increased pressure on manufacturers to cut cost, with the number of pharmaceutical manufacturing jobs shrinking by 15% from December 2007 to June 2009. This may have caused at least some of the best-qualified employees to leave and find employment elsewhere, leaving behind less-qualified workers.
Increased price competition: Some have argued that the Medicare Modernization Act of 2005 (MMA) had an adverse impact on profit margins for generic sterile injectable drugs after it sharply lowered reimbursement rates that providers receive for provider-administered sterile injectables and other Part B drugs. The authors, however, were unable to confirm these assertions due to a lack of price transparency data before 2005. Nevertheless, they maintained that “trends in shortages of drugs affected by the payment reform are similar to the pattern among drugs that should not have been affected by it.” They also noted that “price competition had been a feature of the sterile-injectable industry long before the MMA—something that is consistent with the dynamics we have discussed in the previous section.”
Why do Supply Disruptions Turn Into Shortages?
Next, Woodcock and Wosinka explain several factors that explain why temporarily or permanently halting production on even a single production line so frequently results in a market-wide drug shortage.
High market concentration: A key factor that helps turn a single production line disruption into a drug shortage is high market concentration. Seven firms make up virtually all of sterile-injectable production as measured in standard units. As one narrows the analysis down to specific drug classes, the concentration further increases. Concentration further increases if one drills down to finer market definitions, e.g., specific molecules. For example, in 2008, Teva held a 68% market share of bleomycin, whereas Bedford Laboratories held a 62% share of cytarabine.
Specialized facilities and dedicated lines: One key reason for the high market concentration is that sterile injectable products must be produced in highly specialized facilities. Production must take place in a “clean room” environment with well-defined manufacturing processes and controls to assure that they are sterile, meaning free of contamination from all microorganisms and essentially free of visible particulate matter.
Not only must drugs be produced in specialized facilities, but their production is often committed to specific production lines within those facilities because of the drugs’ chemical properties, presentation form, and potential for cross-contamination. Most drugs’ chemistry requires that they be produced using a complex process of aseptic processing, in which the drug product, container, and closure are subject to sterilization separately and then brought together. A smaller number of drugs need only undergo terminal sterilization, in which the product in its final container is subject to a sterilization process through heat or radiation.
For some, the sterilized product further requires an additional lyophilization step in which water from the solution is removed in a highly controlled environment under high vacuum and extreme cold temperatures to yield a powder. Apart from the differences in production process, certain drug classes such as cephalosporins, penicillins (antibiotics), and some cytotoxics (oncology) require dedicated equipment, and sometimes even dedicated buildings, which then cannot be used to produce other types of drugs.
Finally, manufacturing equipment, due to compatibility issues, may be capable of filling containers of only certain limited size ranges (e.g., different equipment may be needed for a 1- vs. 25-ml vial size) and container types (vials vs. syringes). This further restricts products to specific lines.
Despite the limited fungibility of production driven by dedicated production lines, generic manufacturers do not appear to arrange for backup facilities. The authors noted that of 900 applications for generic sterile-injectable between 2000 and 2011, only 11, or just over 1%, referenced more than one facility for production of the finished drug. This contrasts with branded sterile-injectable applications, almost 20% of which were approved with backup facilities during the same time period. About 5–6% of the studied sterile-injectable ANDAs have subsequently submitted additional production sites. Those submissions occurred on average 4 years after original submission; none had In submitted additional sites within 1 year of the original ANDA.
Because a specific set of products is often committed to select production lines, a process disruption on these lines, especially if they are few, can readily turn into market-wide drug shortages for a whole group of products. Such coproduction of drugs is the reason that drug shortages have “traveled” across therapeutic classes. For example, after one key facility with cytotoxic lines shut down in early 2010, the number of shortages of chemotherapy drugs increased from 4 to 24.
Inventory practices: Generally low inventory levels further increase the risk that a supply disruption will turn into a drug shortage. Such inventory practices reduce expenses by eliminating surplus production and inventory holding costs but increase the risk that even a modest disruption in supply will result in shortage. These low inventory levels are one reason that notifying the FDA early about supply disruptions is important.
Ultimately, the FDA officials argued that “the fundamental problem with injectible shortages is insufficient market reward for quality (including reliability of production) stemming from the buyers’ inability to observe it.” This in turn gives manufacturers strong incentives to minimize quality system investments, especially when faced with pressures brought about by new production opportunities, aging facilities, and the recent economic downturn. Until new incentives are provided to improve quality, it is uncertain whether generic-injectable makers will adequately address quality issues.