A Pandora’s box: sustainable pharmaceutical supply

For the first time in six decades, intravenous penicillin will not be available on hospital shelves in Australia. CSL Limited, who distributes benzylpenicillin in Australia, has been informed by their supplier that due to international floods and disasters, there is a critical global shortage. CSL gave Australian hospitals less than 2 weeks’ notice, and have warned that resumption of local supply is not expected until December 2011.

Shortages of critical life-saving medicines are an increasing threat to public health.1 Although there have been no official medication shortages in Australia in 2011, the true problem is unrecognised here and is escalating internationally. In the United States, there have been 198 drug shortages over the past 3 months,2 more than double that for the entire year of 2007.3 The only organisation in Australia attempting to alert prescribers to impending shortages, the Society of Hospital Pharmacists of Australia, has listed only seven medications to date, while the Therapeutic Goods Administration, which only notifies prescribers once drugs are beyond critically short supply, has listed none.

The reasons behind these shortages are complex and include increased demand, manufacturing delays, recalls, shortages of ingredients, and the realities of business in a competitive international market.1,4 For many medicines, there are only a few sites worldwide, mainly in India and China, that manufacture pharmaceutically active ingredients. Some of these sites are the sole source of multiple active ingredients.5,6 With the opacity of international manufacturing processes, it is very challenging even for governing agencies to understand where medicines are sourced, and subsequently to anticipate geopolitical supply chain vulnerabilities or breaches in quality and safety.5,6 Australia, which produces less than 5% of the pharmaceuticals it consumes,7 is particularly vulnerable.

Shortages in essential medicines occurring at a local hospital level are most commonly a result of international trade and manufacturing issues, as attested by recent shortages in heparin,8 propofol9 and injectable antibiotics.10 However, with increasing instability in climate, a natural disaster locally or distantly could significantly interrupt manufacture and supply chains, and could lead to unexpected disaster-associated mortality globally.

To better understand our local vulnerabilities, we conducted a hypothesis-generating study by analysing consumption and stock reserve of essential medications in our hospital system. The Hunter New England Local Health Network in New South Wales provides hospital care to a population of 413 000 people in a defined geographical area, broadly representative of the Australian demographic.11 We hypothesised that the hospital pharmaceutical reserve was less than 30 days’ supply for a significant proportion of commonly used essential medicines. Further details of our study method are shown in Box 1.

Of the 378 essential medicines on our modified list (Box 1), 3% had less than 7 days’ reserve, 24% had less than 30 days’ reserve, and 26% had more than 3 months’ supply. There were 76 medicines that were high volume and these generally had the least stock reserve, with 51% having less than 30 days’ supply. Medicines used in more acute settings, for instance anti-infectives, analgesics, intravenous fluids and anticoagulants, had relatively small reserve capacity, but all had at least 7 days’ supply (Box 2). There was 34 days’ supply of benzylpenicillin, with 27 defined daily doses prescribed each day in the hospital network.

Medicines used for chronic conditions generally had greater reserve capacity, with most having more than 1 month’s supply. Only one of the 25 antihypertensive agents and one of four statins had less than 30 days’ supply. Two of the five antianginal agents, three of the five antidiabetic agents and four of the six diuretic agents had more than 3 months’ supply. Low-volume medicines, representing 138 agents, generally had large reserve capacity, with 62% having more than 3 months’ supply. HIV medicines and cytotoxic drugs had low reserve capacity (Box 2), a reflection of pharmacy policy designed to reduce cost and expiry of expensive medicines with short shelf lives.

For local disasters, most hospitals would be able to act as community reserves of essential medicines, particularly those used in the treatment of chronic disease, for which most hospitals will have substantial reserve supplies. It has previously been demonstrated that interruption to the supply of medicines to those with chronic health problems contributes significantly to disaster-associated mortality.12 It is imperative that hospital pharmacies are designed to be protected from flooding and damage from natural disasters, that local suppliers are integrally involved in disaster planning, and that there is contingency for hospitals to act as large-scale community dispensaries if required. This is particularly the case for hospitals in northern Australia, where the risks of flooding are greater and where the supply chain is more geographically vulnerable.

Our study brings into question the utility of the National Medical Stockpile. There are more cost-effective ways of sourcing non-surge-related essential medicines; for instance, hospital and community requirements for antihypertensive medicines will not surge in the event of a disaster, but oseltamivir would surge in the event of an influenza pandemic, as would antidotes in the event of mass poisoning. With appropriate national or state linking of hospital pharmacy databases of essential non-surge drugs, similar to the NSW Register of Lifesaving Drugs, medicines could be mobilised rapidly to where they are needed and the stockpile could be significantly downsized.

There is no official mechanism for identifying drug shortages in Australia. Setting up a monitoring system — for example, via a sentinel hospital pharmaceutical supply — could help identify impending drug shortages locally and allow earlier rationalisation to mitigate such events. Legislation could be enacted to compel manufacturers and wholesalers to report compromise of manufacture or supply.

Although many developed countries have registries of efficacious “essential” drugs, there have been no previous attempts to define nationally relevant essential medicines in terms of disaster management and supply chain vulnerability. Our study does not attempt to develop such a list but defines usage of classes of medicines. A national list could be simplified to include a single medicine from each subclass with strategically valuable attributes, for instance, medicines that are manufactured locally. A national list of essential medicines would enable local regulatory bodies such as the Therapeutic Goods Administration to focus on supply chain issues of critical drugs.

International efforts need to focus on sustainable supply of essential medicines and equipment. The WHO needs to consider environmental threats to key international manufacturing plants of essential medicines, and could incorporate such requirements into good manufacturing practice guidelines.

From a longer-term perspective, pharmaceutical supply, similar to food security, is likely to become a more prominent global problem. Governments and international agencies such as the WHO have an obligation to embrace not just quality regulation of pharmaceuticals but also contingency plans for unforeseen global events; events that are increasingly severe as climactic variables become more unpredictable and extreme. From an Australian perspective, this means adopting policy that encourages local manufacturing industries to have some capacity to fill the gaps in essential pharmaceuticals and medical equipment if they ever appear, and it requires national procurement decisions to be based not just on cost but also on sustainability.