“Better data drive better decisions” is a truism that researchers everywhere are all too familiar with. Increasing the availability, usability, and relevance of data is key to tracking performance and informing smarter, more efficient policies—but too often the data we need simply aren’t available, at least not in a useful format. Recently, we’ve been exploring the availability of data (or lack thereof) related to global health commodity markets in the context of CGD’s working group on the Future of Global Health Procurement. To ground the working group’s recommendations, we’re trying to understand the current state of health commodity procurement in low- and middle-income countries (LMICs)—specifically which commodities are procured, by whom, how, and at what price. While there has been progress in improving data availability, the data are spotty and challenges abound. Case in point: there is currently no global repository for data on health commodities, not even for those funded by donors. So, we’re piecing together an array of available data sources, which while not always easily comparable and not without gaps are nevertheless needed to get to the bigger picture.
We’re working within a complex landscape
Health commodity supply chains in LMICs are complex to say the least, generating large amounts of information at each step. In any given country, the myriad products, manufacturers, and formulation packs must be tracked across the value chain. Supply channels typically comprise a complex web of intermediaries between manufacturers and patients. In the private sector, for example, this includes wholesalers, distributors, even sub-distributors, and many retail outlets. Additionally, there are distributor mark-ups and discounts, as well as associated taxes, foreign exchange fluctuations, and stock levels/credit costs to consider.
The other complicating factor is that different stakeholders require different data for different uses. The information a national regulatory body needs is understandably different from what’s required by a government procurement department, for example. And, as countries become richer, they not only progress in terms of the complexity of their health commodity needs but also in terms of the data required to track and ensure those needs are being met.
There are myriad sources and the data are often spotty
In the absence of any global database or repository on health commodities to turn to, we set out to cobble together a range of available data sources, each with its own strengths, gaps, and quirks.
Donor and NGO health commodity data
Donors and NGOs spend roughly $6 billion each year on a relatively small number of health commodities such as vaccines; drugs for HIV, TB, and malaria; bed nets; and contraceptives (based on back-of-the-envelope calculations). In fact, in some 40 low-income and lower-middle-income countries where we have detailed data, donor spending makes up less than 10 percent of the estimated $50 billion health commodity market.
In our search, we found that most of the publicly available health commodity data cover these donor-dependent diseases and program areas. Indeed, the development community has made important strides in tracking financing flows for health commodities and increasing transparency around information on prices and volumes. However, these efforts are not comprehensive, especially when compared to the systemic data collection in the private sector (see below). Tracking donor funding to health commodities across the board is an arduous and cumbersome process. There’s no single database or platform to track the financing, and information on prices and volumes remains sparse.
Some sources we came across include the WHO/Health Action International’s medicine price database, Management Sciences for Health’s International Medical Products Price Guide, WHO/Western Pacific Regional Office’s Price Information Exchange, PEPFAR’s Country Operational Plans, President’s Malaria Initiative Operational Plans, The Global Fund’s Price and Quality Reporting or PQR tool, Médicins Sans Frontières’ Untangling the Web report, and the EURIPID database.
But combining these sources poses several challenges, including, but not limited to, double counting, a lack of common nomenclature for product catalogues, time lag in data capture, outdated data, incomplete data, and a lack of an integrated data platform to support country procurement departments.
Government health commodity procurement data
Beyond the donor-funded disease and program areas, we found that the data landscape gets even more sparse and fragmented. Several LMIC governments do, however, publish details of their tenders and health commodity procurement. The Philippines has a particularly good online system of mapping procurement, the Drug Reference Pricing Index, that shows government procurement down to a single facility and regional medical store level. Other examples we found include South Africa, Indonesia, Jordan, and Kazakhstan. In addition, IQVIA, formerly IMS Health, also captures public sector tender and distribution data for countries like Tunisia, Serbia, South Africa, the Philippines, Zambia, and Uganda.
Lastly, to supplement what’s publicly available, we also requested data from government procurement departments. Out of 40 countries where we contacted representatives from the ministry of health or central medical store, 10 agreed to share data under the promise of anonymity. Of course, we realize this approach is not without limitations—it likely introduces bias towards countries with better data systems, transparency, and the motivation and flexibility to participate in such projects.
Private sector data
We also found private sector data sources, though these often come with a considerable price tag and their own set of drawbacks. MIDAS is an analytics platform that provides standardized and comparable pharmaceutical sales data and medical data from over 90 countries. But data are more focused on the needs of the multinational pharmaceutical industry, with better data coverage in upper-middle-income and high-income markets. In certain instances, IQVIA can disaggregate data at various subnational levels and between the public and private sectors. These databases are more useful from the perspective of comparing procurement patterns across the public and private sectors in LMICs.
Nielson also collects private health commodity data in LMICs. These data focus more heavily on the supermarket and consumer segments of the health commodity landscape, which are useful for over-the-counter and consumer health products like condoms, but less so for prescription pharmaceuticals.
Having good data really does matter
Accurate, relevant, and timely data on the consumption of health commodities across LMICs can help inform public health resource allocation, as well as track performance and efficiencies.
Better demand, consumption, and stock level data, for example, can enable procurement bodies to optimize contracting approaches and ordering practices. Access to information on average costs (and volumes) can simplify product selection for procurement departments and make it easier to identify non-functioning generics markets and, in turn, realize efficiencies from tackling inappropriate price differentials. Technology solutions, including serialization, that generate supply chain data can help track and identify product quality issues. Furthermore, better data on the size and trajectory of the market can potentially help encourage manufactures to register products, especially in smaller, fragmented markets.
Most importantly, data on health commodity procurement make it possible to track pricing, availability, access, and quality—the end goals we care about. These data form the building blocks for understanding what approaches and strategies work well in different settings. This function is not only important for accountability of public spending in LMICs, but also for donor financing of health commodities at the global level.
How can we progress to a more comprehensive picture?
Despite efforts to improve the availability of data, challenges and gaps remain. We’ve found that there are no accepted standards for collecting, publishing, or analyzing data on global health procurement and no database(s) containing up-to-date, reliable, and comparable data on product types, prices, and utilisation volumes. And this challenge becomes even more relevant as a growing number of lower-middle-income countries prepare to lose eligibility for donor assistance while also progressing towards universal health coverage. Without a good view into spending on health commodities, it will be that much more challenging for countries to ensure high-quality, affordable health commodities are available to those who need them the most.
The CGD working group aims to propose how the global community can do better on data and analytics in the area of global health procurement. Also stay tuned for forthcoming analyses drawing on the data we’ve been able to tap into. In the meantime, if you know of additional sources of data on health commodity procurement, please leave us a comment below.
Many thanks to Rachel Silverman for feedback on this post.