As the international community grapples with COVID-19 and the extent to which countries should fund outbreak preparedness, governments also continue to face day-to-day healthcare funding decisions, including for cancer treatments. The rise of the cancer burden is far from a high-income country problem—cancer affects populations in all countries. In 2000, cancer accounted for 3.7 percent of the total burden of disease in lower-middle-income countries; by 2016, it had jumped to 5.6 percent (WHO Disease burden and mortality estimates). In upper-middle-income countries, the burden of cancer is second to cardiovascular disease, but is growing more than twice as fast. As policymakers increasingly take cancer seriously, and allocate a greater proportion of their health expenditure to its treatment, we ask whether cancer treatment is the most cost-effective way for health ministers to spend their money.
In this blog we consider whether cancer should receive special attention compared to other diseases and public health concerns when allocating scarce public health resources in low- and middle-income countries (LMICs). By “special” we mean:
- priority in consideration of cancer treatments as part of value assessments;
- the criteria used to assess whether a treatment represents good value for money; and
- the reimbursement mechanisms.
We argue that the case is weak under all three aspects. Firstly, initiating priority-setting reforms with cancer technologies—be they diagnostic, surgical, radiotherapeutic, or drugs—could introduce significant distortions in a country’s health system. Secondly, having special rules for cancer has thin backing (if at all) in economic theory or ethics. Thirdly, the experience of using special rules for cancer offers strongly cautionary tales. We also build on reflections arising from a recent cancer policy roundtable organised in Amman, Jordan, under the auspices of the Research for Health in Conflict in the Middle East and North Africa programme.
The share of public money that cancer care receives often tends to be based rather on the number of people with the disease than on the number of patients who could be successfully treated—an approach that we find questionable. Allocating public funds worth 10 percent of a country’s entire gross domestic product to reimburse treatment for two types of cancers, as in Senegal, is another story altogether—one that raises the double issue of spending enough and spending well.
Undoubtedly, cancer care deserves attention. But does it deserve this amount of special attention, so much so as to prioritise the reimbursement of cancer treatments at the expense of treatments for other conditions? The seed of the problem is that cancer treatments are increasingly complex and expensive, they tend to be less cost-effective than treatments for other diseases, and their effectiveness varies greatly. Reimbursing cancer treatments from the public purse is a tall financial order that brings uncertain, or at least difficult-to-ascertain, health benefits. Recurrent and capital expenditures for even basic cancer models of care, covering diagnosis and treatment (be this surgical, radiotherapy and/or systemic therapies), are significant costs to payers, for whom the sheer complexity of the cancer care pathway can seem like an impenetrable “black box.”
Initiating priority-setting reforms with cancer technologies–what could go wrong?
It is tempting to start the institutionalisation of priority-setting mechanisms in the health sector–such as health technology assessment (HTA)–with examining cancer drug treatments, as this is where the budgetary pressures are the highest. However, there are important caveats to such an approach.
Firstly, in most LMICs, technical capacity for generating and synthesising evidence—the heart of any explicit priority-setting exercise—is often limited to a small team of analysts, if it exists at all. There are opportunity costs to this scarce resource: should they be predominantly occupied with examining the value of the latest cancer technologies, most of which will likely have uncertain benefits? Or instead focus on proven, low-cost, but not yet covered (in the respective setting) cost-effective technologies, such as certain vaccines, which are likely to save millions of lives? In the absence of a robust topic selection process—which usually takes years to refine and develop—an early focus on cancer can easily steer the priority-setting process away from high-value, proven technologies.
Secondly, inherent resource constraints along the cancer care pathway significantly limit the value of any single such technology in many LMICs. What is the good of affording the latest cancer drug, if the skilled healthcare staff, the required specialised equipment, and the robust referral pathways are not in place, and if the patients present too late? To continue the example of Senegal, despite the high proportion of the budget allocated to cancer care, there is only one medical oncologist in the entire country. Focusing HTA efforts on high-value technologies can divert attention from strengthening health systems as a whole—such as investing in skilled human resources—without which only a fraction of such technologies’ potential benefit can be realised.
Thirdly, focusing HTA on cancer from the outset can make it even more difficult to judge cost-effectiveness of other public health interventions in settings with insufficient data awareness and priority-setting experience, as is often the case in LMICs. Starting the process with two-tiered decision criteria, in which there is one criterion for cancer (let alone with a high cost-effectiveness threshold), and another criterion for “everything else,” can further exacerbate this conundrum. At a recent cancer policy roundtable with a focus on the Middle East the realisation emerged that it is important to start with a decided focus on building the process, and less on the (initial) decision outcome(s). Early decisions will be difficult to justify anyway, so it pays off to focus early efforts in building domestic capacity for generating and engaging with the evidence. There is growing guidance available at global level, for example on how to structure and report economic evaluations, institutionalise HTA processes, and engage in communities of practice. What may seem as initial failures or slow progress, will gradually create opportunities for cross-stakeholder dialogue, and deliberation around the value of considered health interventions, ultimately leading to the internalisation of the HTA language and streamlining priority-setting processes.
Special criteria: should there be special rules for cancer technologies?
The ethical and economic case for using special rules to evaluate cancer technologies has been examined thoroughly before, and it was found wanting. Briefly, there is no a priori reason why cancer patients should be treated differently from other patients. For example, the end-of-life criterion essentially says that it is justifiable to spend more (per unit of benefit, however defined) on patients near death. But it can be argued that other diseases also “produce” terminally ill patients—multiple sclerosis (MS) is one of many. Should there also be special rules for MS?
The burden of proof falls on advocates of special rules for cancer to show either that social values in a given case—say in a particular country or towards a particular population subgroups, like children, or those terminally ill—favour cancer relative to other conditions. A systematic examination of studies conducted among the UK public did not find conclusive support in favour of this argument, or that the effectiveness of cancer treatments in question justifies the cost.
Special funding instruments: What does the international experience say about using special rules for cancer?
The UK’s experience of the Cancer Drugs Fund (CDF), launched in its first iteration in 2010 and significantly revised in 2016, offers a number of cautionary tales not just for expensive new cancer medicines such as immunotherapy but also for new surgical technologies such as robotics. Firstly, after the first few years of operation it became clear that the fund wasn’t sustainable—not in the least because the rising costs were driven by the average cost per patient, not by the number of patients accessing the drugs. In terms of the benefits for patients, an independent review found that only 38 percent (n=18) of drug indications approved through the CDF (prior to the Fund’s rationalization in January 2015) reported a statistically significant improvement in overall survival, whose median was 3.1 months.
Secondly, it is worth noting that the CDF was initially devised as a temporary solution until a value-based approach to pricing branded medicines was rolled out. Still, it survives to this day and serves as a reminder of the difficulties to disinvest in such a “special fund” once it is launched, locking policymakers in a seemingly endless cycle of revisions.
Thirdly, the difficulties of ascertaining the benefit of such special financing arrangements are not to be taken lightly. Purely from a conceptual standpoint, it is near impossible to predict what would have happened to patients if the CDF hadn’t been in place at all. Much more concretely, the promise of using “real-world evidence” by linking and analysing cancer registry data takes considerable time and effort to be realised, even in a setting with an extensive health information infrastructure such as the UK, not to mention the inherent biases in registry data compared to randomised controlled trials.
The way forward
There is little ground in favour of applying special considerations when considering the reimbursement of cancer treatments from public funds, particularly in resource-constrained settings. Parallel funding structures, lax assessment criteria, and undue priority in value assessments can produce profound distortions in the respective health systems.
When proceeding with priority-setting mechanisms as a means for improving the use of public resources, the case is much stronger for investing in transparent, well-explained, deliberative processes that build on the available evidence and consider it from multiple perspectives, applied uniformly across disease areas. For example, in considering cervical cancer, the disease should be situated as an area for judging technologies for pelvic surgery, set within a context that has implications not only for cancer, but also other pelvic conditions, such as post-partum haemorrhage. In the long-run, such decision processes are poised to strengthen the accountability, legitimacy, and quality of spending decisions, leading to better outcomes for the entire population. Treating cancer as “primus inter pares” is unlikely to achieve this system-wide objective.