When the government introduced the lockdown in March the rationale was to slow down the rate of Covid-19 infections so that our healthcare system could be readied to deal with an anticipated surge of patients, and so not be overwhelmed.

Many commentators have pointed out that the Covid-19 pandemic was both a public health crisis and an opportunity to begin the process of bringing health resources into a more rational and equitable system.

In this regard, the government is committed to the National Health Insurance (NHI) policy as a vehicle to achieve universal healthcare. And for some time now the health science community has identified South Africa’s divided and unequal healthcare system – wherein some 80% of the people use the under-resourced public healthcare system – as a barrier to preparing the healthcare system for the surge in infections of Covid-19.

We know that as the pandemic progresses, the demands on the healthcare system will intensify and will result in critical shortages of hospital beds, intensive care unit (ICU) beds, ventilators, and medical staff in the public sector. Existing healthcare needs for non-Covid-19 related emergencies may also be compromised by having resources redirected to Covid-19 related needs (a process that health professionals have identified as already happening).

Since the lockdown, the government has, however, sought to address one immediate problem – procuring ICU beds that come accompanied by ventilators – from the private hospital sector. This differs from countries such as Ireland and Spain where governments temporarily nationalised private beds into the public service. This process of procurement raises all sorts of questions of terms, prices, duration, compensation and so on.

But with the increased demand for government resources for economic relief and other measures, the question may be posed whether this process of procurement is the best way to prepare for the coming surge. This is both a matter of using available public resources optimally – as well as reflecting on whether this is the best way to achieve the desired health outcomes, given the need to make choices.

In this article, we examine three questions:

• Should government be using resources to buy beds from the private sector?
• If so, are ICU beds the best choice – economically and in health terms?
• And should hospital beds be the focus at all?

But, in looking at these we are forced to touch on what ethical principles will underpin our choices.

Framework for allocation of resources

Approaches to decision-making related to the allocation of resources in the face of scarcity commonly include five fundamental values or ethical principles:

1. Equality;
2. Favouring the “worst off”;
3. Utilitarianism – doing the greatest good for the greatest number of people, either by saving the most number of lives and/or saving the largest number of life-years;
4. Promoting and rewarding social usefulness; and
5. “Rule of rescue” – ie, to provide urgent, life-saving treatments “whatever the cost”.

Substantive work has been done in South Africa and by the global health community to assist policymakers in navigating the difficult choices that need to be made in setting health priorities. The World Health Organisation (WHO) has published a “Making Fair Choices on the Path to Universal Health Coverage” report. While we are not in a position to determine which values matter more than others in South Africa’s response to Covid-19, the question we focus on here is the extent to which the allocation of resources towards purchasing private sector ICU capacity satisfies utilitarian values, for which the concept of cost-effectiveness is critical.

ICU beds

There are a range of healthcare interventions to manage the progression of the Covid-19 pandemic. Resources are required to carry out screening, testing, isolation and contact tracing programmes, provision of personal protective equipment (PPE) to health workers, treatment in general/high care wards, and in the most critical cases, treatment in ICU.

A major focus has been whether South Africa has sufficient ICU capacity. The shortfall in ICU beds was quantified in a presentation to the Portfolio Committee on Health on 10 April 2020 where the peak daily demand for ICU beds in the country was projected to be between 4,100 beds (optimistic scenario) and 14,767 (pessimistic scenario) against an ICU bed availability of 3,318 of which 2,140 were in the private sector.

This means that, based on the projected progression of Covid-19 in South Africa and expected utilisation rates of ICU in the management of Covid-19, while there may be sufficient supply of ICU capacity in the private sector for those with medical scheme coverage, there is insufficient supply in the public sector.

ICU care: Costs and outcomes

Intensive care services are very expensive and are one of the largest drivers of hospital costs in the private sector. Even in public hospitals the hospital cost per ICU patient day is estimated to be in the region of R17,000.

Overprovision of ICU beds and subsequent cost escalation through potentially inappropriate use in the private sector was a key finding of the recent Competition Commission Health Market Inquiry. Somewhat ironically, this unfettered ICU-driven cost escalation represents one of the biggest risks to the current medical scheme industry. Discovery Health, the largest medical scheme in the country, recently reported that the average cost of all Covid-19 hospital admissions across its members was R84,708. The average cost of an ICU admission was substantially higher – at R169,525 – and these admissions were also reported to have the “highest variation in cost”. Costs for one member currently still in ICU have already reached R800,000.

While intensive care is expensive, if it were effective at preventing deaths in critically ill Covid-19 patients, it would present a potentially cost-effective use of resources. However, existing evidence unfortunately indicates poor outcomes. A meta-analysis across available publications found a mortality rate of 53% (range 28%-88%) for critically ill patients managed in ICU. This does not mean that ICU is an ineffective intervention, but it does indicate that for a substantial proportion of severely affected Covid-19 patients, admission to ICU is unlikely to be a life-saving intervention, and at best, will only buy a few more days of life at significant financial cost.

Based on a utilitarian approach, there would appear to be little merit in government investing heavily on expanding ICU capacity. Indeed, given the looming shortage of public-sector hospital beds in general during the epidemic surge, a utilitarian ethic suggests that the greatest good for the greatest number may be achieved through investing in field hospitals.

Against this backdrop, MOSAIC – a health economic modelling collective established to provide rapid policy guidance and public engagement for the South African Covid-19 response – carried out an analysis of the cost-effectiveness of ICU in comparison to limiting treatment for these patients to general wards only. The cost-effectiveness of the two approaches hinges on two key factors: the relative difference in cost and the relative difference in effectiveness (measured by the number of what is called “disability-adjusted life years [DALYs]” and the number of deaths averted) for the two approaches. The difference in costs between the two approaches is substantial – the cost for patients limited to general ward treatment is estimated to be R77,000 per case, compared to a cost per case of R102,000 if ICU care is included.

Using best-case assumptions about the relative effectiveness of ICU in comparison to no ICU care suggests that, at most, it would avert 0.40 DALYs and approximately seven in 100 deaths. Put together, the cost-effectiveness of ICU care versus treatment in a general ward is R65,000 per DALY averted. Is this a “best buy” or a good use of available resources?

Appealing to a utilitarian framework, we would consider the opportunity cost of that spend – what else could be purchased and how much health would be gained if that money were used in other ways? Recent landmark work from colleagues at the University of Witwatersrand analysed the historical marginal productivity of South Africa’s public health system, and estimated that on the margin, a spend of R38,465 would avert one DALY within the South African population.

Estimating marginal productivity in any health system is always fraught with uncertainties, but it does provide us with a very useful ballpark for whether a potential investment is a good use of money. To be cost-effective, ICU care would need to be at least 50% more effective than management in a general word. Unfortunately, this is highly unlikely to be the case. Simply put: under a utilitarian framework and based on our estimations, either ICU care in the private sector needs to get a lot cheaper, or a lot better at saving lives for us to be convinced that it represents a good investment relative to the myriad of other potential Covid-19 and non-Covid-19 health investments that the government could make on behalf of the people of South Africa.

How do we proceed?

Based on a utilitarian approach, there would appear to be little merit in government investing heavily on expanding ICU capacity. Indeed, given the looming shortage of public-sector hospital beds in general during the epidemic surge, a utilitarian ethic suggests that the greatest good for the greatest number may be achieved through investing in field hospitals. Moreover, a utilitarian approach requires equal care for those in need of inpatient care for other conditions that are currently crowded out by the provisioning for Covid-19.

However, justification for investing in ICU capacity may be found through the application of other ethical principles such as the “rule of rescue”. This rule calls on society to respond to the extreme risk faced by an identifiable individual, however, it generally only holds when the numbers of patients are small. The difficulty to face with Covid-19 is that while many of us may align with a rule-of-rescue based response, the stark realities of resource constraints call on us to acknowledge that the allocation of resources to build ICU capacity:

1. Would be unsustainable given the costs; and
2. Would lead to a greater loss of life.

If, despite the above, government proceeds to negotiate buying hospital capacity from the private sector, the purchasing arrangement should be structured in a manner that manages the financial risks very carefully. Given the findings of the Health Market Inquiry on supplier-induced demand in the private sector, it is critical that in contracting with the private sector, government does not contract on a fee-for-service basis but manages the financial risk and overall expenditure by transferring and sharing risk with private providers.

The differences in care guidelines and protocols between the public and private sector would also need to be addressed. In its purchasing arrangement from the private sector, the government needs to ensure that:

1. A uniform set of guidelines apply so that the treatment of patients in the public and private sectors is aligned; and
2. That appropriate structures are put in place to ensure that individual healthcare workers do not have to make the difficult decisions on whether to provide care or not.

At a broader level, it needs to be recognised that the utilitarian ethic of doing the greatest good for the greatest number of people is not the only value that should be considered in allocating resources. Experience globally suggests that a multi-ethical framework taking into account context and resource constraints is more likely to result in a fairer allocation of resources. Processes for doing this need to be established, not only for the case of ICU and other care in the Covid-19 era, but for resource allocation more generally.

We also note that the evidence base relating to many features of the Covid-19 pandemic is changing rapidly, making estimations of costs, utilisation, and intervention effectiveness difficult. We will continue to update the model structure and parameters as new information comes to hand and we invite constructive input on how the analysis can be improved and made more useful for informing policy. DM

Geetesh Solanki is specialist scientist at the Health Systems Research Unit, SA Medical Research Council (SAMRC) and an honorary research associate with the Health Economics Unit, University of Cape Town; Tommy Wilkinson is writing in his capacity as a researcher at the Health Economics Unit, UCT; Cynthia Tamandjou Tchuem is a virologist and researcher at the Health Economics Unit, UCT; Sumaiyah Docrat is a health economist at the Alan J Fisher Centre for Public Mental Health, UCT; and Susan Cleary is an associate professor at the Health Economics Unit, UCT.

The opinions expressed here are those of the authors and do not necessarily reflect the views of their institutions.