What is a hospital bed day worth? A contingent valuation study of hospital Chief Executive Officers

An effective and efficient hospital-based program will simultaneously both improve patient outcomes and decrease hospital length of stay. Decreasing length of stay and freeing up, or releasing of hospital beds, represents a cost saving. However, the magnitude of this cost saving is difficult to quantify and is likely to depend on many factors. Hospital beds can have two types of value: (1) how much they cost the hospital to run – the accounting cost (referred to as the hotel cost by the WHO [1]), and (2) the value they have in terms of achieving desired outcomes – the economic (or opportunity) cost. Previous cost-effectiveness analyses most often use the accounting cost of some variant of it [2]. This is predominantly because it is an easier value to calculate and to understand, especially by hospital administrators.

The question that arises is to what extent this is the “right” value to be adopting? Arguably, the second value, the economic cost, is the real value we “should” be interested in. This is because the majority of bed costs are fixed and sunk costs, and therefore the “true” value of releasing a hospital bed is better captured by the extent to which it allows one to achieve other outcomes that the hospital desires, such as treat another patient, reduce waiting lists, and ultimately meet economic targets. An accounting value fails to represent economic opportunity costs, rather accounting conventions are used to recover historical expenditures. WTP may be closer to the economic opportunity cost, and that opportunity cost is a basis for decision making, the efficiency of which is the subject of welfare economics. The true economic value of a bed day is probably lower than the full financial accounting cost. The question is, how much lower?

Previous cost-effectiveness research [3] has shown that the dollar value placed on a hospital bed, average cost versus just consumables, is important. Sensitivity analysis around the bed day value demonstrated that it had a significant impact on the overall decisions derived from the cost effectiveness model. Bed day values often drive the cost savings for cost-effectiveness outcomes and thus if we overvalue bed days, by using arbitrary accounting conventions, programs can seem more cost-effective than they really are. For example, bed days saved add to the costs savings side of a cost-effectiveness analysis and thus a reduction in bed day values will lower costs savings, and assuming all other values are unchanged, will decrease the likelihood of an intervention being cost effective. Specifically, it we use a hypothetical value of AUD $1000 per bed day, and intervention?“X” results in 100 bed days saved, we have a cost saving of AUD $100,000 (in addition to any other costs savings). If,?however, the bed day valuation is lower at AUD $200 we have only a cost saving of AUD $20,000. Therefore, an intervention needs to be less costly or more effective, to offset such a difference. In clear cut cases and with low uncertainty around other model parameters this difference in bed day valuations may not be important. However, arguably there are many cases where there is uncertainty in other parameters and such a fluctuation in values will meaningfully impact on the conclusions to be drawn from the cost effectiveness analysis. The extent to which bed days represent the main portion of the cost savings will directly relate to their overall impact on the outcome. Therefore, there is a need to obtain more accurate measurements for this parameter.

Bed days are a very important cost element in any analysis involving stays in healthcare facilities, therefore an understanding of their value will be important to decision makers trying to maximise the health of patients within a fixed budget. An economic evaluation of beds days could be useful for decision makers because it would help them to maximise patient outcomes and achieve efficiencies in resource use. Longer term it might also serve an educational purpose in that it could enable a shift in current thinking about the resource away from purely financial fixed accounting costs to one where opportunity costs are highlighted to enable sustainable efficiencies.

One can consider two alternative perspectives when thinking about the valuation of hospital beds. First, one could take a broader perspective of the healthcare decision maker who manages waiting lists, and for whom there is a real economic benefit in releasing a bed day for another patient to occupy or for the resource to be used differently. The second perspective is narrower and could be that of a manager working within an Intensive Care Unit (ICU) or hospital. Therefore, who we ask and over what period of time is important.

Previous estimates which have taken a broad perspective include a detailed costing study of an Australian ICU [4] and an economic evaluation which examined spending patterns for Australian public hospital services [5]. Estimates from these studies gave an ICU bed a value of approximately AUD $2600 and a general ward bed a value of AUD $800. These estimates use the average costs over the length of the hospital stay, in a given 12 month period, and do not represent the marginal cost. They include both fixed and variable costs, hence they are unlikely to accurately represent the opportunity cost of this resource.

The alternative perspective considers only the variable cost per bed day. Variable costs are the cash savings that budget holders within the hospital can recover if bed days are not used; they include items such as fluids, dressings and pharmaceuticals. An important issue with the narrow perspective is that costs per bed day decrease over the duration of hospital stay, particularly so for acute beds [6]. We know from previous research that a large proportion (over 80%) of hospital costs are fixed [7] and therefore this variable rate is useful in that it shows what direct savings could be achieved by freeing up a bed. Kahn [8] argues that the cost of an ICU bed is overestimated because of using average costs and not correcting for fixed costs. They estimate that the direct cost savings for the last day of an ICU bed is US$379. In their work they do not measure opportunity costs, which is the value that could be achieved through some alternative use of the resource. Nuti [9] has also demonstrated that some fixed bed costs can actually be considered quasi-fixed on the basis of an alternative use of a bed, the amount of costs recoverable depending on number of beds and the type of intervention. This is exactly the opportunity cost we are interested in measuring. Recent work [10] conducted in Europe using a subset of our methodology has shown also that there is considerable disparity between accounting and economic costs of hospital beds.

For this research we were interested in the broader perspective of the healthcare decision maker reasoning that the Chief Executive Officer (CEO) could, and often does, choose to use this resource for another use when it is released. Therefore, we are interested in the marginal opportunity cost of a hospital bed day. Specifically, we want to know how much CEOs would be willing to pay to free up a bed day in their hospital so they could use it for another purpose, which more accurately represents the choice they actually make. We are also interested in the extent to which this value is comparable with previous cost estimates. To do this we use a contingent valuation method of willingness to pay (WTP) via a direct survey of experts. Whilst we acknowledge this method has some caveats, we chose this approach for two reasons; first because there is no market for such a resource therefore gauging its value through market mechanisms is unrealistic. Second, contingent valuation is a commonly used method in the health domain to elicit values for health gains [11, 12].

This analysis was conducted in the general context of the Australian public health care system and specifically it was performed as part of an overall evaluation of the Australian National Hand Hygiene Initiative (NHHI). Australia’s public hospital system, which provides the majority of acute-care beds, affords free access to hospital care for public patients. It is jointly funded by the Australian federal Government and state/territory governments. However, the public hospitals are run by state and territory governments, and often the CEO of each hospital has discretionary power over how services are run and what investments are made locally. Australian Government funding to the states and territories for public hospitals is made through the National Healthcare Agreement and the National Health Reform Agreement between the Australian Government and the states and territories. One of the main funding models is Activity Based Funding (ABF) which is a way of funding hospitals such that they get paid for the number and mix of patients they treat. If a hospital treats more patients, it receives more funding. However, there is large heterogeneity in funding models and fragmentation of both funding and delivery is the defining characteristic of Australia’s health system [13]. As a result, there are perverse incentives in the system and often this means that decision makers are not incentivised to opt for value-based solutions. Instead doctors and key decision makers in hospitals are often working to maximise revenue and profit. With such power and incentives the value of resources (i.e. beds) could and should be seen as opportunities in which to achieve greater efficiencies.

Within this funding context, the NHHI is a major patient safety programme co-ordinated by Hand Hygiene Australia (HHA) and funded by the Australian Commission for Safety and Quality in Health Care. This program, which commenced in 2009, was designed to improve hand hygiene compliance in every hospital in Australia. It used the WHOs “five moments” program in order to standardise both the training of healthcare workers, and the measurement of compliance, as well as to use a standard definition of Staphylococcus aureus?bacteraemia infections.

We performed a comprehensive evaluation of both the costs [14, 15] and efficacy [16] of the NHHI, along with a full cost effectiveness evaluation [17]. As part of this evaluation we needed to obtain accurate estimates of the cost savings of the program as determined by the number of hospital beds that were able to be released from preventing infections as well as their value [18]. Because most of the costs of running a hospital are fixed in the short term [7], and the incidence of hospital-acquired infections in Australian hospitals is relatively low, only small cash savings will be made from reducing rates of healthcare associated infections with the real cost of healthcare associated infections being the value of the marginal bed day released to some alternative use. It is of paramount importance to get a good estimate of the economic value of releasing a bed, in this case one linked to a reduction in infections. This was the principal aim of the current study. For decision makers, including hospital administrators, the results from this economic evaluation represent whether the program is good value for money and whether they could achieve efficiencies by better investing their resources in another program.

This paper will discuss how we measured the economic value (opportunity cost) of a bed in the context of the NHHI and discuss the implications this has for evaluating the cost-effectiveness of any program or treatment whose intended outcome is to release hospital beds. We also consider the extent to which this value is useful for assessing cost savings.