Everyone knows that when you're really sick, going to the ICU–that place with all the beeping monitors, hypervigilant staff, and high-tech invasive gadgetry–might save your life. But what if in certain situations, transfer to an ICU bed could actually worsen a person's condition, or even hasten their demise?
Hannah Wunsch of Columbia U. suggests it's possible, asking “Is There a Starling Curve for Intensive Care?” in the June 2012 issue of Chest. The Starling curve, you recall, describes the response of the heart as it is filled fuller and fuller with blood: it gradually remodels, expanding to be able to accommodate the higher volume, with improved results initially. Eventually, however, its increased size becomes detrimental, and heart failure results.
There is no question that up to a certain point, “more is more” when it comes to the number of available ICU beds: look no further than to the thousands of people in developing countries who die each year from curable respiratory failure for lack of access to mechanical ventilators. Closer to home, evidence suggests that in the United Kingdom, which has only a fraction of the ICU beds per capita of the United States, delay or denial of entry to the ICU delay or for lack of beds occurs relatively frequently, and is strongly associated with increased 90 day mortality. As England's National Health Service aggressively built out ICU capacity throughout the 2000s, ICU mortality held steady and then markedly fell, suggesting a salutary effect (although other quality improvement initiatives may have contributed).
On the plus side of ICU bed shortages, only 5% of citizens in the UK die in or near an ICU, compared to 20% in the US. Intensive care at the end of life is widely regarded as uncomfortable by health professionals (who do the poking and prodding) and undesirable by most lay people, most of whom say in surveys they would prefer to die at home.
Some have theorized that excess capacity of ICU beds in the US could encourage overenthusiastic, “hair trigger” transfers to the ICU, placing patients at increased risk for unnecessary procedures, testing, drug therapies, and subsequent complications. While this certainly must happen on occasion, there's little good evidence to support the claim of ICU transfers broadly causing excess harm. It's true that most ICU transfers in the US are for short periods of observation only, with only a minority of patients requiring high intensity support like mechanical ventilation ordered vasopressors. But most of these short ICU stays seem to be harmless and pass without incident–or at least, no systematic increased risk of harm has been convincingly demonstrated thus far.
One intriguing signal is the possibility that low ICU bed availability could provide the impetus for more rational end-of-life care discussions. In a recent Canadian study, lack of ICU beds frequently led to a change in goals of care, hopefully in better harmony with the patient and family's desires. When the patient is acutely ill on the medical ward, delete, and ICU beds are available, it's much easier to “transfer first, ask questions later” then to open difficult discussions regarding goals of care and end-of-life preferences. Good evidence also suggests that in areas of the country with high medical care spending at the end of life, patients with advance directives seem to be less likely to experience expensive and undesired ICU care.
Is there a “right” number of ICU beds? Dr. Wunsch:
We do not yet have the sophistication to quantify the many risks and benefits of intensive care and to decide the relative weights in the trade-offs among additional survival, complications, costs, and care experience...[T]he United Kingdom assessed its care across these many domains and concluded that it needed more beds. Up until now, there has been little assessment of what is gained or lost with the relentless addition of ICU beds in the United States.
My own answer would be a simple “no.” It's not a question of a lack of sophistication in research techniques; it's an issue of value judgments, which research can't provide. The heart, you see, only has to pump blood for one person, and its efficiency in doing so can be more or less certainly quantified. In a pinch, you can just look at the whole person and tell the pump's working.
Our national patchwork of ICU beds in the US, on the other hand, serves 300 million people in a multitude of diverse settings, each with its uniquely local culture, infrastructure issues, and competitive environment. At whatever total number of ICU beds we have at any point in time -- for each health system and nationally -- there will be winners and losers, financially and in terms of health outcomes. Hospitals' "relentless addition of ICU beds," while hopefully influenced by a desire to deliver excellent care, is largely driven by financial incentives (including increased efficiency and throughput), and negotiated at a purely local level, almost always without policymaker input. That's how we do things here in the land of the free (for better or worse), and that doesn't seem likely to change anytime soon.
Wunsch H. Is There a Starling Curve for Intensive Care? Chest 2012;141:1393-1399.