Deciding who to admit to a critical care unit

Abstract

Scarce resources may cause doctors to be pessimistic about prognosis and refuse critical care admissions

Chronic obstructive pulmonary disease (COPD) is a leading cause of morbidity and mortality worldwide. The incidence of COPD is rising, and the World Health Organization estimates that it will be the fourth leading cause of death globally by 2030.¹ In this week's BMJ, Wildman and colleagues report the differences between actual survival and survival predicted by a doctor in people with asthma or COPD admitted to intensive care.² This is an important matter to investigate, because people with asthma and COPD who have acute exacerbations that require admission to intensive care have high short term mortality.^{3 4}

Using data from 832 admissions for asthma or COPD in 95 intensive care units and high dependency units in the United Kingdom, the authors found that predicted survival was lower than actual survival (49% v 62%) 180 days after admission. This “prognostic pessimism” was present in the overall sample and for most subgroups of people. The absolute difference between predicted and actual survival was >30% in people with the lowest predicted survival. The authors suggest that the scarcity of intensive care resources in the UK may contribute to doctors' inaccurate predictions of survival because such prognostic pessimism may stop them feeling that they are denying treatable patients potentially life saving treatment. Is such prognostic pessimism a disease that needs treatment (by improving doctors' prognostic skills) or a symptom of an underlying problem with the healthcare system, such as scarce intensive care resources?

Decisions about the use of life sustaining treatment are complex, imprecise, and need to balance the potential risks and benefits to each critically ill person. Predicting the probability of short term survival is important when assessing the benefits of intensive care. Despite knowledge of important prognostic factors,³ previous studies have also shown significant variability in doctors' estimates of survival for people with an exacerbation of COPD who need mechanical ventilation.^{5 6}

Mortality should not be the only consideration when deciding about admission to intensive care. Providing doctors, patients, and families with more accurate estimates of survival during serious illness did not strongly influence the medical decisions made in a large study from the United States.⁷ Quality of life after intensive care is an important consideration also,⁸ especially as—for instance—nearly 90% of seriously ill people would rather die than survive with severe cognitive impairment.⁹ These factors may have had an effect on doctors' predicted prognosis, but this cannot be determined on the basis of data provided in Wildman and colleagues' study.¹ Like predicting patient mortality, the ability of doctors and nurses to predict quality of life after intensive care is unsatisfactory.¹⁰

Making decisions about admission to intensive care is even more complex than determining the benefits and risks to an individual patient when resources are scarce. This may be especially relevant in the UK and southern Europe, where intensive care beds are often lacking.¹¹ The authors speculate that in the face of chronically scarce resources, doctors may develop prognostic pessimism, which leads them to refuse seriously ill patients admission to intensive care. A study comparing admission to intensive care in Canada and the US reported that Alberta had 50% fewer intensive care beds per capita than did western Massachusetts. In the Canadian setting, admission to intensive care was more often denied to elderly patients with chronic medical conditions who were thought unlikely to benefit from such care.¹² Although this illustrates rationing of intensive care on the basis of the availability of resource in Canada, it is unclear whether prognostic pessimism was a factor in the decision making process. Furthermore, the study found no significant difference in hospital mortality despite rationing of intensive care— hospital mortality was not reported in the study by Wildman and colleagues.¹

Future studies of doctors' prognostic accuracy in jurisdictions with fewer limitations in intensive care resources may allow Wildman and colleagues' work to be interpreted within a broader context. This will determine whether prognostic pessimism requires intervention aimed at doctors or at underlying healthcare systems that have inadequate provision of critical care services.