html xmlns=”http://www.w3.org/1999/xhtml” xml:lang=”en”>

The Role of Epidemiology in Critical Care

Hannah Wunsch, MD, MSc

Epidemiology is “the study of the distribution and pattern of health-events, health-characteristics, and their causes or influences in well-defined populations.”¹ The epidemiology of critical care is a relatively new field, given that critical care itself is relatively new, having evolved from the frontlines of World War II and the 1952 polio epidemic in Copenhagen.² Until recently, critical care was viewed as an uncommon extension of regular care in the hospital, delivered by the same physicians who provided the care on the wards or even in the clinic. As critical care developed, it became a separate specialty for physicians in the United States in 1986. With its growth has come the recognition that we need to understand the causes, characteristics, and outcomes of illness for patients receiving this highest level of care.

Why Study the Epidemiology of Critical Care?

Understanding the epidemiology of a group, such as critically ill patients, provides important information relevant to improving delivery of care and outcomes for patients. First, it allows for an understanding of resource use and may allow us to anticipate a population’s needs in the future. Second, epidemiology may shed light on trends that cannot be detected on a small scale, such as typical approaches to care or the harm or benefit from an intervention. Practice and patients may vary substantially from one hospital to another, making small studies impossible to interpret. But findings may be quite clear when examined at the population level. Third, epidemiological data may provide relevant outcome information to assist in prognostication and planning for patients and families. Large-scale analyses of patient groups can also allow for greater precision of estimates regarding both short- and long-term outcomes after critical illness.

Defining the Population

This first issue in the epidemiology of critical care is the question of a well-defined population—something that has often eluded researchers in critical care medicine.³ Some definitions of critically ill populations focus on patients’ admission to a specific type of bed, such as one in an intensive care unit (ICU) that is equipped for organ support or one with a specific nurse-to-patient ratio.⁴ The majority of large databases of critically ill patients (and many studies) use admission to an ICU as the defining characteristic of the patient for inclusion. This is a convenient approach, as it provides a concrete way to decide whether a patient meets criteria for inclusion in the study or database. However, this approach assumes that criteria for admission to an ICU bed are universal, such that the population captured will be homogeneous. Such a supposition is often not correct, leading to heterogeneous groups of patients that may include low-risk surgical patients admitted for monitoring, patients in multiple-organ failure, and patients who are dying.⁵ This approach, using admission to an ICU to study critical care, is probably most valid when examining questions regarding bed allocation and resource use.^6,7 Care in an ICU bed is expensive, and the beds themselves may be a limited commodity in the hospital.⁸ The approach is less robust when the goal is to focus on patient outcomes associated with critical illness, as the heterogeneity of the actual patients then may limit conclusions.⁷

An alternative approach to defining a critically ill population is to use organ failure as the major criterion. The organ failure that is probably used the most frequently to identify a critically ill population is the need for pulmonary support in the form of invasive mechanical ventilation.⁹ For many investigators, the need for invasive mechanical ventilation describes the quintessential critically ill patient. However, even here we can find large disparities; surgical patients may require mechanical ventilation purely for postoperative support rather than any specific intrinsic disease resulting in respiratory failure, and the outcomes for such patients may be better than for patients who have acute respiratory failure secondary to a disease process. For example, in a study of patients mechanically ventilated in Ontario, Canada, the 30-day mortality for mechanically ventilated medical patients was 45%, compared with 29% for surgical patients.¹⁰ Despite this limitation, the common pathway of requiring mechanical ventilation does appear to help define a more homogenous group of patients who are at higher risk of death.¹¹

Yet another approach is to combine disease-specific criteria (such as pneumonia) with the need for organ support (such as invasive mechanical ventilation) to ensure similarities in the cause of the critical illness as well as the resulting common treatment pathway (organ support). Many epidemiological studies in critical care focus on such subgroups.^12,13 However, one must trade off certainty regarding the homogeneity of the population against generalizability to all critically ill patients; limiting a study to such a subgroup may fail to provide a full picture of patterns of care and outcomes.

Burden of Critical Illness

Variation seen worldwide in reported rates of critical illness are likely due to a combination of differences in underlying predisposition to critical illness, differences in resources, and differences in culture.⁴ For example, reported rates of acute lung injury range from 18 to 80 cases per 100,000 person-years,^14,15 and the percentage of ICU patients with reported sepsis varies widely across countries.¹⁶ The underlying predisposition of a population to certain comorbidities may vary, which then predisposes people to develop critical illness. For example, when compared with similar middle-aged people from the United Kingdom, Americans were twice as likely to have diabetes and 3 times more likely to have hypertension.¹⁷ Since a disease such as diabetes increases the likelihood of chronic kidney disease and certain infections, this underlying comorbidity may increase the frequency of some types of critical illness in the US population.

As well as the underlying comorbidities of a population, both resources and culture play a large role in determining the frequency of critical illness.⁴ For example, availability of ICU beds may affect the decisions to intubate patients¹⁸ or to offer intensive care at all.⁶ The availability of ICU beds varies dramatically across countries, ranging from much fewer than 1 per 100,000 people in sub-Saharan Africa up to 20 or more per 100,000 people in the United States and some European countries.^3,19,20 Access to hospital care also may play a role. For example, people without insurance in the United States are less likely to be admitted to a hospital than are those with insurance; for diseases that are usually diagnosed in the hospital (such as severe sepsis), such underlying population concerns may alter estimates.²¹

The approach to end-of-life care also drives differences in the epidemiology of critical care.⁶ With an aggressive approach to end-of-life care, many patients may be admitted to the ICU and be “critically ill” before death, whereas another approach may be to provide palliation, thus bypassing the phase of dying that would constitute the critical illness.^22,23 These differences are evident in comparisons across countries. For example, approximately 20% of people in the United States receive intensive care before death compared with only 5% in the United Kingdom.⁶

Related posts:

Water, and Brain–What Should the Intensivil Know? Paradigms in Sepsis Initiatives in Emergency Preparedness Therapy in the ICU: Practical Management Consideration of the Injured Patient: The Enemy of Good Is Better on Sepsis Resuscitation

Stay updated, free articles. Join our Telegram channel

Join