What are time-limited trials in critical care?

ATS and ACCP elevate the still-undefined concept

Jun 10, 2024

Background

I’m not sure if you’ve noticed this, but patients and families frequently disagree with their physicians about the likely outcomes from critical illness.

For example, among patients undergoing tracheostomy for prolonged respiratory failure, when ICU clinicians predicted only a 6% chance of good functional status at one year, surrogates estimated it at 71%. (Physicians were too negative by one-third: the actual result was 9%.)

Although physicians' prognoses are consistently more accurate than family members', physicians can often be overly pessimistic early in the course of critical illness. Over longer periods of intensive care, clinicians’ prognoses (and those made by machine learning models) tend to increase in accuracy, but are far from perfect.

When surrogates disagree with physicians' prognoses, they often state it's due to a need to maintain hope, that the physician underestimates the patient's resilience, or that supernatural forces will intervene to heal their loved one. In other contexts, we might celebrate such expressions of loyalty, support, and faith. No one is (nor should they be expected to be) “objective” when it comes to a loved one’s possible imminent death.

In the ICU, these disparities in belief and perspective often lead to disagreement between care teams and families over the appropriateness of intensity and duration of care. Whether openly acknowledged or not, these conflicts profoundly affect the care relationship: clinicians who feel compelled to provide ICU care they consider inappropriate (i.e., harmful) experience moral distress, which may manifest as frustration and anger and contribute to burnout syndromes.

Families may perceive ICU teams' questioning the intensity of care as an attempt to limit treatment or as excessive negativity. The physician-surrogate relationship can easily break down, destroying any prospect of productive shared decision-making.

But even in the most mutually respectful and understanding circumstances, what starts out clearly as care in the ICU can over weeks take on a netherworldly, even gruesome tinge. Care teams feel as if they have invisibly slipped into morally problematic terrain, inflicting far more suffering than they relieve, upon a helpless fellow human who is becoming progressively less recognizable as the indignities and ravages of end-stage critical illness erode her personhood. Is this really what she would choose?

“She’s a fighter,” her family might reassure us. “And the white count is down today—that’s a good sign, right?”

Time-Limited Trials of Critical Care

Time-limited trials of intensive care were suggested at least as far back as 2009 as a method of bridging the divide between clinicians and families' perspectives and expectations for ICU care.

In its deceptively simple conception, a patient's surrogates and her physician agree on a period of intensive care, and reassess after that to determine whether continued ICU care or cessation of life-sustaining measures would be more appropriate.

A suggested schema for time-limited trials was suggested by Quill and Holloway in 2011 in JAMA. They proposed disease-specific intervals for TLTs (e.g., 3 days for anoxic encephalopathy to regain pupillary responses after cardiac arrest, 4-7 days for a DNR/DNI patient to receive NIPPV for COPD exacerbation, etc.) that they invented more or less based on opinion and gestalt (wide flexibility and discretion were also advised).

Time-limited trials were reviewed in 2018 in Intensive Care Medicine by Vink et al, and in 2024 in Chest by Kruser et al. While valuable, these reviews also highlight how little literature exists, and its inadequacy to encompass this massively complex and emotionally fraught subject.

PubMed search for “time-limited trial” (43 results 2009-2024)

Studies of Time-Limited Trials in Critical Care

Chang et al (JAMA Int Med 2021)

A before/after prospective cohort quality improvement study was performed at three public hospitals in Los Angeles between 2017 and 2019 (Chang et al, JAMA Internal Medicine 2021). They enrolled 209 patients with advanced chronic illness whom physicians believed to be at risk for futile or harmful ICU interventions, but whose families disagreed and desired aggressive critical care. Clinicians were trained in a structured communication approach offering time-limited trials of ICU care.

In the post-intervention period, there were more family meetings (96% vs 60%), with more discussion of risks and benefits of ICU care (95% vs 35%) and patients' preferences (98% vs 47%).
ICU length of stay fell by a median of 1.3 days, with fewer patients receiving mechanical ventilation or invasive procedures, and with no difference in hospital mortality.

Lecuyer et al (“The ICU Trial”, Crit Care Med 2007)

In a prospective cohort study at one hospital in Paris, 188 cancer patients who developed critical illness midway through their treatment course were admitted to the ICU, with a reappraisal of the intensity of care on day 5.

Only 41 patients in total survived to hospital discharge (21.8%).
Patients who survived five days of mechanical ventilation had an overall survival rate of 40%, however.
Prognostic accuracy improved at day 6, compared to day 3 (using organ failure scores to predict mortality).

There was no control arm, limiting the strength of any conclusions from the data.

At this writing, that’s it: two studies, 14 years apart. (The Chest review identified 18 “empirical studies,” but only these two tested TLTs’ effects on patient outcomes in the ICU.)

ATS Workshop Report on TLTs

In 2024, the American Thoracic Society issued a “workshop report” to define and explore time-limited trials of critical care. This emerged from a Delphi process (a formal iterative method of generating consensus) among >100 stakeholders and 27 committee members, chaired by Dr. Kruser (of the Chest review).

This was only a preliminary effort and wasn’t intended to generate a new consensus or guideline statement. That said, their efforts further underscored the thorniness of the TLT concept and its implications. Committee members couldn’t reach agreement on almost anything (by their preset consensus thresholds).

The most agreed-upon objective for a time-limited trial (at >75% consensus) was "to increase ... readiness to make decisions about limiting or withdrawing life-sustaining interventions."

The unwieldiness and impracticality of the committee’s proposed 16-component schema (or was it 18?) for implementation of TLTs illustrate the challenge of operationalizing the staggering variety of clinical and socio-emotional situations that arise during (what is often) end-of-life critical care.

But Wait …

There are numerous issues to discuss about the TLT concept (how is it different from what we already do, i.e., “let’s reassess after a few days”?; isn’t this at heart just a dressed-up, formalized attempt to pressure more families to withdraw care sooner?; what timeframes would be proposed when we know our prognoses are rough estimates at best?; who has time for all these family meetings?; etc.)

But this post is already too long.

Conclusion

Prolonged, excessive, futile critical care hurts people.

It hurts patients, it hurts their families, and it hurts the people providing the care.

But there’s no alarm that beeps when well-intentioned care with an indeterminate prognosis crosses into something darker.

Even when it seems clear that threshold has been passed, it’s hard for families to stop. It’s hard to let go. It’s hard to say goodbye.

And for many, it feels abhorrent and inconceivable to be the one to say, “stop.” To—it feels like—no matter what the doctors say—cause her death with those words.

More and more patients in the ICU are crossing that unseen threshold—we all sense this.

Frontline clinicians have been leading teams, counseling families, and doing the best for their patients through these increasingly challenging situations. They do so without the needed training or institutional support, relying only on their experience, instincts and personal emotional resources.

Some days, the role feels more like priest than physician.

And so ACCP and ATS signaling interest in and elevating the concept of time-limited trials of critical care is a significant development.

It’s preliminary, it’s rudimentary, but it’s a sign they take the issue seriously, that they will engage despite the challenges.

It’s a start.