Half of patients transferred to long-term acute care facilities (LTACs) on prolonged mechanical ventilation will die within a year, according to a 2010 review. Only a small minority will ever go home without needing significant caregiver assistance. For those over age 65, the prognosis is even worse.
It seems rude, or even cruel, to convey these grim odds to families desperate for hope after their loved one has (just barely) survived a harrowing critical illness. And maybe that’s why we physicians almost never do — as little as 7% of the time, according to surveys of families. But after watching their loved ones struggle further, and often unsuccessfully, through chronic critical illness (since only a minority ever escape the ventilator at LTACs), and often suffering themselves financially and emotionally, most family members say they wish they had heard a more realistic (pessimistic) prognosis from doctors at the outset.
Optimism and hope are universal human values, and a refusal to acknowledge the likelihood of death often seems hard-wired into our medical care system, and into our doctor-brains, trained by that same system. It seems kinder, and is certainly easier, to tell a family Looks like he’s getting better, day by day, or We just can’t predict these things, than to give our honest, deflating opinion. For a long time, there was so little known about outcomes in chronic critical illness that we could say these kinds of things without being consciously misleading.
That may be changing. Researchers on outcomes after chronic critical illness and LTAC admission are starting to collect enough data that could pull away this cloak of false optimism, at least for some patients. Shannon Carson (UNC-Chapel Hill), Jeremy Kahn (Pittsburgh), Gordon Rubenfeld (U. of Toronto) and friends introduce a simple scoring tool in the April Critical Care Medicine that can identify some chronically critically ill patients with a high likelihood of death within one year.
What They Did
Authors performed a retrospective cohort study that included 260 adult patients requiring at least 21 days of mechanical ventilation after critical illness. Patients were selected from the tertiary medical centers of U. of Washington, UCSF, Denver Health, U. Penn, and Duke and included medical, surgical, neurological, and trauma patients. Most required mechanical ventilation for one month or more.
By analyzing patient variables, lab values, hospital course, and outcomes, authors used logistic regression to isolate variables that, when present at 21 days, would independently predict mortality at 1 year.
What They Found
Among the 260 patients mechanically ventilated for > 21 days,
- 28% died in the hospital;
- 12% were discharged home;
- Mortality for the entire cohort was 48% at one year.
Interestingly, this suggests that two-thirds of the patients going to LTACs lived for at least a year, beating the expected mortality of ~50%. My math: consider 100 patients; 28 die in hospital, 12 go home, leaving 60 are to go to LTAC, and since at one year only 20 more of the entire cohort have died, totaling 48 (28+20=48% mortality), that means 40 of the 60 LTAC patients survived in this most-pessimistic scenario (because this assumes none of the patients discharged home died; deaths in their group would improve the LTAC odds further).
They identified four variables that (if present on the 21st day of mechanical ventilation) predicted death within one year, and fit them into a scoring model that weighted them thusly:
- Age >= 65 years: 2 points
- Age 50-64 years: 1 point
- Platelets <= 150,000 on day 21: 1 point
- Vasopressor use on day 21: 1 point
- Hemodialysis at 21 days: 1 point
Adding up a patient’s points gets you what they have coined the “ProVent score” (not to be confused with Provent, the up-and-coming nasal adhesive for obstructive sleep apnea), with observed mortality as follows:
- 0 ProVent points: 20% observed 1-year mortality
- 1: 36% 1-year mortality
- 2: 56% 1-year mortality
- 3: 81% 1-year mortality
- 4 or 5: 100% 1-year mortality (only 14 patients had these scores)
This was an observational cohort, not a “validational” cohort. I imagine the authors are planning on testing their new “ProVent Score” prospectively on a cohort of chronically critically ill patients, and it would be interesting to see how well the model performs.
Although relatively small in scale, this work represents an important step toward better prognosis of chronic critical illness, prolonged mechanical ventilation, and LTAC placement. These issues are growing in importance socially, economically, and politically, but the lack of “revenue upside” in terms of drug and device development has meant researchers must work with a relatively small pool of research funding from public and philanthropic sources.
Carson SS et al. A multicenter mortality prediction model for patients receiving prolonged mechanical ventilation. Crit Care Med 2012;40:1171-1176.