Prolonging Life or Prolonging Death? Two Doctors on Caring for the Critically Sick


February 13, 2015

For doctors, there are rarely easy decisions to be made when it comes to the care of someone who is critically ill. In some cases, a bleak prognosis may require them to advise a patient against continuing their treatment. Other times, they may see cause for optimism, however remote it may seem. Knowing how or when to make that determination is the tricky part.

As writer and renowned surgeon Atul Gawande examines in the recent FRONTLINE documentary Being Mortal, this means doctors often struggle with how to talk to patients about death and dying. In his best-selling book of the same name, he notes that 40 percent of oncologists have admitted to offering treatments they believe are unlikely to work.

We wanted to hear more from doctors who confront this problem every day, so we invited two to join us for a conversation about treating patients who are nearing the end of life. Michael Belkin is chief of vascular and endovascular surgery at Boston’s Brigham and Women’s Hospital and a professor of surgery at Harvard Medical School. Nicholas Sadovnikoff is co-director of the hospital’s surgical intensive care unit (ICU) and also teaches at Harvard Medical. Here’s what they had to say.

Dear Mike:

It is hard to know how to start this discussion, but I thought it would be fruitful to go back to my beginnings as a physician.

As a medical student, I had meticulously avoided the ICU and the sick patients to be found there. I thought them to be too complicated and was intimidated by the high-stakes decisions and interventions they required. It was ironic then that I found myself assigned to the ICU for my first month of internship; I was petrified.

Fortunately, that feeling didn’t last long. I quickly became enamored of the pace, the procedures and ultimately even the sense of the stakes. All of the physiology and pharmacology I learned in medical school was suddenly richly relevant and constantly subject to our monitoring, interventions and manipulations. I loved that we could seemingly snatch patients back from the brink of death, and then with the correct support and treatments, restore them to a functional life. There were few experiences as gratifying as having a patient, previously gravely ill, come walking back through the doors of the ICU to visit and offer thanks. The French word for critical care is réanimation, which I think more accurately reflects what we attempt to do in the ICU.

But even as I celebrated our successes that first month, I also saw that there were patients whom we simply could not save. They remained perched perilously on the precipice of life and death. With our life-sustaining interventions, we could keep some from falling off, but we could not pull them back safely to higher ground. And even when we did manage to get them well enough to leave the ICU for a regular hospital bed, they would be back a few days or weeks later, with either a recurrence of their problem or a new complication.

We colorfully referred to this as the “greenhouse effect,” since like plants in the winter, they cannot survive outside of the life-sustaining greenhouse of the ICU. They are often consigned for weeks or even months to a sort of medical purgatory, attached by tubes in their tracheas to ventilators, with catheters protruding from their necks, chests, abdomens or bladders. When awake they are in constant discomfort, chronically deprived of sleep, and stripped of any dignity, so we often sedate them to the point that they are no longer in communication with their environment.

“But even as I celebrated our successes … I also saw that there were patients whom we simply could not save.”

Even back then, I found myself asking what kind of living this was — perhaps to be tolerated if recovery is the reward — but for many patients, with their permanently failing organ systems, recovery was not to be. In that heady heyday of the newly hatched specialty of critical care — this was mid-1980s — no one much questioned whether we were doing the right thing when applying all of these life-sustaining interventions, but I could see clearly that this was not an end of life I would fancy for myself or for a loved one.

Thirty years later, things have evolved. Patients’ autonomy in making end-of-life choices is more deeply respected, patients and families are somewhat better informed, and palliative care has blossomed as an important specialty that can help optimize that quality of life near its end. Unfortunately profound disagreements sometimes still arise in cases where prognosis is poor, and this can be distressing and morally wrenching to all involved.

This seems like a good point to pause — I am interested in your thoughts and reflections on this topic.

Dear Nick:

I really cannot argue with any of the points you made in your note. As our ability to support life through improved monitoring, advanced pharmacology and more effective life-support techniques has grown, we increasingly run into the dilemma of whether we are prolonging life or prolonging death in the most critically ill patients.

Clinicians like me who admit patients to the ICU after a significant surgery often have a different perspective than the ICU doctors who rotate in and out of the service on a weekly basis. In most cases, as admitting surgeons, we have had the privilege of having in-depth preoperative conversations with the patient and family about their problems and proposed surgery. If we’ve done our job correctly, we have described accurately how arduous their journey may be as they strive for recovery. As you point out, it is also our charge to get to know the patient and their desires and fears about their management during a critical illness. Our personal contract with the patient is to give them the very best surgical and medical care we can to bring them through the surgery with the best possible outcome we can achieve. Thus, when patients become desperately ill, we face an internal conflict. On the one hand, we want to do our very best to pull the patient through to recovery. On the other hand, we have to be able to recognize when it’s time to “let go” and allow the patient to pass with comfort and dignity.

In many cases, our relationships with patients and their families and our understanding about them can temper this decision. Unfortunately, intensivists who rotate in and out of service on a weekly basis only to be confronted with a critically ill patient on a ventilator and other forms of life support do not have this understanding about the patient or the relationships that we have developed. Admittedly, we depend in part on ICU doctors to help us decide when a patient’s chance of recovery is too remote to carry on further.

As you point out, however, there are times when there is genuine disagreement over whether there is enough hope for recovery to carry on. When I have a patient in the ICU for a prolonged period, I follow the patient on a daily basis. I can see subtle changes day-to-day and week-to-week that other physicians might not appreciate. Time and again, I’ve watched a new attending come on service and calculate a critical illness score and determines that a patient has minimal chance of recovery. Of course, those population-based scores break down when applied to individuals. Renal failure in one patient is not the same as renal failure in another. Ventilator dependence for a treatable pneumonia differs from ventilator dependence due to profound weakness or chronic pulmonary disease.

It can be frustrating for an admitting surgeon when the ICU team gets out front with a series of conversations with the family suggesting the situation is hopeless when we believe that there is still reason to push on. Of course, all of this points out the need for frequent and open communication between the various clinicians participating in the care of these critically ill patients. The conversations we have are not only educational, but also critical in maintaining the focus that we are treating human beings and not illnesses.

Dear Mike:

We have much more in common than grounds for disagreement in our approaches, but let me address the points of contention you’ve raised.

First, let me make it clear that I recognize that the relationship between the surgeon and the patient is a unique one, and one that cannot be replicated by relationships with intensivists like me. The ICU practitioner by definition can’t duplicate the type of relationship that a patient may have with their surgeon — relationships that in some cases go back decades — and thus knows the patient less well as a person. In fact we often depend on the surgeon, as well as the patient’s family, to help us understand the patient’s goals and values, to understand what they were like before they became critically ill and unable to speak for themselves.

“When patients become desperately ill, we face an internal conflict. On the one hand, we want to do our very best to pull the patient through to recovery. On the other hand, we have to be able to recognize when it’s time to ‘let go’ and allow the patient to pass with comfort and dignity.”

That said, I would argue that this can be a double-edged sword — that the surgical covenant makes it difficult for the surgeon to have a broad, objective perspective on a patient’s condition. The intensivist lacks the bias that comes with the relationship and can make a more dispassionate assessment. Further, unlike surgeons, who may have many other patients and responsibilities, intensivists spend their time exclusively in the care of the ICU patients. We are intimately aware of the minute-to-minute status of the patient, and we are best positioned to assess their burden of discomfort and suffering. We interact throughout the day with family members and are constantly there to answer their questions and address their fears and concerns. It is a natural occurrence that questions will arise about their loved one’s prognosis, and we can answer with the objective information we possess. It seems to me that both perspectives are needed for the holistic care of a patient: the surgeon bringing the longitudinal perspective and the intensivist tempering it with the immediacy of the current challenges.

I agree with you that scoring systems to assess prognosis have little predictive value in the care of the individual patient, but they do help inform discussions about the magnitude of the physiologic derangements that patients face. Surgeons often use such statistics themselves when discussing the likelihood of death from a surgery, for example. Interestingly, a 2 percent likelihood of operative mortality is likely to be used as a powerful and reassuring statistic, but a 2 percent likelihood of survival in the ICU may be vehemently resisted as “not applying to this individual patient.”

In essence, I strongly agree that we can do the best for our critically ill patients and families by fostering a dialogue that respects both viewpoints, and that reaches a consensus that can be shared with family and loved ones.

Dear Nick:

I’m glad you brought up surgical scoring systems. I can say with certainty that I have never used one throughout my career. Although these scoring systems can identify risk factors which we must consider in prognostication, I have never believed that a multivariate equation derived from different patients in different hospitals with different surgeons can offer superior risk assessment than I and my consultants can while looking at a particular patient to evaluate their operative risk. How can these scoring systems account for the judgments I must make when I evaluate nuances about a particular patient and their anatomy and consider my own experience as I determine whether I can bring the patient safely through to recovery? As I stated earlier, I believe these scoring systems have similar limitations in the ICU.

As you know, the majority of families depend on us to educate them as they make difficult decisions about their loved ones in the ICU. They are often heavily swayed by what we tell them. Both of us know of patients who were desperately ill with grave prognoses who’ve gone on to live meaningful lives after they recovered. In these situations, had our advice to the families been only slightly more pessimistic, it is likely these patients would not be with us today. At the end of the day, I think we can agree that these complex decisions must be made by informed patients and families who are educated and supported by teams of experienced clinicians whose only goal is to do best by the patients and their family.


I certainly agree with you that we can put to rest the idea that scoring systems can readily be applied to individual patients in individual institutions, though I do worry that in our major teaching hospitals, we run the risk of a certain hubris to think of ourselves as essentially above everybody else’s statistics.

I wanted to touch on a couple of last points regarding taking care of profoundly critically ill patients in the ICU. The first is the question of quality of life. While the last thing I would ever want to do is judge what is an acceptable quality of life for a given individual, I think we need to attend to that concept carefully. If the best outcome we can expect for a patient is to be bedridden, institutionalized and dependent for all activities, we need to be certain that that is an outcome acceptable to that person. The important priority is the patient’s values and preferences; the simple objective of survival is not a sufficient goal. As you say, we both know patients who overcame formidable odds to go on to live productive lives.

The problem is that we also both know a much larger number of patients who do not emerge from critical illness, whom we treat aggressively with invasive, life-sustaining technologies, only for them to die after weeks of suffering. I remember one of my instructors during my critical care fellowship once saying, “You have to work really hard on many to save a few.” While this clearly is true, it is not clear that it justifies the suffering and indignity borne by the many who will not be saved. At a minimum, we should be calibrating our treatments according to the patient’s previously articulated preferences, withholding aggressive treatments when they are not aligned with the patient’s goals.

“The ICU practitioner by definition can’t duplicate the type of relationship that a patient may have with their surgeon … That said, I would argue that this can be a double-edged sword — that the surgical covenant makes it difficult for the surgeon to have a broad, objective perspective on a patient’s condition.”

My practice is to start family meetings by first finding out what the family’s understanding is of how sick their loved one is, then to provide a clear and honest assessment of the prognosis, while clearing up any misconceptions. The next step is to have the family tell me as much as possible about the patient as a person. What was she like? What did she like to do? What was important to her? Had she ever talked about what she wanted done if she became this ill? This allows the discussion of the possibility of death to be approached in a compassionate and individualized manner and facilitates making a plan that is respectful of the patient’s preferences.

Finally, I would be remiss not to bring up the topic of resource allocation. While this should never be included in decisions about individual patients at the bedside, ICU care is expensive, and we are facing an ever-growing burden of cost of medical care in our country. While this subject is fraught and complicated, it seems clear that the “blank check” approach to medical expenditures cannot continue. We will need as a society to place limitations on expensive care, and that care that benefits a few at the expense of many will be in the crosshairs for reconsideration. As an important initial step, I believe we will need to do a better job of weaving palliative care strategies into our treatment of patients who are near the end of life, with the goal of optimizing the quality of the time remaining to them.

I want to thank you for your willingness to engage in this discussion and assure you that I have only the highest respect for your skills and integrity as a surgeon.

Best wishes,



You are quite right in that we must consider the patient’s quality of life when making decisions about ongoing care in desperately ill patients. This highlights the need to understand the patient’s as well as family’s wishes in these difficult circumstances.

You are also spot on when you say that we pushed too far on many more patients than the comparatively few who benefit from unexpected dramatic recoveries. This of course, is why we are having this very discussion. While I suspect that there will always be a larger number of patients in which we push too hard, hopefully this discrepancy will diminish as we collectively learn and embrace that sometimes stopping care is the best care.

I agree with you that the topic of resource allocation must be considered. ICU care is the most expensive care we can offer and we can clearly offer more care than our society can pay for. Hopefully, these very discussions will lead to improved use of resources in the ICU. I truly hope, however, that these conversations remain at the 60,000-foot level and thus impact care globally but that they do not come into consideration on individual patients. As we’ve discussed, we must consider the many complex factors as we determine whether to carry on with care in a particular desperately ill patient. I hope that expense of care is not a priority in these individual cases.

Thanks for the dialogue,


Jason M. Breslow

Jason M. Breslow, Digital Editor



In order to foster a civil and literate discussion that respects all participants, FRONTLINE has the following guidelines for commentary. By submitting comments here, you are consenting to these rules:

Readers' comments that include profanity, obscenity, personal attacks, harassment, or are defamatory, sexist, racist, violate a third party's right to privacy, or are otherwise inappropriate, will be removed. Entries that are unsigned or are "signed" by someone other than the actual author will be removed. We reserve the right to not post comments that are more than 400 words. We will take steps to block users who repeatedly violate our commenting rules, terms of use, or privacy policies. You are fully responsible for your comments.

blog comments powered by Disqus

More Stories

Wisconsin Recount Targets 2 Democratic Strongholds, Home to Black and Hispanic Voters
Unlike other places, where the Trump campaign has pursued statewide recounts, the Wisconsin request was limited to Dane and Milwaukee Counties, where Biden held solid leads.
November 25, 2020
7 New Documentaries to Watch over Thanksgiving Break in 2020
Plus, one new podcast to listen to.
November 25, 2020
‘I'm Not A Monster’: Who's Who from the Podcast
Meet the key players in 'I Am Not A Monster,' a special podcast series on how an American mother ended up in the heart of ISIS' self-declared caliphate.
November 24, 2020
On Night of Ginsburg’s Death, McConnell Pushed Trump to Nominate Amy Coney Barrett
An excerpt from FRONTLINE’s ‘Supreme Revenge: Battle for the Court’ goes inside Sen. McConnell’s swift maneuvering to achieve a 6-3 conservative majority.
November 24, 2020