Discussing Prognosis

[Neuronix]

Recently I saw a limited stage SCLC patient treated with chemoRT on OTV for a 9x6cm lung primary. His med onc told him that he had an "excellent chance of cure", and I told him that might be a bit optimistic. The patient seemed shocked by the news, but asked a lot of hard questions. I told the patient all the details as far as the hard numbers about SCLC, the course of the disease, treatment options if there is a failure, etc... I emphasized the benefits of treatment: that chemoRT should substantially prolong his life and has a small chance of cure. At the end of the discussion he seemed grateful for the real information, and I told him not to give up hope and enjoy life. He wanted to know how he could prepare for the future, and I told him he should have a will and a power of attorney for healthcare. I was actually shocked this hadn't been previously discussed.

After the history, physical, and data review, I see the curative consult as a discussion of four things primarily: benefits of radiation, risks of radiation, alternatives to treatment, and logistics of treatment. I always offer and typically provide as part of benefits and risks the percentage of cure we would expect to achieve without radiation and with radiation. For the very educated or motivated patient, I will even bring up the articles on the screen that support what I'm telling them and give them references or copies. I feel it is their right as a patient to know the facts so that they can make proper choices for their own lives and care. For example, if I'm consulted on a stage III breast patient, I often tell the patient that they have a 50% chance of cure with radiation (a gross simplification I know, but not too far off...). I always tell the patient we are going to fight aggressively for that cure.

Today I relayed this viewpoint to my co-residents, and they seemed mortified by my level of candor. I've seen and heard a lot of perspectives on these discussions. More often than not, I see providers who dance around any discussion of hard numbers unless they are directly asked for percentages. Even when directly asked, I have seen providers not provide that information for a number of reasons that I disagree with.

So I wanted to ask the community. In general, how do you handle the discussion of prognosis when you are consulting with the potentially curable patient? In my opinion, for most patients, they use the internet and pull up all the horror stories and misinformation you can get with a Google search. I have found that by the time the patient comes to me, they have often asked their med onc and surgeon for prognosis information without satisfactory answers. There are rare patients who don't want to know their prognosis for various reasons, and certainly that must be respected. Additionally, the data must be conveyed in a way that is understandable and compassionate. But that said, I go under the assumption that this is information that patients want to know, but don't know how to ask for, and is an essential part of the consult. Does anyone agree or disagree?

---

Members don't see this ad.

 

[Gfunk6]

Good question. Generally, I ask my patients point-blank, "do you want to discuss prognosis/survival?" They can accept, decline or identify family members who need to be informed. If you are not the first person to see the patient after a cancer diagnosis (a very common scenario in our field), you may have to contend with a "rosy" picture painted by the Med Onc or Surgeon which you may be loathe to directly contradict if you want to see patient referrals from them again.

It depends a bit on your relationship with the referring attending. I have a great relationship with my Med Oncs in general so I generally can say whatever I want. There are "high ranges" and "low ranges" for survival so you can't necessarily fault the referrer in all cases. However, when the information provided is clearly wrong, I feel ethically compelled to give my two cents.

Wording is key.

WRONG: "Dr. Z is wrong. The median survival of GBM for patients in your prognostic class is 14.2 months."

BETTER: "Well, Dr. Z's estimate is a bit on the optimistic side. In the literature, the median survival for GBM is 14.2 months though it is certainly possible you could do better or worse."

It's all about the art of medicine.

Why do people put nails in coffins?
So the medical oncologist can't give more chemo.

 

[rymd]

Agree with neuronix- conversations about prognosis, along with frank discussion about goals and purpose of the treatment being offered has only lead to excellent interactions, and appreciative patients.
Though as Gfunk points out- it is all in how it is said. If you are going to discuss numbers- let them know that you are describing a median- then explain to them what it means (sorry for the pun)

 

[deleted4401]

Very difficult, indeed. I've only done it for 2 years, and still honing ...

I think statistics get dangerous, especially since even many physicians don't understand means/medians, etc. I also think that when patient's ask for prognosis/survival, although sometimes they want hard numbers, many times they are digging for something specific. I.e. - when you ask what are they trying to find out, and they may say, "Oh, well my daughter's wedding is coming up in 3 months" and you can confidently say "Well, I think you'll most certainly be able to make that" or "Well, that might be tough and you may be sick during that time" or "Unfortunately, I'm not sure that you will, I recommend you spend as much time as possible with your family". A lot of times, that will end it - you've solved what's bothering them. Some will ask more or actually want to hear the numbers, but the majority have some sort of event or checkpoint that they are asking about. When it comes to medians/means, I usually tell them a story about a town with 2 inhabitants, one making $0 a year, and one making $1,000,000 a year and that it's very misleading to say that average income is $500,000. If they seem to understand that and continue to want information, then I give them the numbers. Almost never gets to that point.

I had this whole module during internship from one of the leading experts in palliative care. His name is Bob Arnold, and there is a bunch of stuff online about all of this stuff - breaking bad news, discussing prognosis, talking about futility of further care. This is just one example, but there is a ton out there: http://onlinelibrary.wiley.com/doi/10.3322/canjclin.55.3.164/full

As you fine tune your clinical training (and by the time you finish studying for your oral boards, you'll be probably be the smartest you'll ever be), it might be a good idea to start doing some reading on topics like this. It seems real soft, but I use these techniques daily and it makes those uncomfortable conversations somewhat better.

 

[Mandelin Rain]

"You will either live shorter than, longer than, or exactly 14.2 months."

Population statistics are better left for the boards, than discussions with patients.

 

[Gfunk6]

Population statistics are better left for the boards, than discussions with patients.

Very true.

Folks, listen to the Mandelin Rain.

 

[RadRadRad]

anyone see the first abstract in the big IJROBP astro abstracts for this year. Its regarding patient perceptions of benefit of RT in setting of incurable NSCLC. Looks like many patients are confused about basics such as curative vs palliative intent.

 

[ShirleyT]

I only talk about prognosis if a patient specifically asks, and even then I keep it general b/c statistics are statistics and who knows about an individual patient?

one of my attendings in residency has a med onc friend who was actually sued b/c they told a patient with advanced lung cancer they had a few months to live. Well the patient lived for three years and sued the doc b/c they spent all of their money in the few months he thought he had left to live!!

 

[IndyXRT]

I tailor my discussion by patient. I try to always make sure that the intent of treatment (curative vs. palliative) is clear. If they ask for specific numbers, I give them specific numbers. However, I preface the numbers with a few words about statistics and the fact that they do not accurately predict what is going to happen to any one patient.

Gfunk:
I've heard the line about nails in coffins, but why do you think they bury the coffins in cement lined vaults. . .

 

[Cancerdancer]

I try to discuss prognosis often, but start general and get more specific based on patient sophistication, comfort, and desire. I think a patient with low-risk prostate cancer needs to understand that their 'cancer' is not a death sentence, and the patient with stage IIIb NSCLC deserves to know that, while still 'curable' they face an uphill battle.
I typically ask them something general like 'what do you know about your diagnosis and the aggressiveness of your cancer?' Some answer with a surprising level of sophistication, some are caught totally off guard and clueless. You can help these people, and should, in my opinion.
I then will ask what they want to know. 'Do you want to talk about specific likelihoods of cure or outcomes, or just talk about treatment recommendations?' Most ask for some info...
I'll then say "Well, in general, there are some cancers that are rather low-risk, and we often cure. Some others are more 50/50. And others are an uphill battle. Yours (un)fortunately fits into the XX category. I'm happy to give you more specific numbers, but often people feel knowing that in general is specific enough. How do you feel about that?"
Most stop there. It's rare that they want/need to know about a 95% vs 85% 5 yr OS, or 30% vs 10% 5 yr OS. But I think too often we assume patients know that prostate CA pts generally do well, or that pancreatic/GBM pts do poorly.

 

[seper]

Very dependent on pt's educational level. Where I'm practicing, majority of elderly never went to high school. So I just tell them "you've got a chance of beating this thing if you come for your daily treatments".

 

[Carotenoid]

I am only a few months into my TY, so I am much less experienced than most of you - and I am myself curious about how my views will change with time. For now, however, I of Neuronix' mindset.
While I do agree that some patients/families will not HEAR the diagnosis (no matter how clearly one spells it out - when I was on heme-onc inpatient rotation, we had a gentleman with SCLC that failed both 1st- AND 2nd-line therapy, with expanding brain mets and a maxed-out radiation dose - whose family strongly believed that "he will walk out of here" if only they "stay strong and think positive"). Nonetheless, this stubbornness does not excuse us from informing the patients who want to be informed.

In our ICU we now have a lady with a widely metastatic cervical cancer. Her med onc (whom I have yet to meet because she is abroad on a 3-month teaching stint) had reportedly told the patient and family that "she would only tolerate chemo if she felt better". The "if" part reflected a 1 in a million chance - at the time the conversation took place about a month ago, the patient's performance status was between 3 and 4 according to notes. By the time she was admitted as an inpatient, she was a 4. I tried to convince her to sign a DNR/DNI, but her husband was still holding on the hope that "when Dr. XYZ is back, she is going to give PQR chemo"... and so next time I met the lady and the husband (and the daughters) was on my ICU rotation, where she remains on a ventilator - and the family is now facing the decision of unplugging her.
Could it be that the husband and the patient heard what they wanted to hear, and that in fact the med onc had painted a realistic picture for them? Sure. Nonetheless this tragic case to me personally serves as a perfect illustration of WHY we need to try as hard as we can to convey to the patients their prognosis. Hey, if we are wrong, and they go on to live forever, it is fine. I'd much rather be wrong then not help my patients and families adequately prepare for death/disability.

 

[hot sauce]

I am only a few months into my TY, so I am much less experienced than most of you - and I am myself curious about how my views will change with time. For now, however, I of Neuronix' mindset.
While I do agree that some patients/families will not HEAR the diagnosis (no matter how clearly one spells it out . . . .

In our ICU we now have a lady with a widely metastatic cervical cancer. Her med onc (whom I have yet to meet because she is abroad on a 3-month teaching stint) had reportedly told the patient and family that "she would only tolerate chemo if she felt better". The "if" part reflected a 1 in a million chance - at the time the conversation took place about a month ago, the patient's performance status was between 3 and 4 according to notes. By the time she was admitted as an inpatient, she was a 4. I tried to convince her to sign a DNR/DNI, but her husband was still holding on the hope that "when Dr. XYZ is back, she is going to give PQR chemo"... and so next time I met the lady and the husband (and the daughters) was on my ICU rotation, where she remains on a ventilator - and the family is now facing the decision of unplugging her.
Could it be that the husband and the patient heard what they wanted to hear, and that in fact the med onc had painted a realistic picture for them? Sure. Nonetheless this tragic case to me personally serves as a perfect illustration of WHY we need to try as hard as we can to convey to the patients their prognosis. Hey, if we are wrong, and they go on to live forever, it is fine. I'd much rather be wrong then not help my patients and families adequately prepare for death/disability.

Always be careful judging another doctor based on what a patient says the doctor said. As you said people often hear what they want to hear. Dr. X may have said unfortunately I don't think you will recover from this downturn and more chemo will only make things worse at this point. The family or patient may have then pressed and gone down the what if path and Dr. X may have conceded if things turned around that it would be reasonable to discuss chemo at that point. It is difficult to tell someone that there is no way that you would ever offer any more treatment for their disease.

I also shy away from number unless asked (at least when they are bad). As mentioned by several posters median survival numbers are dangerous. I like Simul's analogy of average income. How many times have we heard the doctor only gave me X months to live but here I am years later. You say a median survival and people often imagine a very sharp peak. When we are attempting definitive treatment, I try to talk more about the numbers in the tail, if they want numbers at all, or at least point out that a small non-zero number of people live much longer than the median. People know that something like a 20% 5 year survival isn't good (just a random number) but it does leave hope without being unrealistic (assuming your number is reasonably accurate).

In cases of a poor prognosis, I try to focus on the positive but patients do need to understand when they have a serious disease so they are not shocked when the most likely outcome occurs. But I do have to be careful as I sometimes find myself heading down the path of saying things that are accurate but too open to overly optimistic understanding.

 

[Carotenoid]

good discussion. Great to know what ppl think on this matter.

 

[Mandelin Rain]

If you ever find yourself saying something like...

"Your 5-year overall survival is X%."

or

"Your median progression free suvival is Y months."

...you're doing it wrong. For so many reasons.

 

[ToasterThief]

I think one of my most uncomfortable moments of med school was going in to a room with a patient with a new diagnosis of GBM with a neurosurg attending. When his confused and worried wife asked the attending what his prognosis was like he said something along the lines of "We're making advances all the time. Whatever I tell you now might be completely wrong in a year or two!"

 

[Neuronix]

When his confused and worried wife asked the attending what his prognosis was like he said something along the lines of "We're making advances all the time. Whatever I tell you now might be completely wrong in a year or two!"

While I agree with Mandelin Rain about not using confusing and marginally relevant statistics in the prognosis discussion (which nobody in this thread has advocated for), I think not answering questions at all in the above quoted way is just as wrong. The prognosis for GBM is not changing quickly. This is a perfect example of even when asked directly, and I would consider that wife to be asking directly, physicians often do not answer the question being asked.

If you need a patient to ask you "What is the 5 year overall survival", it's never going to happen. If they're asking for prognosis, they're asking for something equivalent to "How much time does he have" and "what is the chance of cure". I think it's our obligation to address those concerns. I'm not saying you need to confuse the patient with medical jargon or give them a death sentence. But I think we do need to give them a realistic picture at their level of understanding if they have asked for it.

I have seen avoidance of discussion of prognosis, even when directly asked, much more frequently than I have seen physicians botch the discussion about prognosis. Maybe it's just my own personal experience. If anything, the socially awkward attendings don't want to have these discussions because the discussions are difficult to have. it's far easier emotionally and less potentially offensive if you be vague, avoid upsetting anyone, and leave the discussion for someone else or some later date that never comes.

Of course, the reality is that the first thing the confused and worried wife is going to do is go home, Google "glioblastome multa formi", and read horror stories and a lot of misinformation that probably doesn't apply to them. I think it's better that I leave myself open for an honest discussion and remain open to any questions they might have now or in the future. Of course it's easier to just blow off their questions, and they probably won't ask you again because you are so unhelpful. But, are you really doing your job if that's the case?

Nonetheless this tragic case to me personally serves as a perfect illustration of WHY we need to try as hard as we can to convey to the patients their prognosis.

Your story reminds me of the Atul Gawande article "Letting Go":

http://www.newyorker.com/reporting/2010/08/02/100802fa_fact_gawande

 

[Raygun77]

I think one of my most uncomfortable moments of med school was going in to a room with a patient with a new diagnosis of GBM with a neurosurg attending. When his confused and worried wife asked the attending what his prognosis was like he said something along the lines of "We're making advances all the time. Whatever I tell you now might be completely wrong in a year or two!"

I think it's fair enough though to emphasise, for those stats- savvy patients who actually want numbers- that any 5 year survival data is by definition at least 5 years old...and thus may not reflect practice or outcomes from today onwards (as hopefully your centre incorporates clinical trials etc that do prolong survival)

Keeping on the theme of GBM for example, this updated data by the UCLA neuro-onc team demonstrates that through involvement of clinical trials, close follow up and otherwise good patient care survivals can be dramatically better than 'gold standard' published trials from 2005.

I also think it's fair enough to emphasise the role of 'clinical trial involvement' in improved patient outcomes, since it's a fact that those patients on clinical trials are followed more closely for recurrence, etc, and at the very least are getting gold standard treatment.

But the key is really to ask before you tell imo. Many times i've heard or seen "look doc, I know enough to take the disease seriously but I really don't want to know more", which again is fair enough.

 

[seper]

It takes years for most physicians-in-training to truly understand the meaning of probability of survival, estimated by Kaplan-Meier method.

How much time do you have for each consult?

 

[TarHeelRadOnc]

I think I am in the minority on this, but I think that our patients deserve a detailed discussion of prognosis. This is especially true when the likelihood of toxicity is high and likelihood of cure is relatively low. Agree with previous post that the most important node in the algorithm is treatment intent: definitive vs palliative; and every patient we treat should have at least a cursory understanding of this. As mentioned previously, the prognosis discussion is most challenging in our patients with poor prognosis whom are treated "definitively" (GBM, stage III NSCLC, LS-SCLC, pancreas, etc). For these patients, I use an approach similar to GFunk. I ask if they want to discuss prognosis, and find that roughly 80% do. Agree that mean, median, and Kaplan-Meir are not for consultations. That said, I do use an estimate of long-term survival/cure (ex. 20-25% for stage III NSCLC). Provided the patient has reasonable performance status, I tell them that we give everyone the "benefit of the doubt" and we treat them as if they will be in the 20-25% unless their "body (toxicity requiring treatment cessation) or tumor (on-treatment progression) tell us otherwise." I find this method useful as a way to have a serious conversation about prognosis, while at the same time allowing room for optimism and hope. We have to be careful not to suffocate the latter in our patients.

 

[Palex80]

I think one of the main issues we have to deal with, is how to motivate patients to undergo treatment despite a grim prognosis.

Talking with GBM patients is not such a big issue in my opinion. They have little choices left and we know, that they will benefit a lot from radiation therapy in terms of overall survival and symptom control. Treatment takes quite some time, but it's not really that troublesome for most of them.

The one group of patients, I find hard to talk with, are head & neck patients. These are patients, we wreak havoc upon with our treatment, both in terms of acute and long term toxicity (yes, even with IMRT!) and we often cannot cure. Explaining a cT4 cN2c hypopharynx, that he has to undergo through hell for 7 weeks to have a 20% (?) chance of permanent cure (with questionable quality of life), is tough.

 

[medgator]

I think one of the main issues we have to deal with, is how to motivate patients to undergo treatment despite a grim prognosis.

Talking with GBM patients is not such a big issue in my opinion. They have little choices left and we know, that they will benefit a lot from radiation therapy in terms of overall survival and symptom control. Treatment takes quite some time, but it's not really that troublesome for most of them.

The one group of patients, I find hard to talk with, are head & neck patients. These are patients, we wreak havoc upon with our treatment, both in terms of acute and long term toxicity (yes, even with IMRT!) and we often cannot cure. Explaining a cT4 cN2c hypopharynx, that he has to undergo through hell for 7 weeks to have a 20% (?) chance of permanent cure (with questionable quality of life), is tough.

Not all head and neck patients end up in that category. Probably not even the majority. The discussion is more optimistic for oropharyngeal patients, especially the HPV positive ones.

 

[deleted4401]

I agree about the locally advanced head and neck's like Palex is talking about. The patient's with "T5" disease and 7cm neck nodes - it's aggressive palliation at best, futile toxic hell on earth at worst. The 5 year survival for all comers with larynx cancer is about 50-60%. But the T4s and ones with big neck nodes ... good luck. It's a tough treatment, and we do treat them "curatively" but it's mayhem compared to 60 Gy to the thorax for a stage III lung.

Prognosis is really tough to talk about. It sounds like everyone has a different style. Just have to find what works for you, be honest and clear with the patient without seeming hopeless. Or, say nothing, and say that it is a conversation they should have with their med-onc

-S

 

[medgator]

Or, say nothing, and say that it is a conversation they should have with their med-onc

-S

When it comes to H&N, most of the med oncs I've worked with seem to defer to us. They'll ask me whether they should use platinum or erbitux

 

[Neuronix]

Just wanted to post a recent NEJM article on this topic and commentary in the NY Times.

http://www.nejm.org/doi/pdf/10.1056/NEJMoa1204410

http://well.blogs.nytimes.com/2012/11/19/when-treating-cancer-is-not-an-option/


Edit: Also note this article recently published in the red journal, http://www.redjournal.org/article/S0360-3016(12)00954-6/fulltext "Patient Beliefs About Palliative Radiation Therapy (RT) in Incurable Lung Cancer"

 

[seper]

Very good article and NYT post. For most of our palliative patients, I'd prefer MedOnc's be frank about prognosis.