Keywords: Shared Decision Making, PSA, Prostate Cancer

How to Cite:  Rosenberg, Matt T. “The Role Of Shared Decision Making in Prostate Cancer” November 10, 2017. Accessed Jan 2018.


Dr. Matt T. Rosenberg discusses the logic behind Shared Decision Making and argues against the use of it when treating patients with Prostate Cancer

The Role Of Shared Decision Making in Prostate Cancer


MATT. T. ROSENBERG:  Shared care,  this is something I have a tremendous passion about.  You’ll notice the subtitle, An Efficient Use of Time or Another Exercise in Futility, all right?  Let’s look at this.  You’ll notice my title changed to a concerned doctor.

Dr. Cookson very appropriately mentioned the screening guidelines, all right?  This is what we’re dealing with now.  In 2012, the United States Service Preventive Task Force came out with the D recommendation and through the world into a tizzy. Remember, it wasn’t just in 2012 they did it.  They did it prior to that.  They were telling us that there were some issues going on, and everyone started to get nervous.  They said we shouldn’t be doing it.  

Then in 2017, everybody got excited because they said we’re going to change the recommendation.  We’re going to make it a C.  I remember I called Dr. Crawford that day, and I said, did you see it?  He goes, yeah, it’s a C.  I’m like, they screwed this up even worse, all right?  A C doesn’t help me, all right?  It makes them feel better because the answer to all of these—they’re still putting the onus on me.  

They’re saying shared decision making, shared decision making.  You got to do shared decision making, all right?  That’s great.  It’s great if you’re sitting behind the ivory towers pushing a pencil and not seeing a patient.  It absolutely bites if you’re trying to do the right thing for the patients.  Everyone jumped in on that—the AUA, ASCO, ACS.  Everybody jumped in to shared decision making.  Let’s look at that.  

There are three tenets of shared decision making.  It’s a provision of information.  It’s an elicitation of patient perspective which is talking about your prior experiences, understanding and discussing concerns, delineating preferences, and guiding final decision making, okay?

I’m going to ask my primary colleague in the audience, how do we do this?  Patients come to us in a primary care setting because they want us to help them.  What would you do, Matt, for your family?  What would you do for your wife, your father, your uncle, your brother, all right?  Imagine the shared decision making—imagine this.  I was talking to a student of mine recently.  

This is like me going to the auto shop and saying I need a new muffler.  He goes, well, Dr. Rosenberg, let me explain to you what a muffler does.  I’m like, Jesus.  All right, let me talk to you about what happens if we change the muffler, okay?  Dr. Rosenberg, have you ever changed your muffler before?  All right, and how do you feel about the muffler?  Are you going to miss this?  Are you going to have some—and then I’m not going to actually tell you what to do with your muffler, Dr. Rosenberg.  I’m going to ask you to make that decision, all right?  How obscene does that sound?  

That’s what’s happening with shared decision making.  For the people who’ve recommended it, they have not done primary care.  They don’t have any idea.  David Crawford said in his talk that it’s a 15-minute conversation.  When you and I were doing that video last year with Dan Petroy, a friend of ours who’s a radiation oncologist at Yale, he said, Matt—medical oncologist, excuse me.  He said 22 minutes to do shared decision making.  

I’m looking at that saying awesome.  That’s awesome because I’m doing more of these in primary care than anybody, right?  

You obviously have an extra 22 minutes per patient per day to talk about prostate cancer, all right?  Forget about the fact that you’re a poorly controlled diabetic with an A1c of 9, that you haven’t gotten your eyes checked so quality measures are out the window.  Forget the fact that you can’t afford the medications that I’ve just given you.  

The reality of it is I need 22 minutes to do this in a disease state that I’m not a specialist in, all right?  I am a specialist.  I am the master of everything, right, or the knower of everything—maybe not the master of everything, but I have to know everything, all right?  This is not what I do.  

We talked about it earlier in primary care and internal medicine, we might’ve had a day or two of urology.  In internal medicine, it’s not required.  Are you family medicine or primary care—or is it internal family?  Do you have any training in urology?  Your husband, there you go.   Osmosis, so his—there you go.  You get it through osmosis, but the reality of it is that we don’t have that kind of training.  We’re not the specialists to be able to do this.  It doesn’t make any sense to do this.  

The reality of it is when they looked at shared decision making, it did not work.  We’ve tried this before.  We’ve tried this is colorectal cancer.  We’ve tried this in breast cancer.  We’ve tried it now in prostate cancer.  It fails miserably.  It’s no surprise.  I could’ve saved you a million dollars for a study to tell you that, all right?  

This is a study from 2015 that looked at patients that were eligible for this.  Look at the data.  It’s pretty obvious.  Only 22 % had told us that some doctors recommended it.  Only 14 % had any idea if its saves lives or not.  Ten percents received all the tenets of shared decision making; that’s probably 9 % more than I would’ve expected.  Fifty-five percent are getting the PSA, and that bothers me because really everybody should be getting that.  Shared decision making doesn’t work.  

The reality of it—and I think Dr. Cookson mentioned this—is that we see decreased screening because we’re scared of it.  We’re scared of the PSA.  We’re scared of the rectal exam.  We have decreased screening.  When we have decreased screening, what are we worried about—metastatic disease, all right?  What have we now seen in the literature—more metastatic disease.  What are we going to do?  It’s obvious that we’re doing the wrong thing here.  

The diagnosis of prostate cancer—this is some older studies—it shows that it’s gone down since they’ve adjusted these recommendations.  It’s gone down.  The proponents of no screening are cheering, yay.  Less people are getting screened, and you guys are looking at this saying, crap, more people are getting metastatic cancer because we know people have that, right?  What if we’re wrong?  What if we’re wrong on this stuff?  

I think part of the problem is that we—Dr. Cookson had mentioned before that other disease states would give their right arm for a screener, all right?  In prostate cancer, we’ve taken a different effect.  Think about breast cancer.  We don’t go through shared decision making for breast cancer; you get a mammogram.  You do a breast exam monthly.  

If people didn’t die of breast cancer as much as they did or because people die of breast cancer as much as they did, we screen because it’s the only thing we can do, all right, because they die.  With prostate cancer, you guys in the audience, you do a really good job keeping people alive.  Because people with prostate cancer don’t die as readily as people with breast cancer, we have this question about screening.  

The question really we should ask is, do you want to be a 70-year-old with metastatic prostate cancer getting all the therapies that we’re going to talk about at this meeting or do you want to be a 70-year-old with no cancer at all, all right?  The screening debate is only because we do such a good job with it, and it’s the wrong thing to do.  

The question that Dave and I have worked on for a long time is this simplification of this process.  How do we simplify this?  I was asked this a couple of years ago.  Dave asked me to present this concept.  

The idea is how do we simplify this.  Well, we don’t want needless evaluations.  We don’t want to make unnecessary patients.  Dr. Cookson had mentioned that people get worried because we do too many radical prostatectomies, all right?  Well, actually, I get more worried because we do too many prostate biopsies.  Nobody likes a prostate biopsy, all right?  If they get one, they ain’t getting two, all right?  Patients will come to my office—I’m sure you see that.  They’re like, you know what, don’t even check because I’m not doing this again, all right?  That’s a needless—people get sick from that.  What is the cost?  

Here’s a good one.  What is the cost of urosepsis?  What is the cost of infection after a biopsy to put the patient in the hospital?  Those IV antibiotics are more expensive than my yearly income.  

We don’t want to ask the primary care to counsel outside of their comfort zones.  This whole idea of shared decision making is odd.  It doesn’t make sense.  Why are you asking me to talk to you about something that I’m not a specialist in?  Can we simplify the PSA value?

If we’re going to make primary care happy, we have to have evidence.  We have to have information that is based on evidence and beyond dispute, that the patient has to be presented with a clear framework for decision, and it has to be appropriate for primary care.  Whoever wrote these guidelines, whoever made up this whole idea of shared decision making and screening wasn’t thinking about us, wasn’t thinking about what is the best way for us to treat the masses that don’t get treated, all right?  

It’s a great time to be a primary care provider right now because people can’t get them.  We haven’t advertised for new patients in 20 years.  I saw six new patients yesterday.  I’d like to say they come to me because I’m good.  They come to me because no one else will take them.  No one else is opening the doors.  No one else will see them.  We just don’t have that policy.  We need the stuff that assumes that we can do this in a good period of time.  

We don’t have these two.  The question is, can we simplify screening?  This is just a schematic I want to present because this is how I look at it.   When I talk in lay public lectures, I talk to primary care.  Think of it this way.  This many folks get diagnosed with prostate cancer a year.  This many men die of it, all right?  This means that all of these people get this unnecessary treatment.  That’s the donut.  

This is my target.  This is the sweet spot.  How do I find that sweet spot?  How can I make this process work?  David approached that before with using different things that we have available—biomarkers and MRIs, things like that.  We want to find that patient who is at risk for the aggressive, life-threatening cancer.  

This is the question they should be asking instead of shared decision making.  Can we use what we have right now in 2017 to find a sweet spot?  This is what we know.  Dave already talked about this—1.5.  We have now published two papers on this.  We’ve presented this in a lot of different places.  The nice thing is it’s gaining momentum because it makes sense, all right?  Dave presented a distribution.  Let me show you something else here.  

This is a paper from Aus that I combined with a paper from Vickers which shows very clearly that the risk of prostate cancer starts to go up at 1.5.  Years ago, I remember reading a paper in the White Journal that said that your risk of prostate cancer if your PSA is low—and they picked one—your PSA is less than one, your chance of cancer is essentially zero.  I remember quoting that for years, all right?  It was low.

Patients come in and they’re concerned.  Oh, my god, my PSA is 0.7.  I’m going to die.  No, it’s actually pretty good.  You’re fine, all right?

This study actually did it a little more specifically and said, you know, below 1.5, your risk is very low. If it’s above 1.5, it starts to go up.  It’s still low, but it starts to go up a little bit.  

This is an older paper.  This is a paper from Sweden.  If you look at blue, blue is no cancer.  Red is cancer; that’s your cases.  Look at that.  Look at that.  At 1 or 1.5, you’re there.  What are your chances of cancer?  They’re very, very low.  It starts to go up here.  

Remember that PSA is prostate-specific antigen.  It’s not prostate cancer-specific antigen.  That’s what we have to teach our colleagues here.  A low PSA actually does tell us a lot.  A higher PSA tells maybe we need to think about it.

This is a final graph from that prior study which shows here that when your PSA starts to go up, that’s when we start to see cancer—above that 1 level, above the 1.5.  We have a very good argument we’re making.  

Twenty-seven percent of our patients have a PSA of 1.5 or more which means 77 % have it lower—73 %.  Sorry, I can’t add.  Now we think about shared decision making.  It takes a whole different view.  Wait a minute.  If we use this like a vital sign, maybe I only have shared decision making when I’m concerned, when it’s a problem.  

26 Here’s an analogy in primary care.  Who’s been to their family doctor or their internist any time in the last whatever period of time, a year or so?  Did they check a blood sugar?  Of course, they did.  Did you question them checking their blood sugar?  Of course, you didn’t.  A blood sugar, for the most part—how specific is a blood sugar?  It’s not but it’s screening.  We talked about it with OAB a minute ago, right?  If it’s high, what do I check?  I check an A1c.  I don’t rush you to the endocrinologist for insulin.  I check an A1c, right?  

Guess what?  Now I have my PSAs of risk right here, right—PSAs of risk.  Now I look at it so aha, we have the algorithm.  We have the algorithm that Dave and I published with many, many others.  

Actually, I have funny story.  Do you know why we have 45 on this?  You saw his prior algorithm, he had 50.  I wanted 40 and he wanted 50.  We cut the baby, all right?  After arguing on the phone—seriously, that’s as scientific as we got.  

Why we did this, this is a start.  This isn’t the end product.  This isn’t the guidelines as written by the AUA or ASCO or the United States Preventative Service Task Force.  This is a logical way of thinking about this. What happens is we get a guy, and we screen them.  If your risk, if it’s less than 1.5, your risk of cancer in your life is pretty low.  Great, that’s awesome.  

If it’s 1.5 or above, now, I’m going to look.  Now I’m going to get the A1c—the prostate equivalent of the A1c, which is a biomarker, all right?  Further investigation by a PCP or urologist—I’m going to get a 4K, a phi, a PCA3, a SelectMDx, whatever you’re going to get your hands on.  You’re going to assess risk.  Then if the risk is high, you’re going to evaluate.  This is where shared decision making comes in—right here.  

That’s the most sense.  I only have to have it with the appropriate patients, just like blood sugar.  I’m not going to talk to you about diabetes without checking your sugar.  That’s a total waste of my time.  If your sugar is high in your screening labs, I’ll call you in or you’ll have been there for the physical and say your sugar was 125.  It was fasting, and I’m going to check this.  There’s shared decision making.  Then I’m talking about the risk.

There is a better way to screen for prostate cancer.  We use the PSA as a first-line test.  You get suspicious PSAs which then start with that shared decision making.

What have I learned over time?  What I’ve learned over time is throwing out the PSAs is really throwing out the baby with the bath water.  First of all, I don’t want to get a PSA in somebody who’s not healthy.  If you’re a 50-year-old male coming into my office with a status of post-bilateral BKA breathing through your trach tube but still smoking through it when you remove the oxygen, right, I’m not checking your PSA.  That’s a bad idea.  On the other hand, if you’re an 80-year-old male on your way to go play tennis with your dad, I’m probably going to check your PSA.  I actually have decided I’m insulted by the guidelines when they put age parameters, all right?  

You can’t tell me, as a urologist—sorry, guys, I don’t mean to be disrespectful—you can’t tell me a patient’s health.  I’ll tell you a patient’s health.  That’s what I do.  You tell me prostate health.  You tell me general urial—I’m going to do the overall patient thing.  That’s what I do.  Do I put age on it?  

No, I just went to a birthday party for a patient of mine who just turned 100, right?  She was dancing with me on the floor.  She kissed me; there was some tongue there.  It was really awkward.  A hundred-year-old birthday party for a doctor, that’s a victory dance, right?  We get to do that.  I’m like yes.  Everyone’s like you’re the doctor, you’re my man, right?  All right, but I will assess health, all right?  

We focus on age.  We focus on health and quality measures.  We made a point that if the primary care are not comfortable, then we refer them, but we have to get them comfortable.  Do we need high-volume prostate cancer centers?  We probably don’t.  We probably need education, all right?  We need to focus on those who don’t need treatment.  Maybe we need risk calculators.  The bottom line is actually that we just need to have our eyes open to what this test can do and what it can’t do.  

I wanted to kind of end by showing you what I’ve done in my office.  This is a snapshot of a 12-month period that ended in March of last year, so we’re going to be updating it soon.  What I did is I like using the 4Kscore in my office because it’s easier for me, and I don’t need a DRE before that.  

That’s an important point.  Dave talked about the urine-based studies, okay?  Some of the studies need a rectal exam before.  Primary care do not like doing rectal exams, all right?  I just wrote an editorial that Dave helped me with.  We called it Make the DRE great again.  We just published that a few months ago.  The bottom line is the DRE is more important than just the prostate exam, but you don’t want to give it up.

If you come in for your physical, and I do a prostate exam.  When your labs come back and your PSA is abnormal, so now I get a biomarker.  I’m not going to do two DREs on you, all right?  You ain’t going to be happy.  This is not going to happen.  Maybe if I send you flowers the night before, but it probably is not going to happen.

I use a 4K because of that because you don’t need the DRE for that test.  What we found was these were patients—and I’ll show you the criteria—who had a screening level of the PSA at 1.5 or more.  

Now I want to explain that one for a minute.  What about a stable PSA?  If the PSA has been stable for five years at 3.2, 3.2, 3.2, the chances are that I’m not going to evaluate it, all right?  When we talk about 1.5, that is for the fresh, virgin PSA.  Is everyone following that okay?  All right, good.  

In my office, we had 94 4Ks that were done for the reasons that I’ll talk about in a minute.  Forty-one of them were less than 7.5, so nothing needed to happen.  Nothing needed—I’ll follow up with you in a year.  I’ll check the PSA again.  If it changes, I might re-evaluate, but the risk shows me that it’s low.  Fifty-three had a 4K above 7.5—equal or above.  Those patients were referred off, all right?  

There were 42 consults where 10 were pending at the time of this snapshot.  One refused.  Of those who got it, there was a conversation with the urologist, and a decision was made.

Now we’re working on a paper right now to talk about what decision process has to happen to get a biopsy or to not get a biopsy.  I think we need to clarify that a little bit, but that’s time for another discussion.  The data, as you see, of the 20 biopsies, 15 were negative and 5 were Gleason greater than 7.  That’s a 25 % rate because we were selective here.  

This is what we used as our criteria.  I looked at those 94 patients.  I said if I looked at my old criteria, a PSA of, if older than 60, greater than 2.5 or equal to 2.5, if age less than 60, a nodule, or a velocity change, we would’ve referred 69.  If I use an intelligent algorithm, the one that Dave and I have published, 4Kscore greater than or equal to 7.5 of prostate nodules, 53 were referred, all right?  

Think about the importance of that.  We reduced our referral rate down.   We have made this a cleaner process.  Is it a perfect process?  Absolutely not.  Are you still going to get those patients that get disease as Dave as mentioned?  Yes, you are.  The question is, by making this better, by screening this and using an intelligent way of going about this, are we improving the outcome for the patient?  Are we creating less patients?  I would say we absolutely are.

Finally, what have I learned after many, many years now of working on this.  Prostate cancer screening saves lives.  It does save lives, all right?  It saves people from metastatic disease, and it saves them from getting all the therapies they don’t want from metastatic disease.  

We are the gate keepers.  If you use all these guidelines and you look at what these do, remember that it affects primary care because we’re doing the majority of these.  If you make it difficult for us, the most likely response is I don’t want to deal with it.  

We want to avoid testing in men with little or no gain.  We need to teach that.  When we define the better way of doing it, this is what you do when you go back to your primary care colleagues next week.  Say look, I know there’s a controversy with the PSA, but let me show you why it’s a valuable first-line test.  Let me show you when shared decision making should happen only in that 27 % of patients that have a high PSA.  

There is a sweet spot.  You’ll find it.  You’ll have a consideration for a biomarker then.  You don’t need to worry about it before.  It’s like a vital sign as Dave had mentioned, and 1.5 is hugely important.  If you could teach this to your primary care colleagues, forget about all the various numbers you heard about the PSA.  Focus on 1.5, and you’re going to make their life easier.  Instead of 23 minutes of shared decision making, you’re going to make a 15-second determination of what to do and to do the right thing.  We’re going to be effective with that.  I want to thank you.