How the PSA Guidelines Missed the Mark with Primary Care, and How to Fix Them

Dr. Matt T. Rosenberg spoke at the 24th International Prostate Cancer Update on Saturday, February 22, 2014 on “How the PSA Guidelines Missed the Mark with Primary Care, and How to Fix Them.”

Presentation:

Keywords: decision-making, screening, PSA Guidelines, prostate cancer, education, primary care

How to cite:

Rosenberg, Matt T. “How the PSA Guidelines Missed the Mark with Primary Care, and How to Fix Them.” Grand Rounds in Urology. January 22, 2015. Accessed Apr 2024. https://dev.grandroundsinurology.com/prostate-cancer-matt-t-rosenberg-psa-guidelines/.

How the PSA Guidelines Missed the Mark with Primary Care, and How to Fix Them

As Leonard mentioned, I’m a bit of a hybrid. I’m trained as an urologist as well as a family doc. What I have been doing for the last couple of years is I have been acting as a liaison. I’m trying to teach my colleagues in primary care how to understand urologic disease, and likewise as in this case today I’m trying to teach urologists what primary care is looking at and what we know and in some cases what we don’t know.

Let’s jump right to it. I just want to review what’s out there. This is what we see. The message from the United States Preventative Service Task Force says that PSA screening is a “D” recommendation and physicians should not order a PSA screening unless they are prepared to engage in shared decision-making that enables an informed choice by patients. I’m not going to debate if this is a good decision or not. You guys can do that, and we’ve been doing that for a year. What I want to talk about is the practicality of this. Is this really a realistic thing to be asking of providers, and did the United States Preventative Service Task Force actually look into the reality of what would happen.

Let’s start with our first ARS. There we go. Which of the following statements from the American Academy of Family Practice website is false? Do not routinely screen for prostate cancer using a PSA or a DRE. Two, there is convincing evidence that PSA screening leads to substantial over-diagnosis. Three, many tumors will not harm patients, while the risk of treatments are significant. Four, physicians should not offer or order PSA screening unless they are prepared to engage in shared decision. Five, continue screening for prostate cancer with a DRE but do not routinely screen using a PSA.

Okay, and we see a little bit of a split. Most people said that number one was true, don’t screen with a PSA or DRE. What is on the website is this, do not routinely screen for prostate cancer using the PSA or the DRE, which I find surprising. If you actually read the Task Force recommendations, we don’t talk about that with a DRE. The American Academy jumped to that conclusion and leads to things like this. This is in Forbes magazine. This is what the lay press is showing us. This guy is all over the place talking about this, and he is recommending no more prostate exams, which I don’t think any of us thinks is a very good idea at all.

Let’s go to the next ARS. Shared decision in prostate cancer screening includes which of the following? Provision of information, elicitation of patient’s perspective, guiding final decision-making, four is one and three, and five is one, two, and three.

Great. 77% of you said it’s one, two, and three. That is the answer. If you remember Dr. Partin’s lecture a couple of days ago, he was saying this could be readily done in I think he said half an hour of time. All right, so then that leads us to our next statement, the tenants of shared decision-making. It is all three, provision of information, elicitation of patient perspective, and guiding decision-making. This is an onerous task and the question is do we have the time to do that.

The next question I’ll ask you is who is responsible for this shared decision-making in prostate cancer screening? Is it the patient? Is it the primary care doctor, which is why I’m wearing a tie today? Is it you guys out there in the audience, the urologist? If I showed this to a primary care audience they would be in stitches right now.

The question is who’s responsible for this. For us to answer that question, let’s think about the journey. We’ve been talking about metastatic prostate cancer all week. This is how it’s happens. A patient comes in. They are concerned for whatever reason or it’s a routine visit. They have their office visit. They get a history. I still believe a DRE is appropriate. I still believe a PSA is appropriate, but somewhere along the way something abnormal happens for us to recommend a visit to the specialist to get a biopsy. That specimen ends up with the pathologist and we go forward with that. That is the journey. The United States Preventative Service Task Force tells us we should intervene here, at that point right there at the PSA. Who owns that? That’s the domain of the primary care provider. We’re the ones doing everything up to the point of checking the PSA, including checking the PSA, and then referring them off to you. After we get the PSA, a biopsy and whatever intervention occurs is the domain of the specialist.

What do I have to backup that information? Dave showed this earlier. This is from the Aslani article which is showing that internal medicine and family medicine which makes up primary care, it’s over 88% of all PSAs are coming from this group. When the United States Preventative Task Force is making this recommendation that shared decision-making has to happen, this is the group we’re talking about, 88% right here. That’s who they are focusing on.

Let’s understand what happens with internal medicine. This is a little bit of an eye opener probably for everybody in this room. In internal medicine there is no mandated training in urology. In primary care there is a recommendation for one month of sub-speciality surgical training in ENT, in ophthalmology or in urology. For those of you in academic centers I would ask you how many of you see family practice residents? You probably see medical students, but how many of you actually see family practice residents? How many of you don’t see them routinely come through? I will ask this question in front of a group of primary care. I did this at the AUA once when we were doing the AUA update for primary care. I had the urologist on faculty. I had him turn around to the audience they were speaking to, because they were all up front. I said watch this. There were about 200 people in the audience. I asked how many of them had had any training in urology and five people raised their hands.

When we look at this we are talking about shared decision-making in prostate cancer screening done by a group of people that don’t have the background. It doesn’t mean they’re not smart people. It doesn’t mean they’re not working hard, but they don’t necessarily have that background. If they have to take on this task, how common in shared decision-making in primary care?
This is the Han article, and I’ll just read a couple of highlights here. Discussion of advantages or disadvantages 65% are none. Only 17% did both. Discussion of uncertainty, 88% is none. Extent of shared decision-making, look at the bottom, only 8% did all elements. So, I’ll you the question. How well are they doing shared decision-making process?

The next part of the Han study was the scarier part. If you look at this, he broke it up into three groups, no screening, low intensity screening, and high intensity screening. 44% of his group, and this is 22,000 patients, 44% didn’t have any screening. Of that 44% who had no screening, 88% of them—so you’ve got a number of just under 40%—had no discussion with the doctor. These people are out there and no one is talking to them about anything. That’s kind of concerning.

The conclusions from the Han study is that shared decision-making is uncommon. Most PSA screening occurs with incomplete or no discussion. The absence of shared decision-making applies not only to PSA screening but to the non-screening. Now, he did note that when they had the discussion they limited the high intensity screening, but that was in a minimal amount of patients.

Here’s another study. This is the Volk study. This was a research group of doctors, about 240 of them, and he asked them how they are doing the shared decision-making. Screening without discussion was 24%. Discuss and recommend screening, which was a minimal discussion, was 23%. Discuss and let the patient decide was 47, almost 48%. So, about 50% of the time it was completely inadequate what was going on. He asked them why they were having a problem with this, and there were concerns about malpractice. It was high amongst all groups, a little lower in the group that let the patient decide. They all thought it was standard of care to get a PSA, a little lower in the group that let the patient decide. Of most concern was right here, in the screen without discussion, I don’t have time, 42%. What you’re hearing from primary care is we don’t have the time to have this discussion. In the routine office, in my office the Tuesday before I came out here was busy because I was gone for three days. I saw 70 people. Do you think I had the time to have a 30 minute conversations with my physicals? No. I’m good at this. I’ve got a lot of background in this. I don’t necessarily need 30 minutes but I need some time to have that discussion. My colleagues who are not educated unfortunately the same way I am certainly don’t have that time.

The conclusions from the Volk study were kind of interesting. Physicians who discuss the harms and benefits were more likely to endorse the beliefs that scientific evidence does not support screening. There’s a little bias there. This is the bias that is coming out of primary care. I loved this one. Physicians who do not engage their patients should consider changing their habits. Bad primary care doctor for not having that discussion with patients, you need to change your habits. Well, in my day in seeing 70 patients I don’t know that I have any habits that I could change to add more on. Efforts to educate physicians about shared decision-making process should include countering the beliefs that perpetuate routine screening. Again, that’s inserting some bias that what we’re getting from primary care is you need to be ignoring this, and that’s what scares me. The reality is shared decision-making remains poorly integrated into primary care.

Let’s ask the next question. The guidelines have occurred. How does that affect behavior? This is the Cohn study. What he did was he looked at how they were screening pre-recommendation and post-recommendation. What he did was he compared it to total number of cholesterol screens because that is kind of a routine thing that we do. He found that the recommendations did cause a slight decrease, mainly in the younger group right here, 40 to 49, and in the older group, 70 to 79. There was a little bit of a change, not much if you consider 5.64 to 4.62 or 7.86 to 6.15. So, it’s a little bit of a decrease, not very high. It was significant.

If you looked at it from a breakdown, this is kind of surprising. When you compare family practice to internal medicine you see family practice there was no change in 63%. In internal medicine there was no change in 37%. It decreased in internal medicine 38%. It decreased in 25% in family medicine. It increased in both groups, too. I don’t think we’re seeing a trend here. I think we’re seeing that it’s all over the place and this may not be the right group to be focusing on.

In the Aslani study, and I already referred to this in terms of the number of PSAs that were checked, he looked at it from three different points. He looked at what was happening before the U.S. Preventative Task Force set up a paper saying we may be changing this. That was in 2009. Then he looked at it from after the recommendations were changed. He saw that the slope was going up before there was any concern. Then there was a concern because of the European trial and the American trial came out and said there might be a concern and maybe we shouldn’t be doing this. That went down. You see that from the slopes right here, the post-trial publication.

Then after the actual recommendation came out, it was interesting. There was no significant change in internal medicine. It only went down -0.02. There was an uplift in family practice. They started checking more. It wasn’t significant but it was approaching that. Urology went down. The clue from this is again that maybe we’re attacking the wrong group. This doesn’t belong in primary care. Maybe this belongs more in urology.

What both of those papers concluded was the evidence of decreases in PSA testing after the recommendations were there. They were weak but they were there a little bit. It may reflect influence of the American and European data. The adoption of changes in screening was performed more rapidly by urologists, and I think that’s a clue for us. It’s a clue for us that again the task force may have been pinpointing the wrong group.

Let’s look at it from another angle. What happens when you educate? People have been pushing this. Primary care, what happens when you educate primary care? There are two studies I’m going to reference. The first is the Feng study. What his group did was they looked at three groups of primary care. They took primary care. I know the font here is kind of small. They took primary care doing the routine stuff. They gave them a brochure and said you can hand this to your patients if you want. That’s what we have now. In the second group we’re going to educate them. We’re going to do some web based education for a significant amount of time and educate you on prostate disease. The third group right here, we’re going to educate the patient as well. We’re going to take the time and educate the patient before the visit so that he can talk to the educated physician.

Look at the highlighted areas. Discuss the controversies of prostate cancer screening, 60% for intervention, 39% for control. You can make your own interpretation of the data. It didn’t seem like a major change. Mentions no screening as an alternative, 63% for intervention and 26% for control. What about talking to the patients? How well did that go? Explaining why shared decision-making is important, 34% versus 21%. It’s not a tremendous change there for all of the education they got. Ask the patient their decision, which is one of the tenants of shared decision-making process, 39% versus 28%. What about guiding the patient? 62% of the intervention groups said tell the patient to think about the options versus 39%. Again, you can look at this data. There is a change with education but it’s not a tremendous change.
The conclusion that Dr. Feng made was all groups performed poorly with this. Remember these were groups that got education. They did note that when intervention was done the physicians became a little more neutral, which I think is a benefit, but it was only a trend that he saw in the paper.

In the next study, the Wilkes study, he used a similar algorithm where he looked at primary care that had no education, primary care that had education, and then primary care that had education plus patients who had the education. Let me go through that for a second. What they did was they took patients early to the office and they said we want you to sit and watch this video for an hour. After you watch this video for an hour we’re going to let you talk to the doctor and we’re going to see what happens. That’s the patient education. If you think about that, Lenny, you were just talking about four hours in your clinic for the multimodality approach for metastatic disease. Imagine the patients coming in for an hour just to get education on how to talk to the doctor about prostate screening.

The study objectives, perceived shared decision-making, rates of discussions about prostate cancer screening, and then the physician eliciting patient decision. The results were not very good. Perceived shared decision-making, control versus the medical ed group versus the med ed plus the patient education group were all the same on a zero to five small Likert scale. Rates of discussion of prostate cancer screening, the control was 38%. Med ed was only 3% higher. These were the doctors who were educated. They only did 41%. The med ed and patients, it was a little higher at 65%. In terms of the physician eliciting decision by the patient, neither suggesting for or recommending against the PSA test, control was 15%, med ed 33%, and med ed plus patient was 15%. Your highest was the group with the patient getting some education, but again that data is not tremendously impressive.

This is the scary part of the Wilkes study. These were poor results. He admitted they were poor results in the conclusion. These doctors were reimbursed for this education. When you take the doctors and educate them, that’s one thing. Now we took them, educated them, paid for that, and still the results were not very good. These were all the education from the web based information that they got right here. The conclusion is that he said there were no differences. He was very disappointed in the conclusion. He said this wasn’t good. This wasn’t good. We don’t know why this happened, but we’ve got to do better on that. He said that the physicians lectured the patients as opposed to discussing with it. Okay. He noted that when you did educate there was a leaning, a movement to more neutral counseling, which again is injecting this bias into that.

The scary thing was right here. Intervention just for physicians may not be sufficient. What the United States Preventative Task Force said with shared decision, we can’t achieve our goal by just educating the physician. We have to educate the patient as well. Now we’ve got a real problem. They’re still not addressing what’s going on but now we have to put the onus back on the patient, which in my case in my world is not very reasonable.

Where do we go from here? We have patient centered-centric solutions. This was one of the things that several papers have talked about. We maybe need to send patients information at home. Read up prostate cancer so that we can have a discussion when you come to the office. I can only imagine the folder I would be sending patients. Read up on cancer, read up on diabetes, read up on cholesterol, read up on depression. A month later after they’ve read the material they’d come in. One of the things that they talked about was a video in the waiting room. That was the Arora paper where they talked about we’re going to give you a video based thing you can see on the computer to talk about the cancer screening pros and cons. Again, these are good ideas but in a perfect world we have to ask how practical any of that is.

What about physician-centric solutions? We have brochures. We get handed these things all the time. Whether we give them out, that’s up to us as physicians. We probably should be doing more of that. Web based interactive education. We did that with the studies where we paid the physicians and they didn’t do very well. The question is what sense does that make. Again, it’s a good idea but how practical it is I don’t know.

Let’s think about common sense. This comes from an article from Scherger. This is a really nice article saying that the United States Preventative Service Task Force got it wrong. His comment was this. I can have a discussion with a patient about prostate cancer screening. We could decide not to do it. Five years later they develop metastatic prostate cancer. I’m going to court. They are going to say that I did not educate them well enough. They didn’t get it. It doesn’t matter what they signed, what they did. I’m not educated enough in prostate cancer to have that appropriate discussion. That’s what he’s saying. We would get diced and sliced up there.

The United States Preventative Service Task Force should have criticized the response to the PSA test results, not criticizing the PSA. What we need is we need to come up with a solution. This is a collaboration between the patient in the discussions, the PCP and the urologist. That’s the only way to get the better outcomes.

Who knows where this quote comes from, “I will not cut for stone”? That’s from the Hippocratic Oath. All right? If you read into this, it’s funny. When I was writing the talk I hadn’t intended this, but I was reading another article and I saw this. “I will not cut for stone. I will refer that to the specialist.” That’s what Hippocrates told us to do because as general practitioners you didn’t want me to cut on the patient if I didn’t know what I was doing. I asked the same question. Why should I as a primary care doctor have a discussion about a disease state that I don’t have the appropriate education in? I could get it, but I have to do a lot of other things as a primary care doctor. How practical is that? In remembering how to do PSA screenings we need to remember the Hippocratic Oath. In the current form the recommendations represent a complete losing battle as far as I’m concerned.

Now, I mentioned the problems. How are we going to fix it? We could train primary care on shared decision-making process. We’ve already looked at that. It doesn’t work. I can tell you guys to do it. I can ask all of you to have that one half an hour discussion with every patient about what’s going on. We already have a shortage of urologists. I think this would only intensify the problem and it’s not practical. I think the reality is we have to get rid of shared decision-making and focus on something that makes sense. In this form this makes no sense. I’m not saying that we shouldn’t have discussions with the patients. I’m not saying the data shows us that we may not need to be screening everybody. The way that they are doing it, it makes no sense.

The more common sense comes from what we’ve learned with the AUA guidelines. I think this kind of reaches toward what is good sense, but it doesn’t get us there all the way. We need to avoid PSA tests in men with little to no gain. We need to focus on age, on health, and on quality measures. We need to refer to high volume centers. Maybe we need to focus who has this discussion if the patient wants to have it. Also, we don’t treat those who do not need treatment.

Who has heard of what the American Heart Association set out with the risk calculators for cholesterol? Are you guys familiar with those? We used to treat LDL. Does everyone know that? All right, so if you came into my office with a high LDL I would put you on medication if it was above 100. The American Heart Association in October came out and said we don’t think that’s right. We’re going to put you into a risk calculator and then we’re going to determine your risk. If your risk over a 10 year period is greater than 7.5 I put you on a statin. It’s not perfect. The press has basically said it’s not perfect, but it’s better than what we had. It takes some of the subjectivity out of it.

Maybe what we need is a risk calculator for the PSA. I have a 50-year-old guy who comes into the office. His PSA is 1.2. He has this family history. He has these comorbidities. This is your risk of actually having something bad and succumbing to the disease. But we don’t do the calculator without having the PSA. We need to get the PSA and respect the number.

I’ll end on this. The logic and the evidence provides the answer. Stopping before we get the PSA makes no sense. That ties my hands behind my back. Getting the PSA is appropriate. Putting it into an algorithm is appropriate. Intervening before we get a biopsy is what we need to do. Thank you.