Shared posts

16 Oct 21:04

These prostate cancer screening campaigns are giving men the finger

by Gary Schwitzer
John Epling

When will the DRE die?

Fifty years ago, in a golden moment of television comedy shows, Rowan & Martin’s Laugh-In program regularly featured “The Flying Fickle Finger of Fate” award.  Wikipedia says it “recognized actual dubious achievements by public individuals or institutions.” Do a Google search.  You’ll quickly see how popular this award became.

Yes, I’m dating myself by going back 50 years.  But mine is the generation that often becomes obsessed with being given “the finger” by doctors for digital rectal exams (DRE) looking for prostate cancer.

And it is in that context that we propose to revive The Flying Fickle Finger of Fate award.  It could be awarded to prostate cancer screening promotions that deliver a new Laugh-In about prostate exams. While they are funny, it is no joke that they fail to educate men about the evidence that raises important questions about the DRE.  The following is what you will not learn in these promotional campaigns.

An UpToDate analysis states:

We suggest not performing digital rectal examination (DRE) for prostate cancer screening either alone or in combination with prostate-specific antigen (PSA) screening. Although DRE has long been used to diagnose prostate cancer, no controlled studies have shown a reduction in the morbidity or mortality of prostate cancer when detected by DRE at any age.

There are inherent limitations to the DRE. It can detect palpable abnormalities in the (areas) of the prostate gland where the majority of cancers arise; however, other areas of the prostate where cancer occurs are not reachable by a finger examination. Furthermore, the majority of cancers detected by DRE alone are clinically or pathologically advanced, and (early) stage T1 prostate cancers are nonpalpable by definition.

…there is no high-level evidence that DRE screening improves survival outcomes.

The author of that analysis is Richard Hoffman, MD, MPH, the director of the Division of General Internal Medicine at the University of Iowa Carver College of Medicine.  He wrote to me, “Even the American Urological Association, whose members are probably most adept at performing DREs, admits that evidence is insufficient to justify using DRE as a first-line screening test.”

A systematic review and meta-analysis of the evidence published in the Annals of Family Medicine this year concluded: “Given the considerable lack of evidence supporting its efficacy, we recommend against routine performance of DRE to screen for prostate cancer in the primary care setting.”

After years of going through the perfunctory DRE from my own doctor, I finally challenged him about the evidence.  That was a couple of years ago.  He hasn’t put the glove on to give me the finger ever since.

A new low for an awareness campaign that doesn’t raise awareness

Mike Rowe, TV host of “Dirty Jobs” (no comment), probably had good intentions when he taped a prostate awareness public service announcement. But the end result is no public service because it is filled with misinformation. It features Rowe getting a DRE from his own doctor. In the video, Rowe states that with early detection, the five-year survival rate for prostate cancer is nearly 100 percent. We’ve written about how such five-year survival rate stats are often misleading. Nearly 100%? I recall one observer saying about such a statistic, “Hmmm. Almost as if it’s not a cancer.”  Indeed, early detection with a PSA blood test – also discussed in the video – may mean finding something that would never harm a man in his lifetime but would still end up being treated. So if you’re detecting many things that won’t cause harm, the five-year survival rate is bound to look pretty impressive. Rowe’s doctor delivers concise, catchy, but totally misleading comments about the only two reasons why you should not do a DRE.  “One, the doctor doesn’t have a finger,” he says.  “And the other is if the patient doesn’t have a rectum.”  Yuck it up, then look it up to see evidence-based reasons why you may not want to have a DRE. Rowe compounds the problem by telling men they have “no excuse” not to get a DRE.  That kind of heavy-handed blame game is often employed in screening promotions.  It’s a wrong-headed approach.  Dr. Hoffman wrote to me: “Missing from the amusing video is the part where the doctor tells the patient that getting screened for prostate cancer is not a slam dunk decision.  There are benefits—and harms—to screening and men should be helped to make informed decisions that best reflect their values and preferences.”  In other words, a fully-informed shared decision making discussion between patient and physician is needed, not a “no excuse” mandate.


In Canada, the Prostate Cancer Canada organization has gone all-out with a “Famous Fingers” comedy approach to DRE.  It features famous fingers through time that could be employed in DREs.  Take your pick:  Beethoven, Big Foot, Paul Bunyan, Sherlock Holmes, Winston Churchill, Al Capone, Genghis Khan, Napoleon, Babe Ruth, Frankenstein’s monster.  The campaign is woefully void of evidence.

Such campaigns are giving you the finger

In the end (no puns intended throughout), these campaigns do a disservice to uninformed men.  They generate laughs but present no evidence.  It’s an absurd way to conduct an awareness campaign.  It may raise awareness of the comical creativity of an advertising agency but does nothing to deliver the facts that men need to make an informed decision.

They’ve given you the finger, guys.  You can fight back by arming yourself with knowledge of the evidence. We’ve provided some highlights above. It doesn’t require a PhD to understand what the science shows and what it doesn’t.  If you still choose to pursue the DRE or the PSA blood test, more power to you.  At least you won’t be swayed by a rah-rah, single-minded promotional campaign on behalf of special interests who stand to gain by your business and by further fundraising.



07 Aug 19:48

New High Blood Pressure Guidelines: Back on Track With Lower Treatment Goals, but Implementation Challenges Abound

by Randall S. Stafford
John Epling

"Back on track"??? Reading this abstract makes me dizzy - the author seems to agree with these guidelines while providing a laundry list of reasons why it will be very hard to implement them...nevermind the question of whether we should.

The recently released 2017 High Blood Pressure Guidelines depart from past guidelines in both their approach and recommendations. Developed by multiple health organizations, including the American College of Preventive Medicine, the guidelines continue to define normal blood pressure as <120/80 mmHg, but now define hypertension as ≥130/80 mmHg (previously ≥140/90 mmHg). This change categorizes 101 million Americans (46% of adults) as hypertensive (compared to 32% previously). The guidelines rely heavily on findings from the Systolic Blood Pressure Intervention Trial (SPRINT).
02 Aug 00:48

Development, administration, and validity evidence of a subspecialty preparatory test toward licensure: a pilot study

by John Johnson, Alan Schwartz, Matthew Lineberry, Faisal Rehman and Yoon Soo Park
Trainees in medical subspecialties lack validated assessment scores that can be used to prepare for their licensing examination. This paper presents the development, administration, and validity evidence of a ...
10 Jul 00:25

The Role of the Physician When a Patient Discloses Intimate Partner Violence Perpetration: A Literature Review

by Penti, B., Timmons, J., Adams, D.

Intimate partner violence (IPV) is prevalent and has lasting impacts on the health and well-being of the entire family involved. Primary care physicians often interact with male patients who perpetrate IPV and are in a role potentially to intervene, but there is very little research and guidance about how to address perpetration of IPV in the health care setting. We reviewed the existing literature research related to physicians' interactions with male perpetrators of IPV and summarize the recommendations. If a male patient discloses IPV perpetration, physicians should assess for lethality, readiness to change, and comorbid medical conditions that could impact treatment, such as substance abuse and mental illness. Experts agree that referrals to a Batterer Intervention Program should be the primary intervention. If there are no locally available Batterer Intervention Programs or the patient is unwilling to go, then a physician should refer the abuser to a therapist who has been trained specifically to work with perpetrators of IPV. In addition, physicians should be prepared to offer education about the negative impact of IPV on the victim, on any children, and on the abuser himself. Physicians should address any untreated substance abuse or mental health issues. Referral to couples therapy should generally be avoided. Physicians should continue to have regular follow-up with their male patients to support them in changing their behavior. Further research is needed to assess the role the health care system can have in preventing IPV perpetration.

01 Jun 19:13

Primary Care Outcomes Questionnaire: psychometric testing of a new instrument

by Mairead Murphy, Sandra Hollinghurst, Sean Cowlishaw, Chris Salisbury
John Epling


BackgroundPatients attend primary care for many reasons and to achieve a range of possible outcomes. There is currently no Patient Reported Outcome Measure (PROM) designed to capture these diverse outcomes, and trials of interventions in primary care may thus fail to detect beneficial effects.AimThis study describes the psychometric testing of the Primary Care Outcomes Questionnaire (PCOQ), which was designed to capture a broad range of outcomes relevant to primary care.Design and settingQuestionnaires were administered in primary care in South West England.MethodPatients completed the PCOQ in GP waiting rooms before a consultation, and a second questionnaire, including the PCOQ and seven comparator PROMs, after 1 week. Psychometric testing included exploratory factor analysis on the PCOQ, internal consistency, correlation coefficients between domain scores and comparator measures, and repeated measures effect sizes indicating change across 1 week.ResultsIn total, 602 patients completed the PCOQ at baseline, and 264 (44%) returned the follow-up questionnaire. Exploratory factor analysis suggested four dimensions underlying the PCOQ items: health and wellbeing, health knowledge and self-care, confidence in health provision, and confidence in health plan. Each dimension was internally consistent and correlated as expected with comparator PROMs, providing evidence of construct validity. Patients reporting an improvement in their main problem exhibited small to moderate improvements in relevant domain scores on the PCOQ.ConclusionThe PCOQ was acceptable, feasible, showed strong psychometric properties, and was responsive to change. It is a promising new tool for assessment of outcomes of primary care interventions from a patient perspective.
15 May 20:49

5-a-day fruit and vegetable food product labels: reduced fruit and vegetable consumption following an exaggerated compared to a modest label

by K. M. Appleton and H. J. Pidgeon
John Epling

Unintended consequences, huh?

Food product labels based on the WHO 5-a-day fruit and vegetable (FV) message are becoming increasingly common, but these labels may impact negatively on complementary or subsequent FV consumption. This study ...
14 May 01:20

Is Medical Education a Public or a Private Good?

by Lucey CR.
John Epling

This is a really important question that I don't think we wrestle with openly. In tracing the decline of medical student idealism over the course of their training, I think they are making a switch in this thinking.

The US educational system has 2, at times competing, goals. Education is commonly viewed as a public good, designed to prepare the workforce that the country needs and to educate citizens who contribute to the health of the US democracy. However, education is also seen as a private good, geared toward helping the individual maximize social mobility and personal success. From Virchow in the 19th century to Frenk in the 21st century, thought leaders have embraced the view of medical education as a predominantly public good rather than a private one, maintaining that the purpose of medical education is to improve the health of communities and to decrease the burden of illness and disease. The annual JAMA publication of data describing the demographic composition, geographic distribution, and specialty focus of learners and programs in US undergraduate medical education and graduate medical education provides an opportunity for the medical profession to once again consider whether the medical education community is designed to strike the appropriate balance between providing a public good and a private good.
14 May 01:17

Anscombe’s Quartet: 1980’s Edition

by Method Matters
John Epling

The importance of visualizing data prior to analyzing it.

(This article was first published on Method Matters, and kindly contributed to R-bloggers)

In this post, I’ll describe a fun visualization of Anscombe’s quartet I whipped up recently.

If you aren’t familiar with Anscombe’s quartet, here’s a brief description from its Wikipedia entry: “Anscombe’s quartet comprises four datasets that have nearly identical simple descriptive statistics, yet appear very different when graphed. Each dataset consists of eleven (x,y) points. They were constructed in 1973 by the statistician Francis Anscombe to demonstrate both the importance of graphing data before analyzing it and the effect of outliers on statistical properties. He described the article as being intended to counter the impression among statisticians that ‘numerical calculations are exact, but graphs are rough.’

In essence, there are 4 different datasets with quite different patterns in the data. Fitting a linear regression model through each dataset yields (nearly) identical regression coefficients, while graphing the data makes it clear that the underlying patterns are very different. What’s amazing to me is how these simple data sets (and accompanying graphs) make immediately intuitive the importance of data visualization, and drive home the point of how well-constructed graphs can help the analyst understand the data he or she is working with. 

The Anscombe data are included in base R, and these data (and R, of course!) are used to produce the plot that accompanies the Wikipedia entry on Anscombe’s quartet.

Because the 1980’s are back, I decided to make a visualization of Anscombe’s quartet using the like-most-totally-rad 1980’s graphing elements I could come up with. I was aided with the colors by a number of graphic design palettes with accompanying hex codes. I used the excellent showtext package for the 1980’s font, which comes from the Google font “Press Start 2P.” (Note, if you’re reproducing the graph at home, the fonts won’t show properly in RStudio. Run the code in the standalone R program and everything works like a charm). I had to tweak a number of graphical parameters in order to get the layout right, but in the end I’m quite pleased with the result.

The Code

# showtext library to get 1980's font
# "Press Start 2P"
# add the font from Google fonts"Press Start 2P", "start2p")

width=11,height=8, units='in', res=600)

op par(las=1, mfrow=c(2,2), mar=1.5+c(4,4,.5,1), oma=c(0,0,5,0),
lab=c(6,6,7), cex.lab=12.0, cex.axis=5, mgp=c(3,1,0), bg = 'black',
col.axis = '#F2CC00', col.lab = '#A10EEC', family = 'start2p')
ff y ~ x
for(i in 1:4) {
ff[[2]]"y", i, sep=""))
ff[[3]]"x", i, sep=""))
lmi lm(ff, data= anscombe)
xl substitute(expression(x[i]), list(i=i))
yl substitute(expression(y[i]), list(i=i))
plot(ff, data=anscombe, col="#490E61", pch=21, cex=2.4, bg = "#32FA05",
xlim=c(3,19), ylim=c(3,13)
, xlab=eval(xl), ylab=yl,
family = 'start2p'
abline(lmi, col="#FA056F", lwd = 5)
axis(1, col = '#FA1505')
axis(2, col = '#FA1505')
mtext("Anscombe's Quartet", outer = TRUE,
cex = 20, family = 'start2p', col="#FA1505")

The Plot


In this post, I used data available in R to make a 1980’s-themed version of the Anscombe quartet graphs. The main visual elements I manipulated were the colors and the fonts. R’s wonderful and flexible plotting capabilities (here using base R!) made it very straightforward to edit every detail of the graph to achieve the desired retro-kitsch aesthetic.

OK, so maybe this isn’t the most serious use of R for data analysis and visualization. There are doubtless more important business cases and analytical problems to solve. Nevertheless, this was super fun to do. Data analysis (or data science, or whatever you’d like to call it) is a field in which there are countless ways to be creative with data. It’s not always easy to bring this type of creativity to every applied project, but this blog is a place where I can do any crazy thing I set my mind to and just have fun. Judging by that standard, I think this project was a success.

Coming Up Next

In the next post, I’ll do something a little bit different with data. Rather than doing data analysis, I’ll describe a project in which I used Python to manage and edit meta-data (ID3 tags) in mp3 files. Stay tuned!

To leave a comment for the author, please follow the link and comment on their blog: Method Matters. offers daily e-mail updates about R news and tutorials on topics such as: Data science, Big Data, R jobs, visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...
04 May 19:38

Despite screaming headlines, England’s breast cancer screening computer glitch didn’t kill anyone

by Kevin Lomangino
John Epling

The headlines for this story were disturbing - I don't agree with the hyperbole generated by the modeling of the attributable deaths. But most interesting to me was the fact that preventive screenings were clearly the domain of the health system, seemingly not the responsibility of the GPs. It's a different system.

A bureaucratic snafu in England has generated international headlines bemoaning the deaths of hundreds of women.

Some coverage, such as this CNN story picked up by KSAT in San Antonio, brought visions of a mass casualty event on the scale of a terrorist attack or airline crash:

Dozens of other news outlets chimed in with similarly explosive framing:

The coverage focuses on a computer error at the National Health Service that caused some 450,000 older women (ages 68 to 71) not to receive scheduled invitations for a mammogram.

As a result, the stories claim, between 135 and 270 women may have died from breast cancer that could have been detected and treated earlier with lifesaving results. UK Health Secretary Jeremy Hunt is referenced as the source of these figures.

Lack of mammograms has never killed anyone

What’s the problem with these stories?

For starters, neither breast cancer screening nor any other form of cancer screening has been shown to reduce overall mortality, which is the best measure of whether lives are actually being saved or not.

For this reason, no one can say with any certainty whether mammograms would have saved these women — or whether the lack of a mammogram killed them.

Breast cancer screening has been shown to modestly reduce deaths from breast cancer. (About 1,000 50-year-old women need to be screened for 10 years to prevent one death from breast cancer.) However, this small benefit evaporates when other causes of death, such as cardiovascular disease, are factored in.

Why is this important? It’s possible that for some women, radiation and other toxic therapies that are used to treat screen-detected cancers may ultimately increase the likelihood of dying earlier from some other cause–canceling out any reduction in deaths from breast cancer.

That’s why it’s treacherous to consider lives “saved” by a reduction in deaths from a specific disease. The treatment of that disease may ultimately increase deaths through some other unrelated mechanism.

Overdiagnosis not accounted for

Another problem with this coverage is that it doesn’t adequately acknowledge the risk of overdiagnosis — the detection of cancers that never would have caused health problems or death. Some cancers are slow-growing and might never pose a risk to the patient, while others are very aggressive and may spread despite the earliest possible detection and treatment.

A recent Danish study estimated that about one-third of all mammography-detected breast cancers represent overdiagnosis. This means that women treated for these cancers had no chance to benefit, but experience all the harms of cancer treatment that may include surgery, chemotherapy, and radiation.

‘Epitome of disease-mongering’

Christine Norton, president and co-founder of the Minnesota Breast Cancer Coalition and a breast cancer survivor, called the headline and first paragraph of KSAT’s CNN story “the epitome of disease mongering.”

“Combined they engender fear not only in older women in the UK who were among those not invited to be screened but also in any woman of any age who has thought about the possibility of breast cancer,” she said.

UK Health Minister Jeremy Hunt addresses Parliament.

Norton said it was inaccurate and irresponsible to blame up to 270 deaths on a failure to invite women for breast cancer screening. “The more accurate and responsible statement is the quote in paragraph 5 [of the CNN story] by the UK Health Secretary, Jeremy Hunt, who noted that ‘At this stage it is unclear whether any delay in diagnosis will have resulted in any avoidable harm or death.’ Based on multiple studies, it’s more accurate to state that breast cancer screening results in false positives, over-diagnosis, and over-treatment.”

The point is not that mammograms are useless and should be avoided — it’s that the benefits are probably smaller than most women have been led to believe, and that those benefits are accompanied by significant harms that many women have never been warned about. Women are making decisions about mammograms based on incomplete, imbalanced information. News coverage of mammograms doesn’t help when it reinforces misconceptions about the effectiveness of screening and doesn’t address the harms.

Who will be notified?

According to the CNN coverage, “Health officials will contact the next of kin of women who are thought to have missed a scan and subsequently died of breast cancer, Hunt said. They will apologize and offer a process to establish whether the error led to an earlier death, and if compensation might be payable.”

This statement raises some interesting questions:

  • Will officials also contact women who weren’t notified and let them know that they may have been spared unnecessary and harmful treatment that might have been prompted by a positive screen?
  • Will they also apologize to the women who were notified and whose screen-detected tumors were treated unnecessarily? Will they inform those women of the significant chance that their cancer was overdiagnosed?

None of the superficial stories that I read about this situation asked these legitimate questions.

They missed an opportunity to bring balance to what is typically a one-sided message about breast cancer screening.

Editor’s note: A previous version of this story referred to “the UK’s breast cancer screening computer glitch” in the headline. References to a “UK” glitch have been changed to “England,” since the glitch only affected women in England and not Scotland, Wales, or Northern Ireland.  

22 Apr 18:28

Pseudoscience: The Conspiracy Against Science

by Harriet Hall
An excellent new book examines pseudoscience in 22 essays by prominent scientists from various fields.
17 Oct 01:23

Low Prevalence of Criteria for Early Screening in Young-Onset Colorectal Cancer

by Frank W. Chen, Vandana Sundaram, Thomas A. Chew, Uri Ladabaum
Colorectal cancer (CRC) incidence in adults aged <50 years is increasing in the U.S. despite an overall decline in CRC incidence in the general population.1,2 This trend has received attention in the general and medical press.3 The underlying factors for this trend are unknown.4