More proof medical error is not the third leading cause of death

Over the last 20 years, estimates of the number of deaths caused by medical error have risen from 44,000-98,000 in 1999 [1] all the way up to 440,000 [2] and 251,000 [3]. Despite my efforts [4, 5] and those of others [6, 7] to debunk these guesses, they continue to permeate the lay press. If you Google “third leading cause of death,” you will find countless headlines naming medical error.

The papers claiming medical errors cause so many deaths assume that all complications result from errors and all complications are preventable. They extrapolate their final numbers from small studies not designed to or capable of estimating deaths due to medical error nationwide.

The most recent figures available from the National Hospital Discharge Survey [8] state that the number of hospital deaths dropped from 776,000 in the year 2000 to 715,000 in 2010. It is simply not plausible that 251,000 (35%) or 440,000 (61%) inpatient deaths are due to medical error.

A recent study [9] from Norway found that of 1000 consecutive in-hospital deaths reviewed, only 42 (4.2%) were judged to be probably (greater than a 50% chance) to definitely avoidable.

Are neckties dangerous to your health?

Wearing a necktie significantly decreases cerebral blood flow says a new study in the journal Neuroradiology. This finding caused a minor flurry of activity on Twitter, and as usual, the press sensationalized and misinterpreted the study’s results.

Here’s a headline from the Deccan Chronicle: “Wearing ties hamper [sic] productivity in office; here’s why.” The sub- heading is “Study suggests men who wear T-shirts in the office may produce better work.” T-shirts were not mentioned in the paper. The name of the journal that published it was incorrect in the article too.

Forbes didn’t do much better. It’s lede is “Neckties are stupid. Could they also make you stupid?” The paper said nothing of the kind.

Facial exercises to make you look younger? I don't think so.

One would think a study covered by the New York Times would be both scientifically valid and important. Apparently, that is not always the case.

Under the headline “Facial Exercises May Make You Look 3 Years Younger” is a story about a research letter published in JAMA Dermatology. The Times article concludes with a quote from the lead author, “But for now, it is reasonable to consider contorting and pinching up your face if you wish to try to look younger.”

Is it reasonable? Let’s see if this 1½ page research letter proved its point.

Three new studies confirm germs are everywhere

These ubiquitous germs don’t seem to be harming anyone, but reporting on the studies generates lots of clicks.

For many years kitchen sponges have been known to harbor bacteria. Now comes the news that cleaning those sponges not only doesn’t work, it may make the situation worse by promoting the overgrowth of potentially disease-causing bacteria—for example Moraxella osloensis.

The New York Times reports German researchers found kitchen sponges contained 362 different types of bacteria and as many as 82 billion bacteria per cubic inch of space. The senior author of the study said, “That’s the same density of bacteria you can find in human stool samples” [but not the same types of bacteria] and suggested replacing kitchen sponges frequently.

These revelations were based on bacterial DNA and RNA samples from 14 [yes, just 14] used sponges. Note the use of the word “potentially” to describe the pathogenicity of Moraxella. A PubMed search for this microbe back to 1968 yielded only 82 references, many of which were not pertaining to any human illnesses. The few case reports of infections involved patients who were immunosuppressed.

Good patient safety news you didn’t hear about

In the last five years, there’s been a 21% reduction in hospital acquired conditions (HACs) says a report by the Agency for Healthcare Research and Quality. This means that patients suffered 3.1 million fewer HACs than if the HAC rate had stayed at the 2010 level.

Since 2011, the decrease in HACs has reduced healthcare costs by an estimated $28.2 billion and has saved almost 125,000 lives.

This graphic summarizes the AHRQ findings.

Central line-associated bloodstream infections have fallen by 91%, and postoperative venous thromboembolism by 76%. Here’s a chart that shows the percent decreases in HACs.

The report said the reasons for these improvements “are not fully understood,” but might be due to the following:

• Financial incentives created by the Centers for Medicare & Medicaid Services (CMS) and other payers’ payment policies,
• Public reporting of hospital-level results,
• Technical assistance offered to hospitals by the Quality Improvement Organization (QIO) program, and
• Technical assistance and catalytic efforts of the HHS PfP [[Pay for Performance] initiative led by CMS.

An thorough Google search found a few articles about this important and positive AHRQ report. Were they in the New York Times, Washington Post, Newsweek, US News, or The Daily Mail?

No. the news could only be found on HealthcareIT Analytics, Fierce Healthcare, the website of the Healthcare Association of New York State, Pharmacy Practice News, and HealthcareIT News where I obtained the multicolored graphic above.

Why do you suppose no major media outlet reported the story?

Good news doesn’t get clicks.

Guarded condition

For the last week or so, updates on the condition of the surviving victims of the Orlando mass shooting featured headlines like this:

“6 Orlando shooting victims remain ‘critically ill,’ another 5 patients in ‘guarded’ condition”

Even the trauma center, Orlando Health, tweeted this the other day:

“No surgeries on the victims are scheduled for today. 4 patients are in critical condition, 2 are guarded, and 12 are stable.”

I think we all have an idea of what critical and stable mean [more about this below], but what about guarded?

I’ve never known what guarded means and wondered if it was just me. To find out, I conducted an informal 24-hour Twitter poll.

Here are the results:

By a 2% margin, “I have no idea” was the most common response, but 35% thought guarded condition meant between serious and critical, and 21% thought guarded was better than serious.

When bad research is not critically reported by journalists

Yesterday I posted a critical review of the study "Medical error—the third leading cause of death in the US."

I did not have time to address the media coverage of the paper, but fortunately the website HealthNewsReview did.

Their post started with "Seemingly all the major outlets carried the story, with headlines so alarming that they’d have any conscious hospital patient demanding an immediate discharge."

They called attention to headlines which included the word "now" such as these:

CBS: Medical errors now 3rd leading cause of death in U.S., study suggests
Washington Post: Researchers: Medical errors now third leading cause of death in United States
Nature World News: Medical Errors Now the Third Leading Cause of Mortality in U.S.

They pointed out that the headlines were similar to the title of a press release issued by the PR department at Johns Hopkins and that using "now" in the headline implies that the incidence of deaths caused by medical error has increased.

Many journalists miss major flaws in research papers

Google "jet air dryers" and you will find more than 100 articles about a recent study claiming that jet air dryers dispersed 60 times more "germs" than warm air dryers and 1300 times more "germs" than paper towels.

Here are a few of the juicier headlines:

It wasn't about germs. The study involved measuring the dispersal of a benign virus from the hands of research subjects at various distances and heights from the different hand drying methods.

So what's wrong with it?

Less crowded EDs = less waiting & "big data" isn't so big

Two recent papers caught my eye. As usual, I will preface my critiques with a disclaimer that I have never written a Nobel Prize-winning paper myself.

The first paper, “Emergency department ‘undercrowding’ is associated with decreased waiting times” appeared online in the journal Emergency Medicine Australasia.

The authors looked at emergency department patient numbers and waiting times before and after a nearby tertiary care hospital opened a new emergency department.

Their main finding was that after the new ED opened, their ED saw 28% fewer patients with a concomitant decrease in patient waiting times of 15 minutes from 26 to 11 minutes with p < 0.001, a significant difference.

They concluded, “Wait times are strongly associated with patient presentation numbers.” Furthermore, “Controlling demand may benefit patient processing, flow, and patient perceptions of level of care.”

Appendicitis: Continuing debate about treatment

The other day on Twitter, Kenny Goldberg (@kghealth), a health reporter at KPBS News in San Diego, asked me, "Why all the stories on antibiotics vs. surgery for appendicitis? Are appendectomies inherently dangerous?" My answer was "Great question. The answer is a resounding 'No.' The complication rate is very low."

Yet the papers keep coming.

A new systematic review of all the randomized controlled studies on appendicitis found important shortcomings in all of them. Here are a few:

Bias in selecting patients was a problem in all six of the studies reviewed. Diagnostic criteria for inclusion in the studies were not standardized. Some of the studies enrolled patients with clinically diagnosed appendicitis only. Since some patients may not have had appendicitis, they would probably have improved regardless of how they were treated.

Patients were treated with a variety of antibiotics, Since most of the studies were done in Europe, open appendectomy was the more common surgical intervention. Laparoscopic appendectomy results in fewer complications and shorter lengths of stay than the traditional open procedure.

Follow-up in five of the six studies was one year with only one study following patients as long as a median of 17 months. Rates of recurrent appendicitis necessitating appendectomy ranged from 24% to 60% with an average of 35.4%. What will the recurrence rates be at 3 years? 5 years?

The authors concluded that although more evidence for treating appendicitis with antibiotics has emerged, the comparative effectiveness of that strategy is still unknown. They recommend that patients should be enrolled in clinical trials or registries to help answer this therapeutic question.

The second recent paper involves two issues I have commented about many times—research and medical reporting.

It's a study of 102 pediatric patients between the ages of 7 and 17 with uncomplicated appendicitis as judged by CT scan parameters. After informed consent was discussed, parents were permitted to choose the therapeutic arm, antibiotics or laparoscopic surgery.

Of the 629 patients who presented with acute appendicitis during the study period, only 102 (21%) met the study's inclusion criteria of whom 37 were selected for antibiotic therapy by their parents.

During the median follow-up period of 21 months, 9 (24.3%) patients initially treated with antibiotics had to undergo appendectomy.

I blogged about this study's preliminary results when they were published back in 2014. If you would like more details about its limitations, read that post.

The inadequacies of medical reporting on this paper were rather glaring. Under the headline "Not all kids with appendicitis need surgery. Antibiotics can work just fine," the Boston Globe's new website Stat News said the following:

“'Their parents began to question whether they needed surgery [for appendicitis],' said [lead author] Dr. Peter Minneci, a pediatric surgeon at Nationwide in Columbus, Ohio. Minneci decided to answer the question with a controlled study." Sorry folks, this wasn't a controlled study.

The New York Times reported: "The surgery group had more complications and two of those who chose antibiotics had to be readmitted to the hospital for appendectomies in the first 30 days." This is misleading because although 5 of 65 patients in the surgery group had postoperative complications compared to none of the 9 who eventually had appendectomies in the antibiotic group, the difference was not statistically significant (p = 1.0, Fisher's exact test).

But the most interesting thing about this paper was an entire page explaining why allowing parents to select the therapy was a better method than randomizing patients to one group or the other. It's very clever and must be read to be appreciated.

Here is an excerpt: "The patient choice design allows a therapy to be aligned with the preferences of the patient and his or her family, thereby minimizing the potential negative effects of preferences."

I don't know about you, but if I or anyone in my family had appendicitis, my preference would be for a laparoscopic appendectomy.

Hospitals Mess Up Medications in Surgery—a Lot

Yes, that was the inflammatory headline on Bloomberg Business News last week. It is great click-bait, but factually off base because the research it refers to was done at only one hospital.

Here's what the study found. During 277 operations with 3,671 medication administrations observed at the Massachusetts General Hospital, 193 (5.3%) involved a medication error or an adverse drug event. One or more errors or adverse drug events occurred in 124 (44.8%) of the procedures.

In all, 40 (20.7%) adverse drug events were not preventable—for instance, an allergic reaction to a drug that was not known about before. Of the remainder, “32 (20.9%) of the errors had little potential for harm, 51 (33.3%) led to an observed adverse drug event and an additional 70 (45.8%) had the potential [emphasis added] for patient harm."

Sounds bad, but the Bloomberg article goes on to say "While all the errors observed in the study had the potential to cause harm, only three were considered [potentially] life-threatening, and no patients died because of the mistakes. In some cases, the harm lay in a change in vital signs or an elevated risk of infection."

Do doctors charge too much?

We all know that some doctors’ fees are excessive. I have blogged about this myself citing a neurosurgeon’s $117,000 charge for assisting on a case.

We also know that doctor bashing is a popular sport right now.

In an otherwise reasonable article about high-deductible health insurance on Vox.com, reporter Sarah Kliff’s second paragraph read as follows:

The bolded text was hyperlinked to a Washington Post piece about a study that showed wide variations in hospital charges for appendectomies in California. The study was not about physician fees. No matter how difficult the case was, no surgeon would ever have been paid $186,955 for performing an appendectomy.

Yesterday, I twice asked Ms. Kliff to please correct this grossly misleading paragraph. She acknowledged my request that evening, but as of 9 AM today, nothing had been changed.

Even if Ms. Kliff had correctly identified the hospitals as the culprits, using appendectomy as an example of why patients should shop for the lowest prices was a poor choice.

Nearly every patient with appendicitis does not know he has it until he has gone to an emergency room, seen an ED physician, and had some tests. I doubt most people in this situation would A) ask how much it’s going to cost to have an appendectomy and B) decide to go to another hospital for care. The fact is, hospitals are so secretive about their charges that a patient would be unable to comparison shop especially if the emergency department visit occurred outside of normal working hours.

Even trying to find out the charges for elective surgery remains difficult in 2015.

Physicians—particularly surgeons—have taken a lot of heat recently. We don’t need articles like this to inflame patients (and journalists) even more than they already are.

ADDENDUM 9:45 AM 10/15/15

The article was just changed. The bolded mistaken passage was corrected, but the next sentence (underlined in red) remains the same. Still blaming those "really expensive doctors."

Duke Docs Who Attached Intestine to Vagina Must Stand Trial

Catchy headline, isn't it?

That headline appeared two years ago on the Outpatient Surgery website. Too bad it wasn't accurate.

This case has been a topic on the Internet off and on for a few years. Although the patient's surgery took place in 2008, the malpractice trial did not occur until earlier this month.

Like nearly every news media article about malpractice incidents, details were sketchy and sensationalism was featured.

Also common in cases like this that although many stories appeared about the case before the trial, but few reported the verdict. That is because after deliberations lasting less than one day, the doctors were not found guilty of negligence.

According to a story on the only media outlet reporting the verdict—Courtroom View Network, here's what happened.

The patient underwent surgery for debilitating chronic constipation at Duke University Medical Center in 2008. Postoperatively, she developed a recto-vaginal fistula [a tract or tunnel from the rectum to the vagina] due to a portion of vaginal wall being caught in the surgical staple line when the intestine was reconnected. The problem was corrected by a second operative procedure.

Prior to the first case, the patient had been informed that secondary procedures might be necessary if complications arose.

The malpractice suit was originally denied by a lower court because the plaintiff could not find an expert witness to testify that negligence had occurred, but an appeals court ruled that an expert was not necessary because "even a layperson would be able to determine whether or not negligence occurred," and "It is common knowledge and experience that intestines are meant to connect with the anus, not the vagina, even following a surgical procedure to correct a bowel problem."

Although the second operation solved the patient's problem, she claimed that she had suffered a conversion disorder causing slurred speech, tremors, and weakness. However, subsequent treating physicians felt that her symptoms were inconsistent and that she was "trying to appear disabled."

This case illustrates several important principles about medical malpractice cases.

They often take a long time—an average of 4 to 5 years—to be resolved.

Every less-than-perfect outcome is not necessarily due to negligence. Recto-vaginal fistula is a known complication of this type of surgery.

Stories from 2013, when the appeals court said the case could go to trial and from this year just before the trial started, implied that the surgeons had mistakenly attached the intestine to the vagina which many of us found hard to believe. But without knowing the details, we could only speculate.

The jury did not believe the alleged damage—a conversion reaction—was real.

This lawsuit, which cost both the plaintiff's attorney and the defendants' insurance company a lot of money and dragged two highly competent and respected surgeons through the mud for 7 years, should never have gotten off the ground.

There was a reason that plaintiff couldn't find an expert to testify that negligence occurred. The intestine was never mistakenly "attached to the vagina."

More germs: Planes, desks, and even kisses

Airplanes are so permeated with bacteria that it is truly a wonder that anyone survives a flight. I'll bet you thought it was the air in the cabin, but a recent story in USA Today says otherwise. "The real problems lie on the chair upholstery, the tray table, the armrests and the toilet handle."

What should germophobic passengers do? "First, they should travel with and use an alcohol-based hand sanitizer. They should also travel with a small pack of disinfectant wipes," said the microbiologist who did the study. "The first thing I do when I sit down is to wipe down the armrest and tray table because that's where my arms will be. You need to decontaminate where you'll be spending your time and eating."

From MailOnline: "Millions of us spend our days slaving over a keyboard. But lurking between the keys, hidden on the mouse and nestled in your phone lies [sic] more than 10 million bacteria—400 times more than on the average toilet seat." [Despite what I reported in my last post, the toilet seat will remain the gold standard for comparing contamination levels until it is unseated.]

At least that is way fewer bacteria than the next study found.

"Every Kiss Begins With 80 Million Germs" headlines WebMD about a paper published in the journal Microbiome. WebMD story says, "In one experiment, the researchers gave 21 couples a probiotic drink containing bacteria before they kissed. Swab samples afterwards showed the transfer of those 80 million germs."

From the methods section of the paper: "One of the partners was invited to consume 50 ml of a probiotic yogurt drink containing L. rhamnosus GG, L. acidophilus LA5, and B. lactis BB12 [all non-pathogens]. After 10 seconds, saliva and tongue swabs were collected from this partner (donator) and after a second intimate ["full tongue contact and saliva exchange"] kiss of 10 seconds, saliva and tongue swabs were directly collected from the other partner (receiver)."

Saliva has some antibacterial properties. Maybe the researchers didn't wait long enough to test for bacteria after one of the partners drank the yogurt? People who have 80 million pathogenic bacteria in their mouths probably have bad breath and poor oral hygiene. I doubt they are indulging in 10 second tongue kissing.

And here's an excerpt from the conclusion. "This study indicates that a shared salivary microbiota requires a frequent and recent bacterial exchange and is most pronounced in couples with relatively high intimate kiss frequencies of at least nine intimate kisses per day [my emphasis] or in couples sampled no longer than 1.5 hours after the latest kiss."

Ten second kisses? Nine intimate kisses per day? Other than perhaps high school kids, who is kissing 9 times a day for 10 seconds at a time?

Bottom line? After eating yogurt, wait at least 90 minutes before tongue kissing someone.

That electric hand dryer study was bogus: Here's why

Just about everyone I follow on Twitter commented and/or linked to an article about a study claiming that electric hand dryers spew bacteria all over people using restrooms.

The paper, which appeared online in The Journal of Hospital Infection, said that airborne germ counts near jet air dryers were 27 times higher than counts near paper towel dispensers, and counts near warm air dryers were 4.5 times higher. The authors also coated subjects hands with black paint and measured spatter patterns on surrounding walls and persons dressed in disposable coveralls. And a photo from the study shows the dispersal pattern from a warm air dryer.

So case closed—paper towels are better, right?

I'm not so sure. Instead of reading an article about the paper or just the abstract, I obtained a copy of the whole paper. I also found some comments about it from a spokesperson for a hand dryer manufacturer.

What are the flaws in the study?

Can Google Glass make you a better surgeon?

Advocates of Google Glass in surgery are apparently desperate to find some use for the device.

An article headlined "Google Glass makes doctors better surgeons, Stanford study shows" concluded that the study offered "compelling preliminary evidence that the head-mounted display can be used in a clinical setting to enhance situational awareness and patient safety."

Using an app capable of displaying vital signs on Google Glass in real time, 7 surgical residents recognized critical desaturation in simulated patients having procedures under conscious sedation 8.8 seconds faster than a control group of 7 residents relying on standard monitors. Glass-wearing residents also became aware of hypotension 10.5 seconds before the control group.

Not mentioned in the article but present in a linked abstract of the paper not yet submitted for peer review was this pearl—neither difference was statistically significant.

This evidence is not that convincing. Even if the difference had been statistically significant, it is surely not clinically important.

How seeing vital signs on Google Glass is better than relying on the simple alarms that are built in to every monitor is not clear. Either way, you must stop the operation and look up to see the vital signs.

In a brief video accompanying the article, a surgeon can be seen rather clumsily activating and resetting the app on his Google Glass. The time required to perform these maneuvers apparently was not discussed.

The article, probably written directly from a press release, took a comedic turn with this sentence, "One test demanded that the resident perform a bronchoscopy, in which the surgeon makes an incision in the patient’s throat to access a blocked airway." But bronchoscopy does not involve making an incision in the throat or anywhere else.

If you would like to hear a different side of the Google Glass story, check out this video review from GeekBeatTV entitled "Google Glass is the worst product of all time." You can forward to the 3:45 mark to get past the woes of wearing prescription glasses with Google Glass and hear about the poor battery life, the balky commands, the system crashes, and more.

Should healthcare workers stop shaking hands and "fist bump" instead?

Some well-intentioned researchers from West Virginia University published a small study claiming that substituting a fist bump for a handshake might reduce the transmission of bacteria.

Since many illnesses can be transmitted by contaminated hands, the idea is plausible, but it's a good example of the media misinterpreting a study and misleading naïve readers..

They measured the surface area of open hands and fists in 10 subjects. Not surprisingly surface area of an open hand was significantly greater than that of the fist—30.206 vs. 7.867 sq. in. respectively, p < 0.00001.

They also measured the contact time of handshakes and fist bumps. The handshake took 2.7 times longer than that of the fist bump (0.75 vs 0.28 seconds). [No statistical analysis provided]

Should you wear a hazmat suit when you operate an elevator?

In their quest to culture every object on planet Earth, researchers have found that hospital elevator buttons are more contaminated with bacteria than restroom surfaces.

Of 120 randomly cultured elevator buttons, 73 (61%) grew bacteria. Washroom surfaces were cultured 96 times with 41 (43%) showing microbial growth.

As is customary with papers like this, the media sensationalized the findings with headlines like "Why you should never ever touch that hospital elevator button."

Most stories eventually mentioned the fact that the authors said the majority of bacteria found ”had low pathogenicity," but some, including Vox.com, mrsaidblog.com, newsok.com, and mynews13.com, did not. In fact, the MRSAID blog also confused the benign streptococcus found in this paper with the pathogen that causes strep throat.

I've written several posts about the culturing of various inanimate objects and pointed out that disease transmission has not been documented for almost all surfaces on which bacteria are found. If you click on the "Infection" label to your right, you can read as many as you wish.

Like most papers in this genre, this one has some flaws. You can read the full text here.

It was published in an open-access journal called "Open Medicine." It is not among the top 40 internal medicine journals listed by impact factor. It has no discernible impact factor at all and is not listed PubMed.

The title of the paper "Elevator buttons as unrecognized sources of bacterial colonization in hospitals" overstates the case a bit.

This table shows the types of bacteria found on both the elevator buttons and surfaces in the restrooms.

You can see that pathogens were few. Multiple organisms were found in several instances accounting for some of the numerical discrepancies. I could not figure out how the percentages were calculated. In neither column did the percentages add up to 100%.

Samples of restrooms surfaces were taken "a few months" after the elevator buttons were cultured. The authors conceded that this may have confounded the results.

Washroom sample swabs were taken from the exterior and interior entry-door handles, the privacy latch, and the toilet flusher. They did not swab toilet seats, which previous studies have shown are the gold standard to which all inanimate surfaces should be compared.

It may surprise you to learn that most of the bacteria found on elevator buttons and restroom surfaces can also normally be found ON YOUR HANDS!

Are sutures better than staples for closing skin for C-sections?

According to a recent randomized trial of 746 women having cesarean sections, closing skin incisions with staples resulted in a 10.6% incidence of wound complications, compared to 4.9% of the women closed with sutures. Put another way, suturing was associated with a 57% reduced risk of wound complications.

As usual, naive media, including The New York Times, accepted the study's findings without reservations. Also as usual, I did not.

Here's why.

Check the Y-axis when reading a chart

Here is an interesting way to make statistics more persuasive without really lying.

Take a look at this chart. It was labeled "Hospital readmissions sharply declined."

Now look at this one. The same data are charted, but the decline does not look nearly as sharp.

A common trick is to abbreviate the Y-axis of a chart. Proponents of this will tell you it makes the chart more compact and easier to read. However, the downside is that a small change is made to appear much larger.

Even though the change was said to be statistically significant in this instance, the casual reader would certainly be impressed much more by the sharp decline depicted in the first chart.

I believe the first chart was produced by the Centers for Medicare and Medicaid Services (CMS) to show that its policy on readmissions was working. I was unable to find the original source for it.

This also works with other graphics as shown in this pair of bar charts from the Visualizing Data blog.

I encourage you to look for this type of manipulation when you read research papers. pharmaceutical ads, or any other depiction of data.