Car dating & cognitive dissonance at Grand Rounds

Here's a link to SharpBrains, where yesterday's host Alvaro Fernandez brought together Grand Rounds (a forum for medical bloggers) and Encephalon (a forum for people who blog about the brain and mind). Alvaro offers a tongue-in-cheek, "What a nice surprise! Hello. Nice to meet you" to both groups.

The introduction has already been made.

The need to recognize the inherent fallibility of humans (and design systems that are reliable in spite of the predictable faux pas humans make) was articulated nearly a decade ago in the first IOM report, To Err is Human. Alvaro's invitation, his need to suggest that healthcare professionals dip into the cognitive psychology well, is telling. It's surely part of the reason we've yet to post measurable gains in preventing inadvertent medical error.

It occurs to me that when introductions lead to a relationship, it's because both parties perceive a benefit. It's been ten years, and in the U.S, we're still discussing whether tired residents are really as tired as other tired people. And entertaining other intention-oriented ideas, like "Follow the 5 Rights." This suggests cognitive dissonance between the safety paradigm we have and the one we need. Apparently, "we're just not into you," SharpBrains.

Healthcare remains distinguished from other high consequence industries by the degree of personal vigilance we tolerate and rely on. No matter where you or your organization may be on the journey toward improving patient safety, you should agree to a second date with the folks who study the performance parameters of humans.

Applying lessons learned to healthcare workers and the systems used to deliver care is a necessary step in eradicating the public health problem called "medical error."

Where's the engineer of this train wreck?

Every year in the U.S., 1.5 million people are harmed by medication errors, events that add billions of dollars to the healthcare economy. The medication use system we rely on is a lot like a train sent down the tracks without an engineer. Not only was it not "engineered" in the first place, few people on the train understand everything necessary to ensure its safe operation. Worse, most people on board wrongly assume that someone is, indeed, in charge and sit back, assuming they are free to relax and enjoy the ride.

Maximizing the safety of medication use is no short journey. You start with a complex system that involves licensed individuals, crosses the disciplines of medicine, pharmacy, and nursing, is highly regulated, has deep-pocket special interest groups, and requires a high degree of cooperation and communication amongst professionals and consumers. Oh, and to cement your joy, if you include the word "medication" in a sentence, you've just catapulted the instructions out of the range of the fifth grade reading level that health materials are supposed to be written at. ("Medicine," yes. "Medication," no.)

Medications have to be prescribed (or for OTCs, selected by consumers) dispensed, administered, and, in some way, shape, or form, monitored for effectiveness. For warfarin, the rat poison that mitigates inappropriate blood clotting, monitoring involves serial tests run on blood samples. At the other end of the spectrum are birth control pills, where monthly periods suffice. (Sadly, variability is not the friend of reliability.)

Systems engineers speak of "failure points," predictable places where errors are likely to occur. For example, not knowing that a patient takes warfarin when he or she presents for treatment of another condition is a known failure point that can be predicted to result in medication-related harm.

"Failure finding tasks," "performance shaping factors," and "exposure rates," are engineering concepts that rarely make their way into curricula used to prepare healthcare professionals. Yet when it comes to getting desired health outcomes what healthcare professionals know about how a system works may be as important as knowing how a person's system works.

Systems engineers also learn to design work processes to achieve three distinct safety-sensitive outcomes:

1. prevent errors (operative words: barriers, constraints)


2. discover errors set in motion before they cause harm (operative words: redundancy, "failure finding" tasks)


3. mitigate the potential for errors to cause harm (operative words/concepts: recovery, rescue)

There's a host of things that can be done to improve safety when medications are used. But the first step (and remember, this is always the first step) is to recognize that we--all of us--have a problem. If you prescribe, dispense, administer, or take medications, you have a problem: you're riding in a runaway train.

The good news is that the language of systems engineering can be learned, and concepts adopted and adapted to retrofit the medication use system. And you have a role, no matter where you're coming from. I'm going to begin an occasional series, describing specific actions and activities that can be used to strengthen the system. You'll find these posts are indexed with the label "engineering." All aboard? I hope you'll come back soon!

STLs happen

Acting on an urge to find an outdoor spot to enjoy a warm spring evening, my husband and I recently found ourselves perched on a deck three stories above an old tavern, giving us a bird's eye view of the old courthouse in the small southern city where Larry Flynt, the publisher of Hustler magazine, was gunned down in 1978. I didn't know this fact, although I've lived in nearby communities for over two decades, but my husband did. This Trivial Pursuit-worthy fact, coupled with an old red pick-up truck that was missing the entire driver's side door (but, happily, not the driver) occupied our conversation until I remembered we were right down the street from the new courthouse (where I had been a jury member when a fellow citizen sued the Troy-Bilt company following a chipper-shredder mishap in the mid-1990's). We were destined for good conversation.

• Targeting, then deliberately shooting, a fellow citizen with a handgun
• Driving a motor vehicle absent a key piece of personal protective equipment
• Getting your hand stuck in the business end of lawn and garden equipment

These are three very distinct ways to screw up. (Normally, I'd use the term err--as in To Err is Human--but, hey, it was Friday night.) My husband, of course, would have preferred to continue talking about Larry Flynt, Hustler, or the guy in the truck without the door. But I was determined to talk human factors engineering, the study of how people, given our capabilities and limitations, can be predicted to perform tasks that involve using machines under real world conditions.

"Did you know," I asked, "that the probability of a well-trained, motivated, competent person producing an error while performing a routine task is 1 in 2000?"

Erring while doing something you know how to do is a mistake characterized as a slip, trip, or lapse (STL), and STLs are the most common of human errors. Bringing home Coke instead of Diet Coke or Fruit Punch Juicy Juice instead of Cherry Juicy Juice are classic STLs. Environmental factors, such as similiar packaging, product placement on the shelf, lighting in the grocery store, and distractions during the selection process play a role in STLs. Irrespective of intention, STLs happen, and they happen most often to seasoned people performing tasks and activities they are normally adept at doing. If it happens in the grocery store, it can happen at work.

The chipper-shredder mishap I heard during my jury service could be characterized as a slip, trip, or lapse. (The gentleman who brought the suit was raised on a farm and had used heavy equipment since the time he was a young child.) When an experienced, but busy or distracted, nurse attaches a syringe containing viscous liquid, meant for oral administration, to an IV line and inadvertently administers it intravenously, you've heard another account of a slip, trip, or lapse.

I still haven't gotten over the fact that I learned about slips, trips, and lapses while serving on a jury, instead of during the course of my professional training ten years earlier. How helpful it would have been to know that systems could be designed and engineered in order to compensate for mistakes competent professionals can be predicted to make. (At the risk of beginning a rant, let me point out that in the U.S., we landed a man on the moon three decades before oral syringes--devices that are incompatible with IV tubing and thus prevent competent people from having a 1-in-2000 slip that can kill a person they're trying to cure--became commercially available in hospitals.)

I write about medication safety over at Medscape, on a popular blog called On Your Meds. This forum draws tons of comments from front line clinicians. Most use it to talk about what's it like to use existing processes to deliver the meds (and the care) they want to give. A few weeks ago, On Your Meds received over 8,000 hits in the first 24 hours after a piece entitled, "Medication Misadventures" was posted.

If you're a front line clinician, I encourage you to check out the hundred or so comments in response to "Medication Misadventures," reflecting about the safety problems others perceive, how your system benchmarks with others, and how well you (and your patients) are protected from predictable slips, trips, and lapses.

If you coach, lead, or manage front line clinicians or have a role in funding the systems (equipment, software, and the support services needed to fully operationalize them), I encourage you to read the comments left in response to "Medication Misadventures." Comments are simply electronic footprints, showing clinicians' perceptions about the fitness of the medication use system. These are people (your people, perhaps) interacting with machines under real world conditions. How well is your medication use system engineered to account for human factors?

If you sell medication safety, remember to talk about human factors when you come calling. We're not shooting our fellow citizens with a hand gun. But we may be tempted to drive an old pick-up, and we regularly have to dislodge pieces of cat brier from the chipper-shredders we use to get the job done.

And if you design equipment or systems that will make the medication use system safer, please get back to work. We need you.

Off to mulch..... stay safe and come back soon!

"Here's Your Sign" is not a High-End Risk-Reduction Strategy. Go Figure.

I live in the southern part of the U.S., and, like most of my friends, I like Bill Engvall, one of the funny men on the Blue Collar Comedy tour. Larry the Cable Guy, not so much, but I really think Bill Engvall is funny. If you don't know who he is, here's one of Engvall's jokes (and a link that takes you to the lyrical version of "Here's Your Sign"):

A couple of months ago I went fishing with a buddy of mine, and as we pulled his boat into the dock, I lifted up this big 'ole stringer of bass.

This idiot on the dock goes, "Hey, y'all catch all them fish?"

"Nope. Talked 'em into giving up."

I like Bill Engvall’s take on the human condition. It helps me make it through the Walmart. But when I go to work, I try to leave Bill behind. Here’s why:

In Chipping Away at Risk, I talked about how professional standards in other industries call for the use of the highest feasible strategies to manage predictable risks, and noted that similar thought processes are not yet considered "the norm" in the healthcare culture. In healthcare, it’s easy to draw from our duty-oriented traditions, falling back on what an “A player” wants to do on a good day rather than what a “B player” produces on an average day. (We will not discuss "C players" today.) But Human Factors research tells us that different, more reliable processes are needed to manage predictable risks that arise when people, processes, and equipment converge: "Hey, y'all catch all them fish?"

So here’s a user-friendly list of risk reduction strategies, one that’s widely used by the safety analysts at the Institute for Safe Medication Practices. The strongest error-reduction strategies are listed first, with the less effective options lower on the list:

• Fail-safes & Constraints
• Forcing functions
• Automation & Computerization
• Standardization
• Redundancies
• Reminders & Checklists
• Rules & Policies
• Education & Information
• Suggestions to be more careful or vigilant

And three examples showing how these principles look “on the job”:

1. A patient care unit where the primary fall-prevention intervention involves nursing personnel “keeping a close eye on patients at-risk to fall ” is using a less reliable fall-reduction plan than a unit where nursing vigilance is augmented by standard measures (such as the opportunity to use the bathroom every two hours). This can be predicted because scheduled opportunities to use the bathroom standardizes an intervention while “keep a close eye on them” relies on personal vigilance, a much weaker risk-reduction strategy.
2. A neonatal unit that has a policy stating only 10 units/mL heparin will be stocked in the unit’s automated dispensing cabinet (ADC) has a less reliable risk-reduction plan in place than a neonatal unit where heparin products undergo bar-code scanning prior to delivery to the unit and prior to being prepared for a given patient. Bar-coding is an automated risk-reduction strategy with reliability that trumps both policy statements and the accuracy of humans when “reading the label.”
3. Port-free epidural tubing, especially those with distinguishing colors and features, makes patients safer than using standard IV tubing because the absence of a port is a constraint that can prevent inadvertent administration of parenteral drugs to the patient’s CNS, a tragic occurence that regularly happens when well-educated clinicians become distracted.

I hope the rank order of risk reduction strategies and the clinical examples give you something useful to consider about mitigating on-the-job risks and how to respond when an error occurs.

Stay safe, find some time to fish, and come back soon!

Chipping Away at Risk

I had an unusual introduction to Human Factors (HF) engineering when I served on a jury that heard a chipper-shredder mishap in the mid '90s. I'll forgo the tragic details that gave rise to a suit against the manufacturer, and leave you with just the take-away lesson: if a machine's outer casing is shielding a series of free-swinging blades seated on a spinning Ferris wheel-like device, the likelihood of your hand being sucked in from below (at the small, innocuously-appearing exit shoot) is as great as it is from above (at the larger entry hopper where you toss yard debris). This information becomes even more relevant should be you be tempted to free--even from a distance well away from the exit shoot--a thorny vine caught around the spinning Ferris wheel device.

During the course of the two week trial, HF expert witnesses gave our jury a soup-to-nuts education about chipper-shredders, attempting to get us up-to-speed about design principles, mechanical features, and professional standards manufacturers must conform to.

It did not come as a surprise to HF expert witnesses--neither those testifying for the plaintiff nor those testifying for the defense--that human beings and chipper-shredders had a high potential for yielding tragic outcomes. In fact, a manufacturer's ability to bring a high-hazard product to market, and keep it there, hinges on whether it can be made safe enough to protect people from the predictable mistakes they are likely to make while using it.

While I found the mechanical aspects of chipper-shredder design interesting (even using the information when purchasing a chipper-shredder of my own a few weeks later), I experienced a profound "ah-ha" moment when I realized how differently HF experts evaluated risk and selected risk-reduction strategies, compared to how I did. (At the time, I was an experienced intrapartum nurse and a leader on a 600 births/month maternity service.)

According to HF standards, warnings--even bold ones using pictures with high-contrast color combinations and affixed in strategic locations--are insufficient if a higher-order strategy--like installing a protective grate north of the exit shoot--is feasible. Written directions (think: policies and procedures) similarly fall low on the list. Because written directions and warnings have a high failure rate, their best use is in conjunction with risk-reducing strategies that are more likely to work.

I think my jury service occurred in 1996 or 1997, several years before the publication of To Err is Human, an IOM report that quantified healthcare errors and served as a multi-stakeholder call-to-action. The subsequent 2001 report Crossing the Quality Chasm began to describe specific improvement strategies, previewing successes and borrowing methodologies from human factors-oriented industries, like commercial aviation and nuclear power.

In the mid-'90s, the idea that healthcare workers should do more than, "review policy" and "counsel individual" was revolutionary. But today, it shouldn't be.

Two years ago, Sean Berenholtz and Peter Pronovost, physicians at Johns Hopkins University and leaders in patient safety research, commented on interventions selected to prevent reoccurence of mistakes in healthcare settings, noting, "Unfortunately, weak interventions predominate and are often the same traditional solutions offered in a new package."1

Next time, I'll share more about a rank-ordering of risk reduction strategies that's used to promote medication safety. If I've piqued your interest, you can find a short case study and critique of the use of low-level risk reduction strategies in a June 2008 Pennsylvania Patient Safety Advisory.

In the meantime, stay safe working in your yard!


1 Berenholtz, B. & Pronovost, P. (2007). Monitoring patient safety. Critical Care Clinics, 23, 659-673.