Feeling Silly Saturday

I have a second blog that sits on Medscape's platform. It's a forum where I share medication safety strategies with professionals, and I named it On Your Meds: Straight Talk about Medication Safety. Comments left at On Your Meds are always interesting, often enlightening, funny, and sometimes sad. I think what nurses share there says a lot about the fitness of the systems we use to deliver medications.

I joke that I should have a blog called "Get On Some Meds" or "Stay On Your Meds." Apparently, someone has been listening.

Comments at On Your Meds usually pop up in the week I post something new. So I was surprised to receive a series of notices showing new comments on an old post, entitled, hmmmm, "Not exactly the language of love: Words to identify and prevent errors."

Since error-reporting is a guiding tenet of safety engineering, I thought it might be useful to let others know what happens when you include the words "language of love" and "medication" in the most searchable elements of electronic media, which appears to be the error I made. Can you say "erectile dysfunction spam"?

Unfortunately, none of the cheap generic sildenafil and tadalafil spam I cleared the other day were as amusing as this:



Oh, and by the way: This is not a product endorsement. It's advice about managing a blog. I don't give medical advice, and you should consult a healthcare professional for any problem you might have, including an erection lasting longer than 4 hours. You can find standard safety information about tadalafil and sildenafil here and here.

Please mess with Texas

When I was a Labor & Delivery nurse, and long before I had a better conceptual framework for understanding quality in healthcare, I learned two words that summed up what families valued during their childbirth experiences: "safe" and "satisfying." While these weren't all-inclusive quality markers, they're foundational elements that signaled quality. They still do.

"Satisfying" is a concept that belongs to individuals. The way pain is relieved, how much restoration of function is enough, what a good birth, or a peaceful death look like are outcomes subject to interpretation, outcomes rightly defined by individuals. So long as they're competent to make decisions and their choices don't interfere with others, consumers should own "satisfying."

On the other hand, "safe" can't venture far in territory that's defined by the perception of individuals. Here's one reason why: Each of us is a statistical sample of one. A good outcome may arise because the people and processes effecting the experience are fit, robust, and reliable. In a word, safe. It's also plausible that one good outcome represents a lucky spin of a very damaged wheel. When comparative processes are absent, one outcome tells little about safety. This is why licensing, professional standards, regulatory oversight, benchmarking, quality metrics, error analysis, and peer-review processes are deeply rooted in industries that post excellent safety records. Measures like these unearth information from a variety of sources, revealing data not readily apparent to individual consumers.

Quality assurance processes are far less perfect in healthcare than they are in other high-stakes, high-consequence industries. In the US, over 100,000 people die each year as a result of medical error and healthcare acquired infections, adverse events that cost billions and rob citizens of peace when they are most vulnerable. In healthcare, we're struggling to align individual competency, personal accountability, and good design in a way that's more reliable, better able to deliver the predicted outcome, and, well, safer than it currently is.

Which is why a case in Texas, in which two registered nurses face felony charges for good-faith actions aimed at protecting patients, should concern you. The nurses, both employed at a county hospital, reported concerns about a physician's practice to the Texas Medical Board. Their report included numbers (but not names of patients) that medical board officials would need to identify medical records for review. As a result, both nurses:

• were charged with "misusing official information" (criminal charges arising from the act of disclosing medical record numbers to the Texas Medical Board)
• were fired from their jobs
• face $10,000 fines and up to ten years imprisonment, if convicted

Let me be clear here: A nurse's report of concern about a physician's practice does not constitute a professional sanction. Reports of concern are pieces of data, not judgments. In a healthy system, reports of concern about licensed professionals result in peer-review. In a sick system, reports of concern result in acts that punish those who raise them (and intimidate those who might consider raising similar concerns in the future).

One of the biggest challenges to healthcare safety and quality is under-appreciation and under-reporting of things that have the potential to cause harm. Healthcare is a "no harm, no foul" game, and playing by these rules means we often miss the chance to fix a known problem before actual harm occurs.

In a culture where safety is valued, people who report things that are outside of perceived norms are highly valuable players. But in Winkler County, Texas, they could be criminals.

You can read more about this case, the charges, and contribute to the nurses' legal defense fund, set up by the Texas Nurses' Association, by clicking here.

Lessons from a Sunken Ship

A thing I’ve begun to enjoy about blogging is that it helps me find memories I might otherwise have forgotten and lets me bring people (and pets) into some discussions that might have meaning for you, too. Thanks to all who have sent kind words about Daisy, our dog who is currently receiving hospice care (and recently debuted as the poster child for safe canine medication practices). It’s been a quiet, medicated morning, and I’ve been tooling around the internet and touching base with my parents as I put the finishing touches on the story about ships, Swedes, and safety that I’m sharing today. I hope you’ll enjoy it!

"Those who cannot learn from history are doomed to repeat it."
- George Santayana

On August 10, 1628, only minutes after setting sail on its maiden voyage, the mightiest warship of its time, loaded with a crew of 150, sunk in the Stockholm harbor. The Vasa had been commissioned by King Gustavus Adolphus, Sweden's monarch, who was engaged in a fight with the Poles at the time and desperate to seat a crown jewel in his armada. It’s a well-established fact that the King repeatedly tinkered with the vessel’s design while simultaneously demanding its rapid completion. But these were not the only reasons the Vasa sunk.

Like all disasters, this one had a host of contributory factors setting up the “perfect storm,” that allowed the mighty warship to sink in the Stockholm harbor on a perfectly beautiful, sunny, summer day. An abbreviated but insightful root cause analysis can be found on the official Vasa website. For people who are interested in how culture influences safety, as I am, lessons gleaned from the Vasa are particularly valuable.

It’s relatively easy to see how untested innovation, production pressures, and loss of key leadership contributed to the Vasa’s disastrous voyage. But what's really interesting to me is the Vasa’s failed stability test: In the days before the tragic voyage, the ship had undergone a preliminary test of seaworthiness using the stability testing standards of the day. This involved having a gaggle of men from the shipyard, in this case about 30, run back and forth across the ship’s deck while the ship remained moored. The Vasa’s stability test was halted after just three runs-- long before a satisfactory result was obtained--to prevent the ship from capsizing at the dock.

Nothing further was done to improve the Vasa’s stability before the ship set sail days later.

This sequence of events means that in the interval between the failed test and the maiden voyage, there were at least 30 rank-and-file shipbuilders who knew, who had to have known, that the ship was destined to sink. Do you wonder what they were saying to each other?

I think this particular piece of information captured my imagination when I toured the Vasa Museum several years ago because I know a little something about Swedish sensibilities, having been raised by a first generation Swedish-American whose family flipped back and forth between Sweden and the U.S. in the early 1900’s. Three of my grandparents emigrated from Sweden, and I was born in a small town with a large sub-population of Swedish immigrants. We’re private people, not given to share unsolicited advice (although my cousin once observed that if you sought my father’s advice, he would provide such a detailed explanation that even a novice could fix a Corvair). I joke that if my father asks, “How’s that working out for you?” you’re likely doing something that could cost you a finger.

I don’t know if my father’s sensibilities speak to the culture in the Stockholm shipyard in the 1600’s, and frankly, it probably doesn’t matter. What does matter, and still matters today, is that the Vasa sunk in part because there was no mechanism in place, no recognized, endorsed, or welcomed way, for critical information known by line managers and workers to be heard. My father will help you out, lending his considerable knowledge, time, and skills most generously, but you have to let him know you want to hear from him.

In 2007 (that's 379 years after the Vasa sunk, according to the calculator app in my iPhone), researchers studying how to best identify and respond to healthcare defects giving rise to the epidemic of adverse events that confront us today observed,

“There are many sources to identify defects, including patient safety reporting systems, morbidity and mortality conferences, sentinel events, liability claims, and perhaps most powerfully, asking staff how they think the next patient will be harmed.”1

This concept ain’t an iPhone, folks. Just ask my Dad.

Coming next: Later is better than never (more lessons from the Vasa).

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

LASA: It's Not Just Another Bad Abbreviation

I'm talking about errors associated with look-alike, sound-alike (LASA) drug names today because LASA problems offer concrete examples of risk points that dog clinicians involved in the medication use system. (For purposes of this discussion, my quick-and-dirty working definition of a risk point is "any underlying factor that predisposes to error.")

In the last post, I referred to new research confirming something you probably already know: healthcare professionals struggle with reporting mistakes, and we struggle with the fact that we are fallible when we're involved in errors. When people believe that “bad people” or “good people having a bad day” are individually responsible for most medical errors, it’s easy to see why reporting error and reconciling feelings of personal responsibility become burdensome. But reporting and reconciling become easier when you look for solutions that improve the nature of the process, not the nature of the people. Face it, we’re all going to have a bad day once in awhile, and, unfortunately, not all people are good.

(I recently spent a year as the Safe Medication Management fellow at the Institute for Safe Medication Practices. But, as you can read in my last post, I was tripped up by look-alike packaging of hand sanitizer and hand soap a few weeks back, proving yet again that “knowledge” does not trump “process.”)

Human error is typically a by-product of the systems we practice in, and with LASA errors, it’s hard to miss the risk points. The category of LASA-related errors exists because, frankly, drug names are often similar to one another. It's easy to see how words and phrases like "oxycodone and oxycontin" and "Novolog Mix 70/30 and Novolin 70/30" could be mixed up. Similarities like these regularly give rise to confusion, and yes, error. We see look-alike, sound-alike word confusion in other settings all the time: if you haven't seen "your" erroneously substituted for "you're" recently, you're reading better things than I am! But when word mix-ups have the potential to give rise to medication errors, stronger processes that guard against selecting the wrong one need to be in place.

Next time, I'll share data and some easy-to-access resources for preventing LASA errors. Maybe you have an example of a look-alike or sound-alike error to share? (If you do, tell your story in the “comments,” omitting identifying information. On Florence dot com we neither offer medical advice nor violate HIPAA regulations.)

So, good people, stay safe and come back soon!

A Picture's Worth One Thousand Words

When I made "patient safety" my business, I stepped away from specialty practice in intrapartum and high-risk antepartum nursing care, a decision that is sometimes difficult to explain. Last week, I wrote about how the perception of patient safety as a warm, fuzzy, intention-based goal can get in the way of actionable things--like workflow analysis, process mapping, and harnessing the power of technology--to deliver efficiencies, reliability, and economies of scale.

Another aspect of the career shift has been the risk of becoming a "glass half-empty" kind of girl, a perpetual naysayer who tells earnest, well-intended, and increasingly cash-strapped healthcare professionals, "Really folks, this is simply not enough. Have you forgotten that medical errors are the 8th leading cause of death in the U.S?” Last week, Oprah helped me out, hosting the Quaids and reminding us that "every year in the United States, more people die from medical mistakes than from breast cancer, AIDS and car accidents…combined. It's a major, major health issue that will touch almost every single American at one point in our lives."

I’m not a person who sees the glass is half-empty, nor am I an apologist. So I’ll share here what’s helping me to reconcile the irrefutable mismatch between intention and outcome that is healthcare today.

First, it may be helpful to simply acknowledge that errors are very common in healthcare. So common, in fact, that the Agency for Healthcare Quality and Research has endorsed a taxonomy to describe and categorize them. While this may be shocking at first glance, it’s actually good news: Using a specific nomenclature to describe events and categorize them is an epidemiologic approach to problem solving. Taxonomies are used in the study of other vexing problems (like breast cancer, AIDS, and car accidents). So, it’s reasonable to expect that similar processes would be used to diminish the incidence of our problem: medical errors.

While the charge “First, do no harm,” may resonate with many clinicians, this is a goal statement, not a process map. “Just Do It!” just doesn’t, well, do it when it comes to solving significant threats to health.

If you visit AHRQ’s Patient Safety Network, you’ll find the error taxonomy is searchable by a variety of categories (for example, “care setting”; “clinical area”; “type of error”). The one I use most often is “approach to patient safety” because this query lets me “connect the dots,” seeing how specific strategies (like “patient hand-offs”) are seated within larger motherships (like “Communication Improvement”). The taxonomy maps the current “method to the madness,” and leaves room for new ideas. (You’ll notice that the labels I apply to each post at Florence dot com often include key words from the patient safety taxonomy.)

Second, everyone makes mistakes. We may not mean to, but we do. There is a strong body of evidence suggesting that in the aftermath of an error, healthcare professionals struggle with what actions to take and how to reconcile their feelings about having been involved in an error. And a recent study in the Journal of Patient Safety suggests frontline clinicians remain conflicted about disclosing, discussing, and reporting error, despite efforts to increase transparency, promote reporting, and look at error in context. (If I were to apply a label to the discussion right now, I’d choose: culture of safety.)

It may be easier to start talking about errors that happen in healthcare settings by talking about errors that didn’t. Take a look at the photo below and see if you can guess what happened when I cooked breakfast at my church a few weeks back.

Yes, I washed my hand with a hand sanitizer product intended to be used without water, an activity that neither cleansed nor sanitized my hands.

Obviously, I didn’t read the label. A look at my kitchen sink will help you see why:

(In case you can't read the label--something that's difficult to do even in here in my kitchen--the little-bitty font just above the green leaves says, "Hand Soap.")

Two distinct products that share similar packaging, similar color, and similar placement: an error-prone set-up in the community. And an error-prone set-up at work.

This is not to say that I think the consequences of mixing up products in a community setting and the consequences of mixing up products (particularly medications or cleaning agents) while on-the-job are equivalent. In fact, it’s precisely because the risk of harm is so much greater when error occurs in a healthcare setting that processes on-the-job need to be far more robust than what we typically use at home.

I hope you’ll come back as this discussion evolves! (Feel free to use the comment section to share your thoughts with me and with each other.) And in the meantime, I hope you'll stay safe!

Next time: LASA: It’s not just another bad abbreviation.