Don't throw a skater under the Zamboni when you're already on thin ice

I spent a few days last week with my friend and mentor, Mike Cohen, the President of the Institute for Safe Medication Practices. Although Mike has a goodly number of titles and highly recognizable accomplishments (like being named a MacArthur fellow in 2005), you might enjoy spending time with him for the same reason I do: He's a mensch.

When Mike worked as a clinical pharmacist in a hospital pharmacy over three decades ago, he observed how frequently flaws in the system used to deliver medications set professionals up for failure. Systems engineers use terms like "single fault failures" and "opportunities for recovery" when they design work processes that reliably produce an intended outcome. But it's easier to think about these concepts using things we know from our own experiences.

Let's say we send a group of ice skaters onto a frozen pond. Some skaters in the group are Olympians, some grew up on a lake in Moose Jaw, Saskatchewan, and some are promising beginners. Everyone who skates on your pond knows how to lace up and is able to remain vertical while on the ice. They look good and on days when the ice is frozen solid, all skaters skate safely.

But your ice is not uniform. When the temperature climbs, the ice becomes thin. Dangerous spots are not marked. The skaters from Moose Jaw are better able to spot unsafe ice. They stay off or take action to avoid the risky patches. Statistically, Moose Jaw natives fall through less often than the Olympians and the beginners. But at times, even the Moose Jaw skaters go through the ice.

Variability is the enemy of people interested in engineering safety. Variable conditions make it tougher for beginners to navigate a high stakes process safely. When one of your lighter weight Olympians glides over a thin spot without falling through, beginners erroneously believe the ice is safer than it really is. Variability will trip up seasoned people, too. And when no one has gone through for awhile, it's easy to forget how important it is to mark the weak spots and have a means to pull someone out when the ice cracks.

The medication use system is built on thin ice. It was in the 1970's when Mike Cohen first began imploring healthcare providers to look beyond the characteristics of the last skater who crashed through and it remains thin today.

Here's the evidence: Since 2004, Pennsylvania has required healthcare facilities to report serious adverse, unintended medical events and near-misses to the Pennsylvania Patient Safety Reporting System (PA-PSRS). This reporting occurs irrespective of whether patient harm occurs. Last month, PA-PSRS logged its one millionth report. Over 25% of these events involve medications.

You should infer that similar data could be culled from the state where you live. Remember that, like the people in Pennsylvania, you'll have beginners, Olympians, and Moose Jaw natives on your pond. Ask what measures are in place to ensure the ice is being shored up and the thin spots marked. And you should be interested in ice conditions whether you are a patient or a professional.

In hospitals, things that shore up medication safety fall into broad categories that are objective and observable. In fact, ISMP makes a series of comprehensive medication safety self-assessment tools available for free download on their website. Using these tools, organizations can measure their progress in making medication delivery safer over time.

You should be interested in what happens to skaters who fall, too. At about the same time that Pennsylvania passed the one million mark after five years of error reporting, the citizens of Ohio sent a pharmacist to jail for an inadvertent on-the-job error, one that led to the tragic death of a toddler.

Mike Cohen has published a detailed piece An Injustice Has Been Done that explains how thin the ice in the Ohio case was. It's an impassioned plea about the dangers of punishing people whose primary fault arises from being a fallible human.

In the past five years, Pennsylvania has had 1,000,0000 events reported, each an opportunity to identify a variable that affects the quality of the ice and improve it before harm occurs. (Only 4% of the Pennsylvania reports involve patient harm.) If skaters who stumble know they will be pushed the rest of the way down, who will make a choice to share details about what undermined their performance?

And down the road, when your team can no longer recruit Olympians and Moose Jaw natives, you'll be crossing increasingly thin ice. You won't need particularly sharp blades to push the beginners under.