The trusted source for
healthcare information and
A funny thing happened to Susan Dovey, MPH, as she was conducting a study for The Robert Graham Center for Policy Studies in Family Practice and Primary Care in Washington, DC. She developed a new model for illustrating and analyzing medical errors.
Dovey, an analyst with the center, was lead investigator in a study examining medical errors in an ambulatory setting. As the study progressed, she created a model for illustrating what she calls "Toxic Cascades" and "The Patient Safety Grid." (See illustration.)
Source: The Robert Graham Center, Washington, DC.
"As the data started coming in, we became immersed in them," she says. "It soon became clear that current models . . . were not appropriate for us."
The Toxic Cascade refers to four separate categories of errors, each progressively serious:
1. trickles, e.g., misfiled records;
2. creeks, e.g., prescribing contraindicated medications;
3. rivers, e.g., undiagnosed fractures;
4. torrents, e.g., amputating the wrong leg.
Errors in the less-serious categories progress into the more serious categories in some cases, while in others they do not. The Patient Safety Grid, in turn, incorporates the Toxic Cascade model into four different settings:
2. clinicians’ offices;
3. institutional-based ambulatory care;
The purpose of the model is to help identify problem areas and areas where appropriate intervention can make a difference, Dovey explains. "Perhaps these minor things in primary care can build up to a torrent. We often lose opportunities to stop those from moving much further up the system. Take antihypertensive medications. Most of the patients taking them would not have a heart attack anyway, but some of them would have and didn’t because they were on the appropriate medication."
A "trickle," such as a filing error, is quite common, but if that trickle involves a lab result showing a breast lump to be malignant, inaction could be fatal, she says.
"Having a misfile of an important result that should have been acted upon becomes a creek. If [it is] left without being acted upon until it’s too late, it becomes a river, and if the patient dies, it’s a torrent," Dovey explains.
The Patient Safety Grid was designed to researchers to "take a real macro’ look at the whole scope of medical errors," she says. "What we currently know most about is patient deaths. If a patient dies, we say we can’t let that happen again; we explore the reasons and increase our knowledge. But if they were just harmed, we don’t know as much. In addition, what we currently know most about is what happens in health centers because that’s where most of the research has been done."