Short Link: http://j.mp/b9iezM
[CW: "Clinical Groupware, as a phrase, has fallen out of favor. It was relentlessly marketed with any attention to what it really means. Of course, what it means is more important than what it's called. And the idea of "groupware" for clinicians is still a great idea.]
Last week I highlighted several dozen quotes about groupware, coordination, and workflow, and supplied commentary about their relevance to clinical groupware. This week I thought I’d take a stab at defining “clinical groupware” itself.
But first, what constitutes a good definition?
Southern Rose Buggy Tours
Beaufort, South Carolina
(From my recent vacation there)
While a graduate student in Intelligent Systems, I worked on lexicons for use by natural language processing systems. Lexicography is the science of creating dictionaries, which (I think obviously) includes the science of creating definitions. A good definition meets five requirements. I’ll use “horse” as an example.
- Describes essential, not incidental, attributes of the thing or concept being defined. For example, while many horses are brown, “brown” is not part of the definition of a horse. Fast horses are more valuable than slow horses, but slow horses are horses nonetheless.
- Avoids circularity. A horse cannot be defined simply as a member of the species equus.
- Is not too specific or too general. The definition of a horse should be somewhere between that of a Shetland Pony and a mammal, true of everything that is a horse and falsely applied to everything that is not a horse.
- Avoid obscurity. Use widely understood terms with clear meaning. A horse is a four-legged, solid-hooved, plant-eating domesticated mammal commonly ridden to perform work or obtain entertainment. (If you ask “Is a camel therefore a horse?” you’ve made my point. My definition is too general with respect to the third requirement on this list.)
- A definition should be positive (what a thing or concept *is*) not negative (what a thing or concept is *not*), though sometimes this cannot be avoided (as is the case if blindness is defined to be the absence of vision). A horse is not a camel, but camels should not be invoked in the definition of a horse.
If you can think of exceptions to these requirements, I am not surprised. Creating dictionaries (and lexicons for natural language parsers) is exacting *and* frustrating. Word meanings subtly change from context to context. There are exceptions. I don’t think I ever created a perfect definition. But sometimes you can create a definition good enough for a specific purpose.
Speaking of which, my purpose is to craft a definition of clinical groupware.
I don’t have to start from scratch, since early CSCW (Computer-Supported Collaborative Work) researchers have already done some work for me. In 1982 Peter and Trudy Johnson-Lenz proposed this definition for groupware:
“GROUPWARE = intentional GROUP processes and procedures to achieve specific purposes + softWARE tools designed to support and facilitate the group’s work”
“Clinical” is an adjective modifying the “groupware” noun. “Clinical” means “pertaining to observation and treatment of patients,” so how about the following definition of clinical groupware?
“Intentional care team processes and procedures pertaining to the observation and treatment of patients plus the tools designed to support and facilitate the care team’s work.”
I think this definition of clinical groupware meets all five requirements for a good definition:
- The definition describes essential, not incidental characteristics.
- It doesn’t define “clinical groupware” circularly in terms of either “clinical” or “groupware” (which is why I did not use “software tools”).
- The definition is not too general or too specific. For example, it avoids the undue specificity (in this non-pediatric context) of my use of the Johnson-Lenz definition of groupware to describe pediatric groupware in a previous post.
- It uses terms commonplace within the community of its intended audience.
- The definition is positive. For example, it does not define clinical groupware as being the opposite of traditional EMRs (whatever that would mean, since there are so many different ways to be the opposite of traditional EMRs).
This definition of clinical groupware is based on a well-reviewed definition of groupware. There are many varieties of clinical groupware, but I think the definition is flexible enough to apply to examples of clinical groupware (such as pediatric groupware) while being specific enough to be useful in discussion and communication. It focuses on what clinical groupware *is* and the goals it serves, not *how* it performs them. Being too prescriptive runs the risk of stifling innovation. I am a fan of workflow engines executing process definitions (called “workplans” in the EncounterPRO Pediatric EMR) and radar views supporting shared mental models, but there are other kinds of clinical groupware. I hope the definition can accommodate them as well.
Just in case you skipped to the last paragraph first (which I do sometimes to understand where an article is going before I decide to read it in entirety), here is my proposed definition for clinical groupware:
“Intentional care team processes and procedures pertaining to the observation and treatment of patients plus the tools designed to support and facilitate a care team’s work.”