If any one person deserves author credit for ICD-10-PCS, it is Dr. Robert Mullin. His career included cardiothoracic surgery in the Navy and at St. Raphael’s in New Haven, Connecticut among other places, as well as research in healthcare payment methodologies, beginning with his role in developing the original DRGs in the 1970s. Toward the end of his career, he spent five years working with Rich Averill and a team of coding specialists and physicians, developing the initial version of ICD-10-PCS. PCS was first released by CMS for public comment in 1998.
At the time, when he gave talks about ICD-10-PCS, Dr. Bob was often introduced as “the father of PCS.” As the years went by and we were no closer to ICD-10 implementation, he would change it to ‘grandfather,’ and the audience would laugh, ruefully, because it was a little too true.
Dr. Bob retired about seven years ago, before the ICD-10 proposed rule was published. He was my mentor and friend in the years before he retired, and whatever bedrock understanding of PCS I have, I owe to him. He was a visionary and a “mad scientist”—a term I use for some of my favorite people on the planet.
His design for PCS was radical—a taxonomy of the physical actions (aka root operations) that can be performed on a body part. You can cut out some of a body part (Excision), you can cut out all of it (Resection), or you can obliterate it (Destruction). If the body part is a tube or opening, you can widen the diameter (Dilation), narrow the diameter (Restriction), close it off altogether (Occlusion). If the body part is a tube or reservoir you can reroute the plumbing (Bypass). And so on—31 root operations in all, easily learned by anyone who uses the system.
A classification system is a model of reality. Usually classification systems use existing terms to build that model. The classification of human anatomy is that way—at some point, the names that by convention label the various parts of the body were collected, organized and more or less codified so people who want to talk to each other about an anatomical site have an efficient way to do it. But PCS builds its taxonomy with a twist. PCS uses words from the real world but provides its own definitions. Sometimes they are more precise renderings of common usage and sometimes the PCS definitions of a word are very different from common usage.
Definitions are provided for all root operations because “out there,” the same word can be used to describe very different types of procedures. Using the same word in a different context to mean different things is something language can handle just fine, but classification systems and the coded data they produce should not. Procedures that use a common word to mean different things are distinguished as separate root operations in PCS so that the resulting coded data can make meaningful distinctions for tracking surgical risk, cost, and outcome. Take the word “excision,” for example. In operative reports, physicians will call a procedure “total excision” (Resection in PCS), “partial excision” (Excision in PCS), and “excision of post-operative granuloma” (Extirpation in PCS).
PCS is a system that triangulates on root operation (e.g., Dilation), body part (e.g., coronary artery), and approach (e.g., percutaneous) to locate the procedure at a point in the space of all possible procedures. Additional information stored in the code may specify whether a device was left in the body at the end of the procedure, and in some cases a qualifier provides key information about a particular procedure (for example, that a spinal fusion was done using an anterior technique). PCS was designed to be systematic and regular, to minimize inconsistency in coding and so improve the coded data.
Dr. Bob’s architecture for PCS was equally radical—a set of tables containing the building blocks for PCS codes. Organized by body system, each body system contains the root operation tables and these tables contain the available choices of body part, approach, device, and qualifier for that root operation. This architecture is tailor made for efficient aggregation, database queries, and policies that can define a patient population with ridiculous ease. “All patients who had a laparoscopic procedure on the digestive system” can be stated in this one elegant PCS statement: 0D**4**. In English it says Medical and Surgical section, Gastrointestinal System, all root operations, all body parts, Percutaneous Endoscopic Approach, all devices, all qualifiers. Written as separate codes, the list would be 1,035 codes long.
When we were finished with PCS, according to Dr. Bob we were done—as in, no need for annual updates. He envisioned a system that was a consistent and completely whole, a classification that would not need to be ‘updated’ every year. Its vocabulary intentionally does not try to reflect current procedural terminology (or “CPT” if I wanted to make an acronym out of that). The terms were consciously chosen for their unfashionable, neutral quality as terminology. PCS is intended to have a long shelf life, and be detailed enough to produce meaningful data for the long haul.
Of course, Dr. Bob’s vision will not be realized in its ideal form. PCS has evolved since his time, for reasons both political and practical. People are in the habit of making long, cumbersome lists of codes, even though it is more meaningful to interact with PCS in its native form as a table. And people will continue to demand new codes for their revolutionary device or technique. So the system will evolve over the decades of its useful lifespan.
In Dr. Bob’s vision of the future, the very act of coding would be transformed by PCS, including new hardware. Early in the 2000’s, he described hospitals having an ATM-like screen for procedure coding—no keyboard, only the PCS menu of tables on a standalone screen. The coder could select the desired PCS table and choose a value from each column of a table to build the code. Dr Bob was describing the iPad—back when it was only a twinkle in Steve Jobs’ eye.
To do your jobs, most of you don’t need to know this bit of ancient PCS history (from one of the remaining eyewitnesses). But it is possible that you may be able to do your jobs better if you see ICD-10-PCS on its own terms instead of seeing what it’s not—not ICD-9, and not CPT. PCS is different, by design, and that after all is the point. Dr. Bob was the first to remind people, “Of course PCS is different. If it weren’t different there would be no point in switching.”
Rhonda Butler is a Senior Clinical Research Analyst with 3M Health Information Systems.