[MUSIC] In the last lesson, we learned about ontologies and terminologies. The purpose of this lesson is to consider how standards technologies evolve and change. We will step back from healthcare standards and talk about railway transportation. At least as kids we learned a lot about trains, my kid did. Thus before we get into the details of medical terminologies, we will spend some time describing how humans form standard technology practices. At the end of the lesson, you will be able to describe the evolution of standardized railroads in United States and recognize where the evolution of railroad tracks also is applied to medical terminologies. How did the US railway standards evolve? The worldwide standard gauge for a train is the Stephenson's gauge. This was defined by the English railway engineer George Stephenson back in 1825. He standardized the width of the train tracks for all the trains he engineered to 4 feet and 8.5 inches or 1435 millimeters. This is a strange width. Thus, you might ask, why that standard? Why 4 feet and 8.5 inches of all widths? There's some debate about how this width came to be the standard or the common width. However, many historians think that the width is likely linked to the standard width of the Roman chariot axle. Of all things, why Roman chariot? It happened to be that the Romans were prolific builders and for their chariots, they had standardized the axle width to make repairing axles much simpler. Thus, they standardized the width of the chariot axle. Those axles for all the chariots and carts built thereafter had roughly the same width to them. As a result, the cobblestone and dirt roads that these chariots ran on began to develop deep ruts. In some cases, 6 to 8 or 12 inches. Or in metric terms here, 152.4 to 203.2 or about 304.8 millimeters deep, that would handle only carts of a certain width. Those Roman carts created ruts that were roughly four feet nine inches apart or 1,494 millimeters. You might wonder why it was that the Roman axle was exactly that size? Why were they all created about four feet nine inches or about 1,494 millimeters wide? The best theory is related to the width of a horse. The back side of a horse has a certain width, thus towing a cart the optimal width is roughly four and a half feet wide. Anything larger than that will be tough for horse to pull and anything narrower than that would make the cart risk for tipping over. With this, one could argue that the train track width of four feet and eight and a half inches was determined thousands of years ago by Romans trying to accommodate the invisible width of a horse. Why do we care about the width of a train track? It is important to consider track width because wider train tracks can be more stable. Moreover, wider train tracks might be more comfortable for those riding on them. But trains have to be able to fit two passengers wide on each side of the train with someone walking down the middle with relative ease. So, other than that, who cares? Well, in the later have of the 1800s during he Civil War, it was found that transportation among different railway companies have become a real problem. There was no standard gauge at the time. There was a standard gauge in England, and it started to develop across Europe. However, in the US all the different companies that had built the trains had their own proprietary train tracks and their own train widths. They could have a narrow gauge anywhere from the two feet for small holes, to over six feet in some cases. The varying width tracks became a problem during the Civil War when proprietary use of trains was commandeered for military purposes. Southern states had at least three major train tracks across which they could not transport one set of goods from one track to another. You had to convert them, pick them up and move them to the other track, this made logistics extremely difficult. The tracks were more standardized in the north but not completely standardized. A lot of the difficulty in logistics led to additional challenges for the south in being able to respond to delivering goods and services to the troops. Toward the end of the Civil War, President Lincoln standardized the United States on the 4 feet 8.5 inch Stephenson's standard gauges for all US railroads. This was both initially for the transcontinental railroad. But it followed a lot of standards in the north already. To make this a nationwide standard, the entire country stopped traveling on a Memorial Day weekend in 1886, and in just two days all the trains in the United States were adjusted to fit the standard gauge. It was a monumental effort to organize all of the railroad. Railroad workers, in just those two days, had to shift all the train tracks and all the train axles and all the train engineers to be on the standard gauge. After that, the United States had a standard. I would note that the standard railroad tie has always been about eight and a half feet or 2,590 millimeters regardless of the narrow gauge or broad gauge railroad track. So that really made possible the switch to the standard gauge rail width. So how does the railroad example help us with healthcare? Before we discuss that, let's remind ourselves that human language is metaphorical, symbolic and hence it's dependent on context. People recognize misspellings in terms from other languages, or we can take time to step back and look up the meaning of a strange term or a strange word. Computers on the other hand don't work that way. They can't interpret meaning unless specifically instructed to do so. This is a major hindrance when it comes to aggregating data for analysis. Computer scientists are doing wonderful things with coming up with fuzzy logic and interpreting unstructured data. But we don't have the luxury in clinical data analysis to use fuzzy logic when looking at an individual patient record to determine what the meaning is. What the real coded value is behind the term. Moreover, you can't build evidence-based guidelines for a single patient record. You need to build guidelines for a broad swath of patient records. The only way to do that is to apply a set of rules and guidelines across a category of patients or a category of conditions. In some, healthcare, just like the railroad, works best when data is standardized. Andrew Tanenbaum is a famous computer scientist, and he's well known for one of his quotes. The nice thing about standards is that you have so many to choose from. This paradoxical statement reflects the idea that just like the railroad, there are competing ideas about what the standard should be. As a result, there are times when standardization is slow due to conflicts about which one should be chosen. Clearly, the inventors of one standard have an interest to support their own version. As an example, consider the competing standards of the video industry, where we have VHS tapes, and the ill-fated Betamax tape. This continues today with Blu-ray disc vs HD DVDs. Blu-ray seems to have won the battle, VHS won the battle of its time. Some healthcare terminologies standard are in competition with one another, but it depends on the domain. As we'll learn in the next lessons, SNOMED has its place, ICD has its place for specific objectives, LOINC and others can work in harmony for specific types of concepts. There are some standardization of clinical terms but there's no standard that's going to be adopted for all common data representations. The hope is that as we move forward, the clinical community and the technology we use to apply these standards can advance together in terms use, usability, ease of entry and the highest data quality possible. In our next lessons, we will start to dive much deeper into the specifics of healthcare terminologies.
