1 Spine classification systems to spine disease severity measures: a paradigm shift
The quest to create order where there previously was none can be considered one of mankind’s unique evolutionary achievements. Clearly the ability to differentiate the surrounding world is not germane to humanity but is a basic feature of the cerebrate world at large. The placement of environmental experiences into categories of threat / no threat / unclear or edible / inedible / unsure is a basic survival skill innate to most living creatures. In contrast to these basic qualities, humanity has taken the almost unparalleled step of adding societal structures governed by rules and laws to propel its breathtaking evolutionary pathway. Applying these evolutionary pathway techniques for the advancement of the field of organized medicine required the crucial development of structured observations. These skillful observations enabled clinicians to draw conclusions on the effectiveness of treatment measures that were a step beyond merely performing obvious interventions. The result was the process of classifying disease.
The act of creating disease specific classifications allows an informed person to fit an identified disorder into a larger framework formulated for that particular illness, and thus aids in communication among peers, guides management, and furthers research. If one looks at the definition of the word “classify” in a reference text within the English language, one can find two substantially different associations (see Fig 1-1).
The first obvious meaning of the word expresses that “things or people are assigned to classes or groups.” The less prevalent secondary word association implies that “information is designated as being available only to authorized people for reasons of security.” This contrast in meaning reflects the dichotomous use of classifications in the real life medical world. While classifying disease is intended to be a helpful, open-sourced process, the act of classification also creates an aura of authoritative stature through its use of a codified syntax. In its purely primary application, classifications hold great meaning for the advancement of medicine. Classification systems profoundly enhance the ability to understand and analyze medical challenges, and supports attempts at relaying observed information between individuals in an organized research fashion or even in basic conversational medical communications. Principles for an ideal classification system have been comprehensively summarized by Mirza et al for an ‘Ideal Spinal Injury System’: these include factors such as identification, comprehensiveness, inclusiveness, concise and descriptive terminology treatment guidance and ability to forecast outcomes, among many other features. Even on cursory glance such an ideal system seems unattainable even with the best of intentions [1] (Table 1-1). Nevertheless, this compilation can serve as a helpful reminder on the basic minimal elements necessary to compose a meaningful order.
Unfortunately, the secondary implication of “classifying” disease exists concurrently with varying degrees of prevalence. In the arena of modern medicine, it should seem unthinkable to intentionally codify medical disease assessments for the purposes of creating secrecy or separation, yet it seems safe to say that “codification” had been intentionally utilized by our medical profession for most of its existence. To the present date, many classification systems continue—perhaps unintentionally—to codify their data through use of ambiguous determinants and definitions. This can lead to serious misinterpretations of the comparative effectiveness of one treatment versus another. Thus, it seems only reasonable that the current generation of classification systems also feature the dimensions of transparency and clarity as leading characteristics with the goal of taking any form of secrecy out of previously “classified” medical conditions.
Closely associated with the accessibility of a classification system is the applicability of such a system across borders or medical systems. Speaking a universal language in the area of medical classifications is not just important for research purposes but also relevant to other interested parties, such as patients or insurance carriers. The review of treatment results has for the longest time been left entirely to vested members of the medical community. Dramatic developments of information systems and increasing societal expectations have done away with any shrouds of secrecy that had engulfed medical publications in the past. Results of medical studies now reach the public around the globe in a more timely fashion, and usually after a peer review process to ensure quality of the information being relayed. These significant advances raise the bar for the responsibility of medical researchers to provide hypothesis-driven studies that are adequately powered to ensure quality data and the proper analysis and generalizabilty of results. Again, classifications lay at the heart of attempts at organizing data and then interpreting results with appropriate statistical methods. Undoubtedly these developments place an additional demand on the qualities of a classification system that up to this point has not been adequately considered. The theme of universality of classification systems has received additional emphasis as a result of “globalization” surrounding us in all other aspects of life.
“Globality” identifies the ability of a classification system to transcend a single describing author’s vision and hold a consistently meaningful application to interested parties outside a particular organizational, cultural, or geographic community From its inception onwards, the rating of a medical condition should also clearly be conducted in as objective a fashion as possible, preferably showing the least possible deviation from one rater or reviewer to another. Since we are entering an increasingly result-oriented and business-like phase of medical care around the world, we as members of the medical community need to be able to express the variability of disease states with tools that exceed the dimensions of traditional classification systems [2]. For outcomes measures, we now have simple and sophisticated tools, such as the Oswestry Disability Index (ODI), the Short Form 36 (SF-36), pain scores, and a host of other measures available. It seems safe to say that the quality of “classifications” within the spine realm has not held pace with the developments in the outcomes world at all. The lack of quality of classification systems however directly affects the interpretation of outcomes studies. The authors of most common classifications for the spine almost universally fixate upon radiographic appearances, regardless of important accompanying nonskeletal factors or the general clinical condition of the host, which can profoundly affect treatment outcomes.
The inclusion of such factors is important because they relate to another vital feature of classifications, which one can circumscribe as “predictability”. Consequently, a meaningful classification system should reflect the underlying disease severity in its various categories and furthermore to some degree relate these categories to expected outcomes. The “predictability” demand clearly exceeds the dimensions of virtually any available classification system in the spine arena to date. For example, in the area of traumatology the Gustilo-Anderson open tibia fracture classification bears a direct correlation to the time to bone union, with Type III b and c injuries having much higher rates of complications and longer times to union than types I and II fractures [3, 4]. For spine research the validity of many clinical studies can be questioned without some established and validated disease categorization. For spine trauma, the management of burst fractures may serve as an illustrative example. While there are reasonably distinct definitions for the use of this term itself, important factors, such as the degree of vertebral body comminution, spinal canal compromise, and presence of a lamina fracture are not really addressed in an overarching comprehensive system. Important confounding factors, such as lack of bone quality, obesity or overall injury burden are not addressed by any spine fracture classification system. Until recently the presence of neurologic injury was not seriously addressed, despite the fact that spinal cord integrity weighs heavily on treatment decision-making and literally dominates patient outcomes. Not surprisingly there are many and widely divergent treatment options for burst fractures offered around the world. Attempts at trying to correlate published results are consistently hampered, however, by the difficulty in comparing the study populations from different settings, even though the same term “burst fracture” is used as an inclusion criterion. To no one’s particular surprise, the reported outcomes of almost any form of care provided for a healthy study population with isolated subtle “burst” injury will be much better than that of a study cohort affected by multiple health ailments and a high-grade injury. In the arena of trauma care, this ongoing disconnect of applied and implied terms continues to serve as a starting point of professional discourse on preferred management styles, but rarely leads to interventions into health care delivery by third parties, such as insurance carriers or politicians. The same, however, cannot be said for interventional care of low back pain, which has become a major point of contention between health care politicians and physicians in charge of health care in an era of increasing health care costs and diminishing resources. Here, even more methodologically sound prospective trials have raised questions largely due to lack of offering some form of objectified disease severity assessment tool as part of the baseline data set. To date, our understanding of any possible correlation between baseline Health-Related Quality of Life (HRQoL) data sets, and clinical as well as imaging test results, remains fragmented at best. In a politically charged health care environment, such as we see in the delivery for back pain, this opens the door for manipulative interpretation of scientific results.
The consequences of these developments—an increasingly sophisticated outcomes sciences movement and a stagnated array of classification models based largely on uni-dimensional imaging focused categorization philosophies, leads to an unsatisfactory imbalance rife with the potential for misinterpretation of data and misguided decisions based on these interpretations. A reader’s survey of the present date classification systems available throughout the many fields of spine would likely result in the conclusion that there is an overabundance of rather arbitrarily assembled, narrowly focused categorical systems, which have not been independently scientifically validated and rarely meet more than a few of the criteria proposed by Mirza, or the transparency and globality goals identified earlier.
In light of these observations, it seems to be a logical consequence to rethink the way in which we “classify” diseases and perhaps evolve into an era of scoring disease “severities” instead. Rather than carrying on with attempts at individually reformulating eponymous classification systems, the time has come to systematically rework the way we assess various fields of spine conditions and develop logically developed scoring systems to appropriately express patient conditions that are encountered. This process-based rethinking of classification systems has been demonstrated in an exemplary fashion with the work of the Spine Trauma Study Group starting in 2006. Using an international body of experts, this group built upon existent classification systems, but stepped beyond and has formulated a multi-dimensional disease severity scoring system for the thoracolumbar spine and later for the cervical spine. This experience can serve as a template for the development of other disease severity systems throughout the field of spinal disorders. Regardless of its eventual form, the methodology and thoroughness of the development of the spine trauma scoring systems hopefully will bring to an end the era of seemingly incessant accumulations of eponymous descriptor systems offered throughout spine and introduce the era of scientifically validated forms of structured multi-dimensional data gathering to aid in systematic disease assessment, decision-making, and sound scientific analysis. Hopefully this will be the beginning of many similar efforts. The enclosed best-effort summary of current spinal classification systems and our evaluations will hopefully serve as a starting point towards this cataclysmic change.
1 References
1. Mirza SK, Mirza AJ, Chapman JR, et al (2002) Classifications of thoracic and lumbar fractures: rationale and supporting data. J Am Acad Orthop Surg; 10:364–377. 2. Jauhar S (2008) The pitfalls of linking doctors’ pay to performance. The New York Times; New York: F5. 3. Gustilo RB, Anderson JT (1976) Prevention of infection in the treatment of one thousand and twenty-five open fractures of long bones: retrospective and prospective analyses. J Bone Joint Surg Am; 58:453–458. 4. Henley MB, Chapman JR, Agel J, et al (1998) Treatment of type II, IMA, and 1MB open fractures of the tibial shaft: a prospective comparison of unreamed interlocking intramedullary nails and half-pin external fixators. J Orthop Trauma; 12:1–7.Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
