Under-Utilization of the OTA Fracture Classification in the Orthopaedic Trauma Literature

Abstract

Background

The OTA Fracture Classification is designed to provide a common language and facilitate effective communication among orthopaedic surgeons. We attempted to measure the degree to which this classification is currently being utilized in orthopaedic trauma literature.

Methods

We reviewed all of the articles in the JOT in 2011. We determined which of these articles could have appropriately utilized the 2007 OTA Classification. We calculated the percentage that mentioned and correctly cited this classification system as a reference.

Results

There were 145 articles in 2011. One hundred of these articles were appropriate for classifying a fracture. 38% of these articles utilized the OTA classification in the text. Only 42% of articles mentioning the OTA Classification cited a reference. 38% of these citations used the old (1996) OTA Classification reference, and only 8% overall correctly cited the 2007 OTA Classification reference. 51% of articles mentioned some other classification system; 21 in addition to OTA and 30 instead of the OTA classification.

Conclusions

The OTA Fracture Classification is being used more commonly (38%) but is not routinely used or correctly cited (8%) in articles currently being published in the Journal of Orthopaedic Trauma, despite the fact that it is “required” according to the instructions to authors. We conclude that future authors should utilize and correctly reference the 2007 OTA Classification so that the benefits of a common language can be realized. Routine and consistent utilization of the classification may ultimately lead to more consistency and improved interpretability of treatment outcomes in published orthopaedic trauma research.

Level of Evidence

Level-III case-control study, decision analysis

Introduction

Fracture classification systems are the means by which physicians communicate, characterize fracture patterns, make treatment decisions and determine prognoses. These systems are also useful for reporting and comparing treatment results¹. In general, fracture classification systems should be reliable and valid². The Orthopaedic Trauma Association (OTA) along with the AO Foundation developed a comprehensive fracture classification³, which has gained worldwide acceptance. Its validity has been confirmed by various studies^4,5,6,7. The inter-obser ver and intra-obser ver reliability along with accuracy of this classification system has also been verified⁶. The coding system associated with this classification provides a shorthand form and appears accurate and reliable in clinical practice. The classification is published in a readily available and electronically accessible form⁸. It has been updated ever y ten years to incorporate new knowledge of fractures and classifications. The 2007 version reconciled any differences between the AO and the OTA Classification^3,9.

Although there are numerous fracture classification systems available for specific fracture locations, the OTA Fracture and Dislocation Classification Compendium is the most comprehensive. It applies consistent fracture classification principles to the entire axial and appendicular skeleton. It has incorporated the most useful concepts of individualized location classifications. Individualized location classifications are deficient at providing a common language for fracture classification, which can limit effective communication among orthopaedic surgeons^10,11. Individualized classifications commonly have unknown or poor inter-observer reliability and poor intra-observer reproducibility^12,13.

There are at least three ways in which a fracture classification can be incorporated into an article. One is to mention the classification and stratify the data and results based upon that classification. Another is to use the terminology of the classification to provide a standard and consistent form of communication that is amenable to computerized searching. A third way is to cite a reference to the classification to the reader can easily review on established, validated classification technique. This citation must be explicitly described and readily accessible in print end electronic form so the details do not have to be reproduced in each subsequent article^14,15.

To investigate the possibility that the OTA Classification was somehow deficient, we reviewed how frequently other classifications were utilized. The reason for utilizing other classifications instead of OTA could be that it has inherent deficiencies. If other classifications were utilized in addition to or instead of the OTA Classification, we then analyzed the text to determine whether there was some deficiency in the OTA Classification that the other classification solved. This information would be of value for future revisions of the OTA classification in order to make it as clinically relevant as possible. We hypothesize that the OTA Classification is used, mentioned and cited some of the tune but not always within the orthopaedic trauma literature.

Methods

All of the articles in the; Journal of Orthopaedic Trauma in 2011 were reviewed. We determined which of these articles could have appropriately utilized the OTA Fracture (Classification by eliminating articles that aid not directly deal with fractures or dislocations. We then calculated the percentage of those articles that mentioned the OTA Classification in the title or body of the article. We noted if another classification was used in addition to or instead of the OTA Classification and, if so, made a determination as to why another classification was mentioned. We then identified whether or not the OTA Classification was cited in the references and, if so, whether the 2007 reference was accurately cited or not. We also recorded whether or not another classification that was mentioned was cited in the references.

Results

There were 10 volumes of JOT in 2011 containing 145 articles. One-hundred of t45 articles dealt with fractures or dislocations and the OTA Classification was appropriate for the subject matter of the article. These 100 “should” have mentioned that OTA Classification as it pertained to the study design or methodology However, the OTA Classification was mentioned in 38 of 100 (38%) articles. Thirty-one of 38 (82%) articles mentioned the” OTA Classification” specifically. Seven of 38 118%) articles referred to it as a generic “Fracture Classification” or in another manner¹⁶ [Table 1].

Table 1.

Mention of OTA Classification System in Text

Should Have been Mentioned	Mentioned	Mentioned as “OTA”
100/145 (69%)	38/100	31/38 (82%)

Fifty-one of 100 (51%) articles mentioned another fracture classification system; 30 of 51 (5996) instead of the OTA Classification and 21 of 51 (41%) in addition to the OTA Classification ['Table 2]. The reasons for using another classification were mentioned or deduced and reported as well [Table 3].

Table 2.

Mention of Other Classification System in Text

Total Mentioned	Mentioned Instead of OTA	Mentioned in Addition to OTA
51/100	30/100	21/100

Table 3.

Reasons for Utilizing Other Classification System

Other Classification Mentioned or Cited	Stated Deficiency in OTA Classification	Implied Traditional Classification Standard	Indeterminable
51/100	0	39/51 (76%)	12/51 (29%)

Overall, 16 of 100 (1(51%) articles cited an OTA classification. Sixteen of the 38 (42%) articles that mentioned the OTA Classification cited some version of the OTA Classification as a reference. Six of 16 o38%) article. cited that older 1996 OTA Coding and Classification reference¹⁷ even though the newer classification had been published four year previously³. Ten of 16 (63%) articles cited the 2007 reference of the OTA Fracture Clarssification. Two of these 10 (20%) articles cited the 2007 reference with some error and eight of 10 (80%) or eight out of 100 (896) articles correctly cited the 2007 OTA Fracture Classification reference [Table 4].

Table 4.

Citation of OTA Classification

Mentioned and Cited	Cited 1996 Reference	Cited 2007 Reference	Cited 2007 Reference Correctly
16/38 (42%)	6/16 (38%)	10/16 (63%)	8/16 (8/100 overall)

Forty-seven of 100 (47%) articles cited another classification system as a reference. Forty-eight of 51 (92%) that mentioned another classification cited a reference. We did not investigate the accuracy of the citation) of other classification systems. In comparison, 38 of 100 (38%) articles mentioned the OTA Classification but only 16 of 38 (42%) cited a reference [Tables 4 and 5].

Table 5.

Citation of Other Classification

No Fracture Classification Cited	Cited OTA or Other Classification	Other Classification Cited	Cited Instead of OTA	Cited in Addition to OTA	Cited only OTA
48/100	52/100	47/100	36/100	11/100	5/100

Forty-seven of 100 (47%) of articles cited another classification system as a reference. 11 of 47 cited the other classification in addition to citing OTA. Thirty-six of 47 cited another reference instead of the OTA Classification [Table 5].

Forty-eight of 100 (48%) articles did not cite either OTA or another classification system. Five of 100 (5%) articles cited only the OTA Classification. [Table 5]

Discussion

The rate of utilization and citation of the OTA Fracture Classification in the Journal of Orthopedic Trauma (JOT) was reviewed for the year 2011. This was done in order to determine the degree to which this classification is being used amongst orthopedic traumatologists while communicating scholarly work.

We found that 38% of fracture articles utilized the OTA Classification and mentioned it in the body of the manuscript. This indicates that the classification is being used, but not in all or even most of the articles. The 45 articles that we excluded were anatomical and biomechanical studies related to fractures and fracture fixation where the OTA Classification was not “required” but might have been of some value to mention and cite as a reference. We also found that only 8% of fracture articles in the JOT in 2011 accurately referenced the correct citation for the OTA Fracture Classification. The recommended citation from the JOT is:

Marsh, J. L.; Slongo, Theddy IT; Agel, Julie; Broderick, J. Scott; Creevey, William; DeCoster, Thomas A.; Prokuski, Laura; Sirkin, Michael S.; Ziran, Bruce; Henley, Brad; Audige, Laurent: Fracture and Dislocation Classification Compendium -2007: Orthopaedic Trauma Association Classification, Database and Outcomes Committee. J Orthop Trauma. 2007;21(10 Suppl):S1-133.

The reasons that the classification is or is not used are not entirely apparent. We identified the rate at which other classifications were utilized instead of (30/51 or 59%) or in addition to (21/51 or 41%) the OTA Classification. There did not seem to be many cases where the other classification contained some clinically important parameter that was not captured by the OTA Classification. There did seem to be a sense, on the part of the authors, that the “other” classification was the expected standard for reporting their results. There is certainly a fair amount of consideration given to tradition and inertia when submitting articles for publication. Authors are likely following the standards set by existing published literature as well as their own previous publication experience.

It also appears that some authors may consider the OTA Classification common knowledge and do not see the need for a particular reference; just as one uses words without referencing the dictionary definition of ever y word. This consideration may be particularly true for articles that only use the bone segment aspect of the classification that emphasizes some other aspect of the inter vention. That may, in part, explain why there is such a high rate of referencing a citation for other classifications (92%). Compare this to the 42% rate of referencing the OTA Classification even when it is already mentioned in the text.

An example is the report on regional versus general anesthesia for operative treatment of distal radius fractures¹⁸. This article used OTA Classification terminology (distal radius), perhaps coincidentally. This article did not mention or cite the OTA Classification or any other fracture classification as the emphasis was on the treatment and not the inclusion criteria. Had the authors used the terms “Colles Fracture” instead of “Distal Radius” in the title, they likely would have felt it appropriate to cite a reference to “Colles fracture”. Since the authors used the more appropriate term “distal radius”, they may not have felt the need to reference any particular classification. If the results of the previous study are to be compared to another article describing conscious sedation versus general anesthesia for treatment of “forearm” fractures¹⁹, then the importance of a precise scheme to distinguish between distal radius and radius shaft becomes more evident.

The choice of terminology such as “distal radius” and “radius and ulna shaft” that is built into the OTA Classification is of some importance²⁰. Authors may still choose “Colles fracture” instead of “distal radius fracture” or “both bone forearm fractures” instead of “radius and ulna shaft fractures”. One consequence of using non-standard terminology would be a lack of precision of the inclusion criteria. A “Colles fracture” is not specifically defined anywhere. Additionally, a computer search of all articles relating to “radius and ulna shaft” fractures will not capture all pertinent articles. For example, articles with titles including “both bone forearm, BBFA, forearm, Piedmont, fracture of necessity” or any other colloquial or eponym-derived fracture terminology may not come up when searching for “radius and ulna shaft”. This lack of consistency may lead to unawareness or under-appreciation of some potentially important articles. Furthermore, where would one go to distinguish between a “both bone forearm” fracture and “Colles fracture”? No clear distinction exists in the orthopedic literature, but, the OTA Classification clearly distinguishes “distal radius” from “radius shaft” fractures by the rule of squares.

The methods section of any article should enable subsequent researchers to reproduce the findings. If “forearm fracture” is the inclusion criteria it would be hard for a future researcher to determine whether or not to include distal radius metaphyseal fractures or not. This distinction would introduce an additional variable resulting in two studies of different, rather than the same, patient characteristics. If you want to avoid the problem of comparing apples and oranges then do not make “fruit” the inclusion criteria.

Authors of the 2011 JOT ar ticles who did cite a reference to the OTA Classification often utilized the 1996 version rather than the 2007 version. This is somewhat surprising and disappointing, but may be partially explained by the length of time a manuscript is in preparation and review prior to publication. Four years would seem to be a sufficient period of time to overcome this problem. It may also be explained by the historical tendency of authors to propagate previously cited references. This report highlights the current, correct reference for the OTA Classification and rationale for including it. It is helpful to publish this information to provide the orthopaedic community with objective data of its' current under-utilization as motivation to increase its' utilization in future publications.

The instructions to authors of JOT “require” utilization of the OTA Fracture Classification²¹. It does appear that the presence of this requirement, accompanied by increasing clinical usage, has helped to increase the rate of utilization of the classification in the literature to some degree. Authors that submit a manuscript using another classification appear to have added the OTA Classification in some articles and that appears to be a satisfactory compromise. However, the requirement is not routinely enforced or accomplished. We believe that publication of this article will help remind authors and reviewers of the Journal of Orthopaedic Trauma that this requirement exists. We hope this will, in effect, enhance future compliance and will significantly improve the quality and relevance of orthopedic trauma literature. The more frequently the OTA Fracture Classification is utilized and cited, the more the level of expectation for future submissions will grow. This resulting increase in compliance will hopefully lead to an expansion of this practice to other journals that publish the results of orthopaedic trauma research.

In conclusion, OTA Fracture Classification is being used more commonly (38%) but is still not routinely used or correctly cited (8%) in articles currently being published in the Journal of Orthopaedic Trauma. This is so despite the fact that it is “required” according to the instructions to authors. We feel that authors should utilize and correctly reference the 2007 OTA Fracture Classification so that the benefits of a common language can be realized. Routine and consistent utilization of the classification may ultimately lead to more consistency and improved interpretability of treatment outcomes in published orthopaedic trauma research.

Conflicts of Interest

Dr. DeCoster, TA - OTA Committee on Classification 2005-2011