The article "Intra-Articular Fractures of the Distal End of the Radius in Young Adults," by Knirk and Jupiter, published in The Journal of Bone and Joint Surgery1 in 1986, in its day, was arguably one of the most important works on the management of intra-articular fractures of the distal end of the radius. Prior to the publication of that study, the critical factors that determined successful long-term management of intra-articular distal radial fractures in young patients had not been determined. The finding with the greatest impact on treatment algorithms was that accurate articular restoration was the most critical factor in preventing long-term arthritis in young patients with intra-articular distal radial fractures. However, twenty-three years of advancements in orthopaedic surgery and technology have exposed the methodological flaws of that study. The radiographic analysis incorrectly interpreted fracture lines, and the study failed to use intraobserver and interobserver validation in its analysis. The study also was conducted before the popularization of computerized tomography and wrist arthroscopy. Despite these shortcomings, an updated critical analysis reveals that its conclusions are still germane in today's treatment of distal radial fractures in young adults.
The study by Knirk and Jupiter1 has been cited in 330 research and scholarly articles. Although it is sometimes inaccurately referenced2, it ranks among the most cited manuscripts in the orthopaedic surgery literature. Today, its conclusions not only direct the treatment of distal radial fractures but they also continue to generate hypotheses for outcome studies. Given the impact of the study on the standard of care and orthopaedic research, it is appropriate to critically review its methodology to determine whether the findings remain relevant today.
Intra-articular fractures of the distal end of the radius in the young adult are a distinct group of fractures that can lead to accelerated posttraumatic arthritis. Prior to the publication of the study in 1986, there was an incomplete understanding of the radiographic parameters that were most predictive in determining whether a patient would develop posttraumatic arthritis. Distal radial fractures were managed with the primary goal of extra-articular reduction, with restoration of radial length and maintenance of volar tilt3-7. Knirk and Jupiter conducted the study after observing many young patients who had persistent pain and early advanced degenerative arthritis, despite having distal radial fractures that healed with restored radial length and volar tilt. They recognized that the same group of patients sustained distal radial fractures with radiocarpal incongruity. The primary goal of the study was to determine which radiographic parameters—radiocarpal incongruity, radial length, or volar tilt—were the most critical in the management of distal radial fractures and in predicting the development of posttraumatic arthritis.
The study by Knirk and Jupiter was a retrospective investigation that reviewed the long-term outcomes (mean follow-up of 6.7 years) of forty-three complex intra-articular fractures in forty patients with a mean age of 27.6 years who were managed with various treatment methods. The radiographic parameters of volar tilt, radial length, and radiocarpal incongruity were quantified on the initial injury radiographs with the use of traditional methods. Patient outcomes were subsequently assessed by comparing subjective and objective clinical variables with an arthritis grading scale that measured the presence and extent of posttraumatic arthritis on follow-up radiographs. Radiographic evidence of arthritis developed in 91% of the twenty-four fractures that healed with an articular step-off of =2 mm. In contrast, only two of nineteen fractures that healed with a congruous articular surface eventually had arthritis develop. The study concluded that accurate intra-articular restoration is the most critical factor in achieving a successful clinical result, but restoration and maintenance (extra-articular reduction) of the volar tilt and radial length did not prove to be critical factors.
Notwithstanding the conclusive findings of the study, a comprehensive review of its methodology with use of today's more rigorous standards shows that they are scientifically flawed (Table I). One conspicuous weakness of the study is that it was uncontrolled and lacked any assessment of interobserver or intraobserver agreement. The radiographic parameters of radial length, volar tilt, and articular step-off used to assess outcomes were analyzed by a single reviewer without any controls or way of eliminating chance or other confounding variables. Retrospective studies of radiographs are properly analyzed by multiple reviewers, with a clear method for determining the agreement, or disagreement, between reviewers. Accordingly, the specificity of the radiographic parameters in the study should have been tested by determining the kappa value—a chance-corrected measure of agreement between pairs of observers8. A kappa value nearing 1.0 would have provided more rigorous scientific proof that articular incongruity is the most important factor in determining improved long-term outcomes after the treatment of distal radial fractures. Unfortunately, because there was a single reviewer and no kappa value was determined, it is difficult to argue that the study's conclusions are due to clinical importance rather than to chance alone.
The most important finding of the study was that radiographic evidence of arthritis developed in 91% of patients in whom the distal radial fracture had a residual articular displacement of =2 mm. The methodology used to analyze the radiographs was flawed, however; it could not have accurately determined the degree of intra-articular incongruity or arthritis because it inferred radiographic measurements made from posteroanterior and lateral radiographs with the forearm in 0° of rotation. It has since become well recognized that radiographic measurements of the distal end of the radius that use standard posteroanterior and lateral projections can lead to incorrect interpretation of fracture lines9. A standard posteroanterior radiograph of a normal, volarly tilted articular surface demonstrates a proximal relationship of the volar rim relative to the dorsal rim. In fractures with dorsal angulation visualized in the posteroanterior radiograph, the parallel orientation of the radiographic beam reverses the relationship of the volar and dorsal articular surfaces. Radiographically, the anatomic relationship of fracture fragments and the degree of articular displacement can become distorted and difficult to interpret. Although the lateral radiograph may assist with correlation, it can also lead to misinterpretation. The standard lateral radiograph compared with the 23° tilt lateral projection offers a less reliable assessment of the fracture lines10. The 23° tilt lateral projection positions the articular surface in direct visualization, allowing a more direct profile view of the articular offset in the sagittal view and the apical ridges of the volar and the dorsal rim on the lateral radiograph.
The limitations of the standard lateral and posteroanterior radiographs are particularly germane to the reliability of the findings reported by Knirk and Jupiter since they were analyzed to determine both the degree of articular step-off as well as the degree of articular arthritis. Hence, the reliability of both subjective grading scales used in the study's methodology is questionable (Table II). In particular, one could essentially argue that it is not possible to accurately determine the degree of difference between a 1-mm step-off and a 3-mm step-off with use of standard radiographs. Such variability brings into question the study's finding that 2 mm is the critical amount of intra-articular incongruity that leads to a high rate of posttraumatic arthritis.
Another important study finding was that restoration and maintenance of the volar tilt and radial length did not prove to be critical factors in achieving a successful clinical result. However, variation caused by dorsal or volar angulation also contributes to inaccurate measurements of the radiocarpal interval, radial height, and volar tilt when measured on standard anteroposterior and lateral radiographs. Such variability occurs because these parameters are measured from a reference point at the ulnar border of the radius, which can change anatomically with angular deformity. As Medoff emphasized9, radiographic parameters of the distal end of the radius are more accurately defined when measured from a central reference point that lies halfway between the volar and dorsal corners as visualized on the posteroanterior radiograph. Measurements of radial inclination, radial height, and ulnar variance can change substantially when applying this alternative reference point. Although the study determined that restoration of these measurements had no effect on the radiographic development of arthritis, the validity of the outcome is questionable as the parameters were not measured properly. Because the interpretation of these very parameters was the foundation for the findings in the study, it is difficult to interpret their true relevance in the management of distal radial fractures.
There are a number of other radiographic parameters that the study did not consider and that help to describe the intra-articular injury pattern and may be as important as the degree of articular step-off in determining patient outcomes. Radiographic parameters such as the teardrop angle and anteroposterior distance are newly characterized, and their importance is undefined9. The teardrop angle is an important radiographic measurement because it can herald the presence of residual dorsal deformity and articular incongruity when other parameters may erroneously represent the fracture to be stable and adequately aligned (Fig. 1). In impacted fractures in which there is dissociation of the volar and dorsal surfaces, the volar facet can rotate dorsally into metaphyseal defects, resulting in a severe articular deformity between the dorsal-volar articular surfaces. In such fractures, reduction maneuvers may restore volar tilt and radial length, but substantial decreases in the teardrop angle may remain. Attention to and measurement of the teardrop angle can accurately demonstrate the unstable nature of the fracture, while limited measurement of radial length and volar tilt may ignore such instability.
Similar problems can occur when evaluating another important radiographic parameter on the lateral radiograph, the anteroposterior distance. The anteroposterior distance is the distance between the apices of the dorsal and the volar rim of the lunate facet (Fig. 2). Impaction injuries can cause these apices to splay away from each other, thus increasing the anteroposterior distance (Fig. 3)9. Measurement of the anteroposterior distance can detect resultant incongruity across the sigmoid notch, where measurements of radial length and volar tilt may not accurately represent the instability of the fracture pattern. Because the study by Knirk and Jupiter did not include evaluation of the teardrop angle or the anteroposterior distance, an updated understanding of the correlation between its radiographic findings and its clinical correlation is incomplete.
Another important limitation of the study is that it predates the use of advanced imaging modalities. Although conventional tomography was available at the time, it was not frequently used in the acute evaluation of distal radial fractures11. Subsequent advancements in computerized tomography improved the understanding of and promoted new studies on anatomic relationships in intra-articular distal radial fractures. Those studies showed that standard radiographs fail to depict important injury characteristics in the distal end of the radius12,13 and that computerized tomography better defines fracture patterns14-16. Pruitt et al. demonstrated that, compared with computerized tomography, radiographs underestimate the extent of comminution and degree of articular step-off in the majority of fractures involving the lunate fossa and in a minority of fractures involving the scaphoid fossa13. Two-dimensional computerized tomography is better equipped to characterize fracture fragment diastasis and depression, especially in the center of the scaphoid and lunate fossae, locations usually not visualized well by radiographs. Overall, radiographs are less sensitive in identifying fractures with intra-articular extension—this discrepancy occurs most often with fractures involving the distal radioulnar joint13. Cole et al. showed that computerized tomography was more reliable than plain radiography for quantifying articular incongruities16. Because the majority of the fractures in the study by Knirk and Jupiter were intra-articular fractures with both distal radioulnar and radiocarpal involvement, the radiographic analysis of these fractures failed to depict important characteristics that may have affected the study's outcomes. This shortcoming is important because further studies have demonstrated that the additional information offered from two-dimensional computerized tomography ultimately can influence treatment regimens15.
The use of three-dimensional computerized tomography has further exposed the weaknesses of the study by Knirk and Jupiter. Inferring three-dimensional objects from two-dimensional measurements is inherently limited because it fails to show the variability in the position of the wrist during radiographic imaging, the quality and enhancement of the radiographic image, and the accuracy of the individual making the measurements. Three-dimensional computerized tomography scans, on the other hand, offer reliable, reproducible views of the articular surface and better definition of the degree of displacement and number of fragments compared with radiographs17. In addition, the three-dimensional imaging modality also offers enhanced understanding of osseous detail, thereby providing more accurate quantification of anteroposterior and medial-lateral distances, articular angles, and areas18. Finally, the technology offers superior volumetric and linear measurement accuracy of the carpal bones and the distal radial geometric anatomic relationships19.
More recent studies have investigated whether images based on three-dimensional reconstructions improve fracture characterization or provide more clinically useful information to assist in treatment decisions than do those based on two-dimensional images20,21. Compared with radiographs and two-dimensional computerized tomography, three-dimensional computerized tomography improves both intraobserver and interobserver agreement in the analysis of the presence of intra-articular comminution. Harness et al. found that three-dimensional computerized tomography improves the intraobserver agreement regarding the presence of a central articular fragment as well as coronal plane fracture lines20. This enhanced visualization influences treatment recommendations, resulting in a significantly greater number of decisions for an open approach with a combined dorsal and volar exposure.
Just as advances in imaging have improved the analysis of fracture patterns, the popularization of wrist arthroscopy has provided new insights into the relationship between the carpus and the distal end of the radius in distal radial fractures. Since the publication of the study by Knirk and Jupiter, several investigations have demonstrated that there is a greater degree of associated carpal malalignment with distal radial fractures than previously thought22-25. In a study reviewing thirty distal radial fractures, Hanker demonstrated the presence of carpal instability when he found tears of the scapholunate ligament in 43% of the fractures, tears of the radioscapholunate ligament in 90%, and disruption of the dorsal capsule in 60%22. Similarly, in a series of sixty intra-articular distal radial fractures, Geissler et al. identified the presence of soft-tissue injuries in 68% of the fractures; 43% of the fractures had associated triangular fibrocartilage tears and 32% had associated scapholunate tears23.
Richards et al. supported these findings and showed that there was no correlation between two-dimensional radiographs and arthroscopic findings of interosseous ligament injury in distal radial fractures24. Radiographs were not sensitive in detecting scapholunate tears, as 60% of all scapholunate tears appeared to be normal (a scapholunate interval of =2 mm) on plain radiographs. Conversely, radiographs were nonspecific, as arthroscopic evidence of a scapholunate tear was not detected in half of the patients with a radiographic scapholunate interval of =3 mm.
Wrist arthroscopy ultimately has shown that associated soft-tissue injuries in distal radial fractures are the rule, not the exception. A retrospective review of the radiographs analyzed by Knirk and Jupiter demonstrates that many of the patients had substantial carpal instability patterns associated with intercarpal ligament injury, which not only could have influenced a less optimal outcome but also may have contributed to the development of arthritis (Fig. 4).
One final and important limitation of the study was that, at the time it was published, there were few, if any, validated outcome instruments in hand surgery. Because such instruments (i.e., the Musculoskeletal Function Assessment [MFA] instrument26,27; the Disabilities of the Arm, Shoulder and Hand questionnaire [DASH]28-30; the Patient-Rated Wrist Evaluation31-34; and the Michigan Hand Outcomes Questionnaire35) were developed subsequently, the study did not assess validated functional outcomes and patient satisfaction.
The importance of this exclusion to the management of distal radial fractures was exposed after Catalano et al. investigated the long-term association between the malunion of articular fracture fragments and functional outcomes36. In that study, twenty-one patients under the age of forty-five years with at least 1 mm of residual incongruity of the articular surface were followed for an average of 7.1 years for the development of osteoarthritis and for assessment of their functional outcome36. Subjective outcomes were assessed with use of the MFA, an instrument with established reliability and criterion as well as construct validity26,27. Objective outcomes were evaluated on physical examination by measuring range of motion in flexion-extension, radioulnar deviation, and pronation-supination, as well as grip, lateral key-pinch strength, and three-point pinch strength. Consistent with the findings in the study by Knirk and Jupiter, a strong association was found between the development of osteoarthritis of the radiocarpal joint and residual displacement of articular fragments at the time of osseous union. However, although osteoarthritis of the radiocarpal joint was evident in 76% of the plain radiographs and computerized tomography scans of the wrists, the functional status as determined by physical examination and on the basis of the responses on the outcomes questionnaire did not correlate with the magnitude of the residual step-off and gap displacement at the time of fracture-healing36. In fact, the authors reported an inverse relationship; the patients who reported the worst health-related quality of life and function of the hand were among those who had the least amount of incongruity of the joint at the time of fracture-healing. Ultimately, all patients examined demonstrated a good or excellent functional outcome irrespective of the radiographic evidence of osteoarthritis of the radiocarpal or the distal radioulnar joint or nonunion of the ulnar styloid process36.
While the assumed importance of the study by Knirk and Jupiter was that clinical outcomes would improve with operative reduction of articular incongruities by preventing the development of radiographic signs of osteoarthritis, the study by Catalano et al. demonstrated that these efforts did not necessarily make any difference36. While the study by Catalano et al. had its own set of limitations, the discordance between final radiographic anatomy and final clinical function brings into question the clinical relevance of the study by Knirk and Jupiter.
The same patients were later studied by Goldfarb et al., at the time of the fifteen-year follow-up, and the implications of the study remained the same—radiographic evidence of radiocarpal arthritis after intra-articular distal radial fractures worsened over time; however, despite joint space narrowing and evidence of advanced arthritis, the patients maintained a high level of function37.
The conclusions of these studies notwithstanding, Goldfarb et al. ultimately admitted that these results did not change the way they manage intra-articular distal radial fractures37. Instead, they follow the conclusions from Knirk and Jupiter by continuing "to seek an anatomic restoration of the joint surface to minimize the development of joint arthrosis."37
In conclusion, when Dr. Joseph Buckwalter accepted the presidency of the American Orthopaedic Association in 2000, he promoted the use of rigorous scientific methodology in orthopaedic surgery research to advance improvements in patient care, emphasizing that progress in medicine relies greatly on "rigorous original basic and clinical research."38 However, Dr. Buckwalter dismissed the views of some in the biomedical community that "improvements in medical practice occur only as a result of basic research," noting that both the science and the art of medical practice are important in improving patient care. While rigorous basic science research lays the foundation for progress in medicine, he noted that "clinical observations … stimulate basic scientific investigations that explain these observations… . [If one has the opportunity] to make critical observations, develop hypotheses, and evaluate the clinical value of new scientific results … [he or she] can have a direct and rapid impact on medical practice."38
Viewed through this lens, "Intra-Articular Fractures of the Distal End of the Radius in Young Adults," by Knirk and Jupiter, remains important, notwithstanding multiple flaws in its scientific methodology. It is true that the study lacks interobserver and intraobserver validation. It incorrectly interprets fracture lines. It predates newer imaging technologies. But, despite its shortcomings, the study continues to impact the orthopaedic literature and remains clinically relevant for the very reason that Dr. Buckwalter suggests: Its hypothesis was generated as a result of a critical clinical observation. Over twenty years later, the findings in the study have stood the test of time and continue to transcend the impact of advanced technology and more rigorous research to remain germane in the management of distal radial fractures36,37,39.
However, although multiple investigations have since validated the study's finding that arthritis is inevitable after intra-articular malunion, one conspicuous question has become evident: To what extent are these conclusions clinically relevant? After long-term outcome studies have demonstrated that intra-articular malunion does not necessarily correlate with a loss of clinical function and patient satisfaction, the disconnect between final anatomy and final function, unfortunately, has made the clinician's task more difficult. If anatomic restoration is destiny on the radiograph but not in the life of the patient, where, then, should surgeons draw the line in trying to achieve perfection in the one, knowing that it may not be necessary for a successful outcome in the other?
"Intra-Articular Fractures of the Distal End of the Radius in Young Adults," in its day, changed the way orthopaedic surgeons thought about intra-articular distal radial fractures. However, progress in medicine has exposed its multiple weaknesses, and today a new set of questions are no longer sufficiently answered by the study's findings. The time has come for an updated study, one that meets rigorous modern criteria, to answer these questions and to further direct the future of the management of distal radial fractures. Eventually, that study, too, will be transcended by better techniques, newer research, or additional clinical observations—and so the art of medicine, and the progress of science, will go on.