CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES

Transcription

1 CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES SEPPO E. HONKONEN, MARKKU J. JARVINEN From Tampere University Hospital, Finland We analysed 131 fractures of the tibial condyles in 130 patients, using a modification of the classification of Schatzker, McBroom and Bruce (1979). The patients were reviewed at an average of 7.6 years after the injury. Fifty-five (42%) fractures had been treated conservatively and 76 (58%) operatively. unicondylar and medially tilted bicondylar fractures tended to redisplace into varus position and lateral unicondylar and laterally tilted bicondylar fractures into valgus. There were significant differences when the results were evaluated according to the methods of Hohi and Luck (1956) and Rasmussen (1973). Using our method in conservatively freated cases, the subjective results were acceptable in 49.1%, the functional results in 60.0% and the clinical results in 52.7%. In cases treated by operation the equivalent figures were 57.9%, 73.7% and 52.6%. The poorest results followed displaced medial condylar and medially tilted bicondylar fractures. Varus alignment of the tibial plateau was tolerated worse than valgus alignment. s of the tibia! condyles are caused by a combination of vertical thrust and bending (Kennedy and Bailey 1968) and are often associated with ligament and meniscus injuries. The tilting ofthe articular surfaces and overload cause mala!ignment ofthe leg and secondary overloading, and instability of the knee. The aims of treatment are to obtain a stable, aligned, mobile and painless joint and to prevent post-traumatic osteoarthritis, but it is still controversial as to how these are best attained. Some authors have stated that anatomical restoration of the articular surfaces and stable internal fixation are essential (Rombold 1960; Burn et al 1979; Schatzker, McBroom and Bruce 1979) but others have reported good functional results after conservative treatment or after operative treatment which did not restore the anatomy perfectly (Rasmussen 1973 ; Apley 1979; Sarmiento, Kinman and Latta 1979; Lansinger et a! 1986). One explanation for these contradictory opinions is lack of a good classification. Another is that the methods ofevaluation used were not uniform. In addition, the criteria for an acceptable result may not have been sufficiently stringent for differences between the fracture and treatment groups to have appeared. Classifications of fractures of the tibial plateau are S. E. Honkonen, MD, Orthopaedic Surgeon M. J. J#{228}rvinen, MD, Chief Surgeon Division oforthopaedics, Department ofsurgery, University Hospital, Tampere, Finland British Editorial Society of Bone and Joint Surgery 030l-620X/92/6465 $2.00 J Bone Joint Surg [Br] 1992; 74-B : mostly based upon the site and direction of the fracture lines and the degree of displacement ; medial and lateral fractures are included in the same groups. Schatzker et a! (1979) have stated that it is not correct to speak of tibial plateau fractures collectively, because they differ in their cause, pattern and prognosis. They classify fractures of the lateral condyles into split fractures, split-compression and compression fractures. Because the structure of the medial condyle is stronger than that of the lateral, it is less likely to fragment and Schatzker et al (1979) therefore separate bicondylar fractures into two subgroups. One results from an axial thrust and affects the two condyles similarly, and the hallmark of the other is a fracture line that separates the metaphysis from the diaphysis. This distinction, however, may not be altogether logical. The force that causes a bicondylar fracture is seldom purely axial. If it is applied more on the media! side, the medial plateau subsides more than the lateral and a varus deformity results ; if it is more concentrated on the lateral side, the lateral condyle subsides giving a valgus knee (Fig. 1). A purely axial force causes no change in the plateau angle and no axial deformity of the leg. Hohi and Luck s (1956) functional and anatomical grading system for fractures of the tibia! condyles is widely used with some modifications. Another popular method of evaluation is Rasmussen s (1973) system in which points are given for each item and the result is the sum ofthese points. In both, the criteria for an acceptable functional result are similar. In their radiological scoring scale, Hohl and Luck allow depression of the fragments of up to 3 mm, while Rasmussen allows 5 mm. Rasmussen s system takes no account of degenerative changes. 840 THE JOURNAL OF BONE AND JOINT SURGERY

2 CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES 841 Our aim is to demonstrate the effect ofresidual varus and valgus malalignment on the final result and to discuss the various methods of evaluation. PATIENTS AND METHODS We reviewed the records of all patients admitted to the University Hospital of Tampere between 1977 and 1986 with fractures of the tibia! plateau. There were 212 fractures in 21 1 patients ; 29 patients had died for reasons Fig. la Fig. lb The s of bicondylar fracture : a, laterally tilted ; b, medially tilted; and c, axial. Table I. The classification and number of 13 1 fractures of the tibial plateau and the age (mean and range) of the patients F racture Number Per cent Age (years) 1 Lateral split fracture (21 to 74) 2 Lateral split-compression fracture (31 to 74) 3 Lateral compression fracture (47 to 78) 4 condylefracture (l6to6l) 5 Laterally tilted bicondylar fracture (15 to 72) 6 ly tilted bicondylar fracture (31 to 75) 7 Axial bicondylar fracture (24 to 74) other than the fracture and 52 were too old or ill to attend for review. Most of the cases lost to follow-up had lateral compression fractures, since the average age of these patients when they were injured was more than 60 years. There remained 1 30 patients with 1 31 fractures, 78 women with an average age of 56.1 years (16 to 78) and 52 men with an average age of42.3 years (15 to 74). The average follow-up was 7.6 years (3.3 to 1 3.4). There were 59 (45%) traffic accidents, 22 (17%) home accidents, 14 (1 1%) work-related injuries, 1 1 (8.5%) sports injuries, and 24 (18.5%) other leisure-time injuries. Fifty-one (38.9%) of the fractures were on the right side and 80 (61. 1%) on the left, a similar ratio to that reported by Dovey and Heerfordt (1971). Seventy-six (58%) fractures were treated operatively, Fig. Ic and 55 (42%) conservatively. In the latter patients no effort was made to restore the anatomy of the articular surface. The leg was simply immobilised in a cast from groin to ankle, with the knee in 200 flexion, for 7.4 weeks (4 to 16). Full weight-bearing was allowed after a mean of 9.8 weeks (6 to 16). Those treated operatively were immobilised in a cast for a mean of 6.4 weeks (0 to 12), and became weight-bearing after weeks (4 to 25). In the operated group, rupture or detachment of a meniscus was found in 38 knees (50%); eight menisci were sutured, three partially resected and 27 totally removed. Ruptures or avulsions of ligaments were found in eight knees (1 1% of the operated cases). In six cases the anterior cruciate ligament was avulsed or ruptured; in all but one ofthese it had been sutured or reinserted. The fractures were usually fixed with bolts or screws; plate fixation was used in 16 cases. Cancellous bone grafts taken from the iliac crest were also used in 52 cases (68.4%). No fixation was used in two lateral compression fractures after elevation of the articular surface and bone grafting. In the conservatively treated group, there were four major complications : one pulmonary embolism, two deep-vein thromboses, and one deep infection. In the operatively treated group, there were 1 1 complications: five deep-vein thromboses, three deep infections, two instances of skin necrosis and one peroneal nerve paresis. Secondary operations included three tibia! osteotomies and three total arthroplasties. The result in these cases was assessed before these operations were undertaken. The radiological appearances recorded were localised articular irregularities (step-ofi), changes in the level of a whole plateau, and condylar widening. Tilting was measured in the frontal plane. The subjective, functional, clinical and radiological follow-up results are presented separately. In addition, the total results are compared using the evaluation methods of Hoh! and Luck (1 956) and Rasmussen (1973). The classification, the distribution of the fractures and the average age of the patients in each group are shown in Table I. In the statistical analysis, changes between initial or postoperative and follow-up values (the extent of redisplacement) were compared using the paired t-test. The proportions were tested using the normal approximation for binomial guidelines. Subjective analysis. The grading scale of subjective findings is the most confusing part in the analysis of results. Most examiners have used pain as the only measure, but there are many other characteristics such as stiffness, swelling, limping, muscle weakness, locking, crepitation, giving-way and instability, which may trouble the patient. The frequency and importance of symptoms also vary from patient to patient (Bellamy and Buchanan 1986). In our study, each patient was asked to record the frequency and the importance of each VOL. 74-B, No. 6, NOVEMBER 1992

3 842 S. E. HONKONEN, M. J. JARVINEN symptom, the severity of which was then found by multiplying its frequency by the grade of its importance to the patient (Table II). Patients were asked to report only those symptoms which they regarded as directly due to the tibia! condylar fracture. Functional evaluation. The functional tests were modified from those designed for knee ligament injuries (Marshall, Fetto and Botero 1977; Lysholm and Gillquist 1982). Many of our patients were old and obese and had other reasons for impaired physical activity but their attention was always focused on impairment caused by the tibial condylar fracture. The final functional score was the lowest grade found in any of the five tests used (Table III). Clinical evaluation. The methods of grading the clinical examination are given in Table IV. Muscle atrophy was estimated by measuring the circumference of both thighs 1 5 cm above the joint line. The final clinical score was the lowest grade found in any of the four tests used (Table IV). Radiological evaluation. At review, radiographs of both knees were taken in a standing position. Because of the posterior slope of the tibia! plateau, exact measurement ofarticular surface depression is often impossible (Moore Table II. The scales used to assess the subjective results of fractures of the tibial plateau Frequencyof symptoms Never 1 Importance of symptomst I None 2 Slight 3 Moderate 4 Very Monthly Fortnightly Weekly Daily Extremely symptoms analysed were pain during various activities, swelling, stiffness, weakness, limping, giving-way and crepitation lto5excellent;6tol0good;l2tol5fair;l6to25poor and Harvey 1974). One method is to measure the depth from the remaining intact surface or from a line drawn as an extension of the other tibia! condyle. This method is useless in bicondylar fractures (Schatzker et a! 1979), and does not distinguish between a local impression and depression of a whole condyle. We measured tilting of the tibial plateau in the frontal plane by drawing a line between the deepest points of the weight-bearing area of the two condyles. The angle between that line and the long axis of the tibia! shaft was recorded. Local articular step-offwas measured from the remaining intact articular surface. These methods can be used both in unicondylar and in bicondylar fractures. The normal alignment of the plateau was measured on the radiograph of the uninjured knee. Condylar widening was determined by comparison with the width of the femoral condyles. Secondary degeneration after plateau fracture was recorded as narrowing ofthejoint space (Ahlb#{228}ck 1968). Subchondra! sclerosis was difficult to estimate because fracture healing itself may cause sclerosis. Osteophytes were frequently seen with a normal joint space. If there was narrowing ofthejoint space, relative to the uninjured knee, comparison was made with radiographs taken at Table Ill. Grading of the functional results of fractures ofthe tibial plateau* Walk 1 Normal 2 Slight limp 3 Severe limp or stick 4 Wheelchair Squatting I Normal 2 Impaired 3 < 90 4 Unable Duck-walk I Normal 2 A few steps 3 One step 4 Unable 1 excellent; 2 good; 3 fair; 4 poor Stair climbing I Normal 2 Impaired 3 One at a time 4 Unable Jumping I Normal 2 Impaired 3 Only with the aid of the uninjured leg 4 Unable Table IV. Grading of the clinical criteria of fractures of the tibial plateau* Table V. Grading of the radiological criteria of fractures ofthe tibial plateau* Extension lag Stability 1 None I Normal 2 10 to 5 2 Mediolateral : stable in extension, 3 6 to 10 5#{176} to 10#{176} instability in flexion 4 > l0 Anteroposterior : grade I instability, Lachman or drawer Flexion range test I > Mediolateral : 5#{176} to 10#{176} instability #{176} to 129#{176} in extension 3 90 to 109 Anteroposterior: grade II 4 < 90#{176} instability 4 Mediolateral: > 10#{176} instability Thigh atrophy (cm) extension 1 None Anteroposterior : grade III 2 > 0 to 1 instability 3 > 1 to3 4>3 *1 excellent; 2 good; 3 fair; 4 poor in Plateau tilting (degrees) Varus/valgus tilt (degrees) I None 1 None 2 ltos 2 lto5 36to10 36to10 4>10 4>10 Articular step-off(mm) Condylar widening (mm) 1 None I None 2 lto3 2 ItoS 3 4to6 3 6tolO 4>6 4>10 Degeneration (relative narrowing ofjoint space) I None 2 < 50% 3 > 50% 4 Obliterated #{149}1 excellent; 2 good; 3 fair; 4 poor THE JOURNAL OF BONE AND JOINT SURGERY

4 the time ofinjury. Table V shows the radiological criteria. The final radiological score was the lowest grade found in any of the five measurements used (Table V). RESULTS Radiological appearances. The average and the range of plateau tilt for each fracture group in the conservatively and operatively treated patients are shown in Tables VI and VII. Lateral split fractures kept their position well in both groups. In one conservatively treated case the tibia! plateau ended up in 6#{176} varus due to attrition ofthe medial tibia! condyle from secondary osteoarthritis. Lateral split-compression fractures had only a slight tendency to redisplace into valgus. In one case, during manipulation ofa coexistent femoral fracture, the lateral CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES 843 tibia! condyle was lifted up and the plateau healed in 6#{176} varus. Some displaced lateral compression fractures had a tendency to redisplace after operative reduction. In these osteoporosis of the distal femur and the proximal tibia was often present at the time ofthe accident. Despite good operative correction several fractures of the medial condyle redisplaced into varus, sometimes resulting in a deformity worse than the initial injury. The four such fractures treated conservatively were minimally 3#{176} occurred in seven of 84 (8%) laterally tilted fractures (s 1, 2, 3 and 5) and 1 2 of 28 (43%) medially tilted fractures (s 4 and 6) (p = ). Local irregularities of the articular surface did not change significantly in any fracture groups (Tables VIII and IX). Small bone step-off tended to smooth over eventually. In the whole series, there were ten fractures where the step-off deepened by more than 2 mm. In one osteoporotic lateral compression and one axial bicondylar fracture, the lateral condyle collapsed, leaving a deep and wide hole in the Table VI. The average (and range) of tibial plateau tilt in conservatively treated fractures Plateau til t (degrees) Number At time ofinjury At fohow-up I (-3.Oto +1.0) -2.1 (-6.Oto +2.0) (-4.Oto +6.0) +0.5 (-6.Oto +8.0) (-3.Oto +2.0) -0.7 (-3.Oto +3.0) (-6.Oto +2.0) -4.8 (-7.Oto -2.0) (+l.oto +8.0) +3.0 (+l.oto ) (- 8.0 to - 3.0) ( to - 3.0) (-4.Oto -2.0) -4.9 (-25.Oto +5.0) * + = lateral tilt ; - = medial tilt Table VII. The average (and range) of tibial plateau tilt in operatively treated fractures Number Plateau tilt (degrees) At time of injury Postoperatively At follow-up I (-4.Oto +5.0) -3.0 (-4.Oto -2.0) -3.3 (-4.Oto -2.0) (-2.Oto +20.0) -0.3 (-7.Oto +5.0) +0.8 (-4.Oto +5.0) (-2.Oto ) -0.4 (-2.Oto +2.0) (-3.Oto ) (-6.Oto -2.0) -3.1 (-4.Oto -3.0) -7.1 (-15.Oto -2.0) (+l.oto +8.0) -1.8 (-3.Oto +2.0) -2.8 (-9.Oto + 1.0) (- l5.oto -4.0) -4.1 (- l5.oto +4.0) -6.4 (- l5.oto +4.0) (-2.Oto - 1.0) -2.6 (-5.Oto -2.0) -3.4 (-6.Oto -2.0) * + = lateral tilt ; - = medial tilt displaced but some late media! displacement occurred in three of them. Laterally tilted bicondylar fractures showed little tendency to redisplace but when tilting occurred, it was nearly always into valgus. The one exception was an operated case that was over-corrected to varus and later tilted into 9#{176} varus. The medially tilted bicondylar fractures had a high risk of redisplacement which always resulted in varus deformity. Axial bicondylar fractures kept their position well in both treatment groups. If redisplacement occurred it was into varus except in one case. Plateau displacement during follow-up ofmore than lateral tibial articular surface. The width of the condyles did not change significantly during follow-up (Tables X and XI). In 1 1 of the cases widening exceeded 3 mm, but only one of them ( 7) widened by more than 5mm. Comparison of the evaluation methods. When comparing the final results, considerable differences were found between the evaluation systems. In the conservatively treated group, acceptable functional results, as assessed by the method of Hohi and Luck (1956), were seen in 67.3% of the cases, and by the method of Rasmussen (1973) in 85.5%. Acceptable radiological results were found in 43.7% and in 67.3%, respectively (Table XII). Both functional (p = 0.024) and radiological (p = 0.038) VOL. 74-B, No. 6, NOVEMBER 1992

5 844 S. E. HONKONEN, M. J. JARVINEN results differed significantly using these evaluation methods. With our scoring system, the results were slightly worse than with the Hoh! and Luck method. Acceptable subjective results were found in 49.1%, acceptable functional results in 60.0%, acceptable clinical results in 52.7%, and acceptable radiological results in 49. 1% of the cases (Table XIII). In the operatively treated group, an acceptable functional result was found in 65.8% using the Hoh! and Luck method and in 8!.6% with Rasmussen s method (p = 0.035). Acceptable radiological results were found in 51.4% and in 73.7% of the cases (p = 0.018), respectively (Table XII). Using our method, acceptable subjective results were found in 57.9%, acceptable functional results in 73.7%, acceptable clinical results in 52.6%, and acceptable radiological results in 53.9% (Table XIII). ly tilted fractures, whether treated opera- Table VIII. The average (and range) of measured stepoff* in conservatively treated fractures Step-off (mm) Number At time ofinjury At follow-up (0.Oto3.0) 0.8 (0.Oto4.0) (l.oto 10.0) 3.9 (l.oto 10.0) (2.Oto3.0) 1.0 (0.Oto3.0) (0.Oto4.0) 0.5 (0.Oto2.0) Lateral 6 Lateral 7 Lateral see text (0.Oto 10.0) 0.6 (0.Oto3.0) (0.OtoS.0) 0.0 (0.OtoO.0) (0.OtoS.0) 0.5 (0.0 to 2.0) 3.7 (0.Oto8.0) 0.0 (0.OtoO.0) 3.5 (0.Oto6.0) 0.3 (0.Oto3.0) 2.0 (0.Otol5.0) 0.6 (0.0 to 5.0) p value of difference Table IX. The average (and range) of measured stepoff* in operatively treated fractures Number Step-off (mm) At time ofinjury Postoperatively At follow-up p valuet I (0.Oto4.0) 0.0 (0.OtoO.0) 1.0 (0.Oto2.0) (5.0to25.0) 2.6 (0.Oto7.0) 2.6 (0.Otol.0) (2.Oto2O.0) 1.4 (0.Oto3.0) 3.3 (0.Otol4.0) (0.Oto8.0) 1.3 (0.Oto3.0) 1.6 (0.Otol0.0) Lateral (5.0to25.0) 0.2 (0.0 to 2.0) 1.6 (0.Oto5.0) 0.0 (0.0 to 0.0) 2.1 (0.Oto5.0) 0.2 (0.0 to 2.0) Lateral (0.Otol5.0) 2.2 (0.Oto 14.0) 2.4 (0.Otol.0) 0.5 (0.Oto4.0) 2.1 (0.OtoS.0) 0.9 (0.Oto4.0) Lateral (0.Oto2O.0) 0.0 (0.0 to 0.0) 2.2 (0.OtoS.0) 0.0 (0.0 to 0.0) 1.6 (0.OtoS.0) 0.2 (0.0 to 1.0) *see text tdifference between postoperative and follow-up values tively or conservatively, had significantly worse subjective (p = ), functional (p = ) and clinical (p = ) results than did laterally tilted fractures, although the radiological results in these groups were not significantly different (p = 0.67). The average medial tilt of 6 fractures was - 7.9#{176}, however, and the average lateral tilt of 5 fractures was only - 0.3#{176} (normal is about - 2#{176}), which means that other radiological factors (step-off, degeneration, etc.) equalised the radiological scores. The importance of the direction of plateau tilt is shown in Figure 2. Forty-four knees had normal alignment compared with the opposite knee ; 35 had lateral tilt and 23 had medial tilt of 1#{176} to 5#{176}. Lateral and media! tilt of 6#{176} to 10#{176} was found in 12 and seven knees respectively. In three cases the plateau tilted more than 10#{176} laterally, and in seven cases, more than 10#{176} medially. The functional and subjective results deteriorated more with increasing values of medial tilt than with tilt : : 4#{176} C.) C a Normal Fig. 2 Clinical 0 tilt I. Lateral tilt >10 Normal >10 Normal >10 Tilting of plateau (degrees) The effect of medial or lateral tilting of the tibial plateau on the functional, subjective and clinical results. The percentage of acceptable results is shown on the vertical axis and the magnitude and direction of tilt on the horizontal axis. Confidence intervals (95%) are represented by the vertical bars. THE JOURNAL OF BONE AND JOINT SURGERY

6 CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES 845 lateral tilt. In the group with tilt of only 1#{176} to 5#{176}, the results did not differ markedly (some overlapping of 95% confidence intervals ; Fig. 2), but in those with tilt of 6#{176} to 10#{176}, the difference between the result of varus and of valgus was significant (no overlapping of 95% confidence intervals, Fig. 2). If the plateau tilt was more than 10#{176} in either direction, the result was always unacceptable. The clinical results showed little difference between varus Table X. The average (and range) of widening of the condyles in the conservatively treated fractures Number Condylar widenin g(mm) At time ofinjury At follow-up I I I I. 1 (0.0 to 4.0) 2.0 (0.0 to 5.0) (0.Oto8.0) 3.6 (0.Oto8.0) (0.Otol.0) 0.3 (0.Otol.0) (0.t)to2.0) 1.3 (0.Oto3.0) (2.Oto8.0) 5.1 (2.Oto8.0) (2.Oto6.0) 4.0 (l.oto6.0) (0.Oto 10.0) 4.5 (0.Oto 12.0) 0.07 p value of difference and valgus but the subjective and the functional results showed that a varus deformity is tolerated worse. DISCUSSION s of the tibia! condyles have been classified according to the degree of fracture displacement (Bick 1941 ; Roberts 1968 ; Apley 1979), the anatomical of fracture (Palmer 1951), and its and extent (HohI and Luck 1956) but only the classification of Schatzker et al (1979) separates fractures of the medial and the lateral condyles. Our classification is similar to theirs but we have divided their 6 fractures into medially and laterally tilted bicondylar fractures. In the typical medially tilted bicondylar fracture, the main fracture line runs from the articular surface of the lateral condyle obliquely to the medial cortex of the upper tibia. The media! condyle shifts medially and distally. The lateral cortex may or may not be fractured. In 50% ofour patients there was a fibular fracture as well. In the typical laterally tilted bicondylar fracture, the medial condyle is minimally displaced and the distal part of the fracture line is nearly horizontal, making the media! fragment relatively more stable. The lateral Table XI. Average (and range) of widening of the condyles in operatively treated fractures Number Coodylar widening (mm) At time ofinjury Postoperatively At follow-up p value I (2.Otol3.0) 0.5 (0.Oto2.0) 0.5 (0.Oto2.0) (3.Oto 19.0) 1.5 (0.OtoS.0) 2.2 (0.Otol.0) (0.Oto2.0) 0.2 (0.Oto2.0) 1.1 (0.Oto5.0) (0.Oto 10.0) 1.3 (0.Oto4.0) 2.6 (0.Oto6.0) (8.0 to 14.0) 2.0 (0.0 to 6.0) 2.9 (0.0 to 6.0) (0.Oto2l.0) 1.5 (0.Oto6.0) 1.7 (0.Oto6.0) (4.0 to 19.0) 2.4 (0.0 to 8.0) 2.0 (0.0 to 6.0) 0.4 difference between postoperative and follow-up values Table XII. Acceptable and unacceptable results as assessed by the method of Hohl and Luck (1956) and of Rasmussen (1973), by number and percentage in each group Functional assessment Radiological ameasment Acceptable Unacceptable Acceptable Unacceptabk HohI and Luck Conservatively treated Operatively treated Rasmussen Conservatively treated Operatively treated VOL. 74-B, No. 6, NOVEMBER 1992

7 846 S. E. HONKONEN, M. J. JARVINEN Table XIII. Acceptable and unacceptable results in the conservatively and operatively treated patients, by number and percentage in each group Subjective Functional Clinical Radiological t)v Acceptable Unacceptable Acceptable Unacceptabk Acceptable Unacceptable Acceptable Unacceptable Conservatively treated I / / / / / Total / Operatively treated /() / Total condyle is depressed and displaced laterally. patients the fibula was also fractured, except in one case, in which the proximal tibiofibular syndesmosis was ruptured. In the axial bicondy!ar fracture, the fracture lines usually run from the intercondylar region to the medial and lateral cortices producing a T or Y configuration. As stated by Schatzker et a! (1979), a classification should give information about the mechanism and the prognosis of the injury and provide guidance for treatment; it should also be simple enough for practical use. The Schatzker classification, with our modifications, fulfils these requirements. The Rasmussen (1973) system of evaluation has the disadvantage that summing of the points may hide one unacceptable function, which alone may adversely affect the patient s quality of life. We believe that it gives too optimistic an impression of the outcome in these difficult fractures. Hohl and Luck s (1956) method asserts that for an acceptable result all the conditions must be fulfilled, but their criteria are, perhaps, too lenient. With this method acceptable results have been reported in 65% to 75% of cases (Hoh! and Luck 1956 ; Solonen 1963 ; Porter 1970 ; Dovey and Heerfordt 1971 ; Bakalim and Wilppula 1973), and with the method of Rasmussen in 85% to 95% (Rasmussen 1973 ; Scotland and Wardlaw 1981 ; Lansinger et a! 1986; Duwelius and Connolly 1988). We prefer to classify the modalities separately : the subjective result which measures the frequency and In our importance of the symptoms ; the objective functional result which is how the patient succeeds in some simple tests; the clinical result which records the range of movements, the stability of the knee and the strength of the thigh muscles; and the radiological result which depends upon the alignment of the extremity and the tibial plateau, the regularity of the articular surface and the degree of degenerative change. This method may make comparison of results more complex, but this can be overcome by the use of modern computers. The main conclusion of this study is that bicondylar fractures should be classified into three groups ; medially tilted, laterally tilted and axial. This classification reflects both the mechanism of injury and the prognosis and assists in the planning of treatment of these difficult fractures. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article. REFERENCES Ahlb#{224}ck S. Osteoarthrosis of the knee: a radiographic investigation. Acta Radiol 1968; Suppl 277 :7-72. Apley AG. s of the tibial plateau. Orthop C/in North Am 1979; 10 : Bakalim G, Wilppula E. s of the tibial condyles. Acta Orthop Scand 1973; 44: Bellamy N, Buchanan WW. A preliminary evaluation of the dimensionality and clinical importance of pain and disability in osteoarthritisofthe hip and knee. C/in Rheumato/ 1986; 5: THE JOURNAL OF BONE AND JOINT SURGERY

8 CLASSIFICATION OF FRACTURES OF THE TIBIAL CONDYLES 847 Bick EM. s of the tibial condyles. J Bone Joint Surg 1941; 23: Burn C, Bartzke G, Coldeway J, MuggIer E. s of the tibial plateau. C/in Orthop 1979; 138 : Dovey H, Heerfordt J. Tibial condyle fractures : a follow-up of 200 cases. Acta ChirScand 1971 ; 137 : Duwelius PJ, Connolly if. Closed reduction of tibial plateau fractures. ClinOrthop 1988; 230: HohI M, Luck J. s of the tibial condyle : a clinical and experimental study. J Bone Joint Surg [Am] 1956; 38-A : Kennedy JC, Bailey WH. Experimental tibial plateau fractures : studies of the mechanism and a classification. J Bone Joint Surg [Am] 1968; 50-A : Lansinger 0, Bergman B, K#{246}rner L, Andersson GBJ. Tibial condylar fractures. J Bone Joint Surg [Am] 1986; 68-A :13-9. Lysholm J, Gillquist J. Evaluation of knee ligament surgery results with special emphasis on use of a scoring scale. Am J Sports Med 1982; 10: Marshall JL, Fetto JF, Botero PM. Knee ligament injuries: a standardized evaluation method. C/in Orthop 1977: 123: Moore TM, Harvey JP Jr. Roentgenographic measurement of tibialplateau depression due to fracture. J Bone Joint Surg [Am] 1974; 56-A : Palmer I. s of the upper end of the tibia. J Bone Joint Surg [Br] 1951 ; 33-B:l60-6. Porter BB. Crush fractures ofthe lateral tibial table : factors influencing the prognosis. J Bone Joint Surg [Br] 1970; 52-B : Rasmussen PS. Tibial condylar fractures: impairment of knee joint stability as an indication for surgical treatment. J Bone Joint Surg [Am] 1973; 55-A:l33l-50. Roberts JM. s of the condyles of the tibia : an anatomical and clinical end-result study of one hundred cases. J Bone Joint Surg [Am] 1968; 50-A:1505-2l. Rombold C. Depressed fractures of the tibial plateau : treatment with rigid internal fixation and early mobilization : a preliminary report. J BoneJoint Surg[Am] 1960; 42-A : SarmientoA, Kinman PB, Latta LL. softhe proximaltibia and tibial condyles : a clinical and laboratory comparative study. C/in Orthop 1979; 145: Schatzker J, McBroom R, Bruce D. The tibial plateau fracture : the Toronto experience C/in Orthop 1979 ; 138: Scotland T, Wardlaw D. The use of cast-bracing as treatment for fractures of the tibial plateau. J Bone Joint Surg [Br] 1981 ; 63- B : Solonen KA. s of the tibial condyles. Ada Orthop Scand 1963; 63 :(Suppl) : VOL. 74-B, No. 6, NOVEMBER 1992