Failure of reduction with an external fixator in the management of

Failure of reduction with an external fixator in the management of pelvic ring injuries LONG-TERM EVALUATION OF 110 PATIENTS J. Lindahl, E. Hirvensalo, O. Böstman, S. Santavirta From the Department of Orthopaedics and Traumatology, Helsinki University Central Hospital, Finland

e reviewed 110 patients with an unstable fracture of the pelvic ring who had been treated with a trapezoidal external fixator after a mean follow-up of 4.1 years. There were eight open-book (type B1, B3-1) injuries, 62 lateral compression (type B2, B3-2) and 40 rotationally and vertically unstable (type C1-C3) injuries. The rate of complications was high with loss of reduction in 57%, malunion in 58%, nonunion in 5%, infection at the pin site in 24%, loosening of the pins in 2%, injury to the lateral femoral cutaneous nerve in 2%, and pressure sores in 3%. The external fixator failed to give and maintain a proper reduction in six of the eight open-book injuries, in 20 of the 62 lateral compression injuries, and in 38 of the 40 type-C injuries. Poor functional results were usually associated with failure of reduction and an unsatisfactory radiological appearance. In type-C injuries more than 10 mm of residual vertical displacement of the injury to the posterior pelvic ring was significantly related to poor outcome. In 14 patients in this unsatisfactory group poor functional results were also affected by associated nerve injuries. In lateral compression injuries the degree of displacement of fractures of the pubic rami caused by internal rotation of the hemipelvis was an important prognostic factor. External fixation may be useful in the acute phase of resuscitation but it is of limited value in the definitive treatment of an unstable type-C injury and in type-B open-book injuries. It is usually unnecessary in minimally displaced lateral compression injuries.

W

J Bone Joint Surg [Br] 1999;81-B:955-62. Received 24 November 1997; Accepted after revision 20 May 1999

J. Lindahl, MD, Consultant Orthopaedic Surgeon E. Hirvensalo, MD, Consultant Orthopaedic Surgeon O. Böstman, MD, Consultant Orthopaedic Surgeon S. Santavirta, Professor of Orthopaedic Surgery Department of Orthopaedics and Traumatology, Helsinki University Central Hospital, Topeliuksenkatu 5, FIN 00260 Helsinki, Finland. Correspondence should be sent to Dr J. Lindahl. ©1999 British Editorial Society of Bone and Joint Surgery 0301-620X/99/68571 $2.00 VOL. 81-B, NO. 6, NOVEMBER 1999

Twenty years ago devices for external fixation became popular for the treatment of unstable injuries of the pelvic ring. Compared with conservative treatment they produced 1-5 better results, but it soon became clear that anterior external fixation frames had limitations when used for the most unstable injuries, especially in the posterior part of the 6-10 Methods of open reduction and internal pelvic ring. fixation were therefore introduced, especially for injuries of 11-15 the sacroiliac joint and pubic symphysis. The indications for operation, the methods employed and the results 16 of such treatment have varied. In earlier studies there 10,12,17 were also high rates of complications. The optimal management of different types of injury and the level of acceptable reduction in posterior or anterior disruptions of the pelvic ring are still controversial. We have assessed the long-term radiological and clinical results of unstable injuries of the pelvic ring treated by a trapezoidal external fixator. Special attention was paid to the accuracy of reduction, redisplacement, the final anatomical result and their effects on late symptoms and function. Associated neurological injuries were also evaluated.

Patients and Methods Between January 1982 and December 1993 we treated 132 consecutive patients with an unstable fracture of the pelvic ring by closed reduction and a trapezoidal external fixator. Of these, 110 (62 women and 48 men) were available for physical, functional and radiological examination; 16 had died (six during the first 24 hours, six in the early period, and four later) and six had been lost to follow-up. The mean age of the patients at the time of the injury was 39 years (13 to 87). The mean time of follow-up was 4.1 years (1 to 11). In 62% the cause of fracture was a traffic accident, in 28% a fall from a height, and in 10% other high-energy 12,18 accidents (Table I). According to Tile’s classification, there were 8 open-book (5 type B1 and 3 type B3-1) injuries, 62 lateral compression (31 type B2-1, 29 type B22 and 2 type 3-2) and 40 rotationally and vertically unstable (5 type C1-1, 7 type C1-2, 20 type C1-3, 5 type C2 and 3 type C3) injuries. Four (4%) were open fractures. The mean injury severity score (ISS) was 25 (10 to 19 57). 955

956

J. LINDAHL,

E. HIRVENSALO,

Table I. Mechanism of injury in 110 patients with injuries to the pelvic ring Fall from a height Motor-vehicle accident Car-pedestrian accident Motorcycle accident Major crushing Car-bicycle accident Miscellaneous

Number

Percentage

31 28 27 9 7 4 4

28 25 25 8 6 4 4

We used the trapezoidal external fixator as described by 2,4 Slätis and Karaharju in all patients. Closed reduction was performed before the final tightening of the frame. The stability of the pins was tested clinically and by checking plain radiographs taken in the operating theatre to see that the pins were in bone. Mobilisation on crutches began after six weeks in patients with type-C injuries and during the first week in those with type-B injuries if associated injuries allowed. All patients had a clinical examination with particular attention to their gait, discrepancy in limb-length, hip movement, difficulties in sitting, tilting or obliquity of the pelvis, scoliosis and persistent motor and sensory nerve 20 deficiencies. Pain. The residual pain was graded as: no pain, mild (intermittent, normal activity), moderate (limits activity, abolished by rest) and severe (continuous at rest, intense with activity). The site of pain in the pelvic ring (anterior and/or posterior) was recorded. Radiological analysis. The vertical displacement in type-C injuries with fracture of the sacrum, dislocation or fracturedislocation of the sacroiliac joint or a fracture of the ilium and anterior injury to the pelvic ring, was measured from anteroposterior (AP) radiographs of the pelvis. The AP 21 displacement was determined by CT. Vertical displacement and/or shortening with internal rotation of the hemipelvis in lateral compression injuries, in which there were fractures of the pubic rami and/or symphyseal disruption, and vertical displacement and diastasis in open-book injuries with symphyseal disruption were measured from plain AP radiographs. Radiographs were taken before primary treatment, after reduction and external fixation, after the removal of the external fixation frame, and at the final follow-up visit. The AP displacement of anterior injury to the pelvic ring could not be measured reliably by CT. The radiological result was graded by the maximal residual displacement in the posterior or anterior injury to the pelvic ring as: excellent, 0 to 5 mm; good, 6 to 10 mm; fair, 11 to 15 mm; and poor, more than 15 mm. Evaluation of outcome. The functional outcome was measured using a modification of a scoring system descri22 bed by Majeed which is based on the clinical findings. Ability to work was separated from this assessment giving a maximal total score of 80 points for each patient in order to compare the outcome of different types of fracture and subgroups (Table II). The original scoring system was

O. BÖSTMAN,

S. SANTAVIRTA

modified to focus on the outcome after an unstable fracture of the pelvic ring and not on the handicap caused by multiple injuries. The final score for the clinical outcome was also modified specifically to take account of the outcome after the pelvic injury (Table II). Statistical analysis. The BMDP software package V 7.0 (University of California, Berkeley) was used for the statistical analysis. The data are expressed as frequency and median. The effect of fracture dislocation on late pain was evaluated by Fisher’s exact two-tailed test. We used the Mann-Whitney rank-sum test for comparison of functional results between the subgroups. A p value of less than 0.05 was considered statistically significant. 22

Table II. Functional scoring system modified after Majeed

Points Pain Intense, continuous at rest Intense with activity Tolerable, but limits activity With moderate activity, abolished by rest Mild, intermittent, normal activity Slight, occasional or no pain Maximum Sitting Painful Painful if prolonged or awkward Uncomfortable Free Maximum Sexual intercourse Painful Painful if prolonged or awkward Uncomfortable Free Maximum Walking aids Bedridden or almost Wheelchair Two crutches Two sticks One stick No sticks Maximum Gait unaided Cannot walk or almost Shuffling small steps Gross limp Moderate limp Slight limp Normal Maximum Walking distance Bedridden or few metres Very limited time and distance Limited with sticks, difficult without prolonged standing possible One hour with a stick limited without One hour without sticks, slight pain or limp Normal for age and general condition Maximum Functional outcome (total score) Excellent Good Fair Poor

0 to 5 10 15 20 25 30 30 0 to 4 6 8 10 10 0 to 1 2 3 4 4 0 to 2 4 6 8 10 12 12 0 to 2 4 6 8 10 12 12 0 to 2 4 6 8 10 12 12

78 to 80 70 to 77 60 to 69