Authors: P Fleming, P Connolly, J Rice, D McCormack

The Children’s Hospital, Temple Street, Dublin 1.


Contributors: K Synnott, Selvaraj, P Murray, A Poynton


The case of a 12 year old girl with bilateral recurrent patellar dislocations was presented. Initially the discussion concerned the pathogenesis, natural history and assessment of patello-femoral instability, and subsequently the principles of operative management were discussed. Additional analysis is given in this review, distilled from various texts and an instructional course lecture written by David Dandy.

Pathogenesis and risk factors

The forces acting on the normal patella attempt to displace it laterally, because the tibial tubercle lies lateral to the long axis of the femur and the pull of the quadriceps muscle. When the knee is flexed, the patella slides distally and is engaged by the trochlear groove in the femur. The lateral lip of the trochlea and tension in the medial soft tissues prevent lateral dislocation in the normal situation. Risk factors which predispose to patellar instability can be divided into two main groups:


  1. Factors causing poor engagement of the patella in the trochlea: eg an abnormally high patella (patella alta), patellar dysplasia, or a poorly developed trochlea
  2. Factors which allow the patella to move laterally out of the trochlea, such as a defective lateral trochlear margin or an abnormally shallow trochlear groove. Laxity of the medial retinaculum plays a role in the development of chronic dislocation.


Other reported risk factors include the presence of an osteochondral fragment, a very active individual, generalised ligamentous laxity, excessive femoral anteversion and external rotation deformity of the tibia.


Natural history

The natural history of an acute dislocation is that, overall, approximately one in six patients will develop recurrent dislocation, two in six will have some minor residual symptoms while three in six will be asymptomatic. The incidence of redislocation decreases with age; however in the 11-14 age group it is approximately 60%.


Assessment of recurrent patellar dislocation

Assessment begins with a good history. The circumstances of the original dislocation should be determined, as should the type of movement that precipitates the redislocation - does it occur by simply flexing the knee or is a twisting movement required?


On clinical examination the point at which the patella engages in the trochlea should be noted. Can apprehension be elicited by manualy directing the patella laterally while flexing the knee? Are the patellar retinacula abnormally loose, or does the patient have signs of generalised ligamentous laxity?


Radiographically, valuable information can be obtained. The patellar height can be assessed by either the Insall-Salvati index or the Blackburne-Peel index:

  1. The Insall-Salvati index compares the length of the patella (in its longest diameter on a lateral x-ray) to the length of the patellar tendon. These measurements should be almost the same - patella usually very slightly shorter - and the ratio should be 0.9 -1.1 in males and 0.94 -1.18 in females (adults).Although probably the most widely used index, apparently there is inter-observer error in defining the point of attachment of the tendon to the tibia. Moreover, the length of the patella is not consistently related to the length of the articular surface, which is the more relevant variable.

    2. The Blackburne-Peel index relates the length of the articular surface of the patella to the distance from the patella to the tibial plateau. The normal range for men is 0.85 -1.09, and for women is 0.70 -1.09. This gives a better idea of the relationship between the patella and the femoral trochlea.


Dysplastic condyles and a shallow trochlear groove may be evident on a true lateral x-ray of the knee. The Q-angle (the angle between the quadriceps vector - approximated by a line joining the ASIS to the centre of the patella - and the line of pull of the patellar tendon from the centre of the patella to the tibial tubercle) can be measured on a long-leg film. Some authors accept that while the Q-angle is a useful dynamic concept, an unstable patella will tend to lie more laterally than normal in full extension thereby decreasing the angle.


Arthroscopy is not routinely performed, but (through a medial or suprapatellar approach) allows visualisation of the engagement in the trochlea of the patella. A normal patella should centre in the trochlear groove at between 30o and 60o of flexion.


Surgical treatment

The procedures described to stabilise an unstable patella can be broadly classed as proximal realignments, which alter the tension of the tissues which bind the patella, or distal realignments, which medially transpose the ligamentum patellae. A large number of operations within these broad categories have been described.

Special considerations arise in the paediatric setting. Bony procedures such as Elmslie’s medial transposition of the tibial tubercle and Maquet’s procedure are contraindicated in patients (like our 12 year old patient) who are not yet skeletally mature, because of the unacceptably high risk of genu recurvatum deformity caused by premature fusion of the anterior part of the tibial physis. Only soft-tissue stabilisations should be chosen. Furthermore, stabilisation is important in children - to encourage normal development of the patello-femoral joint and to reduce the risk of osteoarthritis secondary to articular surface lesions caused by recurrent dislocations.


We performed a modified Insall technique (a lateral release combined with lateral ‘double-breasting’ of the medial retinaculum) for proximal realignment and combined this with a Roux-Goldthwait procedure distally (splitting the patellar tendon, detaching the lateral half and transferring it medially beneath the other half). As part of our proximal realignment the medial retinacular flap was lifted with the periosteum of the medial one third of the patella attached. A series of per-operative photographs taken during stabilisation of the right knee was presented.



Patello-femoral arthritis, loss of full flexion and recurrence of dislocation may all complicate surgical stabilisation. Mr P Murray and Mr A Poynton drew attention to the problem of medial dislocation of the patella particularly following radical lateral release. This presents as persistent pain and giving way, and is associated with a positive medial apprehension test. In one series of patients with persistent problems following arthroscopic lateral release, medial dislocation was the cause in 50% of cases. Good results can be obtained by repairing or reconstructing the lateral patello-tibial ligament.