A family history of shoulder instability in first-degree relatives was found in 24 of 100 patients who had been operated on for recurrent anterior shoulder instability. The patients with and without a family history were similar in respect of sex ratio, age at first dislocation and age at operation. The initial dislocation was non-traumatic in 22% of the patients with and in 13% of those without a positive family history (p = 0.3). Postoperative recurrence of instability was experienced by 34% of patients with and 33% of those without a family history (p = 0.9). In those with a family history, 13% of the recurrences were dislocations and 87% were subluxations. In patients without a family history 52% of the recurrences were dislocations and 48% were subluxations (p < 0.05). The contralateral shoulder was unstable in 50% of the patients with a family history and in 26% of those without (p = 0.03).

We used 11 cadaver elbows and a three-dimensional electromagnetic tracking device to record elbow movements before and after implantation of a 'loose-hinged' elbow prosthesis (modified Coonrad). During simulated active motion there was a maximum of 2.7 degrees (+/- 1.5 degrees) varus/valgus laxity in the cadaver joints. This increased slightly after total elbow arthroplasty to 3.8 degrees (+/- 1.4 degrees). These values are lower than those recorded for the cadaver joints and for the prostheses at the limits of their varus/valgus displacements, indicating that both behave as 'semi-constrained' joints under physiological conditions. They suggest that the muscles absorb some of the forces and moments that in a constrained prosthesis would be transferred to the prosthesis-bone interface.

We reviewed 188 patients at one to 20 years (mean 9.5) after surgery for anterior shoulder instability. Twenty-one had shown bilateral instability at the time of surgery, and 26 of the remaining 167 subsequently developed instability of the contralateral shoulder, giving an overall incidence of 24% bilateral involvement. Fourteen of these patients ultimately required bilateral surgery. The onset of contralateral instability was at one month to 15 years (mean 5.7 years) after anterior repair of the operated shoulder, the cumulative incidence increasing with time (p less than 0.01). The incidence was significantly higher in those under 15 years at the time of initial dislocation or under 18 at the time of surgery. One-half of the patients with contralateral instability had signs of posterior instability at follow-up. Other predisposing factors included having sustained the initial injury to the operated shoulder as a result of minimal trauma, and persistence of a sensation of instability in the operated shoulder. Factors which were not statistically significant included sex, dominant side, athletic activity, work history, and whether the initial surgery was for recurrent subluxations or dislocations. The high prevalence of bilateral shoulder instability suggests an intrinsic abnormality such as capsular and ligamentous laxity or muscle imbalance and warrants further investigations.

The anatomy of the cubital tunnel and its relationship to ulnar nerve compression is not well documented. In 27 cadaver elbows the proximal edge of the roof of the cubital tunnel was formed by a fibrous band that we call the cubital tunnel retinaculum (CTR). The band is about 4 mm wide, extending from the medial epicondyle to the olecranon, and perpendicular to the flexor carpi ulnaris aponeurosis. Variations in the CTR were classified into four types. In type 0 (n = 1) the CTR was absent. In type Ia (n = 17), the retinaculum was lax in extension and taut in full flexion. In type Ib (n = 6) it was tight in positions short of full flexion (90 degrees to 120 degrees). In type II (n = 3) it was replaced by a muscle, the anconeus epitrochlearis. The CTR appears to be a remnant of the anconeus epitrochlearis muscle and its function is to hold the ulnar nerve in position. Variations in the anatomy of the CTR may explain certain types of ulnar neuropathy. Its absence (type 0 CTR) permits ulnar nerve displacement. Type Ia is normal and does not cause ulnar neuropathy. Type Ib can cause dynamic nerve compression with elbow flexion. Type II may be associated with static compression due to the bulk of the anconeus epitrochlearis muscle.