is a relatively common disorder in rheumatologic practice. In most cases it is caused by mechanical irritation, and develops frequently in people who have to kneel for long periods of time such as housemaids, nuns, and miners (1). The bursa may be infected either through the skin or by hematogeneous spread (2).
Although erysipelas is common, it is of interest that no case of streptococcal bursitis due to erysipelas has been reported. Balboni is the only author to mention that arthralgia, acute arthritis, and (rarely) suppurative arthritis may be associated with erysipelas, but he fails to give any case report or reference (3).
We report here an unusual case of erysipelas complicated by a septic bursitis. EZ, a 66-year-old white man, was first treated in May 1978 in our department for osteoarthrosis of the right knee. Treatment consisted of indomethacin and physiotherapy. No intraarticular injections were given. The patient became better and was discharged, but a few days later he developed a fever of 39.2"C and chills and noticed a rash above his right knee.
Physical examination revealed a typical erysipelas rash above the right knee. The prepatellar bursa was swollen, full of fluid, and very tender to palpation. There was no swelling or tenderness of the joint itself and no limitation of movement was noticed.
The bursa was aspirated and 10 ml of bloody, purulent fluid was obtained. The total synovial white cell count was 50,000/mm3, with 60% polymorphonuclear leukocytes. Hemolytic streptococci were cultured from the fluid. Other laboratory findings included a Westergren sedimentation rate of 25 mm/hour, total white cell count 9600/mm3, and an antistreptolysin titer of 1/200.
Penicillin therapy was begun at 1,600,000 units intramuscularly daily. The bursa was aspirated and washed out every day with 0.9% saline solution, and penicillin was injected locally into the bursa. Within a few days the patient's fever ceased, the inflammatory symptoms of the skin subsided, and the bursa1 fluid remained sterile after repeated cultures. The patient recovered completely in another 2 weeks, and the bursa did not require surgical removal.
In 1916 Fullerton (4) described the relationship between the knee joint and several of its adjacent bursae with emphasis on their importance in spreading infections. The prepatellar bursa, however, is usually not