The data from the nationwide hospital discharge register was used for
collecting the diagnoses of the thoracic outlet syndrome (TOS) which w
ere combined with the procedure numbers of first rib resection and sca
lenotomy. During the years 1987-1993 the total number of operative per
iods for patients with a TOS diagnosis was 483 for 464 patients. Each
year first rib resections were done significantly (P = 0.001) more oft
en (55.7 SE 7.1; 1.11/100,000) than scalenotomies (13.4 SE 1.8; 0.27/1
00,000). The operation for TOS was most commonly combined with the dia
gnosis of TOS with brachial plexus lesion in 53 %, TOS NUD (not classi
fied) in 21 %, TOS with subclavian artery compression in 19 %, with a
cervical rib in 4 % and TOS with venous compression in 3 %. The large
proportion of the diagnosis TOS NUD clearly shows the need for a bette
r definition for the TOS diagnosis. As long as clear diagnostic criter
ia are lacking, the division of TOS into subgroups is arbitrary. Diagn
ostic division into true neurogenic, major arterial and venous TOS, an
d classifying the rest of the TOS diagnoses under TOS NUD or cervicobr
achiale diffusum is recommended.