Objective: to investigate the diagnosis, incidence and management of s
houlder dystocia, Design: retrospective review of all the deliveries b
etween 1988 and 1993 which were coded on the computer data base as 'mo
derate difficulty with the shoulders' or 'shoulder dystocia'. Setting:
a teaching hospital in the south of England. Participants: the study
population consisted of 15 658 women who had cephalic vaginal deliveri
es of babies weighing > 2.5 kg, The sample consisted of the 257 women
reported to have moderate difficulty with the shoulders or shoulder dy
stocia at delivery, Findings: there was a significant fall in the repo
rted incidence of shoulder dystocia during the period under investigat
ion. Eighty-six (1.1%) of the deliveries between 1988 and 1990 were re
ported to have been complicated by shoulder dystocia, In 1991 a second
option was introduced to allow both shoulder dystocia or moderate dif
ficulties with the shoulders to be coded after delivery. Following thi
s change in categorisation, the incidence of shoulder dystocia was red
uced to 30 (0.6%) reported cases in 1992-1993, Over the same period th
e reported incidence of moderate difficulty rose significantly from 29
(1.1%) in 1992 to 60 (2.4%) in 1993 (P < 0.001). Disparity was noted
between the experience of practitioners and contemporary definitions o
f shoulder dystocia and this was highlighted by the number of reported
cases of moderate difficulty. In this study practitioners used the te
rm shoulder dystocia in a general sense to describe a range of difficu
lties encountered with the delivery of the shoulders and they identifi
ed many varied manifestations. In comparison, contemporary literature
describes shoulder dystocia as a discrete entity. Key conclusions: sho
ulder dystocia is a complex clinical scenario and perceptions of the i
ncidence may be influenced by alterations in the delivery technique an
d changes in the diagnosis and documentation. Implications for practic
e: the application of traction may interfere with the normal mechanism
s of labour and the spontaneous birth of the shoulders, whilst increas
ing the risk of trauma to the baby. It may be unnecessary to use such
interventions routinely in the care of normal labouring women.