Context The GyneFix intra-uterine device has been used in our family planning service since 1997. One of the perceived advantages is its low expulsion rate, as reported by clinical trials.

Objective To calculate expulsion rates in routine clinical use and to look at possible reasons for expulsion.

Design Retrospective casenote analysis and opportunistic client consultation.

Setting A city centre family planning clinic.

Participants The first 1000 GyneFix insertions.

Main outcome measures Parity of client, experience of clinician carrying out insertion, time from device insertion to expulsion.

Results Overall expulsion rate was 7.6%. There was no significant difference in parity of clients experiencing expulsion. Most (4.7%) expulsions were early, occurring within 3 months of insertion. There was considerable variation in early expulsion rate from one clinician to another. Later expulsions also occurred, up to 28 months after insertion. Increasing experience of the inserting clinician led to lower rates of late expulsion. Unnoticed expulsion led to four unplanned pregnancies.

Conclusions The GyneFix expulsion rate in our service is higher than quoted in clinical trials. Early expulsions may be related to insertion technique, representing insufficient implantation of the anchoring knot into the fundal myometrium. Late expulsions also occur, often many months after insertion; the reason for these is unclear. Users should be taught to check for the presence of the thread after each menstrual period and unnoticed expulsion should be confirmed by ultrasound and abdomino-pelvic plain X-ray.