Objective To establish the lowest dose of cyclical dydrogesterone that protects against endometrial hyperplasia induced by continuous 2 mg 17 beta oestradiol, and to study the dose effect on vaginal bleeding and side effects. Design Double-blind, prospectively randomised dose-ranging study. Setting Menopause clinics in the UK and The Netherlands. Subjects Three hundred and seventy-one postmenopausal women with intact uteri, aged 40 to 60. Interventions Administration of six 28-day treatment cycles of continuous daily micronised 17 beta oestradiol with a randomly allocated dose of 5 to 20 mg of dydrogesterone added for the last 14 days of each. Main outcome measures Histological assessment of adequate progestational endometrial response, bleeding patterns and adverse effects. Results The study was completed by 320 subjects (86%). Endometrial transformation occurred in over 94% of those taking 5 mg of dydrogesterone, and in over 97% of those on higher doses, without significant differences between the 10, 15 and 20 mg groups. Acceptable bleeding patterns were found at all doses, with the incidence of withdrawal bleeding rising with increasing dose. The day of onset of bleeding was predictable from cycle to cycle, and occurred later in the 20 mg group than in the others. The incidence of noncyclic bleeding was about 6% at all doses. Withdrawal occurred in 3.3% due to unacceptable bleeding and in 5.4% due to side effects. There was no relation with dose. Conclusions A dydrogesterone-17 beta oestradiol combination hormone replacement therapy confers endometrial protection with an acceptable bleeding pattern and few side effects. At least 10 mg of dydrogesterone for 14 days is required for acceptable endometrial protection.

A 37 year old woman with secondary infertility had been advised by her general practitioner to record her basal body temperature vaginally. One morning she presented to a casualty department complaining that she could not remove the thermometer. Vaginal examination and a radiograph of the pelvis (figure) suggested that the thermometer was in the bladder. This was confirmed at cytoscopy, when the thermometer was removed under general anaesthesia with stone crushing forceps whose jaws were shod with rubber cylinders cut from a latex catheter. The patient made a full recovery and was discharged the next day. Inadvertent insertion of a thermometer into the bladder during recording of vaginal temperature has been reported many times before, and we think it time that doctors stopped advising women to record their basal temperatures Vaginally: oral recording is safe and adequate.-E EYONG, D J BURCH, Department of Obstetrics and Gynaecology, University HospitalofSouthManchester, Manchester M20 8LR. ........ ... :.5 -; ........ ;;.

De Gennes' proposal (see J. Physique Lett., vol.36, p.55 (1975)) that the n to 0 limit of an O(n) symmetric theta 4 theory describes a single polymer chain in solution is extended to allow for the interactions between several polymer chains. To describe m interacting polymers, it is found to be necessary to employ an (m*n)-component field theory before taking the n to 0 limit. Explicit calculation of the second virial coefficient of the osmotic pressure for self-avoiding walks on a Bethe lattice is performed to illustrate the formalism.

The Riedel-Wegner (see Phys. Rev. B, vol.7, p.248 (1973)) phenomenological approach to crossover phenomena is used to calculate the temperature and molecular-weight dependence of the second virial coefficient and expansion factor of polymers in dilute solution both near the Theta temperature and at higher temperatures. The specific heat maximum of a single polymer chain in solution is predicted to diverge logarithmically as its molecular weight goes to infinity. The Theta temperature is assumed analogous to a tricritical point. It is shown that for problems involving the interaction of m polymers, one must work with an (m*n)-component field theory, and then take the limit n to O.