Objective To compare the effects of (a) tibolone, (b) continuous combined oestrogen plus progestogen and (c) placebo on plasma lipid and lipoprotein markers of cardiovascular risk in healthy postmenopausal women. Study design Randomized, single‐centre, placebo‐controlled, double‐blind study. Patients One hundred and one postmenopausal women were randomized (1:1:1) into one of three groups taking daily 2.5 mg tibolone, continuous oral oestradiol‐17β 2 mg plus norethisterone acetate 1 mg daily (E2/NETA) or placebo. Main outcome measures Fasting serum lipid, lipoprotein and apolipoprotein concentrations measured at baseline and after 6, 12 and 24 months of treatment. Results Both tibolone and E2/NETA lowered plasma total cholesterol concentrations relative to placebo. With tibolone, high‐density lipoprotein cholesterol (HDL‐C) was reduced (−27% at 24 months, P < .001), the greatest effect being in the cholesterol‐enriched HDL2 subfraction (−40%, P < .001). Tibolone's effect on HDL concentrations was also apparent in the principal HDL protein component, apolipoprotein AI (−29% at 24 months, P < .001). However, there was no significant effect of tibolone on low‐density or very low‐density lipoprotein cholesterol (LDL‐C and VLDL‐C, respectively). By contrast, the greatest reduction in cholesterol with E2/NETA was in LDL‐C (−22% at 24 months, P = .008). E2/NETA reduced HDL‐C to a lesser extent than tibolone (−12% at 24 months, P < .001). Effects on HDL apolipoproteins were similarly diminished relative to tibolone. E2/NETA had no effect on VLDL‐C or on the protein component of LDL, apolipoprotein B. Conclusion Tibolone reduces serum HDL. E2/NETA reduces LDL cholesterol but not apolipoprotein B, suggesting decreased cholesterol loading of LDL. Any impact these changes may have on CVD risk needs further investigation.

To the Editor We read with great interest the recently published article by Henrot and colleagues 1 on self-compression techniques vs standard-compression in mammography. If feasible, this innovative idea could increase the uptake of mammography screening among women. However, there are limitations in the secondary outcomes of this study, which we would like to address.First, we appreciate the difficulty in blinding participants in this study. However, owing to the subjective nature of the data, such as pain and patient satisfaction, measures could have been taken to overcome the unavoidable bias. Because the aim of the study was to prove noninferiority, one such approach would be to carry out self-compression on one of the patient's breasts and standard compression on the other, as completed in the study by Kornguth and colleagues. 2 This would allow a direct subjective comparison per participant without compromising the objective data collected for the other end points.Henrot and colleagues 1 admitted that the visual analogue scale was a limitation of the study. The mean visual analogue scale score in the self-compression and standard-compression groups was 2 and 3, respectively. Although statistically significant, considering that the study was not blinded, a larger difference would be required to justify this as clinically relevant. Moreover, no significant difference was found with regard to physical and psychological pain in the mammography questionnaire. A modification to the questionnaire would be to include questions relating to patient reattendance and to deduce whether standard compression truly causes more pain and future avoidance of mammograms.The study also involved training the participants in a pretest. We understand that this was vital to ensure that the groups were equally skilled in performing self-compression. However, it suggests that self-compression requires extra training. Even though the study states that it did not focus on the generalizability of these findings, this is a key aspect to consider because training women to perform self-compression may not be practical. Only a clinically significant difference in patient outcome would prompt a change in current practice to incorporate self-compression; with the current results, this is difficult to justify.The authors 1 have achieved their intended primary aim for this study. We believe the next step would be a modified methodology with an improved questionnaire to assess the practicality and to elicit differences in patient satisfaction between the 2 approaches. This is welcomed because it would establish whether self-compression could increase the uptake of screening and thereby benefit public health.

Standardized patient simulation versus didactic teaching alone for improving residents' communication skills when discussing goals of care and resuscitation: a randomized controlled trial.