Sir, We thank Dr. Doumouchtsis and colleagues for their interest in our study, and for pursuing the debate regarding what to do when medical treatment fails 1. The main message of our work was that the B-Lynch technique appears to be an effective procedure with a relatively low morbidity, to control persistent severe postpartum hemorrhage (PPH) following a failure of vessel ligation before considering hysterectomy 2. Our study design did not permit us to determine the place of each conservative surgical procedure (first-, second-, even third-line) in the PPH surgical management algorithm. Nevertheless, in our opinion, there remains insufficient data (only 68 reported cases of B-Lynch suture found in June 2006 by Doumouchtsis and colleagues 1) to widely recommend the B-Lynch suture in the first-line surgical treatment of PPH 2. The aim of the conservative surgical procedure is not only to control PPH with the lowest morbidity, but also to preserve a theoretical functional uterus that will not compromise the patients’ subsequent fertility and obstetrical outcome. First, the main argument to recommend stepwise uterine devascularization as the first-line surgical treatment for PPH is that there are data suggesting that subsequent fertility and pregnancy are not altered following uterine artery ligation and stepwise uterine devascularization 3-7 (and also following hypogastric artery ligation 8), whereas, currently, this parameter has been very poorly studied following B-Lynch suture 9. Second, to date, no adverse events have been reported following uterine artery ligation or stepwise uterine devascularization 3, 4, 7, with the exception of two broad ligament hematomas in a series of 265 uterine artery ligations 7, whereas uterine necrosis requiring hysterectomy has been reported following isolated B-Lynch suture 10, 11. Third, in the absence of controlled trials, systematic reviews are essential in attempting to compare the efficacy of the different conservative surgical procedures 1. Nevertheless, their results should be interpreted with caution. Therefore, in our opinion, it remains unclear whether B-Lynch suture (of which the largest cohort series reported until June 2006 included only 19 patients 1) has really the same overall success rate as uterine artery ligation or stepwise uterine devascularization (of which the largest cohort series reported until June 2006 included, respectively, 265 and 103 patients 1). It is, in fact, questionable (i) to pool the uterine artery ligation/stepwise uterine artery devascularization (success rate of 94.7%, 288/304) with hypogastric artery ligation (success rate of 69.0%, 136/197) in the same category 1, as the hypogastric artery ligation has been found to be considerably less successful than previously thought 12, 13; and (ii) to pool procedures whatever the cause of PPH 1, because the success rate of a conservative procedure for PPH very likely depends on it 14. In particular, the B-Lynch suture has been essentially reported following uterine atony (which is associated with a high rate of success) and rarely in cases of placenta accreta (which is associated with the lowest rate of success), contrary to uterine artery ligation or stepwise uterine devascularization 7, 15. Thus, comparison is unfair if the incidence of placenta accreta is not the same between these procedures. Fourth, stepwise uterine devascularization similar to the B-Lynch suture is a simple and rapid procedure that is accessible to all practitioners. Therefore, in our opinion, stepwise uterine devascularization should currently be the first-line conservative surgical treatment for PPH, in particular as long as the impact on patient's subsequent fertility and pregnancies following the B-Lynch suture is insufficiently assessed [9]. In cases of failed stepwise uterine devascularization, we agree with Dr. Doumouchtsis and colleagues that ‘it would appear reasonable to consider a compression suture prior to (hypogastric artery) ligation’, as hypogastric artery ligation may be associated with severe morbidity, and as it has never previously been assessed in second-line surgical treatment, contrary to the B-Lynch suture 2. Since our study 2, the B-Lynch suture has now become our second-line surgical treatment in our algorithm management for PPH, hypogastric artery ligation being considered before peripartum hysterectomy only by practitioners familiar with the technique, in a hemodynamically stable patient with a desire for subsequent pregnancy, and if the multidisciplinary team considers that the condition is under control. We also totally agree with Doumouchtsis and colleagues that the easiest and less morbid conservative procedure should be preferred in cases of failed medical treatment to control PPH. However, we did not include uterine balloon tamponade in our algorithm for PPH because this algorithm was developed before 2003, the date when the first assessments of this procedure for PPH became available. Moreover, a recent review of the literature documented a total of 673 pelvic arterial embolizations with low morbidity and at least a 91% success rate, in particular in uterine atony, without alteration of the subsequent fertility and pregnancy rates 16. As uterine balloon tamponade has been assessed only in the short term, and in fewer patients (n = 162) with a success rate slightly inferior (84%), in our opinion pelvic arterial embolization should be preferred in the case of persistent bleeding following vaginal delivery, when the embolization unit is located close to the delivery room. Nevertheless, uterine balloon tamponade is certainly an interesting option in low resource settings where arterial embolization is not available, in particular because it can at least provide time to perform the embolization even if success is incomplete. Such a result has been reported by our colleagues from the Evreux Hospital 17, who have routinely transferred their patients to our maternity unit, for arterial embolization in cases of persistent bleeding in hemodynamically stable patients 16.