Core outcome sets for prevention and treatment of postpartum haemorrhage: an international Delphi consensus study.

OBJECTIVE: To develop core outcome sets (COS) for studies evaluating interventions for (1) prevention and (2) treatment of postpartum haemorrhage (PPH), and recommendations on how to report the COS. DESIGN: A two-round Delphi survey and face-to-face meeting. POPULATION: Healthcare professionals and women's representatives. METHODS: Outcomes were identified from systematic reviews of PPH studies and stakeholder consultation. Participants scored each outcome in the Delphi on a Likert scale between 1 (not important) and 9 (critically important). Results were discussed at the face-to-face meeting to agree the final COS. Consensus at the meeting was defined as ≥ 70% of participants scoring the outcome as critically important (7-9). Lectures, discussion and voting were used to agree how to report COS outcomes. MAIN OUTCOME MEASURES: Outcomes from systematic reviews and consultations. RESULTS: Both Delphi rounds were completed by 152/205 (74%) participants for prevention and 143/197 (73%) for treatment. For prevention of PPH, nine core outcomes were selected: blood loss, shock, maternal death, use of additional uterotonics, blood transfusion, transfer for higher level of care, women's sense of wellbeing, acceptability and satisfaction with the intervention, breastfeeding, and adverse effects. For treatment of PPH, 12 core outcomes were selected: blood loss, shock, coagulopathy, hysterectomy, organ dysfunction, maternal death, blood transfusion, use of additional haemostatic intervention, transfer for higher level of care, women's sense of wellbeing, acceptability and satisfaction with the intervention, breastfeeding, and adverse effects. Recommendations were developed on how to report these outcomes where possible. CONCLUSIONS: These COS will help standardise outcome reporting in PPH trials. TWEETABLE ABSTRACT: Core outcome sets for PPH: nine core outcomes for PPH prevention and 12 core outcomes for PPH treatment.

Cervical dilatation patterns of 'low-risk' women with spontaneous labour and normal perinatal outcomes: a systematic review.

BACKGROUND: The call for women-centred approaches to reduce labour interventions, particularly primary caesarean section, has renewed an interest in gaining a better understanding of natural labour progression. OBJECTIVE: To synthesise available data on the cervical dilatation patterns during spontaneous labour of 'low-risk' women with normal perinatal outcomes. SEARCH STRATEGY: PubMed, EMBASE, CINAHL, POPLINE, Global Health Library, and reference lists of eligible studies. SELECTION CRITERIA: Observational studies and other study designs. DATA COLLECTION AND ANALYSIS: Two authors extracted data on: maternal characteristics; labour interventions; the duration of labour centimetre by centimetre; and the duration of labour from dilatation at admission through to 10 cm. We pooled data across studies using weighted medians and employed the Bootstrap-t method to generate the corresponding confidence bounds. MAIN RESULTS: Seven observational studies describing labour patterns for 99 971 women met our inclusion criteria. The median time to advance by 1 cm in nulliparous women was longer than 1 hour until a dilatation of 5 cm was reached, with markedly rapid progress after 6 cm. Similar labour progression patterns were observed in parous women. The 95th percentiles for both parity groups suggest that it was not uncommon for some women to reach 10 cm, despite dilatation rates that were much slower than the 1-cm/hour threshold for most part of their first stage of labours. CONCLUSION: An expectation of a minimum cervical dilatation threshold of 1 cm/hour throughout the first stage of labour is unrealistic for most healthy nulliparous and parous women. Our findings call into question the universal application of clinical standards that are conceptually based on an expectation of linear labour progress in all women. FUNDING: UNDP/UNFPA/UNICEF/WHO/World Bank Special Programme of Research, Development and Research Training in Human Reproduction (HRP), Department of Reproductive Health and Research, World Health Organization, and the United States Agency for International Development (USAID). TWEETABLE ABSTRACT: Cervical dilatation threshold of 1 cm/hour throughout labour is unrealistic for most women, regardless of parity.

Caesarean section surgical techniques: 3 year follow-up of the CORONIS fractional, factorial, unmasked, randomised controlled trial.

BACKGROUND: The CORONIS trial reported differences in short-term maternal morbidity when comparing five pairs of alternative surgical techniques for caesarean section. Here we report outcomes at 3 years follow-up. METHODS: The CORONIS trial was a pragmatic international 2 × 2 × 2 × 2× 2 non-regular fractional, factorial, unmasked, randomised controlled trial done at 19 sites in Argentina, Chile, Ghana, India, Kenya, Pakistan, and Sudan. Pregnant women were eligible if they were to undergo their first or second caesarean section through a planned transverse abdominal incision. Women were randomly assigned by a secure web-based allocation system to one intervention from each of the three assigned pairs. All investigators, surgeons, and participants were unmasked to treatment allocation. In this follow-up study, we compared outcomes at 3 years following blunt versus sharp abdominal entry, exteriorisation of the uterus for repair versus intra-abdominal repair, single versus double layer closure of the uterus, closure versus non-closure of the peritoneum, and chromic catgut versus polyglactin-910 for uterine repair. Outcomes included pelvic pain; deep dyspareunia; hysterectomy and outcomes of subsequent pregnancies. Outcomes were assessed masked to the original trial allocation. This trial is registered with the Current Controlled Trials registry, number ISRCTN31089967. FINDINGS: Between Sept 1, 2011, and Sept 30, 2014, 13,153 (84%) women were followed-up for a mean duration of 3·8 years (SD 0·86). For blunt versus sharp abdominal entry there was no evidence of a difference in risk of abdominal hernias (adjusted RR 0·66; 95% CI 0·39-1·11). We also recorded no evidence of a difference in risk of death or serious morbidity of the children born at the time of trial entry (0·99, 0·83-1·17). For exteriorisation of the uterus versus intra-abdominal repair there was no evidence of a difference in risk of infertility (0·91, 0·71-1·18) or of ectopic pregnancy (0·50, 0·15-1·66). For single versus double layer closure of the uterus there was no evidence of a difference in maternal death (0·78, 0·46-1·32) or a composite of pregnancy complications (1·20, 0·75-1·90). For closure versus non-closure of the peritoneum there was no evidence of a difference in any outcomes relating to symptoms associated with pelvic adhesions such as infertility (0·80, 0·61-1·06). For chromic catgut versus polyglactin-910 sutures there was no evidence of a difference in the main comparisons for adverse pregnancy outcomes in a subsequent pregnancy, such as uterine rupture (3·05, 0·32-29·29). Overall, severe adverse outcomes were uncommon in these settings. INTERPRETATION: Although our study was not powered to detect modest differences in rare but serious events, there was no evidence to favour one technique over another. Other considerations will probably affect clinical practice, such as the time and cost saving of different approaches. FUNDING: UK Medical Research Council and the Department for International Development.

Antenatal care for healthy pregnant women: a mapping of interventions from existing guidelines to inform the development of new WHO guidance on antenatal care.

BACKGROUND: The World Health Organization (WHO) is in the process of updating antenatal care (ANC) guidelines. OBJECTIVES: To map the existing clinical practice guidelines related to routine ANC for healthy women and to summarise all practices considered during routine ANC. SEARCH STRATEGY: A systematic search in four databases for all clinical practice guidelines published after January 2000. SELECTION CRITERIA: Two researchers independently assessed the list of potentially eligible publications. DATA COLLECTION AND ANALYSIS: Information on scope of the guideline, type of practice, associated gestational age, recommendation type and the source of evidence were mapped. MAIN RESULTS: Of 1866 references, we identified 85 guidelines focusing on the ANC period: 15 pertaining to routine ANC and 70 pertaining to specific situations. A total of 135 interventions from routine ANC guidelines were extracted, and categorised as clinical interventions (n = 80), screening/diagnostic procedures (n = 47) and health systems related (n = 8). Screening interventions, (syphilis, anaemia) were the most common practices. Within the 70 specific situation guidelines, 102 recommendations were identified. Overall, for 33 (out of 171) interventions there were conflicting recommendations provided by the different guidelines. CONCLUSION: Mapping the current guidelines including practices related to routine ANC informed the scoping phase for the WHO guideline for ANC. Our analysis indicates that guideline development processes may lead to different recommendations, due to context, evidence base or assessment of evidence. It would be useful for guideline developers to map and refer to other similar guidelines and, where relevant, explore the discrepancies in recommendations and others. TWEETABLE ABSTRACT: We identified existing ANC guidelines and mapped scope, practices, recommendations and source of evidence.

A global reference for caesarean section rates (C-Model): a multicountry cross-sectional study.

OBJECTIVE: To generate a global reference for caesarean section (CS) rates at health facilities. DESIGN: Cross-sectional study. SETTING: Health facilities from 43 countries. POPULATION/SAMPLE: Thirty eight thousand three hundred and twenty-four women giving birth from 22 countries for model building and 10,045,875 women giving birth from 43 countries for model testing. METHODS: We hypothesised that mathematical models could determine the relationship between clinical-obstetric characteristics and CS. These models generated probabilities of CS that could be compared with the observed CS rates. We devised a three-step approach to generate the global benchmark of CS rates at health facilities: creation of a multi-country reference population, building mathematical models, and testing these models. MAIN OUTCOME MEASURES: Area under the ROC curves, diagnostic odds ratio, expected CS rate, observed CS rate. RESULTS: According to the different versions of the model, areas under the ROC curves suggested a good discriminatory capacity of C-Model, with summary estimates ranging from 0.832 to 0.844. The C-Model was able to generate expected CS rates adjusted for the case-mix of the obstetric population. We have also prepared an e-calculator to facilitate use of C-Model (www.who.int/reproductivehealth/publications/maternal_perinatal_health/c-model/en/). CONCLUSIONS: This article describes the development of a global reference for CS rates. Based on maternal characteristics, this tool was able to generate an individualised expected CS rate for health facilities or groups of health facilities. With C-Model, obstetric teams, health system managers, health facilities, health insurance companies, and governments can produce a customised reference CS rate for assessing use (and overuse) of CS. TWEETABLE ABSTRACT: The C-Model provides a customized benchmark for caesarean section rates in health facilities and systems.

Are hypertensive disorders in pregnancy associated with congenital malformations in offspring? Evidence from the WHO Multicountry cross sectional survey on maternal and newborn health.

BACKGROUND: Annually, around 7.9 million children are born with birth defects and the contribution of congenital malformations to neonatal mortality is generally high. Congenital malformations in children born to mothers with hypertensive disorders during pregnancy has marginally been explored. METHODS: Country incidence of congenital malformations was estimated using data on the 310 401 livebirths of the WHO Multicountry Survey which reported information from 359 facilities across 29 countries. A random-effect logistic regression model was utilized to explore the associations between six broad categories of congenital malformations and the four maternal hypertensive disorders "Chronic Hypertension", "Preeclampsia" and "Eclampsia" and "Chronic hypertension with superimposed preeclampsia". RESULTS: The occupied territories of Palestine presented the highest rates in all groups of malformation except for the "Lip/Cleft/Palate" category. Newborns of women with chronic maternal hypertension were associated with a 3.7 (95 % CI 1.3-10.7), 3.9 (95 % CI 1.7-9.0) and 4.2 (95 % CI 1.5-11.6) times increase in odds of renal, limb and lip/cleft/palate malformations respectively. Chronic hypertension with superimposed preeclampsia was associated with a 4.3 (95 % CI 1.3-14.4), 8.7 (95 % CI 2.5-30.2), 7.1 (95 % CI 2.1-23.5) and 8.2 (95 % CI 2.0-34.3) times increase in odds of neural tube/central nervous system, renal, limb and Lip/Cleft/Palate malformations. CONCLUSIONS: This study shows that chronic hypertension in the maternal period exposes newborns to a significant risk of developing renal, limb and lip/cleft/palate congenital malformations, and the risk is further exacerbate by superimposing eclampsia. Additional research is needed to identify shared pathways of maternal hypertensive disorders and congenital malformations.

Caesarean section surgical techniques (CORONIS): a fractional, factorial, unmasked, randomised controlled trial.

BACKGROUND: Variations exist in the surgical techniques used for caesarean section and many have not been rigorously assessed in randomised controlled trials. We aimed to assess whether any surgical techniques were associated with improved outcomes for women and babies. METHODS: CORONIS was a pragmatic international 2×2×2×2×2 non-regular fractional, factorial, unmasked, randomised controlled trial that examined five elements of the caesarean section technique in intervention pairs. CORONIS was undertaken at 19 sites in Argentina, Chile, Ghana, India, Kenya, Pakistan, and Sudan. Each site was assigned to three of the five intervention pairs: blunt versus sharp abdominal entry; exteriorisation of the uterus for repair versus intra-abdominal repair; single-layer versus double-layer closure of the uterus; closure versus non-closure of the peritoneum (pelvic and parietal); and chromic catgut versus polyglactin-910 for uterine repair. Pregnant women were eligible if they were to undergo their first or second caesarean section through a planned transverse abdominal incision. Women were randomly assigned by a secure web-based number allocation system to one intervention from each of the three assigned pairs. All investigators, surgeons, and participants were unmasked to treatment allocation. The primary outcome was the composite of death, maternal infectious morbidity, further operative procedures, or blood transfusion (>1 unit) up to the 6-week follow-up visit. Women were analysed in the groups into which they were allocated. The CORONIS Trial is registered with Current Controlled Trials: ISRCTN31089967. FINDINGS: Between May 20, 2007, and Dec 31, 2010, 15 935 women were recruited. There were no statistically significant differences within any of the intervention pairs for the primary outcome: blunt versus sharp entry risk ratio 1·03 (95% CI 0·91-1·17), exterior versus intra-abdominal repair 0·96 (0·84-1·08), single-layer versus double-layer closure 0·96 (0·85-1·08), closure versus non-closure 1·06 (0·94-1·20), and chromic catgut versus polyglactin-910 0·90 (0·78-1·04). 144 serious adverse events were reported, of which 26 were possibly related to the intervention. Most of the reported serious adverse events were known complications of surgery or complications of the reasons for the caesarean section. INTERPRETATION: These findings suggest that any of these surgical techniques is acceptable. However, longer-term follow-up is needed to assess whether the absence of evidence of short-term effects will translate into an absence of long-term effects. FUNDING: UK Medical Research Council and WHO.

Pre-eclampsia, eclampsia and adverse maternal and perinatal outcomes: a secondary analysis of the World Health Organization Multicountry Survey on Maternal and Newborn Health.

OBJECTIVE: To assess the incidence of hypertensive disorders of pregnancy and related severe complications, identify other associated factors and compare maternal and perinatal outcomes in women with and without these conditions. DESIGN: Secondary analysis of the World Health Organization Multicountry Survey on Maternal and Newborn Health (WHOMCS) database. SETTING: Cross-sectional study implemented at 357 health facilities conducting 1000 or more deliveries annually in 29 countries from Africa, Asia, Latin America and the Middle East. POPULATION: All women suffering from any hypertensive disorder during pregnancy, the intrapartum or early postpartum period in the participating hospitals during the study period. METHODS: We calculated the proportion of the pre-specified outcomes in the study population and their distribution according to hypertensive disorders' severity. We estimated the association between them and maternal deaths, near-miss cases, and severe maternal complications using a multilevel logit model. MAIN OUTCOME MEASURES: Hypertensive disorders of pregnancy. Potentially life-threatening conditions among maternal near-miss cases, maternal deaths and cases without severe maternal outcomes. RESULTS: Overall, 8542 (2.73%) women suffered from hypertensive disorders. Incidences of pre-eclampsia, eclampsia and chronic hypertension were 2.16%, 0.28% and 0.29%, respectively. Maternal near-miss cases were eight times more frequent in women with pre-eclampsia, and increased to up to 60 times more frequent in women with eclampsia, when compared with women without these conditions. CONCLUSIONS: The analysis of this large database provides estimates of the global distribution of the incidence of hypertensive disorders of pregnancy. The information on the most frequent complications related to pre-eclampsia and eclampsia could be of interest to inform policies for health systems organisation.

Antihypertensive drug therapy for mild to moderate hypertension during pregnancy.

BACKGROUND: Mild to moderate hypertension during pregnancy is common. Antihypertensive drugs are often used in the belief that lowering blood pressure will prevent progression to more severe disease, and thereby improve outcome. OBJECTIVES: To assess the effects of antihypertensive drug treatments for women with mild to moderate hypertension during pregnancy. SEARCH STRATEGY: We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (March 2006), the Cochrane Central Register of Controlled Trials (The Cochrane Library 2005, Issue 3), MEDLINE (1966 to November 2005), LILACS (1984 to November 2005) and EMBASE (1974 to November 2005). SELECTION CRITERIA: All randomised trials evaluating any antihypertensive drug treatment for mild to moderate hypertension during pregnancy defined, whenever possible, as systolic blood pressure 140 to 169 mmHg and diastolic blood pressure 90 to 109 mmHg. Comparisons were of one or more antihypertensive drug(s) with placebo, with no antihypertensive drug, or with another antihypertensive drug, and where treatment was planned to continue for at least seven days. DATA COLLECTION AND ANALYSIS: Two review authors independently extracted data. MAIN RESULTS: Forty-six trials (4282 women) were included. Twenty-eight trials compared an antihypertensive drug with placebo/no antihypertensive drug (3200 women). There is a halving in the risk of developing severe hypertension associated with the use of antihypertensive drug(s) (19 trials, 2409 women; relative risk (RR) 0.50; 95% confidence interval (CI) 0.41 to 0.61; risk difference (RD) -0.10 (-0.12 to -0.07); number needed to treat (NNT) 10 (8 to 13)) but little evidence of a difference in the risk of pre-eclampsia (22 trials, 2702 women; RR 0.97; 95% CI 0.83 to 1.13). Similarly, there is no clear effect on the risk of the baby dying (26 trials, 3081 women; RR 0.73; 95% CI 0.50 to 1.08), preterm birth (14 trials, 1992 women; RR 1.02; 95 % CI 0.89 to 1.16), or small-for-gestational-age babies (19 trials, 2437 women; RR 1.04; 95 % CI 0.84 to 1.27). There were no clear differences in any other outcomes. Nineteen trials (1282 women) compared one antihypertensive drug with another. Beta blockers seem better than methyldopa for reducing the risk of severe hypertension (10 trials, 539 women, RR 0.75 (95 % CI 0.59 to 0.94); RD -0.08 (-0.14 to 0.02); NNT 12 (6 to 275)). There is no clear difference between any of the alternative drugs in the risk of developing proteinuria/pre-eclampsia. Other outcomes were only reported by a small proportion of studies, and there were no clear differences. AUTHORS' CONCLUSIONS: It remains unclear whether antihypertensive drug therapy for mild to moderate hypertension during pregnancy is worthwhile.

Conservative management of symptomatic and/or complicated haemorrhoids in pregnancy and the puerperium.

BACKGROUND: Haemorrhoids (piles) are swollen veins at or near the anus, normally asymptomatic. They do not constitute a disease, unless they become symptomatic. Pregnancy and the puerperium predispose to symptomatic haemorrhoids, being the most common ano-rectal disease at these stages. Symptoms are usually mild and transient and include intermittent bleeding from the anus and pain. Depending on the degree of pain, quality of life could be affected, varying from mild discomfort to real difficulty in dealing with the activities of everyday life. Treatment during pregnancy is mainly directed to the relief of symptoms, especially pain control. The so-called conservative management includes dietary modifications, stimulants or depressants of the bowel transit, local treatment, and phlebotonics (drugs that cause decreased capillary fragility, improving the microcirculation in venous insufficiency). For many women, symptoms will resolve spontaneously soon after birth, and so any corrective treatment is usually deferred to some time after birth. Thus, the objective of this review is to evaluate the efficacy of conservative management of piles during pregnancy and the puerperium. OBJECTIVES: To determine the possible benefits, risks and side-effects of the conservative management of symptomatic haemorrhoids during pregnancy and the puerperium. SEARCH STRATEGY: We searched the Cochrane Pregnancy and Childbirth Group Trials Register (30 June 2004). SELECTION CRITERIA: Randomised-controlled trials comparing any of the conservative treatments for symptomatic haemorrhoids during pregnancy and the puerperium (such as dietary modifications, stimulant/depressant of the bowel transit, local treatments, drugs that improve the microcirculation in venous insufficiency) with a placebo or no treatment. DATA COLLECTION AND ANALYSIS: Two review authors independently performed a methodological assessment for deciding which studies to include/exclude from the review and extracted data. MAIN RESULTS: From 10 potentially eligible studies, two were included in this review (150 women). Both compared oral rutosides against placebo. Rutosides seem to be effective in reducing the signs identified by the healthcare provider, and symptoms and signs reported by women, of haemorrhoidal disease. For the outcome no response to treatment: relative risk 0.07, 95% confidence interval 0.03 to 0.20. Regarding perinatal outcomes, one fetal death and one congenital malformation (possible not related to exposure) were reported in the control and treatment group respectively. AUTHORS' CONCLUSIONS: Although the treatment with oral hydroxyethylrutosides looks promising for symptom relief in first and second degree haemorrhoids, its use cannot be recommended until new evidence reassures women and their clinicians about their safety. The most commonly used approaches, such as dietary modifications and local treatments, were not properly evaluated during pregnancy and the puerperium.

Bed rest with or without hospitalisation for hypertension during pregnancy.

BACKGROUND: Bed rest or restriction of activity, with or without hospitalisation, have been advocated for women with hypertension during pregnancy to improve pregnancy outcome. However, benefits need to be demonstrated before such interventions can be recommended since restricted activity may be disruptive to women's lives, expensive, and increase the risk of thromboembolism. OBJECTIVES: To assess the effects on the mother and the baby of different degrees of bed rest, compared with each other, and with routine activity, in hospital or at home, for primary treatment of hypertension during pregnancy. SEARCH STRATEGY: We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (April 2005), the Cochrane Central Register of Controlled Trials (The Cochrane Library, Issue 1, 2005), and EMBASE (January 2002 to December 2004). SELECTION CRITERIA: Randomised trials evaluating bed rest for women with hypertension in pregnancy were selected. DATA COLLECTION AND ANALYSIS: Two review authors assessed trials for inclusion independently, and extracted data. Data were entered into RevMan software and double-checked. MAIN RESULTS: Four small trials (449 women) were included. Three were of good quality. Two trials (145 women) compared strict bed rest with some rest, in hospital, for women with proteinuric hypertension. There was insufficient evidence to demonstrate any differences between the groups for reported outcomes. Two trials (304 women) compared some bed rest in hospital with routine activity at home for non-proteinuric hypertension. There was reduced risk of severe hypertension (1 trial, 218 women; RR 0.58, 95% CI 0.38 to 0.89) and a borderline reduction in risk of preterm birth (1 trial, 218 women; RR 0.53, CI 0.29 to 0.99) with some rest compared to normal activity. More women in the bed rest group opted not to have the same management in future pregnancies, if the choice were given (1 trial, 86 women; RR 3.00, 95% CI 1.43 to 6.31). There were no significant differences for any other outcomes. AUTHORS' CONCLUSIONS: Few randomised trials have evaluated rest for women with hypertension during pregnancy, and important information on side-effects and cost implication is missing from available trials. Although one small trial suggests that some bed rest may be associated with reduced risk of severe hypertension and preterm birth, these findings need to be confirmed in larger trials. At present, there is insufficient evidence to provide clear guidance for clinical practice. Therefore, bed rest should not be recommended routinely for hypertension in pregnancy, especially since more women appear to prefer unrestricted activity, if the choice were given.

Antihypertensive drug therapy for mild to moderate hypertension during pregnancy.

BACKGROUND: Mild-moderate hypertension during pregnancy is common. Antihypertensive drugs are often used in the belief that lowering blood pressure will prevent progression to more severe disease, and thereby improve outcome. OBJECTIVES: To assess the effects of antihypertensive drug treatments for women with mild to moderate hypertension during pregnancy. SEARCH STRATEGY: Relevant trials were identified in the register of trials maintained by the Cochrane Pregnancy and Childbirth Group. In addition, the Cochrane Controlled Trial Register, MEDLINE, and EMBASE were searched. Date of last search: October 2000. SELECTION CRITERIA: All randomised trials evaluating any antihypertensive drug treatment for mild to moderate hypertension during pregnancy, defined whenever possible as systolic blood pressure 140-169 mmHg and diastolic blood pressure 90-109 mmHg. Comparisons were of one or more antihypertensive drug(s) with placebo, with no antihypertensive drug, or with another antihypertensive drug, and where treatment was planned to continue for at least seven days. DATA COLLECTION AND ANALYSIS: Data were extracted independently by two reviewers. MAIN RESULTS: Overall, this review includes 40 studies (3797 women), 24 of which compared an antihypertensive drug with placebo/no antihypertensive drug (2815 women). There is a halving in the risk of developing severe hypertension associated with the use of antihypertensive drug(s) [17 trials, 2155 women; relative risk (RR) 0.52 (95% confidence interval (CI) 0.41 to 0.64); risk difference (RD) -0.09 (-0.12 to -0.06); number needed to treat (NNT) 12 (9 to 17)] but little evidence of a difference in the risk of pre-eclampsia [19 trials, 2402 women; RR 0.99 (0.84 to 1.18)]. Similarly, there is no clear effect on the risk of the baby dying [23 trials, 2727 women; RR 0.71(0.46 to 1.09)], preterm birth [12 trials, 1738 women; RR 0.98 (0.85 to 1.13)], or small for gestational age babies [17 trials, 2159 women; RR 1.13 (0.91 to 1.42)]. There were no clear differences in any other outcomes. Seventeen trials (1182 women) compared one antihypertensive drug with another. There is no clear difference between any of these drugs in the risk of developing severe hypertension, and proteinuria/pre-eclampsia. Other antihypertensive agents seem better than methyldopa for reducing the risk of the baby dying [14 trials, 1010 subjects, RR 0.49 (0.24 to 0.99); RD -0.02 (-0.04 to 0.00); NNT 45 (22 to 1341)]. Other outcomes were only reported by a small proportion of studies, and there were no clear differences. REVIEWER'S CONCLUSIONS: It remains unclear whether antihypertensive drug therapy for mild-moderate hypertension during pregnancy is worthwhile.