Hormone replacement therapy for preventing cardiovascular disease in post‐menopausal women
Abstract
This is a protocol for a Cochrane Review (Intervention). The objectives are as follows:
To determine the efficacy, effectiveness and safety of the HRT for the primary and secondary prevention of cardiovascular diseases in post‐menopausal women.
Background
There is considerable interest in the use of Hormone Replacement Therapy (HRT) in post‐menopausal women and the number of related studies has increased in recent years. The potential benefits claimed for HRT include a wide variety of diseases in women, especially coronary artery disease. Although the bulk of evidence from observational studies suggests that use of exogenous oestrogens is cardio‐protective for post‐menopausal women (Grodstein 1996), the efficacy and effectiveness of HRT in the prevention of cardiovascular diseases is not yet clear (Hulley 1998). Evidence from these observational studies is difficult to interpret because the studies vary in the hormones used, the dose and mode of administration. In addition generalising the findings from observational studies of women at specific ages to all post‐menopausal women is difficult.
The relevance of using data gathered from these observational studies of HRT to predict a reduction in cardiovascular risk is questionable without other supporting evidence. A systematic review would be a reasonable approach contributing to the clarification of the existing uncertainty on this topic.
Objectives
To determine the efficacy, effectiveness and safety of the HRT for the primary and secondary prevention of cardiovascular diseases in post‐menopausal women.
Methods
Criteria for considering studies for this review
Types of studies
Randomised controlled trials comparing HRT with controls (placebo or no treatment) in post‐menopausal women. The studies will have duration of follow up of 6 months or longer. We will also examine separately large prospective cohort studies of at least 1000 women followed up for a minimum of one year.
Types of participants
Post‐menopausal women with or without evidence of cardiovascular disease at baseline.
Types of interventions
1. Hormone Replacement Therapy (HRT), either with oestrogens alone or in different combinations with progestogens will be considered.
2. Placebo (control)
3. No treatment (open control)
Types of outcome measures
· All causes of death
· Major cardiovascular events (angina, acute myocardial infarction and stroke)
· Cardiovascular deaths
· Quality of life measures
. Revascularization procedures (coronary by‐pass, angioplasty, coronary stent)
. Combined vascular events outcome (non‐fatal and fatal CV outcomes + revascularizations)
. Other Cardiovascular related outcomes (venous thrombo‐embolism, hospitalization of coronary heart disease, peripheral arterial disease)
. Intermediate cardiovascular outcomes (blood pressure, blood cholesterol, and coagulation factors)
. Side effects and tolerability of treatment
Search methods for identification of studies
1.1MEDLINE
Electronic search from 1980 to 1999 using the Collaborative Review Group
Search Strategy for cardiovascular diseases (ref ):
#1= HRT
#2= "Oestrogen‐Replacement‐Therapy"/ all subheadings
#3= hormone near replacement
#4= oestrogen near replacement
#5= #1 or #2 or #3 or #4
#6= menopaus*
#7= #5 and #6
#8= "Cardiovascular‐Diseases"/ all subheadings
#9= cardiovascular
#10= stroke
#11= coronary
#12= angina
#13= ischaemic
#14= #8 or #9 or #10 or #11 or #12 or #13
#15= #7 and #14
#16= RANDOMIZED‐CONTROLLED‐TRIAL in PT
#17= CONTROLLED‐CLINICAL‐TRIAL in PT
#18= RANDOMIZED‐CONTROLLED‐TRIALS
#19= RANDOM‐ALLOCATION
#20= DOUBLE‐BLIND‐METHOD
#21= SINGLE‐BLIND‐METHOD
#22= #16 or #17 or #18 or #19 or #20 or #21
#23= TG = ANIMAL
#24= TG = HUMAN
#25= TG = ANIMAL
#26= (TG = ANIMAL) not ((TG = HUMAN) and (TG = ANIMAL))
#27= #22 not #26
#28= CLINICAL‐TRIAL in PT
#29= explode CLINICAL‐TRIALS/ all subheadings
#30= CLIN*
#31= TRIAL*
#32= (CLIN* near TRIAL*) in TI
#33= CLIN*
#34= TRIAL*
#35= (CLIN* near TRIAL*) in AB
#36= SINGL*
#37= DOUBL*
#38= TREBL*
#39= TRIPL*
#40= BLIND*
#41= MASK*
#42= (SINGL* or DOUBL* or TREBL* or TRIPL*) near (BLIND* or MASK*)
#43= (#42 in TI) or (#42 in AB)
#44= PLACEBOS
#45= PLACEBO*
#46= PLACEBO* in TI
#47= PLACEBO*
#48= PLACEBO* in AB
#49= RANDOM*
#50= RANDOM* in TI
#51= RANDOM*
#52= RANDOM* in AB
#53= RESEARCH‐DESIGN
#54= #28 or #29 or #32 or #35 or #42 or #43 or #44 or #46 or #48 or #50 or #52
or #53
#55= #54 not #26
#56= #55 not #27
#57= TG = COMPARATIVE‐STUDY
#58= explode EVALUATION‐STUDIES/ all subheadings
#59= FOLLOW‐UP‐STUDIES
#60= PROSPECTIVE‐STUDIES
#61= CONTROL*
#62= PROSPECTIV*
#63= VOLUNTEER*
#64= CONTROL* or PROSPECTIV* or VOLUNTEER*
#65= (#64 in TI) or (#64 in AB)
#66= #57 or #58 or #59 or #60 or #64 or #65
#67= #66 not #26
#68= #67 not (#27 or #56)
#69= #27 or #56 or #67
#70= #15 and #69
In order to improve the low specificity of Medline, we will review other data bases such as EMBASE using the similar appropriate search strategy. In addition, other data bases will be review:
1.2 THE COCHRANE CONTROLLED TRIALS REGISTER
Data base included in The Cochrane Library (2). We will use the following
search strategy
#1 HRT
#2 Oestrogen‐Replacement‐Therapy*: me
#3 (hormone near replacement)
#4 (oestrogen near replacement)
#5 (((#1 or #2) or #3) or #4)
#6 menopaus*
#7 (#5 and #6)
#8 cardiovascular‐diseases*: me
#9 cardiovascul*
#10 #8 or #9
#11 #7 and #10
The same search strategy than for the CCTR will be used for:
1.3 The data base of Spanish Clinical Trials.
1.4 The McMaster Cardiovascular Clinical Trials Register
1.5 Reference list of papers resulting from this search
1.6 Industry would be contacted to identify unpublished trials and trials in progress
Data collection and analysis
All the articles identified with the above search strategy will be considered as candidates for the review, and registered within the bibliographic data‐ base in ProCite. Such a program will permit the capture the information directly from different electronic data‐ bases, as well as the consultation of summaries, and the addition of notes and commentaries.
Photocopies of selected articles will be obtained from appropriate libraries. Next, we will identify the libraries where to find out the selected articles, and we will seek for a photocopy. Also we will review the abstract books of the eight International Congresses on the Menopause that have been held to date.
INCLUSION OF ARTICLES
Inclusion Criteria:
Only those articles meeting the following criteria will be considered:
1a. clinical trials according to the Cochrane Collaboration criteria, i.e. studies using random allocation methods (explicit or implicit) of individuals (human beings) in 2 or more groups of treatment (including the control group or placebo group).
1b. prospective cohort studies. Only studies with at least 1000 post‐menopausal women at baseline with a minimum of 1 year follow up will be included. In addition, losses to follow up of less than 20 percent will be required. Use of hormone replacement therapy may be determined by self reports or by review of medical records.
2. hormonal replacement treatments administered alone or in combination.
3. inclusion of cardiovascular outcomes of relevance for post‐menopausal women.
Exclusion Criteria:
Therefore, according to the inclusion criteria mentioned above, trials with only surrogate end‐points (i.e. electrocardiographic changes, symptoms, blood pressure, or biochemical changes) will be excluded. Also, those trials including relevant cardiovascular outcomes but of inadequate format (i.e. showing medians or means estimators but not the number of women with such outcome) will be sought from the authors before to be excluded.
Data Extraction:
A specific questionnaire for data extraction will be designed. All the eligible articles will be presented to the reviewers in a blinded fashion, with no information about authors, centres or journals of publication. Two independent reviewers (RGS and LSG) will extract simultaneously the information from the articles. Discrepancies in the results will be solved by consensus.
Statistical Analysis:
For dichotomous outcomes, odds ratios and 95% confidence intervals (CI) will be calculated for each study. For continuous variables (e.g. health‐related quality of life), a weighted mean difference (WMD) or standardised mean difference (SMD) and 95% CI will be calculated for each study. The results of each RCT will be plotted as point estimates with corresponding 95% confidence intervals. For each outcome, a test of heterogeneity will be carried out. In the situation of no heterogeneity, fixed effects meta‐analysis will be performed. If substantial heterogeneity (p<0.1) is detected, the reviewers will look for possible explanations (e.g. population and intervention) for this. If the heterogeneity can be explained, the reviewers will consider (with due caution) the option of presenting the results as sub‐group analyses. If the heterogeneity cannot be explained, the reviewers
will consider the following options: not aggregate the study at all, use a fixed effects model with appropriate cautious interpretation, random effects model with appropriate cautious interpretation, use both fixed and random effects models as an additional aid in explaining the uncertainty around an analysis with heterogeneous studies. Sensitivity analysis will be carried out, excluding studies of low methodological quality.
