Menopause treated with homeopathy - a systematic review

Das Thema oder die Frage, ob Homöopathie eine wissenschaftliche Therapiemethode ist oder nicht, rückt aktuell (März 2019) mehr und mehr in den Vordergrund. Eine weitere Frage, die sich jedoch stellt, ist, ob die individualisierte Homöopathie mit Einzelmitteln eine Heilmethode ist, die mit den von der sogenannten Evidenzbasierten Medizin (EBM) festgelegten Kriterien überhaupt wissenschaftlich "gemessen" werden kann. Diese Frage habe ich unter anderem 2014/2015 in meiner Abschlussarbeit für den Kurs MSc Integrated Healthcare an der University of Central Lancashire, United Kingdom, untersucht.

Nachfolgend meine Systematische Studienübersicht zum Thema "Die homöopathische Behandlung der Menopause." Meine persönliche Tutorin war Dr. Kate Chatfield, der ich hiermit wiederholt meinen persönlichen Dank ausspreche! Mein persönlicher Dank gilt auch meinem Ehemann Ralf, der mich bei diesem Projekt (und auch bei vielen anderen) konstruktiv unterstützt hat. 

 

 

Menopause treated with homeopathy: A systematic review

 

Thesis for the course MSc Integrated Healthcare,

University of Central Lancashire,

Module NU4053

                                  

 

            written by Monika Mueller-Amenitsch, G 20583257

 

Date of submission: 19.04.2015

Word account: 14,490

 

 

Supervisor: Kate Chatfield

External Examiner: Dr Amanda Bingley PhD

 

 


Acknowledgements:

 

I would like to thank my husband who has supported and encouraged me during the challenging conduct of this thesis and throughout the whole MSc course at Uclan. Without him I would not have started this course!

Furthermore, I would like to thank Jean Duckworth, Kate Chatfield, Hazel Partington, Graeme Tobyn and Alison Denham for their academic support, empathy and patience with which they have taught the MSc course during the last 3 years!

I very specially thank Kate Chatfield for the supervision of this thesis. Without her encouragement I would not have dared to write a systematic review following PRISMA guidelines and registering in PROSPERO. Thank you for this wonderful cooperation!

I also thank Gerda Zaalberg for her translation of a study from Dutch into English and the Carstens Stiftung for supplying me with studies I could not access otherwise.

I would like to thank Inge Stadler, Angelika Schumann and Karin Enders, my fellow students, for their friendship during the last 3 years at Uclan.

 

                        Thank you so much!       Monika

 

 

Content

INTRODUCTION.. 1

Background menopause. 1

Menopause-associated symptoms. 3

Hormone therapy. 5

Background homeopathy. 7

The purpose of systematic reviews. 8

Conducting systematic reviews in homeopathy. 11

Motivation for conducting this systematic review.. 13

Different types of systematic reviews. 14

 

MENOPAUSE TREATED WITH HOMEOPATHY: A SYSTEMATIC REVIEW (following PRISMA guidelines) 18

Abstract 18

2. Methods. 23

2.1. Inclusion criteria. 23

2.2. Search strategy. 24

2.4. Quality assessment 26

2.5. Principal summary measures. 26

3. Results. 27

3.1. Characteristics of included studies. 31

3.2. Risk of bias in included studies. 33

3.3. Results of individual studies. 34

3.4. Adverse events. 39

3.5. Synthesis of results. 39

3.5.1. Complex homeopathy. 40

3.5.2. Individualised homeopathy. 42

3.6. Risk of bias across studies. 44

3.7. Flaws across studies. 45

4. Discussion. 47

4.1. Recommendations. 52

4.2. Conclusions. 53

4.3. Limitations. 53

4.4. Funding sources. 53

4.5. Conflicts of interest 54

 

REFLECTION ON THE PROCESS OF CONDUCTING A SYSTEMATIC REVIEW    54

 

 

 

 

 

Figures and tables

Figure 1: PRISMA 2009 Flow Diagram showing the results of the literature search………………………………………………………………………………30

Table 3: Results of included studies…………………………………………….35

 

 

 

 

Appendices

APPENDIX 1, Table 1: Characteristics of included studies…………………..67

APPENDIX 2, Table 2: The Cochrane Collaboration’s tool for assessing risk of bias………………………………………………………………………………71

 

 

 

 

 

 

Abbreviations:                                                                                           

 

ADL – Activities of Daily Living                                                                    

BDI – Beck Depression Inventory

BMI – Blatt Menopausal Index                                                                   

CAM – Complementary and Alternative Medicine

CDSR – Cochrane Database of Systematic Reviews

CI – Confidence Interval

CRD – Centre for Reviews and Dissemination

EORTC QLQ C30 – European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire

EBM – Evidence Based Medicine

FAQ – Final Assessment Questionnaire

FSH – Follicle Stimulating Hormone

GH – General Health

GHHOS – Glasgow Homeopathic Hospital Outcome Scale

Ho – Null hypothesis

HADS – Hospital Anxiety and Depression Scale

HFRDIS = Hot Flash Related Daily Interference Scale

HFS – Hot Flash Score

HFSS – Hot Flash Severity Score

HRSD – Hamilton Rating Scale for Depression

HT – Hormone Therapy

IH – Individualised Homeopathy

KMI – Kupperman Menopausal Index

LDL – Low Density Lipoprotein

LH – Luteinising Hormone

MRS (II) – Menopause Rating Scale (II)

MSQ – Menopausal Symptom Questionnaire

MYMOP – Measure Yourself Medical Outcome Profile

NHS – National Health Service

PRISMA – Preferred Reporting Items for Systematic Reviews and Meta-Analyses

QoL – Quality of Life

QUOROM – Quality of Reporting of Meta-Analyses

RCT – Randomised Controlled Trial

SD – Standard Deviation

SF-36 QoL – SF-36 Quality of Life Score

SSRI – Selective Serotonin Reuptake Inhibitor

UK – United Kingdom

WA – Washington


INTRODUCTION

 

The number of aging women is rapidly growing worldwide (Wilken-Jensen & Ottesen, 2003). With increasing life span, women spend nowadays one-third of their lifetime under menopause (Moreira et al., 2014). This implies that a change of the medical approach may be necessary as demand for the treatment of menopause-associated symptoms increases (Wilken-Jensen & Ottesen, 2003). Menopausal women belong, for several reasons, to one of the biggest patient groups using alternative and complementary medicine (CAM) (Wilken-Jensen & Ottesen, 2003). A pilot data collection study within the five National Health Service (NHS) Homeopathic Hospitals in the United Kingdom (UK) showed that menopausal symptoms belonged to the most commonly treated complaints (Thompson et al., 2008). The growing number of postmenopausal women and their great interest in using CAM, especially homeopathy, were important motivating factors in the decision to undertake a systematic review on menopause treated with homeopathy. This topic is timely and highly reflective of current need. 

 

 

Background menopause

 

Menopause is defined as the time of the final menstrual period, one which is not followed by any further ovary-controlled menstruation for 12 months, and can only be identified retrospectively (De Gruyter, 1994)). Post-menopause follows menopause and lasts approximately 10 years (Groth & Stolzenberg, 1995).

Menopause is characterised by amenorrhoea and not necessarily by any other symptoms (Utian, 1994) and generally occurs between the ages 48 – 51 years (De Gruyter, 1994). Climacteric, also known as peri-menopause, is the time in the life of a woman when changing from a productive phase into an infertile one (Nayak et al., 2011) and this results from the cessation of the cyclical function of the ovaries (De Gruyter, 1994). During this time the ovaries’ reaction to the Follicle Stimulating Hormone (FSH) and Luteinising Hormone (LH) that are produced in the pituitary gland diminishes gradually and eventually ceases altogether. The function of FSH is to stimulate the growth of follicles in the ovaries and the follicles produce oestrogen (Kiener, 2000). The result of the decreased reaction to FSH and FH are a smaller number of ovulations, irregularities of the menstrual cycle and a fall of the ovaries’ production of progesterone and oestrogen. At the same time the concentration of FSH and LH in the blood increases since no positive oestrogen feedback from the ovaries is taking place (Kiener, 2000). In addition to this natural menopausal process involving a deficiency of oestrogen, women with breast cancer can also enter into an ‘artificial’ climacteric phase as a result of the treatment they receive for the cancer. Breast cancer is one of the most common cancers worldwide (Rada et al., 2010). Since oestrogen may promote the growth of oestrogen receptor- positive breast cancers, patients are commonly treated with endocrine medicaments having an anti-oestrogen effect such as tamoxifen and aromatase inhibitors (Rada et al., 2010) such as Anastrozole (Jones & Buzdar, 2004). Breast cancer patients may also be treated with chemotherapy which commonly impairs the natural function of the ovaries (Rada et al., 2010). Consequently, reasons for lack of oestrogen in breast cancer patients can be follicle cell death and ovarian ablation induced by chemotherapy, medicaments withdrawing oestrogen or the natural process of menopause itself (Thompson & Reilly, 2003).

After hysterectomy with bilateral oophorectomy, or bilateral oophorectomy alone, in women of child-bearing age menopause starts immediately and is called ‘surgical menopause’ (Northrup, 1995). 

 

Menopause-associated symptoms

 

The timing of the menopause differs between women and not every woman develops impairing symptoms during menopause (Katz, 1997).  According to Kiener (2000) the primary menopause-related symptoms include the following:

-       Vasomotor symptoms such as hot flushes and night sweating, experienced by 75 % of menopausal women.

-       Psychological and emotional symptoms such as fatigue, mood swings, crying, anxiety, irritability, sleeping difficulties, lack of concentration, depression and nervousness.

-       In the urogenital region the mucous membrane of the vagina and vulva is getting thinner. This may lead to vaginal dryness and vaginitis which may cause increased urination and dyspareunia. Some women experience incontinence, cystitis and low libido.

-       Bone mass/density decreases and approximately 25% of postmenopausal women suffer from severe osteoporosis.

-       Various other symptoms such as dizziness, palpitations, paraesthesia, tachycardia, diarrhoea, constipation, arthralgia and weight gain may arise.

 

After the menopause women suffer more often from cardiovascular diseases such as apoplexy. Further typical menopause-associated symptoms include menstrual disturbances such as irregular and/or heavy bleeding (Katz, 1997). Penny, Razlog, Deroukakis, and Johnston (2009) complete this symptom list with loss of elasticity in the skin, mild hirsutism, myalgia, cold extremities and formication.

Nencioni et al. (1999) found, in their cross-sectional analysis of 9309 women, that from pre- to post-menopause significant increases in total cholesterol, low density lipoprotein (LDL) and triglycerides take place. Increased lipid levels have been linked with a higher risk of cardiovascular disease (Rousseau, 2001).

Breast-cancer patients being treated with anti-oestrogens such as tamoxifen and Anastrozole are known to develop more intense menopausal symptoms such as hot flushes and sweat due to oestrogen deficiency (Jones & Buzdar, 2004). However Gupta et al. (2006) conducted a cross-sectional survey in Solihull Hospital in the West Midlands, UK, from December 2003 – May 2004, to identify symptoms due to oestrogen deficiency in women who had been treated for breast cancer within the last 5 years. They found no association between tamoxifen use and the intensity of hot flushes, sweats, heart complaints or sleeping difficulties. However, an important finding from this survey was that women taking antidepressant treatment developed menopausal symptoms, including hot flushes and sweats, more often and more severely.

Subsequently, the issue of whether women who are receiving breast cancer treatment suffer from more intense vasomotor symptoms than those women who undergo a natural menopause remains a matter of debate. It is, however, clear that they suffer from menopausal symptoms like sudden heat in the face, neck and chest and that hot flushes are typical (Rada at al., 2010).

 

Hormone therapy

 

In the 1960s the treatment of all menopausal symptoms with oestrogen was considered as effective and imperative (Wilson & Wilson, 1963). Also, in the following years, hormone therapy (HT) was regarded as beneficial in the treatment of menopausal symptoms, improving quality of life, preventing osteoporosis and cardiovascular disease (Speroff et al., 1999). More recently Bordet, Colas, Marijnen, Masson, and Trichard (2008) describe how in this day and age, hot flushes are the main reason for menopausal women to take HT.

On the other hand, in the middle 1970s, oestrogen was found to increase the risk of endometrial cancer (Smith, Prentice, Thompson, & Herrmann, 1975). For this reason, from 1976 onwards, oestrogen was administered in combination with progesterone to protect against endometrial cancer (Rousseau, 2001). Later it was discovered that HT with oestrogen alone, or oestrogen combined with progesterone, given to menopausal women in their 50s and 60s, increased breast density and the risk of developing breast cancer. Furthermore, the combined use of HT raised the risk of breast cancer to a higher degree than oestrogen as single medication (Dixon, 2003).

Women also reported other, less severe, side-effects from HT like breast enlargement and tenderness (Thompson & Relton, 2009) and Wilken-Jensen and Ottesen (2003) described the adverse effects of increased risk of venous thromboembolism and vaginal bleeding.

Because of the link between breast cancer and oestrogen, as well as oestrogen with added progesterone, breast cancer survivors are advised not to take HT to avoid recurrences of the cancer (Rousseau, 2001). Moreover, due to the potential for side effects, the use of HT is not recommended for all women and remains a controversial topic. Katz (1997) describes how half of studied women being prescribed HT discontinued within 6 months. Many women decide against HT for fear of cancer, undesirable side effects and also, for some, the opinion that menopause is a natural process (Wilken-Jensen & Ottesen, 2003). In a survey conducted between July 1993 and June 1995, Reynolds, Obermeyer, Walker, and Guilbert (2002), found that one of the main reasons for stopping HT under post-menopause was the fear of weight gain.

The contra-indication of HT in breast cancer survivors, and the various potential side-effects of HT, are plausible reasons for women to seek alternative treatment modalities when suffering from menopause-related ailments.

 

Background homeopathy

 

The medical system of homeopathy was founded by Samuel Hahnemann (1755-1843), a German medical doctor, and, the name he awarded this system, ‘homeopathy’ is derived from the Greek expressions ‘homoios’ and ‘pathos’ which translate as similar and suffering. Suffering similarly captures the fundamental principle of homeopathy, that likes can be treated with likes. Hahnemann was not the first to speak of this idea, the notion of similars had previously also been described by Hippocrates and Paracelsus. However, it was Hahnemann who introduced the use of similars as an essential part of a medical science (Ullman, 2014). There are three primary principles that underpin the practise of homeopathy and set it apart from other forms of medicine. According to Ullman (1997) the principle of similars means that a substance that can provoke symptoms in a healthy person can be used (in small doses) to cure the same symptoms in an ill patient. The second principle of homeopathy that remedies are prescribed for patterns of symptoms (totality) and not simply for isolated symptoms. As a result a homeopath administers remedies that are individually adapted for the overall condition of a sick person. This means that two persons with the same diagnosis may be given different homeopathic remedies since prescriptions are based upon the entire symptom picture and not the medical diagnosis. The third homeopathic principle is that of potentisation; homeopathic remedies are made from substances that are consecutively diluted and potentised by succussion. Through this process the original substance diminishes gradually and the more a homeopathic remedy is diluted and succussed, the more potent it becomes (Ullman, 1997). The paradox of potentisation has generated much scepticism and debate about homeopathy (Cooper & Relton, 2010). On the other hand homeopathic remedies are not known to produce harmful side-effects, since they contain extremely low levels of the material substance (Katz, 1997). During consultations homeopaths focus on the patient’s individual symptoms to prescribe a similar remedy. Women seeking help for menopause-related symptoms will be prescribed a homeopathic remedy adapted to their personal complaints, perhaps fitting, since each woman experiences menopause differently (Katz, 1997). The absence of side-effects and the individualised medication are attractive notions and hence it is not difficult to understand why some menopausal women seek homeopathic help. In a recent retrospective survey of 563 menopausal women with vasomotor symptoms in the USA, who had stopped HT, nearly half were using CAM and homeopathy was among the most frequently used (Kupferer, Dormire, & Becker, 2009).

 

The purpose of systematic reviews

 

A systematic review is a procedure of systematically identifying relevant research papers and, from these papers, systematic quality assessments and result condensations are made. The intention of a systematic review is to appraise and interpret all available research evidence being of importance to a formulated question. Study quality is evaluated in a standardised manner and the studies fulfilling the inclusion criteria are systematically synthesised. The purpose of a systematic review is to find an answer to a specific research question on the basis of the best obtainable evidence (Glasziou, Irwig, Bain, & Colditz, 2003). In a systematic review all steps conducted should be transparent and therefore reproducible. Through the use of transparent and systematic methods bias is minimised and reliable findings are provided from which conclusions can be drawn (Liberati et al., 2009). Today systematic reviews have become an important tool for guiding decision making in health care. Clinicians need them to keep themselves informed in their domain (Swingler, Volmink, & Ioannidis, 2003). Systematic reviews are considered to be the “gold standard” in terms of research evidence as they synthesise the findings of several studies examining the same questions (Boland, Cherry, & Dickson, 2013). Systematic reviews can be regarded as basis for future research since they frequently serve as the origin for the development of clinical practice guidelines (Moher, Liberati, Tetzlaff, Altman, & the PRISMA Group, 2009). They contribute to practice decisions in all areas of health care. Systematic reviews are essential for policy makers since they supply evidence for the assessment of risks, benefits, and harms of procedures in health care (Liberati et al., 2009).

 

According to Glasziou et al. (2003) systematic reviews have two main advantages. Firstly, they enhance the potential to study consistency of results through data combination. The statistical power is significantly improved through combination of all the studies that seek to answer a specific question. Secondly, similar effects across different study settings give evidence of stability and transferability of the findings to other settings. Inconsistencies of study settings may improve the generalisability and transferability of data and therefore they can be regarded as advantageous.

 

Disadvantages of systematic reviews might be the influence of small biases and small effect sizes. All individual studies have flaws which need to be assessed. When conducting a systematic review there is the danger of allowing even small biases to result in an obvious effect (Glasziou et al., 2003).

Systematic reviews have become a fundamental tool in Evidence Based Medicine (EBM). EBM aims to promote diligent, distinct and cautious use of current best evidence in deciding about the care of individual patients. The practice of EBM seeks to integrate individual knowledge of clinical experts with the best obtainable external clinical evidence from systematic research (Sackett, Rosenberg, Muir Gray, Haynes, & Richardson, 1996). For EBM randomised controlled trials (RCTs) are widely regarded as the gold standard for assessing effectiveness of interventions in health research and the only means of establishing efficacy (Verhoef et al., 2005). Double-blind RCTs are assumed to be free of human bias (Kaptchuk, 2001). Therefore, historically systematic reviews have been RCT focused to provide best evidence, especially if the research question investigates the effectiveness of a treatment method. Other study designs, such as qualitative methods, are difficult to include in systematic reviews since their data are difficult to analyse. This can be regarded as a disadvantage when undertaking a systematic review in homeopathy since RCTs are not unanimously agreed upon as appropriate for testing the effectiveness of homeopathic interventions (Relton, O’Cathain, & Thomas, 2008).

 

Conducting systematic reviews in homeopathy

 

CAM interventions, such as homeopathy, often consist of complex whole systems of health care that are vastly different in approach to conventional biomedicine. RCTs are designed to test specific effects of standardised treatments in groups of people and this design may not necessarily meet the requirements of individualised interventions such as homeopathy (Verhoef et al., 2005). In homeopathy remedies are individually adapted for the overall condition of patients and treatments may differ between people with the same diagnosis. Homeopathic assessment typically involves a lengthy interview in which the patient is asked about many aspects of their health, current and historical, and themselves. This interview process undoubtedly carries the potential for therapeutic benefit just on its own (Weatherly-Jones, Thompson, & Thomas, 2004). One particular problem with understanding research in homeopathy is that the specific effects of homeopathic medicine and the non-specific effects of consultations cannot be easily separated and regarded independently of each other (Weatherly-Jones et al., 2004). Placebo-controlled RCTs, however, seek to measure specific effects of the treatment and any differences between the treatment and placebo arm are considered to result from the specific effects of treatment alone (Weatherly-Jones et al., 2004). Therefore it is questionable whether RCTs are appropriate research designs for trials of individualised homeopathy treatment, which represent real world practice and thus provide external validity. Weatherly-Jones et al. (2004) suggested that for homeopathic interventions modified designs are needed that take into account the complexity of treatment.

 

Further problems for systematic reviews in homeopathy are the existence of studies in which all patients of one treatment group receive the same homeopathic single remedy or complex homeopathic remedy. Such studies are usually included, especially if they are RCTs, even though they do not represent real world practice. Their results are synthesised with those trials of individualised homeopathy (IH) meeting the inclusion criteria as if they were as one and the same. In the systematic review of Kassab, Cummings, Berkovitz, van Haselen and Fisher (2009) for example, in which the effect of homeopathic medicines for adverse effects of cancer treatments were investigated, results of studies with individualised, clinical and combination or complex homeopathy were pooled together. This pooling of heterogeneous data generally leads to flawed or poor results unfavourable for homeopathy. Furthermore, the lack of large scale trials in homeopathy weakens the quality of available data and the scarcity of research in special areas often leads to the assumption that there is no evidence that homeopathy has an effect in those fields.

Motivation for conducting this systematic review

 

There are many problems facing the current body of research in homeopathy and consequently the conduct of systematic reviews on this field. However, I am interested in research on homeopathy for several reasons. I am a health care practitioner and medical homeopath having had my own practice since 1999. During my MSc course at the University of Central Lancashire I learned that many studies on homeopathy, especially RCTs, have negative results ending up with the conclusion that the clinical effects of homeopathy are more or less placebo effects. This notion is especially highlighted by systematic reviews and meta-analyses such as that of Shang et al. (2005) which examined 110 trials of homeopathy and 110 matched trials of allopathic medicine. From my own practice I believe that homeopathy works when administering the individual or similar remedy. Therefore I started to wonder why homeopathy is assessed so negatively in research and I became very interested in analysing study designs of homeopathic trials asking myself questions such as: Do these trials correspond to real world practice? Can the effect of homeopathy be ascertained scientifically with current EBM standards? These questions became important to me during my MSc course and therefore I chose to undertake a systematic review of homeopathy for my thesis. The topic menopause is very common in my practice. Many women seek help for menopause-related symptoms. At the moment I am 46 years old and menopause is something I will experience personally in the near future. When deciding a topic for my systematic review I could not find any existing systematic review on menopause treated with homeopathy. This is a further reason why I have chosen this topic. I believe, and hope, that my systematic review will make an important contribution to research and that it will be published. Therefore I chose to conduct a systematic review according to the statement for Preferred Reporting Items for Systematic reviews and Meta-Analysis (PRISMA) to ensure scientific quality and the potential for publication in a peer reviewed journal.

 

Different types of systematic reviews

 

The Cochrane Database for Systematic Reviews (CDSR) is the most important supply for systematic reviews in health care. Systematic reviews accepted by Cochrane are peer reviewed and internationally considered to be of the highest quality in EBM (MacLehose, 2014). Cochrane reviews all have the same rigorous structured format and the use of Review Manager Software facilitates this (Higgins & Green, 2011). The Cochrane collaboration was founded in 1993 and is registered as a charity in the UK. It is an international not-for-profit organisation, an independent network of health practitioners, researchers, patient advocates and others, responding to the challenge of providing research evidence serving as informing decisions about health care (“The Cochrane Collaboration”, 2014). To help make such well-informed decisions the Cochrane collaboration prepares, maintains and promotes the accessibility of systematic reviews of the evidence. One of the important principles of Cochrane reviews is to minimise bias through scientific rigour and to avoid conflicts of interest. (Higgins & Green, 2011). The Cochrane Central Register of Controlled Trials (CENTRAL) is a bibliographic database providing extraordinarily concentrated reports of RCTs (Foxlee, 2014). The Cochrane Handbook for Systematic Reviews of Interventions provides guidance on structured methodology and format for the preparation and maintenance of Cochrane Intervention reviews and Cochrane Overviews of reviews (Higgins & Green, 2011). If accepted, systematic reviews of interventions following Cochrane guidelines are published in full online in the CDSR, in The Cochrane Library which is published by Wiley-Blackwell, on the internet as well as on CD-ROM (Higgins & Green, 2011). To reduce the potential for bias in the review process, assessments should be made independently of the results of the studies included. For Cochrane reviews it is therefore a prerequisite to establish and document methods that are to be used prior to the review process (Higgins & Green, 2011). This is ensured by the publication of a review protocol before having the relevant studies at one’s disposal since this minimises the review authors’ biases, assures transparency of methods and processes and thus allows peer review of the planned methods (Light & Pillemer, 1984). The development of a protocol for a Cochrane review is similar to that of a single trial. A Cochrane review should follow the review protocol and all changes in it have to be reported and documented. Protocols for Cochrane reviews are published before the accomplished systematic review in the CDSR (Higgins & Green, 2011).

 

Outwith Cochrane, other systematic reviews, also considered to be of high importance, are those following the PRISMA statement. Meta-analysis, the use of statistical techniques in a systematic review to summarise the results of the included studies, is incorporated into many systematic reviews, but not all (Moher et al., 2009). PRISMA is an update of the guidelines for the reporting of meta-analyses of RCTs, called the statement for Quality Of Reporting Of Meta-analyses (QUOROM) (Moher et al., 2009). The QUOROM Statement was developed in 1996 and published in 1999 (Liberati et al., 2009). The PRISMA Statement was developed in the year 2005 with 29 participants, called the PRISMA group (Moher et al., 2009). The statement was revised 11 times by the PRISMA group before it was approved for publication in 2009 (“PRISMA: Transparent reporting of systematic reviews and meta-analyses”, 2014). The aim of PRISMA is to guarantee clear presentation of the planning, procedure and results of a systematic review (Liberati et al., 2009). PRISMA also focuses on RCTs, like the Cochrane collaboration, but other types of research can also be included when conducting a PRISMA review (“PRISMA: Transparent reporting of systematic reviews and meta-analyses”, 2014).  Review authors following PRISMA guidelines have to observe a 27-item checklist ensuring transparent reporting of a systematic review as well as a four-phase flow diagram. The aim of the flow diagram is to make the process of the literature search transparent by presenting the numbers of identified records, excluded articles and included studies (Liberati et al., 2009). Systematic reviews accepted by PRISMA are published in peer reviewed journals. Journals endorsing PRISMA are listed on the PRISMA Web site (www.prisma-statement.org) (Moher et al., 2009). Before conducting a PRISMA review it has to be registered in PROSPERO, an international prospective register of systematic reviews in health and social care, part of the Centre for Reviews and Dissemination (CRD), University of York, UK. In the registration form the main features from the review protocol about the design and conduct of a systematic review are submitted and kept as a permanent record. Thus, transparency is provided and publication bias reduced. The key aims of PROSPERO are to avoid duplication and to minimise bias in systematic reviews. Registration records are made available to the public on an open database. If during the review process an alteration of plans is taking place, registration information has to be changed and will be published. After having accepted the registration of a systematic review PROSPERO provides a unique registration number for each registered review which can be cited in publications and records (“University of York: Centre for Reviews and Dissemination”, 2013). PRISMA considers a review protocol indicating objectives and methods of a systematic review as important (Liberati et al., 2009). If a review protocol exists, the author has to list places where the protocol is registered and where it can be accessed during the registration process and before starting with the PRISMA review (“University of York: Centre for Reviews and Dissemination”, 2013).

 

For EBM systematic reviews that do not follow particular guidelines, or provide a study protocol and/or registration form that is available to the public prior to the review process, are deemed to be of lower quality evidence as they carry a high potential for bias. The chance of such a review being published in peer reviewed journals or databases is very low. 

 

 

MENOPAUSE TREATED WITH HOMEOPATHY: A SYSTEMATIC REVIEW (following PRISMA guidelines)

 

Keywords: Systematic review, menopause, climacteric, homeopathy.

 

Abstract

 

Purpose: To examine whether any kind of homeopathic treatment (individualised, complex, clinical, isopathic) reduces or relieves menopause-related symptoms in women with a diagnosis of natural or artificial menopause (due to breast-cancer treatment) independent of age.

Methods: A systematic review of single or double-blinded controlled trials  comparing individualised, clinical, complex, therapeutic and isopathic homeopathy with conventional treatment, other kind of homeopathy or placebo in the treatment of menopausal symptoms in women of any age. Registration Number: CRD42014013999. English and non-English articles from 1990 to 2014 were searched on electronic databases including PubMed, ScienceDirect (1990 – Present), Ebscohost.com (1995 – 2013), OVID (Journals and eBooks) (1990 – Present), EMBASE (1996 – Present), MEDLINE (with Full-Text) (1997 – 2013), Academic Search Complete (1995 – 2012), AMED (Allied & Complementary Medicine Database) (1990 – 2010), BioMed Central, CINAHL Complete (1997 – 2013), The Cochrane Library, ProQuest Health & Medical Complete (1997 – Present), SpringerLink (1997 – Present), Wiley Online Library (1990 – Present), http://www.opengrey.eu (for grey literature), https://clinicaltrials.gov/ (for unpublished and ongoing research), www.dh.gov.uk (for unpublished and ongoing research), cam-quest.org. Search terms used were homeopathy, homoeopathy, menopause and climacteric.

Results: From 548 records retrieved and screened 6 full articles were included in the review. One study was assessed as having a high risk, two as having an unclear risk and three as having a low risk of bias. All studies included were placebo-controlled and double-blinded. Two studies tested the effect of complex homeopathy (one of them used an anthroposophic medication), one study investigated the effect of individualised homeopathy (IH) and a complex homeopathic remedy, and two studies tested the effect of IH. Duration and sample size differed tremendously and 3 studies were under-powered. Although the findings of the studies under review have revealed mixed results there is sufficient evidence to state that both complex and IH may have specific treatment effects over and above placebo in the treatment of menopause.

Conclusions: This review has identified only a small number of heterogeneous studies with two types of homeopathy testing treatment effects on menopausal symptoms. Further work in this field needs to be done with sufficiently powered studies. To adequately examine the therapeutic context of the whole individualised homeopathic approach we need more pragmatic trials as described by Verhoef et al. (2005). For studies with participating women taking tamoxifen or newer oestrogen inhibitors we suggest sub-group analyses.

 

     

1. Introduction                         

In general, natural menopause occurs between the ages of 48 – 51 years (De Gruyter, 1994). The number of women aged 40 – 49 years has increased by 58% in the last 50 years in developed countries and by 32% in Europe (“International Menopause Society”, 2009). The growth of the menopausal population will bring about new challenges implying a change in medicine (Wilken-Jensen & Ottesen, 2003). Menopause is defined retrospectively and is characterised by 12 months of amenorrhoea after the final menstrual period (De Gruyter, 1994). The peri-menopause, or climacteric, is the phase in between the fertile stage and menopause (Penny et al., 2009). Artificial menopause can be induced by breast cancer treatment through the anti-oestrogen effects of endocrine medicaments such as tamoxifen or aromatase inhibitors, or chemotherapy treatment which can impair the natural function of the ovaries (Rada et al., 2010). After menopause the ‘post-menopause’ period can last around 10 years (Groth & Stolzenberg, 1995). Typical menopausal symptoms include hot flushes, sleep and mood disturbance, joint pains and fatigue (Thompson, 2010). Marked sweating is another typical menopause-associated symptom. However, each woman experiences menopause and post-menopause in their own way and some women are almost free from symptoms (Katz, 1997). Menopausal and postmenopausal women account for a large proportion of CAM users (Wilken-Jensen & Ottesen, 2003). Generally hormone therapy (HT) is considered as a safe and effective modality for the treatment of menopause-related symptoms but many women experience unpleasant side-effects (Katz, 1997). According to Wilken- Jensen and Ottesen (2003) HT increases the risk of thromboembolism. Furthermore it has been found that the long-term use of HT significantly increases the risk of developing endometrial and breast cancer (Dixon, 2003). Therefore HT is not regarded as appropriate treatment in women with a history of breast cancer (Rada at al., 2010). Menopausal women who cannot take HT, or who choose not to, seek alternative treatments for menopause-related symptoms. Homeopathy is the alternative of choice for many (Kupferer et al., 2009). Homeopathy was founded by the German medical doctor Samuel Hahnemann in the late eighteenth century and nowadays it is one of the most-used forms of medicine in the world and the fastest growing. Worldwide 300 – 500 million patients in over 80 countries use it (“Canadian Consumers Centre for Homeopathy”, 2011). Homeopathy is based on the principle of similars which asserts that substances that cause symptoms in healthy persons can be used to stimulate healing in sick people displaying similar symptoms (Jonas & Jacobs, 1996). The word ‘homeopathy’ is used to describe different forms of treatment, some administered by a trained practitioner, some self-prescribed, some involving single remedies and some combinations of many different substances (Thompson, 2010). ‘Classical’ or IH involves the in-depth assessment of patients, typically through a lengthy consultation, in order to identify the homeopathic remedy that is known to exhibit the greatest similarity in symptom pattern to that of the patient (Vithoulkas, 1980). This single remedy is normally prescribed in a highly diluted and potentised form to stimulate healing whilst avoiding side-effects. ‘Complex’ homeopathy differs from this approach as more than one homeopathic preparation is used in a fixed combination for a particular condition or illness in a standardised manner. Further kinds of homeopathy are ‘clinical’ homeopathy where the same single homeopathic remedy is prescribed for a group of patients with the same clinical complaint; and isopathy, where the homeopathic medicine is actually prepared from the causative factor, for example, the homeopathic remedy Coffea (from coffee beans) being used to treat the adverse effects of coffee abuse (Linde et al., 1997). There are some indications that the use of individualised treatment by a homeopath can have clinical benefits. For example, an audit undertaken between 2001 and 2003 by Relton and Weatherley-Jones (2005), in a community menopause clinic in the UK, found that menopausal women, not taking HT, reported significant benefits from individualised homeopathic treatment. However, to date no systematic review of experimental research has been published that investigates the effectiveness of homeopathy in the treatment of menopause-related symptoms. The purpose of this paper is to present the first systematic review of this area. All forms of homeopathy were included although classical homeopathy, with the administration of single remedies, is the most widespread representing real world practice of homeopaths (Blackstone, 1993).

 

To examine whether any kind of homeopathic treatment (individualised, complex, clinical, isopathic) reduces or relieves menopause-related symptoms, all controlled trials comparing homeopathy with placebo, conventional treatment or other kind of homeopathy were reviewed. Included were studies that assessed the effectiveness of homeopathy in the treatment of menopause-associated symptoms in women with a diagnosis of natural or artificial menopause (due to breast cancer treatment) independent of age. All studies with potencies from 3X onwards were included.

 

2. Methods

 

The systematic review was conducted according to PRISMA statement criteria (Liberati et al., 2009). It was registered in PROSPERO (registration number CRD42014013999). The review record was amended on December 15, 2014: Secondary outcomes were broadened since almost all relevant studies found at that point used scales and scores as mentioned in primary outcomes (= menopause scales, quality of life scales, diaries, visual analogue scales, individual menopause-related symptom scales). Therefore all primary outcomes were added to criteria for secondary outcomes.

 

2.1. Inclusion criteria

 

All studies were included if they met the following inclusion criteria: Randomised, prospective, single or double-blinded studies; from 1990 till present; with control as placebo, conventional treatment, other kind of homeopathy; assessing the effectiveness of different kinds of homeopathy (individualised or classical, complex, clinical, isopathic); in the treatment of menopause-related symptoms such as hot flashes, marked sweating, depression, vaginal dryness, low libido, osteoporosis, fatigue, sleep disturbances, palpitations, emotional imbalance, etc.; in women of any age with a diagnosis of natural or artificial menopause (as for example caused by breast cancer treatment). No restrictions on language were set; papers in languages other than English and German were translated. Forms of homeopathic administration included were liquids, tablets, external applications and olfactory.

 

2.2. Search strategy

 

The following electronic databases were searched in September 2014 by Monika Mueller-Amenitsch: PubMed, ScienceDirect (1990 – Present), Ebscohost.com (1995 – 2013), OVID (Journals and eBooks) (1990 – Present), EMBASE (1996 – Present), MEDLINE (with Full-Text) (1997 – 2013), Academic Search Complete (1995 – 2012), AMED (Allied&Complementary Medicine Database) (1990 – 2010), BioMed Central, CINAHL Complete (1997 – 2013), The Cochrane Library, ProQuest Health & Medical Complete (1997 – Present), SpringerLink (1997 – Present), Wiley Online Library (1990 – Present), http://www.opengrey.eu (for grey literature), https://clinicaltrials.gov/ (for unpublished and ongoing research), www.dh.gov.uk (for unpublished and ongoing research), cam-quest.org. The search strategy combined the terms “menopause” or “climacteric” and “homeopathy” or “homoeopathy”. The search strategy on PubMed was “menopause” AND “homeopathy” OR “menopause” AND “homoeopathy” OR “climacteric” AND “homeopathy” OR “climacteric” AND “homoeopathy”.

 

 

2.3. Study selection and data extraction

 

Titles of papers retrieved by the search were assessed for inclusion according to above mentioned inclusion criteria and articles which clearly did not fulfil the eligibility criteria were excluded. Potentially relevant abstracts and full texts were assessed by both reviewers and any discrepancies resolved through discussion via skype or e-mail. Remaining studies were included in the systematic review and full text copies were obtained. Reference lists of included papers were checked for further relevant studies. The data from included studies was extracted by one reviewer and independently cross-checked by the second to minimise error and reduce potential bias. Disagreements were resolved by discussion between the two review authors. Information was extracted from each included trial on: (1) study type (randomised, blinded), (2) study methods (including number and characterisation of participating women, details of type and mode of administration, details on dosage, duration of intervention), (3) type of outcome measure (timing and measurement including the level of reduction of menopause-associated symptoms, improvement in quality of life score (using a validated scale), length of follow-up, unintended effects of treatment, change in levels of Serum FSH, oestrogen and lipid profile), (4) publication year and journal of application.

 

 

 

 

2.4. Quality assessment

 

 

Each study that met the inclusion criteria was rated with the Cochrane Collaboration’s tool for assessing risk of bias as listed in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins & Green, 2011). This instrument provides a standardised means to assess study quality by assigning a judgement of “low risk” of bias, “high risk” of bias, or “unclear risk” of bias. The criteria covered by this quality assessment tool include (A) random sequence generation (selection bias), (B) allocation concealment (selection bias), (C) blinding of participants and personnel (performance bias), (D) blinding of outcome assessment (detection bias), incomplete outcome data (attrition bias), (E) selective reporting (reporting bias), (F) other sources of bias. Risk of bias was assessed independently by the two reviewers. Any disagreements were resolved by discussion via e-mail or skype. Studies were characterised as having a low risk of bias if all of the seven criteria in the risk of bias table were assessed as low risk.

 

2.5. Principal summary measures

 

The pre-specified primary outcome measures of treatment effect were: menopause and quality of life scales; improvement of general and global symptoms recorded in validated diaries, and any scores from validated visual analogue scales; individual menopause-related symptoms such as hot flushes, emotional imbalances, depression, palpitation, insomnia, vaginal dryness, etc.

3. Results

 

In the electronic search a total of 548 citations were found. After removing the duplicates (n = 58) 490 remained. Of these, 477 records were excluded because after reviewing the abstracts it was found that they clearly did not meet the inclusion criteria. The full text of the remaining 13 studies were read to assess eligibility. Of these, 7 studies did not meet the inclusion criteria as described.

 

 

Excluded studies

1.    An experimental study without control, conducted in South Africa, in which a complex homeopathic remedy was compared with a complex phytotherapeutic product in the treatment of typical climacteric symptoms (Penny, Razlog, Deroukakis, & Johnston, 2009). Excluded because phytotherapy had not been defined as comparator in our inclusion criteria.

2.    An open, multi-national prospective, pragmatic and non-comparative observational study evaluating the effect of IH on hot flashes in menopausal women. 5 % of the patients additionally received non-homeopathic medication (notably minerals) and/or food supplements (notably soy-based) (Bordet, Colas, Marijnen, Masson, & Trichard, 2008). This study, conducted in France, was excluded because of lack of control group.

3.    An investigation into homeopathic remedies as placebo alternatives, conducted in Poland, testing a complex homeopathic remedy as an alternative to placebo in the treatment of menopause-related vegetative and emotional disturbances. The author of this study (Wasilewski, 2004) regarded the use of multi-ingredient homeopathic drugs as a particularly interesting alternative to the placebo hypothesising that they might have the same high effectiveness as placebo (30 – 50%) in the treatment of depressive disorders. Since this systematic review aimed to investigate the effect of homeopathic treatment and not whether homeopathy can be considered a placebo alternative, both review authors agreed that this study did not fulfil the specified inclusion criteria.

4.    A prospective observational trial, conducted in the UK, investigating IH in the treatment of symptoms of oestrogen withdrawal in breast cancer patients (Thompson & Reilly, 2003). Excluded because it lacked a control group.

5.    An open, multicentre, prospective, observational study conducted in India, to assess the usefulness of IH in distress during climacteric years (Nayak et al., 2011). Excluded because it lacked a control group.

6.    A pilot outcome study, conducted in the UK, ascertaining the effect of IH in the treatment of hot flushes (Clover & Ratsey, 2002). Excluded because it lacked a control group.

7.    A pilot cohort multiple randomised controlled trial conducted in the UK assessing the clinical and cost effectiveness of individualised homeopathic treatment in women with menopausal hot flushes (Relton, O’Cathain, & Nicholl). Excluded because the control group received no intervention which was not defined as comparator in our inclusion criteria.

 

Six studies met the inclusion criteria and were included in the systematic review. No further studies were found from the references of the included studies. One unpublished, but accomplished, study was found. Its results were obtained on https://clinicaltrials.gov/. One study, published in Dutch, was translated into English. The following figure (Figure 1) shows the flow of information through the different stages of the systematic review:  

 

 

 Figure 1: PRISMA 2009 Flow Diagram showing the results of the literature search

Records identified through database searching
( n = 548)

 

Additional records identified through other sources
(n =  0 )

Records after duplicates removed
(n = 58 )

Records screened
(n = 490 )

Records excluded
(n = 477 )

Full-text articles assessed for eligibility
(n = 13 )

Full-text articles excluded, with reasons
(n = 7 )

Studies included in qualitative synthesis
(n = 6 )

Studies included in quantitative synthesis (meta-analysis)
(n = 0 )

 

 

 

 3.1. Characteristics of included studies

 

Six double-blinded, controlled trials (five placebo-controlled and one compared with conventional treatment and placebo), had a total of 484 participating women. Only the full text of one study could not be obtained in English or German (Bekkering, Van den Bosch, & Van den Hoogen, 1993). It was translated from Dutch into English. All other papers were obtained in English.

Three studies with a total of 215 participating women evaluated the effect of complex homeopathy in the treatment of peri – and postmenopausal symptoms (von Hagens et al., 2012; Bekkering et al., 1993; Colau, Vincent, Marijnen, & Allaert, 2012). The name of the complex homeopathic remedy in the study of Bekkering et al. (1993) was Famosan tabletten ® (presumably A. Vogel, Netherlands) and that of Colau et al. (2012) was BRN-01/Acthéane ® (Laboratoires Boiron, France).

In the study of Hagens et al. (2012) a homeopathic anthroposophic complex medication called Ovaria comp. Globuli velati® (WALA Heilmittel GmbH, Germany) was tested. This was a three-armed and three phased trial. Each group experienced one placebo period and two remedy periods in different sequences. Each period lasted 12 weeks. At the end of the study, the total changes were compared between the three groups after 2 x 12 weeks of remedy and 1 x 12 weeks of placebo in different chronological orders. There were no washout periods in between.

One study, with 83 participating women, tested the effectiveness of individualised and complex homeopathy for menopausal symptoms in breast-cancer survivors (Jacobs, Herman, Heron, Olsen, & Vaughters, 2005). In this study all three arms (combination, single remedy, placebo) were compared with each other. The name of the complex homeopathic remedy was Hyland’s Menopause® (Standard Homeopathic Company, USA).

Two studies, with a total of 186 participating women, investigated the effect of IH in the treatment of menopausal, peri – and postmenopausal symptoms (Thompson, Oxon, Montgomery, Douglas, & Reilly, 2005; Macías-Cortés, Aguilar-Faisal, & Asbun-Bojalil, 2013). The study of Macías-Cortés et al., 2013 was not published at the point of this review in December 2014. However, the results of the study were available and could be retrieved on https://clinicaltrials.gov/. Participants in the study of Thompson et al. (2005) were breast-cancer survivors, having menopausal symptoms from oestrogen withdrawal.

All studies were randomised except for that of Bekkering et al. (1993) which was a semi cross-over design. Primary and secondary outcome measures for all studies were menopause and quality of life scales, anxiety and depression scales and patient diaries. Only the Jacobs et al. study (2005) included biochemical testing, with FSH level set as a secondary outcome. In all studies, except for Bekkering at al. (1993), adverse events were monitored. However, none of the studies examined adverse events as primary or secondary outcomes. The duration of interventions was very different and ranged from 6 weeks (“ClinicalTrials.gov”, 2015) to one year (Jacobs et al., 2005). The mean age of participating women was not available for all studies but ranged from 49 to 55.5 years where detailed. In the Colau et al. study (2012) there was no clear information about age as it simply stated that ‘women aged 50 years or older’ were included. Bekkering et al. (1993) provided no information about age. Sample sizes varied tremendously and ranged from 5 to 133 participating women.

The aims of the studies and the particular aspects of menopause that were assessed also differed. Jacobs et al. (2005), Hagens et al. (2012) and Thompson et al. (2005) examined menopausal symptoms in general. Bekkering et al. (1993) aimed to objectify the positive effect of a homeopathic complex remedy on menopausal symptoms that had already been taken by the participating women before baseline. Colau et al. (2012) evaluated the efficacy of a homeopathic complex remedy on the reduction of the hot flash score. Macías-Cortés et al. (2013) examined the efficacy, and safety, of IH versus placebo or fluoxetine, a selective serotonin reuptake inhibitor (SSRI), in menopausal women with moderate to severe depression. Baseline characteristics were similar between groups in 5 studies. Colau et al. (2012) reported a higher baseline hot flash severity score in the placebo group. However, the different values were adjusted using Cole’s least mean square method. Detailed characteristics of the six included studies such as specifications of medications, study setting, primary and secondary outcome measures, etc. are shown in Table 1, APPENDIX 1.

 

3.2. Risk of bias in included studies

 

All 6 included studies were rated according to the Cochrane Collaboration’s tool for assessing risk of bias as previously described (Higgins & Green, 2011). The overall risk of one study was deemed to be high (Bekkering et al., 1993). The study of Hagens et al. (2012) had an unclear risk of bias for one subdomain and that of Jacobs et al. (2005) for one key domain. The other three studies were assessed as having a low risk of bias (Colau et al., 2012; Thompson et al., 2005; “ClinicalTrials.gov”, 2015). Table 2, APPENDIX 2, shows a summary of the risk of bias assessment for the 6 studies. The proportion of information from studies at high or unclear risk of bias was assessed sufficient to have an effect on interpretation of results.

 

3.3. Results of individual studies

 

Results of individual studies are shown in table 3.

For confirmation, on January 28, 2015 an e-mail was sent to Dr. Macías-Cortés since it was not clear if the result tables on https://clinicaltrials.gov/ showed the final scores after 6 weeks of treatment, or changes from baseline. Her answer was that tables showed final scores and so we could compare changes between baseline and after 6 weeks of treatment. Dr. Macías-Cortés sent us the complete results of the study which showed that evaluations were made after 4 and 6 weeks. On https://clinicaltrials.gov/ only final scores after 6 weeks were shown.

 

 Table 3: Results of included studies

 

Authors/Sample size

Primary outcome measures

Secondary outcome measures

Hagens et al./

102 women

No statistically significant difference in total score of MRS II between remedy and placebo after 12 weeks.

Total reduction of MRS II after 24 weeks of remedy in group 1 (R/R/P) and group 2 (P/R/R) was not clinically relevant.

Changes in total score of MRS II after 36 weeks reached clinical relevance in group 1: -5.0 (mean) (6.3 (SD)) , 95% CI -7.5 to

-2.5 and group 2: -5.9 (mean) (6.2 (SD)), 95% CI – 8.7 to -3.1.

Almost no decrease of symptoms after 36 weeks in group 3 (R/P/R): -5.0 (mean) (5.7 (SD)), 95% CI -2.9 to 1.9.

Bekkering et al./

5 women

Mean scores of BMI not significant between groups.

Mean scores of BMI not significant between groups.

Colau et al./

108 women

Comparison of HFS using AUC between groups: Intervention group [82.3 (mean) +/- 49.4 (SD), 95% CI 68.3 to 96.4] significantly lower than placebo group [113.0 (mean) +/-88.2 (SD), 95% CI 88.2 to 137.8].

Adjusted baseline values using Cole’s least mean square method (baseline HFS was higher in the placebo group): HFS still significantly lower in intervention group (88.2+/-6.5) than in placebo group (107.2+/-6.4).

 

No significant difference of HFRDIS  (10 dimensions and QoL) between the two groups except for the concentration dimension which was significantly better in the intervention group at week 12 (p < 0.05). Reduction in the HFRDIS score was significant in each group but did not differ significantly between groups.

Reduction of MRS was significant for each group but did not differ significantly between the two groups. Reduction in distress and the number of night sweats as measured using a VAS was similar between the two groups. Morisky-Green score showed that there was poorer compliance with treatment in the placebo group than in the intervention group: Difference not statistically significant. 

Jacobs et al. (2005)/

83 women

No significant difference in the HFSS or in the total hot flashes among the 3 groups. However, the IH group had a lower severity score and fewer hot flashes during the first 3 months of the study (p = 0.1).

Result of sub-analyses: In the no-tamoxifen users of the complex group there was a statistically significant increase in the HFSS as compared to placebo (p = 0.01) and a highly significant difference when compared to the IH group (p < 0.001). This is in contrast to figure 3: Time plots show that HFSS decreases in the complex and IH group and increases in the placebo group. A marked increase in the complex group can only be seen at month 9!

There was a highly significant increase in the total number of hot flashes in the no-tamoxifen users of the complex group compared to placebo (p = 0.006) and compared to the IH group (p = 0.002).

HFSS score of no-tamoxifen users in the IH group remained stable over time.

No statistically significant differences between the 3 groups in the KMI score or in individual symptoms of the KMI except for a significant increase in headaches in the no-tamoxifen users of the complex group at 6 months (p = 0.04) and 12 months (p = 0.03).

Positive trend toward a lower KMI score in the IH group compared to placebo (p = 0.1).

Comparison of SF-36 QoL found a positive trend in the physical functioning component (p = 0.1) during the study period in both homeopathy groups. On the SF-36 the GH score was significantly increased in both homeopathy groups compared to placebo (p = 0.02 IH versus placebo, p = 0.03 complex versus placebo).

No differences in FSH levels before or after the study period.

 

Thompson et al./

53 women

No evidence of any difference between groups for MYMOP activities of daily living (ADL) or overall profile scores. Both groups showed large and clinically important improvements over the study period: MYMOP ADL score and MYMOP overall profile score of all follow-up visits showed in both groups a difference of mean – 0.4 (a change of 0.8 on MYMOP is considered to be clinically relevant).

Mean differences between baseline and final follow-up at 16 weeks in the intervention group: -1.6 (mean), 95% CI -0.9 to -2.3 (MYMOP activity score) and -1.3 (mean), 95% CI -0.8 to 1.8 (MYMOP overall profile score). Corresponding differences in placebo group: -1.5 (mean), 95% CI -0.9 to -2.1 and -1.2 (mean), 95% CI -0.7 to -1.7.

No difference between the 2 groups of MSQ, patient diaries, HADS, FAQ measured at repeated follow-up visits. Some evidence of an interaction between time of follow-up and trial arm (IH) for assessment of overall health on the EORTC QLQ-C30: Mean scores in the intervention group continued to increase at both follow-up visits at which this variable was measured; scores in the placebo group decreased between the two follow-up visits. Adjusted difference between the 2 groups: EORTC/overall health: -0.2 (mean), 95% CI -0.8 to 0.5. EORTC/overall quality of life: -0.1 (mean), 95% CI -0.7 to 0.6.

GHHOS showed in both groups clinically relevant improvements in symptoms and mood disturbances with 75% of women reporting a marked or major improvement. Adjusted difference between the 2 groups as measured with GHHOS at final follow-up visit only was 0.03 (mean), 95% CI -1.1 to 0.6.

There was a suggestion that women in the active remedy group were less satisfied with the homeopathic approach but perceived greater helpfulness, at least for their main symptom, than women in the placebo group. Participants in both groups were equally likely to think that they had received active remedy.

 

Macías-Cortés et al./

133 women

HRSD: Difference at week 6 between IH and placebo: 5.0 (mean) +/-0.8 (SD), 95% CI 3.1 to 7.0 and between fluoxetine and placebo: 3.2 (mean) +/- 0.81 (SD), 95% CI 1.31 to 5.25.

 

HRSD Change from baseline at week 6:

IH: 11.3 (mean).

Fluoxetine: 8.9 (mean).

Placebo: 5.7 (mean)

BDI: Statistical significant difference at week 6 between IH and placebo: 3.4 (mean) +/-1.71 (SD), 95% CI -0.73 to 7.61 and difference at week 6 between fluoxetine and placebo: 1.28 (mean) +/-1.73 (SD), 95% CI -2.93 to 5.5.

BDI change from baseline at week 6:

IH: 14.3 (mean)

Fluoxetine: 10.8 (mean)

Placebo: 11.5 (mean)

 

Responder rates (= decrease of 50% or more from baseline score using HRSD after 6 weeks treatment) at week 6 between IH and placebo: p-value <0.05, OR 0.11, 95% CI 0.03 to 0.34 and between fluoxetine and placebo: p-value <0.05, OR 0.19, 95% CI 0.06 to 0.56.

Change of responder rates from baseline after 6 weeks:

IH: 24

Fluoxetine: 19

Placebo: 5

 

GS: Difference at week 6 between IH and placebo: 8.6 (mean) +/- 2.42 (SD), 95% CI 2.72 to 14.52 and between fluoxetine and placebo: 3.6 (mean) +/- 2.37 (SD), 95% CI -2.33 to 9.6.

Change of GS from baseline after 6 weeks:

IH: 17.2 (mean)

Fluoxetine: 10.4 (mean)

Placebo: 11.1 (mean)

 

Remission rates (= HRSD < 7 points after 6 weeks of treatment) at week 6 between IH and placebo: p-value 0.10, OR 0.26, 95% CI 0.05 to 1.32 and between fluoxetine and placebo: p-value 0.11, OR 0.27, 95% CI 0.05 to 1.39.

Change of remission rates from baseline after 6 weeks:

IH: 7

Fluoxetine: 7

Placebo 2

 

 

 

 

Abbreviations list:

 

ADL – Activities of Daily Living

AUC – Area Under the Curve

BDI – Beck Depression Inventory

BMI – Blatt Menopausal Index

CI – Confidence Interval

EORTC QLQ C30 – European Organisation for Research and Treatment of

                                                Cancer Quality of Life Questionnaire

FAQ – Final Assessment Questionnaire

FSH – Follicle Stimulating Hormone

GH – General Health

 

GHHOS – Glasgow Homeopathic Hospital Outcome Scale

GS – Greene’s Scale

HADS – Hospital Anxiety and Depression Scale

HFRDIS – Hot Flash Related Daily Interference Scale

HFS – Hot Flash Score

HFSS – Hot Flash Severity Score

HRSD – Hamilton Rating Scale for Depression

IH – Individualised Homeopathy

KMI – Kupperman Menopausal Index

MRS – Menopause Rating Scale

MRS II – Menopause Rating Scale II

MSQ – Menopausal Symptom Questionnaire

MYMOP – Measure Yourself Medical Outcome Profile

OR – Odds Ratio

P – Placebo

QoL – Quality of Life

R – Remedy

SD – Standard Deviation

SF-36 QoL – SF-36 Quality of Life Score

VAS – Visual Analogue Scale

 

 

 3.4. Adverse events

 

Hagens et al. (2012) reported 85 adverse events from 96 participants (88.5%). 3 adverse events were serious, but were not related to trial medication. Colau et al. (2012) reported 5 adverse events in the intervention group, and 4 in the placebo group, including one severe in each group. Serious adverse events were not considered to be related to the medication investigated. Statistical analyses of Jacobs et al. (2005) revealed increased hot flashes and headaches in the homeopathic complex no – tamoxifen group. In the trial of Thompson et al. (2005) approximately one quarter of the women experienced adverse events in both groups. There was no evidence of any difference between the two groups. “ClinicalTrials.gov” (2015) reported no serious adverse events. The highest percentage of other adverse events was found in the fluoxetine group (52.17%) followed by the IH group (43.18%) and the placebo group (37.21%).

3.5. Synthesis of results

 

Because the type of homeopathic interventions, sample sizes and study durations varied markedly in the studies included, meta-analysis would not have been appropriate and hence the decision was taken to describe the studies, their results, their limitations and applicability in a narrative, qualitative manner.

 

Six placebo-controlled, double-blinded trials with a total of 484 participating women met the inclusion criteria. All studies except for one (Bekkering et al., 1993) were randomised. Three studies, with a total of 215 women, tested complex homeopathic medicines for peri – and postmenopausal symptoms (Hagens et al., 2012; Bekkering et al., 1993; Colau et al., 2012). Hagens et al. (2012), with a total of 102 women, assessed the effect of an anthroposophic complex homeopathic medicine called Ovaria comp. Globuli velati® (WALA Heilmittel GmbH, Germany) in a three arm, three phase trial. Jacobs et al. (2005), with 83 women, investigated the effect of IH and a complex homeopathic remedy called Hyland’s Menopause® (Standard Homeopathic Company, USA) on menopause-related symptoms due to carcinoma in situ or stage I-III of breast cancer. All three arms (IH/complex remedy/placebo) were directly compared with each other. Two studies, with a total of 186 women, tested IH for menopausal symptoms (Thompson et al., 2005; “ClinicalTrials.gov”, 2015). Thompson et al. (2005) investigated IH for menopausal symptoms due to oestrogen withdrawal in breast cancer. “ClinicalTrials.gov” (2015) compared IH with fluoxetine in the treatment of menopausal and postmenopausal disorders with moderate to severe depression.

 

3.5.1. Complex homeopathy

 

The study by Hagens et al. (2012) was assessed as having an unclear risk of bias in the subdomain incomplete outcome data. In this study, changes in the Menopause Rating Scale II (MRS II) which was the primary outcome measure, after 12 weeks, did not show any statistical difference between remedy and placebo. Neither did evaluation of MRS II after 24 weeks (a secondary outcome measure). However, total reduction of MRS II (a secondary outcome measure) after 36 weeks of treatment reached clinical relevance in the two groups who had taken the homeopathic remedy continuously for 24 weeks (phases 1&2 or phases 2&3). In the group in which the two remedy periods of 12 weeks (phases 1&3) were interrupted by 12 weeks of placebo (in phase 2), almost no decrease of symptoms could be seen after 36 weeks. Compliance was poor: 64.5% of total patients took at least 75% of the recommended dose during the first treatment period versus 80% during the last two periods.

The study by Colau et al. (2012) was assessed as having low risk of bias. In this study of a complex homeopathic medication the primary outcome measure, Hot Flash Score (HFS), was significantly lower (mean reduction 56.7%) in the intervention group than in the placebo group after 12 weeks of treatment. This was still the case after adjusted baseline values using Cole’s least mean square method. In secondary outcome measures one sub-dimension (concentration dimension) was significantly better in the intervention group when compared with placebo. Reduction of other secondary outcome measures (HFRDIS, MRS) was significant in both groups but did not differ significantly between groups. Compliance in the intervention group was better than in the placebo group, but not statistically significant.

The study by Bekkering et al. (1993) was judged as being at high risk of bias. This study, conducted over 4 months, had only 5 participants. It was underpowered for meaningful statistical analysis. Mean scores of Blatt Menopausal Index (BMI) did not differ significantly between groups. However, the participating women had already taken the remedy before baseline and reported positive effects. The aim of this study was not to examine if the remedy had an effect, but to ‘objectify the positive effect’ previously experienced. 

 

 

3.5.2. Individualised homeopathy

 

In the three trials that included IH all participating women had an initial homeopathic consultation and follow-ups with a blinded homeopath. The study by Jacobs et al. (2005) in which two forms of homeopathy (IH and complex) were investigated, was assessed as having an unclear risk of bias in the domain selective reporting. Homeopathic follow-ups took place at 2 months intervals. There were no significant differences in the Hot Flush Severity Score (HFSS), or in the total number of hot flushes (primary outcome measures) among the three groups after the study period of one year. However, in comparison to the other groups the IH group had a lower HFSS and fewer hot flushes during the first three months of the study. Sub-analyses showed that the no-tamoxifen users of the complex group had a statistically significant increase of HFSS compared to placebo and a highly significant increase when compared to IH. However, time plots indicted that HFSS decreased in the complex and single remedy group and increased in the placebo group. A marked increase in the combination group can only be seen at month 9. Total numbers of hot flushes of the same sub-group show a highly significant increase when compared with the other two groups. Kupperman Menopausal Index (KMI) (secondary outcome measure) did not show any statistically significant differences between the three groups except for a significant increase in headaches in no-tamoxifen users in the complex group at 6 and 12 months. However, there was a positive trend toward a lower KMI score in the IH group compared to placebo. General Health score (GH) improved significantly in both homeopathy groups compared to placebo (secondary outcome measure). In both homeopathy groups comparison of SF-36 Quality of Life Score (SF-36 QoL) showed a positive trend in the physical functioning component during the study period.

The only trial testing IH against placebo alone lasted for 16 weeks (Thompson et al., 2005). It was deemed as having a low risk of bias. All 53 participating women had 4 follow-up visits at 4-weekly intervals with a homeopath. Both groups showed large and clinically relevant improvements over the study period for Measure Yourself Medical Outcome Profile (MYMOP) Activities of Daily Living (ADL) (primary outcome measure). However, there was no evidence of any difference between the groups. There was some evidence of an interaction between the time of follow-up and trial arm (IH) for assessment of overall health (secondary outcome measure). In both groups clinically relevant improvements in symptoms and mood disturbances were seen for the Glasgow Homeopathic Hospital Outcome Scale (secondary outcome). No difference was found between the two groups for Menopausal Symptom Questionnaire (MSQ), patient diaries, Hospital Anxiety and Depression Scale (HADS) and Final Assessment Questionnaire (FAQ) (secondary outcome measures). The study failed to show that any specific effect of IH added over and above the non-specific effects of the homeopathic consultations plus placebo. In both groups participants were equally likely to believe they had received active remedy, thus suggesting a treatment effect of the homeopathic consultation.

The three-armed trial of “ClinicalTrials.gov” (2015) was assessed as having a low risk of bias. All participants had a homeopathic consultation follow-up at week 4. After the study period of 6 weeks the primary outcome measure (Hamilton Rating Scale for Depression (HRSD)) indicted a difference between IH and placebo, and the null hypothesis, that the means would be  the same across IHT, fluoxetine and placebo, was rejected. The same was the case after evaluation of Greene’s Scale at week 6. Between IH and placebo the secondary outcome measure of Beck Depression Inventory (BDI) at week 6 indicted the statistical significant ANOVA result (p<0.05) which suggested rejecting the global null hypothesis. When comparing baseline values with final scores in outcome measure data tables, after 6 weeks of treatment, the IH group experienced the highest degree of change in all measures, except for remission rates at 6 weeks (participants with a score of < 7 in HRSD, secondary outcomes). There, IH and fluoxetine scores were equal.

 

3.6. Risk of bias across studies

 

The methodological quality of the included studies was ascertained using the Cochrane Collaboration’s tool for assessing risk of bias (Higgins & Green, 2011). Three of the 6 included studies were of low risk of bias (Colau et al., 2012; Thompson et al., 2005; “ClinicalTrials.gov”, 2015). Two of these found significant effects compared with placebo (Colau et al., 2012; “ClinicalTrials.gov”, 2015).

Two further studies were of unclear risk of bias (Hagens et al., 2012; Jacobs et al., 2005). Hagens et al. (2012) with an unclear risk of bias in one sub-domain (attrition bias/secondary outcomes) reported clinically relevant changes in secondary outcome measures after 36 weeks. Jacobs et al. (2005) reported significant results in one domain of secondary outcomes of both homeopathy groups compared to placebo.

The study by Bekkering et al. (1993) was deemed at high risk of bias since it was assessed as having a high risk of bias in three domains (random sequence generation, selective reporting, other sources of bias) and an unclear risk of bias in the domain incomplete outcome data but found no significant benefit of the homeopathic treatment over and above placebo.  

 

3.7. Flaws across studies

 

Three of the studies were underpowered (Bekkering et al, 1993; Thompson et al., 2005; “ClinicalTrials.gov”, 2015). Bekkering et al. (1993) had only 5 participants. For the Thompson et al. study (2005) statisticians calculated that 30 women would be needed in each of the 2 groups to give 80% power at a 5% level of significance to detect a 1.5 difference in MYMOP assuming a standard deviation of 2 units. Only 53 participants were randomised. For the Macías-Cortés et al. study (2013) statisticians calculated that to detect an effect size = 4.5, in a three group design (1:1:1), using F-Test, with a 5% risk of type 1 error (α) and 83% power, 63 participants per group would be required considering also a 10% drop-out rate. Only 133 participants were randomised.

 

Bekkering et al. (1993) did not provide detailed information about dosage and no information about contents of the remedies taken. In reference 1 of the paper “Laboratorium Biohorma” was mentioned in connection with Famosan ®, the study medication. On their website “Biohorma natuurlijk & gesond” http://www.biohorma.nl/over-biohorma/historie-biohorma.php reviewers found the manufacturer of the medication, A. Vogel. On A. Vogel’s website http://www.avogel.nl/homeopathie/famosan-tabletten.php contents and dosage of Famosan ® were found. However, it is not clear whether the contents of 2014 correspond to those during conduction of the study.     

The authors of the Hagens et al. study (2012) themselves considered the absence of washout periods between treatment phases as limitation since the potential impact of hang-over effects during the second and third treatment periods could not be excluded. Thus interpretation of results from the second and third periods, or from overall effects, was made more difficult.

Contents of this study’s medication, Globuli velati®, also raises questions. For example, for one of the contents, Apis regina tota, no homeopathic proving (drug trial) exists and this remedy is not known in IH. We therefore assumed this remedy is mainly associated with anthroposophic theories rather than homeopathy. Another concern was that Argentum metallicum does not belong to the very important remedies for the treatment of menopause such as Sepia, Pulsatilla or Lachesis but it is more used for metrorrhagia during menopause and ailments of ovaries such as chronic infection, pain and swelling (Seideneder, 1997).

In the Jacobs et al. study (2005), the complex remedy, Hyland’s Menopause®, was taken continuously over one year (1 tablet 3 times/day). This dosage was in contrast to the over-the-counter instructions for this medication which is, ‘to take it until symptoms subside or to stop after 7 days if symptoms get worse’. This issue was regarded as major flaw by the authors of the study. The number of headaches significantly increased in no-tamoxifen users of the complex group. Authors concluded that this sub-group might have undergone a proving. According to Vithoulkas (1980) a homeopathic proving can occur when a homeopathic remedy is given frequently in low doses causing symptoms it is meant to cure. The other members of this group did not experience this phenomenon which may have been due to an antidoting effect of tamoxifen. In IH it is common practice to stop medication when old symptoms get worse or when new symptoms arise (Meyer-Koenig, 1995). It is questionable why participants with increased headaches were not told to stop the remedy. 

 

4. Discussion

 

This systematic review has examined the effects of homeopathic treatment on naturally or artificially induced menopausal symptoms. Altogether 6 placebo-controlled studies, with either complex or IH treatment, were included. One of the studies included in this review was assessed as having a high risk of bias and for 2 trials the risk of bias was unclear. Duration and sample sizes differed tremendously and three studies were underpowered.

 

Type of homeopathy

Of high importance is that results are interpreted in the light of different forms of homeopathy. For IH, which is rooted in traditional principles, single homeopathic remedies are individually selected after a detailed, and often lengthy, homeopathic consultation (Vithoulkas, 1980). During this consultation it is not unusual for the patient to disclose and discuss aspects of themselves in a manner that may in itself be of therapeutic benefit (Eyles, Leydon, & Brien, 2012). IH is reflective of the real world practice of classical homeopaths and thus trials investigating the effectiveness of IH offer a higher external validity than those assessing the effect of complex homeopathy in which all participants take one and the same remedy (Matthiessen, 2011). Knowledge of the sphere of action of individual homeopathic remedies comes from homeopathic drug provings but complex homeopathic treatments, where different single homeopathic remedies are mixed together, most commonly do not have any drug proving. Assumptions are made about the potential scope of action based upon the individual components without proper testing. Hence complex homeopathic treatment is often accused of contravening the individualised and holistic principles of homeopathy (Blackstone, 1993).

 

 

Synthesised results

Although the findings of the studies under review have revealed mixed results there is sufficient evidence to state that both complex and IH may have specific treatment effects over and above placebo. One of complex homeopathy reported significant results in primary outcome measure (Colau et al., 2012). A further study of complex homeopathy (Hagens et al., 2012), and one of complex and IH (Jacobs et al., 2005) reported clinically relevant results in some secondary outcome measures. In the latter study positive differences in trends were seen in the IH group in primary and secondary outcomes. One study of complex homeopathy did not show any significant differences at all (Bekkering et al., 1993) and in one IH study clinically relevant improvements were detected in both groups (Thompson et al., 2005). Lastly, one study of IH detected statistically significant benefits over placebo in one secondary outcome measure as well as further important differences in other outcomes (“ClinicalTrials.gov”, 2015). This was the shortest study lasting only 6 weeks.

 

The tamoxifen factor

Two of the studies included tamoxifen-users (Thompson et al., 2005; Jacobs et al., 2005). Only Jacobs et al. (2005) made sub-group analyses between tamoxifen and no-tamoxifen users: The no-tamoxifen users of the complex remedy group differed in occurrence of headaches and it has been suggested that they may have undergone a homeopathic proving of the remedy. The authors assumed that the tamoxifen had antidoted the homeopathic complex remedy since the tamoxifen-users, of the same group, did not develop this phenomenon.

In the trial of Thompson et al. (2005) 79% of the participating women used tamoxifen and no sub-group analysis was undertaken. If tamoxifen has the potential to antidote homeopathic treatment, as suggested by Jacobs et al. (2005), then it seems reasonable to suppose that this may have contributed to the lack of significant difference between treatment and placebo groups in the Thompson study!

 

Specific and non-specific effects

Three trials tested the effects of IH (Jacobs et al., 2005; Thompson et al., 2005; “ClinicalTrials.gov”, 2015). In these studies all participants received initial homeopathic consultations and follow-ups. In all studies the homeopaths were blinded. Interestingly, in the study by Thompson et al. (2005) both groups showed large and clinically relevant improvements in primary outcomes (MYMOP), and participants of both groups were equally likely to think that they had received the active remedy. The improvements in both groups have been attributed to the therapeutic value of the homeopathic consultation, which raises the issue of whether the placebo group can be considered as having no treatment (Moerman & Jonas, 2002). Participants had one initial homeopathic consultation and 4 homeopathic follow-up visits. Weatherley-Jones et al. (2004), on consideration of this issue, suggests that in order for clinical trials of homeopathy to represent real world practice accurately, modified study designs are necessary to take into account the complexity of the homeopathic consultation. The specific effects of homeopathic medicine, and the non-specific effects of consultations cannot be regarded as independent of one another. Therefore, Walach (2000) suggests to consider the whole setting of homeopathy as curative to make clinical research in homeopathy more informative. In complex interventions such as homeopathy the consultation alone may have therapeutic effects and, in addition, there is very little research investigating the sole effect of homeopathic consultations (Thompson, 2004). Weatherley-Jones et al. (2004) asserts that efficacy of IH cannot be adequately measured in placebo-controlled RCTs because of this factor. For assessment of the effect of IH more pragmatic trials, offering external validity by representing real world practice, are needed (Thompson & Relton, 2009). One important example of such an approach was the study of Relton et al. (2012) which was not included in this systematic review because the control group received usual care. In this pilot cohort multiple RCT, a cohort of women with menopausal hot flushes was recruited through an observational study. From this cohort patients were screened, and those eligible for individualised homeopathic treatment identified. Of these, only those randomly selected to the homeopathy group were informed about homeopathic treatment and offered treatment. In common with real world practice, only patients receiving the homeopathy had homeopathic consultations.

Weatherley-Jones et al. (2004) reported about difficulties of treating homeopaths involved in double-blinded studies in which both the intervention and placebo group had homeopathic consultations. The homeopaths suffered confusion as to how to decide on the next prescription, especially when well-indicated remedies did not work. In studies with greater model validity, such as that of Relton et al. (2012), this kind of confusion is avoided and more objective assessments of IH can be made.

“ClinicalTrials.gov” (2015) reported specific effects of IH over and above placebo in spite of the fact that all participants received consultations and homeopaths were blinded. This may have been due, at least in part, to the shorter duration of this study, and that participants only met their homeopath twice.

A recent systematic review of randomised placebo-controlled trials of IH with homeopathic consultations in both groups reported that medicines prescribed in such studies may have small, specific treatment effects (Mathie et al., 2014).

 

4.1. Recommendations

 

However, in agreement with recommendations by Jacobs et al. (2005) it is suggested that in trials which testing complex homeopathy and IH subjects should be randomised to the IH group in one arm and the complex group in another and results of each arm should be compared with placebo.

 

Sub-group analysis

For studies with participating women taking tamoxifen or newer oestrogen inhibitors we suggest sub-group analyses to assess their possible antidoting effect of homeopathic remedies.

Pragmatic studies

To adequately examine the therapeutic context of the whole individualised homeopathic approach we need more pragmatic trials as described by Verhoef et al. (2005).

 

4.2. Conclusions

 

To make definite conclusions of the effect of IH and complex homeopathy on menopause-related symptoms we need sufficiently powered studies. Small and specific treatment effects were found. We agree with Thompson et al. (2005) that studies on IH with homeopathic consultations for the active and placebo groups should take enhanced context effects into account in power calculations.

 

4.3. Limitations

 

Only a small number of trials was included in the review that differed in type of homeopathic interventions, sample size and duration. Review of results are limited by the inclusion of studies with small samples. In the translation of the paper of Bekkering et al. (1993) minor errors cannot be excluded.

 

4.4. Funding sources

 

None.

4.5. Conflicts of interest

 

No conflicts of interest declared.

 

 

REFLECTION ON THE PROCESS OF CONDUCTING A SYSTEMATIC REVIEW

 

It was a great challenge for me to conduct a systematic review following the rigorous structured format demanded in the PRISMA guidelines (Liberati et al., 2009). Before starting with the systematic review I wondered if this rigid structure would limit my creative flow of writing. However, during the process I found that the 27-item PRISMA checklist was very helpful since it clearly indicated what was important to examine. It helped me not to lose myself in secondary details and to avoid writing a review of excessive length. Now I believe that my systematic review is clearly structured and easy to read.

 

Systematic reviews in homeopathy

To undertake a systematic review of high quality, with the framework of current EBM standards, I only included studies of homeopathy with a control. As a consequence, not all of the studies included represented the real world practice of classical homeopaths working with IH. Some studies with positive results had to be excluded. For this systematic review I found studies that tested complex homeopathy for the treatment of menopause-associated symptoms that met the prerequisites and thus they had to be included in the analysis. Such studies fall under the umbrella of ‘homeopathy’, but they do not meet the gold standard of homeopathic care which is, individualised treatment by a homeopath, with in-depth interviews, to find the matching homeopathic remedy to the totality of symptoms (Blackstone, 1993). For me it was confusing to consider such studies as being of the same value as studies investigating IH. Other researchers undertaking systematic reviews in homeopathy should be aware that studies testing complex or clinical homeopathy have to be included, potentially creating flaws in the result. Most reviews do not make any distinction between the different forms of homeopathy. My advice for future reviewers would be to clearly define the various homeopathic approaches and to present the results in this light. Hopefully, in doing so, more and more pragmatic trials investigating the holistic effect of IH will be conducted.  

 

Excluded studies of IH showing evidence of treatment effects

Four studies of IH representing real world practice with very positive effects had to be excluded because they lacked a control. In all studies participants had homeopathic consultations and follow-ups.

1.    The prospective observational study by Thompson and Reilly (2003) investigating IH in the management of symptoms of oestrogen withdrawal in women with breast cancer. Tamoxifen-users (55%) were included. 44% of the participants were on medication other than tamoxifen, including antidepressants and clonidine. The study duration was 3 years (June 1997 to June 2000). Significant improvements were seen in symptom scores and the primary endpoint ‘the effect on daily living scores’, as well as for quality of life and quality of health as measured by the EORTC-QOL score. The high levels of anxiety and depression improved significantly over the study period. Satisfaction with the homeopathic treatment was high. In tamoxifen-users significant improvements were seen as well, especially in hot flushes. After this overall positive study of IH by Thompson and Reilly was completed, a double-blind RCT of the homeopathic approach for the same group of patients was undertaken in cooperation with other authors (Thompson et al., 2005). This study was discussed in my systematic review.

 

2.    The pilot study by Clover and Ratsey (2002). This study tested IH in the treatment of hot flushes. The effect of the homeopathic approach was tested on frequency and severity of hot flushes in women with, and without, a history of breast cancer. Tamoxifen and no-tamoxifen users were included. The duration of the study was one year. 73% of all patients reported clinically relevant improvement. Tamoxifen-users also showed overall benefit of IH, especially in the severity of hot flushes. However, compared with the other groups, they had lower scores in reduction of the frequency of hot flushes.

 

 

3.    The open, multi-national prospective, pragmatic and non-comparative observational study by Bordet et al. (2008). They evaluated homeopathic treatments for hot flushes and their effect on quality of life in menopausal women. 5% of patients were additionally prescribed non-homeopathic medication (notably minerals) and/or food supplements (notably soy-based). The study lasted 5.5 months. 99 homeopathic physicians from 8 countries took part and they remained totally free regarding to their prescriptions and treatment choices. They used single individual homeopathic remedies. The study suggested that homeopathic treatment for hot flushes in menopausal women is effective and revealed a significant reduction in the frequency of hot flushes by day and night and a significant reduction in the daily discomfort they caused. Ninety percent of the women reported disappearance or lessening of their symptoms. The additional consumption of minerals and/ or food supplements did not seem to affect the results.  

 

4.    The open, multicentre, prospective, observational trial by Nayak et al. (2011). They ascertained the usefulness of individual homeopathic therapy in the management of distressing symptoms encountered during climacteric years in women and also the changes in the levels of FSH and lipid profile after one year of treatment. The overall duration of the study was 4 years. The mean score of typical menopause-associated symptoms such as hot flushes, night sweats, palpitation, etc. at the end of the study was statistically significant reduced. Statistically significant results were also seen in serum levels of cholesterol, triglycerides, and very-low density lipoprotein. The reduction of FSH levels experienced in 98 of 175 patients was not statistically significant. Values of high-density lipoprotein and low-density lipoprotein were not statistically significant either.

 

The overall positive results of these 4 studies of individualised homeopathic treatment of menopause-related symptoms, offering external validity, could not be included in my systematic review since they did not meet requirements for inclusion. If they too had been pooled in the analysis the result of my systematic review would be much clearer and the evidence of homeopathy would appear in a much more positive light.  

 

The pilot ‘cohort multiple randomised controlled trial’ by Relton et al. (2012) offered very high external validity and as well as meeting high quality research standards since it had a control. However, the control group neither had homeopathic treatment, nor homeopathic consultations. Unfortunately, I had to exclude this paper since this comparator was not defined as inclusion criterion. IH was tested for 36 weeks to assess the clinical and cost effectiveness of treatment for women with menopausal hot flushes who cannot take HT. After random selection to the treatment group, patients were given information about the homeopathic intervention. Then patients were free to refuse it. As in routine real world healthcare, information about treatment was only given to patients being treated. Results showed that the pilot data for seven of the eight outcome measures at 36 weeks adjusted for baseline value favoured the treatment group. Only the MYMOP Wellbeing score outcome measure favoured the group without treatment. However, there were baseline differences between groups in severity of hot flushes. Another flaw was that ITT analysis also included those patients who did not accept the treatment as well as those who did. Authors stated this as potential type II error and pointed the need for CACE analysis which did not appear to have been undertaken.

This study, despite of obvious flaws in statistical analysis and baseline imbalances between groups, indicated some benefit from IH and offered high external validity. Inclusion of this trial in my systematic review would also have strengthened the evidence of IH in the treatment of menopause-related symptoms. 

 

What did I learn from the process?

It was very interesting to read and analyse so many differently designed studies on various kind of homeopathy. I started to ask myself why in research individualised, complex, clinical and isopathic homeopathy is regarded as one and the same issue. I thought about ‘What is homeopathy?’ When reflecting on this question I believe that to make research in homeopathy clearer and more distinct, researchers should start to think about the deeper meaning of the term ‘homeopathy’. I consider the definition of Relton et al. (2008) as a very good one which suggests that in research the term ‘homeopathy’ should only be used to the whole ‘therapeutic system of homeopathy’. Thus, such confusion could be avoided.

In agreement with Weatherley-Jones et al. (2004) I found that studies deemed to be of higher quality in terms of the EBM hierarchy of evidence tended to show less-significant results than studies of lower quality. I consider this issue a paradox. In studies without control, which are regarded as being lower quality, even participants using allopathic medicines showed improvements when using IH (Thompson & Reilly, 2003; Clover & Ratsey, 2002). It seems that there is still a gap between high quality research and an appropriate investigation of the whole-systems healthcare of homeopathy. After having conducted this systematic review I agree more than ever with Weatherley-Jones et al. (2004) that more relevant and meaningful pragmatic studies of comparative effectiveness need to be designed to analyse the whole homeopathic approach.

 

 

Reference list:

Bekkering, G. M., Van den Bosch, W. J., & Van den Hoogen, H. J. (1993). Bedriegt schone schijn? Een onderzoek om de gerapporteerde werking van een homeopathisch middle te objectiveren. Huisarts Wet, 36(12), 414-415.

Bell, I., Lewis, D., Brooks, A., Schwartz, G., Lewis, S., Walsh, B., & Baldwin, C. (2004). Improved clinical status in fibromyalgia patients treated with individualized homeopathic remedies versus placebo. Rheumatology, 43(5), 577-582.

Blackstone, V. (1993). Single or multiple prescribing – a debate. British Homeopathic journal, 82(1), 37-52.

Boland, A., Cherry, M. G., & Dickson, R. (2014). Doing a systematic review: A student’s guide. London, United Kingdom: Sage.

Bordet, M. F., Colas, A., Marijnen, P., Masson, J.L., & Trichard, M. (2008). Treating hot flushes in menopausal women with homeopathic treatment: Results of an observational study. Homeopathy, 97(1), 10-15.

Canadian Consumers Centre for Homeopathy. (2011). Retrieved from http://www.homeocentre.ca/

Castelo-Branco, C. (2009). International Menopause Society. Retrieved from http://www.imsociety.org/

ClinicalTrials.gov. (2015). Retrieved from https://clinicaltrials.gov/ct2/results?term=homeopathy+AND+menopause&Search=Search

Clover, A., & Ratsey, D. (2002). Homeopathic treatment of hot flushes: A pilot study. Homeopathy, 91(2), 75-79.

Colau, J., Vincent, S., Marijnen, P., & Allaert, F. (2012). Efficacy of a non-hormonal treatment, BRN-01, on menopausal hot flashes. Drugs in R and D, 12(3), 107-119.

Cooper, K. L., & Relton, C. (2010). Homeopathy for insomnia: A systematic review of research evidence. Sleep Medicine Reviews, 14(5), 329-337.

De Gruyter, W. (1994). Pschyrembel: Klinisches Wörterbuch (257th ed.). Berlin, Germany: Wörterbuch-Redaktion.

Dixon, J. M. (2003). Hormone replacement therapy and the breast. Surgical Oncology, 12(4), 251-263.

Eyles, G., Leydon, G. M., & Brien, S. B. (2012). Forming connections in the homeopathic consultation. Patient Education and Counseling, 89(3), 501-506.

Foxlee, R. (2014). The Cochrane collaboration: Cochrane central register of controlled trials (CENTRAL). Retrieved from http://www.chochrane.org/editorial-and-publishing-policy-resource/cochrane-central-register-controlled-trials-central

Glasziou, P., Irwig, L., Bain, C., & Colditz, G. (2003). Systematic reviews in health care: A practical guide (2nd ed.). Cambridge, United Kingdom: Cambridge University Press.

Groth, S., & Stolzenberg, R. (1995). Wechseljahre: Eine Broschuere zur Selbsthilfe (5th ed.). Berlin, Germany: Oktoberdruck.

Gupta, P., Sturdee, D. W., Palin, S. L., Majumder, K., Fear, R., Marshall, T., & Paterson, I. (2006). Menopausal symptoms in women treated for breast cancer: The prevalence and severity of symptoms and their perceived effects on quality of life. Climacteric, 9(1), 49-58.

Hahnemann, S. (1992). Organon der Heilkunst: Textkritische Ausgabe der 6. Auflage. Heidelberg, Germany: Haug Verlag.

Higgins, J. P., & Green, S. (2011). Cochrane handbook for systematic reviews of interventions version 5.1.0. The Cochrane collaboration. Retrieved from http://www.cochrane-handbook.org

Jacobs, J., Jimenez, L. M., Gloyds, S. S., Gale, J. L., & Crothers, D. (1994). Treatment of acute childhood diarrhea with homeopathic medicine: A randomized clinical trial in Nicaragua. Pediatrics, 93(5), 719-725.

Jacobs, J., Springer, D. A., & Crothers, D. (2001). Homeopathic treatment of acute otitis media in children: A preliminary randomized placebo-controlled trial. Pediatric Infectious Disease Journal, 20(2), 177-183.

Jacobs, J., Herman, P., Heron, K., Olsen, S., & Vaughters, L. (2005). Homeopathy for menopausal symptoms in breast cancer survivors: A preliminary randomized controlled trial. The Journal of Alternative and Complementary Medicine, 11(1), 21-27.

Jonas, W. B., & Jacobs, J. (1996). Healing with homeopathy. New York, NY: Warner Books.

Jones, K. L., & Buzdar, A. U. (2004). A review of adjuvant hormonal therapy in breast cancer. Endocrine-Related Cancer, 11(3), 391-406.

Kaptchuk, T. J. (2001). The double-blind, randomized, placebo-controlled trial: Gold standard or golden calf? Journal of Clinical Epidemiology, 54(6), 541-549.

Kassab, S., Cummings, M., Berkovitz, S., van Haselen, R., & Fisher, P. (2009). Homeopathic medicines for adverse effects of cancer treatments (Review). The Cochrane Library, ?(2), 1-44. doi: 10.1002/14651858.CD004845.pub2.

Katz, T. (1997). Homeopathic treatment during the menopause. Complementary Therapies in Nursing & Midwifery, 3(2), 46-50.

Kiener, C. (2000). Das MSD Manual (6th ed.). Muenchen/Jena, Germany: Urban & Fischer.

Kupferer, E. M., Dormire, S. L., & Becker, H. (2009). Complementary and alternative medicine use for vasomotor symptoms among women who have discontinued hormone therapy. Journal of Obstetric, Gynecologic, and Neonatal Nursing, 38(1), 50-59.

Liberati, A., Altman, D. G., Tetzlaff, J., Mulrow, C., Gotzsche, P. C., Ioannidis, J. P.,…Moher, D. (2009). The Prisma statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: Explanation and Elaboration. British Medical Journal, 339(b2700), doi: http://dx.doi.org/10.1136/bmj.b2700

Light, R. J., & Pillemer, D. B. (1984). Summing up: The science of reviewing research. Cambridge, MA: Harvard University Press.

Linde, K., Clausius, N., Ramirez, G., Melchart, D., Eitel, F., Hedges, L., & Jonas, W. (1997). Are the clinical effects of homeopathy placebo effects? A meta-analysis of placebo controlled trials. Lancet, 350(9081), 834-843.

Macías-Cortés, E. C., Aguilar-Faisal, L., & Asbun-Bojalil, J. (2013). Efficacy of individualized homeopathic treatment and fluoxetine for moderate to severe depression in peri- and postmenopausal women (HOMDEP-MENOP): Study protocol for a randomized, double-dummy, double-blind, placebo-controlled trial. (2013). Trials, 14(105), 1-8. doi:10.1186/1745-6215-14-105

MacLehose, H. (2014). The Cochrane collaboration: Cochrane database of systematic reviews (CDSR). Retrieved from http://www.cochrane.org/editorial-and-publishing-policy-resource/cochrane-database-systematic-reviews-cdsr

Mathie, R. T., Lloyd, S. M., Legg, L. A., Clausen, J., Moss, S., Davidson, J. R., & Ford, I. (2014). Randomised placebo-controlled trials of individualized homeopathic treatment: Systematic review and meta-analysis. Systematic Reviews, 3(142): 10.1186/2046-4053-3-142

Matthiessen, B. (2011). Homeopathy in healthcare: Effectiveness, appropriateness, safety, costs. Berlin Heidelberg, Germany: Springer-Verlag.

Meyer-Koenig, P. (1995). Leitfaden fuer den Umgang mit Q-Potenzen. Goettingen, Germany: Burdorf Verlag fuer homoeopathische Literatur.

Moerman, D. E., & Jonas, W. B. (2002). Deconstructing the placebo effect and finding the meaning response. Annals of Internal Medicine, 136(6), 471-476.

Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G., & Prisma Group. (2009). Preferred reporting items for systematic reviews and meta-analyses: The Prisma statement. Annals of Internal Medicine, 151(4), 264-269.

Moreira, A. C., Silva, A. M., Santos, M. S., & Sardao, V. A. (2014). Phytooestrogens as alternative hormone replacement therapy in menopause: What is real, what is unknown. The Journal of Steroid Biochemistry and Molecular Biology, 143(9), 61-71.

Nayak, C., Singh, V., Singh, K., Singh, H., Gupta, J., Lamba, C. D.,…Luxmi, K.S. (2011). Management of distress during the climacteric years by homeopathic therapy. The Journal of Alternative and Complementary Medicine, 17(11), 1037-1042.

Nencioni, T., Ciammella, M., Bonaccorsi, G., Morano, D., Viglino, S., Grassini, E.,…Petronio, P. (1999). The effect of menopause on blood lipid and lipoprotein levels. Artherosclerosis, 147(1), 147-153.

Nelson, H. D., Kimberley, K. V., Haney, E., Rongwei, F., Nedrow, A., Miller, J.,…Humphrey, L. (2006). Nonhormonal therapies for menopausal hot flashes: A systematic review and meta-analyis. JAMA, 295(17), 2057-2071.

Northrup, C. (1995). Frauenkoerper Frauenweisheit(2nd ed.). Muenchen, Germany: Zabert Sandmann.

Penny, B. L., Razlog, R., Deroukakis, M., & Johnston, G. (2009). A comparative study of Klimakt-Heel and Femolene Ultra in the management of typical climacteric symptoms. Health SA Gesondheid, 14(1), 144-148.

PRISMA: Transparent reporting of systematic reviews and meta-analyses. (2014). Retrieved from http://www.prisma-statement.org/index.htm

Rada, G., Capurro, D., & Pantoja, T. (2010). Non-hormonal interventions for hot flushes in women with a history of breast cancer (Review). The Cochrane Library, ?(9), 1-65. doi:10.1002/14651858.CDOO4923.pub2

Relton, C., & Weatherley-Jones, E. (2005). Homeopathy service in a National Health Service community menopause clinic: Audit of clinical outcomes. The Journal of the British Menopause Society, 11(2), 72-73.

Relton, C., O’Cathain, A., & Thomas, K. J. (2008). “Homeopathy”: Untangling the debate. Homeopathy, 97(3), 152-155.

Relton, C., O’Cathain, A., & Nicholl, J. (2012). A pilot ‘cohort multiple randomised controlled trial’ of treatment by a homeopath for women with menopausal hot flushes. Contemporary Clinical Trials, 33(5), 853-859.

Reynolds, R. F., Obermeyer, C. M., Walker, A. M., & Guilbert, D. (2002). The role of treatment intentions and concerns about side effects in women’s decision to discontinue postmenopausal hormone therapy. Maturitas, 43(3), 183-194.

Rousseau, M. E. (2001). Evidence-based practice in women’s health: Hormone therapy for women at menopause. Journal of Midwifery & Women’s Health, 46(3), 167-180.

Sackett, D. L., Rosenberg, W. M., Muir Gray, J. A., Haynes, R. B., & Richardson, W. S. (1996). Evidence based medicine: What it is and what it isn’t. British Medical Journal, 312(71), doi: http://dx.doi.org/10.1136/bmj.312.7023.71

Seideneder, A. (1997). Mitteldetails der homoeopathischen Arzneimittel: Materia medica synthetica. Ruppichteroth, Germany: Similimum Verlag.

Shang, A., Huwiler-Muentener, K., Nartey, L., Jueni, P., Doerig, S., Sterne, J. A.,…Egger, M. (2005). Are the clinical effects of homeopathy placebo effects? Comparative study of placebo-controlled trials of homeopathy and allopathy. The Lancet, 366(9487), 726-732.

Smith, D. C., Prentice, R., Thompson, D. J., & Herrmann, W. L. (1975). Association of exogenous estrogen and endometrial carcinoma. The New England Journal of Medicine, 293(23), 1164 – 1167.

Swingler, G. H., Volmink, J., & Ioannidis, J. P. (2003). Number of published systematic reviews and global burden of disease: Database analysis. British Medical Journal, 327(7423), 1083-1084.

The Cochrane Collaboration. (2014). Retrieved from http://www.cochrane.org/

Thompson, E. A., & Reilly, D. (2003). The homeopathic approach to the treatment of symptoms of oestrogen withdrawal in breast cancer patients: A prospective observational study. Homeopathy, 92(3), 131-134.

Thompson, E. A., Oxon, B. A., Montgomery, A., Douglas, D., & Reilly, D. (2005). A pilot, randomized, double-blinded, placebo-controlled trial of individualized homeopathy for symptoms of estrogen withdrawal in breast-cancer survivors. The Journal of Alternative and Complementary Medicine, 11(1), 13-20.

Thompson, E. A., Barron, S. J., Mathie, R. T., Baitson, E. S., Berkovitz, S. R., Fisher, P.,…Roniger, H. (2008). Towards standard setting for patient-reported outcomes in the NHS homeopathic hospitals. Homeopathy, 97(3), 114-121.

Thompson, E. A., & Relton, C. (2009). Designing clinical trials of homeopathy for menopausal symptoms: A review of the literature. Menopause International, 15(1), 31-34.

Thompson, E. A. (2010). Alternative and complementary therapies for the menopause: A homeopathic approach. Maturitas, 66(4), 350-354.

Thompson, T. D. (2004). Can the caged bird sing? Reflections on the application of qualitative research methods to case study design in homeopathic medicine. BMC Medical Research Methodology, 4(4), 4-9.

Ullman, D. (1997). Homoeopathie fuer Kinder: Erkrankungen bei Kindern naturgemaess behandeln (11th ed.). Muenchen, Germany: Scherz.

Ullman, D. (2014). Homeopathic educational services: A condensed history of homeopathy. Retrieved from https://www.homeopathic.com/Articles/Introduction_to_Homeopathy/A_Condensed_History_of_Homeopathy.html 

University of York: Centre for reviews and dissemination. (2013). Retrieved from http://www.crd.york.ac.uk/prospero/

Utian, W.H. (1994). Ovarian function, therapy-orientated definition of menopause and climacteric. Experimental Gerontology, 29(3-4), 245-251.

Verhoef, M. J., Lewith, G., Ritenbaugh, C., Boon, H., Fleishman, S., & Leis, A. (2005). Complementary and alternative medicine whole systems research: Beyond identification of inadequacies of the RCT. Complementary Therapies in Medicine, 13(3), 206-212.

Vithoulkas, G. (1980). The science of homeopathy. New Delhi, India: B. Jain Publishers (P) Ltd.

Von Hagens, C., Schiller, P., Godbillon, B., Osburg, J., Klose, C., Limprech, R., & Strowitzki, T. (2012). Treating menopausal symptoms with a complex remedy or placebo: A randomized controlled trial. Climacteric, 15(4), 358-367.

Walach, H. (2000). Magic of signs: A non-local interpretation of homeopathy. British Homeopathic journal, 89(3), 127-140.

Wasilewski, B. W. (2004). Homeopathic remedies as placebo alternatives – verification on the example of treatment of menopause-related vegetative and emotional disturbances. Science and Engineering Ethics, 10(1), 179-188.

Weatherley-Jones, E., Thompson, E. A., & Thomas, K. J. (2004). The placebo-controlled trial as a test of complementary and alternative medicine: Observations from research experience of individualised homeopathic treatment. Homeopathy, 93(4), 186-189.

Wilken-Jensen, C., & Ottesen, B. (2003). The aging woman: The role of medical therapy. International Journal of Gynecology and Obstetrics, 82(3), 381-391.

Wilson, R. A., & Wilson, T. A. (1963). The fate of the non-treated postmenopausal woman: A plea for the maintenance of adequate oestrogen from puberty to the grave. Journal of the American Geriatric Society, 34(11), 1-12.

 

 APPENDIX 1, Table 1: Characteristics of included studies

 

Remedy/

Potency

Complex, anthroposo-phic; Apis regina tota 4X, Arg.met. 5X, Ovaria bovis 4X;

3x10 globuli/d, sublingual

Complex.

Presumably

Ambra 4X, Cimic. 6X, 

Lach.12X,

 Sang. 12X,

Sep. 6X, Zinc. 8X.

presumably 3x2 tablets/d.

Complex. Cimic. 4CH, Arn. 4CH, Glon. 4CH, Lach. 5CH, Sang. 4CH. 2 tablets/d

Duration/

follow-up

36 weeks, 3  treat-ment periods à 12 weeks without washout periods/

12,24,36 weeks

4 months, 2 months verum, 2 months placebo/

no infor-mation

12 weeks (after collection of baseline data)/only information on data collection

Outcome

Measure

Primary: MRS II from baseline to 12 weeks.

Secondary:

Changes in 3 subscores of MRS II after 12 weeks,

MRS II after 24 + 36 weeks

BMI

Primary: HFS

Secondary: HFRDIS, MRS, VAS, Morisky-Green score

Meno-pausal status

Postmeno-pausal 62,8%;

Perimeno-pausal

37,2%

Climac-teric

Amenorr-hoea for more than 12 months = postmeno-pausal

Age range

Mean age: 53.5 years

(SD 4.5)

No infor-mation

50 years or older

Sample size

102 women, 3 groups (RRP=33, PRR=35, RPR=34)

5 women

108 women, 2 groups à 54

Control

Identical placebo made of saccharose only

Placebo similar to verum (similar alcohol percentage)

Placebo identical to verum, including only saccharose (75%), lactose (24%), magnesium stearate E572(1%)

Study design

Randomised, placebo-controlled, double-blind, three-armed

Double-blind, placebo-controlled semi cross-over survey

Multicentre, randomised, double-blind, placebo-controlled

Study setting

University Women’s

Hospital, Heidelberg, Germany

General practice, Rheden, Netherlands, presumably of Bekkering

In 35 active centres in France (gynecolo-gists in private practice)

Authors/

Source

Hagens et al./

Climacteric 2012, 15(4),

358-367.

Bekkering et al./

Huisarts Wet 1993, 36(12), 414-415.

Colau et al./ Drugs in R and D 2012, 12(3), 107-119.

Remedy/

Potency

Single remedy: IH, individual dosage.

Complex: Amyl nitrate 3X, Sang. 3X, Lach. 12X, 1 tablet 3x/d  

Single remedy: IH/

individual dosage

Single remedy: IH, individual dosage. Fluoxetine

( = SSRI) 20mg/d

Duration/

follow-up

1 year/

homeo-pathic consul-

tations at

2 months

intervals

(all patients)

16 weeks/

4 follow-up visits (homeo-pathic) at 4-weeks intervals

(all patients)

6 weeks/

homeo-pathic consul-tations at week 4 (all patients)

 

Outcome

Measure

Primary: HFSS (frequency, times, severity).

Secondary:

Total no of hot flashes, KMI, SF-36 QoL, FSH level

Primary: MYMOP.

Secondary:

MSQ, EORTC QLQ-C30,

patient diaries,

HADS, FAQ, GHHOS

Primary:

HRSD.

Secondary:

BDI, responder rates, GS, remission rates.

Meno-pausal status

Meno-pause due to car-cinoma in situ or Stage I-III breast cancer (inclusion of tomoxifen users)

Meno-pause due to oestrogen withdrawal in breast cancer

(inclusion of tamoxifen- users)

Meno-pausal and postmeno-pausal disorders with moderate to severe depression

Age range

Mean age: 55.5 years

Mean age: 52.6 years

Mean age: 49 years (SD 5.8)

Sample size

83 women (complex or combination n=30, single remedy n=26, placebo n=27)

53 women (intervention group n=28, placebo group n=25)

133 women.

Group 1 = IH n=44, group 2 = Fluoxetine n=46, group 3 = placebo n=43

Control

Placebo identical in taste, appearance and odour to verum

Placebo identical to verum

Placebo identical to verums (IH, fluoxetine)

Study design

Randomised, double-blinded, placebo-controlled

Pilot, randomised, double-blinded, placebo-controlled

Randomised, placebo-controlled, double-blind, double-dummy, three-arm,  superiority trial

Study setting

Private medical clinic, Seattle, WA

Outpatient department of a NHS homeopa-thic hospital

(Glasgow Homeopa-thic Hospital)

Juárez de México Hospital in the outpatient service of homeopathy

Authors/

Source

Jacobs et al./

The Journal of Alternative and Complement-ary Medicine 2005, 11(1), 21-27.

Thompson et al./The Journal of Alternative and Complement-ary Medicine 2005, 11(1), 13-20.

Macías-Cortés et al./Trials 2013, 14(105), 1-8 (protocol).

clinicaltrials.

gov/ (results)

 

 

Abbreviations list:

Arg.met. – Argentum metallicum

Arn. – Arnika montana

BDI – Beck Depression Inventory

BMI – Blatt Menopausal Index

CH – Centesimal Hahnemannian

Cimic. – Cimicifuga racemosa

d – day

EORTC QLQ C30 – European Organisation for Research and Treatment of

                                                Cancer Quality of Life Questionnaire

FAQ – Final Assessment Questionnaire

FSH – Follicle Stimulating Hormone

GHHOS – Glasgow Homeopathic Hospital Outcome Scale

Glon. – Glonoinum

GS – Greene’s Scale

HADS – Hospital Anxiety and Depression Scale

HFRDIS – Hot Flash Related Daily Interference Scale

HFS – Hot Flash Score

HFSS – Hot Flash Severity Score

HRSD – Hamilton Rating Scale for Depression

IH – Individualised Homeopathy

KMI – Kupperman Menopausal Index

Lach. – Lachesis muta

mg – milligram

MRS – Menopause Rating Scale

MRS II – Menopause Rating Scale II

MSQ – Menopausal Symptom Questionnaire

MYMOP – Measure Yourself Medical Outcome Profile

NHS – National Health Service

P – Placebo

R – Remedy

Sang. – Sanguinaria canadensis

SD – Standard Deviation

Sep. – Sepia officinalis

SF-36 QoL – SF-36 Quality of Life Score

SSRI – Selective Serotonin Reuptake Inhibitor

WA – Washington

VAS – Visual Analogue Scale

Zinc. – Zincum metallicum

 

 

          

APPENDIX 2, Table 2: The Cochrane Collaboration’s tool for assessing risk of bias                                                      

Hagens et al. (2012)

Domain

Review authors’ judgement

Support for judgement

Random sequence generation

(selection bias)

 

Low risk.

102 women were randomised to

treatment groups by assignment to the next patient number of the appropriate stratum. Quote: “Randomisation was stratified by menopausal status (post-menopause versus peri-menopause).”

Allocation concealment

(selection bias)

 

Low risk.

Quote: “Lists for stratified randomised allocation to the three treatment groups with block length of 6 were created by an independent biometrician.”

Blinding of participants and

personnel

(performance bias)

 

Low risk.

Quote: “double-blind.” Appearance of verum and placebo identical. Lists for stratified randomised allocation, created by an independent biometrician, were sent to the manufacturer who packed and labelled the trial medication with batch numbers, ascending patient numbers, treatment weeks, respectively for the two strata. Unblinding of patients did not occur throughout the trial.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double blind.” Comment: Probably done and see comment under blinding of participants and personnel above.

Incomplete outcome data

(attrition bias)

 

Unclear risk for secondary outcome measure only.

The main analysis of the primary efficacy endpoint was done based on the full analysis set (FAS) involving all randomised patients that received at least one dose of the study medication and had the primary efficacy endpoint documented. After allocation 6/102 women withdrew consent and were not included in the analysis of primary outcomes. Further 24/102 were lost for analysis of combined treatment effects (reasons: lost to follow-up, discontinuation of intervention, exclusion from analysis/combined treatment effects). Altogether 30/102 women were lost for combined treatment effects = secondary outcomes. The calculation of the combined treatment effect was the mean (standard deviation, SD) calculated on the basis of patients for whom data at baseline as well as the 3rd period were available. Missing outcome data for analysis of combined treatment effects were unbalanced in numbers across the 3 groups. All 3 groups had 1 placebo period and 2 remedy periods in different sequences. Thus withdrawals are not listed separately for each group.  

Selective reporting

(reporting bias)

 

Low risk.

All of the outcomes described in the methods were reported.

 

Other sources of bias.

 

 

 

 

Low risk.

 

The study appears to be free of other sources of bias.

 

Total risk of bias within study.

 

Unclear risk of bias for one subdomain (attrition bias, secondary outcomes).

 

 

 

 

 

Bekkering et al. (1993)

Domain

Review authors’ judgement

Support for judgement

Random sequence generation

(selection bias)

 

.High risk.

In short intervals and independent of each other the 5 participants consulted Bekkering requesting a prescription of Famosan (the medication tested).

Allocation concealment

(selection bias)

 

Low risk.

A code was given to each packing (placebo/verum) by a colleague who kept the key to the code.

Blinding of participants and

personnel

(performance bias)

 

Low risk.

 

Quote: “double-blind.” The pharmacy manufactured a placebo remedy with an alcohol percentage similar to the homeopathic remedy. The coded packings were returned to the pharmacy in 5 clusters of 4 (2 homeopathic, 2 placebo). After the codes had been revealed, both the research group and the placebo group could be identified.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double-blind.”

See comment under blinding of participants and personnel.

Incomplete outcome data

(attrition bias)

 

Unclear risk.

Scores of 1/5 participants were not taken into account because of incomplete questionnaire.

Selective reporting

(reporting bias)

 

High risk.

Recording of climacteric complaints according to Blatt Menopausal Index (BMI) with specification of 13 menopausal complaints. Outcomes only presented in detailed manner for hot flashes and transpiration.

Other sources of bias

 

 

High risk.

The same participants took Famosan before baseline and had reported positive effects. They were no blank pages any more. The sequence of remedy/placebo (= 2 months each) among participants was not indicated. No distinction between primary and secondary outcomes.

Total risk of bias within study

High risk of bias for three key domains and unclear risk of bias for one key domain.

 

 

 

 

Colau et al. (2012)

Domain

Review authors’ judgement

Support for judgement

Random sequence generation

(selection bias)

 

Low risk.

Carried out centrally by Laboratoires Boiron and generated using the random function of SAS (version 9.2) software.

Allocation concealment

(selection bias)

 

Low risk.

See comment above. Allocation was equilibrated by blocks of 6 in that each investigator/centre was allocated 3 active treatment units and 3 placebo treatment units assigned in a double-blind, randomised fashion. Treatments were dispensed according to the randomisation list.

Blinding of participants and

personnel

(performance bias)

 

Low risk.

Quote: “double-blind.” Placebo tablets were identical in appearance to the active tablets. Laboratoires Boiron held the key to the randomisation list in a sealed envelope, which was not opened until the end of the study.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double-blind.” See comment above. The key to randomisation was used only after freezing of the database and finalisation of the statistical analyses.

Incomplete outcome data

(attrition bias)

 

Low risk.

54 participants were randomised in each group (active and placebo). Statistical analysis was carried out on the intention-to-treat (ITT) population, defined as all patients who took at least one dose of the study treatment and had at least one post-enrolment evaluation. In the active group there were 4/54 and in the placebo group 3/54 withdrawals before taking treatment. They were not included in the ITT sample.

 

Selective reporting

(reporting bias)

 

 

Low risk.

 

 

All the outcome measures mentioned have been reported in the pre-specified way.

Other sources of bias

 

 

Low risk.

The study appears to be free of other sources of bias.

Total risk of bias within study

Low risk of bias for all key domains.

 

 

 

 

Jacobs et al. (2005)

Domain

Review authors’ judgement

Support for judgement

Random sequence generation

(selection bias)

 

Low risk.

Quote: “Randomisation was done using computer-generated random numbers in blocks 4 and 6 and was known only to the homeopathic pharmacist.”

Allocation concealment

(selection bias)

 

Low risk.

Quote: “The code was not broken until after initial data analysis was completed (triple blinded)”. The homeopathic pharmacist randomised the subjects to one of the 3 treatment groups. See comment under random sequence generation.

Blinding of participants and

personnel

(performance bias)

 

Low risk.

Quote: “double-blinded.” Study medications (placebo, verum single remedy, verum combination remedy) were identical in taste, appearance, and odour and they were dispensed in identical containers. None of the homeopathic practitioners, study personnel, or co-investigators knew which subjects had been randomised to which group.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double-blinded.”

See comment above. Quote: “The code was not broken until after initial data analysis was completed (triple blinded).”

Incomplete outcome data

(attrition bias)

 

Low risk.

Although altogether 28/83 (33.7%) participants withdrew all randomised participants were analysed. 

Selective reporting

(reporting bias)

 

Unclear risk.

All the outcome measures mentioned have been reported in the pre-specified way. However, the reported results for combination remedy do not match with detail in figure 3.

Other sources of bias

 

Low risk.

The study appears to be free of other sources of bias.

Total risk of bias within study

Unclear risk of bias for one key domain.

 

 

 

 

Thompson et al. (2005)

Domain

Review authors’ judgement

Support for judgement

Random sequence

generation

(selection bias)

 

Low risk.

Patients were randomly assigned using a random number table kept by the pharmacy.

Allocation concealment

(selection bias)

 

Low risk.

See comment above.

Quote: “At no point could investigators foresee assignments and continued to be blinded throughout the trial.”

Blinding of participants

and personnel

(performance bias)

 

Low risk.

Quote: “double-blinded.”

Helios forwarded medicines directly to the patients’ homes ensuring blinding. Changed prescriptions were sent in accordance with initial randomisation. Placebo was identical to the active remedies with similar frequency of administration over the study period.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double-blinded. At no point could investigators foresee assignments and continued to be blinded throughout the trial.” See comment above.

Incomplete outcome

data

(attrition bias)

Low risk.

All 53 randomised participants were analysed on an intention-to-treat basis using appropriate regression models. Withdrawals were also analysed (intervention group 5/28; placebo group 3/25). Reasons for withdrawals were reported.

Selective reporting

(reporting bias)

 

Low risk.

All the outcome measures mentioned have been reported in the pre-specified way.

Other sources of bias

Low risk.

 

The study appears to be free of other sources of bias.

Total risk of bias within study

Low risk of bias for all key domains.

 

 

 

 Macías-Cortés et al. (2013)/”ClinicalTrials.gov” (2015)

Domain                               

Review authors’ judgement

Support for judgement

Random sequence

generation

(selection bias)

 

Low risk.

Simple randomisation in a 1:1:1 ratio using a computer-generated random allocation sequence, by a statistician not further involved in the study.

Allocation concealment

(selection bias)

 

Low risk.

See comment above. Assignment of participants in sequential order to the treatment groups. The randomisation list was sent to the research pharmacist at the start of the study who randomly delivered the treatment according to the allocation sequence in one of the 3 groups.

Blinding of participants and personnel

(performance bias)

 

Low risk.

Quote: “double-dummy, double-blind.” Placebos seem identical to their corresponding verum formulations. The randomisation list was kept strictly confidential. Participants, the homeopathic doctor, the psychologist and the statistician remained blinded to the identity of the 3 treatment groups until the end of the study.

Blinding of outcome

assessment

(detection bias)

 

Low risk.

Quote: “double-dummy, double-blind.” See comment above.

Incomplete outcome

Data

(attrition bias)

 

Low risk.

All participants under randomisation were included in the primary efficacy population (ITT population), regardless whether or not they adhered to the treatment protocol or provided complete data sets. Only patients who withdrew their consent to use their personal data were excluded from the analysis. No withdrawals of consent listed. Losts to follow-up were included in the analysis.

Selective reporting (reporting bias)

 

Low risk.

According to the protocol primary and secondary outcomes are to be measured at weeks 4 and 6. However, according to published results primary and secondary outcomes were only measured at week 6! Dr. Macías-Cortés has sent me an e-mail on 28.01.2015 with results after 4 and 6 weeks. Everything has been reported in the pre-specified way.

 

 

Other sources of bias

 

Low risk.

 

The study appears to be free of other sources of bias.

 

Total risk of bias within study

 

Low risk of bias for all key domains.

 

 

 

 

Abbreviations list:

MYMOP – Measure Yourself Medical Outcome Profile