Elsevier

Social Science & Medicine

Volume 57, Issue 12, December 2003, Pages 2325-2341
Social Science & Medicine

A review of psychosocial interventions in infertility

https://doi.org/10.1016/S0277-9536(03)00138-2Get rights and content

Abstract

Counselling has been strongly recommended by numerous governmental, medical and community associations to help infertile people. The purpose of this review was to determine whether psychosocial interventions improved well-being and pregnancy rates, and to identify the kinds of interventions that were most effective. A systematic search identified all published and unpublished papers in any language and any source that (1) described a psychosocial intervention and (2) evaluated its effect on at least one outcome measure in an infertile population. A total of 380 studies met the first criteria but only 6.6% (n=25) of these were independent evaluation studies. Analysis of these studies showed that psychosocial interventions were more effective in reducing negative affect than in changing interpersonal functioning (e.g., marital and social functioning). Pregnancy rates were unlikely to be affected by psychosocial interventions. It was also found that group interventions which had emphasised education and skills training (e.g., relaxation training) were significantly more effective in producing positive change across a range of outcomes than counselling interventions which emphasised emotional expression and support and/or discussion about thoughts and feelings related to infertility. Men and women were found to benefit equally from psychosocial interventions. Directions for future research on the evaluation of psychosocial interventions are discussed.

Introduction

The provision of psychosocial interventions for infertile couples has been recommended since the advocacy work of Barbara Eck Menning (1980) directed research attention to emotional distress as a consequence of infertility rather than, as had been the emphasis until then, a cause of infertility. Her recommendation to provide psychological services to infertile couples has been reiterated by regulatory bodies in several countries (Bruhat, 1992; Human Fertilisation and Embryology Authority (HFEA), 1995), various associations involved in the care of infertile couples both at a professional (cf. Hammer-Burns & Covington, 1999; Boivin & Kentenich, 2002) and community level (e.g., ISSUE, CHILD, RESOLVE) as well as those of numerous mental health professionals working with infertile couples (cf. Bresnick & Taymor, 1979; Menning, 1980). Moreover, the recommendation is consistent with the interest infertile people themselves have expressed in receiving more psychosocial help (Laffont & Edelmann, 1994; Sundby, Olsen, & Schei, 1994).

Despite widespread belief in the worthiness of such counselling, relatively few studies have evaluated the effectiveness of psychosocial interventions in the field of infertility. Moreover, to this author's knowledge no review exists of this research base. The scarcity of evaluative studies is a significant problem within the psychology research base especially now that interventions in medical settings are expected to be evidence-based (Sackett, Straus, Richardson, Rosenberg, & Haynes, 2000). A review is therefore timely. The aim of the present review was to examine the outcome studies that do exist with the intent of appraising this research and providing direction for future research on the evaluation of effective psychosocial interventions.

The review was organised around three main questions that have been the subject of debate in the infertility psychosocial literature. These were:

  • 1.

    Do psychosocial interventions improve well-being?

  • 2.

    Do psychosocial interventions increase pregnancy rates?

  • 3.

    Are some interventions more effective than others?

A systematic search identified the studies needed to answer these questions. In this type of methodology the search procedure attempts to identify all research evidence concerned with the primary question, in this case, the effectiveness of psychosocial interventions for infertility (Sackett et al., 2000). These could be any published or unpublished papers from any source in any language. Accordingly, both electronic and manual searches were used to identify studies that (1) described a psychosocial intervention and (2) evaluated the intervention.

The results are presented in the same order for each question. A description of study results is provided in terms of outcomes as well as the percentage of studies finding positive results. A commentary section discussing the results is then presented.

An initial broad search of the psychological and medical research base (i.e., PsycINFO, BIDS (International Bibliography for Social Sciences), Medline, EMBASE (Excerpta Medica Database) showed that 1,957 articles and books had been published on the psychological aspects of infertility since 1966 (i.e., title or abstract included the search terms psychology and infertility or their variants). This set of articles and books was then further examined to extract only those that included intervention, group, therapy, counselling (or variants thereof) in their title or abstract, and to exclude duplicates. Abstracts were read to identify those studies that evaluated the effectiveness of at least one psychosocial intervention on at least one outcome measure. In addition, the reference list of tagged papers was examined to identify further outcome studies that might not have been picked up by the electronic search. This secondary search yielded N=380 studies.

Of the 380 studies 345 (90.8%) were excluded from the review for various reasons. These were that the study (1) mentioned psychosocial intervention(s) but did not include any evaluation (n=290, 76.3%); (2) described a case study which could not be generalised (n=47, 12.4%); (3) evaluated non-specific patient-centered care delivered as part of routine care (e.g., “tender loving care”, Stray-Pedersen & Stray-Pedersen, 1984) (n=5, 1.3%) and; (4) results could not be interpreted due to the lack of statistical or other methodological detail (n=3).

The final sample consisted of 35 studies, of which only 25 were independent evaluations on separate populations. This set of independent studies represents 6.6% of the potential pool of 380 studies. Table 1 provides summary details for the 25 independent studies selected. The studies were grouped according to the type of intervention evaluated. The three basic categories were (1) counselling interventions; (2) focussed educational interventions, and (3) comprehensive educational programmes. The feature that distinguished educational programmes (focussed or comprehensive) from counselling was the therapeutic objective. If the main aim of the intervention was to impart knowledge or provide skills training then the intervention was classed as educational. If, in contrast, the main aim of the intervention was emotional expression and support, and/or discussion of thoughts and feelings related to infertility (as cause or consequence) then the intervention was classed as counselling. The difference between focussed and comprehensive interventions was in the range of information or skills training provided to participants with focussed interventions providing one main skill (e.g., coping or relaxation training) and comprehensive programmes providing a range (e.g., coping and relaxation training). It is acknowledged that the categories were not wholly independent and that information could be provided in counselling interventions and/or emotional expression in educational programmes. However, the categories were sufficiently different in their emphasis with respect to these therapeutic interventions to warrant a separation.

Table 1 shows selected characteristics of the interventions. As shown in Table 1 the counselling interventions could further be sub-divided into three types. The first and oldest types evaluated the effects of both short and long-term psychoanalytic or psychodynamic psychotherapy used to alleviate psychic conflicts, often originating in childhood, believed to be blocking pregnancy. The second group of studies used ‘infertility counselling’ which focused more directly on reactions to infertility and on discussions about the impact of infertility on various domains, for example, marital and sexual relations or on feelings of masculinity and femininity. The final set of counselling studies followed the theoretical work of Beck (1976) and aimed to identify, reality-test and correct distorted cognitions and beliefs about infertility.

The second category of interventions listed in Table 1 focussed on educational programmes which mainly incorporated one or two educational activities including coping training, stress reduction, sex therapy and receiving preparatory information about medical tests or treatments. The final category of studies was also educational but assessed the effectiveness of more comprehensive and structured educational psychosocial interventions. For example, the Behavioral Medicine Program for Infertility (BMPI, also known as the mind/body program) (Domar, Seibel, & Benson, 1990) is a 10-week group program that includes, for example, cognitive-restructuring, methods for emotional expression, relaxation training, nutrition and exercise. Similarly, the intervention designed by Clark and colleagues (Clark et al., 1995; Clark, Thornley, Tomlinson, Galletley, & Norman, 1998) for obese anovulatory women includes not only education about nutrition and medical topics but also group support and stress management techniques.

In addition to the difference in the type of intervention, the studies presented in Table 1 also differed with respect to other intervention characteristics. Duration described the number of weeks necessary to carry out the intervention. While educational interventions were time-limited, many of the counselling interventions were open-ended with duration determined by the needs of the client. For these interventions an average number of weeks was recorded in Table 1. It should be noted that the number of weeks in all but two studies (i.e., Holzle, Brandt, Lutkenhaus, & Wirtz, 2002; Takefman, Brender, Boivin, & Tulandi, 1990) was equal to the number of sessions. About the same number of studies delivered the intervention to individuals (in a group or individual format) (n=11) with slightly fewer studies using a couple format (n=13). One study presented data for both individual and couple interventions (Strauss, Hepp, Stading, & Mettler, 2002). When the intervention was directed to individuals it was almost always delivered to women only. Finally, sample size differed depending on the study with the range between 10 and 134 people in intervention studies versus 5–48 people in the comparison group.

According to evidence-based medicine, the methodology that yields the least biased evidence for the effectiveness of an intervention is the randomised controlled trial where consecutive patients are randomly assigned to experimental and control conditions (Khan, Riet, Popay, Nixon, & Kleijnen, 2001). This type of design is effective because it controls for non-specific factors that may influence the responses of treated and untreated groups on outcome measures. Randomisation and the use of control groups ensures that differences between groups on outcome measures are due to intervention effects rather than to other factors not controlled as part of the experiment.

None of the studies reviewed met all the criteria of evidence-based medicine and almost all showed some shortcomings with regard to good practice for evaluation studies. Randomisation protocols were not adhered to in several studies (e.g., Domar, Clapp, Slawsby, Dusek, Kessel, & Freizinger, 2000a; Domar, Clapp, Slawsby, Kessel, Orav, & Freizinger, 2000b; McQueeney, Stanton, & Sigmon, 1997; Takefman et al., 1990) and/or high refusal and/or attrition rates compromised randomisation when it was used (Wischmann et al., 2001a; Connolly et al., 1993; Clark et al., 1995). Although many studies used a comparison group it often comprised of people who refused to participate, dropped out of treatment or were fertile (McQueeney et al., 1997; Clark et al. 1995; O’Moore, O’Moore, Harrison, Murphy, & Carruthers, 1983). These were not adequate controls because such individuals may differ markedly from participants. Indeed, four of five studies comparing participants who withdrew to those who continued found pre-treatment differences on demographic and/or psychological variables (Clark et al., 1995; Stewart et al., 1992; Domar et al., 2000b; Holzle et al., 2002). Almost all studies used at least one ‘in-house’ questionnaire designed specifically for the study but for which they provided no psychometric evaluation. In 16% of studies this was the only measure used to assess outcomes. Together these methodological shortcomings indicated the need to take into account the quality of studies when reviewing study findings.

Table 2 shows methodological characteristics of the studies reviewed where information was available. Table 2 shows that the final sample represented, on average, about 59% of the sample initially recruited (or randomised) to the study groups (based on studies providing attrition information). The types of participant included in the intervention group were for the most part infertility patients or infertile people from the community. Of the 25 studies reviewed, 36% (n=9) did not use any comparison group, 44% (n=11) used a routine care, waiting list or other control group (e.g., support only) whereas the remainder compared study findings to women who did not participate in the intervention for various reasons (e.g., dropped out, refused to attend counselling). The majority of studies using control groups (81.8%, n=9 of 11 studies) used randomisation procedures to allocate participants to intervention or control groups. The majority of studies (60%, n=15) included a pre-to-post intervention comparison, though seven studies (28%) used only a single assessment (i.e., post-only) to evaluate the intervention. Finally, the majority of studies used both validated questionnaire measures as well as questionnaires developed specifically for the study (i.e., in-house measures).

In light of the variability in methodological characteristics, it was decided that review findings should be considered not only for the group of studies as a whole but also for those studies which were deemed of ‘better quality’. Studies of better quality were defined as studies those that used a control group, either routine care, wait list or support-only controls, and that used either random assignment and/or a pre-to-post design to account for the influence of uncontrolled factors on intervention effects. In total 11 of the 25 studies (44%) reviewed could be said to meet these requirements. Nine studies (36%) included random assignment to a control or intervention group (Connolly et al., 1993; Domar, Clapp, Slawsby, Dusek, Kessel, & Freizinger (2000a), Domar, Clapp, Slawsby, Kessel, Orav, & Freizinger (2000b); Emery et al., 2001; Liswood, 1995; Sarrel & deCherney, 1985; Strauss, & Mettler, 2002; Takefman et al., 1990; Wallace, 1984; Wischmann et al., 2001a). The other two studies (8%) included a routine care or wait-list control group and a pre-to-post design (without random assignment) (Tuschen-Caffier, Florin, Krause, & Pook, 1999; Stewart et al., 1992). The findings of the review were reported in terms of the overall sample of 25 studies and again when only considering the findings of these 11 studies.

Table 3, Table 4, Table 5 present the results extracted from the studies selected. It was not possible to present results in terms of effect sizes, as is often the case in meta-analyses or systematic reviews because many of the studies did not provide sufficient statistical detail to be able to compute effect sizes.1 The only exception was for pregnancy rates as all studies provided the number of people who achieved a pregnancy in intervention versus comparison groups (when used). Consequently, effect sizes were presented for pregnancy data. For the remaining variables the Tables show whether an outcome was assessed (i.e., presence of a square) and if it was, whether the outcome showed a positive intervention effect (i.e., filled square) or a non-significant intervention effect (i.e., unfilled square). Positive interventions were defined as significant pre-to-post comparisons and/or significant experimental versus control group comparisons which favoured interventions. In none of the studies was a negative intervention effect (i.e., deterioration) observed. Studies identified as better quality were indicated in these tables by the use of superscript ‘a’ beside the author name.

Section snippets

Are psychosocial interventions beneficial to well-being?

A wide range of measures was used to assess the effect of psychosocial interventions on well-being and the findings depended to a certain extent on the outcome measure used.

Do psychosocial interventions increase pregnancy rates?

In total 15 studies investigated pregnancy as an outcome variable. However, all but one study (Clark et al., 1995) sampled individuals currently undergoing fertility treatment introducing the possibility that pregnancies were due to medical treatment rather than the psychosocial intervention. Consequently, in this section only the better quality studies assessing impact on pregnancy (n=8 of 11 studies) and using a control group were examined.2

Are some interventions more effective than others?

The original intention in this section was to compare individual interventions. However, this was not possible because too few replications of the same intervention (e.g., coping training, sex therapy) existed and in fact, only one intervention was evaluated more than once on independent samples (i.e., Domar et al., 1990, mind/body program). As a result of the limited replications, this question addressed the broader intervention categories, that is, counselling versus educational

Recommendations for future research

The main way to advance research in this area would be to carry out well-controlled studies that would evaluate the effectiveness of the intervention features identified in this review as being potentially effective. The review suggested that the more successful interventions lasted between 6 and 12 weeks with a follow-up period of at least 6 months and had a strong educational and skills training component and/or group format which emphasised medical knowledge and acquisition of stress

Conclusion

Thirty years of research has produced 25 independent studies evaluating psychosocial interventions for infertile people and of these, only eight met minimum requirements for good quality studies. The analyses carried out in this review have revealed the limited nature of this evidence base. Despite the lack of sound evaluation, the same time period saw the publication of almost 400 papers strongly recommending psychosocial interventions. Therefore, the main take home message from this review,

Acknowledgements

I gratefully acknowledge Tewes Wischmann and Bernhard Strauss who helped find and translate German studies and Ulrike Hahn who, additionally, commented on a first draft of this paper. Many thanks also to Deborah Lancastle for help in identifying potential studies and to Janet Takefman who also commented on the first draft of the manuscript.

References (66)

  • A.M O’Moore et al.

    Psychosomatic aspects in idiopathic infertilityEffects of treatment with autogenic training

    Journal of Psychosomatic Research

    (1983)
  • P.M Sarrel et al.

    Psychotherapeutic intervention for treatment of couples with secondary infertility

    Fertility and Sterility

    (1985)
  • B Stray Pedersen et al.

    Etiologic factors and subsequent reproductive performance in 195 couples with a prior history of habitual abortion

    American Journal of Obstetrics and Gynecology

    (1984)
  • F van Balen et al.

    Assessment of a telephone helpline on infertility provided by a patient association

    Patient Education and Counselling

    (2001)
  • L.M Wallace

    Psychological adjustment to and recovery from laparoscopic sterilization and infertility investigation

    Journal of Psychosomatic Research

    (1985)
  • A.T Beck

    Cognitive therapy and the emotional disorders

    (1976)
  • H Bents

    Verhaltenstherapeutische Paartherapie bei Kinderwunschpatienten [Behavioral couple therapy for patients with infertility]

    Jahrbuch fur Medizinische Psychologie

    (1991)
  • B.J Berg et al.

    Psychiatric morbidity in the infertile populationA reconceptualization

    Fertility and Sterility

    (1990)
  • Bernstein, J., Potts, N., & Mattox, J. (1985). Assessment of psychological dysfunction associated with infertility....
  • J Boivin et al.

    ESHRE monographsGuidelines for counselling in infertility

    (2002)
  • J Boivin et al.

    Why are infertile patients not using psychosocial counselling?

    Human Reproduction

    (1999)
  • J Boivin et al.

    The effects of female sexual response in coitus on early reproductive processes

    Journal of Behavioral Medicine

    (1992)
  • A Bowling

    Measuring healthA review of quality of life measurement scales

    (1997)
  • K.H Brandt et al.

    Auswirkungen von Kurzzeitpsychotherapie auf den Erfolg in einem in-vitro-Fertilisierung. [effects of short-term psychotherapy on the success of an in vitro fertilization/embryo transfer program]

    Wiener Medizinische Wochenschrift

    (1991)
  • E Bresnick

    A holistic approach to the treatment of the crisis of infertility

    Journal of Marital and Family Therapy

    (1981)
  • M.A Bruhat

    Recommendations of the French College of Obstetricians and Gynaecologists of the diagnosis, treatment, cost and results of the treatment of infertility health services

    Human Reproduction

    (1992)
  • G Christie et al.

    Individual and group psychotherapy with infertile couples

    International Journal of Group Psychotherapy

    (2000)
  • A.M Clark et al.

    Weight loss results in significant improvement in pregnancy and ovulation rates in anovulatory obese women

    Human Reproduction

    (1995)
  • A.M Clark et al.

    Weight loss in obese infertile women results in improvement in reproductive outcome for all forms of fertility treatment

    Human Reproduction

    (1998)
  • J Cohen

    A power primer

    Psychological Bulletin

    (1992)
  • K.J Connolly et al.

    An evaluation of counselling for couples undergoing treatment for in-vitro fertilization

    Human Reproduction

    (1993)
  • A.D Domar et al.

    The impact of group psychological interventions on distress in infertile women

    Health Psychology

    (2000)
  • A.D Domar et al.

    Healing mind, healthy womanUsing the mind/body connection to manage stress and take control of your life

    (1996)
  • Cited by (281)

    • Infertility

      2023, Encyclopedia of Mental Health, Third Edition: Volume 1-3
    View all citing articles on Scopus
    View full text