Background

One in eight Australian women will develop breast cancer and approximately 30% will require mastectomy [1]. Having the option of breast reconstruction (BR) has a positive impact on the quality of life of many women with breast cancer, including improved psychological well-being, body image, vitality, femininity and sexuality [2–9]. A few studies, however, have reported either no significant differences in particular outcomes for women with or without BR [10, 11], or significant aesthetic dissatisfaction with BR outcomes in women with post-operative complications [12].

Not all women requiring or choosing mastectomy will want BR [13] and individual women have their own reasons for choosing immediate BR (IBR), delayed BR (DBR) or no BR (NBR). Several studies have investigated the predictors of BR. Factors associated with an increased likelihood of BR uptake in Australia include higher socioeconomic status and levels of health insurance coverage [14], less remote geographic location [15], younger age, non-Indigenous status and less co-morbidity [16], higher levels of education, having children, not requiring radiotherapy [17] and less distance to travel [18]. In other nations, studies have associated higher rates of BR with a stronger self-concept of body image (France [19] and Switzerland [20]), a focus on appearance (Hong Kong [21]), younger age (United States of America (USA) [22], United Kingdom (UK) [23] and Switzerland [20]), depression (UK [23]), the role of partners (Japan [24]), and the influence of the surgeon (England [25]). Being of white race has been shown to be strongly associated with increased uptake of BR in the United States [26–28].

Understanding these predictors is useful for planning health services, but a more personal understanding of the decision making process is needed to support individual women to make a choice that is most aligned with their own values and needs so they can minimise the likelihood of decisional regret [29, 30]. Women considering BR following mastectomy for breast cancer face a potentially difficult choice that involves weighing up oncological, practical and personal factors at a very stressful time. Some women will want to make this decision on their own, others with the help of loved ones or clinicians [21, 31–33]. Some women will find the decision much easier than others [34, 35], some may not want to discuss all options [20, 25] and some may have unrealistic expectations [36, 37].

This systematic review has two aims. The first is to explore women’s preferences regarding BR. Understanding how and why women come to their decision will help health professionals better support them through the decision-making process. It will also help to inform the development of specific educational strategies for consumers and health professionals (including plastic, breast and oncoplastic breast surgeons and breast care nurses) who discuss the options with patients undergoing mastectomy.

The second aim is to test the reliability of a set of domains used in a previous study to categorise women’s reasons for their choice of BR [37]. This review assessed the suitability of these domains for a much larger and more diverse population of women considering BR following mastectomy for breast cancer.

Methodology

This study is part of a larger project investigating various aspects of BR. The literature search, therefore, was initially broad and for this specific study was narrowed to articles relevant to decision-making/reasons for choice.

The following databases were searched: Australasian Medical Index; ATSI Health; Cochrane Central Register of Controlled Trials; Cochrane Database of Systematic Reviews; Cochrane Database of Effect; Cochrane Health Technology Assessment; Embase, Health & Society; Informit Health Collection; Medline; Medline In-Process (Premedline); PsycInfo; Rural Health. Articles published up until 29 February 2016 were identified using the following search criteria: Initial general search terms were: (“breast cancer” OR “ductal carcinoma in situ” OR “mastectomy”) AND “breast reconstruction”. Specific search terms were as follows: “reason*” OR “choice*” OR “motivat*” OR “preference*” OR “access*” OR “decision*” OR “patient reported outcomes measure*” OR “PROM*”. Articles were imported into an Endnote library.

Inclusion criteria were original studies reporting: mastectomy for DCIS or invasive cancer AND immediate, delayed or no breast reconstruction AND implant or autologous breast reconstruction AND included patient-reported information about reasons for choice.

Exclusion criteria were as follows: studies of prophylactic mastectomy only; comparison of BCS and mastectomy only; review articles containing no original data; articles containing information about predictors of BR, or factors associated with BR, without patient-reported reasons for choice; abstracts without full text details; articles not written in English.

Figure 1 illustrates the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) framework used in this review. All studies satisfying the selection criteria were included, the first relevant publication dating back to 1979. References from included publications were checked, and citations of the original publications were also checked where available. Abstracts from these publications were reviewed by KF and MB to ensure they satisfied the selection criteria. Thirty studies were included in the review.

Fig. 1
figure 1

Systematic review process using an adaptation of the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) schema

Full size image

Assessment of quality

Study quality was assessed using an amended quality assessment tool for qualitative studies from the Alberta Heritage Foundation for Medical Research [38] (see Online Resource 1). This tool is appropriate for evaluating studies that are mostly descriptive (or level III studies using Daly et al.’s terminology) [39]. Questions related to the verification procedures and reflexivity were replaced because we are reviewing patient-reported responses, rather than assessing the objectivity of the authors (see Online Resource 2). Justification for our amendments is included with the Online Resource 2. Each paper was independently rated by two authors (KF and MB) on 12 items and allocated a score out of 24 (a higher score reflecting better quality). KF and MB then reviewed the independent ratings and agreed on a consensus quality score for each publication.

Classification into domains

Women’s reasons for choosing to have or decline BR were categorised into eight domains as previously described [37]. This earlier prospective study used modified statements from Reaby’s 1998 article [34] to compile three separate questionnaires asking why women chose IBR, DBR or NBR. Analysis of the 69 statements was simplified by categorising them into eight domains, based on pragmatic consensus views of the investigators and the findings suggested that they were appropriate for that study population [37].

For this systematic review, these original domains were refined slightly from the earlier study to ensure coverage of all reasons identified in this review. Titles of the domains “feeling normal” and “feeling good” were amended to “feeling/looking normal” and “feeling/looking good” and the “expectations” title was amended to “relationship expectations” to differentiate these reasons from cosmetic or functional expectations.

Results

Studies included in the review

Table 1 provides an overview of the 30 included studies, including study aim and time frame, recruitment and study setting, number of participants and participation rate, design/methods, tools used, limitations/bias, reasons, conclusions and overall quality score. It shows a predominance of studies from the US (n = 11), with four from Australia, two each from Japan, France, England and the Netherlands and one each from Hong Kong, Switzerland, Sweden, Finland, South Africa, Malaysia and Poland. Studies were conducted between the late 1970s and the end of 2014. The number of participants totalled 4269, ranging from 11 to 501 per study. There was a large gap in time between mastectomy and survey/interview, which ranged from 2 weeks to 31 years. The quality score ranged from 11 to 22 out of a possible 24.

Table 1 Overview of included studies
Full size table

The majority of studies (67%) used retrospective methods: 13 employed retrospective questionnaires only [20, 22, 24, 33, 40, 42, 48–50, 53–55]; five used retrospective interviews only [25, 29, 31, 32, 43]; and two used a combination of retrospective questionnaires and interviews [21, 34]. Four studies used prospective questionnaires [19, 23, 35, 45] and one a prospective patient audit [51]. Five studies employed both prospective and retrospective assessments, three of these using questionnaires [37, 44, 46], one interviews [47] and one questionnaires and interviews [52].

Clinical and demographic information

Table 2 provides information on the stage of cancer, mean age and age range, other demographic information provided, time since mastectomy and BR, adjuvant breast cancer treatments and the type and timing of surgery. The type/stage of cancer was not stated in 12 of the studies [19, 22, 25, 31, 34, 43, 45, 47, 51, 53–55]. Where specified, cancer stage/type ranged from stage 0 (ductal carcinoma in situ) to stage IV (metastatic breast cancer).

Table 2 Clinical information
Full size table

Age data were difficult to compare across studies. Overall, the reported mean age ranged from 39.5 to 58 years (for IBR and DBR) and from 51.7 to 67 years (for NBR). Age was not stated at all in two studies [32, 51]; median age was given in five studies [20, 37, 46, 48, 52] and median age range in two studies [42, 45]. Other studies used the mean age. Where more than one group was examined (22 studies) age by BR group (IBR, DB, NBR) was only specified in eight studies [22, 23, 29, 33, 34, 37, 40, 54].

Reasons for BR by domains

Table 3 provides a summary of the patient-reported reasons for choice of IBR, DBR and NBR, as discussed in the 30 publications included in this review, and categorised within their relevant domains. This content analysis approach, whereby the 103 identified reasons were categorised into eight previously determined domains, or groupings of similar reasons, allows further analysis as to the relevant contribution of each major group of reasons within the literature. Table 4 illustrates the proportion of appropriate studies each domain was relevant to. The main patient-reported reasons for wanting BR—either immediate or delayed—relate to the desire of women to feel and look ‘normal’ (70% of publications) or ‘good’ (85% of publications). Being practical (83%) was a major determinant of choice for all women, even those who chose to not have BR, while the influence of others was less commonly cited (60% of publications). Fifty-two percent of publications cited relationship expectations as relevant for choice of IBR or DBR, while fear was influential in either delaying or declining BR in the vast majority of studies (91%). Reasons for choice of immediate versus delayed BR were provided in 57% of studies, and the majority of studies (80%) of women’s reasons for declining BR cited BR as being unnecessary.

Table 3 Patient-reported reasons for choosing IBR, DBR or NBR
Full size table

BR techniques have improved over the years, increasing the range of BR options for women, as well as the possibility of having BR either as an immediate or delayed procedure. However, the reasons for wanting to have BR have generally not changed. The oldest study in this review, published in the US in 1979, identified women’s motivations for undergoing BR [42]. These motivations fitted very neatly into the domains formed to categorise women’s reasons for choosing either IBR or DBR in 2016 [37], as did the reasons provided by women in all studies included in this review. Our findings indicate that women’s reasons for wanting or not wanting BR are similar across time, cultural backgrounds and health system settings. They also suggest the domains used are appropriate and comprehensive, as Table 4 demonstrates.

Table 4 Domains covered by publications
Full size table

Thematic analysis

This comprehensive review identified three recurrent themes across studies.

Lack of information about BR options

Lack of information was a common theme. In some studies, women were not aware of the full range of BR options. In other studies, there were participants who remained unaware that BR even existed [5, 43]. Eighty percent of studies included in this review commented on the lack of information women had about BR options. For example, one study reported only 57% of participants had heard about BR [23], another stated 45.5% were unaware of BR as an option [29], and another noted 18% of NBR participants were not aware of BR [40]. A study of Japanese women [50] reported around 25% of mastectomy patients did not receive any information or explanation about BR, or did not fully understand the information they received. The study by Nelson [33] found almost 20% of participants reported no discussion of BR prior to mastectomy, 41% had no discussion about the advantages and disadvantages of different BR options and 49% had not discussed the option of IBR prior to choosing DBR.

Lack of genuine choice

A second major finding is the lack of genuine choice women have been offered when making the difficult decision to have or not have BR. Only one study offered all women all BR options [37], while three studies only provided reasons for women choosing to not have BR [42, 48, 55]. It was possible to clearly identify a lack of choice in BR options in 13 of the remaining 26 studies (50%) that cited reasons for BR [19, 21, 23, 25, 26, 29, 33, 34, 44, 47, 50, 51, 54] but this rate is likely to be higher, given complete information about the type and/or timing of BR offered was not reported in half the studies.

Reasons for this lack of choice include either the surgeon not offering BR, or only offering a specific type of BR. In one study, 207 of 405 women who agreed to participate (51%) were excluded from the analysis because “medical considerations dictated their surgical approach” [21]. A South African study reported 52 out of 135 women (38.5%), previously judged suitable for locoregional surgery, were not offered BR because of combinations of factors including morbid obesity, ptotic breasts, significant comorbidity, advanced age and expected post-mastectomy radiotherapy [PMRT] [51].

In some studies, women were able to choose BR, but not the type or timing of reconstruction. For example, participants in the Ananian study [19] were given a choice of BR or no BR. However, if they opted for BR, the surgeon decided what type of BR they would have [19]. Similarly, women in the Contant study [44] were given an information module and in a follow-up consultation within two weeks, the “recommended surgical treatment” (implant-based IBR) was discussed [44]. Thirty percent of DBR patients in another study had not been given any choice about the timing, despite almost half of them (45%) stating post-operatively that they would have chosen IBR if given the option [33].

Shameem’s study also highlighted the role of the surgeon, with 33.6% of NBR participants stating their surgeon was against reconstruction, close to the same proportion that were offered BR (36.8%) [29]. Hequét found that the surgeon not offering BR was the main reason for NBR in 18% of cases and a contributing reason in a further 20% [48]. A low-quality study concluded that lack of information was not an important factor in decision-making, although 15% of women identified it as a reason, and 10% as the reason for not undergoing BR [55].

Health system limitations

Finally, the review identified a range of health system factors that may also limit a woman’s access to BR, such as cost (hospital, surgery, implants, time off work) [29, 50]. Even in France and Australia, where BR is available free of charge in the public system, cost was noted as an important issue for not pursuing BR [37, 42, 48]. In these countries, women face a trade-off between decreased hospital costs and increased waiting times [42, 48] and the cost of time off work is also a consideration.

Lack of insurance coverage [26], non-availability of BR in their local area [42], restrictive national clinical guidelines [29], and shortage of operating time and hospital facilities [25, 29, 51] have also been cited as health system barriers to BR.

Discussion

Quality of included studies

Half of the studies (n = 15) scored 18 or more (75%) in their quality assessment. The heterogeneity of studies in this systematic review precluded any detailed analysis by type, timing or setting of BR. The majority of studies (67%) were retrospective, with the potential for recall bias increasing the longer the time since mastectomy. However, a prospective methodology did not automatically equate to a higher quality score, nor did a combination of questionnaire and interview methods.

One-third of included articles did not have reasons for choice as their primary outcome (n = 10) and the earlier studies were written at a time when only delayed implant-based BR was locally available (pre-2006; n = 9). However, many were also high quality, large studies, where all BR options were technically available, even if all options were not offered to all clinically eligible participants.

Appropriateness of domains

The categorisation of 103 reasons discussed in the 30 articles into eight domains appears to be a valid means of analysing a great range of data. We acknowledge the use of domains we have previously created to categorise reasons for choice of BR in this study leaves us open to criticisms of selection bias. To negate this criticism, we have been transparent in our allocation of reasons to each domain (Table 3). Every reason cited in the 30 papers has been included, although the exact wording may vary, as many of the same reasons were referred to in several publications.

Policy and practice implications

This systematic review of women’s reasons for the choice of IBR, DBR or NBR has provided an insight into the range of issues that women value when making this major decision at a difficult time in their lives. Several studies have suggested that up to 50% of women undergoing mastectomy for breast cancer would choose to have BR if it was available [13, 23, 56], indicating that there is a large unmet demand for BR services. The potential benefits of BR are well known [2–9] and its value has been recognised in clinical recommendations in Australia [57], the USA [58] and the UK [59].

However, failure of practice to keep up with long-existing policy recommendations was highlighted in 2016 when Cancer Australia, the peak Australian Government body responsible for providing national leadership in cancer services, published a document entitled Influencing best practice in breast cancer [60]. Part of this document stated that it is “not appropriate to perform a mastectomy without first discussing with the patient the options of immediate or delayed breast reconstruction” [60]. Such a statement would not have been necessary if guidelines were being followed.

Failure to address barriers to genuine choice in medically suitable women who wish to have BR has equity implications. Offering all clinically eligible women a genuine choice of BR options resulted, in one metropolitan practice in Sydney, in a rate of BR uptake of 41%, without any corresponding increase in adverse outcomes [61]. This figure is a marked increase compared with previous estimates of the national Australian BR rate (9% [62], 12% [63] and 16% [18]).

Research has identified significant geographical, racial, age and socioeconomic variation in uptake of BR within Australia that cannot be explained by women’s preferences for BR [14–18]. Similar variations have been documented in the UK [2] and the USA [26]. The provision of BR services requires careful planning and a ‘big picture’ view in order to avoid unintended consequences. For example, a policy of minimum volume thresholds, aimed at improving cancer care in France, has led to IBR being largely limited to specific cancer centres and regional teaching hospitals, raising concerns about geographical equity of access to BR [64].

In this review, many women were not provided with information about breast reconstruction, and information was not always presented in a manner that women could understand. It is unclear whether this was the result of omission of information by the surgeon or whether women did not recall a discussion that occurred. In 1998, Reaby [34] reported that 12% of her Australian participants who underwent mastectomy between 1986 and 1992 were not aware of BR. A 2012 survey of more than 700 Breast Cancer Network Australia (BCNA) members found 10% of women who had a mastectomy had never discussed BR with their surgeons and were not offered it [41], suggesting that little had changed in Australia in over 20 years. A more recent survey of BCNA members, published in 2015 [54], reported only 61% of participants had an initial discussion about BR prior to mastectomy, meaning IBR was not an option for 39% of these women. This is despite only 2% of these respondents stating they were unaware of BR [54]. These results suggest a sizeable difference exists between BR knowledge and BR availability in Australia.

This finding of inadequate provision of information for women considering BR is supported by other studies [65–67]. However, two of the studies included in the review argued for the selective provision of information. Begum et al., in their small retrospective study of autologous BR, found that some women seek information and others avoid it, and noted the need to tailor information delivery to individuals [25]. A larger study of 101 DBR and NBR patients found that 39.5% of participants did not want information regarding the opportunities of BR, either at the time of initial diagnosis and mastectomy, or later [20]. The authors argue that counselling about BR should be available, but not obligatory, because the majority of NBR patients overcome negative attitudes towards their mastectomy quickly and are uninterested in BR. However, they noted that women who are ambivalent about BR need to be offered intensive counselling about their choice [20].

A common reason for not offering women IBR has been the predicted need for PMRT [68–70]. This in turn highlights the “significant variability” both among patients, and between patients and providers, about what matters to women most [71]. Surgeons’ concerns about the potential negative aesthetic impact of PMRT on implants may not correlate with the patient’s desire to have implant-based IBR anyway. Contant’s study reported, despite 50% of reconstructions resulting in “complications or complaints”, 70% of women were satisfied with their reconstruction and only 12% would never choose IBR again [44]. Similarly, Brennan found women are willing to accept the potential adverse outcomes of implant-based IBR following PMRT in exchange for its benefits including enhanced body image during treatment and the possible avoidance of more complicated and costly delayed autologous BR [72].

Clinicians have an ethical duty to discuss all clinically appropriate BR options with their mastectomy patients and to ensure that women have an opportunity to consider the options. Frequently, as these are complex discussions, a number of pre-operative consultations may be required as well as discussion with other professionals in the team (such as nurses or psychologists). The surgeon must also be willing to work in a multidisciplinary team with other plastic or oncoplastic surgeons, or refer the women elsewhere for treatment if she chooses a locally unavailable option.

Health system barriers do exist, but they are not impenetrable. By understanding women’s reasons for wanting or not wanting BR, or for choosing to delay that decision, clinicians can help women to make a choice that is most aligned with their individual patient’s values and needs. It will also help to inform the development of specific educational strategies for consumers and health professionals (including plastic and oncoplastic surgeons) who discuss the options with patients undergoing mastectomy.

Limitations

As with most systematic reviews, the included articles report on a range of outcomes and rely on various instruments to assess them. The majority of articles have employed qualitative methodology, which is the appropriate format to address questions about women’s preferences. The major aim of this study is to better understand the reasons underlying women’s choices, and the use of open questions and/or ad hoc instruments may limit the generalisability of findings. Indeed, the diversity in study designs and populations has prevented analysis about differences in reasons between age groups, ethnicity or health systems, and our findings should be interpreted carefully. Despite this heterogeneity, however, there is a strong consistency in women’s reasons across studies.

Secondly, the majority of the 30 studies have come from western, high-income countries, with the exceptions of Malaysia and Poland. This is likely to reflect a paucity of BR research, and possibly BR itself, in low-income countries, and again constrains the generalisability of our findings to other settings.

A final possible limitation is the lack of validation for our assessment of study quality. The quality score does not directly reflect a particular study’s limitations but rather provides a standardised measure to compare the same range of attributes across studies. A study with several specific limitations may still score highly on general attributes, while a high-quality study may not. However, the same is true of any quality measure used to assess heterogeneous studies.

Strengths of this review include rigorous methodology, incorporating a systematic search strategy and pre-defined inclusion criteria and analysis plan. The large number of participants (over 4200) across 30 studies from 13 countries provides a depth of coverage which contributes to a better understanding of the most salient issues for women faced with this decision.

Conclusion

This systematic review has provided a synthesis of the reasons underlying women’s preferences regarding BR. It has demonstrated a consistency of views about what matters most to women faced with the difficult decision to have or not have BR following mastectomy for breast cancer. It has also demonstrated the utility of categorising reasons into domains. Women and clinicians may be able to focus on these eight domains in their consideration of BR options. Highlighting the paucity of genuine choice of BR, and its implications for equitable delivery of BR services, will help lay the foundations for further research and policy advocacy into the most feasible means of increasing women’s choice of BR options.