Gender bias in shared decision‐making among cancer care guidelines: A systematic review

Abstract Background In cancer care, the promotion and implementation of shared decision‐making in clinical practice guidelines (CPG) and consensus statements may have potential differences by gender. Objective To systematically analyse recommendations concerning shared decision‐making in CPGs and consensus statements for the most frequent cancers exclusively among males (prostate) and females (endometrial). Search Strategy We prospectively registered the protocol at PROSPERO (ID: RD42021241127). MEDLINE, EMBASE, Web of Science, Scopus and online sources (8 guideline databases and 65 professional society websites) were searched independently by two reviewers, without language restrictions. Inclusion Criteria CPGs and consensus statements about the diagnosis or treatment of prostate and endometrial cancers were included from January 2015 to August 2021. Data Extraction and Synthesis Quality assessment deployed a previously developed 31‐item tool and differences between the two cancers analysed. Main Results A total of 176 documents met inclusion criteria, 97 for prostate cancer (84 CPGs and 13 consensus statements) and 79 for endometrial cancer (67 CPGs and 12 consensus statements). Shared decision‐making was recommended more often in prostate cancer guidelines compared to endometrial cancer (46/97 vs. 13/79, 47.4% vs. 16.5%; p < .001). Compared to prostate cancer guidelines (mean 2.14 items, standard deviation 3.45), compliance with the shared‐decision‐making 31‐item tool was lower for endometrial cancer guidelines (mean 0.48 items, standard deviation 1.29) (p < .001). Regarding advice on the implementation of shared decision‐making, it was only reported in 3 (3.8%) endometrial cancer guidelines and in 16 (16.5%) prostate cancer guidelines (p < .001). Discussion and Conclusions We observed a significant gender bias as shared decision‐making was systematically more often recommended in the prostate compared to endometrial cancer guidelines. These findings should encourage new CPGs and consensus statements to consider shared decision‐making for improving cancer care regardless of the gender affected. Patient or Public Contribution The findings may inform future recommendations for professional associations and governments to update and develop high‐quality clinical guidelines to consider patients' preferences and shared decision‐making in cancer care.


| INTRODUCTION
The selection of the best diagnostic approach or treatment in cancer care must be personalized 1,2 given the vast quantity of strategies, screening techniques and therapeutical practices currently available. 3 These decisions require a high level of patient participation. 4 It has been purported that gender bias exists with the preferences of men being given greater priority than those of affected women. 5 The participation of patients concerning the best diagnostic or treatment approach for their own disease through shared decisionmaking (SDM) is currently considered essential in achieving sustainable, high-quality cancer care. 4,[6][7][8] This is important because different diagnostic or treatment options with similar potential may lead to different results depending on the patient's preferences and values. 4,9 SDM has been shown to increase patient satisfaction, 4 cost-effectiveness 4 and reduce negligence claims. 10 Therefore, in many developed countries, SDM is legally compulsory, [10][11][12] and professional medical associations widely recommend it. [13][14][15] The systematic implementation of SDM in cancer care faces several obstacles, [16][17][18] and it is still poor. 19,20 Despite various proposed strategies to promote SDM, 9,21 clinical practice guidelines (CPGs) and consensus statements generally fail to recommend it, as recently suggested for breast cancer. 22,23 It is important to address the possible existence of a gender bias in SDM recommendations. This could be hypothesized for guidelines concerning cancers that exclusively affect biological males versus those that exclusively affect biological females. Particularly, major implications can result from treating prostate cancer, such as disruptions to urinary, bowel or sexual function. Due to the significant tradeoffs with prostate cancer screening and treatment, SDM has been strongly encouraged. 2,14 In fact, according to the US Preventive Services Task Force, screening of prostate cancer using the prostate-specific antigen (PSA) presents grade C of evidence for men aged 55-69 years (meaning that the decision should be individualized), and grade D for men older than 69 years, which has led to a reduction in the screening. 24 Similarly, the treatment of endometrial cancer can result in significant consequences, such as loss of fertility for premenopausal females, urinary or faecal incontinence or early menopause, among others. Decisions regarding hormonal treatment after oophorectomy for perimenopausal females remain uncertain. The significant counterparts of the screening and surgical or hormonal treatment of this pathology, also make SDM highly recommendable for its diagnosis and treatment. Again, the US Preventive Services Task Force highlights that there is no standard or routine screening test for endometrial cancer and all of them have risks and side effects, including periodic pelvic examination. 25 We systematically reviewed the characteristics of CPGs and consensus statements concerning SDM in the diagnosis and treatment of the most frequent cancer exclusively affecting males, that is, prostate cancer, and the most frequent cancer exclusively affecting females, that is, endometrial cancer.

| PATIENTS AND METHODS
The systematic review was conducted following prospective protocol registration (Prospero ID: CRD42021241127) and was reported according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses 26 (Supporting Information: Appendix 1). For comparison, we selected the most frequent exclusively male cancer (prostate cancer) 27 and the most frequent exclusively female cancer (endometrial cancer, also known as uterine cancer, carcinoma of the uterine corpus or adenocarcinoma of the endometrium). 27

| Search strategy and data source
We conducted a systematic search covering from January 2015 to Appendix 2), and references from systematic reviews and other studies on this topic were analysed.

| Study selection and data extraction
We covered CPGs and consensus statements on diagnosis and therapeutic management of prostate or endometrial cancer, developed by professional societies, organizations or government agencies. Guidelines on the management of cancer complications (e.g., castration-resistant prostate cancer, or Lynch syndrome for endometrial cancer) were also included. Obsolete documents updated in more recent years from the same organization, documents for Duplicate documents were removed. The management of the information (selected documents) for the review was facilitated using EndNote ® version 20 (Clarivate Analytics).

| Quality assessment
The 31-item tool 23 for quality assessment of CPGs and consensus statements on SDM was used, originally based on items identified from the AGREE II 28  ). The quality assessment was divided into 13 domains (Supporting Information: Appendix 3). No formal score or cut point for defining quality was considered, as recommended by the authors of the tool. 23

| Statistical analyses
First, a descriptive analysis of quality assessment items concerning SDM was conducted separately for prostate and endometrial cancer.
Second, differences between both groups were analysed using T tests, and χ 2 tests for quantitative and qualitative variables, respectively. When χ 2 conditions for applications were not met, Fisher exact tests were applied.

| Factors associated with SDM
Only 59 (33.5%) guidelines included information on SDM. Table 2 shows the characteristics of the guidelines stratified by the presence  Complete information on differences between prostate and endometrial cancer guidelines is shown in Table 3. When applying the 31item tool 23 for assessing SDM compliance of the data extraction items (Figure 2), we showed important differences depending on the cancer type. Although compliance with the items was low for both types of cancer, most of them were much lower for endometrial

| DISCUSSION
Our thorough systematic review of clinical practice guidelines and consensus statements regarding prostate and endometrial cancer diagnosis and treatment found that recommendations concerning SDM were not universal for both types of cancer. Importantly, however, they were significantly weaker for endometrial versus prostate cancer, demonstrating a gender bias in SDM for cancer care.
We found that recommendations on SDM were more frequent in recent guidelines, but important items regarding SDM reporting (especially those regarding advice on SDM implementation) were missing across the time horizon.
We chose two diseases (prostate cancer and endometrial cancer) for which SDM is especially recommended. The risks and benefits involved in treatment decisions are uncertain and should be individualized. As an example, according to the most recent prostate cancer guideline provided by the European Association of Urology, 41 radical prostatectomies should not be denied on the grounds of age alone, but the stage of the disease, the frailty of the patient and the consensus between specialists and the patient should guide the final decision. This is also applicable to active surveillance, watchful waiting or radiotherapy, among other treatment options. Postoperative incontinence and erectile dysfunction are common problems following surgery, around 20% and 70%, respectively. 41 Therefore, risks and benefits must be considered and discussed, and the management of complications should be equally approached.
Similarly, surgery as a treatment for endometrial cancer can also lead to fertility loss, urinary incontinence or early menopause. According to the most recent European guideline, 80  . Future specific studies should analyze and discuss whether the differences found in our study might reflect actual gender bias in cancer care.
A key strength of our study was a global perspective with a large number of clinical practice guidelines and consensus statements included. We did not restrict our search to specific languages or data source limitations. Nevertheless, it should be noted that gender bias is not equal across the world, which might influence the results of this review. We tried to approach this point by comparing the frequency of SDM in different continents, but no important differences were observed. One perceived limitation of our study is the subjective  100 We also showed a lower frequency of SDM in non-European gynaecological guidelines, which suggests that further information and dissemination on SDM benefits should be especially strengthened in these contexts.
We showed that SDM is increasingly being covered in guidelines in the most recent years. Most of the guidance methodological handbooks for updating clinical practice guidelines recommend that the time between updates should be 2 or 3 years, therefore older guidelines run the risk of being outdated. 114 We only covered prostate and endometrial cancer guidelines for comparing a potential gender bias in SDM in cancer care, as the most frequent exclusively male and female cancers requiring SDM according to the recommendations, due to feasibility criteria. Sex disparities in this regard should be confirmed by studying other exclusively-men cancers (e.g., testicular cancer) and other exclusively female cancers (e.g., cervical, or ovarian cancer). Potential differences in recommendations may reflect a bias in the statement of clinical evidence for men versus women (e.g., grade of recommendation of screening for both pathologies). Although elucidating that gap is not within the scope of this project, we recommend approaching this point in future research, not only for cancer care.
Our results suggest that SDM should be introduced in endometrial cancer guidelines, and also reinforced in prostate  Thus, there is a gender bias that merits further investigation and correction to achieve equality in improving cancer care.

AUTHOR CONTRIBUTIONS
Each author certifies that he/she has made a direct and substantial contribution to the conception and design of the study, the