Defining Outcomes in Facial Gender-Affirming Surgery: A Systematic Review to Inform Core Outcome Set Development

1. Introduction

Facial gender-affirming surgery (FGAS) targets one of the most important anatomic regions contributing to gender dysphoria in transgender and non-binary individuals [1]. For some patients, particularly those assigned male at birth, FGAS may be the most impactful surgical intervention for alleviating dysphoria [1]. As access to FGAS has grown and surgical techniques have evolved, research on outcomes has expanded in parallel. Even so, how these outcomes are defined and reported varies substantially across studies [1,2,3,4,5,6].

This heterogeneity is not unique to FGAS, but it presents particular challenges in a field where patient-centered constructs—such as gender congruence, psychosocial well-being, and satisfaction—are central to evaluating success. Some studies emphasize technical or anatomic precision, others rely on surgeon-reported aesthetic impressions, and only a minority incorporate validated patient-reported outcome measures. Psychosocial outcomes, functional changes, and long-term patient experiences are mentioned far less frequently [1]. As a result, the literature is fragmented, difficult to synthesize, and limited in its ability to guide evidence-based care or generate best practices.

Core Outcome Sets (COSs) offer a structured solution to these challenges. COSs define a minimum group of outcomes that should be reported consistently across studies evaluating a particular intervention. They have been successfully implemented in reconstructive surgery, including breast reconstruction after mastectomy through initiatives such as BRAVO [21,22]. The Core Outcome Measures in Effectiveness Trials (COMET) Initiative has further emphasized the value of COSs in reducing heterogeneity, strengthening comparative research, and ensuring that outcomes reflect the priorities of both patients and clinicians [8].

To inform COS development for FGAS, it is necessary to first characterize the full range of outcomes currently reported in the literature. Initial observations reveal studies that variously prioritize surgical measurements, aesthetic judgment, or limited satisfaction metrics, often without standardized definitions of success [18]. Understanding where outcome reporting is consistent, where it is lacking, and which measurement tools are being used is essential for building a unified and patient-centered framework.

This study undertook a systematic review of 334 FGAS publications to map current outcome reporting practices, identify gaps in measurement, and clarify the domains emphasized across the literature. The goal is to provide the foundation needed for subsequent consensus-building and COS development, ultimately improving clarity, comparability, and clinical relevance in FGAS outcomes research.

2. Materials and Methods

Search Strategy and Selection Criteria: A comprehensive literature search was performed in PubMed to identify studies reporting outcomes of facial gender affirming surgery published through March 15, 2025; the search was last updated on March 15, 2025. The search strategy included terms related to transgender and gender diverse populations (for example transgender, gender dysphoria, gender incongruence) and facial gender affirming procedures (for example facial feminization surgery, rhinoplasty, mandibular contouring, tracheal shave). Filters were used to exclude studies focused on aesthetic procedures for cisgender patients, facial trauma, or congenital craniofacial conditions. This systematic review was conducted in accordance with the PRISMA 2020 guidelines, and all reporting elements recommended in the PRISMA checklist have been incorporated. A PRISMA flow diagram (Figure 1) was used to document study selection and ensure transparency [23].

Search Strategy Details: The complete PubMed search strategy, including all Boolean operators and field tags, is provided in Table A1, as required by PRISMA 2020 Item 7. Citation searching was not performed. Search strategies were not adapted from prior reviews. The search strategy was not peer reviewed.

Study Screening and Eligibility: Titles and abstracts of all identified records were screened independently by four reviewers (SR, SR, GA, RC). Full texts were reviewed to determine final eligibility. Eligible study designs included retrospective and prospective studies, case series, case reports, and reviews with extractable data. Narrative reviews without primary extractable data, editorials, and non-surgical studies were excluded. Studies were included if they evaluated surgical facial gender affirming procedures and reported clinical, functional, aesthetic, psychosocial, quality of life, or patient reported outcomes. Studies were excluded if they focused only on non-surgical interventions such as hormone therapy, voice training, or hair transplantation, were non-English, were cadaveric studies, or were narrative reviews without primary data. Studies that could not be obtained after institutional access and interlibrary loan attempts were excluded. Review articles and commentaries that contained extractable outcome data were retained and handled separately to avoid duplication. When a review included patients who also appeared in primary data sources, only the primary data were counted. Any disagreements during screening or selection were resolved by discussion among the review team. This review was not prospectively registered with PROSPERO or any other registry.

Data Extraction and Outcome Measures: Data were extracted using a standardized template. Extracted information included publication year, country, study design, sample size, study duration, participant demographics, surgical focus, and outcome domains. Outcome categories included aesthetic, functional, psychosocial, clinical measurement, satisfaction, health related quality of life, complications, and other domains. Information on outcome measurement methods was also recorded, including whether outcomes were reported by patients or physicians, whether validated instruments were used, follow-up duration, and the methods used to report satisfaction.

Psychosocial outcomes and health related quality of life were defined clearly for this review because studies used these terms inconsistently. Psychosocial outcomes referred to mental and social dimensions of well-being such as emotional adjustment, social relationships, and self-esteem. Health related quality of life was defined more broadly to include physical, mental, emotional, and social functioning. While psychosocial outcomes can contribute to health related quality of life, the two categories were extracted separately.

Two reviewers (MM and KS) independently performed data extraction, and disagreements were resolved by joint review.

Quality Assessment: Methodological quality of non randomized studies was assessed using the Methodological Index for Non Randomized Studies criteria. Risk of bias for non-randomized studies was assessed using the MINORS tool by two reviewers independently. Reporting bias and certainty assessments were not performed because this review did not involve meta-analyses or graded bodies of evidence. Items assessed included prospective data collection, appropriateness of endpoints, unbiased assessment, and sample size justification. Studies for which these criteria did not apply, such as descriptive papers, were not assigned MINORS scores but were retained for outcome mapping when primary data were present.

Data Analysis: Extracted data were organized in a structured spreadsheet. Descriptive statistics were used to summarize study characteristics, outcome domains, the use of validated measurement tools, and trends in outcome reporting. Pearson’s chi square tests were used to compare categorical variables such as study design and outcome domains. Bartlett’s test was used to examine variability in follow-up duration across studies. McNemar’s test was used to compare proportions of physician reported and patient reported outcomes. All analyses were performed in RStudio version 12.1 (Posit, Boston, MA). A significance threshold of p < 0.05 was used. Because this review did not include pooled effect estimates, no summary effect measures (e.g., risk ratios, mean differences) were calculated. No meta-analysis was performed because included studies were heterogeneous in design, outcome measures, and reporting style. Data were synthesized descriptively and grouped by outcome domain, anatomical region, study design, and measurement method. No statistical transformations were applied, and all outcomes were reported as described in the source studies.

3. Results

3.1. Study Characteristics

Of the 334 publications identified, 231 met inclusion criteria, with a combined total of 30,936 participants. The included studies consisted of retrospective studies (n=80, 34.6 percent), review articles (n=79, 34.2 percent), case reports or case series (n=24, 10.4 percent), prospective studies (n=18, 7.8 percent), and commentaries (n=12, 5.2 percent). Publication years ranged from 1990 to 2025, with the highest number of publications emerging in 2024 (n=57, 24.7 percent). Studies were conducted across 30 countries, with the largest proportion from the United States (n=159, 68.5 percent).

The average duration of observational studies was 6.1 ± 7.1 years (range 0.1 to 38 years). Sample sizes varied substantially, with a median of 36 participants (IQR 15 to 168). Demographic reporting was inconsistent. Among studies reporting adequate details, transgender women were the most represented group (92.6 percent, n=214), followed by transmasculine participants (14.3 percent, n=33) and nonbinary participants (5.2 percent, n=12). The mean reported participant age was 34.5 ± 8.8 years.

3.2. Procedures and Facial Regions Prioritized

Surgical focus varied across studies. The upper third of the face was addressed in 151 studies (65.4 percent), the midface in 134 studies (58.0 percent), and the lower third in 139 studies (60.2 percent). Panfacial interventions were described in 99 studies (42.9 percent). Procedures involving the neck, including chondrolaryngoplasty, appeared in 47 studies (20.3 percent).

Outcomes by anatomical region showed consistent patterns. Studies focusing on the upper third most often used aesthetic outcomes and cephalometric or CT based measurements. Midface studies most frequently emphasized aesthetic assessments, rhinoplasty specific anthropometric measurements, and surgeon-rated appearance. Lower third studies often reported complications and objective measurements related to bony contouring, along with patient satisfaction. Panfacial studies were the most likely to include psychosocial outcomes and quality-of-life measures, although these remained inconsistently applied.

3.3. Outcome Domains

Outcome reporting varied widely. Aesthetic outcomes were the most frequently reported (61.5 percent, n=142), followed by complications (60.2 percent, n=139), satisfaction (51.5 percent, n=119), psychosocial outcomes (45.9 percent, n=106), clinical measurements (36.4 percent, n=84), reoperation or readmission rates (20.8 percent, n=48), and health related quality of life (21.6 percent, n=50).

The category labeled other outcomes occurred in 23.8 percent of studies (n=55). This category included perioperative experience, intraoperative variables such as anesthesia duration and blood loss, postoperative factors such as opioid use, and measures related to speech or vocal quality when applicable.

No included study reported using the GENDER-Q, which was recently developed and published after the inclusion period for most studies.

A chi square goodness of fit test showed significant variability in how outcome domains were distributed across studies (χ2 = 115.8, p < 0.001). Additional chi square analyses demonstrated strong associations between psychosocial outcomes and health related quality of life (p < 0.001), satisfaction and both psychosocial outcomes and health related quality of life (p < 0.001), and anatomic measurement with aesthetic outcomes (p < 0.001).

Among studies reporting patient satisfaction (n=85), 36.8 percent used patient ratings, 36.8 percent assessed quality of life or social functioning, 16.5 percent used observer-rated appearance, and 0.8 percent used artificial intelligence based assessments.

3.4. Primary Outcomes

When primary outcomes were explicitly stated, aesthetic objectives were the most common, followed by functional outcomes such as voice quality, complications, or reoperation rates. Fewer studies focused on patient-centered outcomes such as satisfaction, gender congruence, or psychosocial status. These were often reported qualitatively, without validated instruments or formal scoring systems. Objective metrics such as cephalometrics appeared inconsistently and were usually paired with subjective assessments. Only a minority of studies defined success in measurable terms.

3.5. Outcome Tracking and Measurement

Reporting of follow-up timing showed substantial variability. Across all studies that reported both first and last follow-up, the average times were 3.7 months (range 0.25 to 12 months) and 23.6 months (range 1.5 to 204 months), respectively. Bartlett’s test demonstrated significant variability in both early follow-up (χ2 = 11.5, p < 0.01) and late follow-up (χ2 = 223.2, p < 0.0001), indicating inconsistent tracking practices.

Physician-reported outcomes appeared in 149 studies (64.5 percent), which was significantly higher than patient-reported outcomes (97 studies, 42.0 percent; p < 0.001). A subset of 69 studies (29.9 percent) incorporated both physician and patient perspectives. Among patient satisfaction studies, most relied on general statements rather than quantitative scales. Only 35 studies provided numerical satisfaction scores, with an average satisfaction level of 89.2 ± 9.5 percent (range 69.3 to 100.0 percent).

3.6. Validated Measurement Tools

Only 26 studies used validated outcome instruments. FACE-Q was the most common (n=13), followed by PROMIS (n=2) and GRBAS (n=2). No studies used GENDER-Q. This reflects a substantial gap in standardized patient-reported outcome measurement across facial gender affirming surgery.

3.7. Quality Assessment

Methodological quality as measured by the MINORS criteria varied. The median score for non comparative studies was 9 (IQR 7 to 11), indicating low to moderate quality. Comparative studies had an average score of 16 (IQR 14 to 18), indicating moderate quality. Common limitations included lack of prospective data collection, absent control groups, and incomplete reporting of follow-up and statistical justification.

3.8. Risk of Bias Assessment

Methodological quality varied across studies. The median MINORS score for non-comparative studies was 9 (IQR 7–11), indicating low to moderate quality. Comparative studies had a median score of 16 (IQR 14–18), representing moderate quality. Common limitations included lack of prospective data collection, absence of control groups, and incomplete follow-up reporting.

Figure 3. Heterogeneity of Outcome Measurement Methods Across Domains in Facial Gender-Affirming Surgery Studies. The size of each bubble indicates the relative frequency with which each domain-method pairing was reported across included studies. For instance, large bubbles along the “Quality of life, social reintegration” column suggest that this patient-reported method is the most common for assessing psychosocial and HRQoL outcomes. In contrast, domains like “Reoperation/Readmission” or “Complication” are more frequently reported via objective measurements or alternative methodologies. This visually underscores both the diversity and inconsistency in how outcomes are defined and measured in FGAS research. Notably, many domains are assessed through multiple methods, reinforcing the need for standardization in outcome reporting.

Figure 3. Heterogeneity of Outcome Measurement Methods Across Domains in Facial Gender-Affirming Surgery Studies. The size of each bubble indicates the relative frequency with which each domain-method pairing was reported across included studies. For instance, large bubbles along the “Quality of life, social reintegration” column suggest that this patient-reported method is the most common for assessing psychosocial and HRQoL outcomes. In contrast, domains like “Reoperation/Readmission” or “Complication” are more frequently reported via objective measurements or alternative methodologies. This visually underscores both the diversity and inconsistency in how outcomes are defined and measured in FGAS research. Notably, many domains are assessed through multiple methods, reinforcing the need for standardization in outcome reporting.

Table 1. Outcome Domains Across Included Studies. This table summarizes the distribution of outcome domains assessed across the included studies. Outcome domains were categorized as either internal (patient-centered)—such as psychosocial functioning, quality of life, or satisfaction—or external (appearance-focused)—including metrics like facial aesthetics or surgical symmetry. While some studies incorporated a combination of internal and external outcomes, the majority prioritized external domains, reflecting a broader trend in the literature toward emphasizing aesthetic results over patient-reported experiences or functional outcomes.

Outcome Domain	Papers Reported (%)
Psychosocial	43.19
Measurements	37.09
Aesthetics	60.56
Reoperation/Readmission	21.60
Complication	62.44
HRQOL	22.54
Other	25.35

Table 1. Outcome Domains Across Included Studies. This table summarizes the distribution of outcome domains assessed across the included studies. Outcome domains were categorized as either internal (patient-centered)—such as psychosocial functioning, quality of life, or satisfaction—or external (appearance-focused)—including metrics like facial aesthetics or surgical symmetry. While some studies incorporated a combination of internal and external outcomes, the majority prioritized external domains, reflecting a broader trend in the literature toward emphasizing aesthetic results over patient-reported experiences or functional outcomes.

Outcome Domain	Papers Reported (%)
Psychosocial	43.19
Measurements	37.09
Aesthetics	60.56
Reoperation/Readmission	21.60
Complication	62.44
HRQOL	22.54
Other	25.35

Table 2. Number of Outcome Domains Assessed per Study. This table illustrates how many outcome domains each study included in its evaluation. The majority of studies assessed only one or two domains, highlighting a narrow focus in outcome reporting. Few studies captured a broader range of outcomes, suggesting limited multidimensional assessment. This trend underscores the need for more comprehensive evaluation frameworks that integrate both patient-centered and aesthetic outcomes in future research.

Outcomes Reported	Papers Reported (%)
Quality of Life / Reintegration into Society	35.21
Observer-rated Appearance	17.37
Artificial Intelligence Rating	0.94
Quantitative Anatomical Measurement	25.82
Other	45.07

Table 2. Number of Outcome Domains Assessed per Study. This table illustrates how many outcome domains each study included in its evaluation. The majority of studies assessed only one or two domains, highlighting a narrow focus in outcome reporting. Few studies captured a broader range of outcomes, suggesting limited multidimensional assessment. This trend underscores the need for more comprehensive evaluation frameworks that integrate both patient-centered and aesthetic outcomes in future research.

Outcomes Reported	Papers Reported (%)
Quality of Life / Reintegration into Society	35.21
Observer-rated Appearance	17.37
Artificial Intelligence Rating	0.94
Quantitative Anatomical Measurement	25.82
Other	45.07

4. Discussion

This review demonstrates considerable variability in how outcomes are defined, measured, and reported within the facial gender affirming surgery literature. Although the number of publications in this area has increased substantially in recent years, there remains no shared framework for determining what constitutes success in these procedures [8]. Most studies evaluate aesthetic or technical endpoints, and these are often based on surgeon judgment rather than standardized, transparent criteria. Patient centered outcomes such as psychosocial functioning, gender congruence, and quality of life are incorporated less frequently and are measured inconsistently when included. This lack of uniformity limits the field’s ability to compare findings across studies, perform meta-analyses, or establish benchmarks that reflect patient priorities.

Only 26 of 232 studies used validated patient reported outcome instruments, which represents a small minority in the current evidence base. Although many studies identified primary outcomes, most did not provide definitions that reflected patient experience or expectations. Measures of psychosocial well-being or long-term functional change were uncommon, and the methodological quality of studies was generally low to moderate based on MINORS criteria. These limitations have practical consequences. Heterogeneous outcome reporting has been cited in policy documents from private insurers as a reason for restricting coverage of facial gender affirming surgery, since the inconsistency makes it difficult to demonstrate reliable improvements in gender dysphoria or related domains [4,5]. Without consistent measurement, it is also challenging to develop best practices in perioperative care, rehabilitation, or multidisciplinary coordination, including emerging approaches such as surgical prehabilitation programs [24,25,26].

The variability in outcomes is also influenced by the diversity of procedures encompassed by facial gender affirming surgery. The upper third, midface, lower face, and neck involve distinct anatomic goals and functional considerations, and studies focusing on these regions often use different outcome metrics. This variation reflects genuine differences in surgical objectives but also contributes to fragmentation in the literature. For the purposes of Core Outcome Set (COS) development, consolidating these procedures into a single review allows for a comprehensive evaluation of all outcomes currently being measured. However, a future COS will likely need to incorporate domain specific modules to capture the unique goals of different facial regions while still maintaining a minimum set of shared outcomes.

The development of a COS offers a structured solution to these problems and establishes a minimum group of outcomes that should be reported consistently in all studies evaluating an intervention [27]. It does not restrict additional outcomes but ensures that essential domains are always included. For facial gender affirming surgery, a COS would help unify reporting practices, strengthen comparative research, and support patient-centered clinical decision making.

Outcome instrument selection will be an important component of COS development. Generic instruments such as PROMIS offer strong psychometric rigor, population-level comparability, and flexibility across domains. These advantages make PROMIS valuable for assessing constructs such as mental health functioning, social participation, and overall well-being. However, PROMIS instruments do not capture face-specific concerns or gender affirmation metrics. In contrast, FACE-Q provides detailed assessment of facial appearance and surgical change but is procedure specific and cannot be compared against broader populations. GENDER-Q, published in late 2024 and designed for gender affirming care more broadly, addresses many of the gaps identified in this review, including gender congruence, satisfaction, and psychosocial impact. Even so, it is not specific enough for facial gender affirming surgery, and there remains a need for a COS focused specifically on FGAS. No studies in our dataset used GENDER-Q, which reflects that many included publications predated its release, as our literature search was conducted in March 2025. A future COS for facial gender affirming surgery will likely require both generic instruments to capture broader changes in well-being and condition specific tools to assess appearance-related and gender specific outcomes.

Beyond improving standardization, the development of a COS has important implications for equity. Transgender and nonbinary individuals remain underrepresented in surgical outcomes research, and inconsistent reporting risks reinforcing this underrepresentation [28]. A COS would require future studies to consistently include outcomes that matter to these communities, supporting a more inclusive evidence base. Standardized, transparent outcome reporting may also strengthen policy and insurance coverage decisions by providing clear, reproducible data on the effects of facial gender affirming surgery.

Advancing this work will require a structured, multi phase consensus process. The next step involves qualitative work with key stakeholders, including transgender patients, surgeons who perform facial gender affirming surgery, primary care clinicians, mental health professionals, speech specialists, and community advocates. These discussions will help identify outcomes that are most meaningful across perspectives. These findings can then be integrated into a Delphi process, which uses iterative rounds of rating and feedback to reach consensus among a diverse panel of experts. This strategy has been used successfully in the development of the GENDER-COS project for genital gender affirming surgery [29] and provides a model for establishing a COS for facial gender affirming surgery.

This review has several limitations. The search was limited to studies accessible through PubMed and to the plastic surgery literature. Other specialties, particularly otolaryngology and oral and maxillofacial surgery, also contribute to facial gender affirming surgery, and their relative underrepresentation is a limitation of the available evidence. Inconsistent demographic reporting across studies also limited the ability to analyze outcomes by age, race, or gender identity. Finally, many studies lacked complete information regarding follow-up duration, validated instrument use, or primary endpoint definitions, which may reduce the strength of cross-study comparisons.

This review also has limitations related to the review process. Only one database (PubMed) was searched, which may have missed relevant studies indexed elsewhere. The review was not prospectively registered, and no protocol was prepared, which may introduce procedural bias. Additionally, data extraction was limited by inconsistent reporting across studies, and the inability to obtain several full texts may have introduced selection bias.

5. Conclusions

In this review, we examined the existing literature on outcomes in facial gender affirming surgery to create a comprehensive overview of how outcomes are currently defined and measured. By mapping the full range of reported domains, instruments, and measurement practices, this work provides the first structured foundation needed to guide the development of a COS for facial gender affirming surgery. The findings highlight areas of inconsistency, underreporting, and limited use of validated tools, all of which underscore the need for greater standardization. The next phase will require direct engagement with patients, clinicians, and other stakeholders to build consensus on which outcomes should form the core of future research. Establishing a COS will help create more consistent, comparable, and patient-centered outcome reporting for facial gender affirming surgery moving forward.