Racial and ethnic differences in self-reported satisfaction with fertility clinics and doctors

To investigate if differences in self-reported satisfaction with fertility clinics and doctors differ by race/ethnicity. We used cross-sectional survey data from FertilityIQ online questionnaires completed by patients receiving US. fertility care from July 2015 to December 2020. Univariate and multivariate logistic and linear regression analyses were performed to assess association of race/ethnicity on patient-reported clinic and physician satisfaction. Our total sample size included 21,472 unique survey responses (15,986 Caucasian, 1856 Black, 1780 LatinX, 771 East Asian, 619 South Asian, 273 Middle Eastern, 187 Native American self-reported). When adjusting for potential confounders (demographic and patient satisfaction), we found that Black patients rated their doctors more highly (odds ratio (OR) 1.30, 95% confidence interval (CI) 1.04–1.62 p = 0.022 logistic and Coefficient 0.082, 95% CI 0.013–0.15 p = 0.02 linear), while other ethnic groups did not show significant differences compared to Caucasian patients. East Asians had borderline lower satisfaction with clinic satisfaction in logistic regression (OR 0.74 95% CI 0.55–1.00 p = 0.05), while significant differences were not found for other ethnic groups for clinic satisfaction. In summary, some but not all minority groups differed in their self-reported perception of satisfaction with fertility clinic and doctors compared to Caucasian patients. Cultural differences towards surveys may contribute to some of these findings, and satisfaction by racial/ethnic group may also be modified by results of care.


Introduction
Prior literature has demonstrated disparities in assisted reproductive technology (ART) outcomes by racial/ethnic group. Minorities and particularly Black patients have been found to have worse outcomes when compared to Caucasian patients in multiple studies [1,2]. An analysis of 38,309 cycles using Society for Assisted Reproductive Technology Clinic Outcome Reporting System (SART CORS) data found increased miscarriage rate and decreased live birth rate for Black women compared to White women, even after controlling for uterine and tubal factor disease [3]. Studies have also reported that even in cycles with preimplantation genetic testing (PGT), Black race was associated with lower live birth rate in frozen embryo transfer cycles (adjusted risk ratio [aRR], 0.82; 95% confidence interval [CI] 0.73-0.92) and a higher risk of clinical loss (aRR 1.59; 95% CI 1.28-1.99) [4]. Another study reported that Black recipients with donor embryo transfer cycles had a lower probability of live birth than White donors and 1 3 recipients (aRR, 0.84; 95% CI 0.71-0.99) [5]. Disparities in assisted reproductive technology (ART) outcomes for Asian and Hispanic groups have also been reported, with a large SART study finding that adjusted odds of pregnancy were reduced for Asians (adjusted odds ratio (aOR) 0.86, 95% CI 0.80-0.93). This study also reported that live birth odds were reduced for multiple minority groups: Asian (aOR 0.90, 95% CI 0.82-0.97), Black (aOR 0.62, 95% CI 0.56-0.68), and Hispanic (aOR 0.87, 95% CI 0.79-0.96) [6]. Studies on South Asian populations have reported inconsistent information on disparities, including some analyses reporting worse outcomes in this population, while others have reported equivalent outcomes [7,8]. Studies on Hispanic patients have been limited by sample size, though some have reported that live birth rates from ART may be decreased for this population [6]. Literature on patient perspective of self-reported fertility care satisfaction (in general and also by race/ethnicity) is extremely limited. A study of 1460 patients at an academic fertility center in Illinois found that Black women (14.7%) were more likely to report race as a barrier to fertility treatment compared with other races, though satisfaction with fertility care was not assessed [9]. A prior FertilityIQ analysis of 7456 women from 2015 to 2018 reported a number of factors associated with positive patient-reported experience [10]. Other literature has investigated patient satisfaction in fertility care, with studies generally reporting high general patient satisfaction (94%) [11][12][13][14]. In other fields of medicine, several studies have suggested that minority groups (including East Asian and Black patients) may report lower satisfaction than White groups [15,16], though not all studies have found these effects [17]. In general, the literature on patient perspective and satisfaction by race/ethnicity for fertility care and other medical fields is very limited, and this is an important area for future research.
Given the known disparities in ART outcomes and the lack of literature on patient satisfaction in fertility care by race/ethnicity, we aimed to use the large, multi-racial Fer-tilityIQ database to investigate if differences in self-reported satisfaction with fertility clinics and doctors also differ by race/ethnicity, using seven distinct categories categorized by FertilityIQ (Caucasian, Black, East Asian, South Asian, LatinX, Middle Eastern, Native American). This is one of the first studies of this subject matter of racial/ethnic satisfaction with fertility care, and the largest to date.

Materials and methods
This is a cross-sectional survey study using online-based FertilityIQ questionnaires, which were voluntarily completed by English-speaking patients seeking care at US fertility centers. FertilityIQ (www. ferti lityiq. com) is an online website that contains extensive fertility-related resources (with both free and paid resources), including clinic and physician satisfaction questionnaires, courses on fertility and assisted reproductive technology topics, and information from fertility experts. All surveys were conducted in English from July 2015 to December 2020. FertilityIQ has voluntary surveys available on most United States fertility clinics and doctors. This study was determined to be exempt from Institutional Review Board by the University of California, San Francisco, as this did not involve human subjects and only deidentified survey data were used for analysis.

Study and participants
The FertilityIQ questionnaire has been described in prior literature [10]. To summarize, the online survey contains 115 questions on a variety of fertility-related topics including patient demographics, fertility diagnoses and treatment, and clinic and doctor information. The full survey is available at www. ferti lityiq. com. The survey includes both mandatory and optional questions (with the majority being optional), as well as skip logic where applicable. The survey contains a variety of question formats, including multiple choice, free text, and Likert scale, depending on the question of interest. All surveys were voluntarily filled out by patients with experience at U.S. fertility clinics. We removed entries where one person reviewed the same clinic more than once, and used the latest or most recent assessment. FertilityIQ had access to personal and identifying information prior to providing the data to our investigators. The FertilityIQ team removed duplicate reviews and those where an email address was connected to reviews of two doctors. In the case of reviews for more than one doctor, the most recent review was included and earlier reviews were excluded from the dataset. The data were then de-identified before being provided to investigators. Additionally, less than 2% of all reviews were from respondents who provided multiple clinic reviews, and this population was not excluded from the results in order to capture as many clinic reviews as possible. In most versions of the FertilityIQ survey, participants are instructed to only post reviews for the doctor who has been primarily responsible for their care during that cycle. Patient responses and clinic experiences were not verified.

Racial/ethnic and outcomes classification
We classified patients into the following self-reported racial/ ethnic groups based on the FertilityIQ questionnaire: Caucasian, Black, East Asian (EA), South Asian (SA), LatinX, Middle Eastern (ME), and Native American (NA), according to the ethnicities surveyed in the questionnaire. Multi-racial or unknown race were excluded from the analysis. Race/ ethnicity of the treating physician was not available in the data.
In terms of classification of doctor and clinic satisfaction outcomes, we used methodology similar to a prior analysis on FertilityIQ general patient satisfaction [10]. Patient and clinic satisfaction was classified according to the answer to the question ''Would you recommend this fertility clinic/ physician to a best friend?" which was answered on a 1-10 scale (1 = no, 10 = very likely to recommend). This is a required question on the FertilityIQ questionnaire. We evaluated doctor and clinic satisfaction as both categorical and continuous outcomes. For categorical outcomes, a score of 9 or 10 was categorized as a positive patient satisfaction outcome. Per prior literature, this positive patient satisfaction outcome was chosen based on business literature on net promotor score (NPS) in customer surveys, which is designed to ''limit the promoter designation to only those who are the most enthusiastic consumers to avoid 'grade inflation' that occurs with traditional customer-satisfaction assessments'' [10,18]. For continuous outcomes, the exact numerical score (1-10) was used in analysis.

Statistical analysis
We investigated the outcomes of clinic and doctor satisfaction as both categorical and continuous outcomes, using logistic and linear regression models, respectively. We also compared the baseline characteristics of the different racial groups using Chi-squared, Fisher's exact test, or ANOVA, as appropriate. We studied the outcomes using univariate and multivariate models. In our multivariate models, we adjusted for both demographic confounders, as well as other confounders which may affect satisfaction with fertility clinics/ doctors based on prior literature. Demographic confounders included age (continuous), insurance status (Yes/No), gender (Male, Female, Other, Decline to state), lesbian/gay/bisexual/transgender/and queer + (LGBTQ +) status (Yes/No), income ($0-49 K, $50-99 K, $100-199 K, $200-499 K, $500 K +), education (less than high school, high school (HS)/GED, some college, bachelor's degree, master's degree, professional degree, doctorate), state, and profession (multiple choice question: architect, business executive, consultant, doctor, engineer, entrepreneur, graphic designer, investment banker, investor, lawyer, marketer, nurse, operations, other, professor, sales representation teacher, writer).
Patient satisfaction confounders included questionnaire elements such as results of treatment, treated as number/ human, good communicator, trustworthy, billing, wait time, convenience, understand, adjustment of plan, presence of primary RN, error made, frequency of seeing doctor (patient satisfaction questions were either a Likert scale 1-5, Yes/No, or multiple choice). In particular, we included the results of treatment as a confounder (full question: Did you get the results that you wanted while you were under the care of [Field-Doctor]?), as this is likely to affect patient perception of the doctor and clinic. We were not able to directly assess treatment response as a confounder, as patients were undergoing multiple types of fertility treatments which make it difficult to apply a uniform outcome measure. Additional details on these satisfaction questions have been provided in Supplementary Appendix. These patient satisfaction confounders were chosen based on factors found to be significantly associated with a positive fertility clinic experience using previous FertilityIQ data [10]. Variables included were found to not be collinear in the analysis.
All tests were two sided with significance at the alpha = 0.05 level. Statistical analyses were performed using Stata Version 15 (Stata Corporation).

Results
Our total sample size included 21,472 unique responses (15, In terms of baseline characteristics, the groups were overall similar on age (p = 0.61, ranging from mean of 35.72 years for Native American patients to 36.76 years for Middle Eastern patients). The groups significantly differed on all other demographic and patient satisfaction characteristics (p < 0.05), though not all differences were clinically significant. Over 90% of participants identified as female and non-LGBTQ + . Caucasian, EA, SA, and Middle Eastern groups reported higher income with a larger percentage of respondents over $200 K per year, and a higher percentage of respondents with college degree or higher. Unexplained infertility was the most common diagnosis across all races, followed by male factor and PCOS. Racial/ethnic groups also differed significantly on all patient satisfaction confounders (p < 0.01), though not all differences appeared to be clinically significant and may have been driven by the large sample size. Patient satisfaction characteristics differed significantly between groups, likely due to larger sample size as most values did not appear qualitatively different. Most Likert scale values were greater than 4 (scale 1-5 with 5 being maximum). Clinical errors were reported by 25.1-34.9% of respondents (which may be partially due to patient perception), which encompassed a wide variety of errors as defined by the questionnaire (see Supplementary Appendix). Respondents reporting that they obtained their desired treatment result ranged from 36.9% to 44.0%, with an additional 20-30% reporting that they were not sure or it was too early to know; as infertility success rates can be limited by multiple factors (most notably age), these responses are not surprising.      In the multivariate adjusted analysis for clinic satisfaction (Table 3C, D), we found that EA had borderline lower satisfaction with clinic in logistic regression (OR 0.74, 95% CI 0.55-1.00 p = 0.05), while significant differences were not found for other ethnic groups. For clinic satisfaction, as a linear outcome, no group significantly differed from Caucasian patients. In the multivariate adjusted analysis, most patient satisfaction confounders were highly significant, while most demographic confounders were not significant with the exception of insurance and certain income groups in terms of association with doctor and clinic satisfaction outcomes.
In a separate analysis adjusting for demographic confounders only, we found that EA rated both their physician and clinics lower, though other groups were not significantly different (with the exception of SA rating their clinics lower on categorical satisfaction only) (Supplementary Material).

Discussion
In summary, in this large cohort of self-reported fertility care satisfaction, we found that after adjustment of confounders, Black patients reported higher doctor satisfaction and East Asians reported borderline lower clinic satisfaction. Despite some differences in unadjusted satisfaction among multiple racial/ethnic groups, most differences did not persist after adjustment for demographic and patient satisfaction confounders. These patterns are particularly interesting given the known disparities in ART outcomes, with Black patients consistently being found in literature to have worse outcomes than other racial groups [1]. Overall, our data are reassuring in that minority groups generally do not appear to report worse satisfaction scores with fertility care, though more investigation is needed on both fertility patient satisfaction as well as outcome disparities in outcomes by race/ ethnicity.

Comparison with existing literature
Prior literature has consistently reported disparities in ART outcomes, with minorities and particularly Black patients having poorer outcomes when compared to Caucasian patients due to unclear etiologies [1-3, 6, 19]. Studies have demonstrated poorer outcomes for Black women even in PGT and donor cycles [4,5]. Disparities in ART outcomes for Asian and Hispanic groups have also been reported [6,[20][21][22][23]. It is important to note that the data on this subject of racial/ethnic disparities in ART outcomes may be limited by incomplete report of race/ethnicity in datasets used for analysis on this subject [24], leading to possible bias in the results. Additionally, studies have typically reported minority groups of Black, Asian, and Hispanic, even though these groups may contain substantial heterogeneity and additional subgroups. Literature on patient perspective on self-reported fertility care satisfaction by race/ethnicity is extremely limited, despite the more extensive literature on disparities in outcomes as above. A study of 1460 patients at an academic fertility center in Illinois found that Black women (14.7%) were more likely to report that race was a barrier to getting fertility treatment compared with White (0.0%), Hispanic (5.1%), and Asian (5.4%) women, though satisfaction with fertility care was not assessed [9]. A prior FertilityIQ analysis of 7456 women from 2015 to 2018 reported a number of factors associated with positive patient-reported experience (including positive results, being treated like a human instead of number, good communication and expectation setting by doctor, shorter wait times, scheduling, and billing satisfaction) [10]. This study did evaluate race/ethnicity as a possible covariate. In the univariate (unadjusted) analysis (published in online supplemental material), they did find that there was a difference in report of a positive experience based on race (p = 0.0002) with similar findings to our study-Black patients were more likely to report a positive experience and South or East Asian patients were less likely to do so (compared to White patients). However, race was not found to be a significant predictor of satisfaction in their multivariable model. Our analysis has expanded upon this earlier data with more detailed racial stratifications, a substantially larger sample size, and using a more concise list of confounders. In addition, our study examined satisfaction with both doctors and clinics separately, while the prior study looked at clinic satisfaction only as a measure of positive patient-reported experience.
A few other studies have investigated patient satisfaction in fertility care, with an older study of 1499 patients in the Netherlands reporting high general patient satisfaction (94%), with waiting times, information provision, and emotional support the least positive aspects of care [11]. Prior literature includes smaller studies that have investigated multiple aspects of satisfaction with fertility care, including general patient satisfaction and provider perception [12][13][14], though study on fertility satisfaction by ethnicity is very limited.
In other fields of medicine, a large study of around 250,000 Press-Ganey surveys found that Asian, younger and female patients provided less favorable ratings than other race/ethnicity older and male patients [15]. This is in line with our findings of EA reporting lower clinic satisfaction than other groups. Other large studies have suggested that minority groups may report lower satisfaction than Caucasian groups with medical care, including a European study of 138,878 cancer patients [16] and a study of 7795 patients     on satisfaction with physician, nursing, and overall care during hospitalization [17]. In contrast, a smaller study of 527 surveys found no differences in outpatient pediatric surgical care by race/ethnicity or socioeconomic status. These findings from other studies may not be completely generalizable to fertility care given the differences in patient population and higher percentage of self-pay patients in fertility. In general, given the limited research, more research in patient satisfaction by race/ethnicity in medicine (not just fertility care) is warranted. Our analysis is the largest and one of the first to investigate patient satisfaction in fertility care by race. We found that despite disparities in ART outcomes, Black patients were the only group to rate their physician satisfaction significantly higher than other groups. EA were the only group to rate their clinic satisfaction as significantly lower than other groups, while other disparities were not seen after adjustment for confounders. As expected, there were baseline differences between the groups in terms of demographic and clinic satisfaction characteristics. Though some disparities were found on univariate analysis, after adjustment of all confounders, most differences in unadjusted satisfaction scores no longer persisted with the exceptions above. Interestingly, Black patients did not report lower satisfaction despite literature suggesting worse ART outcomes in this population. Overall, data are reassuring in that minority groups do not appear to have worse satisfaction with fertility care, with the exception of EA for clinic satisfaction. However, more investigation is needed into the underlying disparities in fertility care for minority populations.

Strengths and limitations
The strengths of this analysis include the large sample size, detailed information available on potential confounders, and the novelty of the subject matter. FertilityIQ is the largest database of patient satisfaction on fertility care in the United States, and our large sample size is an asset in allowing seven categories of racial/ethnic stratification. Additionally, the prior literature on patient satisfaction by race/ethnicity in fertility care is extremely limited, and ours is by far the largest study in this area. We were also able to investigate racial/ethnic groups including Middle Eastern and Native American which have traditionally not been as frequently studied (though sample sizes were limited for these groups), and were also able to study East Asian and South Asians separately (as this is a heterogeneous group which has often been studied as one racial category). FertilityIQ also collects extensive demographic information which was used in multivariate analysis, as some of these factors may be confounders for patient satisfaction.
The limitations of the study include its cross-sectional nature, self-reported voluntary survey data (which may be subject to bias in terms of recall bias or who chooses to fill out an online survey), and small but statistically significant differences that may not be clinically significant (as a result of large sample size). The survey was available online but not sent out to patients, so patients had to choose to go to the website to voluntarily, which may lead to response bias and inability to assess response rate. Additionally, there may be cultural differences in terms of who chooses to fill out a survey and how different groups respond to survey questions, or cultural differences in expectations of care. It is important to note that given known disparities in access to care, this survey can only be interpreted in the context of those who were able to obtain fertility care. The survey was also limited to English-speaking patients, which is a potential bias in the analysis by not capturing non-English-speaking patients. We also had to exclude multi-racial respondents from the analysis due to difficulty analyzing multi-racial data; however, as the population of multi-racial individuals will continue to increase, is it important to develop ways to adequately analyze this population. Though the data must be interpreted with caution, our large sample size gives us confidence in our results for the different racial/ethnic groups studied.

Conclusion
In summary, some but not all minority groups differed in their self-reported perception of satisfaction with fertility clinic and doctors compared to Caucasian patients. On fully adjusted analyses, some of these differences were no longer apparent, though Black patients continued to rate doctors more highly, while EA rated clinics lower. Cultural differences towards surveys may contribute to some of these findings, and satisfaction by racial/ethnic group may also be modified by results of care.

Supplementary Information
The online version contains supplementary material available at https:// doi. org/ 10. 1007/ s00404-023-07043-3. Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.