The objectives were to describe the demographic characteristics of children with Fragile X syndrome (FXS) and to determine predictors of attendance at Fragile X (FX) clinics. Findings from the Community Support Network (CSN) and Our Fragile X World (OFXW) samples showed that children who attended FX Clinics were mostly male, high-school aged or younger, and white, non-Hispanic. Using logistic regression models, awareness about FX Clinic services, guardian education, and income (CSN), and child age, family income, and total number of co-occurring conditions (OFXW) were predictors of clinic attendance. Demographic and child characteristics accounted for a large portion of the explained variance. Importantly, symptom severity and parent knowledge about services were independent predictors beyond the demographic characteristics of families.
There has been an increase in outreach efforts to increase health services use by children and adults from different racial and ethnic groups and socioeconomic backgrounds (Behrens, Lear, & Price, 2013; Roux et al., 2012; Wildman & Langkamp, 2012). With improvements in access, including closing geographic distances, telephone appointments, and telemedicine, there is a need to ensure that these underrepresented groups (hereafter represented by the acronym URG) are reached and obtain the same level of services as the broader population. While outreach has improved care for some pediatric populations affected by adverse health outcomes (Fifield et al., 2010; Katz, Volkening, Butler, Anderson, & Laffel, 2014), it remains difficult to reach URG affected by disorders and conditions requiring behavioral and mental health services. In particular, these include developmental disabilities, such as autism spectrum disorder (ASD) and intellectual disability (ID; Chiri & Warfield, 2012; Levine & Sullivan, 2014; Pena-Ricardo, Schrager, Williams, & Vanderbilt, 2014; Zablotsky, Kalb, Freedman, Vasa, & Stuart, 2014), such as FXS: Martin et al., 2013), the focus of this article.
FXS is the most common cause of inherited ID (Hersh & Saul, 2011; Martyn et al., 2013), and is the most common known single-gene cause of autism (Muhle, Trentacoste, & Rapin, 2004; Schaefer & Mendelsohn, 2013). FXS is caused by a mutation in the FMR1 gene on the X chromosome that interferes with the production of its protein (FMRP, an RNA-binding protein involved in regulation of translation). FMRP is important for cognitive ability and function. Absence or significant reduction of FMRP levels resulting from the FMR1 full mutation causes intellectual disability, behavioral and learning challenges, and various physical characteristics in individuals with FXS (Boyle & Kaufmann, 2010; Hagerman & Hagerman, 2002). Although FXS occurs in both genders, males are more frequently affected than females and generally with greater severity (Bailey, Raspa, Olmsted, & Holiday, 2008; Crawford, Acuna, & Sherman, 2001; Hustyi et al., 2014).
The Fragile X Clinical and Research Consortium (FXCRC) is a group of specialty clinics across the United States (US) that began in 2006 to treat individuals with FXS, primarily children in pediatric settings, but also adults at many of the clinics. The primary goals of the FXCRC are to provide best-practice care to individuals with FXS and their families, and to collaborate on research to advance knowledge and care in FXS. The primary research project of the FXCRC is a registry and database project called FORWARD (Fragile X Online Registry With Accessible Research Database) that was initiated in 2011 and collects longitudinal data on the natural history of FXS including medical, behavioral, and neurodevelopmental features. The FORWARD project is funded through a cooperative agreement with the Centers for Disease Control National Center for Birth Defects and Developmental Disabilities (CDC-NCBDDD).
The FXCRC was initiated by the National Fragile X Foundation (NFXF), the leading support and education organization with the mandate to engage the broader population of families and individuals affected by a fragile X-associated disorder (FXD). In addition to the FXCRC, the Community Support Network (CSN), the national volunteer program of the NFXF, serves to connect and support individuals and families affected by a FXD. Currently, the CSN is comprised of approximately 50 local groups across the US, with about five to ten families per group. The FXCRC, NFXF, and CSN have a vested interest in improving access to services for URG.
Based on pilot data collected from 2009 through 2011 (the forerunner to FORWARD), the FXCRC clinic population was predominately Caucasian (85%), with 7% identifying as African-American, 6% as Latino/Hispanic, 1% as Asian and 1% as other (Kidd et al., 2013). According to the 2010 U.S. Census Bureau, approximately 64% of Americans identified themselves as non-Hispanic/Latino White, 12.6% as non-Hispanic/Latino Black or African-American, 16.3% as Hispanic/Latino, and 5% as Asian (https://www.census.gov/prod/cen2010/briefs/c2010br-02.pdf). Studies of diverse populations within the US have not been large enough to determine whether the prevalence of FXS differs among ethnic/racial groups. (Coffee et al., 2009; Crawford et al., 2002; Tassone et al., 2012). Thus, racial and ethnic minorities appear to be underrepresented at FXCRC clinics.
In addition to race/ethnicity, we seek to understand other factors that might be associated with attendance at Fragile X (FX) Clinics, including socioeconomic status (SES), distance to a FX Clinic, or access to insurance. Lack of awareness of the FX Clinics and their services or of the needs/care of children with FXS may play a role. Finally, a child's primary care provider's knowledge of FXS and his or her relationship with the family may be related to clinic attendance.
In order to better ascertain whether increasing access to services is being achieved for URG with FXS, we collected data from several sources to characterize families who have at least one child with FXS. The ultimate objectives of this report are to determine whether attendance at clinics is representative of the broader FXS community and to determine predictors of attendance at clinics. With this knowledge, it may be possible to identify potential gaps that can be addressed to improve access for all families with FXS.
The surveys and registries that were used to collect data on families with FXS include four sources of data that encompass the most comprehensive data sources of individuals with FXS and their families. To date, there is no comparison of the demographic descriptions, symptomatic and behavioral features, and health-seeking experiences of these families and their children with FXS. This is one effort to describe what gaps may exist in access to services and predictors of non-attendance at clinics that might be approached with outreach efforts. Data from two surveys were used to compare families who attended a FX Clinic to those who did not. These included a) a survey given to members of the FX parent support groups across the U.S. (the CSN groups described above) in 2014; and b) the Our Fragile X World (OFXW) national survey using data collected in 2012.
The CSN groups are support groups that exist to hold community events, offer support to families, bring families together, and fundraise for FXS services and research. They are generally led by volunteer parents of a child with FXS. The structure and some seed funding for events is provided by the NFXF.
The subjects were male or female members of their local CSN groups, 18 years or older, who were parents/guardians of children/adult offspring with a full mutation or FXS. The study was approved by the institutional review board (IRB) of Emory University and consent was obtained at entry to the online survey. Only one family member was asked to complete the survey on the youngest person with FXS in the household. The youngest person was chosen as the index person for the survey, as clinic attendance is weighted towards younger children due to clinics being located primarily in pediatric settings. Subjects were recruited through the CSN Program Coordinator (HUR), by emailing a letter template which described the purpose and procedures of the study to CSN Leaders in September 2014. The CSN Leaders then emailed the letter template with the survey URL included to each member family in their local group, with up to three follow-up requests as needed. This cross-sectional survey was administered via Survey Monkey (SurveyMonkey Inc., Palo Alto, CA), with modified settings to ensure that identification was removed so that this could be considered an anonymous survey.
For the CSN survey, we calculated participation rate as the number of responders divided by the approximate number of informally counted families who were active members of the CSN prior to initiation of the survey, or 375 families. There were 344 families in the community who responded to the CSN Survey, leading to a 92% response rate. Of those who responded, 26 families did not have children with the full mutation of FXS and were excluded. Participation rates were not possible to calculate for OFXW, FORWARD or the The Fragile X Research Registry (FXRR) due to their study designs.
There were 28 questions regarding demographic characteristics (age of child, race/ethnicity, rural/suburban/urban residence, guardian education, marital status, employment, insurance coverage and household income), severity of FXS, whether child was first child diagnosed, and attendance at, knowledge of, and attitudes toward FX Clinics. It was important that CSN Leaders did not know the identification of members completing the survey so that unfiltered responses on sensitive data items could be given regarding education, service use, and socioeconomic circumstances.
OFXW is a survey research registry of more than 1,000 families who have a child with FXS. Data reported here are from a national survey of families conducted in 2012. The OFXW Survey was reviewed and approved by the IRB of RTI International. Families were recruited to the registry through invitation letters mailed by the NFXF and FRAXA Research Foundation and announcements posted on their web sites. A total of 459 families who had a child with the full mutation 3 years of age or older completed the items asking about attendance at FX Clinics. One parent completed the survey on behalf of the family; for this study, we included data on the youngest child in the family. Respondents chose to complete a web-based survey (94%) or a phone interview (6%). Consent was obtained prior to the start of the survey.
When enrolling in the survey research registry, respondents provided demographic information about the family (e.g., family income), as well as information about each child such as date of birth, gender, and genetic status. Per IRB guidelines, respondents were able to answer questions for all minor children and adult males 18 and older, but only for adult females 18 and older for whom they had legal guardianship. Families also reported on a variety of characteristics for each child, such as whether the child had ever been diagnosed or treated for several co-occurring conditions (i.e., attention, hyperactivity, aggression, self-injury, autism, seizures, anxiety, depression, developmental delay). Other items asked about attendance at specific FX Clinics and characteristics of the child's primary care provider.
The demographics of families who attend FXCRC clinics in the FORWARD Study and in the Fragile X Research Registry (Abbeduto, Mallick Seltzer, & Piven, 2011), a collaborative national contact registry used for research study recruitment, are also provided below. These two sources of data provide an additional profile of families attending FX clinics or engaging in research, respectively. We note that there is overlap of participants among all of these sources of data and thus we did not attempt to conduct statistical comparison among sources, only within a source of data. The CSN Survey attempted to quantify this overlap. Among those who completed the CSN survey, about a third of the families reported that they also participated in OFXW, FXRR, or FORWARD; about 7–13% of the total CSN survey sample were in only one of these. Whether the same individuals were represented in the different data points presented in the paper was not known. The CSN Survey was designed specifically to obtain data on issues of clinic attendance whereas OFXW was broader in its purpose. FORWARD and FXRR did not have internal data available to compare those who did and did not use FX Clinics. Thus CSN and OFXW are used to identify barriers and facilitators of clinic attendance. Some questions overlap in these surveys and some are unique to each. These are reported in the following general content sections.
FORWARD is a multisite, observational study, funded by the Centers for Disease Control and Prevention. The project includes data from individuals with FXS and their family members, collected by 25 FX Clinics that were affiliated with the FXCRC at the time of this study. Clinics received IRB approval from their associated institutions. FORWARD collects demographic information on families with FXD through a one-time registry form. The longitudinal database collects more extensive clinician- and parent-reported data on those diagnosed with FXS. Data are collected primarily from those ages 0–24 years, although individuals of any age can participate. Clinicians collect individual data primarily during the patient's regular FX Clinic visit. Data used in this study were collected for the longitudinal database from September 7, 2012 to August 31, 2014 and linked data from the registry form registry form. These data have gone through extensive quality control and form the FORWARD Dataset Version 1.0.
The Fragile X Research Registry is a collaborative national contact registry used for research study recruitment. It was established in 2009 by combining the resources of the Intellectual and Developmental Disabilities Research Center (IDDRC) Registry Cores at the University of North Carolina at Chapel Hill (UNC-CH) and the University of Wisconsin at Madison. This project was reviewed and approved by the IRB of the University of North Carolina at Chapel Hill. The results of the development process were compiled in a summary document (Abbeduto et al., 2011) to be disseminated to groups seeking to establish registries. In addition to merging their existing registries, the sites launched a new web enrollment site in June, 2010. Registrants consent on-line and provide contact and demographic information, basic diagnostic data (e.g., fragile X genetic status, additional IDD or medical diagnoses) and study participation history which are used to determine eligibility for research studies. The Registry staff makes a follow-up contact to verify information submitted online, confirm authorization to provide consent, and obtain laboratory reports confirming Fragile X genetic status.
Definition differences between studies
The term “guardian” was used differently between studies. While the CSN and FORWARD used this term with the accompanying adjective “primary”, the OFXW and FXRR used guardian for the respondent, but not with the clarification of whether the individual was the primary guardian or caregiver. The respondents for all of the studies are likely to be the mother or father (biological or adoptive parent) reporting on behalf of the family versus other caregivers. The employment field may have pertained to the mother or father depending on the interpretation of the question. This field would likely be most affected by guardian status, given that there may be less employment for the mother due to caregiving of a special needs child. This could yield differences across studies for employment, in particular.
The CSN and OFXW surveys form the bulk of the analyses for this article; the FXRR and FORWARD Projects are limited to Table 1 analyses.
The frequency of demographic or other characteristics is presented as a number (percentage) or mean ± standard deviation according to study population. For the CSN and OFXW surveys, data were stratified by attendance at clinics and tests of associations within a survey for attribute variables used the chi-square statistic for dichotomous variables, the Cochrane-Armitage trend test for ordinal variables, or the t-test for continuously distributed data.
Multiple logistic regression models were used to examine factors that predicted clinic attendance using the CSN and OFXW surveys. For each survey, we used comparable variables and their definitions as best as possible (these are noted in tables). CSN survey data were examined over three models; OFXW survey data were examined over four models. In both sets of analyses, blocks of variables were entered using a hierarchical structure across three steps, and only statistically significant predictors were retained in the final models.
In the first step, a block of demographic variables associated with the child and their family, including the age, gender, and race/ethnicity of the child, family income, highest level of education achieved, employment, and marital status were entered into the model. A block of variables associated with the child's health and behavior, including the number of co-occurring conditions, and variables associated with overall health, symptom severity, and behavior, were added in step two, adjusting for variables that were statistically significant in the previous step. Finally, we entered variables, one at a time, that were associated with knowledge, attitudes and beliefs about FXS and specialty clinics or the relationship between the primary care provider and the parent/guardian in the third step. Each variable in the third step was adjusted by variables found significant in the previous two steps. Analyses were performed using the Statistical Analysis System v9.4 (SAS Inc., Cary, NC) and Stata IC (StataCorp LP, College Station, TX). The Wald chi-square test was included to test whether the group of new variables significantly increased the fit of the model. Odds ratios were used to predict the odds of clinic attendance, and 95% confidence intervals (CI) were used around estimates of the odds ratio.
Overall Comparison of Demographics
Overall, the CSN population of the youngest children with FXS was male, high school-aged on average, and largely self-reported non-Hispanic white (Table 1, All Subjects/CSN column). Family characteristics suggested high percentages of married respondents, those with a college education or more, part- to full-time employment, and relatively high incomes (nearly two-thirds with greater than or equal to $75,000 annual incomes). This compares similarly to the OFXW and FXRR populations on the available demographic data (Table 1, All Subjects/OFXW and FXRR columns). Region of residence was not included in the table, but residence was similar across populations, except for a higher southern population in FXRR. CSN parent/guardians were more highly educated and more likely to be employed than those in the other studies.
Demographics and Attendance at FX Clinics
In terms of clinic attendance, the FORWARD population, by definition, includes only subjects who have attended FXCRC clinics. The comparison here between study populations suggests that FORWARD consists of more racially and ethnically diverse families and single parent families with lower household incomes than families reported to attend clinics from the CSN and OFXW samples (Table 1, Attended clinic columns). The FORWARD dataset included more individuals attending clinics who reported lower parental education levels (HS degree or less [9.7%] and some college [24.6%]) when compared with the CSN sample (HS degree or less [6.9%] and some college [19.3%]), but not the OFXW sample (HS degree or less [4.3%] and some college [28.3%]). The age of FORWARD participants was also lower as compared to all other groups, which is likely due to the targeting of a younger age group for FORWARD.
Comparisons, related to attendance and non-attendance at a clinic, were also made within the CSN and OFXW samples. In the OFXW sample, children who were younger were more likely to attend a clinic (p < 0.001, Table 1). In both samples, families with higher levels of parental education (both p < 0.002) and higher parental incomes (CSN, p < 0.001; OFXW, p = 0.03) were more likely to have a child who attended a clinic (Table 1). Also, in the CSN Survey, the marital status was significantly different with clinic attendees more often being with a partner (p < 0.01). The residence of the family (urban, rural, suburban) was not different between attendees and non-attendees in the CSN survey (p > 0.27), although the top reason for not attending clinics was that there were no FX Clinics nearby (32%).
Attributes of Children and Attendance at FX Clinics
The first section of Table 2 (Child Characteristics) summarizes the frequencies of the variables used to describe the attributes of the child's health and behavior in the CSN and OFXW studies. In the OFXW data, the total number of co-occurring conditions was statistically associated with clinic attendance (p < 0.005, Table 2). Within the CSN population, severity of symptoms or behaviors (mild, moderate, severe) was not associated with clinic attendance (p > 0.27). Of note, anecdotal reports from families suggested that children would not attend clinics due to a variety of factors (sensory issues, anxiety), however, only 1% of families claimed children refusing attendance as a barrier. In the CSN survey, attending a clinic did not depend on whether the index child was the first child diagnosed in the family; 80% of attendees versus 71% of non-attendees, were the first children diagnosed in the family (p > 0.16).
Families' Relationship With Provider/Clinic and Knowledge, Attitudes, and Beliefs About Fragile X Clinics
In the second section of Table 2 (Parent Characteristics), the OFXW survey asked a series of questions related to the characteristics of primary care providers (PCP) including whether they were knowledgeable about FXS and how they interacted with the family (Table 2). Based on univariate statistics, these features were not associated with clinic attendance.
The CSN survey asked respondents about the families' knowledge, attitudes, and beliefs about FX Clinics (Table 2). Using simple univariate comparisons, there was a statistically significant positive association between clinic attendance and families who a) thought that a general visit to a FX Clinic is important, b) were aware of services that an FX Clinic provides, c) knew about FXS, and d) were able to explain it to their primary care doctor.
The CSN survey also found that a large proportion of families who attended FXS clinics would a) return to the last FX Clinic that they visited (83%), b) recommend that clinic to others (93%), and c) come back to that clinic for regular yearly visits even if no current problems existed (84%). With respect to insurance coverage, the CSN survey found that there was very little difference for acceptance and use of insurance between the type of clinic visit (i.e., whether it was for the initial evaluation, follow-up visit for specific problems, or a routine annual follow-up). Roughly 4–6% of visits for each visit type were denied by insurance. There was an association between household income by cost barrier to clinic attendance. Families with lower household incomes were more likely to state that they “can't get insurance coverage” or “can't afford it” (p < 0.005), although numbers were small.
Among CSN respondents who had not visited a clinic, they were asked to check the reasons that described why they had not attended. The top three most frequently checked reasons were 1) 29% of parents stated that their child was well cared for by the primary provider, 2) 18% were unsure of the benefit of the clinics, and 3) 17% did not know about clinics.
Modeling the Association Between Attributes and Clinic Attendance
Multivariate logistic regression models were used to identify the best fit for potential predictors of attendance at clinics using the data from the CSN and OFXW surveys. The strategy was similar, although the specific variables included and their definitions were sometimes different. Data from the CSN survey regarding child (age, gender, race/ethnicity) and parent/guardian (household income, education, employment status, marital status) characteristics were entered into Step 1 (Table 3). Household income and parent/guardian education remained statistically significant and were then included in Step 2 with the addition of the characteristics specific to the child as they related to symptomatology. After adjusting for income and education, the categorization of child symptom severity into mild, moderate, or severe was not statistically significant at the p = 0.05 level and thus did not add any explanatory value to clinic attendance.
Models 1 through 3 incorporated, characteristics of the families' knowledge, attitudes, and beliefs, again adjusting for education and income. A family's belief that it was important to visit a specialty clinic (Model 1) and their knowledge base about FXS (Model 3) were not predictive of attendance (both p > 0.60). Only Model 2 with the addition of awareness about services that a FX Clinic provides (p < 0.001) and adjusting for education and family income, added explanatory value to the prediction of clinic attendance with 29% of the variance explained by this model. The Wald test suggested that this model had the best fit of all CSN Survey models tested in Table 3.
The first step of the multivariate logistic regression using the OFXW data examined the demographics of the child and family (age, gender, income, race, employment status, and marital status). Family income, child age, and gender were statistically significant predictors, with individuals from families with higher incomes, younger children, and male children at significantly higher odds of attending a FX Clinic (Table 4). The second step entered a block of variables describing the attributes of the child, including total number of co-occurring conditions dichotomized at less than five vs. five or more conditions, overall health, mood, ability to interact, ability to listen, ability to adapt, and overall thinking, reasoning, and learning ability after adjusting for significant variables identified in step 1 (age, gender, and income). Age and family income remained statistically significant predictors, as well as total number of co-occurring conditions, with individuals who had more five or more co-occurring conditions at increased odds of attending a clinic. Step 3 completed each of the four models (Models 1 through 4), in which characteristics of the child's primary care provider were individually added to each model after controlling for the variables found to be statistically significant in previous models. However, these were not statistically significant predictors of clinic attendance. The Wald test suggested that step 2, which included the demographics and child functioning variables, offered the best fit for predicting the odds of clinic attendance.
Summary of Findings
The goal of this article was to identify factors associated with attendance at FX Clinics in order to better understand the utilization of services among different subgroups of individuals and families. The purpose of the FX clinics is to provide best practice care in the evaluation and treatment of individuals with FXS and their families, including therapies, genetic counseling and referrals. This study is important, as it identifies populations that may be targeted to help broaden availability of these services, and identifies potential limitations of interpreting results from studies from different registries. Findings across the CSN and OFXW samples showed that those who attended FX Clinics were mostly male, high-school aged or younger, white and non-Hispanic. These individuals came from families in which the parents were married, had at least a college degree, worked full-time, and had incomes of $100,000 or more. The demographic breakdown of the FORWARD sample largely mirrored the CSN and OFXW clinic attendee samples. However, the FORWARD data showed higher percentages of females, individuals who were black or Hispanic, those from households with single parents, lower education levels, or lower incomes. This suggests that FORWARD may be the best source of URG with FXS to date. This might be due to the more urban settings of the FORWARD clinics which may result in more successful outreach to URG or the ability to recruit individuals in person. Another possibility may be that FORWARD participants have an existing relationship with their provider which may mean they are more comfortable participating in a research registry. Overall, the race/ethnicity frequencies among all four studies suggest that there is a lack of diversity in clinics and in research surveys. This lack could be due to several reasons. First, there may be differences in the prevalence of FXS among racial/ethnic groups; additional studies need to be done to determine if this is true. If major differences do not exist, either URG are not being diagnosed with FXS or they not accessing FX Clinics. In this study, we focused on understanding potential barriers to clinic attendance to try to address the latter explanation.
In terms of predictors of attendance at specialty clinics, demographic and child functioning characteristics accounted for a large portion of the variance explained in the models. Across the two samples, gender (OFXW), age (OFXW), education (CSN), and income (CSN, OFXW) were related to clinic attendance. Research among those with autism spectrum disorders has found that racial and ethnic minorities, older children, and those from families with lower education levels have had difficulty in accessing services (Broder-Fingert, Shui, Pulcini, Kurowski, & Perrin, 2013; Magana, Lopez, Aguinaga, & Morton, 2013; Thomas, Ellis, McLaurin, Daniels, & Morrissey, 2007). In a study using data from the National Survey of Children's Health, children from families with lower incomes were more likely to be diagnosed with learning disabilities and behavior problems, but had comparable general health outcomes to children from higher income families (Larson & Halfon, 2010). More broadly, our results speak to other findings that show disparities in diagnosis and treatment among ethnic minorities and those from lower incomes when accessing general healthcare (Geiger, 2003). Poor insurance coverage has also led to difficulty in accessing health care services or specialty care among low-income families (Krauss, Gulley, Sciegaj, & Wells, 2003).
Symptom severity was an additional predictor for the OFXW data but not the CSN data. This factor was considered relevant a priori given that parents may seek more expert help with co-occurring conditions, such as seizures, autism, or problem behaviors. However, prior research shows that families with low-income or -education level who have children with special health care needs may not recognize the need for such services (Porterfield & McBride, 2007).
In the CSN sample, awareness of clinic services was the strongest predictor, with the odds of attending a clinic for those who stated that they had an increased awareness of clinic services was more than 8 times that of those who did not attend a clinic. These results parallel previous research on lack of service utilization among those families with children with special health care needs due to decreased knowledge regarding services available and the perceived need for specialty services (Magana et al., 2013; Porterfield & McBride, 2007). From Table 2, over 50% of non-attendees were not aware of clinic services, indicating that even among families involved in a social support group, there is room to improve on education about clinic services available to assist with behavioral problems, genetic counseling, consult on medication, and to provide speech and other therapies under a single umbrella.
There are recognized limitations to the sampling and design of each study. Both the CSN and OFXW study relied on volunteers for participation, although attempts were made to “capture” the whole CSN population of families with children with FXS. That is, we were able to send the survey link to all known members of the CSN, and thus had the potential to obtain full or 100% participation. Volunteers are self-selected and tend to be different from those who do not participate, although we had no information on the demographic characteristics of families who did not participate. Also, the homogeneity within surveys with respect to race/ethnicity and some SES attributes limited our ability to do more granular investigations. This also may be related to the fact the CSN and OFXW samples were recruited from parent organizations and most completed the surveys online. Thus, for both reasons, the results presented may not be representative of all families with FXS. FORWARD is designed to be a consecutive sample of clinic attendees with FXS. However, given the clinic-based recruitment model for FORWARD, only those attending a clinic and whom consent to participate are represented in the sample. Limited demographic data regarding those who attend clinic but are not participating in FORWARD has been collected, with a view towards addressing the representativeness of the sample.
There was the potential for selection bias in the CSN or OFXW surveys, as those without email addresses or computers could not be included, and lower SES or URG families may be less likely to be included. The surveys may be subject to some information bias in recall of information about satisfaction with specific clinics and services if visits were not very recent. In addition, the reporting by guardians on a single child may not reflect the breadth of experiences with other children and services, or at other times in their history. This might affect the predictive ability of child symptomatology, provider relationship, and knowledge, attitudes, and beliefs on clinic attendance. Finally, there may have been overlap of study participants in the various samples. Because of this, analyses regarding clinic attendance were performed within samples to maintain independence of observations.
Despite these limitations, these data provide important information about who is accessing specialty care at FX Clinics and help to define barriers. The results of this study may have value to help identify specialty clinic use and needs of individuals with other genetic conditions, especially those with an intellectual or developmental disability. Additional outreach to increase the awareness of the FX Clinics and their value remains important for all URGs. Possible ways to increase attendance at clinics for URG could be scholarships for those with financial barriers or telemedicine visits for those who are unable to travel to the clinic. Future research efforts should focus on understanding the barriers and challenges for specific subgroups in accessing FX Clinics. Also it will be important to determine whether there are barriers to diagnosis within subgroups.
This study was funded, in part, by the National Fragile X Foundation. We thank the custodians of FORWARD, OFXW, and FXRR for their contribution of resources to the project. In particular, we thank Dr. Ted Brown as the PI for FORWARD and Dr. Howard Andrews for analysis and oversight of FORWARD data. We thank the Community Support Network for the assistance of their leaders in this project. For OFXW, we thank Don Bailey, Anne Edwards, and Ellen Bishop for their contributions to the project. Importantly, we are indebted to the families affected by Fragile X syndrome. We could not have completed this paper without their good will in participating in the multiple research studies on Fragile X syndrome.