Sex and racial/ethnic disparities in children presenting for short stature evaluation: in-depth analysis at a single center
Article information
Abstract
Purpose
To assess disparities in children referred for short stature evaluation and to evaluate the effectiveness of physician interventions on referral rates.
Methods
Retrospective chart review was conducted on children referred to a pediatric endocrinology center for short stature evaluation between 1/1/2022 and 12/31/2023. Interventions for participating physicians included an educational lecture and electronic medical record alert. Six-month pre- and postintervention referral rates for short stature from a general pediatrics practice were assessed.
Results
There were 747 children (68% males) with a predominance of non-Hispanic White (NHW) children (64%). Females presented at a younger age (P<0.001), lower height (P<0.001), and a greater height deficit (P=0.002) than males. Hispanic children presented with greater height deficits than NHW and non-Hispanic Black (NHB) children (all P<0.05). In those with heights <-2 standard deviations (SDs) (n=192) from the mean, there were no significant sex differences; however, Hispanic children continued to have greater height deficit than NHW and NHB children (all P<0.02). There was no sex difference in those who underwent growth hormone stimulation testing (GHST); however, NHW children comprised the largest racial group. After implementing the interventions in the general pediatrics practice, short stature referral rates improved (15 of 118 referrals [13%] to 25 of 81 referrals [31%], P=0.002).
Conclusions
Disparities in overall short stature referrals were less evident in the subset of children with heights <-2 SD from mean. There was no significant sex bias in GHST, but racial/ethnic disparities remained. Improvement in referring and evaluating females and children from minority groups is still crucial as they remain under referred.
Highlights
· Sex and racial/ethnic disparities present in overall short stature referrals to a single pediatric endocrinology center were less evident in the subset of children with heights less than -2 standard deviations. With implementation of two interventions at an adjacent general pediatrics practice that serves a predominantly Hispanic population, we improved short stature referral rates from their practice to our pediatric endocrinology center from 13% to 31%.
Introduction
Short stature evaluation is a multistep process that begins with concern from the child, caregivers, or primary care provider (PCP). A referral to endocrinology is typically initiated by the PCP. An endocrinologist may pursue laboratory evaluation, bone age assessment, and/or growth hormone stimulation testing (GHST). Sex and racial/ethnic disparities have been reported in the primary care evaluation, subspecialty referral, and evaluation and treatment of children with short stature. Females are screened for growth hormone deficiency (GHD) by PCPs half as often as males, and Black children are less likely to be referred to an endocrinologist than non-Hispanic White (NHW) children [1]. Prior studies have documented a predominance of White males being referred for short stature evaluation [2,3]. In pediatric growth hormone (GH) treatment registries, females and Black children are much less likely to be diagnosed with and treated for GHD [4,5].
The aims of this study were to investigate the sex and racial/ethnic differences among all children referred to a pediatric endocrinology center for evaluation of short stature and to determine whether various physician interventions would improve referral rates from an associated general pediatrics practice.
Materials and methods
1. Study design
Retrospective chart review was performed on all children who presented to a single academic pediatric endocrinology center for short stature evaluation between 1/1/2022 and 12/31/2023. The endocrinology center receives referrals from many pediatric and family medicine practices. Race/ethnicity distribution of the community includes: 45% NHW, 24% Hispanic, 13% non-Hispanic Black (NHB), 13% Asian, 4% Mixed, and 1% Other [6]. Data extracted from the electronic medical record (EMR) of the initial evaluation included sex, race/ethnicity, age, height, and parental heights. The height measurement was converted to height z-score standardized for age and sex based on the Centers for Disease Control (CDC) growth charts [7]. Sex-adjusted midparental heights (MPH) were calculated and converted to height z-scores relative to the CDC growth charts. The height deficit was calculated as the difference between the child’s height and the MPH z-scores. The decision to perform GHST was individualized to each child. The provocative GHST agents used at our center are arginine and glucagon. GH concentrations are measured at baseline and every 30 minutes for a total of 180 minutes. GHD is diagnosed if peak GH concentration is less than 10 ng/mL. Peak GH concentration and GHST outcome were extracted if available.
In addition, we focused interventions on an academic general pediatrics practice that is part of the same health system and serves primarily Hispanic and Black families. In October 2023, an educational lecture was given to the resident physicians who work in the practice. Normal and abnormal growth patterns, height measurement techniques, definition of short stature, and GHD diagnosis and treatment options were reviewed with a highlight on sex and racial disparities in short stature referrals. In November 2023, the EMR of the general pediatrics practice went live with an alert that informed PCPs if at the visit their patient's height was greater than 2 standard deviations (SDs) below the mean for age and sex to prompt PCPs to consider endocrine referral. We compared endocrine referral rates for short stature 6 months pre (June-November 2023) and post (December 2023-May 2024) educational interventions.
2. Statistical analysis
Descriptive summaries of demographics and clinical characteristics were reported as frequency (proportions) for categorical variables and median (interquartile range) for continuous variables. Two-tailed 2-sample test of proportions and chi-square test were used for categorical variables. The Wilcoxon rank-sum test was used for continuous variables. P-values <0.05 were considered significant. Analyses were performed using R ver. 4.3.1 (R Foundation for Statistical Computing, Vienna, Austria).
3. Ethical statement
The Institutional Review Board of the Icahn School of Medicine at Mount Sinai reviewed the study and determined it met the exemption criteria. Written informed consent was not required.
Results
There were 747 new children who presented for short stature evaluation, 506 males (68%) and 241 females (32%) with a male:female ratio of 2:1. The median age at initial visit was 11.8 years (8.9–13.8 years) and median height was -1.5 (-2.0 to -0.9) SD from mean. Table 1 denotes demographic and clinical characteristics, with a predominance of NHW children (64%) and males across all racial/ethnic groups (Fig. 1).
Demographic and clinical characteristics of endocrinology center population, stratified by race/ethnicity
Male:female ratio of each racial/ethnic group among children referred for short stature evaluation.
1. Sex
Females presented at a younger age than males (11.0 [7.8–13.2] years vs. 12.1 [9.5–14.0] years, respectively; P<0.001). Females also presented with a shorter height than males (-1.7 [-2.2 to -1.2] SD vs. -1.4 [-1.9 to -0.8] SD, respectively; P<0.001) and with a greater height deficit (1.1 [0.5–1.6] SD vs. 0.8 [0.2–1.4] SD, respectively; P=0.002).
2. Race/ethnicity
There was no difference in age at initial visit between the NHW, NHB, and Hispanic children. Among the NHW children, females presented at a younger age and with a lower height than males (P<0.001); these sex differences were not present within the other racial/ethnic groups.
Compared to NHW children, NHB children had a greater MPH (P=0.04) but there was no difference in height or height deficit. Hispanic children had lower height and MPH and a greater height deficit than NHW children (all P<0.01). While there was no difference in height between Hispanic and NHB children, Hispanic children had a lower MPH and greater height deficit (all P<0.05).
3. Heights <-2 SD
Of the 747 children, 192 (26%) had heights <-2 SD from mean. The percentages of males (55%) and females (45%) did not differ significantly (P=0.05); there was no sex difference in age or height deficit. These children presented at a younger age than those with heights ≥-2 SD from mean (10.8 [7.2–13.8] years vs. 11.8 [9.4–13.8] years, respectively; P=0.03).
A greater proportion of Hispanic children presented with heights <-2 SD than heights ≥-2 SD from mean (28% vs. 15%, P<0.001). In contrast, fewer NHW children presented with heights <-2 SD than heights ≥-2 SD from mean (57% vs. 66%, P=0.02).
In the children with heights <-2 SD from mean, there was no difference in age or height between the racial/ethnic groups. Hispanic children had lower MPH and greater height deficit than NHW (all P<0.001) and NHB (all P<0.02) children.
4. GHST
In total 174 (23% of all children) underwent GHST. The proportions of males (24%) and females (22%) who underwent GHST did not differ. Females who underwent GHST presented for initial evaluation at a younger age and lower height than males who underwent GHST (all P<0.01). Peak stimulated GH concentrations did not significantly differ by sex.
NHW children comprised the largest proportion of individuals who underwent GHST (76%; Table 2). A smaller proportion of Hispanic children underwent GHST than NHW children (9% vs. 28%, respectively; P<0.001). There was no racial/ethnic difference in age, height, MPH, or height deficit for children who underwent GHST.
Growth hormone stimulation testing (GHST) and growth hormone deficiency (GHD) diagnosis, stratified by race/ethnicity
5. GHD diagnosis
Of the 174 children who underwent GHST, 34% were diagnosed with GHD (41 males, 18 females). The proportion of children who underwent GHST and were diagnosed with GHD did not significantly differ by sex (34% males vs. 35% females). The male:female ratio of children diagnosed with GHD is 2.3:1, similar to the ratio of all children who presented for initial evaluation.
In comparison to 25% of Hispanic children who underwent GHST and were diagnosed with GHD, 35% of NHW and 40% of NHB children who underwent GHST were diagnosed with GHD (Table 2). There was no significant difference in these proportions.
Of the 747 children presenting for short stature evaluation, 130 were diagnosed with a causative condition other than GHD(Table 3).
Diagnoses of the 130 children with a causative condition other than growth hormone defi-ciency
6. Referral rates from general pediatrics practice
In the 6 months prior to interventions, there were 118 referrals from the general pediatrics practice, 15 of which were for short stature (13%). In the 6 months subsequent to interventions, 25 of 81 total referrals were for short stature (31%, P=0.002). The percentage of females referred increased from 13% to 36%, males decreased from 87% to 64% although these changes were not statistically significant. Of the children referred for short stature, 58% were Hispanic, 15% NHB and 3% NHW. There was no difference in racial/ethnic distribution pre- versus postinterventions. There was no seasonal difference within the preintervention interval or postintervention interval.
Discussion
There was a marked predominance of males and NHW children presenting to our center for short stature evaluation. While there was no racial/ethnic difference in age at initial evaluation, Hispanic children were overall shorter than NHW children and had a greater height deficit than both NHW and NHB children. These sex and racial/ethnic disparities are consistent with prior studies [2,8-11].
The decision to undergo short stature evaluation is multifactorial and can be influenced by caregivers’ attitudes and level of concern. In one study, parents believed that short men suffer in self-esteem and personal success whereas short women do not [12]. Grimberg et al. [13] reported comments from parent focus groups held to explore height-related decision making, which included consideration for males to be tall and Black families with a higher threshold to consider short stature a concern than White families. Another study of parents reported males had a higher acceptable height cutoff for short stature than females [14]. When examining sex disparities by race/ethnicity, we noted a male predominance across all groups. Overlooking females and children from minoritized groups for evaluation of growth impairment risks a delay in treatable illnesses that can impair growth and health outcomes (e.g., Turner syndrome, celiac disease). Untreated GHD not only affects adult height but has consequences on body composition, metabolic profile, and bone mineral density [15].
Although in our overall cohort (1) there was male predominance, (2) females presented at a younger age, and (3) females were shorter and had a greater height deficit, in our subset of children with heights <-2 SD from mean, the percentages of males and females did not significantly differ and there was no sex difference in age or height deficit. It is reassuring that we did not see sex disparities in the shorter children who might have been more likely to have organic disease, and they presented at a younger age than those with heights ≥-2 SD from mean, reducing the chance of delaying diagnosis.
In our subset of children with heights <-2 SD from mean, Hispanic children continued to have a lower MPH and greater height deficit than NHW and NHB children. Hispanic children have a higher prevalence of short stature which could explain the differences in height and MPH compared to NHW and NHB children in our study [16-18]. However, the greater height deficit cannot only be attributed to familial short stature and instead may reflect the under-evaluation and under-referral by PCPs, raising concern for potentially biased decision making. Explicit and implicit provider biases have been reported [19,20]. It is crucial for providers to address structural racism in healthcare that may be perpetuating inequitable health outcomes in minoritized groups [21].
GHST is a key step in the diagnosis for GHD. Kamoun et al. [8] reported more males than females underwent GHST but no sex difference in subsequent GH prescription. In contrast, Tanaka et al. [9] did not find a sex difference in the proportion of children who underwent GHST or with GHD diagnosis. They also reported the male:female ratios of those with short stature and those diagnosed with GHD were not significantly different. Our results did not document a sex difference in GHST or GHD diagnosis, suggesting sex did not play a significant role in our endocrine providers’ decision making.
Racial/ethnic disparity remained in the rates of GHST. Despite no significant difference in age, height, MPH, or height deficit in children who underwent GHST, more NHW children underwent GHST than Hispanic children in our study. In a single center of 7,425 children presenting for growth evaluation, NHW children were 1.4 times more likely than NHB children and 1.7 times more likely than Hispanic children to undergo GHST [3]. The etiology of this disparity in our population is unclear but potential reasons may include endocrine providers’ bias in recommending GHST and insufficient parental education on the importance of GHST. Studies are needed to evaluate the clinical decision making of pediatric endocrinologists. We do not have data on the families’ preferred language or what language was utilized during the visits, so it is possible that although interpreter services are available, language barrier may pose a potential cause for lower rates of GHST in Hispanic children.
In our study we had too few Hispanic and NHB children with GHD to compare their clinical characteristics to NHW children with GHD. In a single center study, NHB children treated with GH had greater height deficits than treated NHW and Hispanic children [3]. Of patients enrolled in the Pfizer International Growth Study, Black patients treated for idiopathic GHD had greater height deficits and lower stimulated peak GH concentrations than White patients, suggesting that Black patients were more severely affected at presentation [22].
Physician-focused education about short stature and an embedded EMR alert increased short stature referral rates from a predominantly Hispanic population. Although not statistically significant, rates of females referred increased from 13% to 36%. We anticipate that this rate will continue to improve over time. Future studies to expand and implement these interventions at other general pediatrics practices may improve the disparities present in short stature referrals.
There are limitations to this study. Height was the only growth parameter assessed from the first visit; data on growth velocity, pubertal status, and radiographic/laboratory evaluation were not readily available. These variables may have influenced provider decisions and could contribute to the understanding of the sex and racial/ethnic disparities. While the endocrinology center receives referrals from many pediatric and family medicine practices, this study focused on only one practice. The personnel who obtain the height measurements in the pediatric practice were not reeducated. Lastly, this is a single center study.
In conclusion, the very small percentage of NHB children who were referred for evaluation of short stature is cause for considerable concern. Sex and racial/ethnic disparities seen in our pediatric endocrinology center are less evident in children with heights <-2 SD from mean. We did not find a significant sex bias in those who underwent GHST and were diagnosed with GHD but marked racial/ethnic disparities remained. Interventions to educate and alert pediatricians about short stature resulted in an increase in referral rates. Continued efforts are needed to improve referrals of females and children from minoritized groups.
Notes
Conflicts of interest
No potential conflict of interest relevant to this article was reported.
Funding
This study was supported by the 2022 Improving the Care of Patients with Disorders of Impaired Growth Quality Improvement Grant from Pfizer Inc. The funder had no role in the design, data collection, data analysis, and reporting of this study.
Data availability
The data that support the findings of this study can be provided by the corresponding author upon reasonable request.
Author contribution
Conceptualization: VW, JS, THL, CD, GP, CK, RR; Data curation: VW, JS; Formal analysis: VW; Funding acquisition: VW, JS, RR; Methodology: VW, JS, THL, CD, GP, CK, RR; Project administration: JS; Writing - original draft: VW; Writing - review & editing: VW, JS, THL, CD, GP, CK, RR