If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL, United StatesStanley Manne Children's Research Institute, Ann & Robert H. Lurie Children's Hospital, Chicago, IL, United States
Department of Pediatrics, University of Colorado Anschutz Medical Campus, Aurora, CO, United StatesChildren's Hospital Colorado, Aurora, CO, United States
Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, United StatesChildren's Hospital of Philadelphia, Philadelphia, PA, United States
Department of Pediatrics, Northwestern University Feinberg School of Medicine, Chicago, IL, United StatesStanley Manne Children's Research Institute, Ann & Robert H. Lurie Children's Hospital, Chicago, IL, United States
Diagnosis of cystic fibrosis (CF) through newborn screening (NBS) improves health and survival of people with CF. Racial and ethnic disparities in CF outcomes could be reduced by early diagnosis. It has not previously been reported whether there is equitable age at first evaluation by race and ethnicity. We conducted an exploratory study of the Cystic Fibrosis Foundation Patient Registry to assess disparities in follow-up of infants with CF in a country with universal NBS for CF.
•
Infants described as Black/African American, American Indian/Native Alaskan, Asian and/or from other race groups, and/or of Hispanic ethnicity had a later age at first event (AFE: sweat test, clinical evaluation, or hospital stay) than White, not Hispanic infants and a lower weight-for-age z-score averaged between 12 and < 24 months.
•
A nested cohort study showed that AFE, Black race, CFTR variant class I-III, prematurity and public insurance were associated with lower WFA z-score at 1 year of age.
•
Later evaluation of infants with CF from historically marginalized groups may worsen long standing health disparities. Improving detection of CFTR variants, reducing bias against diagnosis of CF in infants from minoritized groups, and improving timeliness of follow-up for all infants with an out-of-range CF NBS test may reduce these disparities.
Abstract
Objective
We evaluated whether implementation of cystic fibrosis (CF) newborn screening (NBS) leads to equitable timeliness of initial evaluation. We compared age at first event (AFE, age at sweat test, encounter and/or care episode) between infants categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or Hispanic and/or other (Group 1) to White and not Hispanic infants (Group 2).
Methods
This retrospective cohort study from the Cystic Fibrosis Foundation Patient Registry (CFFPR) included infants born 2010-2018. Race and ethnicity categories followed US Census definitions. The primary outcome was AFE; the secondary outcome was weight for age (WFA) z-score averaged 12 to < 24 months. We compared distributions by Wilcoxon rank-sum test and proportions by Chi-square or Fisher's exact tests. A nested cohort study used a linear mixed effects model of variables that affect WFA, chosen a priori, to evaluate associations with 1-year WFA z-score.
Results
Among 6354 infants, 21% were in Group 1. Group 1 median AFE was 31 days (IQR 19, 49) and Group 2 was 22 days (IQR 14,36) (p< .001). Median WFA z-score at 1-2 years was lower in Group 1. In 3017 infants with complete data on variables of interest, AFE, Black race, CFTR variant class I-III, prematurity and public insurance were associated with lower 1-year WFA z-score.
Conclusions
Differences in AFE for infants with CF from historically marginalized groups may exacerbate long standing health disparities. We speculate that inequitable identification of CFTR gene variants and/or bias may influence timeliness of evaluation after an out-of-range NBS.
] In the US, the risk factors best defined are low SES and Hispanic ethnicity, both associated with worse lung disease and decreased survival. Low SES, but not Hispanic ethnicity, is associated with worse nutrition. Diagnosis and early therapy through newborn screening (NBS) prevents severe complications in the first months of life [
Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group.
]. Newborn screening is generally conducted by public health department laboratories that are linked to follow-up programs. Most use a single sample, two-step algorithm. Immunoreactive trypsinogen (IRT), a pancreatic zymogen elevated in the blood of infants with CF [
] is measured; when elevated, DNA is extracted to test a panel of common CFTR variants. A test is out-of-range when elevated IRT and 1-2 pathogenic CFTR variants are detected. Screening results are generally sent to primary care offices, who refer infants with out-of-range tests to facilities with expertise in diagnostic testing and treatment. While rapid referral is recommended for all infants with an out-of-range test, the presence of one, instead of two, CFTR variants may raise less clinical concern and delay evaluation.
Most NBS laboratories use CFTR variant panels that detect most variants in people of northern European ancestry. While race is a social construct, CFTR variant distribution differs by demographic race and ethnicity. About 90% of people with CF in the US categorized as White and not Hispanic have at least one copy of the most common variant, F508del [
CFTR mutation distribution among U.S. Hispanic and African American individuals: evaluation in cystic fibrosis patient and carrier screening populations.
]. Those categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race groups, and/or Hispanic ethnicity less often have F508del [
CFTR mutation distribution among U.S. Hispanic and African American individuals: evaluation in cystic fibrosis patient and carrier screening populations.
CFTR mutation distribution among U.S. Hispanic and African American individuals: evaluation in cystic fibrosis patient and carrier screening populations.
]. In 2019, a parent contacted the Cystic Fibrosis Foundation, asking if disparities in timeliness of diagnosis might affect Black/ African American babies. Also expressed was possible healthcare provider bias against a CF diagnosis in people perceived to have ancestry outside of Europe [
]. We identified no published data on timeliness related to demographics for any genetic disorder on the NBS panel and committed to studying this issue in a planned evaluation of outcomes of CF NBS in the US.
Infants with CF had improved health during years spanning NBS implementation [
Outcomes of infants born during the first 9 years of CF newborn screening in the United States: a retrospective Cystic Fibrosis Foundation Patient Registry cohort study.
]. Our objective for this study was to explore whether infants with CF who are demographically categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity had a later age at first testing or clinical evaluation than White and not Hispanic infants and whether this would be associated with differences in growth. We described subgroups with specific attention to the paradoxical finding that while older children and adults with CF and Hispanic ethnicity have worse lung disease, nutrition is well-preserved [
], this cohort included participants in the CFFPR born 2010-2018 with age at diagnosis and first CF Center event at 0 to 365 days. The CFFPR includes 81-84% of people with CF in the US; methods have been described [
]. Demographic race and ethnicity categories use US Census definitions. Race categories were Black/African American, White, American Indian/Alaska Native, Asian, Native Hawaiian/Pacific Islander, two or more races, none of these or unknown. Ethnicity categories were Hispanic, not Hispanic, or unknown. Trained staff at CF Centers enter data without specific instructions for data collection, or indication of whether self-report is used for demographic data.
The primary outcome was age at first event (AFE), calculated using the earliest date of a sweat test, and/or clinical encounter, and/ or care episode lasting > 24 hours at a CF center (usually a hospitalization). More than one event could occur on the same day. This clinically meaningful measure was chosen a priori because the CF center-reported median age at diagnosis is lower than median AFE. The independent variable was racial and ethnic category. Group 1 included infants in categories Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity, a composite chosen based on the primary objective. Group 2 included infants described as White and not Hispanic. We described subgroups in major categories of a. Black/African American, American Indian/Native Alaskan, Asian and/or other race without Hispanic ethnicity; b. White with Hispanic ethnicity; and c. Black/African American, American Indian/Native Alaskan, Asian and/or other race with Hispanic ethnicity.
We explored a secondary outcome, weight-for-age (WFA) z-score, using World Health Organization growth charts, averaged at all visits between 12 and 24 months of age for univariable analysis. We chose this time frame due to rapid growth in the first two years of life, giving a more stable measurement than a single time point. We also compared height-for-age (HFA) z-score between 12 and 24 months, use of nutritional supplements (via oral or enteral feeding), clinic visits and hospitalizations, and rates of Pseudomonas aeruginosa infection.
We explored other demographic and clinical data that could influence AFE and health [
]. SES measures included parental educational attainment (5 categories spanning less than high school to a graduate degree), health insurance type, median income by zip code (MIZ) and household size. The MIZ was defined by US Census Bureau 2013-2017 American Community Survey 5-Year Estimates. Preterm birth and low birth weight are associated with higher rates of insufficient sweat collection for quantitative chloride testing [
] and lower WFA; we examined gestational age (full term or preterm, defined as <37 weeks GA) and birth weight percentile. Other data relevant to nutrition included CFTR genotype and pancreatic insufficiency. In CFFPR, CFTR variants are classified based on whether the defect causes absent, dysfunctional, or decreased function, as described by McKone [
]. Class I, II and III variants are associated with pancreatic insufficiency, class IV-V are associated with pancreatic sufficiency, and rare variants are often unclassified [
]. We defined pancreatic insufficiency as use of pancreatic enzyme replacement therapy (PERT) during the first year of life. There is significant missing data in CFFPR on the more accurate fecal pancreatic elastase [
Outcomes of infants born during the first 9 years of CF newborn screening in the United States: a retrospective Cystic Fibrosis Foundation Patient Registry cohort study.
Following initial analyses, we conducted a nested cohort study of a subgroup of infants to consider the impact of variables, selected a priori on nutritional status at one year of age. This cohort included only participants with complete data on variables of interest. To further explore these associations, we included race and ethnicity as separate variables, and included gestational age at birth, meconium ileus/ intestinal obstruction, initial sweat chloride value, genotype class, insurance, MIZ category (in tertiles), and CF center (nested in geographic region of birth) on WFA Z-score closest to the first birthday, recorded at 11-15 months.
The CFFPR is approved by local Institutional Review Boards. Written informed consent is obtained from parents or legally authorized representatives for each child's participation.
3. Statistical analysis
Descriptive data were summarized for groups and sub-groups. Tests compared data from Group 1 to Group 2, using complete case analysis. Observations with missing data on a variable were excluded from the corresponding test. Continuous variables were compared by the Wilcoxon rank-sum test and proportions by the Chi-square or Fisher's exact test. A p-value of <.05 was considered statistically significant. Adjustments for multiple comparisons were not made, given the exploratory nature of the study and our focus on primary and secondary outcomes [
For the nested cohort study, a linear mixed effects model with a random program effect, nested in geographic region, assessed the impact of a priori defined variables on WFA z-score, accounting for clustering of patients within programs. The program effect was added to account for CF Center practice variation.
Outcomes of infants born during the first 9 years of CF newborn screening in the United States: a retrospective Cystic Fibrosis Foundation Patient Registry cohort study.
], among 6879 infants born 2010-2018 and enrolled in CFFPR, 154 (2.2%) had a date of diagnosis before birth, i.e., prenatal identification, and 371 (5.4%) had AFE >365 days of age. Demographics of the remaining 6354 infants are summarized in Table 1. The distribution of race and ethnicity categories within the US Census region of birth reflected the general population. Group 1 comprised 21% of infants, who were more often insured by Medicaid, less often had private insurance, and had lower MIZ than Group 2. There were other statistically significant differences that were numerically small. Some SES variables, including parental education, had significant missing data, so while differences were seen in distributions, they may not be representative of the population. Subgroup demographics are described in major categories in Supplemental Table 1.
Table 1Demographics of infants with CF born 2010-2018 classified by race and ethnicity (US Census 2010 Definitions).
Patient Characteristic
Group 1(N=1335)
Group 2 (N=5019)
P value
Gender, N (%)
F
660 (49)
2430 (48)
M
675 (51)
2589 (52)
Race, N (%)
Black/ African American
411 (31)
White
635 (48)
5019 (100)
Other+
289 (22)
-
American Indian/ Alaska Native
21 (7)
-
Asian
21 (7)
-
Native Hawaiian/Pacific Islander
6 (84)
-
None of the above
243 (84)
-
Hispanic ethnicity
No
423 (32)
4776 (95)
Yes
842 (63)
-
Unknown
70 (5)
243 (5)
Health insurance, 0-365 days, N (%)
Any Health insurance++
985 (99)
4091 (100)
.01
More than one type of insurance
119 (12)
650 (16)
.002
Private Insurance Policy
266 (27)
2235 (54)
<.001
Medicaid
640 (64)
1922 (47)
<.001
State special needs program
153 (15)
391 (10)
<.001
Other++
46 (5)
217 (5)
.40
Missing
338
997
Household income based on zip code, US dollars
Median, in thousands (N, IQR)
51 (40, 66)
56 (45, 72)
<.001
Lowest median income quartile in cohort
409 (34)
1047 (23)
<.001
Missing
126
377
US Census Region of Birth
Northeast
163 (12)
612 (12)
<.001
Midwest
185 (14)
1281 (26)
South
380 (29)
1461 (29)
West
289 (22)
689 (14)
Unknown/Foreign
318 (24)
976 (20)
Father's education (first year infant enrolled)
<.001
Less than High School
87 (18)
156 (7)
High School diploma or equivalent
202 (42)
722 (30)
Some College
98 (21)
513 (21)
College Graduate
66 (14)
789 (33)
Masters/Doctoral level degree
24 (5)
222 (9)
Missing
858
2617
Mother's education (first year infant enrolled)
<.001
Less than High School
98 (185)
153 (6)
High School diploma or equivalent
195 (37)
668 (27)
Some College
130 (25)
546 (22)
College Graduate
78 (25)
923 (37)
Masters/Doctoral degree
28 (5)
228 (9)
Missing
806
2403
Number of people in household (participant included, first year in registry)
<.001
2
28 (3)
53 (2)
3
259 (30)
1346 (37)
4
258 (30)
1212 (33)
5
167 (19)
621 (17)
6
72 (8)
234 (6)
7
39 (5)
93 (3)
8 or more
40 (5)
72 (2)
Missing
472
1388
Group 1: infants categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity. Group 2: infants categorized as White and not Hispanic. +other race categories can include one or more other race. ++ health insurance categories are not mutually exclusive. +++ includes military health plans, Medicare, Indian Health Service, and others.
Diagnostic findings and AFE are summarized in Table 2. Median AFE was later and more widely distributed in Group 1: 31 (IQR 19,49) compared to 22 (IQR 14,36) days in Group 2 (p <.001). Both groups' most frequent first event was sweat testing, but Group 1 less often had an encounter (e.g., clinical evaluation) on the same day. Birth weight was missing in more than half of infants. Group 1 infants less often had a positive NBS or prenatal screening, but the difference was small. In contrast, Group 1 more often had DNA analysis reported as a diagnostic finding, suggesting identification of CFTR variants not reported on the NBS test. Meconium ileus incidence was not significantly different between groups. Most infants had no symptoms at diagnosis, but Group 1 had more respiratory symptoms and failure to thrive. Most infants had two CFTR variants reported, but distribution of CFTR variants was different between groups: 67% in Group 1 and 86% in Group 2 had known class I-V variants, suggesting more rare variants in Group 1. More Group 2 patients were prescribed PERT during the first year of life.
Table 2Diagnostic and Clinical Findings in infants with CF born 2010-2018 classified by race and Hispanic ethnicity (US Census 2010 Definitions).
Patient Characteristic
Group 1(N=1335)
Group 2(N=5019)
P value
Age at diagnosis, days, median (IQR)
21 (7, 42)
13 (3, 26)
<.001
Age at first CF event, days, median (IQR)
31 (19, 49)
22 (14, 36)
<.001
First CF Event
Sweat Test N (%)
927 (69)
3117 (62)
<.001
Encounter N (%)
577 (43)
2650 (53)
<.001
Care episode > 24 hours N (%)
100 (8)
407 (8)
0.46
CFTR Variant Class N (%)
<.001
Class I-III
709 (53)
3598 (72)
Class IV, V
185 (14)
701 (14)
Other Variant class, both alleles known
388 (29)
585 (12)
One allele missing or unknown
37 (3)
52 (1)
Both alleles missing or unknown
16 (1)
83 (2)
Sweat testing
Age at first sweat test median (IQR)
36 (24, 70)
28 (18, 55)
< .001
Missing N (%)
99 (7)
486 (9)
Initial sweat test QNS++ N (%)
75 (6)
242 (5)
0.30
Missing N (%)
109 (8)
504 (10)
Sweat test value
Initial sweat chloride, mmol/L, median (IQR)
92 (73, 101)
95 (84, 103)
< .001
Missing N (%)
184 (14)
745 (15)
PERT during first year of life
<.001
Taking PERT N (%)
975 (82)
4156 (87)
Missing
145
266
WHO birth weight z-score (≥37 GA Weeks)
z-score category
0.002
<10th percentile
95 (17)
271 (12)
>=10th - <25th percentile
113 (21)
406 (18)
>=25th - <50th percentile
134 (24)
542 (24)
>=50th percentile
209 (38)
1007 (45)
Median (IQR)
-0.30 (-0.99, 0.51)
-0.07 (0.75, 0.57)
<.001
Missing
784
2793
Signs and symptoms at diagnosis
Meconium Ileus N (%)
161 (15)
692 (14)
0.10
Positive NBS N (%)
1063 (80)
4154 (83)
0.008
Prenatal screening N (%)
25 (2)
191 (4)
<.001
DNA analysis N (%)
315 (24)
977 (20)
<.001
Respiratory N (%)
61 (5)
126 (3)
<.001
Failure to thrive/ malnutrition
117 (9)
238 (5)
<.001
Steatorrhea/ loose stools N (%)
56 (4)
179 (4)
0.28
Group 1: infants categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity. Group 2: infants categorized as White and not Hispanic. ᵇ ++QNS responses are yes or no. ᶜ 0.8% of Group1 infants and 0.3% of Group 2 infants had neither answer recorded. Abbreviations: IQR, interquartile ratio; QNS, quantity not sufficient PERT, pancreatic enzyme replacement therapy.
Fig. 1 shows the distribution of AFE (panel a) and average median WFA (panel b) and HFA z-scores (panel c) at 12–24-months. Nutritional outcomes are summarized in Table 3. At initial clinic visit, there was no difference in median WFA z-score, but Group 1 had lower median HFA z-score and a higher frequency of infants with HFA z-score < 10th percentile. There was no difference in use of supplemental feedings. Because nutritional abnormalities are more frequent in pancreatic insufficient infants, we performed a limited post-hoc comparison of AFE and nutrition outcomes, restricted to infants prescribed PERT, summarized in Table 4. There was a similar difference in median AFE in this analysis, but Group 1 less often had AFE before 30 days of age. There was a similar difference in WFA and HFA z-scores between groups.
Fig. 1Age at first CF center Event and nutritional outcomes in infants with CF born 2010-2018 demographically categorized as Black/African American, American Indian/Native Alaskan, Asian, and/or other race, and/or Hispanic (Group 1) or White and not Hispanic infants (Group 2). a. Distribution of Age at First Event b. Distribution of World Health Organization weight-for-age z-score, average between 12-24 months of age c. Distribution of World Health Organization height-for-age z-score, average between 12-24 months of age.
Table 3Nutritional Outcomes of infants with CF born 2010-2018 classified by race and ethnicity (US Census 2010 definitions).
Group 1(N=1335)
Group 2(N=5019)
P value
Nutritional status at first clinic visit, N (%) reported
WFA z-score category
0.197
<10th percentile N (%)
362 (41)
1333 (38)
>10th - <25th percentile N (%)
158 (18)
718 (21)
>25th - <50th percentile N (%)
174 (20)
741 (21)
>50th percentile N (%)
179 (21)
702 (20)
WFA z-score median (IQR)
-1 (-1.81, -0.19)
-0.92 (-1.72, -0.17)
0.175
missing
462
1525
HFA z-score category
N=873 (65)
N=3491 (70)
0.004
<10th percentile N (%)
319 (37)
1132 (32)
>10th - <25th percentile N (%)
175 (20)
615 (18)
>25th - <50th percentile N (%)
176 (20)
750 (22)
>50th percentile N (%)
203 (23)
994 (28)
HFA z-score Median (IQR)
-0.89 (-1.76, -0.07)
-0.68 (-1.55, 0.12)
<.001
missing
462
1525
Supplemental feeding first year N (%) reportedb
N=1175 (88)
N=4715 (94)
Oral supplements N (%)
476 (41)
1873 (40)
0.69
Enteral tube supplements N (%)
82 (7)
357 (8)
0.49
missing
160
304
Nutritional status, 12 to <24 months N (%) reported
WFA z-score, median (IQR)
-0.02 (-0.72, 0.69)
0.11 (-0.53, 0.68)
< 0.001
missing
217
663
HFA z-score, median (IQR)
-0.66 (-1.50, 0.08)
-0.54 (-1.25, 0.14)
< 0.001
missing
217
672
Group 1: infants categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity to Group 2: infants categorized as White and not Hispanic. bReported as yes/no. Abbreviations: WFA, weight for age by World Health Organization growth charts; HFA, height for age by World Health Organization growth charts.
Table 4First event and nutritional outcomes in pancreatic insufficient infants with CF born 2010-2018 classified by race and Hispanic ethnicity (US Census 2010 Definitions).
Group 1 N=975
Group 2 N=4156
P value
Age at first CF event, days, median (IQR)
29 (17,45)
21 (13,33)
<.001
Age at first event, category, N (%)
< 30 days
509 (52)
2928 (70)
< .001
30-59 days
298 (31)
755 (18)
≥ 60 days
168 (17)
473 (11)
First CF Event
Sweat Test N (%)
611 (63)
2387 (57)
.003
Encounter N (%)
483 (50)
2386 (57)
< .001
Care episode > 24 hours N (%)
98 (10)
401 (10)
0.70
Nutritional outcomes
Weight-for-Age, average at 12-24 months
WHO WFA z-score Median (IQR)
-0.11 (-0.75,0.59)
0.062 (-0.57,0.65)
< .001
Missing
100
440
Height-for-Age, average at 12-24 months
WHO HFA z-score Median (IQR)
-0.79 (-1.59,0)
-0.59 (-1.29, 0.78)
< .001
Missing
100
446
Group 1: infants categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity. Group 2: infants categorized as White and not Hispanic.
Healthcare utilization and infection during the first year of life are summarized in Supplemental Table 2. The median number of CF Center visits was 8 for Group 1 and 9 for Group 2. The rate of hospitalizations for pulmonary exacerbation was higher in Group 1, but the rate of Pseudomonas aeruginosa infection was not.
Fig. 2 shows the consort diagram for 3017 infants with complete data for a priori defined variables included in the nested cohort study. The linear mixed effects model in Table 5 demonstrates that AFE, prematurity, Black race, class I-III variants, and public insurance were associated with lower WFA z-score at 1 year of age. Higher MIZ, compared to median MIZ, was associated with higher WFA z-score.
Fig. 2Consort diagram showing disposition of cohort participants for inclusion in the multiple mixed linear effects model.
This exploratory investigation demonstrates that infants demographically categorized as Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity, are older at AFE and have poorer early nutritional outcomes than those categorized as White and not Hispanic. We found that lower WFA z-score at one year was associated with later AFE, Black/African American race, and public insurance. The descriptive nature of this study, significant missing data for some variables, and the plethora of influences on early life health in CF allow only hypothesis generation regarding the association between AFE and early life nutritional outcomes. Prospective studies that assess barriers to prompt evaluation and implementation of CF care to demographically representative infants should be undertaken. Enhancing collection of sociodemographic data, including reduction of missing variables and self-report of race and ethnicity (where allowed) may lead to targeted interventions in countries or regions with established registries. The experience of people from minoritized groups, whose children were diagnosed with CF after implementation of newborn screening, is critically important to consider through engagement efforts and qualitative studies.
These findings may be clinically important. Early life nutritional deficits in CF [
Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group.
]. The clinical significance of a median difference 8-9 days between groups cannot be assessed using retrospective data. Variation in onset of severe pancreatic insufficiency after birth in CF [
] contribute to early life nutrition in CF. We previously noted that 40% of all infants in this cohort had WFA < 10th percentile at their first clinical encounter [
Outcomes of infants born during the first 9 years of CF newborn screening in the United States: a retrospective Cystic Fibrosis Foundation Patient Registry cohort study.
Pulmonary findings in infants with cystic fibrosis during the first year of life: results from the Baby Observational and Nutrition Study (BONUS) cohort study.
]. Group 1 infants more frequently had symptoms at presentation. The larger proportion of Group 1 infants with AFE > 30 days suggests risk for early, severe complications of CF, including hyponatremic dehydration [
], both sometimes fatal in the first weeks of life. Thus, inequalities in timely evaluation may worsen sociodemographic health disparities in CF.
Group 1 less often had a clinical encounter as part of the first CF event and more often had DNA analysis, suggesting that < 2 CFTR variants were detected by NBS. CFTR panels for NBS vary by state and have changed over time. Since CFFPR does not distinguish whether variants are detected by NBS or other tests, this requires further study. Some states perform CFTR sequencing, but only after one variant is detected on a panel [
], reduces risk of false negative tests, and can be performed in one step, but is not widely used. Since extended gene sequencing detects CFTR variants of unknown clinical significance, an increase in cystic fibrosis screen positive, inconclusive diagnosis (CFTR-related metabolic syndrome in the United States) infants is expected with this approach [
]. Strategies to reduce burdens on families and health care systems, and ongoing efforts to characterize pathogenicity of rare variants, are required. With any algorithm, all infants with out-of-range CF NBS should be evaluated promptly, regardless race, ethnicity, or whether 1 or 2 variants is reported. Increased coordination of follow-up of out-of-range NBS results between primary care providers, CF Centers, and state NBS programs help to achieve this goal.
Given the well-known influence of low SES on health outcomes and mortality in CF and in the general population, it may be especially important to reduce barriers to timely visits at CF Centers arising from geographic and structural barriers to access. We found that Group 1 had lower SES based on several measures. The association between Medicaid insurance and poorer nutrition outcomes should be interpreted in the context of the Medicaid program's role of insuring children with lower SES; Medicaid has unequivocally reduced infant mortality for disorders detected by NBS [
]. Other factors associated with low SES are likely to contribute to worse health outcomes. A study from the United Kingdom showed that NBS does not provide clinical benefits to infants of lower SES [
], but did not report data on timeliness of evaluation. Furthermore, countries with universal health care provision show persistence of adverse health outcomes associated with low SES [
]. Nevertheless, reducing effects of race and ethnicity-related bias in NBS detection of CFTR variants and increasing health care professional knowledge that CF occurs in all populations can reduce compounding health effects of racism [
Limitations of this study include that racial and ethnic categories were limited to US Census definitions and were not systematically self-reported. Use of composite comparison groups 6was a pragmatic decision, based on prior literature and the small numbers in some groups, that does not fully reflect the great diversity of the CF population. Interpretation of the mixed model assumes that those with complete data are representative of registry participants overall. The cohort was enrolled before approval of any CFTR modulator for infants < 1 year of age, and increased disparities in CF outcomes are expected based on reduced eligibility for these disease-modifying therapies in people with CF who are categorized as Black/African American, American Indian/Native Alaskan, Asian, and/or Hispanic [
Further analyses of the relationship between AFE and nutritional outcomes, and of processes that can influence timeliness of referral, are needed. Because pre-symptomatic treatment is the overarching goal of NBS, we advocate for quality improvement activities that improve timeliness of CF diagnosis for all infants, with a focus on equity in diagnostic evaluation and initiation of care.
6. Conclusion
We identified disparities in first CF Center evaluation and clinical outcomes in infants demographically categorized as Black/African American, American Indian/Native Alaskan, Asian, and/or Hispanic compared to those categorized as White and not Hispanic. Timely evaluation and care for all infants with an out-of-range CF NBS is essential for health equity in CF, and further research is needed to better understand these findings and identify interventions. Evaluation of timeliness of care after out-of-range NBS tests for other disorders is also needed to assess whether these findings are unique to CF or occur more broadly. Quality improvement efforts to assure pre-symptomatic diagnosis and treatment should not be delayed.
Funding
Supported by the Cystic Fibrosis Foundation (MCCOLL19QI0 to SAM)
Contributors’ statement
Susanna A. McColley, MD, contributed to conceptualization/design, methodology, investigation, funding acquisition, and resources; drafted the initial manuscript; and edited the manuscript in preparation for submission.
Stacey L. Martiniano, MD, contributed to conceptualization/design, methodology, investigation, and reviewing and editing the manuscript.
Clement L. Ren, MD, MBA, contributed to conceptualization/design, methodology, investigation, and reviewing and editing the manuscript.
Marci K. Sontag, PhD, contributed to conceptualization/design, methodology, investigation, and reviewing and editing the manuscript.
Karen Rychlik, MS, contributed to conceptualization/design, methodology, investigation, formal analysis, and reviewing and editing the manuscript.
Lauren Balmert, PhD, contributed to conceptualization/design, methodology, investigation, formal analysis, and reviewing and editing the manuscript.
Alexander Elbert, PhD, contributed to conceptualization/design, methodology, investigation, data curation, formal analysis, and reviewing and editing the manuscript.
Runyu Wu, MS, contributed to conceptualization/design, methodology, investigation, data curation, formal analysis, and reviewing and editing the manuscript.
Philip M. Farrell, MD, PhD, contributed to conceptualization/design, methodology, investigation, and reviewing and editing the manuscript.
All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
Definitions
Event: Sweat test, care episodes, or encounter.
Date of diagnosis: CFFPR reported date of diagnosis reported by CF Center.
Care episode: An event with start and end dates that are not the same (e.g., a hospitalization).
Encounter: An event that starts and end on the same date (e.g., an ambulatory visit).
Group 1: infants demographically categorized as being from Black/African American, American Indian/ Native Alaskan, Asian, and/or other race, and/or Hispanic ethnicity.
Group 2: infants demographically categorized as White and not Hispanic.
Article summary
Infants from historically marginalized groups, categorized using US Census race and ethnicity definitions, were older at first evaluation for cystic fibrosis.
What's known on the subject
Early CF diagnosis improves long-term health, but variability in timeliness of evaluation could reduce benefits of newborn screening (NBS). There may be bias in CF diagnosis based on race or ethnicity.
What this study adds
Infants with CF categorized Black/African American, White, American Indian/Alaska Native, Asian, Native Hawaiian/Pacific Islander, two or more races, none of these or unknown and/or of Hispanic ethnicity were older at initial CF Center evaluation than White, non-Hispanic infants.
Declaration of Competing Interest
The authors have no conflicts of interest relevant to this article to disclose. Dr. Albert and Mr. Wu are employees of the Cystic Fibrosis Foundation.
Acknowledgements
We gratefully acknowledge Marisol Vazquez, MPH, for project management and manuscript preparation assistance; and Jennifer Taylor-Cousar, MD, MSCS, for her review and suggestions on the manuscript.
Early diagnosis of cystic fibrosis through neonatal screening prevents severe malnutrition and improves long-term growth. Wisconsin Cystic Fibrosis Neonatal Screening Study Group.
CFTR mutation distribution among U.S. Hispanic and African American individuals: evaluation in cystic fibrosis patient and carrier screening populations.
Outcomes of infants born during the first 9 years of CF newborn screening in the United States: a retrospective Cystic Fibrosis Foundation Patient Registry cohort study.
Pulmonary findings in infants with cystic fibrosis during the first year of life: results from the Baby Observational and Nutrition Study (BONUS) cohort study.
00628-2&id=gr1.jpg){kind=link}
00628-2&id=gr2.jpg){kind=link}