- Email: [email protected]
Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study
Accepted Manuscript Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study Erina Lie, MD, Kevin J. Psoter, PhD, Katherine B. Püttgen, MD PII:
S0190-9622(18)32652-5
DOI:
10.1016/j.jaad.2018.09.041
Reference:
YMJD 12833
To appear in:
Journal of the American Academy of Dermatology
Received Date: 14 May 2018 Revised Date:
10 September 2018
Accepted Date: 24 September 2018
Please cite this article as: Lie E, Psoter KJ, Püttgen KB, Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study, Journal of the American Academy of Dermatology (2018), doi: https://doi.org/10.1016/j.jaad.2018.09.041. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Article Type: Original article Title: Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study
RI PT
Authors: Erina Lie, MDa; Kevin J. Psoter, PhDb; Katherine B. Püttgen, MDa
EP
TE D
Manuscript: 2500 words Capsule summary: 50 words Abstract: 200 words References: 35 Table count: 5 Figure count: 0
M AN U
Address correspondence to: Erina Lie, MD Johns Hopkins University School of Medicine Department of Dermatology 200 N. Wolfe St, Unit 2107 Baltimore, MD 21287 (E): [email protected] (P): 267-608-8222 | (F): 410-614-9308
SC
Affiliations: aJohns Hopkins School of Medicine, Department of Dermatology, Baltimore, MD; b Johns Hopkins School of Medicine, Department of Pediatrics, Baltimore, MD
IRB: Reviewed and approved by Johns Hopkins University IRB; approval #IRB00039855
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Funding Source: None
Financial Disclosure: Dr. Katherine Püttgen has served as a consultant to Pierre Fabre. The remaining authors have no financial relationships relevant to this article to disclose Conflicts of Interest: The authors have no conflicts of interest relevant to this article to disclose Keywords: pediatric dermatology; infantile hemangioma; access to care; socioeconomic status; system-based practice; cohort study
ACCEPTED MANUSCRIPT 2 38
ABSTRACT
39 Background: Early specialist evaluation during rapid proliferative growth of complicated
41
infantile hemangiomas (IH) is crucial. Health disparities and barriers of access-to-care for
42
children with IH have not been examined.
43
Objective: Investigate whether socioeconomic status (SES) is associated with age at
44
subspecialist presentation for IH evaluation.
45
Method: Retrospective cohort study of 804 children presenting to a large academic hospital.
46
Primary outcome was age at initial presentation. Covariates included demographic,
47
socioeconomic, geographic, and clinical characteristics. Medicaid or Children's Health Insurance
48
Program (CHIP) were proxies for lower SES. Analysis of covariance, χ2 tests, and generalized
49
ordered logistic regressions were performed.
50
Results: Children with lower SES had higher odds of presenting after 3 months of age (OR 2.11,
51
95% CI 1.31-3.38). In the subset that qualified for institutional care management program
52
(ICMP), no risk factors were associated with delayed presentation.
53
Limitations: Use of insurance and economic distress as proxies for SES; exclusion of uninsured
54
children may underestimate racioethnic effects; single academic center study limiting
55
generalizability.
56
Conclusions: Children with IH and lower SES were more likely to present later to specialists,
57
but those enrolled in an ICMP did not, suggesting that integrated ICMPs may mitigate disparities
58
and delayed access-to-care for IH among lower SES populations.
AC C
EP
TE D
M AN U
SC
RI PT
40
ACCEPTED MANUSCRIPT 3 59 60
INTRODUCTION Infantile hemangiomas (IH) are common vascular birthmarks with a unique life-cycle involving rapid proliferation in the first months of life and eventual involution by 3.5 and 4 years
62
of age in the vast majority.1-4,7 The most rapid growth occurs within 5.5-7.5 weeks of age and
63
over 80% of growth occurs by 3 months of age.3,4 More than 10% of children require specialist
64
evaluation to manage or prevent complications such as disfigurement, ulceration, visual, airway
65
or auditory compromise, feeding difficulties, hepatic dysfunction, infection, or bleeding.1 IH that
66
are facial, large, or segmental are associated with higher morbidity and risk of complications
67
requiring medical management.5,6 Moreover, approximately 50% may leave residual skin
68
changes, including telangiectasias, fibrofatty tissue, anetoderma, redundant skin, disfiguring
69
scars, or destruction of anatomic landmarks after involution.7 For infants at risk for
70
complications and disfiguring sequelae, early diagnosis by pediatricians and timely referral to
71
vascular anomalies specialists in early infancy is crucial since early therapeutic intervention may
72
prevent complications, improve outcomes, and positively impact the burden of disease and
73
quality of life.3,6,8-11 Disparities in healthcare access are well documented in dermatology and
74
pediatric primary care,12-18 but to our knowledge, health disparity and barriers to care for children
75
with IH have not been examined. Identifying factors associated with access to subspecialty
76
services may optimize the care, health outcomes, and quality of life of children with IH. This
77
study aimed to determine whether children of particular racioethnicity or socioeconomic status
78
(SES) present at an older age to subspecialty care for IH evaluation.
SC
M AN U
TE D
EP
AC C
79
RI PT
61
ACCEPTED MANUSCRIPT 4 80
METHODS
81
Study Population The study population included all children younger than 18 years with a state of
83
Maryland residential zip-code presenting to a single academic hospital specialty-care center for
84
initial or follow-up evaluation with a diagnosis of IH as classified according to the International
85
Classification of Diseases, ninth or tenth revisions,i during the 3-year period of July 1, 2013 to
86
June 30, 2016. Patients presenting initially to the following subspecialties were included:
87
dermatology, surgery, plastics, gastroenterology, ophthalmology, and otolaryngology. Children
88
with non-IH vascular anomalies, confirmed by chart review, and with unknown date of prior
89
presentation to outside specialty-care for IH evaluation were excluded. Accounting for the
90
introduction of widespread oral propranolol therapy, we excluded patients with initial
91
presentation prior to 2010. Demographic and clinical data were abstracted from electronic
92
medical records and photographs taken from each patient’s initial presentation for IH evaluation.
93 94
Outcome
TE D
M AN U
SC
RI PT
82
The primary outcome was age at initial presentation for subspecialty IH evaluation. Age
96
at presentation was classified into ordered categorical variables indicating longer time to initial
97
presentation for IH in the following manner: (1) early presentation, defined as first clinical
98
presentation prior to three months of age; (2) delayed presentation for presentation between three
99
and six months of age; or (3) late presentation as presentation after six months of age.
AC C
EP
95
100
i
Included ICD-9: 228.00, 228.01, 228.04, 228.09, 747.60, 757.32 or ICD-10: D18.00, D18.01, D18.03, D18.09, D18.1, D49.2, Q82.5, Q85.8, Q85.9, Q89.8, Q89.9, Q10.7, Q27.30, Q27.8, Q27.9, Q28.9, Q38.0, I78.1, I99.9, L98.8
ACCEPTED MANUSCRIPT 5 101 102
Covariates The following individual-level covariates were investigated as potential risk factors for later presentation: gender (male vs. female), race (African American vs. other), ethnicity
104
(Hispanic vs. non-Hispanic), and distance to the nearest institution-affiliated vascular anomalies
105
specialist. Distances were calculated using ArcGIS, Version 10.4 (ESRI, Redlands, CA, USA) as
106
the shortest Euclidean distance from the patient’s residential zip-code centroid to the nearest
107
such practice in the state of Maryland.
SC
108
RI PT
103
Medical insurance (primary payor) served as individual-level proxy for SES, where coverage through the Maryland medical assistance (MA) program or the state Children’s Health
110
Insurance Program (CHIP) program was defined as lower SES.19 Patients with commercial
111
insurance or Tricare, who did not receive state or federal MA, and who were self-pay were
112
classified as private insurance or higher SES. Additionally, we identified individuals in the Johns
113
Hopkins HealthCare LLC (JHHC), an institution-sponsored program for select payors to
114
facilitate healthcare and insurance management services. Partner payors encompassed managed
115
care organizations (MCO), government programs and employee health insurance, which
116
included both MA/CHIP and non-MA/CHIP payors. Distressed community index (DCI) score, a
117
measure of economic distress and prosperity based on seven indicators: education, housing,
118
unemployment, poverty rate, income, change in employment, and change in business
119
establishments,20 corresponding to each patient’s zip-code of residence, served as community-
120
level proxy for SES. Higher DCI scores correspond to higher levels of distress.
TE D
EP
AC C
121
M AN U
109
ACCEPTED MANUSCRIPT 6 122 123
Statistical analyses Demographic and clinical characteristics were compared in this retrospective cohort between early, delayed and late presenters using analysis of covariance and χ2 tests for
125
continuous and categorical covariates, respectively. We compared these characteristics between
126
children with higher and lower SES, and between children with non-JHHC and JHHC payor
127
programs using t-tests with unequal variances and χ2 tests.
Initially univariate generalized ordered logistic regression with a partial proportional odds
SC
128
RI PT
124
assumption21 was used to evaluate the association of a priori identified characteristics with the
130
categorical outcome, time at presentation (early, delayed, or late). These initial analyses also
131
considered clinical characteristics, including the hemangioma severity scale (HSS) score,ii
132
history of ulceration or bleeding, number of lesions, and lesion type (segmental vs. non-
133
segmental).5 These models fitted two separate logistic regressions, one at each cut point of the
134
outcome variable: early vs. delayed/late presenters and early/delayed vs. late presenters.
135
Multivariable generalized ordered logistic regressions were then constructed and included all
136
demographic and clinical factors. The proportional odds assumption (i.e. parallel lines
137
assumption) was evaluated using the variable specific and global Brant tests. For variables in
138
which the proportional odds assumption was violated, separate point estimates were presented.
139
Preplanned subgroup analyses were performed to evaluate whether risk factors differed
AC C
EP
TE D
M AN U
129
140
based upon payor type and JHHC membership. Due to limitations of small sample sizes in one or
141
more subcategories, subgroup analyses for race and ethnicity were not performed. Regression
142
model results are presented as odds ratio (OR) with corresponding 95% confidence intervals
143
(CI). A two-sided P<0.05 was considered statistically significant. All analyses were performed ii HSS is a validated scoring tool developed by Haggstrom AN, et al. that numerically quantifies the severity of IH lesions, accounting for size, location, pain, risk of complications, and disfigurement (see Reference #5 for original publication on this topic)
ACCEPTED MANUSCRIPT 7 144
using STATA Version 14.1 (StataCorp, College Station, TX, USA) and generalized ordered
145
logistic regressions were fitted using the gologit2 function for STATA.22
AC C
EP
TE D
M AN U
SC
RI PT
146
ACCEPTED MANUSCRIPT 8 147 148
RESULTS A total of 804 children met inclusion criteria and comprised the study population of which 219 (27.2%), 284 (35.3%), and 301 (37.4%) were classified as early, delayed and late
150
presenters, respectively. The mean age at presentation was 1.9 months (SD 0.7), 4.3 (SD 0.9) and
151
21.1 (SD 29.4), respectively.
RI PT
149
Table I describes the demographic and clinical characteristics of the study population,
153
overall and by time of IH presentation. Most were female (66.4%), Caucasian (72.9%), non-
154
Hispanic (82.1%), had higher-SES payors (77.0%) and had non-JHHC payors (83.6%). In
155
general, demographic characteristics did not differ significantly between presentation groups.
156
However, the early-presenting group had a greater proportion of children with higher-SES payor
157
(early 83.1% vs. delayed 73.2% vs. late 76.1%, p=0.030) while the delayed-presenting group had
158
a higher proportion of Hispanics (early 6.9% vs. delayed 11.6% vs. late 6.0%, p=0.048). Several
159
clinical characteristics were significantly different, including higher severity score (early 7.9 vs.
160
delayed 6.6 vs. late 5.2, p<0.001) and proportion of segmental IH (early 8.7% vs. delayed 4.2%
161
vs. late 2.3%, p=0.003). Table II details the demographic and clinical characteristics at initial
162
subspecialty presentation by primary payor (insurance provider) and by participation in JHHC.
M AN U
TE D
EP
163
SC
152
In unadjusted analyses, lower-SES and JHHC payor groups were associated with later presentation, whereas patients with higher severity score and segmental IH were more likely to
165
present earlier (Table III). After adjustment for clinical characteristics, lower-SES remained
166
significantly associated with early vs. delayed/late presentation (OR 2.11, 95% CI 1.31-3.38).
167
Clinically, increasing severity score was associated with lower odds of presenting after 3 and 6
168
months (OR 0.90, 95% CI 0.87-0.93) and having more than one IH lesion was associated with
169
lower odds of presenting after 6 months of age (OR 0.72, 95% CI 0.52-1.00).
AC C
164
ACCEPTED MANUSCRIPT 9 170
Multivariable analyses demonstrated that among children with non-JHHC payors, having lower-SES payors was associated with higher odds of presenting after 3 months (OR 2.37, 95%
172
CI 1.33-4.22). Increasing severity score and having more than one IH were associated with an
173
11% (95% CI 7-14%) lower odds of presenting after 3 and 6 months and 31% (95% CI 1-51%)
174
lower odds of presenting after 6 months of age (Table IV). In contrast, children with JHHC
175
payors had no demographic or clinical risk factors that were significantly associated with later
176
presentation.
SC
177
RI PT
171
Among children with higher-SES payors, no demographic risk factors were significantly associated with time-of-presentation in multivariable analyses; however, increasing severity
179
score (OR 0.90, 95% CI 0.87-0.94) was associated with later presentation after both 3 and 6
180
months, while having more than one IH (OR 0.66, 95% CI 0.46-0.96) was associated with lower
181
odds of presenting after 6 months of age (Table V). Higher DCI score was associated with lower
182
odds of presenting after 6 months (OR 0.85, 95% CI 0.75-0.97) in children with lower-SES
183
payors. Higher severity score (OR 0.85, 95% CI 0.78-0.93) and segmental IH (OR 0.06, 95%CI
184
0.01-0.34) were associated with higher odds of presenting after 6 and 3 months of age,
185
respectively.
TE D
EP AC C
186
M AN U
178
ACCEPTED MANUSCRIPT 10 187 188
DISCUSSION Several studies suggest that the first 4 weeks of life is the optimal age of referral to specialist care, yet there is often a delay in presenting high-risk IH to specialists with an average
190
initial evaluation frequently occurring at age 4 months or later.2-4,24 This discrepancy means
191
children often miss the optimal treatment window during the rapid proliferative phase when
192
therapy, if necessary, is most efficacious and associated with the lowest risk of long-term
193
sequelae.2-4,25 Reasons for this delay may be multifactorial and are not well-studied.
SC
194
RI PT
189
In our cohort, children with lower SES were more likely to present after 3 months of age relative to their higher-SES counterparts. This is consistent with other studies showing that
196
children covered by Medicaid-CHIP face greater barriers to specialty care than those with
197
commercial insurance.26-29 Medicaid-insured children generally come from households with
198
lower income, education, and health literacy that may contribute to this disparity. Additionally,
199
subspecialists tend to have limited acceptance of public insurance relative to primary care
200
providers due in part to lower reimbursement and longer time-to-payment for Medicaid
201
beneficiaries relative to private insurance.26-28 Chaudhry and colleagues reported a <20%
202
Medicaid-insured acceptance rate for new patients seeking atopic dermatitis evaluation with
203
dermatologists.30,31 Dermatology and otolaryngology are among the top three specialties with
204
limited access for Medicaid-CHIP patients relative to privately-insured patients.26,28
TE D
EP
AC C
205
M AN U
195
We found no difference in time to presentation based on gender, race, ethnicity, distance
206
to the nearest subspecialty clinic, membership in our institution’s managed-care program or DCI
207
score. Reassuringly, clinical characteristics including higher severity score and greater number of
208
IH were more likely to present earlier regardless of SES status.
ACCEPTED MANUSCRIPT 11 209
Children with Medicaid-CHIP who were enrolled in our institution-facilitated managedcare program (JHHC lower-SES payors) did not present later than their peers with JHHC higher-
211
SES payors suggesting that facilitating care-management may mitigate against the disparity in
212
timely access-to-care for IH evaluation in children of lower SES. One possible explanation is
213
that the JHHC program facilitates appointment scheduling to subspecialists within the institution.
214
Our dermatology clinics impose a maximum limit on time-to-appointment for all patients with
215
IH, further facilitating timely access to appointments. Many patients have pediatricians within
216
the institution who often have existing relationships with or work in close proximity to pediatric
217
subspecialists, which may also facilitate better communication and identification of referral
218
needs. It is worth nothing that many pediatric subspecialists are housed within academic medical
219
centers, which have been shown to attenuate disparities in access-to-care for children with
220
MA/CHIP.28,29 Establishing effective relationships with pediatricians is also of paramount
221
importance for facilitating early identification and referral to subspecialty care.
SC
M AN U
TE D
222
RI PT
210
In our cohort, lower-SES children living in more distressed residential areas were more likely to present before 6 months relative to their fellow lower-SES peers residing in less
224
distressed areas. Our institution is located in one of the most distressed neighborhoods in the
225
state serving a majority of very low SES population. It is likely that there was a higher
226
proportion of children with high DCI score residing in closer proximity to our facilities, thus
227
contributing to better access to referring pediatrician and subspecialty services. Further studies to
228
investigate this relationship and characterize the elements of DCI score that contribute to time-
229
of-presentation for IH evaluation would be beneficial. The importance of safety net providers
230
(SNPs), which encompass academic medical centers, community health centers, and providers of
231
specialized services, in improving access for poorly-insured patients is well-established.28,29,32
AC C
EP
223
ACCEPTED MANUSCRIPT 12 This finding suggests the need to support these institutions and defend funding for long-term
233
sustainability, especially with threats to the Medicaid and Medicare Disproportionate Share
234
Hospital allotments for SNPs, unclear sustainability of long-term CHIP funding, and the
235
continued discussion of possible adjustments to the Patient Protection and Affordable Care Act
236
of 2010.32,33
237
RI PT
232
Our findings highlight the importance of SNPs in attenuating inequalities in healthcare access, calling attention to an important way in which academic medical institutions can help
239
bridge the access gap and foster more equal healthcare access for children. We suggest a
240
potential role for institution-facilitated care-management programs (ICMPs) to mitigate
241
disparities in timely access-to-care for lower SES pediatric populations. Further investigation on
242
the impact of such services on cost and health outcome is warranted.
M AN U
243
SC
238
Examining follow-up care and patient outcome fall beyond the scope of this study, but is an important next step to further understand the consequence of health disparities. Further studies
245
comparing referral patterns among various conditions sensitive to early subspecialty evaluation
246
may potentially reveal ubiquitous factors contributing to pediatric health disparities.
247
249
Limitations
We used insurance status as an individual-level proxy and DCI scores as a community-
AC C
248
EP
TE D
244
250
level proxy for SES due to unavailable individual household income data. Uninsured patients
251
were excluded from this study due to unavailable data. Hispanics and non-Caucasians have
252
higher likelihoods of being uninsured and having lower income, and Hispanics are more likely to
253
have lower access to primary care,34,35 so exclusion of uninsured children may underestimate the
254
effect of racioethnic factors. Patients with pediatrician referral who did not schedule an
ACCEPTED MANUSCRIPT 13 appointment for subspecialty evaluation were not captured. Additionally, our study considered
256
access-to-care specific to the local healthcare market in Maryland at a single academic medical
257
center, which may limit generalizability. Likewise, access to specialty care may be dictated by
258
the number of specialists or centers in a particular region, thus limiting timely access regardless
259
of payor type. Heterogeneity of clinical practice and referral patterns by individual physicians
260
within the institution and region may also affect time-of-presentation. Finally, despite our efforts
261
to adjust for meaningful confounders, residual confounders could exist. We considered but
262
excluded additional variables including those for non-English speakers, need for interpreter
263
services, and presence of internal IH or PHACE syndrome, due to the lack of complete data.
M AN U
SC
RI PT
255
AC C
EP
TE D
264
ACCEPTED MANUSCRIPT 14 265 266
CONCLUSIONS Earlier evaluation of complicated IH by a specialist is associated with better clinical outcome and may prevent complications, disfigurement, and functional impairment.1-3,6,8-11,23 We
268
found that children with lower SES were more likely to present after 3 months of age than their
269
higher-SES counterparts, but that children enrolled in an ICMP did not present later. It is
270
reassuring that children in our cohort with severe and multiple IH were more likely to present
271
earlier regardless of SES status. Our findings suggest that integrated ICMPs may mitigate against
272
disparities in timely access-to-care for IH evaluation, particularly for children with lower SES.
SC
RI PT
267
M AN U
273 274 275 276
280 281 282 283 284 285 286 287
EP
279
AC C
278
TE D
277
ACCEPTED MANUSCRIPT 15 288
AUTHORSHIP DECLARATION
289 Dr. Katherine Püttgen conceptualized and designed the study, supervised data collection, and
291
critically reviewed and revised the manuscript. Dr. Erina Lie coordinated and performed data
292
collection, carried out the initial analyses, drafted the initial manuscript, and reviewed and
293
revised the manuscript. Prof. Kevin Psoter provided guidance and oversight on statistical
294
analyses, and reviewed and revised the manuscript. All authors approved the final manuscript as
295
submitted and agree to be accountable for all aspects of the work.
SC
RI PT
290
M AN U
296 297 298 299
303 304 305 306 307 308 309 310
EP
302
AC C
301
TE D
300
ACCEPTED MANUSCRIPT 16 311
ABBREVIATIONS
ACA – Affordable Care Act 2010
314
CHIP – Children’s Health Insurance Program
315
CI – confidence interval
316
DCI – distressed community index
317
FY – fiscal year
318
JHHC – Johns Hopkins HealthCare LLC
319
HSS – hemangioma severity scale
320
IH – infantile hemangioma
321
ICD – International Classification of Diseases
322
ICMP – institutional care management program
323
MA – medical assistance
324
MCO – managed care organizations
325
OR – odds ratio
326
SD – standard deviation
327
SES – socioeconomic status
328
SNP – safety net provider
330 331 332 333
M AN U
TE D
EP
AC C
329
SC
313
RI PT
312
ACCEPTED MANUSCRIPT 17 334
REFERENCES
335 336
1. Hoeger PH, Harper JI, Baselga E, et al. Treatment of infantile haemangiomas: recommendations of a European expert group. Eur J Peds. 2015 Jul;174(7):855-65.
338
2. Anderson KR, Schoch JJ, Lohse CM, Hand JL, Davis DM, Tollefson MM. Increasing
RI PT
337
incidence of infantile hemangiomas (IH) over the past 35 years: Correlation with decreasing
340
gestational age at birth and birth weight. J Am Acad Dermatol. 2016 Jan; 74(1):120-6.
342 343 344 345
3. Tollefson MM, Frieden IJ. Early growth of infantile hemangiomas: what parents’ photographs tell us. Pediatrics. 2012 Aug; 130(2):e314-20.
M AN U
341
SC
339
4. Chang L, Haggstrom AN, Drolet BA, et al. Growth characteristics of infantile hemangiomas: implications for management. Pediatrics. 2008 Aug;122(2):360-7. 5. Haggstrom AN, Beaumont JL, Lai JS, et al. Measuring the severity of infantile hemangiomas: instrument development and reliability. Arch Dermatol. 2012 Feb;
347
148(2):197-202.
348
TE D
346
6. Haggstrom AN, Drolet BA, Baselga E, et al. Prospective study of infantile hemangiomas: clinical characteristics predicting complications and treatment. Pediatrics. 2006 Sep;
350
118(3):882-7.
EP
349
7. Baselga E, Roe E, Coulie J, et al. Risk Factors for Degree and Type of Sequelae After
352
Involution of Untreated Hemangiomas of Infancy. JAMA Dermatol. 2016 Nov 1;
353
152(11):1239-1243.
354 355
AC C
351
8. Boccara O, Méni C, Léauté-Labreze C, et al. Haemangioma family burden: creation of a specific questionnaire. Acta Derm Venereol. 2015 Jan; 95(1):78-82.
ACCEPTED MANUSCRIPT 18 356
9. de Graaf M, Totte J, Breugem C, van Os-Medendorp H, Pasmans S. Evaluation of the Compliance, Acceptance, and Usability of a Web-Based eHealth Intervention for Parents of
358
Children with Infantile Hemangiomas: Usability Study. JMIR Res Protoc. 2013 Dec 17;
359
2(2):e54.
360
RI PT
357
10. Sagi L, Zvulunov A, Lapidoth M, Ben Amitai D. Efficacy and safety of propranolol for the treatment of infantile hemangioma: a presentation of ninety-nine cases. Dermatology. 2014;
362
228(2):136-44.
SC
361
11. Chamlin SL, Mancini AJ, Lai JS, et al. Development and Validation of a Quality-of-Life
364
Instrument for Infantile Hemangiomas. J Invest Dermatol. 2015 Jun; 135(6):1533-9.
368 369 370 371 372 373 374
13. Federico MJ, Liu AH. Overcoming childhood asthma disparities of the inner-city poor. Ped Clin North Am. 2003 Jun; 50(3):655-75.
TE D
367
117(5):426-34.
14. Flores G, Tomany-Korman SC. Racial and ethnic disparities in medical and dental health, access to care, and use of services in US children. Pediatrics. 2008 Feb; 121(2):e286-98. 15. Dougherty D, Chen X, Gray DT, Simon AE. Child and adolescent health care quality and
EP
366
12. Carter-Pokras O, Baquet C. What is a "health disparity"? Public Health Rep. 002 Sep-Oct;
disparities: are we making progress? Acad Pediatr. 2014 Mar-Apr; 14(2):137-48. 16. Findley K, Williams DR, Grice EA, Bonham VL. Health Disparities and the Microbiome.
AC C
365
M AN U
363
Trends Microbiol. 2016 Nov; 24(11):847-50.
375
17. Jamerson T, Sylvester R1, Jiang Q, et al. Differences in Cardiovascular Disease Risk Factors
376
and Health Behaviors Between Black and Non-Black Students Participating in a School-
377
Based Health Promotion Program. Am J Health Promot. 2017 Jul; 31(4):318-24.
ACCEPTED MANUSCRIPT 19 18. Smith DF, Boss EF. Racial/ethnic and socioeconomic disparities in the prevalence and
379
treatment of otitis media in children in the United States. Laryngoscope. 2010 Nov;
380
120(11):2306-12.
381
19. Maryland Health Connection. Medicaid Basics and Benefits.
RI PT
378
382
https://www.marylandhealthconnection.gov/shop-and-compare/medicaid-basics-and-
383
benefits/. Accessed: September 19, 2017.
385
20. Economic Innovation Group. "The Distressed Communities Index" Maryland Zipcode Level
SC
384
DCI Scores. http://eig.org/dci. Published: November 3, 2015. Accessed: June 26, 2017. 21. Hosmer Jr DW, Lemeshow S. Applied logistic regression. John Wiley & Sons. 2004.
387
22. Williams R. Generalized ordered logit/partial proportional odds models for ordinal dependent
388
M AN U
386
variables. Stata Journal. 2006 Mar 1; 6(1):58.
23. Hemangioma Investigator Group. Haggstrom AN, Drolet BA, Baselga E, et al. Prospective
390
study of infantile hemangiomas: demographic, prenatal, and perinatal characteristics. J
391
Pediatr. 2007 Mar; 150:291-4.
392
TE D
389
24. Wedgeworth E, Glover M, Irvine AD, et al. Propranolol in the treatment of infantile haemangiomas: lessons from the European Propranolol In the Treatment of Complicated
394
Haemangiomas (PITCH) Taskforce survey. Br J Dermatol. 2016 Mar;174(3):594-601.
395
25. de Graaf M, Knol MJ, Totté JE, et al. E-learning enables parents to assess an infantile
397 398 399 400
AC C
396
EP
393
hemangioma. J Am Acad Dermatol. 2014 May; 70(5):893-8. 26. Bisgaier J, Rhodes KV. Auditing access to specialty care for children with public insurance. N Engl J Med. 2011 Jun 16; 364(24):2324-33. 27. Prindaville B, Antaya RJ, Siegfried EC. Pediatric dermatology: past, present, and future. Pediatr Dermatol. 2015 Jan-Feb; 32(1):1-12.
ACCEPTED MANUSCRIPT 20
403 404 405 406 407
Access for Children. Arch Pediatr Adolesc Med. 2012 Apr; 166(4):304-10. 29. Skinner AC, Mayer ML. Effects of insurance status on children's access to specialty care: a systematic review of the literature. BMC Health Serv Res. 2007 Nov 28; 7:194.
RI PT
402
28. Bisgaier J, Polsky D, Rhodes KV. Academic Medical Centers and Equity in Specialty Care
30. Chaudhry SB, Armbrecht ES, Shin Y, et al. Pediatric access to dermatologists: Medicaid versus private insurance. J Am Acad Dermatol. 2013 May; 68(5):738-48.
31. U.S. Government Accountability Office. Medicaid and CHIP: most physicians serve covered
SC
401
children but have difficulty referring them for specialty care. Report to Congressional
409
Committees. http://www.aap.org/workforce/GAO_Report_Medicaid_CH_June2011.pdf.
410
Accessed: July 7, 2017.
413 414 415 416
United States. N Engl J Med. 2016 Nov 3; 375:1790-6.
33. Mead KH, Brantley E, Zur J, Goldberg DG. Collaboration Across the Safety Net in the Era
TE D
412
32. Chokshi DA, Chang JE, Wilson RM. Health Reform and the Changing Safety Net in the
of Health Reform: Opportunities or Obstacles? Med Care Res Rev. 2017 Jun; 74(3):286-310. 34. Office of Disease Prevention and Health Promotion. Healthy People 2020, 2020. https://www.healthypeople.gov/. Accessed: September 19, 2017.
EP
411
M AN U
408
35. Berdahl TA, Friedman BS, McCormick MC, Simpson L. Annual report on health care for
418
children and youth in the United States: trends in racial/ethnic, income, and insurance
419
disparities over time, 2002-2009. Acad Pediatr. 2013 May-Jun; 13(3):191-203.
420
AC C
417
ACCEPTED MANUSCRIPT 21 Table I. Characteristics of Children Evaluated for Infantile Hemangioma at Initial Subspecialty Presentation Age at IH Presentation Early (N = 219) Delayed (N = 284)
EP
a
Late (N = 301)
P value
21.1 (29.4)
n/a
4.3 (0.9)
148 (67.6) 71 (32.4)
183 (64.4) 101 (35.6)
172 (78.5) 14 (6.4) 33 (15.1)
199 (70.1) 24 (8.5) 61 (21.5)
15 (6.9) 186 (84.9) 18 (8.2) 32.3 (29.0)
33 (11.6) 218 (76.8) 33 (11.6) 35.4 (33.1)
18 (6.0) 256 (85.0) 27 (9.0) 33.2 (34.1)
182 (83.1) 37 (16.9)
208 (73.2) 76 (26.8)
229 (76.1) 72 (23.9)
192 (87.7) 27 (12.3) 28.8 (25.5)
236 (83.1) 48 (16.9) 29.5 (25.7)
244 (81.1) 57 (18.9) 27.9 (23.9)
7.5 (5.5)
6.6 (4.3)
5.2 (3.4)
<0.001
181 (82.6) 38 (17.4)
239 (84.2) 45 (15.8)
264 (87.7) 37 (12.3)
0.241
158 (72.2) 61 (27.8)
188 (66.2) 96 (33.8)
226 (75.1) 75 (24.9)
0.056
200 (91.3) 19 (8.7)
272 (95.8) 12 (4.2)
294 (97.7) 7 (2.3)
0.003
RI PT
1.9 (0.7)
TE D
203 (67.4) 98 (32.6)
0.679
215 (71.4) 25 (8.3) 61 (20.3)
0.283
SC
Characteristic Overall (N = 804) Demographics Age, mean (SD), month 9.9 (20.0) Gender Female 534 (66.4) Male 270 (33.6) Race Caucasian 586 (72.9) African American 63 (7.8) Otherb 155 (19.3) Ethnicity Hispanic or Latino 66 (8.2) Not Hispanic or Latino 660 (82.1) Unidentified 78 (9.7) Distance to clinic, mean 33.7 (32.4) (SD), km c Payor Higher SES 619 (77.0) Lower SES 185 (23.0) JHHC program No 672 (83.6) Yes 132 (16.4) d DCI Score, mean (SD) 28.7 (25.0) Clinical Characteristics Severity score, mean (SD) 6.4 (4.5) Ulceration or bleeding No 684 (85.1) Yes 120 (14.9) Number of IH lesion One 572 (71.1) More than one 232 (28.9) Type of IH lesion Non-segmental 766 (95.3) Segmental 38 (4.7)
M AN U
421 422
0.048 0.540
0.030
0.128 0.726
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; MA/CHIP = medical assistance/Children’s Health Insurance Program; SD = standard deviation a Statistically significant P-values based upon analysis of variance or Chi squared tests for continuous and categorical variables, respectively, in bold b Other includes Asian (n = 30), American Indian or Alaska Native (n = 1), Native Hawaiian or Pacific Islander (n = 1), two or more races (n = 28), and unidentified (n = 95) c Lower SES includes insurance through the Maryland MA/CHIP program; Higher SES includes commercial non-MA/CHIP and self-pay insurances d DCI score is an estimated measure of economic distress and prosperity for each zip-code based on seven indicators: education, housing, unemployment, poverty rate, income, change in employment, and change in business establishments. Higher scores correspond to higher levels of distress
423
ACCEPTED MANUSCRIPT 22 Table II. Characteristics at Initial Infantile Hemangioma Presentation by Type of Insurance Provider (Primary Payor) and Membership in the Johns Hopkins HealthCare LLC Payor Program, Respectively
9.8 (23.1)
0.941
9.7 (18.9)
11.3 (24.7)
0.387
182 (29.4) 208 (33.6) 229 (37.0)
37 (20.0) 76 (41.1) 72 (38.9)
0.030
192 (28.6) 236 (35.1) 244 (36.3)
27 (20.5) 48 (36.4) 57 (43.2)
0.128
415 (67.0) 204 (33.0)
119 (64.3) 66 (35.7)
0.492
511 (82.6) 24 (3.9) 84 (13.6)
75 (40.5) 39 (21.1) 71 (38.4)
<0.001
28 (4.5) 526 (85.0) 65 (10.5) 32.0 (26.3)
38 (20.5) 134 (72.5) 13 (7.0) 39.5 (47.1)
569 (91.9) 50 (8.1) 23.5 (21.3)
103 (55.7) 82 (44.3) 46.2 (28.3)
6.2 (4.3)
SC
10.0 (19.0)
90 (68.2) 42 (31.8)
0.639
515 (76.6) 46 (6.9) 111 (16.5)
71 (53.8) 17 (12.9) 44 (33.3)
<0.001
42 (6.3) 559 (83.2) 71 (10.6) 33.1 (29.8)
24 (18.2) 101 (76.5) 7 (5.3) 37.1 (43.4)
<0.001
569 (84.7) 103 (15.3) 26.6 (23.7)
50 (37.9) 82 (62.1) 39.5 (28.3)
7.1 (5.1)
0.026
6.4 (4.4)
6.7 (5.0)
0.390
538 (86.9) 81 (13.1)
146 (78.9) 39 (21.1)
0.007
578 (86.0) 94 (14.0)
106 (80.3) 26 (19.7)
0.092
433 (69.9) 186 (30.1)
139 (75.1) 46 (24.9)
0.172
470 (69.9) 202 (30.1)
102 (77.3) 30 (22.7)
0.089
592 (95.6) 27 (4.4)
174 (94.0) 11 (6.0)
0.373
639 (95.1) 33 (4.9)
127 (96.2) 5 (3.4)
0.822
M AN U
444 (66.1) 228 (33.9%)
EP
Demographics Age, mean (SD), month Age at presentation Early (<3 months) Delayed (3-6 months) Late (>6 months) Gender Female Male Race Caucasian African American c Other Ethnicity Hispanic or Latino Not Hispanic or Latino Unidentified Distance to clinic, mean (SD), km JHHC Program No Yes DCI Score, mean (SD)d Clinical Characteristics Severity score, mean (SD) Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
Primary Payor Lower-SES (N = 185)
Age at IH Presentation Membership in JHHC Payor Program b b P value Non-JHHC JHHC P value (N = 672) (N = 132)
RI PT
Higher-SES (N = 619)
a
TE D
Characteristic
AC C
424 425 426
<0.001 0.005
<0.001
<0.001 0.188
<0.001 <0.001
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; MA/CHIP = medical assistance/Children’s Health Insurance Program; SD = standard deviation a Lower SES includes insurance through the Maryland MA/CHIP program; Higher SES includes commercial non-MA/CHIP and self-pay insurances b Statistically significant P-values based upon t tests with unequal variances or Chi squared tests for continuous and categorical variables, respectively, in bold c Other includes Asian (n = 30), American Indian or Alaska Native (n = 1), Native Hawaiian or Pacific Islander (n = 1), two or more races (n = 28), and unidentified (n = 95) d DCI score is an estimated measure of economic distress and prosperity for each zip-code based on seven indicators: education, housing, unemployment, poverty rate, income, change in employment, and change in business establishments. Higher scores correspond to higher levels of distress
ACCEPTED MANUSCRIPT 23 Table III. Univariate and Multivariable Generalized Ordered Logistic Regression Analysis Evaluating the Association of Demographic and Clinical Characteristics with Age at First Infantile Hemangioma Presentation
REF 0.98 (0.74-1.30)
REF 1.19 (0.74-1.90)
REF 1.05 (0.62-1.77)
REF 1.30 (0.71-2.37) 1.00 (0.66-1.53) 1.00 (0.96-1.04) REF 1.67 (1.12-2.48) REF 1.41 (1.00-1.98) 0.99 (0.94-1.04) 0.90 (0.88-0.93)
0.60 (0.34-1.06)
1.09 (0.77-1.52)
REF 0.36 (0.19-0.66)
REF 0.81 (0.50-1.31) 1.02 (0.66-1.58) 1.00 (0.96-1.04)
REF 2.11 (1.31-3.38)
1.22 (0.81-1.85)
REF 1.37 (0.92-2.02) 0.95 (0.90-1.01)
0.90 (0.87-0.93)
REF 0.74 (0.52-1.0)
REF 1.07 (0.76-1.51)
SC
REF 1.00 (0.77-1.31)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) Payor Higher SES Lower SES JHHC program No Yes DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
M AN U
Predictors
Adjusted OR (95% CI) Early vs. Early/Delayed Delayed/Late vs. Late
RI PT
Unadjusted OR (95% CI) Early vs. Early/Delayed Delayed/Late vs. Late
TE D
427 428 429
0.73 (0.53-1.01)
REF 0.92 (0.63-1.34) REF 1.09 (0.76-1.55)
0.72 (0.52-1.00)
REF 0.59 (0.30-1.16)
430
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for payor status, and number of IH lesions in both unadjusted and adjusted regressions and Hispanic ethnicity in unadjusted analyses
ACCEPTED MANUSCRIPT 24 Table IV. Subgroup Analysis (Non- Johns Hopkins HealthCare LLC vs. Johns Hopkins HealthCare LLC Payors) – Multivariable Generalized Ordered Logistic Regression Analysis of Demographic Predictors for Age at First Infantile Hemangioma Presentation, Adjusted for Clinical Characteristic Confounders
REF 1.37 (0.67-2.84)
REF 1.11 (0.60-2.04)
REF 1.14 (0.38-3.44)
REF 2.37 (1.33-4.22) 0.94 (0.88-1.01) 0.89 (0.86-0.93) REF 1.01 (0.66-1.54)
REF 1.15 (0.47-2.80) 0.86 (0.19-3.92) 1.00 (0.92-1.08)
M AN U
REF 0.69 (0.38-1.25) 1.01 (0.64-1.61) 1.00 (0.96-1.05)
SC
REF 0.91 (0.67-1.23)
REF 1.05 (0.72-1.53) REF 0.68 (0.33-1.37)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) Payor Higher SES Lower SES DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
JHHC Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
RI PT
Non-JHHC Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
Predictors
1.15 (0.70-1.90)
REF 1.34 (0.61-2.95) 0.97 (0.85-1.10) 0.95 (0.87-1.04)
TE D
431 432 433 434
0.69 (0.49-0.99)
REF 0.68 (0.30-1.52) REF 1.00 (0.44-2.25) REF 0.12 (0.01-1.55)
435
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for payor status and number of IH lesions in regression analyses of NonJHHC payors; no proportional odds assumption was violated in regression analyses of JHHC payors
ACCEPTED MANUSCRIPT 25 Table V. Subgroup Analysis (Higher Socioeconomic Status vs. Lower Socioeconomic Status Payor) – Multivariable Generalized Ordered Logistic Regression Analysis of Demographic Predictors for Age at First Infantile Hemangioma Presentation, Adjusted for Clinical Characteristic Confounders
REF 0.77 (0.42-1.41)
REF 1.08 (0.50-2.37)
REF 1.13 (0.51-2.51)
REF 1.30 (0.75-2.26) 0.96 (0.89-1.03) 0.90 (0.87-0.94) REF 0.83 (0.53-1.29)
REF 0.76 (0.36-1.61) 1.16 (0.35-3.78) 1.03 (0.97-1.10)
M AN U
REF 0.77 (0.39-1.52) 1.02 (0.63-1.64) 0.98 (0.93-1.04)
SC
REF 1.04 (0.76-1.43)
REF 1.05 (0.71-1.56) REF 0.66 (0.36-1.63)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) JHHC Program No Yes DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
Lower-SES Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
RI PT
Higher-SES Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
Predictors
TE D
436 437 438 439
0.66 (0.46-0.96)
REF 1.43 (0.78-2.61) 1.06 (0.91-1.24)
0.85 (0.75-0.97)
0.96 (0.89-1.05)
0.85 (0.78-0.93)
REF 1.39 (0.66-2.94) REF 1.09 (0.5-2.16)
REF 0.06 (0.01-0.34)
0.91 (0.15-5.36)
440
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for number of IH lesions in regression analyses of higher SES children; proportional odds assumption was violated for DCI score, severity score, and type of IH lesion for lower SES children
ACCEPTED MANUSCRIPT
CAPSULE SUMMARY
•
What is already known on this topic: Children with Medicaid face greater barriers of
RI PT
access-to-care. Delayed management of complicated infantile hemangiomas compromises outcomes. •
What this article adds to our knowledge. Lower socioeconomic status is associated with
SC
delayed care for infantile hemangioma, but not among children with institutional caremanagement services.
M AN U
How this information impacts clinical practice and/or changes patient care. Institutionfacilitated managed-care programs may moderate socioeconomic disparities in accessing
EP
TE D
timely specialty care and improve outcomes.
AC C
•
S0190-9622(18)32652-5
DOI:
10.1016/j.jaad.2018.09.041
Reference:
YMJD 12833
To appear in:
Journal of the American Academy of Dermatology
Received Date: 14 May 2018 Revised Date:
10 September 2018
Accepted Date: 24 September 2018
Please cite this article as: Lie E, Psoter KJ, Püttgen KB, Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study, Journal of the American Academy of Dermatology (2018), doi: https://doi.org/10.1016/j.jaad.2018.09.041. This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
ACCEPTED MANUSCRIPT 1 Article Type: Original article Title: Lower Socioeconomic Status is Associated with Delayed Access to Care for Infantile Hemangioma, a Cohort Study
RI PT
Authors: Erina Lie, MDa; Kevin J. Psoter, PhDb; Katherine B. Püttgen, MDa
EP
TE D
Manuscript: 2500 words Capsule summary: 50 words Abstract: 200 words References: 35 Table count: 5 Figure count: 0
M AN U
Address correspondence to: Erina Lie, MD Johns Hopkins University School of Medicine Department of Dermatology 200 N. Wolfe St, Unit 2107 Baltimore, MD 21287 (E): [email protected] (P): 267-608-8222 | (F): 410-614-9308
SC
Affiliations: aJohns Hopkins School of Medicine, Department of Dermatology, Baltimore, MD; b Johns Hopkins School of Medicine, Department of Pediatrics, Baltimore, MD
IRB: Reviewed and approved by Johns Hopkins University IRB; approval #IRB00039855
AC C
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37
Funding Source: None
Financial Disclosure: Dr. Katherine Püttgen has served as a consultant to Pierre Fabre. The remaining authors have no financial relationships relevant to this article to disclose Conflicts of Interest: The authors have no conflicts of interest relevant to this article to disclose Keywords: pediatric dermatology; infantile hemangioma; access to care; socioeconomic status; system-based practice; cohort study
ACCEPTED MANUSCRIPT 2 38
ABSTRACT
39 Background: Early specialist evaluation during rapid proliferative growth of complicated
41
infantile hemangiomas (IH) is crucial. Health disparities and barriers of access-to-care for
42
children with IH have not been examined.
43
Objective: Investigate whether socioeconomic status (SES) is associated with age at
44
subspecialist presentation for IH evaluation.
45
Method: Retrospective cohort study of 804 children presenting to a large academic hospital.
46
Primary outcome was age at initial presentation. Covariates included demographic,
47
socioeconomic, geographic, and clinical characteristics. Medicaid or Children's Health Insurance
48
Program (CHIP) were proxies for lower SES. Analysis of covariance, χ2 tests, and generalized
49
ordered logistic regressions were performed.
50
Results: Children with lower SES had higher odds of presenting after 3 months of age (OR 2.11,
51
95% CI 1.31-3.38). In the subset that qualified for institutional care management program
52
(ICMP), no risk factors were associated with delayed presentation.
53
Limitations: Use of insurance and economic distress as proxies for SES; exclusion of uninsured
54
children may underestimate racioethnic effects; single academic center study limiting
55
generalizability.
56
Conclusions: Children with IH and lower SES were more likely to present later to specialists,
57
but those enrolled in an ICMP did not, suggesting that integrated ICMPs may mitigate disparities
58
and delayed access-to-care for IH among lower SES populations.
AC C
EP
TE D
M AN U
SC
RI PT
40
ACCEPTED MANUSCRIPT 3 59 60
INTRODUCTION Infantile hemangiomas (IH) are common vascular birthmarks with a unique life-cycle involving rapid proliferation in the first months of life and eventual involution by 3.5 and 4 years
62
of age in the vast majority.1-4,7 The most rapid growth occurs within 5.5-7.5 weeks of age and
63
over 80% of growth occurs by 3 months of age.3,4 More than 10% of children require specialist
64
evaluation to manage or prevent complications such as disfigurement, ulceration, visual, airway
65
or auditory compromise, feeding difficulties, hepatic dysfunction, infection, or bleeding.1 IH that
66
are facial, large, or segmental are associated with higher morbidity and risk of complications
67
requiring medical management.5,6 Moreover, approximately 50% may leave residual skin
68
changes, including telangiectasias, fibrofatty tissue, anetoderma, redundant skin, disfiguring
69
scars, or destruction of anatomic landmarks after involution.7 For infants at risk for
70
complications and disfiguring sequelae, early diagnosis by pediatricians and timely referral to
71
vascular anomalies specialists in early infancy is crucial since early therapeutic intervention may
72
prevent complications, improve outcomes, and positively impact the burden of disease and
73
quality of life.3,6,8-11 Disparities in healthcare access are well documented in dermatology and
74
pediatric primary care,12-18 but to our knowledge, health disparity and barriers to care for children
75
with IH have not been examined. Identifying factors associated with access to subspecialty
76
services may optimize the care, health outcomes, and quality of life of children with IH. This
77
study aimed to determine whether children of particular racioethnicity or socioeconomic status
78
(SES) present at an older age to subspecialty care for IH evaluation.
SC
M AN U
TE D
EP
AC C
79
RI PT
61
ACCEPTED MANUSCRIPT 4 80
METHODS
81
Study Population The study population included all children younger than 18 years with a state of
83
Maryland residential zip-code presenting to a single academic hospital specialty-care center for
84
initial or follow-up evaluation with a diagnosis of IH as classified according to the International
85
Classification of Diseases, ninth or tenth revisions,i during the 3-year period of July 1, 2013 to
86
June 30, 2016. Patients presenting initially to the following subspecialties were included:
87
dermatology, surgery, plastics, gastroenterology, ophthalmology, and otolaryngology. Children
88
with non-IH vascular anomalies, confirmed by chart review, and with unknown date of prior
89
presentation to outside specialty-care for IH evaluation were excluded. Accounting for the
90
introduction of widespread oral propranolol therapy, we excluded patients with initial
91
presentation prior to 2010. Demographic and clinical data were abstracted from electronic
92
medical records and photographs taken from each patient’s initial presentation for IH evaluation.
93 94
Outcome
TE D
M AN U
SC
RI PT
82
The primary outcome was age at initial presentation for subspecialty IH evaluation. Age
96
at presentation was classified into ordered categorical variables indicating longer time to initial
97
presentation for IH in the following manner: (1) early presentation, defined as first clinical
98
presentation prior to three months of age; (2) delayed presentation for presentation between three
99
and six months of age; or (3) late presentation as presentation after six months of age.
AC C
EP
95
100
i
Included ICD-9: 228.00, 228.01, 228.04, 228.09, 747.60, 757.32 or ICD-10: D18.00, D18.01, D18.03, D18.09, D18.1, D49.2, Q82.5, Q85.8, Q85.9, Q89.8, Q89.9, Q10.7, Q27.30, Q27.8, Q27.9, Q28.9, Q38.0, I78.1, I99.9, L98.8
ACCEPTED MANUSCRIPT 5 101 102
Covariates The following individual-level covariates were investigated as potential risk factors for later presentation: gender (male vs. female), race (African American vs. other), ethnicity
104
(Hispanic vs. non-Hispanic), and distance to the nearest institution-affiliated vascular anomalies
105
specialist. Distances were calculated using ArcGIS, Version 10.4 (ESRI, Redlands, CA, USA) as
106
the shortest Euclidean distance from the patient’s residential zip-code centroid to the nearest
107
such practice in the state of Maryland.
SC
108
RI PT
103
Medical insurance (primary payor) served as individual-level proxy for SES, where coverage through the Maryland medical assistance (MA) program or the state Children’s Health
110
Insurance Program (CHIP) program was defined as lower SES.19 Patients with commercial
111
insurance or Tricare, who did not receive state or federal MA, and who were self-pay were
112
classified as private insurance or higher SES. Additionally, we identified individuals in the Johns
113
Hopkins HealthCare LLC (JHHC), an institution-sponsored program for select payors to
114
facilitate healthcare and insurance management services. Partner payors encompassed managed
115
care organizations (MCO), government programs and employee health insurance, which
116
included both MA/CHIP and non-MA/CHIP payors. Distressed community index (DCI) score, a
117
measure of economic distress and prosperity based on seven indicators: education, housing,
118
unemployment, poverty rate, income, change in employment, and change in business
119
establishments,20 corresponding to each patient’s zip-code of residence, served as community-
120
level proxy for SES. Higher DCI scores correspond to higher levels of distress.
TE D
EP
AC C
121
M AN U
109
ACCEPTED MANUSCRIPT 6 122 123
Statistical analyses Demographic and clinical characteristics were compared in this retrospective cohort between early, delayed and late presenters using analysis of covariance and χ2 tests for
125
continuous and categorical covariates, respectively. We compared these characteristics between
126
children with higher and lower SES, and between children with non-JHHC and JHHC payor
127
programs using t-tests with unequal variances and χ2 tests.
Initially univariate generalized ordered logistic regression with a partial proportional odds
SC
128
RI PT
124
assumption21 was used to evaluate the association of a priori identified characteristics with the
130
categorical outcome, time at presentation (early, delayed, or late). These initial analyses also
131
considered clinical characteristics, including the hemangioma severity scale (HSS) score,ii
132
history of ulceration or bleeding, number of lesions, and lesion type (segmental vs. non-
133
segmental).5 These models fitted two separate logistic regressions, one at each cut point of the
134
outcome variable: early vs. delayed/late presenters and early/delayed vs. late presenters.
135
Multivariable generalized ordered logistic regressions were then constructed and included all
136
demographic and clinical factors. The proportional odds assumption (i.e. parallel lines
137
assumption) was evaluated using the variable specific and global Brant tests. For variables in
138
which the proportional odds assumption was violated, separate point estimates were presented.
139
Preplanned subgroup analyses were performed to evaluate whether risk factors differed
AC C
EP
TE D
M AN U
129
140
based upon payor type and JHHC membership. Due to limitations of small sample sizes in one or
141
more subcategories, subgroup analyses for race and ethnicity were not performed. Regression
142
model results are presented as odds ratio (OR) with corresponding 95% confidence intervals
143
(CI). A two-sided P<0.05 was considered statistically significant. All analyses were performed ii HSS is a validated scoring tool developed by Haggstrom AN, et al. that numerically quantifies the severity of IH lesions, accounting for size, location, pain, risk of complications, and disfigurement (see Reference #5 for original publication on this topic)
ACCEPTED MANUSCRIPT 7 144
using STATA Version 14.1 (StataCorp, College Station, TX, USA) and generalized ordered
145
logistic regressions were fitted using the gologit2 function for STATA.22
AC C
EP
TE D
M AN U
SC
RI PT
146
ACCEPTED MANUSCRIPT 8 147 148
RESULTS A total of 804 children met inclusion criteria and comprised the study population of which 219 (27.2%), 284 (35.3%), and 301 (37.4%) were classified as early, delayed and late
150
presenters, respectively. The mean age at presentation was 1.9 months (SD 0.7), 4.3 (SD 0.9) and
151
21.1 (SD 29.4), respectively.
RI PT
149
Table I describes the demographic and clinical characteristics of the study population,
153
overall and by time of IH presentation. Most were female (66.4%), Caucasian (72.9%), non-
154
Hispanic (82.1%), had higher-SES payors (77.0%) and had non-JHHC payors (83.6%). In
155
general, demographic characteristics did not differ significantly between presentation groups.
156
However, the early-presenting group had a greater proportion of children with higher-SES payor
157
(early 83.1% vs. delayed 73.2% vs. late 76.1%, p=0.030) while the delayed-presenting group had
158
a higher proportion of Hispanics (early 6.9% vs. delayed 11.6% vs. late 6.0%, p=0.048). Several
159
clinical characteristics were significantly different, including higher severity score (early 7.9 vs.
160
delayed 6.6 vs. late 5.2, p<0.001) and proportion of segmental IH (early 8.7% vs. delayed 4.2%
161
vs. late 2.3%, p=0.003). Table II details the demographic and clinical characteristics at initial
162
subspecialty presentation by primary payor (insurance provider) and by participation in JHHC.
M AN U
TE D
EP
163
SC
152
In unadjusted analyses, lower-SES and JHHC payor groups were associated with later presentation, whereas patients with higher severity score and segmental IH were more likely to
165
present earlier (Table III). After adjustment for clinical characteristics, lower-SES remained
166
significantly associated with early vs. delayed/late presentation (OR 2.11, 95% CI 1.31-3.38).
167
Clinically, increasing severity score was associated with lower odds of presenting after 3 and 6
168
months (OR 0.90, 95% CI 0.87-0.93) and having more than one IH lesion was associated with
169
lower odds of presenting after 6 months of age (OR 0.72, 95% CI 0.52-1.00).
AC C
164
ACCEPTED MANUSCRIPT 9 170
Multivariable analyses demonstrated that among children with non-JHHC payors, having lower-SES payors was associated with higher odds of presenting after 3 months (OR 2.37, 95%
172
CI 1.33-4.22). Increasing severity score and having more than one IH were associated with an
173
11% (95% CI 7-14%) lower odds of presenting after 3 and 6 months and 31% (95% CI 1-51%)
174
lower odds of presenting after 6 months of age (Table IV). In contrast, children with JHHC
175
payors had no demographic or clinical risk factors that were significantly associated with later
176
presentation.
SC
177
RI PT
171
Among children with higher-SES payors, no demographic risk factors were significantly associated with time-of-presentation in multivariable analyses; however, increasing severity
179
score (OR 0.90, 95% CI 0.87-0.94) was associated with later presentation after both 3 and 6
180
months, while having more than one IH (OR 0.66, 95% CI 0.46-0.96) was associated with lower
181
odds of presenting after 6 months of age (Table V). Higher DCI score was associated with lower
182
odds of presenting after 6 months (OR 0.85, 95% CI 0.75-0.97) in children with lower-SES
183
payors. Higher severity score (OR 0.85, 95% CI 0.78-0.93) and segmental IH (OR 0.06, 95%CI
184
0.01-0.34) were associated with higher odds of presenting after 6 and 3 months of age,
185
respectively.
TE D
EP AC C
186
M AN U
178
ACCEPTED MANUSCRIPT 10 187 188
DISCUSSION Several studies suggest that the first 4 weeks of life is the optimal age of referral to specialist care, yet there is often a delay in presenting high-risk IH to specialists with an average
190
initial evaluation frequently occurring at age 4 months or later.2-4,24 This discrepancy means
191
children often miss the optimal treatment window during the rapid proliferative phase when
192
therapy, if necessary, is most efficacious and associated with the lowest risk of long-term
193
sequelae.2-4,25 Reasons for this delay may be multifactorial and are not well-studied.
SC
194
RI PT
189
In our cohort, children with lower SES were more likely to present after 3 months of age relative to their higher-SES counterparts. This is consistent with other studies showing that
196
children covered by Medicaid-CHIP face greater barriers to specialty care than those with
197
commercial insurance.26-29 Medicaid-insured children generally come from households with
198
lower income, education, and health literacy that may contribute to this disparity. Additionally,
199
subspecialists tend to have limited acceptance of public insurance relative to primary care
200
providers due in part to lower reimbursement and longer time-to-payment for Medicaid
201
beneficiaries relative to private insurance.26-28 Chaudhry and colleagues reported a <20%
202
Medicaid-insured acceptance rate for new patients seeking atopic dermatitis evaluation with
203
dermatologists.30,31 Dermatology and otolaryngology are among the top three specialties with
204
limited access for Medicaid-CHIP patients relative to privately-insured patients.26,28
TE D
EP
AC C
205
M AN U
195
We found no difference in time to presentation based on gender, race, ethnicity, distance
206
to the nearest subspecialty clinic, membership in our institution’s managed-care program or DCI
207
score. Reassuringly, clinical characteristics including higher severity score and greater number of
208
IH were more likely to present earlier regardless of SES status.
ACCEPTED MANUSCRIPT 11 209
Children with Medicaid-CHIP who were enrolled in our institution-facilitated managedcare program (JHHC lower-SES payors) did not present later than their peers with JHHC higher-
211
SES payors suggesting that facilitating care-management may mitigate against the disparity in
212
timely access-to-care for IH evaluation in children of lower SES. One possible explanation is
213
that the JHHC program facilitates appointment scheduling to subspecialists within the institution.
214
Our dermatology clinics impose a maximum limit on time-to-appointment for all patients with
215
IH, further facilitating timely access to appointments. Many patients have pediatricians within
216
the institution who often have existing relationships with or work in close proximity to pediatric
217
subspecialists, which may also facilitate better communication and identification of referral
218
needs. It is worth nothing that many pediatric subspecialists are housed within academic medical
219
centers, which have been shown to attenuate disparities in access-to-care for children with
220
MA/CHIP.28,29 Establishing effective relationships with pediatricians is also of paramount
221
importance for facilitating early identification and referral to subspecialty care.
SC
M AN U
TE D
222
RI PT
210
In our cohort, lower-SES children living in more distressed residential areas were more likely to present before 6 months relative to their fellow lower-SES peers residing in less
224
distressed areas. Our institution is located in one of the most distressed neighborhoods in the
225
state serving a majority of very low SES population. It is likely that there was a higher
226
proportion of children with high DCI score residing in closer proximity to our facilities, thus
227
contributing to better access to referring pediatrician and subspecialty services. Further studies to
228
investigate this relationship and characterize the elements of DCI score that contribute to time-
229
of-presentation for IH evaluation would be beneficial. The importance of safety net providers
230
(SNPs), which encompass academic medical centers, community health centers, and providers of
231
specialized services, in improving access for poorly-insured patients is well-established.28,29,32
AC C
EP
223
ACCEPTED MANUSCRIPT 12 This finding suggests the need to support these institutions and defend funding for long-term
233
sustainability, especially with threats to the Medicaid and Medicare Disproportionate Share
234
Hospital allotments for SNPs, unclear sustainability of long-term CHIP funding, and the
235
continued discussion of possible adjustments to the Patient Protection and Affordable Care Act
236
of 2010.32,33
237
RI PT
232
Our findings highlight the importance of SNPs in attenuating inequalities in healthcare access, calling attention to an important way in which academic medical institutions can help
239
bridge the access gap and foster more equal healthcare access for children. We suggest a
240
potential role for institution-facilitated care-management programs (ICMPs) to mitigate
241
disparities in timely access-to-care for lower SES pediatric populations. Further investigation on
242
the impact of such services on cost and health outcome is warranted.
M AN U
243
SC
238
Examining follow-up care and patient outcome fall beyond the scope of this study, but is an important next step to further understand the consequence of health disparities. Further studies
245
comparing referral patterns among various conditions sensitive to early subspecialty evaluation
246
may potentially reveal ubiquitous factors contributing to pediatric health disparities.
247
249
Limitations
We used insurance status as an individual-level proxy and DCI scores as a community-
AC C
248
EP
TE D
244
250
level proxy for SES due to unavailable individual household income data. Uninsured patients
251
were excluded from this study due to unavailable data. Hispanics and non-Caucasians have
252
higher likelihoods of being uninsured and having lower income, and Hispanics are more likely to
253
have lower access to primary care,34,35 so exclusion of uninsured children may underestimate the
254
effect of racioethnic factors. Patients with pediatrician referral who did not schedule an
ACCEPTED MANUSCRIPT 13 appointment for subspecialty evaluation were not captured. Additionally, our study considered
256
access-to-care specific to the local healthcare market in Maryland at a single academic medical
257
center, which may limit generalizability. Likewise, access to specialty care may be dictated by
258
the number of specialists or centers in a particular region, thus limiting timely access regardless
259
of payor type. Heterogeneity of clinical practice and referral patterns by individual physicians
260
within the institution and region may also affect time-of-presentation. Finally, despite our efforts
261
to adjust for meaningful confounders, residual confounders could exist. We considered but
262
excluded additional variables including those for non-English speakers, need for interpreter
263
services, and presence of internal IH or PHACE syndrome, due to the lack of complete data.
M AN U
SC
RI PT
255
AC C
EP
TE D
264
ACCEPTED MANUSCRIPT 14 265 266
CONCLUSIONS Earlier evaluation of complicated IH by a specialist is associated with better clinical outcome and may prevent complications, disfigurement, and functional impairment.1-3,6,8-11,23 We
268
found that children with lower SES were more likely to present after 3 months of age than their
269
higher-SES counterparts, but that children enrolled in an ICMP did not present later. It is
270
reassuring that children in our cohort with severe and multiple IH were more likely to present
271
earlier regardless of SES status. Our findings suggest that integrated ICMPs may mitigate against
272
disparities in timely access-to-care for IH evaluation, particularly for children with lower SES.
SC
RI PT
267
M AN U
273 274 275 276
280 281 282 283 284 285 286 287
EP
279
AC C
278
TE D
277
ACCEPTED MANUSCRIPT 15 288
AUTHORSHIP DECLARATION
289 Dr. Katherine Püttgen conceptualized and designed the study, supervised data collection, and
291
critically reviewed and revised the manuscript. Dr. Erina Lie coordinated and performed data
292
collection, carried out the initial analyses, drafted the initial manuscript, and reviewed and
293
revised the manuscript. Prof. Kevin Psoter provided guidance and oversight on statistical
294
analyses, and reviewed and revised the manuscript. All authors approved the final manuscript as
295
submitted and agree to be accountable for all aspects of the work.
SC
RI PT
290
M AN U
296 297 298 299
303 304 305 306 307 308 309 310
EP
302
AC C
301
TE D
300
ACCEPTED MANUSCRIPT 16 311
ABBREVIATIONS
ACA – Affordable Care Act 2010
314
CHIP – Children’s Health Insurance Program
315
CI – confidence interval
316
DCI – distressed community index
317
FY – fiscal year
318
JHHC – Johns Hopkins HealthCare LLC
319
HSS – hemangioma severity scale
320
IH – infantile hemangioma
321
ICD – International Classification of Diseases
322
ICMP – institutional care management program
323
MA – medical assistance
324
MCO – managed care organizations
325
OR – odds ratio
326
SD – standard deviation
327
SES – socioeconomic status
328
SNP – safety net provider
330 331 332 333
M AN U
TE D
EP
AC C
329
SC
313
RI PT
312
ACCEPTED MANUSCRIPT 17 334
REFERENCES
335 336
1. Hoeger PH, Harper JI, Baselga E, et al. Treatment of infantile haemangiomas: recommendations of a European expert group. Eur J Peds. 2015 Jul;174(7):855-65.
338
2. Anderson KR, Schoch JJ, Lohse CM, Hand JL, Davis DM, Tollefson MM. Increasing
RI PT
337
incidence of infantile hemangiomas (IH) over the past 35 years: Correlation with decreasing
340
gestational age at birth and birth weight. J Am Acad Dermatol. 2016 Jan; 74(1):120-6.
342 343 344 345
3. Tollefson MM, Frieden IJ. Early growth of infantile hemangiomas: what parents’ photographs tell us. Pediatrics. 2012 Aug; 130(2):e314-20.
M AN U
341
SC
339
4. Chang L, Haggstrom AN, Drolet BA, et al. Growth characteristics of infantile hemangiomas: implications for management. Pediatrics. 2008 Aug;122(2):360-7. 5. Haggstrom AN, Beaumont JL, Lai JS, et al. Measuring the severity of infantile hemangiomas: instrument development and reliability. Arch Dermatol. 2012 Feb;
347
148(2):197-202.
348
TE D
346
6. Haggstrom AN, Drolet BA, Baselga E, et al. Prospective study of infantile hemangiomas: clinical characteristics predicting complications and treatment. Pediatrics. 2006 Sep;
350
118(3):882-7.
EP
349
7. Baselga E, Roe E, Coulie J, et al. Risk Factors for Degree and Type of Sequelae After
352
Involution of Untreated Hemangiomas of Infancy. JAMA Dermatol. 2016 Nov 1;
353
152(11):1239-1243.
354 355
AC C
351
8. Boccara O, Méni C, Léauté-Labreze C, et al. Haemangioma family burden: creation of a specific questionnaire. Acta Derm Venereol. 2015 Jan; 95(1):78-82.
ACCEPTED MANUSCRIPT 18 356
9. de Graaf M, Totte J, Breugem C, van Os-Medendorp H, Pasmans S. Evaluation of the Compliance, Acceptance, and Usability of a Web-Based eHealth Intervention for Parents of
358
Children with Infantile Hemangiomas: Usability Study. JMIR Res Protoc. 2013 Dec 17;
359
2(2):e54.
360
RI PT
357
10. Sagi L, Zvulunov A, Lapidoth M, Ben Amitai D. Efficacy and safety of propranolol for the treatment of infantile hemangioma: a presentation of ninety-nine cases. Dermatology. 2014;
362
228(2):136-44.
SC
361
11. Chamlin SL, Mancini AJ, Lai JS, et al. Development and Validation of a Quality-of-Life
364
Instrument for Infantile Hemangiomas. J Invest Dermatol. 2015 Jun; 135(6):1533-9.
368 369 370 371 372 373 374
13. Federico MJ, Liu AH. Overcoming childhood asthma disparities of the inner-city poor. Ped Clin North Am. 2003 Jun; 50(3):655-75.
TE D
367
117(5):426-34.
14. Flores G, Tomany-Korman SC. Racial and ethnic disparities in medical and dental health, access to care, and use of services in US children. Pediatrics. 2008 Feb; 121(2):e286-98. 15. Dougherty D, Chen X, Gray DT, Simon AE. Child and adolescent health care quality and
EP
366
12. Carter-Pokras O, Baquet C. What is a "health disparity"? Public Health Rep. 002 Sep-Oct;
disparities: are we making progress? Acad Pediatr. 2014 Mar-Apr; 14(2):137-48. 16. Findley K, Williams DR, Grice EA, Bonham VL. Health Disparities and the Microbiome.
AC C
365
M AN U
363
Trends Microbiol. 2016 Nov; 24(11):847-50.
375
17. Jamerson T, Sylvester R1, Jiang Q, et al. Differences in Cardiovascular Disease Risk Factors
376
and Health Behaviors Between Black and Non-Black Students Participating in a School-
377
Based Health Promotion Program. Am J Health Promot. 2017 Jul; 31(4):318-24.
ACCEPTED MANUSCRIPT 19 18. Smith DF, Boss EF. Racial/ethnic and socioeconomic disparities in the prevalence and
379
treatment of otitis media in children in the United States. Laryngoscope. 2010 Nov;
380
120(11):2306-12.
381
19. Maryland Health Connection. Medicaid Basics and Benefits.
RI PT
378
382
https://www.marylandhealthconnection.gov/shop-and-compare/medicaid-basics-and-
383
benefits/. Accessed: September 19, 2017.
385
20. Economic Innovation Group. "The Distressed Communities Index" Maryland Zipcode Level
SC
384
DCI Scores. http://eig.org/dci. Published: November 3, 2015. Accessed: June 26, 2017. 21. Hosmer Jr DW, Lemeshow S. Applied logistic regression. John Wiley & Sons. 2004.
387
22. Williams R. Generalized ordered logit/partial proportional odds models for ordinal dependent
388
M AN U
386
variables. Stata Journal. 2006 Mar 1; 6(1):58.
23. Hemangioma Investigator Group. Haggstrom AN, Drolet BA, Baselga E, et al. Prospective
390
study of infantile hemangiomas: demographic, prenatal, and perinatal characteristics. J
391
Pediatr. 2007 Mar; 150:291-4.
392
TE D
389
24. Wedgeworth E, Glover M, Irvine AD, et al. Propranolol in the treatment of infantile haemangiomas: lessons from the European Propranolol In the Treatment of Complicated
394
Haemangiomas (PITCH) Taskforce survey. Br J Dermatol. 2016 Mar;174(3):594-601.
395
25. de Graaf M, Knol MJ, Totté JE, et al. E-learning enables parents to assess an infantile
397 398 399 400
AC C
396
EP
393
hemangioma. J Am Acad Dermatol. 2014 May; 70(5):893-8. 26. Bisgaier J, Rhodes KV. Auditing access to specialty care for children with public insurance. N Engl J Med. 2011 Jun 16; 364(24):2324-33. 27. Prindaville B, Antaya RJ, Siegfried EC. Pediatric dermatology: past, present, and future. Pediatr Dermatol. 2015 Jan-Feb; 32(1):1-12.
ACCEPTED MANUSCRIPT 20
403 404 405 406 407
Access for Children. Arch Pediatr Adolesc Med. 2012 Apr; 166(4):304-10. 29. Skinner AC, Mayer ML. Effects of insurance status on children's access to specialty care: a systematic review of the literature. BMC Health Serv Res. 2007 Nov 28; 7:194.
RI PT
402
28. Bisgaier J, Polsky D, Rhodes KV. Academic Medical Centers and Equity in Specialty Care
30. Chaudhry SB, Armbrecht ES, Shin Y, et al. Pediatric access to dermatologists: Medicaid versus private insurance. J Am Acad Dermatol. 2013 May; 68(5):738-48.
31. U.S. Government Accountability Office. Medicaid and CHIP: most physicians serve covered
SC
401
children but have difficulty referring them for specialty care. Report to Congressional
409
Committees. http://www.aap.org/workforce/GAO_Report_Medicaid_CH_June2011.pdf.
410
Accessed: July 7, 2017.
413 414 415 416
United States. N Engl J Med. 2016 Nov 3; 375:1790-6.
33. Mead KH, Brantley E, Zur J, Goldberg DG. Collaboration Across the Safety Net in the Era
TE D
412
32. Chokshi DA, Chang JE, Wilson RM. Health Reform and the Changing Safety Net in the
of Health Reform: Opportunities or Obstacles? Med Care Res Rev. 2017 Jun; 74(3):286-310. 34. Office of Disease Prevention and Health Promotion. Healthy People 2020, 2020. https://www.healthypeople.gov/. Accessed: September 19, 2017.
EP
411
M AN U
408
35. Berdahl TA, Friedman BS, McCormick MC, Simpson L. Annual report on health care for
418
children and youth in the United States: trends in racial/ethnic, income, and insurance
419
disparities over time, 2002-2009. Acad Pediatr. 2013 May-Jun; 13(3):191-203.
420
AC C
417
ACCEPTED MANUSCRIPT 21 Table I. Characteristics of Children Evaluated for Infantile Hemangioma at Initial Subspecialty Presentation Age at IH Presentation Early (N = 219) Delayed (N = 284)
EP
a
Late (N = 301)
P value
21.1 (29.4)
n/a
4.3 (0.9)
148 (67.6) 71 (32.4)
183 (64.4) 101 (35.6)
172 (78.5) 14 (6.4) 33 (15.1)
199 (70.1) 24 (8.5) 61 (21.5)
15 (6.9) 186 (84.9) 18 (8.2) 32.3 (29.0)
33 (11.6) 218 (76.8) 33 (11.6) 35.4 (33.1)
18 (6.0) 256 (85.0) 27 (9.0) 33.2 (34.1)
182 (83.1) 37 (16.9)
208 (73.2) 76 (26.8)
229 (76.1) 72 (23.9)
192 (87.7) 27 (12.3) 28.8 (25.5)
236 (83.1) 48 (16.9) 29.5 (25.7)
244 (81.1) 57 (18.9) 27.9 (23.9)
7.5 (5.5)
6.6 (4.3)
5.2 (3.4)
<0.001
181 (82.6) 38 (17.4)
239 (84.2) 45 (15.8)
264 (87.7) 37 (12.3)
0.241
158 (72.2) 61 (27.8)
188 (66.2) 96 (33.8)
226 (75.1) 75 (24.9)
0.056
200 (91.3) 19 (8.7)
272 (95.8) 12 (4.2)
294 (97.7) 7 (2.3)
0.003
RI PT
1.9 (0.7)
TE D
203 (67.4) 98 (32.6)
0.679
215 (71.4) 25 (8.3) 61 (20.3)
0.283
SC
Characteristic Overall (N = 804) Demographics Age, mean (SD), month 9.9 (20.0) Gender Female 534 (66.4) Male 270 (33.6) Race Caucasian 586 (72.9) African American 63 (7.8) Otherb 155 (19.3) Ethnicity Hispanic or Latino 66 (8.2) Not Hispanic or Latino 660 (82.1) Unidentified 78 (9.7) Distance to clinic, mean 33.7 (32.4) (SD), km c Payor Higher SES 619 (77.0) Lower SES 185 (23.0) JHHC program No 672 (83.6) Yes 132 (16.4) d DCI Score, mean (SD) 28.7 (25.0) Clinical Characteristics Severity score, mean (SD) 6.4 (4.5) Ulceration or bleeding No 684 (85.1) Yes 120 (14.9) Number of IH lesion One 572 (71.1) More than one 232 (28.9) Type of IH lesion Non-segmental 766 (95.3) Segmental 38 (4.7)
M AN U
421 422
0.048 0.540
0.030
0.128 0.726
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; MA/CHIP = medical assistance/Children’s Health Insurance Program; SD = standard deviation a Statistically significant P-values based upon analysis of variance or Chi squared tests for continuous and categorical variables, respectively, in bold b Other includes Asian (n = 30), American Indian or Alaska Native (n = 1), Native Hawaiian or Pacific Islander (n = 1), two or more races (n = 28), and unidentified (n = 95) c Lower SES includes insurance through the Maryland MA/CHIP program; Higher SES includes commercial non-MA/CHIP and self-pay insurances d DCI score is an estimated measure of economic distress and prosperity for each zip-code based on seven indicators: education, housing, unemployment, poverty rate, income, change in employment, and change in business establishments. Higher scores correspond to higher levels of distress
423
ACCEPTED MANUSCRIPT 22 Table II. Characteristics at Initial Infantile Hemangioma Presentation by Type of Insurance Provider (Primary Payor) and Membership in the Johns Hopkins HealthCare LLC Payor Program, Respectively
9.8 (23.1)
0.941
9.7 (18.9)
11.3 (24.7)
0.387
182 (29.4) 208 (33.6) 229 (37.0)
37 (20.0) 76 (41.1) 72 (38.9)
0.030
192 (28.6) 236 (35.1) 244 (36.3)
27 (20.5) 48 (36.4) 57 (43.2)
0.128
415 (67.0) 204 (33.0)
119 (64.3) 66 (35.7)
0.492
511 (82.6) 24 (3.9) 84 (13.6)
75 (40.5) 39 (21.1) 71 (38.4)
<0.001
28 (4.5) 526 (85.0) 65 (10.5) 32.0 (26.3)
38 (20.5) 134 (72.5) 13 (7.0) 39.5 (47.1)
569 (91.9) 50 (8.1) 23.5 (21.3)
103 (55.7) 82 (44.3) 46.2 (28.3)
6.2 (4.3)
SC
10.0 (19.0)
90 (68.2) 42 (31.8)
0.639
515 (76.6) 46 (6.9) 111 (16.5)
71 (53.8) 17 (12.9) 44 (33.3)
<0.001
42 (6.3) 559 (83.2) 71 (10.6) 33.1 (29.8)
24 (18.2) 101 (76.5) 7 (5.3) 37.1 (43.4)
<0.001
569 (84.7) 103 (15.3) 26.6 (23.7)
50 (37.9) 82 (62.1) 39.5 (28.3)
7.1 (5.1)
0.026
6.4 (4.4)
6.7 (5.0)
0.390
538 (86.9) 81 (13.1)
146 (78.9) 39 (21.1)
0.007
578 (86.0) 94 (14.0)
106 (80.3) 26 (19.7)
0.092
433 (69.9) 186 (30.1)
139 (75.1) 46 (24.9)
0.172
470 (69.9) 202 (30.1)
102 (77.3) 30 (22.7)
0.089
592 (95.6) 27 (4.4)
174 (94.0) 11 (6.0)
0.373
639 (95.1) 33 (4.9)
127 (96.2) 5 (3.4)
0.822
M AN U
444 (66.1) 228 (33.9%)
EP
Demographics Age, mean (SD), month Age at presentation Early (<3 months) Delayed (3-6 months) Late (>6 months) Gender Female Male Race Caucasian African American c Other Ethnicity Hispanic or Latino Not Hispanic or Latino Unidentified Distance to clinic, mean (SD), km JHHC Program No Yes DCI Score, mean (SD)d Clinical Characteristics Severity score, mean (SD) Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
Primary Payor Lower-SES (N = 185)
Age at IH Presentation Membership in JHHC Payor Program b b P value Non-JHHC JHHC P value (N = 672) (N = 132)
RI PT
Higher-SES (N = 619)
a
TE D
Characteristic
AC C
424 425 426
<0.001 0.005
<0.001
<0.001 0.188
<0.001 <0.001
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; MA/CHIP = medical assistance/Children’s Health Insurance Program; SD = standard deviation a Lower SES includes insurance through the Maryland MA/CHIP program; Higher SES includes commercial non-MA/CHIP and self-pay insurances b Statistically significant P-values based upon t tests with unequal variances or Chi squared tests for continuous and categorical variables, respectively, in bold c Other includes Asian (n = 30), American Indian or Alaska Native (n = 1), Native Hawaiian or Pacific Islander (n = 1), two or more races (n = 28), and unidentified (n = 95) d DCI score is an estimated measure of economic distress and prosperity for each zip-code based on seven indicators: education, housing, unemployment, poverty rate, income, change in employment, and change in business establishments. Higher scores correspond to higher levels of distress
ACCEPTED MANUSCRIPT 23 Table III. Univariate and Multivariable Generalized Ordered Logistic Regression Analysis Evaluating the Association of Demographic and Clinical Characteristics with Age at First Infantile Hemangioma Presentation
REF 0.98 (0.74-1.30)
REF 1.19 (0.74-1.90)
REF 1.05 (0.62-1.77)
REF 1.30 (0.71-2.37) 1.00 (0.66-1.53) 1.00 (0.96-1.04) REF 1.67 (1.12-2.48) REF 1.41 (1.00-1.98) 0.99 (0.94-1.04) 0.90 (0.88-0.93)
0.60 (0.34-1.06)
1.09 (0.77-1.52)
REF 0.36 (0.19-0.66)
REF 0.81 (0.50-1.31) 1.02 (0.66-1.58) 1.00 (0.96-1.04)
REF 2.11 (1.31-3.38)
1.22 (0.81-1.85)
REF 1.37 (0.92-2.02) 0.95 (0.90-1.01)
0.90 (0.87-0.93)
REF 0.74 (0.52-1.0)
REF 1.07 (0.76-1.51)
SC
REF 1.00 (0.77-1.31)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) Payor Higher SES Lower SES JHHC program No Yes DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
M AN U
Predictors
Adjusted OR (95% CI) Early vs. Early/Delayed Delayed/Late vs. Late
RI PT
Unadjusted OR (95% CI) Early vs. Early/Delayed Delayed/Late vs. Late
TE D
427 428 429
0.73 (0.53-1.01)
REF 0.92 (0.63-1.34) REF 1.09 (0.76-1.55)
0.72 (0.52-1.00)
REF 0.59 (0.30-1.16)
430
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for payor status, and number of IH lesions in both unadjusted and adjusted regressions and Hispanic ethnicity in unadjusted analyses
ACCEPTED MANUSCRIPT 24 Table IV. Subgroup Analysis (Non- Johns Hopkins HealthCare LLC vs. Johns Hopkins HealthCare LLC Payors) – Multivariable Generalized Ordered Logistic Regression Analysis of Demographic Predictors for Age at First Infantile Hemangioma Presentation, Adjusted for Clinical Characteristic Confounders
REF 1.37 (0.67-2.84)
REF 1.11 (0.60-2.04)
REF 1.14 (0.38-3.44)
REF 2.37 (1.33-4.22) 0.94 (0.88-1.01) 0.89 (0.86-0.93) REF 1.01 (0.66-1.54)
REF 1.15 (0.47-2.80) 0.86 (0.19-3.92) 1.00 (0.92-1.08)
M AN U
REF 0.69 (0.38-1.25) 1.01 (0.64-1.61) 1.00 (0.96-1.05)
SC
REF 0.91 (0.67-1.23)
REF 1.05 (0.72-1.53) REF 0.68 (0.33-1.37)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) Payor Higher SES Lower SES DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
JHHC Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
RI PT
Non-JHHC Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
Predictors
1.15 (0.70-1.90)
REF 1.34 (0.61-2.95) 0.97 (0.85-1.10) 0.95 (0.87-1.04)
TE D
431 432 433 434
0.69 (0.49-0.99)
REF 0.68 (0.30-1.52) REF 1.00 (0.44-2.25) REF 0.12 (0.01-1.55)
435
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for payor status and number of IH lesions in regression analyses of NonJHHC payors; no proportional odds assumption was violated in regression analyses of JHHC payors
ACCEPTED MANUSCRIPT 25 Table V. Subgroup Analysis (Higher Socioeconomic Status vs. Lower Socioeconomic Status Payor) – Multivariable Generalized Ordered Logistic Regression Analysis of Demographic Predictors for Age at First Infantile Hemangioma Presentation, Adjusted for Clinical Characteristic Confounders
REF 0.77 (0.42-1.41)
REF 1.08 (0.50-2.37)
REF 1.13 (0.51-2.51)
REF 1.30 (0.75-2.26) 0.96 (0.89-1.03) 0.90 (0.87-0.94) REF 0.83 (0.53-1.29)
REF 0.76 (0.36-1.61) 1.16 (0.35-3.78) 1.03 (0.97-1.10)
M AN U
REF 0.77 (0.39-1.52) 1.02 (0.63-1.64) 0.98 (0.93-1.04)
SC
REF 1.04 (0.76-1.43)
REF 1.05 (0.71-1.56) REF 0.66 (0.36-1.63)
EP
Demographics Gender Female Male Race Other African American Ethnicity Non-Hispanic Hispanic Undefined Distance to clinic (10km) JHHC Program No Yes DCI Score Clinical Characteristics Severity score Ulceration or bleeding No Yes Number of IH lesion One More than one Type of IH lesion Non-segmental Segmental
Lower-SES Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
RI PT
Higher-SES Adjusted OR (95% CI) Early vs. Early/Delayed vs. Delayed/Late Late
Predictors
TE D
436 437 438 439
0.66 (0.46-0.96)
REF 1.43 (0.78-2.61) 1.06 (0.91-1.24)
0.85 (0.75-0.97)
0.96 (0.89-1.05)
0.85 (0.78-0.93)
REF 1.39 (0.66-2.94) REF 1.09 (0.5-2.16)
REF 0.06 (0.01-0.34)
0.91 (0.15-5.36)
440
AC C
Abbreviations: IH = infantile hemangioma; SES = socioeconomic status; JHHC = Johns Hopkins HealthCare LLC; DCI = distressed community index; SD = standard deviation; OR = odds ratio; CI = confidence interval; REF = reference value Note: proportional odds assumption was violated for number of IH lesions in regression analyses of higher SES children; proportional odds assumption was violated for DCI score, severity score, and type of IH lesion for lower SES children
ACCEPTED MANUSCRIPT
CAPSULE SUMMARY
•
What is already known on this topic: Children with Medicaid face greater barriers of
RI PT
access-to-care. Delayed management of complicated infantile hemangiomas compromises outcomes. •
What this article adds to our knowledge. Lower socioeconomic status is associated with
SC
delayed care for infantile hemangioma, but not among children with institutional caremanagement services.
M AN U
How this information impacts clinical practice and/or changes patient care. Institutionfacilitated managed-care programs may moderate socioeconomic disparities in accessing
EP
TE D
timely specialty care and improve outcomes.
AC C
•