Low Rates of Preventive Healthcare Service Utilization Among Adolescents and Adults With Down Syndrome

Abstract

Introduction:

People with Down syndrome have health risks that require specific lifelong preventive healthcare. With increasing life expectancy, people with Down syndrome also face health conditions typical of their unaffected peers and thus need coordinated healthcare. The purpose of this study is to describe rates of age/gender- and Down syndrome-specific preventive healthcare activities among adolescents and adults with Down Syndrome.

Methods:

Using Medicaid claims (2006–2010) in California, Colorado, Michigan, and Pennsylvania, the cohort was defined as people with Down syndrome aged ≥12 years seen by primary care providers and enrolled in Medicaid for ≥45 of 60 months without dual Medicare enrollment (n=3,501). Age focus-consistent primary care providers were defined as having a focus concordant with a patient’s age: 12–17 years, child or mixed focus; ≥26 years, adult or mixed focus; 18–25 years, any focus. Differences in healthcare activities were evaluated using Pearson’s chi-square, Fisher’s exact, and Kruskal-Wallis tests. Analyses were performed in 2015–2017.

Results:

Seventy-nine percent of the cohort had an age focus-consistent primary care provider. However, 40% of adults aged ≥26 years received care from child-focused primary care provider. Only 43% with an age focus-consistent provider had ≥1 well examination (age focus-inconsistent primary care provider: 35%, p<0.001). Most preventive activities had poor rates (<50%) regardless of age focus consistency between provider and patient age or whether they were age/gender- or Down syndrome-specific (well examinations, vaccinations, sleep apnea, hearing, breast, cervical, and colon cancer screenings). Lipids, vision, and thyroid screenings reached moderate levels (50% to <80%).

Conclusions:

Rates of age/gender- and Down syndrome-specific preventive recommendations were low among adolescents and adults with DS, regardless of the age focus consistency of their primary care provider. This represents a significant opportunity to improve primary care in this vulnerable population.

INTRODUCTION

Down syndrome (DS) is the leading identifiable genetic cause of intellectual disability, occurring in 1:700 to 1:1,000 live births.^1–4 Historically, the majority of children with DS did not survive childhood.^4–8 Thanks to medical advancements, >80% of people with DS now reach adolescence with a median life expectancy in their mid-50s.^4–6,9–15 Owing to their genetic condition, people with DS have increased risks for multiple comorbidities that can present during childhood (e.g., intellectual disability, language difficulties, hypotonia, congenital heart disease [CHD]) or develop throughout their lives (e.g., hypothyroidism, hearing loss, vision abnormalities, obesity, obstructive sleep apnea, dementia).^16–23

There are well-established DS guidelines for preventive care during childhood (age ≤21 years) and clear recommendations during adulthood.^16–22 Previous work by Jensen et al.²⁴ within a local health system demonstrated suboptimal adherence to most preventive recommendations in adults with DS, regardless of the field of their primary care provider (PCP). Toler²⁵ documented that preventive care is frequently neglected in women with DS. Similar work by Santoro and colleagues^26,27 within pediatric settings showed poor baseline adherence to the majority of guidelines for children with DS. Yet, people with DS are more engaged in society and reach higher levels of achievement than ever before, incurring risks and developing health issues more typical of their unaffected peers.²⁸ Consequently, preventive care among adolescents and adults with DS must incorporate both age/gender- and DS-specific preventive healthcare recommendations. However, little is known about national patterns of preventive care within this population.

This retrospective cohort study examines age/gender- and DS-specific preventive healthcare patterns among a cohort of adolescents and adults with DS in four U.S. states. Based on prior research²⁴ and the authors’ experiences as PCPs, it is hypothesized that preventive care will be suboptimal for adolescents and adults with DS, but that patients whose PCP’s focus was consistent with their age would have higher rates of recommended age/gender-specific care.

METHODS

Study Sample

This study was approved by the Colorado Multiple IRB (13–2072). Data were obtained from the Center for Medicare and Medicaid Services Medicaid Analytic Extract files for all Medicaid beneficiaries in Colorado, California, Michigan, and Pennsylvania from 2006 to 2010 (N=28,752,018). The analysis was restricted to Medicaid beneficiaries with DS who were aged ≥12 years by January 1, 2006. First, patients with DS were identified by ICD-9-CM=758.0²⁹ within their claims during the study years (n=20,277). Patients whose DS diagnosis only occurred during obstetric visits were not classified as having DS as this typically reflects DS in their fetus³⁰ (excluded n=11). Codes used to identify the domains throughout this study are available upon request.^{10,16–20,24,31–58}

All medical encounters are not necessarily captured within claims for clients on Medicaid managed care, those with restricted benefits, private insurance, enrolled in a state child health insurance program, or Medicare dual enrolled. The authors therefore restricted cohort selection to patients enrolled in Medicaid fee-for-service (including primary care case management) for ≥75% of the study frame (≥45 of 60 months) to reliably capture encounters (n=9,758). Patients dually enrolled in Medicaid and Medicare were excluded (n=5,807) as investigators could not assess their services billed under Medicare. Each state was impacted differently by the exclusion of dually enrolled patients: Colorado retained 37% of their cohort, Michigan 6%, Pennsylvania 15%, and California 56%. Patients without clinical visits with a least one PCP were excluded from analysis (n=450). This led to a final cohort of 3,501 adolescents and adults with DS.

Measures

Patient study age was assigned by age at the study midpoint: July 1, 2008. Age categories were based on study age: 12–17 years, adolescents; 18–25 years, transition aged; ≥26 years, adults. Age-specific preventive healthcare activities were evaluated for all patients meeting the specified age criteria at any time during the 5-year study. Rates of recommended activities were assessed by patient age at each visit, not study age.

The PCPs were identified as providers billing ≥10 well examinations in a calendar year. PCPs were identified as: (1) child-focused, (2) adult-focused, or (3) mixed-focused, based on the proportion of their well examinations with child- or adult-specific billing codes. PCPs with ≥80% of well examinations billed as well child were categorized as child-focused; providers with ≥80% of well examinations billed as well adult were considered adult-focused. PCPs not meeting either category were assigned the status of “mixed-focused”. Each patient was attributed to a PCP type (child, adult, or mixed focus or none) based upon the type of PCP they encountered most frequently. In the case of ties, the PCP type from the most recent well examination was retained. For patients without well examinations, the PCP type from the most recent claim was retained. Age focus-consistent PCPs were defined as having a focus concordant with the patient’s age. This includes a child- or mixed-focused PCP for a person aged 12–17 years or an adult- or mixed-focused PCP for a person aged ≥26 years. A person aged 18–25 years would be age focus consistent if seen by any PCP focus. Although the authors anticipated the training backgrounds of PCPs would be as follows, the poor fidelity of the provider specialty field in Medicaid data did not allow for clear identification of PCP specialty: child-focused, Pediatrics; adult-focused, Internal Medicine or Geriatrics; mixed-focused, Family Medicine or combined Internal Medicine and Pediatrics. The term “age focus consistent” was used to designate PCP type as recommendations were in place during the study years making 21 years the upper age limit for patients seen by Pediatricians (child-focused PCPs).⁵⁹

The Rural-Urban Commuting Area Codes approximation^60–62 was used to assign patient ZIP codes as urban or rural. ZIP codes without Rural-Urban Commuting Area Codes were assigned as urban/rural based upon classification of surrounding ZIP codes or the county containing the majority of the ZIP code.

Identification of comorbidities outside of DS was made by the presence of ≥1 inpatient or ≥2 outpatient claims with the diagnosis code of interest.⁶³ Patients with DS with only 1 outpatient claim were not excluded as they were found to have similar demographics, diagnoses, and healthcare utilization to those patients with ≥1 claim or with an inpatient claim containing a DS diagnosis code.

This study evaluated age/gender- and DS-specific preventive healthcare activities that were measurable by ICD-9/CPT codes, could occur within 5 years, and had near uniform acceptance during the study years (2006–2010). For each activity, patients received “credit” if the activity occurred at least once during this 5-year study. Age/gender-specific preventive recommendations included: well person examination, influenza vaccination, cholesterol screening (age ≥18 years), cervical cancer screening (women aged 21–65 years), and adolescent vaccinations (age 12–18 years: tetanus, diphtheria, and pertussis; human papillomavirus; meningococcal).^35,36,58,64 Authors also observed screening activities based on personal risk for osteoporosis (age ≥40 years), breast cancer (women aged ≥40 years), colorectal cancer (age ≥50 years), diabetes, and pneumococcal vaccination. Screening ages for cholesterol, cervical cancer, colorectal cancer, and adolescent vaccinations reflect recommendations from the Agency for Healthcare Research and Quality’s National Guideline Clearinghouse during the study years.^35,36 This study followed the more conservative approach to breast cancer screening (age ≥40 years) from the American College of Obstetricians and Gynecologists.⁶⁵ Rates of osteoporosis screening were followed starting at age 40 years in this population owing to high rates of low bone mineral density in people with DS of both genders starting in their 40s.^66,67 This study additionally observed patterns of pneumonia vaccination and diabetic labs in this cohort due to increased risk of complications of respiratory illness and high rates of obesity, respectively, in people with DS.^10,17,68 For DS-specific recommendations, the authors assessed the presence of thyroid, vision, and hearing screening, as well as screening tailored to risk/symptoms for acquired cardiac valve disease and obstructive sleep apnea.^69–71 Rates of each activity were classified as good (≥80%), moderate (50% to <80%), and poor (<50%).^24,69–71

Statistical Analysis

Patient characteristics and care patterns were compared between patients with DS by PCP focus and age consistency of a PCP’s focus using Pearson’s chi-square, Fisher’s exact, or Kruskal-Wallis tests. Preventive care patterns were additionally stratified by patient study age and compared by age consistency of PCP’s focus with patient study age. Analyses were performed in 2015–2017 using SAS, version 9.4.

RESULTS

At total of 3,501 adolescents and adults with DS met inclusion criteria, with a median age of 25 (IQR=19–35) years. Nearly 22% of the cohort were adolescents (aged 12–17 years), 32% were transition aged (18–25 years), and 47% were adults (aged ≥26 years). Fifty-two percent of the cohort was male; 86% resided in urban settings. The cohort was 39% Hispanic, 38% White, 7% Black, 5% Asian, and 4% Native Hawaiian/Pacific Islander, with 6% other/unknown (Table 1). Most of the cohort (83%) resided in California, with 8% in Colorado, 7% in Pennsylvania, and 2% in Michigan. The overwhelming majority was enrolled in Medicaid for the entire study (median=60 months, IQR=60, 60).

Table 1.

Patient Characteristics

Characteristic	All patients, % (n) or median (IQR)	Child PCP, % (n) or median (IQR)	Mixed PCP, % (n) or median (IQR)	Adult PCP, % (n) or median (IQR)	p-value	Age focus-consistent PCP, % (n) or median (IQR)	Age focus-inconsistent PCP, % (n) or median (IQR)	p-value
Total	3,501	51.6% (1,808)	24.4% (854)	24.0% (839)		78.9% (2,763)	21.1% (738)
Male	52.4% (1,835)	55.8% (1,009)	54.0% (461)	43.5% (365)	<0.001	52.4% (1,448)	52.4% (387)	0.988
Study period age, years					<0.001			<0.001
12–17 (adolescent)a	21.5% (754)	30.6% (554)	12.8% (109)	10.8% (91)		24.0% (663)	12.3% (91)
18–25 (transition)b	31.9% (1,117)	33.6% (607)	30.9% (264)	29.3% (246)		40.4% (1,117)	NA
≥26 (adult)c	46.6% (1,630)	35.8% (647)	56.3% (481)	59.8% (502)		35.6% (983)	87.7% (647)
Median age (IQR) in years	25.0 (18.8–35.4)	21.8 (17.2–30.9)	27.7 (21.3–37.5)	29.6 (22.3–39.8)	<0.001	22.8 (18.2–31.8)	33.7 (27.8–41.8)	<0.001
Race					<0.001			<0.001
White	38.2% (1,338)	35.6% (643)	45.9% (392)	36.1% (303)		36.5% (1,009)	44.6% (329)
Hispanic	39.3% (1,376)	44.4% (802)	31.9% (272)	36.0% (302)		40.9% (1,130)	33.3% (246)
Black	7.3% (255)	6.6% (119)	8.1% (69)	8.0% (67)		7.4% (205)	6.8% (50)
Asian	5.1% (179)	3.4% (62)	4.2% (36)	9.7% (81)		5.5% (151)	3.8% (28)
Native Hawaiian/Pacific Islander	3.7% (131)	3.8% (69)	3.3% (28)	4.1% (34)		3.8% (105)	3.5% (26)
Unknown/other	6.3% (222)	6.3% (113)	6.7% (57)	6.2% (52)		5.9% (163)	8.0% (59)
Urban residence	86.3% (3,023)	86.0% (1,554)	80.1% (684)	93.6% (785)	<0.001	86.2% (2,383)	86.7% (640)	0.739
State					<0.001			<0.05
CA	82.7% (2,896)	80.9% (1,462)	74.2% (634)	95.4% (800)		81.6% (2,254)	87.0% (642)
CO	8.1% (285)	11.7% (211)	7.1% (61)	1.5% (13)		Non-CA 18.4% (509)a	Non-CA 13.0% (96)a
MI	2.3% (79)	2.4% (44)	2.6% (22)	1.5% (13)		Non-CA 18.4% (509)d	Non-CA 13.0% (96)a
PA	6.9% (241)	5.0% (91)	16.0% (137)	1.5% (13)		Non-CA 18.4% (509)a	Non-CA 13.0% (96)a
Months enrolled in Medicaid (IQR)	60.0 (60.0–60.0)	60.0 (59.0–60.0)	60.0 (60.0–60.0)	60.0 (60.0–60.0)	0.003	60.0 (60.0–60.0)	60.0 (60.0–60.0)	0.194

Notes: Boldface indicates statistical significance (p<0.05) from Pearson’s χ² test, Fisher’s exact test, or the Kruskal-Wallis test.

^aAmong adolescents, 73.5% attended child-focused PCPs, 14.4% attended mixed-focus PCPs, and 12.1% attended adult-focused PCPs. Also, 87.9% attended age focus-consistent PCPs, while 12.1% attended age focus-inconsistent PCPs.

^bAmong transition age group, 54.3% attended child-focused PCPs, 23.7% attended mixed-focus PCPs, and 22.0% attended adult-focused PCPs. Also, 100% attended age focus-consistent PCPs.

^cAmong adults, 39.7% attended child-focused PCPs, 29.5% attended mixed-focus PCPs, and 30.8% attended adult-focused PCPs. Also, 60.3% attended age focus-consistent PCPs, while 39.7% attended age focus-inconsistent PCPs.

^dNon-California cell collapsed for age-appropriate/age-inappropriate as at least one of the states has n≤10 and cannot be reported per the data use agreement.

NA, not applicable; PCP, primary care provider; CA, California; CO, Colorado; MI, Michigan; PA, Pennsylvania.

Fifty-two percent of the cohort received care from child-focused PCPs, followed by 24% each from mixed- and adult-focused PCPs. Providers from each PCP type cared for patients in each age category (Table 1, Figure 1). Among adolescents with DS, 88% received care from a PCP whose focus was consistent with their age (74% child focus, 14% mixed focus). Owing to the definition of age focus consistency, all transition-aged patients received care from an age focus-consistent PCP (54% child-focused, 24% mixed-focused, and 22% adult-focused). The proportion of patients receiving care from adult- or mixed-focused PCPs increased with increasing patient age. However, 40% of adults aged ≥26 years with DS received care from child-focused PCPs.

Figure 1.

Patterns of PCP focus among adolescents and adults with Down syndrome by patient age. PCP, primary care provider

People in the cohort were medically complex (Table 2) with a median of 5 organ systems chronically affected (IQR=3–7). The most frequent of these are: neurologic (76%), mental/behavioral health (74%), respiratory (65%), endocrine (57%), digestive (51%), and dermatologic (50%). Eleven percent had encounters for hypertension, 10% for diabetes, and 23% for hyperlipidemia. Among comorbidities commonly associated with DS, 14% had encounters for CHD, 5% Eisenmenger syndrome, and 2% pulmonary hypertension. Twenty-seven percent had encounters for hypothyroidism and 15% for obstructive sleep apnea.

Table 2.

Medical Complexity of Cohort of Adolescents and Adults With Down Syndrome

Medical Complexity	All patients, % (n) or median (IQR), n=3,501	Child-focused PCP, % (n) or median (IQR), n=1,808	Mixed-focus PCP, % (n) or median (IQR), n=854	Adult-focused PCP, % (n) or median (IQR), n=839	p-value	Age focus-consistent PCP, % (n) or median (IQR), n=2,763	Age focus-inconsistent PCP, % (n) or median (IQR), n=738	p-value
Highlighted comorbidities common either to Down syndrome or adulthood
Congenital heart disease	13.5% (471)	15.9% (288)	11.6% (99)	10.0% (84)	<0.001	13.8% (381)	12.2% (90)	0.26
Eisenmenger syndrome	4.7% (165)	5.9% (106)	3.5% (30)	3.5% (29)	0.004	4.6% (127)	5.1% (38)	0.529
Hypothyroidism	27.2% (953)	24.2% (437)	28.7% (245)	32.3% (271)	<0.001	26.5% (732)	29.9% (221)	0.061
Hypoxia	4.4% (155)	4.9% (88)	4.8% (41)	3.1% (26)	0.100	4.4% (121)	4.6% (34)	0.789
Obstructive sleep apnea	15.3% (536)	16.3% (294)	15.2% (130)	13.3% (112)	0.153	15.3% (422)	15.4% (114)	0.907
Pulmonary hypertension	1.9% (66)	2.0% (36)	2.0% (17)	1.5% (13)	0.714	1.8% (49)	2.3% (17)	0.347
Dementia	0.9% (33)	—a	—a	—a	NS	0.8% (21)	1.6% (12)	0.031
Diabetes	10.2% (356)	9.4% (170)	11.0% (94)	11.0% (92)	0.301	8.9% (247)	14.8% (109)	<0.001
Hyperlipidemia	22.7% (796)	17.1% (310)	24.1% (206)	33.4% (280)	<0.001	21.1% (584)	28.7% (212)	<0.001
Hypertension	10.6% (370)	8.3% (150)	11.1% (95)	14.9% (125)	<0.001	9.3% (257)	15.3% (113)	<0.001
Chronically affected organ systems
Number of organ systems chronically affected	5.0 (3.0–7.0)	5.0 (3.0–7.0)	5.0 (3.0–7.0)	5.0 (3.0–7.0)	<0.001	5.0 (3.0–7.0)	5.0 (4.0–7.0)	<0.001
Neurologic	75.8% (2,653)	77.7% (1,404)	72.0% (615)	75.6% (634)	0.006	75.9% (2,096)	75.5% (557)	0.828
Mental/behavioral health	73.7% (2,580)	73.3% (1,325)	72.4% (618)	75.9% (637)	0.214	72.3% (1,997)	79.0% (583)	<0.001
Respiratory	65.0% (2,276)	66.2% (1,197)	63.0% (538)	64.5% (541)	0.252	65.5% (1,811)	63.0% (465)	0.199
Endocrine	57.3% (2,005)	52.2% (943)	60.0% (512)	65.6% (550)	<0.001	55.2% (1,525)	65.0% (480)	<0.001
Digestive	50.7% (1,776)	52.3% (946)	46.4% (396)	51.7% (434)	0.013	49.3% (1,361)	56.2% (415)	<0.001
Dermatologic	50.2% (1,758)	49.7% (898)	49.5% (423)	52.1% (437)	0.461	50.1% (1,384)	50.7% (374)	0.777
Musculoskeletal	39.8% (1,393)	37.4% (677)	41.5% (354)	43.1% (362)	0.011	38.3% (1,059)	45.3% (334)	<0.001
Cardiovascular	38.0% (1,330)	38.3% (692)	35.5% (303)	39.9% (335)	0.158	36.2% (1,000)	44.7% (330)	<0.001
Genitourinary	31.0% (1,087)	27.7% (500)	32.8% (280)	36.6% (307)	<0.001	30.0% (828)	35.1% (259)	0.007
Hematologic	20.0% (699)	17.8% (322)	19.1% (163)	25.5% (214)	<0.001	17.7% (490)	28.3% (209)	<0.001

Notes: Boldface indicates statistical significance (p<0.05) from Pearson’s χ² test, Fisher’s exact test, or the Kruskal-Wallis test.

^aExact cell sizes of ten or less (or their complement) and related statistics are not shown per the data use agreement.

NS, not significant (p≥0.05); PCP, primary care provider.

Using the framework of good (≥80%), moderate (50% to <80%), and poor (<50%) rates for each preventive activity,^24,69–71 the following patterns occurred at least once in the 5-year study period (Figure 2).

Figure 2.

Preventive care Patterns by age consistency of PCP focus.

Notes: All data in this figure represent screening patterns occurring at least once in 2006–2010. Age focus-consistent PCPs are defined as having a focus consistent with a patient’s study age: aged ≤18 years, child-focused; aged ≥26 years, adult-focused. Persons aged 18–25 years are considered to be in transition and therefore can be appropriately seen by child-focused, adult-focused, or mixed-focus PCPs. Overall indicates total cohort.

^ap<0.05 for comparisons between patients seeing age focus-consistent and age focus-inconsistent PCPs.

^bPersonalized screening.

^cExact cell sizes of ten or less (or their complement) and related statistics are not shown per the data use agreement.

PCP, primary care provider; Tdap, tetanus, diphtheria, acellular pertussis; HPV, human papillomavirus; CHD, congenital heart disease.

All age categories had poor rates (<50%) of well examinations. Only 41% of the cohort had ≥1 well examination (1 well examination, 20%; ≥2 well examinations, 21%). More patients with an age focus-consistent PCP had a well examination (43%) than those with an age focus-inconsistent PCP (35%, p<0.001) (Figure 2). More adolescents (aged 12–17 years) receiving care from age focus-consistent PCPs (child- or mixed-focused) had a well examination (49%) compared with those cared for by age focus-inconsistent PCPs (adult-focused: 25%, p<0.001) (Appendix Table 1). Within the transition-aged population (18–25 years), no differences were observed in well examination patterns by PCP type (44% from child- or adult-focused PCPs vs 48% from mixed-focus PCPs, p=0.253). Rates of well examinations among adults with DS did not differ by age focus consistency of PCPs (age focus consistent, 38%; age focus inconsistent, 36%; p=0.466) (Appendix Table 1).

Similarly, there were poor rates (<50%) of all recommended vaccinations (Figure 2). The authors observed no difference in rates of ≥1 influenza vaccination during the study period (age focus-consistent PCP, 34%; age focus-inconsistent PCP, 37%; p=0.227) (Figure 2). Annual rates of influenza vaccination ranged from 11% (2006–2007) to 16% (2009–2010) (Appendix Table 1). Among those aged 12–18 years, rates of tetanus, diphtheria, and pertussis; human papillomavirus (female patients), and meningococcal vaccination were similarly low with higher rates among patients with age focus-inconsistent (i.e. adult-focused) PCPs (Figure 2). Investigators observed low rates of pneumonia vaccination (2.7%), the decision for which would be prompted by individual risk assessment.

Regarding age/gender-specific activities, moderate (50% to <80%) rates of cholesterol screening were observed for patients aged ≥18 years in most subgroups, increasing to 80% among adults with age focus-consistent PCPs (Figure 2, Appendix Table 1). Diabetic lab studies (HbA1c or blood glucose levels) occurred in 37% of the cohort. All procedure-based services occurred at low rates (<50%) during this 5-year study. Breast cancer screening (female patients aged ≥40 years) was highest at 46%, followed by screening for cervical cancer (female patients aged ≥21 years) at 31%, colorectal cancer (people aged ≥50 years) at 26%, and osteoporosis (people aged ≥40 years) at 7%. The only significant difference by age focus consistency of PCP was in cervical cancer screening (age focus-consistent PCPs, 29%; age focus-inconsistent PCPs, 37%; p=0.014) (Appendix Table 1).

Recommendations for primary care in people with DS include annual screening for thyroid, vision, and hearing abnormalities.^16–20 The authors observed moderate rates of thyroid (72%) and vision (55%) screening, with an inconsistent impact of PCP age focus consistency upon these activities (Figure 2). Poor rates of hearing screenings were observed (18% overall), ranging from 10% (adults with age focus-consistent PCPs) to 44% (adolescents with age focus-inconsistent PCPs) (Appendix Table 1). Among the DS-specific screenings tailored to risk/symptoms, <5% of the cohort had a sleep study to evaluate for sleep apnea. More than 90% of the cohort with both DS and a history of CHD had an echocardiogram during this study period. Only 22% of those with DS but without CHD had an echocardiogram to evaluate for acquired cardiac valve disease (Appendix Table 1).

As California contributed 83% of the cohort, the authors conducted a subgroup analysis to evaluate its impact upon the findings. Without California data, 15 of the 18 domains followed in this study remained within the same categories of good, moderate, and poor rates (Appendix Figure 1). Rates of well person examinations and influenza vaccinations increased into the moderate range at 66% and 52%, respectively, but remained below recommended frequencies. Rates of vision examination decreased to 46%.

DISCUSSION

In this cohort of adolescents and adults with DS, child-focused primary care relationships were observed well into adulthood and poor rates (<50%) of most preventive healthcare activities were noted regardless of age focus consistency of PCPs or whether the activity was age/gender- or DS-specific.

Forty-percent of adults with DS (aged ≥26 years) in this cohort received care from child-focused PCPs. There are several possible explanations for this observation. Pediatric (child-focused) training during our study years contained an explicit focus on the care of people with DS that theoretically increases their comfort in caring for people with DS.⁷² Contemporaneous national survey data indicate that only half of general internists feel prepared to care for transition-aged adults with childhood-onset chronic disease.⁷³ Additionally, longitudinal relationships with PCPs who know their patients with DS well have advantages for provider-patient/caregiver interactions.⁵⁹ However, this study observed similar rates of DS-specific activities (both annual and personalized screenings) among providers across age categories, regardless of age focus consistency. Ironically, age focus-inconsistent providers (adult-focused in adolescents, child-focused in adults) demonstrated higher rates of adolescent vaccinations and of cervical cancer screening (Figure 2, Appendix Table 1).

Only 41% of the cohort received at least one well examination during these 5 years. Strikingly, all vaccination rates were <50% regardless of PCP age focus consistency. Although 35% of the cohort had ≥1 influenza vaccination during this 5-year study, annual rates were only 11%–16%. By contrast, adult U.S. influenza vaccination rates were 41% (2009–2010).⁷⁴ This is particularly important as respiratory infection is the leading cause of death among adults with DS after CHD.^10,68 Additionally, people with DS have immunologic abnormalities that increase their risk for serious infections.^75,76 Thus, many DS-specific recommendations include annual influenza vaccination.^16,17

This study observed good rates of laboratory-based cholesterol screening (63%). Within procedure-based activities, screening for breast cancer was the most common (46%), followed by cervical cancer (31%), colorectal cancer (26%), and osteoporosis (7%). By comparison, 2010 U.S. screening rates were 72% for breast cancer, 83% for cervical cancer, and 59% for colorectal cancer, with 21% of U.S. women aged 50–64 years receiving osteoporosis screening in the same time period.^77,78 These findings are unexpected as people with DS have much higher risk of developing osteoporosis (25%–50% prevalence among adult men and women with DS^79–82) than breast cancer (<1% prevalence^83,84). In fact, there is ongoing debate regarding the utility of screening mammography in women with DS.⁸⁴ Given case reports of individuals with DS and breast cancer, however, the authors argue that family history and clinical presentation should guide breast cancer screening decisions in women with DS.⁸⁵

People with DS are a vulnerable population who benefit from well-supported syndrome-specific recommendations.^16–19,22 Among the activities recommended annually for people with DS, thyroid and vision screening reached moderate rates (50% to <80%), whereas hearing screening occurred in <20% of this cohort. This is particularly concerning as rates of hearing loss in adults with DS range from 73% to 100%.⁸⁶ Preventive activities tailored to symptoms among adolescents and adults with DS include screening for obstructive sleep apnea and acquired cardiac valve disease.^17,18 Prevalence estimates for obstructive sleep apnea range from 30% to >90% in DS.^87–94 However, <5% of the cohort had a sleep study. Prevalence estimates for acquired cardiac valve disease in adolescents and adults with DS range from 8% to 46%.^{44,48–50,95–97} In this study, >90% of people with DS and CHD had an echocardiogram. However, only 22% of the cohort without CHD underwent echocardiogram to screen for acquired cardiac valve disease.

Limitations

This retrospective cohort study has several possible limitations. First, the ability to identify conditions and healthcare activities is limited to documentation in claims data and may underestimate conditions or screening practices. Second, the Medicaid population is inherently more complex than commercially insured individuals. However, approximately 80% of people with DS have Medicaid.⁹⁸ Thus, these data offer a reasonable estimation of national healthcare patterns in persons with DS. Third, the algorithm attributing PCP focus is based on proxy coding behaviors owing to poor fidelity in the provider specialty code within Medicaid data. Thus, the PCP focus variable may not fully capture the specialties of all providers. Fourth, as discussed above, California contributed 83% of the cohort. The subgroup analysis excluding California data showed that the trends identified in this study remained qualitatively within the same categories of poor, moderate, or good rates with the following exceptions: Without California patients, the categories for well examinations and influenza vaccinations improved from poor to moderate and decreased for vision screening from moderate to poor (Appendix Figure 1). Given the overall qualitative consistency of the findings and the lack of statistical power to evaluate differences in trends without the California population, the authors retained California patients with DS in this analysis. These limitations notwithstanding, this is among the largest cohort studies of adolescents and adults with DS, representing healthcare patterns in 4 distinct states. The findings are consistent with previous work²⁴ and, the authors believe, provide an accurate depiction of trends in preventive healthcare for adolescents and adults with DS living in the U.S.

CONCLUSIONS

This study of adolescents and adults with DS observed poor rates of most recommended preventive healthcare activities regardless of the age focus consistency of their PCP or whether the activity was age/gender- or DS-specific. This study demonstrates that suboptimal preventive healthcare is nearly universal among adolescents and adults with DS. Future work should evaluate perspectives regarding primary care delivery to and potential interventions for this population, as enhancing preventive care is essential to improving the health outcomes and well-being of people with DS.