Maternal and Child Health Journal

, Volume 16, Issue 9, pp 1771–1778

Factors Associated with a Medical Home Among Children with Attention-Deficit Hyperactivity Disorder

Authors

    • Department of Health Outcomes and PolicyUniversity of Florida
  • Melanie Hinojosa
    • Department of Health Outcomes and PolicyUniversity of Florida
  • Jacqueline Baron-Lee
    • Department of Health Outcomes and PolicyUniversity of Florida
  • Dan Fernandez-Baca
    • Department of Health Outcomes and PolicyUniversity of Florida
  • Ramon Hinojosa
    • Rehabilitation Outcomes Research Center, North Florida/South Georgia Veterans Health System
  • Lindsay Thompson
    • Department of PediatricsUniversity of Florida
Article

DOI: 10.1007/s10995-011-0922-6

Cite this article as:
Knapp, C.A., Hinojosa, M., Baron-Lee, J. et al. Matern Child Health J (2012) 16: 1771. doi:10.1007/s10995-011-0922-6

Abstract

Providing a medical home to children with Attention-Deficit Hyperactivity Disorder (ADHD) is challenging. Little is known about the factors associated with having a medical home for these children, or how comorbidities affect having a medical home. Our study aims are: (1) identify factors associated with having a medical home and five sub-components of a medical home and (2) determine the effect of medical home on several outcomes for children with ADHD. The sample included 5,495 children with ADHD from the 2007 National Survey of Children’s Health. Descriptive and multivariate analyses were conducted. Children with ADHD alone and children with ADHD plus a physical diagnosis had greater frequencies of having a medical home, or meeting the five sub-components, than children with ADHD plus a mental diagnosis. Multivariate results show that children with ADHD plus a physical and/or mental comorbidity were 24–63% more likely to be without a medical home compared to children with only ADHD. Having a medical home also had a bearing on several child health outcomes. Having a medical home was significantly associated with being less likely to have an unmet health need and having fewer missed school days; but also being less likely to have received needed mental health care (P < .05). Our results suggest that there are differences in parent’s perceptions of receiving care among children with ADHD. Pediatric medical home projects and policies should acknowledge that children with ADHD often have comorbidities making their care more complex. These complexities should be addressed during practice transformation and setting reimbursement policies.

Keywords

ADHDMedical homeChild health outcomesComorbiditiesSurvey

Introduction

Attention-deficit disorder and Attention-Deficit Hyperactivity Disorder (ADHD) are among the most commonly diagnosed neurobehavioral disorders in childhood [1]. Approximately 9.5% of children ages 4–17 have been diagnosed with ADHD and this percentage has increased over time [2]. ADHD is characterized by chronic behavioral patterns of inattention, hyperactivity, and/or impulsivity [3]. Treatment regimes include psychostimulant medications, behavioral therapy, or both. Children with ADHD often have, or are at risk for, other mental health concerns such as autism spectrum disorder, suicidal ideation, depression, bipolar disorder, and aggression [46].

A Patient Centered Medical Home has been recommended as a model of care that can facilitate improvements in health outcomes and coordination of care for children with complex conditions, including ADHD. The American Academy of Pediatrics (AAP), Maternal Child Health Bureau, Healthy People 2010, and the American Academy of Family Physicians support the medical home model [710]. Definitions vary, but the AAP defines the components of a medical home as accessible, continuous, comprehensive, family-centered, coordinated, compassionate, and culturally effective [7]. Although many states and health care organizations have begun to endorse and provide care to children under the medical home model, evidence on its effectiveness is currently limited. A 2008 systematic review of the evidence on medical home for children with special health care needs (CSHCN) by Homer et al. [11] concluded that evidence to support that medical homes improve health outcomes is modest [11]. The authors noted that most of the studies focused on children with asthma and many studies only assessed one component of the medical home model.

An even smaller literature exists on the effectiveness of medical homes for children with ADHD. Toomey et al. [12] compared parental-report of having a medical home, and its sub-components, for children with asthma and ADHD [12]. Results suggest that children with ADHD are less likely than children with asthma to have a medical home, and, as severity of either condition increased, the likelihood of having a medical home decreased. A 2009 study of 43 primary care practices found that a medical home was associated with reductions in hospitalizations for children with comorbidities such as asthma and ADHD [13]. Although these studies on children with ADHD provide valuable insights, more information is needed to understand how having a medical home (and meeting each of the sub-components) varies by comorbidities and what affect a medical home has on more diverse sets of health outcomes for children with ADHD.

Given this paucity of evidence, our study has two aims. First, using parental responses from the 2007 National Survey of Children’s Health (NSCH) we compare the factors associated with having a medical home or meeting the five NSCH sub-components of a medical home for children with ADHD versus those with ADHD plus comorbidities. Second, we investigate the impact that having a medical home has on four child health outcome measures including: (1) missed school days, (2) number of preventive health care visits, (3) unmet health care needs, and (4) receipt of needed mental health treatment. For each of these study aims we focus only on children in the NSCH dataset who currently have ADHD, accounting for comorbidities. We hypothesize that children with ADHD plus physical and mental diagnoses (as compared to children with ADHD alone) will be more likely to not have a medical home, not meet the five NSCH sub-components of a medical home, and have poorer child health outcomes.

Subjects and Methods

Sample

Survey data from the 2007 NSCH were used in this study [14]. The NSCH is a random-digit dial survey that collects data on children’s physical and mental health, insurance, access to health care, medical home, neighborhoods, and family health for parents/caregivers of children ages 0–17. Overall, 91,642 respondents representing between 1,725 and 1,932 households from each state and the District of Columbia completed the survey. Our sample included the 5,495 children in the study whose parents indicated that they currently had an ADHD diagnosis. Children were included in the sample if their parent responded affirmatively to the question on the NSCH which asks if your child currently has ADHD.

Measures and Analyses

A dichotomous medical home indicator was generated by the survey developers from parental responses on the five medical home sub-components. The five sub-components of a medical home as defined by the survey developers are: (1) having a personal doctor, (2) having a usual source of care, (3) receiving family-centered care, (4) having no problem getting referrals, and (5) having effective care coordination. Each of these five sub-components is a dichotomous variable.

Descriptive statistics were estimated to determine: (1) the frequency of children with ADHD, and (2) the frequency of children with ADHD, and other comorbid physical or mental health diagnoses, who had a medical home and met the five sub-components. Comorbid physical diagnoses were assessed by items on the NSCH which ask the child’s parent if the child currently has speech problems, Tourette’s syndrome, asthma, diabetes, epilepsy, hearing or vision problems, bone, joint, or muscle problems, or injuries related to the brain. Comorbid mental health diagnoses were assessed by items which ask if the child currently has autism, developmental delay, depression, and/or anxiety. Mutually exclusive ADHD categories were created for children with ADHD alone, ADHD with comorbid mental health diagnoses (i.e., ADHD plus mental diagnosis), ADHD with comorbid physical diagnoses (i.e., ADHD plus physical diagnosis), or ADHD with comorbid mental and physical diagnoses (i.e., ADHD plus mental plus physical diagnosis).

Logistic multivariate regressions were also conducted to determine whether several variables predicted if a child’s parent reported having a medical home or if the parent reported the medical home as meeting the five sub-components. The dependent variable is a medical home composite. This composite is based on the NSCH dataset and information on how the composite was defined are available online (http://www.childhealthdata.org/learn/NSCH). Independent variables included: (1) ADHD category, (2) child’s gender, (3) child’s age, (4) child’s race/ethnicity, (5) the respondent’s relation to child, (6) insurance status, (7) single or two-parent household, (8) number of times moved, (9) federal poverty status level, (10) mother or father’s mental health (binary variable based on parental report, 1 = ‘excellent’ or ‘good’, 0 = ‘fair’ or ‘poor’), and (11) caregiver strain. Caregiver strain was a composite score created from questions about: (1) how well parents can talk and share ideas, (2) how well they are coping with parental demands, (3) how often they feel their child is much harder to care for than other children, (4) how often they feel that the child does things that bother them a lot, and (5) how often they felt angry with their child. Each variable is dichotomous (1 = yes, 0 = no). The composite score represents a sum of the dichotomized positive (1 = yes) and negative (0 = no) responses to each question. Scores ranged from 0 to 5, whereby higher scores indicate greater caregiver strain (Cronbach α = .73).

Multivariate analyses were then used to measure the effect of having a medical home and the four ADHD categories on the four child health outcomes of interest: (1) days missed from school, (2) number of preventive care visits, (3) receipt of mental health treatment, and (4) unmet health care need. Two of the outcome variables (i.e., unmet health care need and receipt of mental health treatment) were also dichotomized (1 = yes, 0 = no) dependent variables and set equal to one if the outcome was met, and zero otherwise.

Regression models specific for nominal data were used to estimate the effects of medical home and ADHD category on number of days missed from school and number of preventive care visits. Because the dependent variables exhibit overdispersion, negative binomial regressions were performed. STATA was used to conduct the analyses [15].

Results

Sample Characteristics

Table 1 summarizes the sample characteristics of the parents of children with ADHD who participated in the NSCH survey. Most children in the sample had ADHD without comorbidities (49.4%), were White (73.2%), and male (70.9%). Most were two parent homes (50.6%), from families with income > 400% of the federal poverty status level (33.7%), and mothers (62.2%) and fathers (70.5%) with excellent or good mental health. Mean age of the children in the sample was 12.75 years (SD = 3.43) and average caregiver strain was 2.13 (SD = 1.61).
Table 1

Sample characteristics as reported by parents of children with ADHD in the NSCH (N = 5,495)

Variable

Sample characteristics

%

n

ADHD category

ADHD alone

49.4

2,712

ADHD plus mental diagnosis

18.1

996

ADHD plus physical diagnosis

16.4

900

ADHD plus mental plus physical

16.1

887

Race/ethnicity

Hispanic

8.2

445

White non-hispanic

73.2

3,972

Black non-hispanic

10.4

560

Other

8.2

442

Gender of child

Male

70.9

3,897

Female

29.1

1,594

Household type

Single parent

49.4

2,701

Two parent

50.6

2,762

Poverty status

0–99%

15.7

863

100–199%

18.2

998

200–399%

32.4

1,780

≥ 400%

33.7

1,854

Relationship of respondent

Mother

74.6

4,104

Father

15.4

844

Other

10.0

547

Mother’s mental health

Excellent/good

62.2

3,036

Fair/poor

37.8

1,849

Father’s mental health

Excellent/good

70.5

2,607

Fair/poor

29.5

1,093

Insurance coverage

Insured

95.3

5,232

Uninsured

4.7

257

Born in the USA

Yes

99.0

5,403

No

1.0

52

 

Mean (Standard Deviation (SD)

Age of child

12.75 (3.43)

Number of times moved

2.43 (2.51)

Caregiver strain

2.13 (1.61)

Frequency of Children with ADHD Who Have a Medical Home

Table 2 shows that about half of children with ADHD alone reported having a medical home, but that frequency decreased with mental health comorbidities. Interestingly, more children with ADHD plus physical diagnosis (49.7%) had a medical home than children with ADHD alone (48.2%) and with children with ADHD plus mental diagnosis. The same trends were consistently seen in all of the five sub-components of medical home whereby children with ADHD alone or ADHD plus physical diagnosis had a higher frequency of meeting the sub-component than children with ADHD plus mental diagnoses.
Table 2

Frequency of children with ADHD who have a medical home and those who meet the medical home sub-components

 

Has a medical home

Meets medical home sub-component

Has personal doctor

Has a usual source of care

Has family centered care

Has no problems getting referrals

Has effective care coordination

ADHD category

ADHD alone

48.2

94.9

96.1

67.3

76.1

56.0

ADHD plus mental diagnosis

36.1

93.2

95.6

59.4

73.6

47.6

ADHD plus physical diagnosis

49.7

95.7

96.5

70.7

79.5

58.6

ADHD plus mental plus physical

30.1

94.6

93.9

54.1

68.8

42.5

Factors Associated with Having a Medical Home

Results from the multivariate logistic regressions for having a medical home or meeting the five sub-components are presented in Table 3. The data show that children with ADHD plus a physical, mental, or combined diagnosis were 24–63% less likely than children with ADHD alone to have a medical home (Odds Ratio (OR) = .76, .54, and .37, respectively). Results for the five sub-components indicated that children with ADHD plus mental diagnosis were significantly less likely to have family-centered care or effective care coordination than children with ADHD alone. Children with ADHD plus both physical and mental diagnoses were 43% more likely to have problems getting referrals versus children with ADHD alone.
Table 3

Odds ratios of having a medical home and medical home sub-components

Variable

Has a medical home

Medical home sub-component

Has a personal doctor

Has a usual source of care

Has family centered care

Has no problems getting referrals

Has effective care coordination

Odds ratio

Odds ratio

Odds ratio

Odds ratio

Odds ratio

Odds ratio

ADHD category

ADHD plus mental diagnosis

.54**

.84

1.35

.74*

.69

.64**

ADHD plus physical diagnosis

.76**

1.02

1.25

1.05

1.00

.83

ADHD plus mental plus physical

.37**

.89

.91

.55**

.57*

.47**

Race/ethnicity

Hispanic

.72*

.96

.47*

.43**

.55*

.75

Black non-hispanic

.42**

.48*

.31**

.81

.78

.64*

Other

1.03

.91

1.61

1.05

.75

.88

Gender of child

Female

.99

1.03

1.49

1.06

1.10

1.10

Household type

      

Two parent

1.02

1.23

.93

1.04

1.18

.98

Poverty status

0–99%

1.03

1.18

.59

.82

1.04

1.21

100–199%

1.04

.90

.73

.83

.79

1.07

≥400%

.93

1.61*

2.68**

1.21*

.73

.85

Relationship of respondent

Father

.77*

.97

.50*

.81*

1.10

.84

Other

.33*

.74

.77

.43*

.38

.29*

Mother’s mental health

Excellent/good

1.18*

.94

1.45

1.19*

1.22

1.22

Father’s mental health

Excellent/good

1.50**

1.18

1.18

1.30*

1.27

1.39

Insurance coverage

Insured

1.51*

3.22**

2.58*

1.55*

.99

1.20

Born in the USA

Yes

1.09

1.00

1.33

1.68

1.44

.83

Age of child

.98

1.00

.97

.98

1.00

1.00

Number of times moved

.98

.91*

.94

.97

.98

.95*

Caregiver strain

1.25**

1.00

1.04

1.16**

1.08

1.28**

** P < .005; * P < .05. Referent groups: ADHD alone, white non-hispanic, male, single parent, 200–399% poverty status, maternal respondent, fair/poor mother and father mental health

These multivariate results also show that several other, non-clinical factors were associated with having a medical home or meeting the five sub-components. The most consistent associations were from race/ethnicity (non-Whites were least likely to have a medical home or meet the five sub-components), insurance status (having insurance was associated with having a medical home and meeting the five sub-components), caregiver strain (higher levels of strain were associated with having a medical home and meeting the five sub-components), and relationship of respondent to child (non-maternal respondents were less likely to report their child having a medical home or meeting the five sub-components).

Effect of Having a Medical Home on Four Child Health Outcomes

Table 4 presents the estimated odds ratios and coefficient parameters for regressions for four child health outcomes. An odds ratio of 1 indicates that an event (e.g., having an unmet health care need) is equally likely to occur in a referent group (e.g., ADHD diagnosis category). An odds ratio greater than 1 indicates that an event is more likely to occur in a referent group whereas an odds ratio less than 1 indicates that an event is less likely to occur in a referent group. Results from the first logistic regression show that having a medical home was associated with being less likely to have an unmet health care need (OR = .24, P < .005). Having ADHD plus any other comorbidity (i.e., mental, physical, or both) resulted in being 1.8–2.8 (P < .005) times more likely to have reported an unmet health care need compared to those with ADHD alone. Results from a second logistic regression show that having a medical home was associated with being less likely to have received needed mental health care (OR = .64, P < .005). Having ADHD plus any other mental health comorbidity resulted in being approximately 2.5–3 times more likely to have received needed mental health care. Beyond the medical home and ADHD categorical variables, several other patterns were consistently realized in the logistic regressions. For example, paternal respondents and being insured were associated with positive outcomes (i.e., lower odds of unmet needs and greater odds of needed mental health care) across both regressions.
Table 4

Logistic and negative binomial regressions results for four child health outcomes

Variable

Child health outcomes

Has an unmet health need

Received mental health care

Number of school days missed

Number of preventative care visits

Odds ratio

Odds ratio

Coefficient

Coefficient

Has a medical home

    

Yes

.24**

.64**

−.24**

−.02

ADHD category

ADHD plus mental diagnosis

1.80**

2.45**

.25**

.28**

ADHD plus physical diagnosis

1.76**

1.14

.37**

.25**

ADHD plus mental plus physical

2.79**

3.09**

.47**

.44**

Race/ethnicity

Hispanic

1.38

.92

−.04

.00

Black non-hispanic

1.02

.74

−.28*

−.14

Other

1.53*

1.33

−.10

−.15*

Gender of child

Female

1.10

1.08

.17**

.10*

Household type

Two parent

1.36*

1.07

.17**

.17**

Poverty status

0–99%

.91

.91

−.06

.23*

100–199%

1.13

.81

−.05

.06

≥400%

.77

1.31**

−.13*

−.11*

Relationship of respondent

Father

.64*

1.25*

−.14*

.04

Other

.96

.64

−.09

.09

Mother’s mental health

Excellent/good

.77

.81*

−.15**

.04

Father’s mental health

Excellent/good

.68

1.00

−.02

−.09

Insurance coverage

Insured

.53

1.94**

.02

.08

Born in the USA

Yes

.90

1.00

.08

−.01

Age of child

.97

.97

.03**

−.02**

Number of times moved

1.11**

1.05**

.04**

.01

Caregiver strain

.93

.81**

−.01

.65

Model type

Logistic

Logistic

Negative binomial

Negative binomial

** P < .005; * P < .05. Referent groups: ADHD alone, white non-hispanic, male, single parent, 200–399% poverty status, maternal respondent, fair/poor mother and father mental health

Results from the negative binomial regressions yielded important nuances in the care that children with ADHD receive. While having a medical home was associated with 24% fewer school days missed, interestingly there was no association between having a medical home and the number of preventive care visits. Not surprisingly, however, children with ADHD plus any other comorbidity were associated with 25–47% greater number of school days missed and 25–44% more preventive care visits.

In addition to the medical home and ADHD categorical variables, several other patterns were realized across the two negative binomial regressions. For instance, black and non-Hispanic children, those who are ≥400% of the federal poverty status level, those who have mothers with excellent/very good mental health, and those with paternal respondents were associated with a fewer number of school days missed. On the other hand, older and female children, those who moved more times, and children with two-parent households were associated with a greater number of school days missed. Additionally, older children, those who are ≥400% of the federal poverty status level, and children of ‘other’ races were associated with fewer preventive health care visits. Lastly, female children, those who have a two-parent household, and those who are 0–99% of the federal poverty status level were associated with greater numbers of preventive health care visits.

Discussion

This study of children with ADHD investigated the factors associated with having a medical home and the five NSCH associated sub-components, as well as, the relationships between having a medical home and four child health outcome measures. While the medical home model may help attain improved health, it is increasingly important to understand how it serves children with specific and varying needs. A novel contribution of our study is that we controlled for, and can comment on, the differential effects of, physical and mental comorbidities for children with ADHD. Results of this study contribute to the literature in four ways.

First, we were able to enumerate the frequency of children with ADHD who have comorbidities. Although the literature is clear that children with ADHD often have comorbidities, our analysis suggests that more than one-half of all the children with ADHD have an additional physical diagnosis. Moreover, 35% of those children also have a mental health diagnosis. These results have implications for clinical practice patterns. Given that the majority of children with ADHD are managed by their primary care physician, the high frequencies of comorbidities implies that there are additional competing demands on available resources to be able to manage their care [16]. Future research should investigate if specific programs for CSHCN or ADHD Outpatient Clinics would offer more efficient models for providing care and would better reach the goals of being an ADHD medical home [17]. Future studies should also further investigate the complex issue of whether specialists should function as an extension of the medical home model.

Second, our results on the frequency of children with ADHD who have a medical home, or meet the five sub-components of a medical home, corroborate and expand the literature. In our study we found that 48% of children with ADHD alone have a medical home. Comparing this percentage to all children in the NSCH who have a medical home (57.5%) and all children in the 2005–2006 National Survey of Children with Special Health Care Needs (NS-CSHCN) who have a medical home (47.1%), our results demonstrate that children with ADHD are similar to the general population of CSHCN [2]. Crabtree et al. note that there is significant fragmentation between the physical and mental health care systems that is not always well addressed in medical home models [18, 19]. This separation undermines core values of the medical home model and can lead to poor health outcomes. Our results also expand the literature in that we show that having additional comorbidities affect having a medical home and meeting the five sub-components. Having ADHD plus a physical diagnosis actually resulted in better outcomes than mental diagnosis or having ADHD alone. Finally, our results suggest that family-centered care and effective care coordination are the two of the five sub-components that are most lacking for children with ADHD.

Third, our regression analyses on the factors that affect having a medical home or meeting the five sub-components are comprehensive. We found that children with ADHD plus any comorbidity (i.e., physical, mental, or both) were less likely to have a medical home when adjusting for several child and household factors. Toomey et al.’s [12] study, which used data from the 2005 to 2006 NS-CSCHN, found that children with ADHD were less likely to have a medical home than children with asthma [12]. It is difficult to determine if our results confirm or contradict this findings since we included asthma in our list of comorbid physical conditions, but we did not focus specifically on the physical diagnosis of asthma. Future research should compare results by individual physical diagnoses and mental diagnoses. Our regression results also suggest that children with ADHD and a mental health diagnosis are less likely to receive effective care coordination. Care coordination requires a single locus of planning and coordination, which may prove difficult for children with multiple mental health diagnoses. Perhaps care coordination is not being provided due to its expense. Stille and colleagues [20] note that for a medium size practice the estimated costs of care coordination are $22,000–$33,000 [20]. Thus, policies aimed at improving care coordination for these children may need to include an enhanced payment to cover these costs [21]. More research is needed to understand our finding that having a medical home is associated with lower odds of having received mental health care. Perhaps this refers to mental health received outside of the primary care setting.

Fourth, our study adds to the evidence on the effect of medical homes on child health outcomes. None of the 33 studies in Homer’s et al. [11] review focused on children with ADHD or other mental health diagnoses. Our results somewhat corroborate those in the literature in that children with a medical home are more likely to have their needs met [22]. We found that children with ADHD and a mental health comorbidity were approximately more than 2.5–3 times likely to have received needed mental health care and were associated with 28–44% more preventive care visits. These results suggest that regardless of the model of care, positive outcomes are being achieved. Finally, our results corroborate others who have found that having a medical home, or the five sub-components of a medical home, are associated with fewer missed school days [23]. However, children with ADHD plus mental and physical diagnoses were associated with more missed school days.

Several study limitations merit attention. First, there are other sub-components, of a medical home other than the five employed in the NSCH that are not addressed in the dataset. Cultural sensitivity and compassionate care are key components of the AAP’s definition of a medical home that are not included in the NSCH. Second, there are other important child health outcomes such as immunizations or weight management that were not considered. Third, we cannot comment on how the severity of illness might affect the results. For example, a child with ADHD plus a speech problem might face different barriers to having a medical home than a child with ADHD plus epilepsy. Fourth, there are certainly many other physical and mental comorbid diagnoses that were not asked about on the NSCH. Moreover, survey answers cannot shed light on whether or not the child’s ADHD is the primary diagnosis. Fifth, the medical home construct used in the NSCH is parent-reported which may affect the implications. For example, if a parent perceives an unmet need which could be due to access at a systematic level, and not the practice level, he or she may also rate other components of the medical home lower (e.g., care coordination).

Conclusion

Despite these limitations, our findings on the factors associated with a medical home for children with ADHD corroborate and extend those of prior studies on CHSCN. We found that although children with ADHD have a similar frequency of having a medical home as other CSHCN, when a comorbidity is factored into the equation, this frequency falls. Particularly, children with ADHD and mental health comorbidities should be targeted in interventions designed to improve family-centered care and effective care coordination. Results from demonstration projects that solely focus on children with mental health diagnoses, such as the ABCD initiatives currently underway in Illinois, should help to understand how their experiences differ from other CSHCN [24]. Finally, our results suggest that having a medical home is associated with positive child health outcomes such as fewer missed school days and unmet health care needs. It is expected that the expansion of the medical home model will only become more pronounced under the Affordable Care Act. As a result, studies such as ours, which focus on children with diverse illnesses, will become increasingly important.

Conflict of interest

None of the authors have a conflict of interest.

Copyright information

© Springer Science+Business Media, LLC 2011