Admitting what is needed: How the health system and society can reduce hospitalizations for children with medical complexity

THE PROBLEM

Hospitalization is a relatively common occurrence for patients with medical complexity. In fact, high health care use is a defining feature of children with medical complexity (CMC). ¹ CMC are a heterogeneous and growing population, unified by conceptual similarities: multisystem and frequently progressive chronic conditions leading to substantial functional limitations, family‐identified service needs, and high healthcare use. ¹ Annually, approximately 13%–20% of CMC are hospitalized ² , ³ and 50%–80% of CMC healthcare spending goes to hospital care. ¹ , ³ , ⁴ However, some hospitalizations experienced by CMC are avoidable. ⁵ Substantial variation in hospital utilization for CMC across regions, ⁶ conditions, ³ and emergency departments ⁷ suggests opportunities surround us.

Avoiding hospitalization for CMC is beneficial for most stakeholders. Fewer hospitalizations are in the interest of payers and aligned with those health systems engaged in risk‐sharing or capitated arrangements for CMC populations. For families, hospital admissions are extremely challenging periods. In this study, we—a hospitalist, a parent of a child with medical complexity, and a complex care pediatrician—review causes of hospitalization and propose recommendations to reduce hospitalizations for CMC (Table 1).

Table 1.

Recommendations to reduce hospitalizations for children with medical complexity

Shifting the focus of clinical care

1. Identify clinical quality measures that are linked to hospitalization

2. Design and test interventions with family partners that strengthen family capacity and address child susceptibility to serious acute illness

3. Quantify and reduce inequities in access, retention, and benefits from complex care models for diverse populations

4. Uncover active ingredients of complex care that lead to reductions in hospitalization and develop strategies to implement these activities in general outpatient settings

5. Use conceptual models and include analyses of intervention mechanisms in research

Needed social and policy focus

6. Evaluate the impact of policies to support family caregivers on hospitalization rates

7. Advocate for policies that support care for children with medical complexity at home and in the community, including adequate quantities of supplies/equipment, respite, and reimbursement/workforce to reliably meet family needs

Important unknowns

8. Develop adaptive interventions that can respond to the dynamic, changing nature of child and family circumstances

9. Establish an evidence base to treat common causes of acute illness or chronic disease exacerbation through rigorous clinical intervention research

10. Clarify complex care's vision, for example, as a specialty service for those with medical complexity to access, or as a set of effective activities implemented broadly throughout redesigned pediatric care

NOT SIMPLY MORE CLINICAL CARE

We might assume that the clinical needs intrinsic to medical complexity mean that clinical care drives CMC outcomes (such as hospitalization) to a greater extent than for general populations, where clinical care explains only 20% of population health. ⁸ However, clinical care might, perhaps counterintuitively, play a similarly limited role in determining outcomes even for CMC.

For example, common healthcare quality measures can have weak or limited relationships with hospitalization risk for CMC. Family‐reported measures of ambulatory care access, experience, and process appear to poorly predict subsequent hospitalizations for CMC, even though several of these measures predict hospitalizations for children with less complex chronic illnesses. ² Ambulatory care sensitive conditions (ACSC, e.g., asthma, pneumonia, urinary tract infection) are generally considered manageable through high‐quality outpatient care and therefore indicate potentially preventable hospitalization. However, research involving CMC with ACSC hospitalizations challenges this conceptualization. For example, ambulatory care measures may be poor predictors of ACSC hospitalizations for CMC despite predicting ACSC hospitalizations for children with noncomplex chronic illness. ⁹ Receipt of outpatient care in the week prior to admission is similar for CMC whether admissions are for ACSC or not. ¹⁰ Measuring different ACSCs selected for CMC (such as constipation, dystonia, medical device complications, etc.) might reveal more relationships between clinical care and hospitalization.

SHIFTING THE FOCUS OF CLINICAL CARE

We suspect, however, that when linking healthcare quality to hospitalizations for CMC, we may simply need to focus on different constructs (Recommendation 1). When families of CMC describe experiences leading up to hospitalization, they identify health system factors, child susceptibility to serious illness, and family capacity and resources as key determinants of hospitalization. ¹¹ Our interpretation is that family capacity and resources along with child susceptibility, likely play outsized roles in determining CMC hospitalization risk relative to conventional health system factors; these require deliberate attention (Recommendation 2). Traditional outpatient practices are often ill‐equipped to powerfully influence such determinants without paradigm shifts.

A growing body of research underscores how innovative clinical interventions can, in fact, reduce CMC hospitalizations. Complex care programs combine a rich tapestry of care coordination and medical care with family‐centered and longitudinal multidisciplinary relationships built on trust. ¹² These programs plausibly affect both child susceptibility to serious illness and family capacity to a greater extent than a typical general practice because of a combination of greater time, multidisciplinary staffing, and acquired expertise. However, despite evidence demonstrating fewer hospitalizations after enrollment into complex care programs, ¹³ , ¹⁴ we still do not understand how they exert this impact. ¹⁵ Moreover, data are needed to quantify inequities in access to, retention in, or benefits from complex care enrollment (Recommendation 3). As it is unclear if complex care can be practically scaled to reach all CMC, understanding how these programs reduce hospitalizations, that is, what the mechanisms and key ingredients are, is necessary to translate effective activities within non‐complex care settings ¹³ (Recommendations 4 and 5). Longitudinal mixed‐methods research could generate a theoretically sound conceptual model of mechanisms through which complex care lowers hospitalizations. As one example, an innovative design might use ecological momentary assessment ¹⁶ combined with ethnography among families of children and clinicians who are and are not in complex care programs to study the phenomenon of hospitalization. Causal pathways hypothesized from the model could be tested using hybrid effectiveness‐implementation, ¹⁷ quasi‐experimental, ¹⁸ or even rigorous quality improvement designs ¹⁹ as new programs are started or as children are enrolled into established programs.

Emerging experimental evidence highlights several promising clinical activities, such as home visiting, ²⁰ , ²¹ tailored health crisis planning, ²⁰ and digital interventions to monitor and manage health at home. ²² , ²³ CMC families may be able to predict when hospitalization is imminent, ²⁴ , ²⁵ and harnessing this insight might allow focused interventions to be delivered during critical periods when an illness trajectory remains modifiable. Each of these interventions can provide windows into what causes hospitalization, and which constructs should be measured to judge care quality in the future (Recommendations 1 and 5).

NEEDED SOCIAL AND POLICY FOCUS

Clinical innovations likely offer only part of the solution, and we believe social and policy investments are important complements to eliminating avoidable hospitalizations for CMC. In fact, avoiding the hospital in lieu of care at home for people with medical complexity was a cornerstone of landmark bipartisan legislation in the 1980s and 1990s to expand access to Medicaid home and community‐based services (HCBS) waivers for long‐term services and supports for individuals with disabilities. At that time, it was inconceivable and financially impossible to have a child at home with a mechanical ventilator. However, through advocacy from Katie Beckett's and other courageous families, what can be managed by families at home has changed dramatically. Arguably, a societal obligation exists to appropriately support the creation of sophisticated health systems in homes of CMC through policy, funding, and services.

Amidst the many positives that come with caring for CMC at home, the challenges caregivers face can be overwhelming. In addition to the traditional demands of parenting, families of CMC manage complex health systems. ²⁶ Many must forego employment ²⁷ and most deliver unsafe quantities of health care to CMC (average > 50 h/week, with many delivering around‐the‐clock care). ²⁸ The supply of support personnel needed for reliable daily CMC care is inadequate, even when approved by insurance or mandated through policy. ²⁹ This includes frequent turnover and massive workforce shortages for all home health workers, such as respite, private duty nurses, certified nursing assistants, and personal care assistants. Only 3% of CMC insured by Medicaid receive formal, professional home health support services to facilitate living at home. ³ Among patients needing to establish home care nursing, nursing shortages have been observed to account for 27 excess hospital days per patient. ³⁰ The RAISE (Recognize, Assist, Include, Support and Engage) Family Caregiver Act, passed in 2018, directs Health and Human Services to develop a national family caregiving strategy, including respite options, care coordination, and financial security. Newer proposed policies such as the HCBS Access Act would make HCBS mandatory via Medicaid, thus protecting these services from being removed at the state level. Paying trained family caregivers, for example, as certified nursing assistants, ³¹ is a necessary strategy, but not sufficient to solve the critical workforce shortage facing all individuals with disabilities.

Although common sense suggests that addressing these well‐recognized challenges of CMC caregiving at home would likely prevent hospitalizations and improve other important outcomes, actual research evidence is sparse. Among children with autism spectrum disorders, HCBS waivers have been associated with fewer unmet needs. ³² Among older adults, HCBS appears associated with fewer hospitalizations, ³³ and removal of HCBS has been associated with increased hospitalization rates. ³⁴ Additional policy evaluations should link CMC hospitalization rates with policy implementation across states. Future studies could apply new measures of pediatric home care quality ³⁵ to identify associations between quality and hospitalization rates. Such data can help strengthen financial arguments proposed to offset the policy investments ³ , ²⁹ (Recommendations 6 and 7).

Importantly, CMC experience substantially greater negative social determinants of health than non‐CMC. For example, 51% live in families with income less than 200% of the federal poverty level. ³⁶ While caring for CMC is difficult for any family, it is substantially more difficult in the face of poverty, racism, housing instability, adverse childhood experiences, mental illness, and other challenges. Undoubtedly, these social determinants influence child susceptibility to acute illness and family capacity and resources. Larger studies of hospital utilization illustrate crude associations with demographics, including race/ethnicity or public insurance. ³⁷ CMC living in communities with greater resources, such as access to health care, clean air, quality schools and safe housing spend less time in the emergency department and fewer days hospitalized. ³⁸ The influence of more specific social determinants of health and hospitalizations for CMC comes primarily from smaller studies, often among CMC in complex care programs, making associations harder to interpret. ³⁹ , ⁴⁰ Avoidable hospitalizations are a likely consequence, but confirmatory research is limited.

IMPORTANT UNKNOWNS

There are additional important unknowns to eliminating avoidable hospitalizations for CMC. First, the population of CMC is heterogeneous, and there are presumably inconsistent intervention effects, where effectiveness in one subgroup may not translate to effectiveness in another subgroup. Second, hospital risk for CMC is dynamic. One difficult year may not predict another, ⁴¹  and understanding how to tailor interventions to match changing risks is essential (Recommendation 8). Third, evidence guiding clinical treatment for common conditions prompting CMC hospitalization is scant. Clinical trials to identify effective treatments for high‐priority clinical conditions facing CMC are needed, and hospitalization may often be an appropriate trial outcome (Recommendation 9). Finally, there is an opportunity to clarify the national vision for pediatric complex care, which begs many questions, including whether we aspire for all CMC to have the opportunity to access this specialized model or whether we aspire to have a future where all pediatric care integrates the most meaningful aspects learned from specialized complex care models (Recommendation 10).

CONCLUSION

Although CMC are a relatively small pediatric population, this topic has broad relevance. Disability and medical complexity can happen to any individual at any time in their lives. Interventions that replace avoidable hospitalizations with highly supportive home services likely improve other meaningful clinical and social outcomes, and may translate to individuals with less complex chronic illnesses. Robust home and community support will likely lead to fewer hospitalizations and broader societal benefits.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Chen LP, Gerber DM, Coller RJ. Admitting what is needed: How the health system and society can reduce hospitalizations for children with medical complexity. J Hosp Med. 2023;18:90‐94. 10.1002/jhm.12948