Associations between comorbidity-related functional limitations and pneumonia outcomes

Abstract

Background:

Underlying comorbidities are common in children with pneumonia.

Objective:

To determine associations between comorbidity-related functional limitations and risk for severe pneumonia outcomes.

Design, Setting, and Participants:

We prospectively enrolled children <18 years with and without comorbidities presenting to the emergency department with clinical and radiographic pneumonia at two institutions. Comorbidities included chronic conditions requiring daily medications, frequent healthcare visits, or which limited age-appropriate activities. Among children with comorbidities, functional limitations were defined as none or mild, moderate, and severe.

Main Outcomes and Measures:

Outcomes included an ordinal severity outcome, categorized as very severe (mechanical ventilation, shock, or death), severe (intensive care without very severe features), moderate (hospitalization without severe features), or mild (discharged home), and length of stay (LOS). Multivariable ordinal logistic regression was used to examine associations between comorbidity-related functional limitations and outcomes, while accounting for relevant covariates.

Results:

A cohort of 1116 children, including 452 (40.5%) with comorbidities; 200 (44.2%) had none or mild functional limitations, 93 (20.6%) moderate, and 159 (35.2%) had severe limitations. In multivariable analysis, comorbidity-related functional limitations were associated with the ordinal severity outcome and LOS (p < .001 for both). Children with severe functional limitations had tripling of the odds of a more severe ordinal (adjusted odds ratio [aOR]: 3.01, 95% confidence interval [2.05, 4.43]) and quadrupling of the odds for longer LOS (aOR: 4.72 [3.33, 6.70]) as compared to children without comorbidities.

Conclusion:

Comorbidity-related functional limitations are important predictors of disease outcomes in children with pneumonia. Consideration of functional limitations, rather than the presence of comorbidity alone, is critical when assessing risk of severe outcomes.

INTRODUCTION

Pneumonia is a common reason children present to the Emergency Department (ED) and one of the most common reasons for hospitalizations in the United States.¹ Children with comorbidities, including asthma, other chronic lung diseases, cardiovascular disease, neurological disorders, and others are at increased risk for pneumonia and poor clinical outcomes.² These children have longer hospitalizations, more frequent need for intensive care, and more often suffer severe respiratory complications, as compared to otherwise healthy children. The clinical impact of individual comorbidities, however, is highly variable, and characterizations based solely on the presence or absence of comorbidities may be less meaningful at the bedside. For example, a child with cerebral palsy may be severely affected with spastic quadriplegia, feeding and swallowing difficulties, and chronic respiratory problems, or that child may be minimally affected with little to no increased respiratory morbidity. When challenged by acute pneumonia, the former is likely predisposed to a more severe illness course and is at higher risk for less common pathogens, whereas pneumonia in the less severely affected child may be more similar to a child with no comorbidities.

As an alternative, comorbidity-related functional limitations depict to what degree a comorbidity may negatively influence a child’s activities of daily living. A better understanding of an individual’s functional limitations, therefore, may be a better predictor of disease outcomes in children with pneumonia as compared to the presence of comorbidities alone. However, we are unaware of prior studies examining associations between functional limitations and disease outcomes for children with pneumonia. Using prospectively collected data from children with pneumonia presenting for emergency care at two US children’s hospitals, we sought to define associations between the presence and degree of comorbidity-related functional limitations and acute pneumonia outcomes.

METHODS

Study design and participants

This analysis was nested within a prospective observational cohort of 1116 children with community-acquired pneumonia presenting for emergency care at two US children’s hospitals. Study participants included children <18 years of age presenting to the ED at the Monroe Carell Jr. Children’s Hospital at Vanderbilt (Nashville, TN) or Primary Children’s Hospital at the University of Utah (Salt Lake City, UT). Children were enrolled during a pilot phase at Vanderbilt (December 2014 to January 2017) and later across both institutions (September 2017 to May 2019) using identical enrollment criteria. Similar to other pneumonia studies,^3,4 children presenting with acute signs and/or symptoms of infection (e.g., fever), AND respiratory illness (e.g., cough, tachypnea), AND radiographic evidence of pneumonia were considered eligible for inclusion. Radiographic pneumonia was determined based on the clinical radiology report and confirmed by the principal investigator at each site. When radiographic evidence of pneumonia was equivocal, only those cases with clinical documentation supporting a diagnosis of pneumonia were included. The parent study was focused on community-acquired pneumonia and excluded children if they had a tracheostomy, cystic fibrosis, severe immunosuppression, hospitalization within the preceding 7 days, or had a clear alternative diagnosis. After informed consent was obtained, children and/or their caregivers were interviewed to obtain the child’s past medical history (including comorbidities and functional limitations), history of present illness, and other pertinent information using a standardized questionnaire. Following discharge, a chart review was performed to collect details of the hospital course, including receipt of antibiotics for the current illness, complications, and outcomes. The study was approved by each institution’s Human Research Protections Program.

Comorbidity and functional limitations

The Centers for Disease Control and Prevention defines chronic diseases as, “conditions that last 1 year or more and require ongoing medical attention or limit activities of daily living or both.”⁵ For our study, comorbidities were operationally defined as any condition present or expected to be present for 12 months or more and for which any of the following were also true: (1) required daily medications; (2) required contact with specialty health care providers or frequent contact with a primary health care provider; or (3) limited the ability to perform age-appropriate activities or ability to interact with peers of the same age at least some of the time. With the exception of asthma, comorbidities were grouped by body system for reporting; those present in <5% of the cohort were collapsed into “Other.”

Data regarding functional limitations were ascertained using two items from the 2009 to 2010 National Survey of Children and Youth with Special Healthcare Needs⁶: “During the past 12 months (or since birth if <1 year), how often have your child’s health conditions affected (his/her) ability to do things other children (his/her) age do?” and “Do these conditions affect (his/her) ability to do things a great deal, some, or very little?” We postulated these items chosen best reflected the presence and extent of global functional limitations.

The primary exposure for analysis was comorbidity-related functional limitations, a single categorical variable that combines the presence of comorbidities and associated functional limitations into four mutually exclusive groups. Children with no comorbidity served as the reference. Children with comorbidities were classified according to their functional status: no or mild, moderate, and severe. No or mild functional limitations included children whose caregiver reported their comorbidities never affected their abilities and those whose comorbidities sometimes, usually, or always affected their abilities but to a very little extent; moderate included children whose abilities were sometimes, usually, or always affected to some extent and those whose abilities were sometimes affected a great deal; severe included children whose abilities were usually or always affected a great deal.

Outcomes

The primary outcome was a four-level ordinal disease severity measure based upon the most severe outcome experienced during the index encounter (mild, moderate, severe, and very severe). Children discharged from the ED were classified as mild. Hospitalized children not requiring intensive care were considered moderate. Those requiring intensive care were classified as severe or very severe with the latter consisting of those children who required invasive mechanical ventilation, developed shock requiring vasoactive medications, or died. We also evaluated length of stay (LOS), measured in hours from the time of ED triage to discharge from the ED or hospital, as a secondary outcome.

Statistical analysis

Descriptive statistics were provided as frequencies (proportions) for categorical variables and medians (interquartile ranges) for continuous variables. Kruskal–Wallis and Pearson’s χ² tests were used to compare baseline characteristics across the four mutually exclusive groups representing comorbidity-related functional limitations. Two separate models were used to evaluate associations between comorbidity-related functional limitations and our ordinal disease severity outcome as well as hospital LOS (in hours) using multivariable ordinal logistic regression. This approach is useful for modeling discrete ordinal outcomes as well as continuous outcomes (continuous data may also be considered ordinal).^7,8 Similar to binary and linear regression models, ordinal logistic regression estimates a single β-coefficient or odds ratio for each parameter meaning that the change in odds associated with a one-unit change in any given predictor variable is the same at each level of the outcome scale. In our analysis, each model was adjusted for age (in months), gender, race (non-Hispanic Black, non-Hispanic White, Hispanic, Asian, other), insurance (public, private, both, none), presence of parapneumonic effusion requiring drainage, and distant site of infection. Restricted cubic spline functions were applied on age to allow for nonlinear associations. The proportional odds assumption was assessed graphically and confirmed as valid for both models (data not shown).⁷ We also conducted sensitivity analyses for both models whereby children with baseline home oxygen therapy or noninvasive ventilatory support (which may necessitate higher levels of care regardless of illness severity; n = 39), and those in whom pneumonia was not the primary reason for admission (e.g., a child with diabetic ketoacidosis triggered by acute pneumonia; n = 61) were excluded. Finally, we also evaluated two alternative models for each outcome that excluded functional limitations but included comorbidities (yes/no or 0, 1, 2, 3+). The original models outperformed each of these models as indicated by a lower Akaike information criterion (data not shown). All analyses were performed using R version 4.0.3 with a two-sided p < .05 indicating statistical significance.

RESULTS

Study population

The study enrolled 1116 children (median age: 33 months), including 452 (40.5%) with one or more comorbidities. Children with comorbidities were older, more likely to identify as male and non-Hispanic White, and to have public insurance than children without comorbidities (Table 1). Children with comorbidities were also more likely to receive antibiotics. The most common comorbidities included asthma (13.7%), other pulmonary disorders (7.5%), neurologic disorders (12.6%), cardiovascular disorders (8.2%), and metabolic disorders (8.5%).

TABLE 1.

Baseline characteristics

			Comorbidity
	All participants (N = 1116)	No comorbidity (N = 664)	No or mild functional limitation (N = 200)	Moderate functional limitation (N = 200)	Severe functional limitation (N = 159)	p value
Age (months)	43.0 [19.0, 92.0]	33.0 [16.0, 71.0]	55.5 [26.0, 94.2]	62.0 [27.0, 112.0]	96.0 [36.0, 162.0]	<.001
Prematurity						<.001
Yes	18.4% (204)	13.9% (92)	25.3% (50)	27.2% (25)	23.6% (37)
No. of comorbidities						<.001
0	59.5% (664)	100.0% (664)	0.0% (0)	0.0% (0)	0.0% (0)
1	21.9% (244)	0.0% (0)	74.0% (148)	60.2% (56)	25.2% (40)
2	9.7% (108)	0.0% (0)	17.0% (34)	23.7% (22)	32.7% (52)
3+	9.0% (100)	0.0% (0)	9.0% (18)	16.1% (15)	42.1% (67)
Sex						.327
Male	53.7% (599)	54.4% (361)	53.5% (107)	59.1% (55)	47.8% (76)
Female	46.3% (517)	45.6% (303)	46.5% (93)	40.9% (38)	52.2% (83)
Race						<.001
Non-Hispanic Black	11.6% (130)	9.8% (65)	14.5% (29)	19.4% (18)	11.3% (18)
Non-Hispanic White	63.8% (712)	60.4% (401)	67.5% (135)	63.4% (59)	73.6% (117)
Hispanic	16.9% (189)	19.9% (132)	12.0% (24)	14.0% (13)	12.6% (20)
Asian	2.0% (22)	2.4% (16)	2.5% (5)	0.0% (0)	0.6% (1)
Other	5.6% (63)	7.5% (50)	3.5% (7)	3.2% (3)	1.9% (3)
Insurance						<.001
Public	36.5% (395)	30.7% (197)	36.8% (71)	41.8% (38)	56.7% (89)
Private	45.5% (492)	49.6% (318)	45.6% (88)	41.8% (38)	30.6% (48)
Both	11.1% (120)	11.5% (74)	11.9% (23)	8.8% (8)	9.6% (15)
None	6.9% (75)	8.1% (52)	5.7% (11)	7.7% (7)	3.2% (5)
Antibiotica						.018
Yes	79.7% (889)	77.4% (514)	80.0% (160)	79.6% (74)	88.7% (141)
Effusion drainage						.126
Yes	2.6% (29)	3.5% (23)	1.5% (3)	0.0% (0)	1.9% (3)
Distant infectionb						.878
Yes	4.5% (50)	4.1% (27)	5.0% (10)	5.4% (5)	5.0% (8)
Daily medications taken at home	0 [0, 2]	0 [0, 0]	1 [0, 3]	2 [1, 4]	5 [2, 7]	<.001

Note: Data are presented as % (no.) for categorical variables and as median [interquartile range] for continuous variables; comparisons were made across the four mutually exclusive groups representing comorbidity-related functional limitations using Pearson’s χ² and Kruskal–Wallis tests.

^aAntibiotic prescribed for the current illness.

^bDefined as the presence of another site of infection (e.g., osteomyelitis, meningitis, etc.).

Among children with comorbidities, 200 (44.3%) had no or mild functional limitations, 93 (20.6%) had moderate limitations, and 159 (35.2%) had severe functional limitations (Table 1). Children with moderate or severe functional limitations were also more likely to have more than one comorbidity. Functional limitations differed across individual comorbidities (Figure 1). Most children with neurologic comorbidities had severe functional limitations (75.7%) and other comorbidities (78.5%), while more than half of children with asthma had none or mild limitations.

FIGURE 1.

Functional limitations for selected comorbidities. The y-axis represents the proportion of comorbidity-related functional limitations classified as none or mild, moderate, and severe. The x-axis represents the categories of comorbidities present in the study population. *Other includes psychiatric, gastrointestinal, renal, hematologic, endocrine, rheumatologic, and oncologic comorbidities. **Other pulmonary excludes asthma.

Associations between comorbidity-related functional limitations and outcomes

For the ordinal severity outcome, 166 (14.9%) were considered mild, 695 (62.3%) moderate, 200 (17.9%) severe, and 55 (4.9%) very severe (Table 2). There was a significant association between functional limitations and ordinal severity outcome (p < .001) (Figure 2). In multivariable analyses, the odds of experiencing a more severe outcome doubled for those with moderate functional limitations (adjusted odds ratio [aOR]: 2.15, 95% confidence interval [CI] [1.36, 3.38]), whereas those with severe functional limitations had a tripling of the odds for more severe outcomes compared with those without comorbidities (aOR: 3.02 [2.05, 4.43]).

TABLE 2.

Pneumonia outcomes by comorbidity-related functional limitations

			Comorbidity
	All participants (N = 1116)	No comorbidity (N = 664)	No or mild functional limitations (N =200)	Moderate functional limitations (N = 93)	Severe functional limitations (N = 159)	p value
Ordinal severitya						<.001
Mild	14.9% (166)	17.0% (113)	18.0% (36)	9.7% (9)	5.0% (8)
Moderate	62.3% (695)	62.5% (415)	65.5% (131)	61.3% (57)	57.9% (92)
Severe	17.9% (200)	16.4% (109)	14.0% (28)	23.7% (22)	25.8% (41)
Very severe	4.9% (55)	4.1% (27)	2.5% (5)	5.4% (5)	11.3% (18)
Length of stay (h)	45.0 [22.0, 82.2]	43.0 [21.0, 68.0]	41.0 [20.0, 73.2]	51.0 [32.0, 94.0]	84.0 [43.5, 154.0]	<.001

Note: Data are reported as percentage (frequency) and median [interquartile range].

^aAn ordinal severity measure based upon the most severe outcome during the encounter. Children discharged from the ED were mild; hospitalized children without intensive care were moderate; those requiring intensive care were severe or very severe, with the latter being those requiring invasive mechanical ventilation, vasoactive medications, or who died.

FIGURE 2.

Adjusted odds ratios for ordinal severity outcome and length of stay by functional limitations. Multivariable ordinal logistic regression models were used to evaluate associations between comorbidity-related functional limitations with our outcomes of disease severity and hospital length of stay, adjusting for age, gender, race, insurance, presence of parapneumonic effusion requiring drainage, and distant site of infection.

Overall, the median LOS was 45 h (IQR: 22.0–82.2) (Table 2). In multivariable analysis, functional limitations were significantly associated with LOS (p < .001) (Figure 2). Children with moderate functional limitations (median LOS: 51 h; IQR: 32.0–94.0) as well as those with severe functional limitations (median LOS: 84 h; IQR: 43.5–154.0) had increased odds of longer LOS (aOR: 2.52 [1.7, 3.73] for moderate functional limitations; aOR: 4.72 [3.33, 6.7] for severe functional limitations) as compared with those without comorbidities (median LOS: 43.0; IQR: 21.0–68.0). In our sensitivity analyses excluding children with home oxygen or noninvasive respiratory ventilatory support and those whose primary reason for admission was not related to pneumonia, results were materially unchanged and associations remained significant for both the ordinal severity and LOS outcomes (data not shown).

DISCUSSION

In this study, including data from 1116 children with pneumonia at two US children’s hospitals, 40% of children had one or more chronic comorbidities, with one in five children having comorbidities associated with moderate to severe functional limitations. Children with more severe functional limitations were at increased risk for more severe in-hospital outcomes, including admission to intensive care, respiratory failure, and longer LOS.

Children with moderate or severe functional limitations demonstrated the strongest associations with more severe outcomes, with 29%–36% requiring intensive care unit (ICU) care as compared to 20% or less for those with no comorbidities or comorbidity with no or mild functional limitations. These children were also more likely to have multiple comorbidities. In contrast, children with comorbidity but no or only mild functional limitations had similar outcomes to those with no comorbidities. Reasons for these differences are likely multifactorial, reflecting true differences in illness severity, potential increased provider discomfort or uncertainty, and the need for increased care coordination for children with moderate to severe functional limitations. Regardless of the reasons, functional limitations represent an important determinant of pneumonia disease outcomes.

The severity of functional limitations was also associated with the type of comorbidity. Approximately half of the children with asthma had no or only mild functional limitations. In contrast, most children with neurologic comorbidities had severe functional limitations. Prior studies demonstrate that children with neurologic disorders who are hospitalized with pneumonia have longer LOS, more complications, more ICU admissions, and utilize more resources as compared to those without neurological disorders.⁹ Variation in outcomes is also seen in this population in those with aspiration versus nonaspiration pneumonia,¹⁰ and in hospital-level resource utilization.¹¹ Functional limitations may help explain some of this observed variation.

To our knowledge, no prior studies have evaluated functional limitations as a predictor of pneumonia severity. Several studies have highlighted that children with comorbidities are at increased risk for more severe in-hospital outcomes.^2,12–15 Leyenaar et al.² previously demonstrated that children with complex chronic conditions who were hospitalized with pneumonia had longer LOS, a 50% increase in the odds for antibiotic escalation and pneumonia-related complications, and a quadrupling of the odds for readmission as compared to those without comorbidities. Irrespective of diagnosis, children with complex chronic conditions and those with multiple chronic conditions are at increased risk for ICU mortality, readmissions, and overall hospital resource utilization.^12–17 In the current study, the majority of children with multiple comorbidities also had moderate to severe functional impairment (69% in those with two or more comorbidities and 82% of those with three or more comorbidities). Although not directly comparable due to differences in the population studied and methods employed, it is likely that most children with complex and/or multiple comorbidities in these prior studies similarly had high degrees of functional impairment.

Children with medical complexity account for nearly one-third of all pediatric acute care hospitalizations and 40% of discharges from US children’s hospitals.^10,18 Efforts to optimize pneumonia outcomes and improve overall care delivery in these children are critically important. Improved understanding of functional limitations among children with comorbidities could facilitate improved clinical decisionmaking, reduce outcome disparities, and inform research and quality efforts. For example, clinical practice guidelines that tailor recommendations according to functional status (and thus the risk for severe outcomes) may reduce unwarranted variation around triage disposition, resource utilization, and other management decisions, thereby improving the quality and value of care. Similarly, directing care coordination and other services to bridge the transition from hospital to home to those at highest risk for poor outcomes may reduce disparities in reutilization and enhance the quality of life.^19–21 In the research context, assessments of functional status could reduce potential confounding effects in observational studies, target interventions to the highest risk populations in randomized trials, and enhance predictive accuracy in prognostic studies.

This study has several potential limitations. Comorbidities represent a broad group of many different disorders, all of which may not impart a similar risk for poor pneumonia outcomes. Our definition of functional limitations was based on a caregiver report and has not previously been validated. We sought to minimize misclassification using a standardized interview and rigorously defined case definitions. Our outcome measures are imperfect and may be influenced by factors other than disease severity. There may be other outcome differences that we did not examine, such as cost, other resource utilization measures, and posthospitalization morbidity. This study was performed on children presenting for emergency care at two tertiary care, free-standing children’s hospitals, and the prevalence and severity of comorbidities, disease outcomes, and other population characteristics may differ from other settings that may limit the generalizability of our findings.

CONCLUSION

In our study, chronic comorbidities and functional limitations were associated with pneumonia disease severity. Those with moderate to severe functional limitations were at the highest risk for severe outcomes, and this population represented approximately half of all children with chronic comorbidities. Consideration of functional limitations is critical when assessing expected outcomes. Further study in this population is needed to develop tailored approaches to care based on the presence and severity of comorbidity-related functional limitations that minimizes morbidity and optimizes resource utilization.