Depression and Anxiety Symptoms During and After Pediatric Asthma Hospitalization

We conducted a prospective cohort study in a large children’s hospital in the mid-Atlantic United States that serves as a quaternary referral center and urban community hospital. Patients were eligible to participate if they were at least 7 years old and <18 years old and had been admitted to an inpatient unit with an asthma exacerbation. Patients were ineligible to participate if they were admitted to the ICU, were in custody of social services, or had intellectual disability or cognitive impairment that prevented them from completing study questionnaires. Patients were also ineligible to participate if they or their consenting caregiver did not have sufficient English proficiency to complete the informed consent process and questionnaires because interpreter services were not available to the study team.

Trained research assistants screened patients admitted for asthma exacerbation for eligibility on transfer to an emergency department observation unit or medical floor. Research assistants approached eligible participants and their parent(s) or legal guardian (hereafter “caregivers”) when available during hospitalization, including during evening hours (5–10 pm) and on weekends. The research assistant explained the study and completed a written informed consent process with caregivers and obtained assent from all child participants. In accordance with state law, the research assistant also obtained informed consent for use of mental health information from participants aged 14 to 17 years and from the caregiver for children younger than 14 years. The study was approved by the hospital’s institutional review board.

All child participants completed standardized self-report scales (described below), and caregivers completed a self-report questionnaire and demographic form. All questionnaires were completed on a tablet by using REDCap (Research Electronic Data Capture).¹¹  Child participants were compensated $20 for completing study questionnaires in the hospital.

Starting in the sixth week after hospital discharge, a trained research assistant contacted the patient and caregiver to complete follow-up questionnaires (ie, the same pediatric self-report scales completed at baseline) via telephone or e-mailed questionnaire, depending on each participant’s stated preference. No participant actively withdrew from the study, but participants who were not reached after 3 contact attempts were considered lost to follow-up. Child participants who completed the follow-up questionnaire were compensated an additional $20.

Child participants completed the pediatric self-report Patient-Reported Outcomes Measurement Information System (PROMIS) 8-item Depression Scale, short form version 1.0, and pediatric self-report PROMIS 8-item anxiety scale, short form version 1.0, which were developed as part of a National Institutes of Health initiative.^12–14  The questionnaire asks how frequently the child has experienced a given symptom in the past 7 days, with response options on a 5-point scale (never, rarely, sometimes, often, and always), and scores are converted from a raw summary score to a scaled score centered around a population norm of 50 (ie, T scores). In keeping with recommendations, depression and anxiety symptom scores at or below the population norm of 50 were classified as normal, scores of 51 to 55 were classified as mildly elevated, scores of 56 to 65 were classified as moderately elevated, and scores >65 were classified as severely elevated.¹⁵  The PROMIS depression and anxiety instruments measure symptom severity (as opposed to examining whether a child meets criteria for a psychiatric diagnosis), and previous research has revealed that symptom scores change over time in response to changes in overall health status.^9,10,16–19  For child self-reported PROMIS depression scores, a change of 3 points in the symptom score corresponds to a clinically meaningful change.²⁰  For anxiety, no specific cut points have been defined; however, a general guideline for interpretation of PROMIS measures is that a change of 0.5 SD corresponds to a clinically meaningful change.¹⁵ 

The intensity of the child’s asthma exacerbation was categorized on the basis of a hospital-specific asthma treatment pathway as severe (requiring continuous albuterol) or moderate (requiring intermittent albuterol) at the time of emergency department presentation.²¹  Asthma exacerbation severity was obtained from structured data fields in the hospital’s electronic health record (EHR). To capture asthma treatment status in the period after the hospitalization, we determined if the patient was prescribed or directed to take an inhaled corticosteroid at hospital discharge on the basis of discharge prescriptions and communication with the discharging clinician. Documented mental health or psychiatric diagnoses were obtained from the patient’s diagnosis and problem lists in the EHR, which captures active diagnoses associated with the hospitalization and with previous outpatient visits that took place within the hospital’s clinical network (ie, affiliated treatment sites).

We summarized demographic and clinical characteristics using medians, frequencies, interquartile ranges, and percentages. We converted raw depression and anxiety symptom scores to scaled scores (T score = 50, SD = 10) on the basis of guidelines in the PROMIS manual.²²  In analyses of patient trajectories, scores were dichotomized as either normal or elevated to retain power because of small cell sizes among the subgroups of patients with elevated scores and to reduce multiple comparisons. Symptom scores for depression and anxiety that fell at or below a population norm of 50 were classified as normal, and scores >50 were classified as elevated.

We compared in-hospital and follow-up depression and anxiety scores using paired t tests and used the Pearson correlation coefficient to examine the relationship between depression and anxiety symptoms during hospitalization and again at follow-up. We also examined patient trajectories by capturing the number of patients falling into each mental health symptom severity category at baseline compared with follow-up and constructed Sankey diagrams using the networkD3 package in R to visually display individual patient trajectories. All other analyses were conducted in Stata 15 (Stata Corp, College Station, TX). P values <.05 were considered statistically significant.

We approached 190 patients during hospitalization and enrolled 135 (71%). Of those, 39 (29%) were lost to follow-up, resulting in a final sample of 96 child-caregiver dyads. The mean time from date of discharge to completion of follow-up questionnaires was 45.9 days (SD = 22.4). There were no significant differences at baseline in age, sex, race, ethnicity, Medicaid insurance status, or depression and anxiety symptoms during hospitalization for those who did and did not complete follow-up questionnaires (P > .05 for all comparisons).

Table 1 reveals demographic and clinical characteristics for the enrolled sample. The sample included a high proportion of patients (91%) who were of Black, Hispanic, or other race or ethnicity (not white). The most common insurance type was Medicaid (70%), and a range of family incomes were represented. Among caregivers, 93% reported completing high school or beyond and 30% graduated college. Among study participants, 69% were classified in the emergency department as having a severe asthma exacerbation requiring continuous albuterol. The median hospital length of stay was 2 days (interquartile range 1–2).

During hospitalization, 37% of the 135 children who enrolled in the study had depression scores in the elevated range and 47% had anxiety scores in the elevated range. Depression and anxiety scores were positively correlated during hospitalization (r = 0.74). Table 2 reveals the proportion of children with scores classified in the normal, mild, moderate, and severe depression and anxiety symptom ranges at baseline.

Table 3 reveals comparisons of depression and anxiety symptoms during hospitalization and at follow-up and the change in scores over time for the 96 patients who completed follow-up measures. The mean within-person change in depression symptoms from hospitalization to follow-up was −5.5 points (SD = 10.8; P < .001), representing a clinically meaningful reduction in symptoms. The mean within-person change for anxiety symptoms was −8.3 points (SD = 11.8; P < .001), also representing a clinically meaningful reduction in symptoms. Depression and anxiety scores were positively correlated at postdischarge follow-up (r = 0.62).

Supplemental Table 4 reveals the number of patients with an elevated (>50) depression or anxiety symptom score at each time period, and Fig 1 illustrates these data with patient symptom trajectories for the 96 patients who completed follow-up measures. All possible trajectories were present: no elevated symptoms at hospitalization or follow-up, persistent elevated symptoms, elevated symptoms that resolved over time, and new symptoms that emerged over time. Overall, 27% of patients had elevated depression or anxiety symptoms during hospitalization but no longer had elevated symptoms at postdischarge follow-up, whereas 9% of patients had no significant symptoms during hospitalization but reported elevated symptoms at follow-up.

We also examined findings among subgroups based on their reporting of depression or anxiety symptoms during hospitalization. Among the subgroup of 28 patients initially reporting both depression and anxiety during hospitalization, 18% continued to report both depression and anxiety symptoms 1 to 2 months after discharge, 18% reported either depression or anxiety, and 64% reported no symptoms at follow-up. Among the subgroup of 45 patients with no reported depression or anxiety symptoms during hospitalization, 20% had elevated symptoms after discharge.

We found that more than half (53%) of the children and adolescents hospitalized for an asthma exacerbation in our study reported elevated depression or anxiety symptoms during hospitalization, despite the fact that only 11% of patients in our final sample had a documented mental health condition. At 1 to 2 months’ follow-up, 26% of patients reported elevated depression or anxiety symptoms, indicating that some symptoms had resolved by that time. Patients reporting elevated symptoms of both depression and anxiety during hospitalization were most likely to report persistent depression and anxiety symptoms at follow-up.

Our findings are consistent with previous studies revealing that depression and anxiety symptoms are common among children with asthma^2,7,23  and expand our understanding of mental health symptom burden in children hospitalized for asthma exacerbation. Most previous research has been focused on documentation of diagnoses of mood and anxiety disorders rather than measuring symptom burden. Our findings reveal that many children and adolescents experience mental health symptoms during hospitalization, and most of them had no formal mental health diagnoses. Our study sample was predominantly composed of Black and Hispanic patients insured by Medicaid, who are at higher risk of asthma and asthma hospitalization^24,25  and may also be at risk for mental health disparities.²⁶  Thus, our results help extend findings from previous research to priority populations that have been historically underrepresented in clinical trials and other research.^27–29  In addition, the finding that acute mental health symptoms were common may be of particular interest for the general pediatric hospitalist because children requiring hospitalization for other issues, particularly those with an acute on chronic presentation (such as irritable bowel syndrome or diabetic ketoacidosis) may experience similar distress.

Regulatory bodies increasingly require mental health screenings during medical encounters,³⁰  and hospitals are beginning to compare performance on mental health measures. Our findings suggest that hospital-based screening initiatives are likely to detect acute symptoms and reactions to acute general medical illness. As with physical pain, young people are likely to benefit from assistance in managing their anxiety and mood symptoms as part of their hospital care, regardless of whether those symptoms persist beyond their stay or are part of a more chronic mood or anxiety disorder. Several hospital-based coping tool kits exist for children and parents,^31–34  and future researchers could investigate whether implementing these tools more broadly could benefit children with elevated depression and anxiety symptoms during hospitalization for asthma or other conditions. Previous work has revealed that when mental health consultation is needed, involving mental health consultants earlier reduces hospital length of stay.³⁵  When mental health specialists are not available, a previous qualitative study with hospitalized young people experiencing mental health symptoms found that they valued compassionate clinicians; staying occupied with activities; physical comforts, such as warm blankets; and knowing what to expect, suggesting that general patient-centered care can reduce distress.³⁶  Future research to investigate what depression and anxiety symptom thresholds warrant specific interventions or mental health consultation could help ensure that patients with elevated symptoms receive the right support at the right time.

It is also important to note that not all children who struggle with mental health symptoms during an acute hospitalization will have persistent symptoms over time. Our finding that a substantial portion of children no longer had elevated symptoms at postdischarge follow-up is consistent with previous research revealing that depression and anxiety symptom burden is responsive to changes in a child’s physical asthma symptoms.¹⁹  Yet 20% of the children in our sample who did not report elevated depression and anxiety symptoms during hospitalization did report elevated symptoms 1 to 2 months after discharge, which could suggest that some children had a delayed response to the stress of acute illness exacerbation and hospitalization (ie, medical posttraumatic stress^37,38 ) or that they experienced a new stressor. Given that we observed various trajectories, a process to monitor symptoms post hospitalization is likely to be important, both to avoid overtreatment and because new or persistent depression and anxiety symptoms influence overall well-being and can affect adherence to asthma therapies.^3,39  This highlights the importance of educating families about when to seek follow-up mental health care (eg, if symptoms persist or new or serious symptoms, such as suicidal thoughts, emerge) and also points to the value of incorporating mental health specialists into hospital-based and outpatient medical care teams. When this is not possible, families are likely to benefit from messaging that acknowledges that coping with a chronic health condition can be challenging and that anxiety and depression symptoms are common and treatable.

Future research to understand the reasons for elevated depression and anxiety symptoms during and after asthma hospitalization can inform clinical care to help address these symptoms. Questions of particular importance for future research will include the following: understanding risk factors for elevated depression and anxiety symptoms, whether hospital-based interventions can improve depression and anxiety symptoms, how ongoing asthma symptoms post hospitalization correlate with asthma self-management and overall well-being, and what constitutes optimal postdischarge outpatient care for mental health symptoms.

This study has several limitations. First, in keeping with routine hospital practice, we relied on EHR-documented history of psychiatric conditions rather than formal psychiatric evaluations to identify children with a history of a mental health diagnosis, and the accuracy of this information may vary because of variations in clinical practice.⁷  In addition, whereas the EHR captured diagnostic information for children receiving primary care or specialty mental health care within the hospital’s care network, we did not have access to information from outside providers. Second, we did not have information on whether patients received mental health treatment after discharge. As a result, we are unable to determine if symptoms that resolved did so because of treatment or did so without treatment. Third, we offered participants the option to complete follow-up rating scales by telephone or e-mail. This was to maximize convenience for families, but it introduces a possibility of measurement bias if mode of completion influenced disclosure of symptoms. A study of mode effects in self-report versus telephone-administered PROMIS pediatric scales found high correlations between scores obtained with the 2 modes, with small differences that would not be expected to affect overall interpretation of results.⁴⁰  Fourth, we were not able to include families who preferred a language other than English because trained interpreters were not available to the study team. Fifth, a high proportion of families were lost to follow-up. The most common reason for failed telephone follow-up was that the phone numbers provided to the study team were for prepaid, no-contract cellular phones, and the phone numbers were no longer in service at follow-up. Future studies may benefit from anticipating and troubleshooting potential contact difficulties in advance, such as acquiring contact info from multiple caregivers.

Anxiety and depression symptoms are common during hospitalization for acute asthma exacerbation and persist after discharge in many patients. Our findings suggest that efforts to assess symptom burden during hospitalization could be most effective at helping patients if they are accompanied with a process to monitor symptoms post hospitalization.

We thank John Sebastian de Armas, Helen Fetaw, Lila Gutstein, Jessica Andrews, Alexis Kalargheros, and Madelyn Goode for their assistance with data collection, and we thank Cadence Bowden for her assistance preparing the manuscript for publication.