BACKGROUND AND OBJECTIVES

The coronavirus disease 2019 (COVID-19) pandemic contributed to the public health crisis for pediatric mental health. We characterized our local patient population presenting with suicidality or suicide attempts before and after the pandemic by examining: 1. frequencies of hospitalizations for suicidality to determine whether they differed by age, legal sex, race and ethnicity, or socioeconomic status; 2. average length of stay and discharge disposition; 3. 7-, 30-, and 365-day reutilization rates; and 4. admission trends during COVID-19 surges.

METHODS

Retrospective data between March 2018 and March 2022 was analyzed, including patients ages 10 to 17 years hospitalized for either suicidality or a suicide attempt at 1 freestanding tertiary care pediatric medical hospital in the Midwest. Encounters were divided into 2 categories on the basis of the COVID-19 pandemic: “Prelockdown” (March 1, 2018–March 12, 2020) and “postlockdown” (March 13, 2020–March 31, 2022). Patients were limited to 1 presentation pre- and postlockdown. We analyzed frequencies using means and SDs, categorical data using χ2 and Fisher’s exact tests, and continuous data with t tests.

RESULTS

A total of 1017 encounters were included, stratified into pre- and postlockdown groups for analysis (909 encounters, 889 unique patients). There was a significant difference in 365-day reutilization pre- and postlockdown when analyzing re-presentation to the emergency department (P = .025) and hospital readmission (P = .006). Admissions incrementally increased after the COVID-19 alpha variants in September 2020 and again after the delta variant in August 2021.

CONCLUSIONS

The COVID-19 pandemic further intensified the already critical issue of pediatric mental health, demonstrating increased reutilization in the year after their initial presentation and an increase in admissions after the alpha variant.

The rising suicide rates among America’s youth pose a growing public health concern. Before the coronavirus disease 2019 (COVID-19) pandemic, pediatric hospitals reported a concerning trend in hospitalizations related to psychiatric illness.13  From 2008 to 2015, the total number of hospitalizations for suicidality more than tripled, and the percentage of all pediatric hospitalizations for these concerns doubled across 31 children’s hospitals.2  The COVID-19 pandemic amplified the rates of mental health disorders, and increasing evidence suggests that the depression, anxiety, and social isolation associated with the COVID-19 pandemic have led to increased suicidality among youth.49  At our institution, hospital medicine providers perceived an increase in patients presenting with suicidality, suicide attempts (SAs), and/or suicide death after the pandemic’s onset. Therefore, we sought to better understand local trends of patients presenting with suicidality preceding and during the COVID-19 pandemic at our institution and to understand whether the unique COVID-19 variant surges impacted admissions for suicidality. Although earlier studies explored how school closures affected emergency department (ED) visits and hospitalizations for psychiatric disorders,9  understanding of the influence of the specific COVID-19 variant surges on hospitalizations for suicidality remains limited.

Moreover, we sought to explore the relationship between social factors and youth suicidality at our institution because literature has shown that the COVID-19 pandemic disproportionately affected racial and ethnic minority groups, as well as those with lower socioeconomic status (SES).4,6,1015  A growing body of evidence has highlighted significant racial and ethnic disparities in suicide trends during the pandemic.4,6,1315  Notably, a prepandemic study investigating suicide rates among children from 2001 to 2015 unveiled differences between white and Black children in risk of suicide death at different ages. Among younger children (aged 5–12 years), Black youth exhibited significantly higher incidence of suicide than white children, whereas older white children had higher suicide rates.16  The existing literature on social factors motivated us to understand the unique characteristics of patients presenting to our hospital with suicidality.

Additional interconnected factors we aimed to understand included length of stay (LOS), discharge disposition, and reutilization rates, with a specific focus on understanding the potential impact of the COVID-19 pandemic on these aspects. We conjectured that a larger proportion of patients were discharged from the hospital than to an inpatient psychiatric facility after the onset of the pandemic because of multiple institutional-specific and county-specific factors such as: We do not have an associated inpatient psychiatric or behavioral unit, the location of our county’s inpatient psychiatric facility changed, adolescent capacity at other community facilities decreased shortly after the pandemic, and notable pandemic-related staffing shortages occurred.

Although existing literature in both adult and pediatric contexts before the pandemic suggests high reutilization rates among patients hospitalized with suicidality,1720  there is a paucity of literature specifically focused on readmission rates to acute care hospitals subsequent to initial hospitalizations for suicidality. The most comprehensive, multicenter, retrospective, pediatric study to date demonstrated a 30-day readmission rate of 8.5% (with 94.5% of the readmissions being for a psychiatric problem) after a medical hospitalization for suicidality,20  whereas several other studies demonstrated 30-day readmission rates of 8 to 22%.2125  Because it is known that there is an increased risk of readmissions for patients with a previous hospitalization or ED visit,2125  our goal was to ascertain our local readmission rates, assess how the time elapsed from the initial encounter affected these rates, and determine whether readmissions increased after the onset of the pandemic. Specifically, we aimed to investigate readmission rates for mental behavioral health (MBH)-related complaints over 7, 30, and 365 days because we felt this comprehensive approach may provide insight into the various aspects of our system that could be optimized. For example, readmissions within 7 days may offer a signal that these patients were not ready for a discharge and may have benefitted from inpatient psychiatric care, whereas readmissions within 30 days could indicate that patients were not adequately connected with postdischarge care. Of note, the 30-day readmission rates are most often reported in the psychiatric readmission literature, as recommended per the National Committee on Quality Assurance,2024,26,27  providing an additional reason to examine the 30-day time period. Although multiple factors could account for readmissions within 365 days, we analyzed these with the rationale that readmissions within 1 year could indicate persistent and severe depression, worsening mental health after the pandemic, and/or inadequate community resources, among a multitude of other issues.

Overall, our goal was to thoroughly characterize a specific patient cohort of youth ages 10 to 17 years hospitalized for suicidality at our medical tertiary care hospital, examining retrospective data between March 2018 and March 2022. To achieve this, our objectives were as follows:

  1. Compare the frequencies of hospitalizations for suicidality or SA (with possible suicide death) before and after the onset of the COVID-19 pandemic to determine whether they differed by age, sex, race and ethnicity, or SES.

  2. Determine average LOS and understand how the pandemic may have impacted trends in discharge disposition.

  3. Examine and compare 7-, 30-, and 365-day reutilizations rates.

  4. Assess admission trends during the COVID-19 surges, specifically the alpha variant in November 2020, delta variant in August 2021, and οmicron variant in January 2022.

We conducted a retrospective, single-center study that included patients aged 10 to 17 years hospitalized on the acute care floor or ICU for suicidality between March 2018 and March 2022. Our children’s hospital is a freestanding, tertiary care medical hospital in the Midwest and has 306 inpatient beds (with no associated inpatient psychiatric or behavioral unit), and is a referral center for the Southwestern part of our state along with portions of 2 bordering states. Our county does have a mental health ED, but given that our hospital serves multiple counties, families often bring their children to our ED when they are experiencing a mental health crisis. Our ED has the ability to transfer patients directly to inpatient psychiatric facilities. If there is not bed availability in the community, patients are hospitalized for mental health boarding. Given that we frequently care for MBH boarding patients, we have care guidelines for these patients to provide standardized care to this population. Our psychiatry and psychologist colleagues are available for consultation as needed, but do not see every patient hospitalized with suicidality.

To capture this patient population, electronic health record (EHR) data were extracted from Epic (Verona, Wisconsin) via an EHR-generated report that included hospitalized patients with a behavioral health diagnosis as defined by our institution’s executive MBH team. Diagnoses were manually reviewed to only include diagnoses that indicated suicidality (including suicidal risk, ideation, and/or plan) and SAs (including but not limited to intentional ingestion and intentional self-harm with suicidal intent). Patients with subsequent suicidal death from their encounter were also included. The EHR-generated report included encounter information (admission/discharge date and time, LOS, admitting and discharge specialty, and primary diagnosis) and patient demographics (age, race and ethnicity, legal sex, insurance type, and zip code). Legal sex is the patient’s sex assigned at birth and was extracted from our EHR-generated report and is referred to as sex for the remainder of this manuscript. No patients were excluded from this analysis. This study was approved by our institution’s institutional review board.

To assess differences pre- and postlockdown, our data were split into the following 2 groups on the basis of the timing of the COVID-19 pandemic: Prelockdown (March 1, 2018–March 12, 2020) and postlockdown (March 13, 2020–March 31, 2022). The beginning of our postlockdown period was chosen as March 13, 2020, the day our state-ordered lockdown occurred, which mandated all schools to close. Our prelockdown period and postlockdown period were each 24.5 months. Four COVID-19 variant surges were individually assessed on the basis of local health department data: Novel coronavirus 19 (March 2020), alpha (September 2020), delta (August 2021), and omicron (January 2022).

All encounters meeting inclusion criteria were examined in terms of LOS, age of presentation, and discharge disposition. To assess differences pre- and postlockdown, each patient could have 1 index admission in both the pre- and postlockdown periods. Reutilizations were grouped into “none,” “once,” or “more than once” categories. The more than once category was grouped for accuracy to avoid overestimating reutilizations. Seven-, 30-, and 365-day reutilization data were obtained through manual chart review by independent reviewers to identify and verify that re-presentation to the ED or hospital were for an MBH-related complaint (included but not limited to ingestion, self-harm, suicidality, SA, and/or death, aggressive behavior, or a behavioral medication issue). This was limited to presentations/facilities with shared EHR access for subsequent encounters. Any other presentations for an acute medical issue were excluded.

To accurately assess demographic information (race and ethnicity, sex, insurance type, and zip code-based SES), each patient meeting inclusion criteria was only counted once (referred to as unique patients). Including race and ethnicity as our goal was to identify if individuals with a particular self-reported race and ethnic background were disproportionally affected by the pandemic. Race and ethnicity exist as 1 designation in our EHR and was assigned by survey self-report upon health care encounter. There are not definitions or criteria regarding the categories provided to caregivers completing the demographic information.

Zip code-based SES (low, medium-low, medium, medium-high, high) were assigned based on the Milwaukee County health report, a population health framework generated by the Center for Urban Population Health. Of note, our hospital is located within Milwaukee County, which has 35 zip codes contained and is home to one-eighth of Wisconsin’s residents. Milwaukee county contains the largest city in Wisconsin, Milwaukee, which is 1 of the most racially segregated cities in the United States.28  Those patients from outside Milwaukee County were also included (Table 1). Discharge disposition was categorized by our EHR into: Home, psychiatric facility, and other (against medical advice, police custody, or deceased).

TABLE 1

Patient Demographics and Comparison Pre- and Postlockdown (N = 889)

Patient DemographicsTotal N (%)Prelockdown N (%)Postlockdown N (%)P
Race and ethnicity N = 889 N = 340 N = 549  
 American Indian or Alaska Native 5 (0.6%) 0 (0.0%) 5 (0.9%) .053 
 Asian American 15 (1.7%) 5 (1.5%) 10 (1.8%) 
 Black or African American 179 (20.1%) 65 (19.1%) 114 (20.8%) 
 Hispanic or Latino or Latinx 147 (16.5%) 64 (18.8%) 83 (15.1%) 
 Multiracial 43 (4.8%) 9 (2.6%) 34 (6.2%) 
 Native Hawaiian or other Pacific Islander 1 (0.1%) 1 (0.3%) 0 (0.0%) 
 Unknown 68 (7.6%) 31 (9.1%) 37 (6.7%) 
 White or Caucasian 431 (48.5%) 165 (48.4%) 266 (48.5%) 
Sex N = 889 N = 340 N = 549  
 Female 709 (79.8%) 270 (79.4%) 439 (80.0%) .842 
 Male 180 (20.2%) 70 (20.6%) 110 (20.0%) 
Insurance type N = 889 N = 340 N = 549  
 Commercial 390 (43.9%) 148 (43.5%) 242 (44.1%) .354 
 Medicaid 487 (54.8%) 185 (54.4%) 302 (55.0%) 
 Self-pay 12 (1.3%) 7 (2.1%) 5 (0.9%) 
Zip code-based SES N = 889 N = 340 N = 549  
 Low 112 (12.6%) 40 (11.8%) 72 (13.1%) .011 
 Medium-low 113 (12.7%) 40 (11.8%) 73 (13.3%) 
 Medium 92 (10.3%) 31 (9.1%) 61 (11.1%) 
 Medium-high 56 (6.3%) 18 (5.3%) 38 (6.9%) 
 High 37 (4.2%) 9 (2.6%) 28 (5.1%) 
 Outside Milwaukee County limits 479 (53.9%) 202 (59.4%)a 277 (50.5%)a 
Milwaukee County zip code-based SES N = 410 N = 138 N = 272  
 Low 112 (27.3%) 40 (29.0%) 72 (26.5%) .76 
 Medium-low 113 (27.4%) 40 (29.0%) 73 (26.8%) 
 Medium 92 (22.4%) 31 (22.5%) 61 (22.4%) 
 Medium-high 56 (13.7%) 18 (13.0%) 38 (14.0%) 
 High 37 (9.0%) 9 (6.5%) 28 (10.3%) 
Patient DemographicsTotal N (%)Prelockdown N (%)Postlockdown N (%)P
Race and ethnicity N = 889 N = 340 N = 549  
 American Indian or Alaska Native 5 (0.6%) 0 (0.0%) 5 (0.9%) .053 
 Asian American 15 (1.7%) 5 (1.5%) 10 (1.8%) 
 Black or African American 179 (20.1%) 65 (19.1%) 114 (20.8%) 
 Hispanic or Latino or Latinx 147 (16.5%) 64 (18.8%) 83 (15.1%) 
 Multiracial 43 (4.8%) 9 (2.6%) 34 (6.2%) 
 Native Hawaiian or other Pacific Islander 1 (0.1%) 1 (0.3%) 0 (0.0%) 
 Unknown 68 (7.6%) 31 (9.1%) 37 (6.7%) 
 White or Caucasian 431 (48.5%) 165 (48.4%) 266 (48.5%) 
Sex N = 889 N = 340 N = 549  
 Female 709 (79.8%) 270 (79.4%) 439 (80.0%) .842 
 Male 180 (20.2%) 70 (20.6%) 110 (20.0%) 
Insurance type N = 889 N = 340 N = 549  
 Commercial 390 (43.9%) 148 (43.5%) 242 (44.1%) .354 
 Medicaid 487 (54.8%) 185 (54.4%) 302 (55.0%) 
 Self-pay 12 (1.3%) 7 (2.1%) 5 (0.9%) 
Zip code-based SES N = 889 N = 340 N = 549  
 Low 112 (12.6%) 40 (11.8%) 72 (13.1%) .011 
 Medium-low 113 (12.7%) 40 (11.8%) 73 (13.3%) 
 Medium 92 (10.3%) 31 (9.1%) 61 (11.1%) 
 Medium-high 56 (6.3%) 18 (5.3%) 38 (6.9%) 
 High 37 (4.2%) 9 (2.6%) 28 (5.1%) 
 Outside Milwaukee County limits 479 (53.9%) 202 (59.4%)a 277 (50.5%)a 
Milwaukee County zip code-based SES N = 410 N = 138 N = 272  
 Low 112 (27.3%) 40 (29.0%) 72 (26.5%) .76 
 Medium-low 113 (27.4%) 40 (29.0%) 73 (26.8%) 
 Medium 92 (22.4%) 31 (22.5%) 61 (22.4%) 
 Medium-high 56 (13.7%) 18 (13.0%) 38 (14.0%) 
 High 37 (9.0%) 9 (6.5%) 28 (10.3%) 

N = 889 patients; this is the number of unique patients meeting study inclusion criteria and were only included in our analysis once.

a

Indicates which demographic groups differed significantly pre- to postlockdown.

Data were reported as mean and SD or as total number (N) and percentages. Categorical data were analyzed using χ2 or Fisher’s exact tests, as appropriate. Continuous data were analyzed using paired t tests. A P value < .05 was considered significant. SPSS Version 28.0 (IBM) was used for analysis.

When examining LOS, age at presentation, and discharge disposition, every encounter was examined (1017 encounters), referred to as our “Entire Cohort.” After allowing each patient to 1 index admission within our pre- and postlockdown comparison, there were 909 encounters (889 unique patients) included in our pre-/postlockdown analysis (Fig 1).

FIGURE 1

Hospitalizations and encounters for suicidality during study period. This figure shows the timeline of hospitalizations and encounters included for patients presenting with suicidality and/or SA during our study period. It is divided into the Entire Cohort meeting inclusion criteria (March 1, 2018–March 31, 2022), how that cohort was split into pre- and postlockdown numbers, and how those encounters were reviewed in terms 7-, 30-, and 365-day reutilization through 1 year after our study period (March 1, 2018–March 31, 2023). Our state-ordered lockdown occurred on March 13, 2020, closing all of the schools, and this date was the beginning of our postlockdown cohort.

FIGURE 1

Hospitalizations and encounters for suicidality during study period. This figure shows the timeline of hospitalizations and encounters included for patients presenting with suicidality and/or SA during our study period. It is divided into the Entire Cohort meeting inclusion criteria (March 1, 2018–March 31, 2022), how that cohort was split into pre- and postlockdown numbers, and how those encounters were reviewed in terms 7-, 30-, and 365-day reutilization through 1 year after our study period (March 1, 2018–March 31, 2023). Our state-ordered lockdown occurred on March 13, 2020, closing all of the schools, and this date was the beginning of our postlockdown cohort.

Close modal

The average LOS was longer postlockdown (48.28 ± 54.30 postlockdown versus 38.23 ± 24.75 hours prelockdown, P < .001). Younger patients presented postlockdown (14.71 ± 1.74 years versus 15.06 ± 1.61 years, respectively; P = .002). Patients were discharged from the hospital more frequently postlockdown (31.8% post versus 24.3% pre; P < .032).

Pre-/postlockdown encounters were limited to 1 index admission pre- and postlockdown (909 encounters). They were compared regarding reutilization rates for MBH complaints at 7, 30, and 365 days from index presentation. The most common MBH-related re-presentation was an intentional ingestion with suicidal intent. There was no significant difference in 7- and 30-day ED visits or readmissions pre- versus postlockdown (Table 2). However, there were significant differences in those who presented to the ED and those readmitted over 365 days pre- and postlockdown. Regarding 365-day reutilization rates for ED visits, 10.9% of the patients presented more than once postlockdown compared with 5.6% prelockdown (P = .025). Regarding 365-day readmissions, 4% of the patients were readmitted more than once postlockdown compared with 0.6% prelockdown (P = .006).

TABLE 2

Rates of Reutilization for Mental/Behavioral Health Reasons Pre- and PostLockdown (N = 909)

Reutilization RatesN (%)P
7-d ED revisit (total N = 909) None Once More than once  
 Prelockdown (N = 340) 338 (99.4%) 2 (0.6%) — .225 
 Postlockdown (N = 569) 560 (98.4%) 9 (1.6%) — 
30-d ED revisit (Total N = 909) None Once More than once  
 Prelockdown (N = 340) 326 (95.9%) 11 (3.2%) 3 (0.9%) .359 
 Postlockdown (N = 569) 534 (93.8%) 30 (5.3%) 5 (0.9%) 
365-d ED revisit (total N = 909) None Once More than once  
 Prelockdown (N = 340) 279 (82.1%) 42 (12.4%) 19 (5.6%)a .025 
 Postlockdown (N = 569) 440 (77.3%) 67 (11.8%) 62 (10.9%)a 
7-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 340 (100%) 0 (0%) — .297 
 Postlockdown (N = 569) 566 (99.5%) 3 (0.5%) — 
30-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 334 (98.2%) 6 (1.8%) — .397 
 Postlockdown (N = 569) 554 (97.4%) 15 (2.9%) — 
365-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 312 (91.8%) 26 (7.6%) 2 (0.6%)a .006 
 Postlockdown (N = 569) 495 (87.0%) 51 (9.0%) 23 (4.0%)a 
Reutilization RatesN (%)P
7-d ED revisit (total N = 909) None Once More than once  
 Prelockdown (N = 340) 338 (99.4%) 2 (0.6%) — .225 
 Postlockdown (N = 569) 560 (98.4%) 9 (1.6%) — 
30-d ED revisit (Total N = 909) None Once More than once  
 Prelockdown (N = 340) 326 (95.9%) 11 (3.2%) 3 (0.9%) .359 
 Postlockdown (N = 569) 534 (93.8%) 30 (5.3%) 5 (0.9%) 
365-d ED revisit (total N = 909) None Once More than once  
 Prelockdown (N = 340) 279 (82.1%) 42 (12.4%) 19 (5.6%)a .025 
 Postlockdown (N = 569) 440 (77.3%) 67 (11.8%) 62 (10.9%)a 
7-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 340 (100%) 0 (0%) — .297 
 Postlockdown (N = 569) 566 (99.5%) 3 (0.5%) — 
30-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 334 (98.2%) 6 (1.8%) — .397 
 Postlockdown (N = 569) 554 (97.4%) 15 (2.9%) — 
365-d hospital readmission (total N = 909) None Once More than once  
 Prelockdown (N = 340) 312 (91.8%) 26 (7.6%) 2 (0.6%)a .006 
 Postlockdown (N = 569) 495 (87.0%) 51 (9.0%) 23 (4.0%)a 

N = 909 patients; these are patients the pre-/postlockdown analysis, limited to one index admission in both the pre- and postlockdown periods.

a

Indicates which demographic groups differed significantly pre- to postlockdown.

A frequency analysis using the pre-/postlockdown group examined discharge disposition and whether they had none, 1, or >1 reutilization (Table 3). Results showed that fewer patients were discharged to a higher level of psychiatric care postlockdown versus prelockdown for those with no or 1 reutilization. However, a significant proportion of patients who reutilized more than once were discharged to a psychiatric facility postlockdown. More patients were discharged from the hospital across each category postlockdown than prelockdown (Table 3).

TABLE 3

365-Day Reutilization Rates From Initial Discharge Disposition Pre- and Postlockdown (N = 909)

365-d readmission or emergency revisit disposition (Total N = 909)N (%)
Higher level of psychiatric care (N = 639) None Once More than once 
 Prelockdown (N = 252, 74.1%) 200 (79.4%) 37 (14.7%) 15 (6.0%) 
 Postlockdown (N = 387, 68.0%) 290 (74.9%) 47 (12.1%) 50 (12.9%) 
Home (N = 263) None Once More than once 
 Prelockdown (N = 84, 24.7%) 75 (89.3%) 5 (6.0%) 4 (4.8%) 
 Postlockdown (N = 179, 31.5%) 148 (82.7%) 19 (10.6%) 12 (6.7%) 
Other (left AMA, in police custody, or deceased) (N = 7) None Once More than once 
 Prelockdown (N = 4, 1.2%) 3 (75%) 1 (25%) 0 (0%) 
 Postlockdown (N = 3. 0.5%) 2 (66.7%) 1 (33.3%) 0 (0%) 
365-d readmission or emergency revisit disposition (Total N = 909)N (%)
Higher level of psychiatric care (N = 639) None Once More than once 
 Prelockdown (N = 252, 74.1%) 200 (79.4%) 37 (14.7%) 15 (6.0%) 
 Postlockdown (N = 387, 68.0%) 290 (74.9%) 47 (12.1%) 50 (12.9%) 
Home (N = 263) None Once More than once 
 Prelockdown (N = 84, 24.7%) 75 (89.3%) 5 (6.0%) 4 (4.8%) 
 Postlockdown (N = 179, 31.5%) 148 (82.7%) 19 (10.6%) 12 (6.7%) 
Other (left AMA, in police custody, or deceased) (N = 7) None Once More than once 
 Prelockdown (N = 4, 1.2%) 3 (75%) 1 (25%) 0 (0%) 
 Postlockdown (N = 3. 0.5%) 2 (66.7%) 1 (33.3%) 0 (0%) 

N = 909 patients; these are patients the pre-/postlockdown analysis, limited to one index admission in both the pre- and postlockdown periods. AMA, against medical advice.

Demographic data of the unique patients (N = 889) are reported in Table 1. There was no significant difference in terms of race and ethnicity, sex, or insurance type pre- and postlockdown. White or Caucasian patients were the largest percentage of patients in our analysis, followed by those who identify as Black or African patients and Hispanic or Latino or Latinx patients. When examining zip code-based SES, those patients from outside Milwaukee County limits presented more frequently prelockdown (59.4%) compared with postlockdown (50.5%) (P = .011). Though not significant, when analyzing patients only from within Milwaukee County, 54.7% patients were from a zip code that was in a low to medium-low SES group (Table 1).

A chronological visual representation of total MBH-related encounters annotated with the COVID-19 variant time frames was generated (Fig 2). MBH-related hospitalizations increased after the alpha variant, returned to prelockdown historical levels during summer 2021, and then dramatically increased following the delta variant.

FIGURE 2

Hospitalizations for intentional SA and/or suicide ideation with COVID-19 variants. Admissions for suicide ideation or attempt incrementally increased after the COVID-19 alpha variants in September 2020 and again after the delta variant in August 2021.

FIGURE 2

Hospitalizations for intentional SA and/or suicide ideation with COVID-19 variants. Admissions for suicide ideation or attempt incrementally increased after the COVID-19 alpha variants in September 2020 and again after the delta variant in August 2021.

Close modal

The findings of our study add to the growing body of evidence suggesting that the COVID-19 pandemic has further intensified the already critical issue of pediatric mental health and its profound impact on reutilization for MBH complaints. Our findings showed an increase in reutilization for MBH reasons within 1 year after an initial hospitalization for suicidality and/or SA. Although previous studies have examined the pandemics’ effects on suicidality and/or SA and ED and medical hospitalization admissions,46,8,9,15,29,30  we are unaware of any other studies that have included 5 years of data spanning the pandemic, with the 365-day reutilization data included.

Younger patients presented postlockdown, which was also reported in another study examining suicidality and/or SA and the pandemic.4  Although the age difference was <1 year, the social and psychological effects of the COVID-19 pandemic may have accelerated suicidal thoughts or actions in those predisposed to MBH conditions in their adolescent years.

Our analysis did not reveal any statistically significant racial or ethnic differences in patients presenting before or after the lockdown period. However, when comparing the pre- and postlockdown periods, it appears there was an increase in patients presenting with suicidality who self-identify as multiracial. However, because our race and ethnicity data were generated via self-report, and patients who identify as multiracial may represent many different races and ethnicities, as well as our small sample size and single-center study, we acknowledge patients presenting with suicidality need to be further studied to better understand and characterize the demographics of these patients.

When examining the geographical distribution of our patients, those outside our county limits presented less frequently postlockdown which may indicate that our inner-city population may have been more greatly affected by the pandemic. From our zip code-based SES analysis, we did not identify a difference pre- and postpandemic but patients with low and medium-low SES represented the highest proportions of patients with suicidality, which is supported by previous literature on the relationship between SES and suicidality.13,16,31,32  This should be a population of focus to enhance preventive mental health care and follow-up efforts.

Other factors that may have been affected by the pandemic are LOS and reutilization. The average LOS was longer postlockdown, which is likely multifactorial. The medical and psychiatric needs of the patients presenting postpandemic may have been higher, necessitating a longer LOS, as also observed in the literature.33,34  Additionally, because of fewer team members delivering care in person (many of our social work team members were only performing virtual safety assessments at the beginning of the pandemic), LOS could have increased because of discharge planning needs of these patients, as well (ie, time taken to arrange transfer of care to an outpatient psychiatric facility, lack of bed availability at inpatient psychiatric facilities prolonging LOS, and then devising an appropriate safety plan as an alternative discharge solution).

Our local inpatient psychiatric facilities decreased their bed availability because of multiple factors, the largest one being pandemic-related staffing shortages making it more difficult to transfer patients to inpatient psychiatric facilities. However, we did transfer our most acute patients to inpatient psychiatric facilities, regardless of the timing for a bed to become available. With the lack of inpatient psychiatric beds, it is not surprising that there was a statistical significance of patients being discharged from the hospital more frequently postlockdown. In addition, longer medical hospitalizations with mental health consultation (for some of these patients) may have allowed for emotional stabilization and an ability for these patients to be discharged from the hospital. Patient and family preference may have also factored in this shift, because of the fears of contracting COVID-19 at the facilities. Although some may believe that the shift in discharging patients from the hospital after being hospitalized for an SA may not be in the best interest of the patients, we believe that it is a testament to our highly skilled MBH providers (psychiatrists, psychologists, and social workers) and their ability to safety-plan with lower risk patients. Notably, as seen in Table 3, a larger proportion of patients with multiple reutilizations postlockdown were discharged to a higher level of care, indicating their complexity and ongoing mental health needs.

Interestingly, in contrast to other studies examining 30-day reutilization rates, our 30-day reutilization rates (both ED and readmissions) for MBH-related complaints both pre- and postlockdown were lower, ranging from ∼1.4% to 5% for 1 re-presentation to either the ED or hospital setting, in comparison with the 8% to 22% 30-day reutilization rates reported in the literature.2025  Our pre- and postlockdown results revealed no significant differences in 7- and 30-day reutilization. However, there was a noteworthy increase in reutilization rates over a 365-day period lockdown, with a significant increase postlockdown, once again suggesting the profound impact of the COVID-19 pandemic on this population. The reutilization seen within 1 year could indicate persistent and severe depression, ongoing stressors, worsening mental health related to the pandemic, and/or inadequate community resources, among a multitude of other issues. Overall, our high rates of reutilization for MBH complaints indicates a sustained vulnerability in the year after initial presentation.

There was chronological association between the COVID-19 variant surges and admission frequency, with the admissions increasing after the alpha variant surge and then, after a brief trough, sky-rocketed after the delta surge in August of 2021. It is unclear whether the large surge was a cumulative effect of the previous surges, or delayed effects of the initial lockdown. It is also unclear how seasonal change contributed to our data, because previous literature has revealed seasonal trends in pediatric MBH presentations to the ED.35  Nonetheless, this temporal depiction suggests a link between the dynamics of the pandemic and the mental health crisis in our youth, also supported by other evidence on postlockdown effects on overall mental health and the pandemic’s effects on suicidality and SAs.

Limitations of our study included our data being from a single tertiary care hospital in a Midwestern city, which may or may not be generalizable to other populations. To collect accurate reutilization data, manual chart review was performed and was reliant on 4 different reviewers’ data collection. We recognize that reutilization may be underreported given that not every institution has access to an EHR, which is how reutilization data were captured. Furthermore, by narrowing our reutilization data to 1 encounter pre- and 1 postlockdown, we may have inadvertently missed multiple presentations beyond the initial 365 days that we performed for chart reviews. Additionally, the timeline of our COVID-19 surges are different than other Southern states. Our race and ethnicity data relied upon patients and/or family’s self-report of their race and ethnicity because the question may be asked multiple different ways when it was imported into the EHR. Also, we used sex at the time of chart review, and our health system and EHR are still behind in appropriately identifying transgender and nonbinary youth.

On the basis of our results that, postlockdown, more patients were both (1) being discharged from the hospital and (2) re-presented to our medical hospital for MBH-related issues within 365 days of discharge, suggest that future efforts should target those patients being discharged from the hospital after their initial hospitalization. Although the need to ensure these patients are established with mental health care posthospitalization is paramount, future research efforts could specifically examine if there was a relationship between the intensity of outpatient care and re-presentation or what specific social or demographic factors were related to readmission. In addition, as mentioned above, additional research is needed locally to determine which racial and ethnic groups are being disproportionately affected and then targeted interventions can be developed.

In conclusion, our comprehensive, retrospective analysis of patients presenting with suicidality demonstrated that patients presenting to our medical tertiary care hospital after the COVID-19 pandemic had longer hospitalizations, were discharged from the hospital back to their homes, and had more re-presentations to our hospital for MBH-related issues, as compared with those presenting 2 years before the pandemic’s onset. In addition, our results suggest that the later COVID-19 variants resulted in increased admissions for suicidality. Overall, our findings underscore the profound impact the COVID-19 pandemic has had on our youth and provide us with data to drive future targeted interventions for the patients hospitalized with suicidality.

We thank Dr Robert Treat, professor in academic affairs, for assisting us with statistical inquiries. We also thank Dr Sarah Bauer for internal review of our manuscript. We also thank Ms Daliborka Radic, inpatient clinical analytics manager, for her work in initial data collection. Lastly, we thank Children’s Hospital and our Department of Pediatrics leadership for support of our initiative.

Ms Dellazoppa and Dr Vepraskas contributed to the conceptualization of the study design, performed data collection, participated in qualitative data analysis, drafted the initial manuscript, and reviewed and assessed statistical analysis; Dr Zaspel contributed to the conceptualization of the study design, performed data collection, participated in qualitative data analysis, and reviewed and revised the manuscript; Ms Porada contributed to the conceptualization of the study design, performed data collection, participated in qualitative data analysis, reviewed and revised the manuscript, and created and performed statistical analysis; Dr Bourgeois performed data collection, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and are accountable for all aspects of the work.

FUNDING: No external funding.

CONFLICT OF INTEREST DISCLOSURES: The authors have indicated they have no conflicts of interest relevant to this article to disclose.

1
Bardach
NS
,
Coker
TR
,
Zima
BT
, et al
.
Common and costly hospitalizations for pediatric mental health disorders
.
Pediatrics
.
2014
;
133
(
4
):
602
609
2
Plemmons
G
,
Hall
G
,
Doupnik
S
, et al
.
Hospitalization for suicide ideation or attempt: 2008–2015
.
Pediatrics
.
2017
;
141
(
6
):
e20172436
3
Titus
L
,
Stephany
A
,
Porada
K
,
McFadden
V
.
Intent, substance, and care: characteristics of adolescent ingestion hospitalizations
.
Hosp Pediatr
.
2021
;
11
(
2
):
160
166
4
Bridge
JA
,
Ruch
DA
,
Sheftall
AH
, et al
.
Youth suicide during the first year of the COVID-19 pandemic
.
Pediatrics
.
2023
;
151
(
3
):
e2022058375
5
Bruns
N
,
Willemsen
L
,
Stang
A
, et al
.
Pediatric ICU admissions after adolescent suicide attempts during the pandemic
.
Pediatrics
.
2022
;
150
(
2
):
e2021055973
6
Friedman
J
,
Godvin
M
,
Shover
CL
,
Gone
JP
,
Hansen
H
,
Schriger
DL
.
Trends in drug overdose deaths among US adolescents, January 2010 to June 2021
.
JAMA
.
2022
;
327
(
14
):
1398
1400
7
Goldstick
JE
,
Cunningham
RM
,
Carter
PM
.
Current causes of death in children and adolescents in the United States
.
N Engl J Med
.
2022
;
386
(
20
):
1955
1956
8
Raffa
BJ
,
Schilling
S
,
Henry
MK
, et al
.
Ingestion of illicit substances by young children before and during the COVID-19 pandemic
.
JAMA Netw Open
.
2023
;
6
(
4
):
e239549
9
Zima
BT
,
Edgcomb
JB
,
Rodean
J
, et al
.
Use of acute mental health care in US children’s hospitals before and after statewide COVID-19 school closure orders
.
Psychiatr Serv
.
2022
;
73
(
11
):
1202
1209
10
Clark
E
,
Fredricks
K
,
Woc-Colburn
L
,
Bottazzi
ME
,
Weatherhead
J
.
Disproportionate impact of the COVID-19 pandemic on immigrant communities in the United States
.
PLoS Negl Trop Dis
.
2020
;
14
(
7
):
e0008484
11
Khanijahani
A
,
Iezadi
S
,
Gholipour
K
,
Azami-Aghdash
S
,
Naghibi
D
.
A systematic review of racial/ethnic and socioeconomic disparities in COVID-19
.
Int J Equity Health
.
2021
;
20
(
1
):
248
12
Magesh
S
,
John
D
,
Li
W
, et al
.
Disparities in COVID-19 outcomes by race, ethnicity, and socioeconomic status: a systematic-review and meta-analysis
.
JAMA Netw Open
.
2021
;
4
(
11
):
e2134147
13
Office of the Surgeon G. Publications and Reports of the Surgeon General
.
Protecting Youth Mental Health: The US Surgeon General’s Advisory
.
Washington, D.C.
:
US Department of Health and Human Services
;
2021
14
Rider
E
,
Ansari
E
,
Varrin
P
,
Sparrow
J
.
Mental health and wellbeing of children and adolescents during the covid-19 pandemic
.
BMJ
.
2021
;
374
:
n1730
15
Schnitzer
PG
,
Dykstra
H
,
Collier
A
.
The COVID-19 pandemic and youth suicide: 2020–2021
.
Pediatrics
.
2023
;
151
(
3
):
e2022058716
16
Bridge
JA
,
Horowitz
LM
,
Fontanella
CA
, et al
.
Age-related racial disparity in suicide rates among US youths from 2001 through 2015
.
JAMA Pediatr
.
2018
;
172
(
7
):
697
699
17
Beck
AT
,
Steer
RA
,
Kovacs
M
,
Garrison
B
.
Hopelessness and eventual suicide: a 10-year prospective study of patients hospitalized with suicidal ideation
.
Am J Psychiatry
.
1985
;
142
(
5
):
559
563
18
O’Connor
RC
,
Smyth
R
,
Ferguson
E
,
Ryan
C
,
Williams
JM
.
Psychological processes and repeat suicidal behavior: a four-year prospective study
.
J Consult Clin Psychol
.
2013
;
81
(
6
):
1137
1143
19
Suominen
K
,
Isometsä
E
,
Suokas
J
,
Haukka
J
,
Achte
K
,
Lönnqvist
J
.
Completed suicide after a suicide attempt: a 37-year follow-up study
.
Am J Psychiatry
.
2004
;
161
(
3
):
562
563
20
Doupnik
S
,
Rodean
J
,
Zima
BT
, et al
.
Readmissions after pediatric hospitalization for suicide ideation and suicide attempt
.
J Hosp Med
.
2018
;
13
(
11
):
743
751
21
Bardach
NS
,
Doupnik
SK
,
Rodean
J
, et al
.
ED visits and readmissions after follow-up for mental health hospitalization
.
Pediatrics
.
2020
;
145
(
6
):
e20192872
22
Feng
JY
,
Toomey
SL
,
Zaslavsky
AM
,
Nakamura
MM
,
Schuster
MA
.
Readmission after pediatric mental health admissions
.
Pediatrics
.
2017
;
140
(
6
):
e20171571
23
Heslin
KC
,
Weiss
AJ
.
Hospital Readmissions Involving Psychiatric Disorders, 2012. Healthcare Cost and Utilization Project (HCUP) Statistical Briefs
.
Washington, D.C
.:
Agency for Healthcare Research and Quality
;
2006
24
Hoffmann
JA
,
Krass
P
,
Rodean
J
, et al
.
Follow-up after pediatric mental health emergency visits
.
Pediatrics
.
2023
;
151
(
3
):
e2022057383
25
Leon
SL
,
Cloutier
P
,
Polihronis
C
, et al
.
Child and adolescent mental health repeat visits to the emergency department: a systematic review
.
Hosp Pediatr
.
2017
;
7
(
3
):
177
186
26
Gay
JC
,
Zima
BT
,
Coker
TR
, et al
.
Postacute care after pediatric hospitalizations for a primary mental health condition
.
J Pediatr
.
2018
;
193
:
222
228.e1
27
National Committee for Quality Assurance
.
HEDIS measures and technical resources
. Available at: https://www.ncqa.org/hedis/measures/. Accessed January 23, 2024
28
Frey
W
.
Black-white segregation edges downward since 2000, census shows
. Available at: https://www.brookings.edu/articles/black-white-segregation-edges-downward-since-2000-census-shows/. Accessed January 25 2024
29
Hill
RM
,
Rufino
K
,
Kurian
S
,
Saxena
J
,
Saxena
K
,
Williams
L
.
Suicide ideation and attempts in a pediatric emergency department before and during COVID-19
.
Pediatrics
.
2021
;
147
(
3
):
e2020029280
30
Krass
P
,
Dalton
E
,
Doupnik
S
,
Esposito
J
.
US pediatric emergency department visits for mental health condition during the covid-19 pandemic
.
JAMA Netw Open
.
2021
;
4
(
4
):
e218533
31
Maffly-Kipp
J
,
Eisenbeck
N
,
Carreno
DF
,
Hicks
J
.
Mental health inequalities increase as a function of COVID-19 pandemic severity levels
.
Soc Sci Med
.
2021
;
285
:
114275
32
Mullins
N
,
Kang
J
,
Campos
AI
, et al
.
Major Depressive Disorder Working Group of the Psychiatric Genomics Consortium
;
Bipolar Disorder Working Group of the Psychiatric Genomics Consortium
;
Eating Disorders Working Group of the Psychiatric Genomics Consortium
;
German Borderline Genomics Consortium
;
MVP Suicide Exemplar Workgroup
;
VA Million Veteran Program
.
Dissecting the shared genetic architecture of suicide attempt, psychiatric disorders, and known risk factors
.
Biol Psychiatry
.
2022
;
91
(
3
):
313
327
33
Ibeziako
P
,
Kaufman
K
,
Scheer
KN
,
Sideridis
G
.
Pediatric mental health presentations and boarding: first year of the COVID-19 pandemic
.
Hosp Pediatr
.
2022
;
12
(
9
):
751
760
34
Leith
T
,
Brieger
K
,
Malas
N
, et al
.
Increased prevalence and severity of psychiatric illness in hospitalized youth during COVID-19
.
Clin Child Psychol Psychiatry
.
2022
;
27
(
3
):
804
812
35
Marshall
R
,
Ribbers
A
,
Sheridan
D
,
Johnson
KP
.
Mental health diagnoses and seasonal trends at a pediatric emergency department and hospital, 2015–2019
.
Hosp Pediatr
.
2021
;
11
(
3
):
199
206