To evaluate whether admission on weekends affects the length of stay (LOS) for patients hospitalized with somatic symptom and related disorders (SSRDs).
Data from 2012–2018 was obtained for all patients aged 4 to 21 years (N = 5459) with a primary discharge diagnosis of SSRDs from 52 tertiary care pediatric hospitals in the United States. We obtained patient demographics, admission date and/or time, LOS, procedure count, and comorbid conditions. We defined a weekend as 3 pm Friday to 3 pm Sunday. The Wilcoxon rank test was used for unadjusted analysis. Multiple logistic regression was used to estimate the odds of having LOS >1 day, >2 days, >3 days, and >4 days in weekend versus weekday groups.
Weekend admission significantly correlated with increased LOS (P < .001). Compared with weekdays, a weekend admission was associated with increased odds of having LOS >1, >2, and >3 days. This remained statistically significant while adjusting for the number of chronic conditions, procedures, and individuals with Black or Hispanic ethnicity compared with White ethnicity. LOS was not associated with sex or age of the patients.
Patients with SSRDs admitted on the weekend have an increased LOS compared with those admitted on a weekday. This may be due to a decrease in multidisciplinary care available during weekends. In future studies, researchers should aim to better understand the specific factors that contribute to this disparity and test interventions that may close the gap in care, including expanding to 7-day services, increasing mental health resources, and working to decrease the need for inpatient admissions.
Somatic symptom and related disorders (SSRDs) are highly prevalent among the pediatric population. It is estimated that somatic complaints contribute to 15% to 20% of the burden of yearly pediatric health care costs in the United States, involving multiple unnecessary laboratory tests, imaging, subspecialist referrals, and hospitalizations.1,2 Although data are limited on prevalence, particularly in nonprimary care settings, estimates range from 17% to 50% of school-aged children and adolescents and potentially 20% of children aged 3 to 5 years have had, or do have, an SSRD.3 Despite this prevalence, SSRDs are currently underrecognized, and even after appropriate diagnosis, a chasm exists between the patient needs and the health care system’s availability to care for patients with these disorders.
Ideally, patients with SSRDs are identified on an outpatient basis and are managed with close liaison between medical and mental health professionals.4–6 If in the hospital setting, the evaluation of SSRD calls for early recognition by health care providers and early multidisciplinary assessment, with engagement of the child’s family, school, and primary care provider.5–8 Delays in mental health involvement lead to patients and families perceiving their care as being handed off to mental health specialists and further stigmatizes the condition.9 This, coupled with the risk for iatrogenic injury while in the hospital, reveals the need for prompt, early mental health involvement and discharge from the inpatient setting.
Occasionally, cases present in which hospitalization is indicated to optimize care or get buy-in from major stakeholders, particularly parents. When the medical workup is extensive, this can perpetuate the disease process and impose a heavy burden on health care systems.10–13 A review of 60 hospitalized patients with SSRDs by Gao et al revealed that 90% of admissions followed presentation in the emergency department, with a median admission length of 3 days, although 23% of admissions were longer than 6 days.10
Studies have also revealed that the earlier mental health providers are involved in a patient’s care, the shorter the length of stay (LOS) and decrease in cost.8,14 In 2019, Ibeziako et al published national guidelines for the inpatient management of SSRDs and recommended “early interdisciplinary assessment.” However, it has been revealed that generally all hospital staff experience a sharp decline in workforce numbers across a weekend.15 Particularly, psychology, psychiatry, physical therapy and/or occupational therapy, and social work services frequently have decreased staffing in the hospital on the these days15–22 yet are integral to the multidisciplinary care for patients with SSRDs and are necessary to ensure a shorter LOS.6
Even with diagnoses that require admission for primary medical indications, such as failure-to-thrive, pediatric leukemia, and heart failure, studies have revealed an increase in LOS with admission on a weekend.23–25 In this study, we aim to evaluate whether admission on weekends affects the LOS, and thereby quality of care, for patients hospitalized with SSRDs in the Pediatric Health Information System (PHIS) database.
Methods
Data for this study were obtained from PHIS, an administrative database that contains inpatient, emergency department, ambulatory surgery, and observation encounter-level data from >52 not-for-profit, tertiary care pediatric hospitals in the United States. These hospitals are affiliated with the Children’s Hospital Association (Lenexa, KS). Data quality and reliability are assured through a joint effort between the Children’s Hospital Association and participating hospitals. Portions of the data submission and data quality processes for the PHIS database are managed by IBM Watson Health (Ann Arbor, MI). For the purposes of external benchmarking, participating hospitals provide discharge and/or encounter data, including demographics, diagnoses, and procedures. Nearly all of these hospitals also submit resource use data (eg, pharmaceuticals, imaging, and laboratory) into PHIS. Data are deidentified at the time of data submission, and data are subjected to a number of reliability and validity checks before being included in the database. For this study, data from 52 hospitals were included.
Data Extraction
Our inclusion criteria were patients between the ages of 4 and 21 years, primary diagnosis of SSRD (Table 1), and inpatient or observation status at the time of discharge during 2012–2018. We extracted data from these admissions, including sociodemographic characteristics, temporal characteristics (admission month, weekend versus weekday, and season), number of procedures, secondary and/or coexisting diagnosis, and complex chronic conditions.26 We removed subsequent admissions from the same patient number after their first diagnostic admission. We included only those patients with a LOS >1 day because >1-day admissions are a necessity for patients with an SSRD diagnosis to include a multidisciplinary approach.
Demographics
. | Weekday (n = 4365) . | Weekend (n = 1094) . | P . | Test . |
---|---|---|---|---|
LOS, mean (SD) | 2.83 (4.01) | 3.23 (6.77) | <.001 | Wilcoxon rank sum |
Admission age, y, mean (SD) | 13.87 (2.87) | 14.02 (2.95) | .111 | t test |
Chronic condition count, mean (SD) | 0.41 (0.94) | 0.37 (0.89) | .139 | Wilcoxon rank sum |
Procedure count, mean (SD) | 0.48 (1.04) | 0.49 (1.01) | .470 | Wilcoxon rank sum |
Sex, n (%) | .778 | χ2 | ||
Female | 3141 (72) | 780 (71.3) | ||
Male | 1222 (28) | 313 (28.6) | ||
Race and/or ethnicity, n (%) | .058 | Fisher’s exact | ||
White, non-Hispanic | 2517 (57.7) | 621 (56.8) | ||
Hispanic or Latino, nonwhite | 522 (12) | 134 (12.2) | ||
Black | 912 (20.9) | 212 (19.4) | ||
Asian | 64 (1.5) | 12 (1.1) | ||
Pacific Islander | 12 (0.3) | 1 (0.1) | ||
American Indian | 16 (0.4) | 9 (0.8) | ||
Unknown | 322 (7.4) | 105 (9.6) |
. | Weekday (n = 4365) . | Weekend (n = 1094) . | P . | Test . |
---|---|---|---|---|
LOS, mean (SD) | 2.83 (4.01) | 3.23 (6.77) | <.001 | Wilcoxon rank sum |
Admission age, y, mean (SD) | 13.87 (2.87) | 14.02 (2.95) | .111 | t test |
Chronic condition count, mean (SD) | 0.41 (0.94) | 0.37 (0.89) | .139 | Wilcoxon rank sum |
Procedure count, mean (SD) | 0.48 (1.04) | 0.49 (1.01) | .470 | Wilcoxon rank sum |
Sex, n (%) | .778 | χ2 | ||
Female | 3141 (72) | 780 (71.3) | ||
Male | 1222 (28) | 313 (28.6) | ||
Race and/or ethnicity, n (%) | .058 | Fisher’s exact | ||
White, non-Hispanic | 2517 (57.7) | 621 (56.8) | ||
Hispanic or Latino, nonwhite | 522 (12) | 134 (12.2) | ||
Black | 912 (20.9) | 212 (19.4) | ||
Asian | 64 (1.5) | 12 (1.1) | ||
Pacific Islander | 12 (0.3) | 1 (0.1) | ||
American Indian | 16 (0.4) | 9 (0.8) | ||
Unknown | 322 (7.4) | 105 (9.6) |
Data Analysis
Administrative data were obtained for all patients aged 4 to 21 years with a primary admission diagnosis of SSRD. Demographic characteristics, day of admission, LOS, complex conditions, number of procedures, and comorbid conditions were analyzed. Patients were classified as weekend and weekday admission by definition of weekend as Friday at 3 pm to Sunday at 3 pm. The Wilcoxon rank test was used for unadjusted analysis. LOS was measured as the difference between discharge date and admission date, with a minimum value of 1 day (ie, same day or 1-day gap was considered 1 day, 2 days difference considered 2 days, etc). We created 4 logistic regression models assessing impact of weekend and/or weekday admission, race and ethnicity, sex, age, chronic condition count, and procedure count on LOS at cut points (>1 day versus ≤1 day, >2 days versus ≤2 days, >3 days versus ≤3 days, and >4 days versus ≤4 days). Data analysis was performed by using R version 4.0.2.
Results
There were 5459 admissions with a primary diagnosis of SSRD, with 4365 (80%) admissions occurring on a weekday and 1094 (20%) on the weekend. The median (interquartile range) LOS was 2 (1–3) days (Table 1). Weekend admission was significantly associated with an increased LOS (P < .001, Wilcoxon rank test) (Table 1). The multivariable logistic regression results are presented in Table 2. Broken down further, the odds of having LOS >1 day was associated with weekend admission (P < .001; odds ratio [OR] = 1.30, 95% confidence interval [CI]: [1.13–1.50]), adjusting for chronic condition count (P < .001; OR = 1.16, 95% CI: [1.09–1.24]), procedure count (P < .001; OR = 1.74, 95% CI: [1.59–1.89]), and Black (P = .03, OR = 1.18 95% CI: [1.02–1.36]) and Hispanic (P = .002; OR = 1.33, 95% CI: [1.12–1.60]) ethnicity compared with White ethnicity. Chronic condition count was adjusted for known association with longer LOS with greater complexity of the patient,27 and procedure count was adjusted to avoid skewing of data because of contribution of admission length from procedure recovery time.
Multivariable Models Comparing LOS Cut Points
. | >1 d/≤1 d . | >2 d/≤2 d . | >3 d/≤3 d . | >4 d/≤4 d . | ||||
---|---|---|---|---|---|---|---|---|
OR (95% CI) . | P . | OR (95% CI) . | P . | OR (95% CI) . | P . | OR (95% CI) . | P . | |
Weekend | 1.3 (1.13–1.5) | <.001 | 1.56 (1.36–1.8) | <.001 | 1.3 (1.1–1.53) | .002 | 1.16 (0.96–1.4) | .128 |
Chronic condition count | 1.16 (1.09–1.24) | <.001 | 1.15 (1.09–1.23) | <.001 | 1.24 (1.16–1.33) | <.001 | 1.23 (1.14–1.32) | <.001 |
Procedure count | 1.73 (1.59–1.89) | <.001 | 1.59 (1.49–1.7) | <.001 | 1.51 (1.42–1.61) | <.001 | 1.47 (1.38–1.57) | <.001 |
Hispanic or Latino, nonwhite | 1.33 (1.11–1.6) | .002 | 1.35 (1.12–1.61) | .001 | 1.23 (1–1.52) | .049 | 0.94 (0.73–1.21) | .652 |
Black or Black/Hispanic | 1.18 (1.02–1.36) | .025 | 1.16 (0.99–1.34) | .058 | 1.09 (0.91–1.3) | .332 | 1.02 (0.83–1.24) | .851 |
Asian or Asian/Hispanic | 0.72 (0.45–1.16) | .180 | 1.1 (0.65–1.79) | .717 | 0.83 (0.43–1.49) | .553 | 0.71 (0.31–1.43) | .375 |
Pacific Islander or Pacific Islander/Hispanic | 1.13 (0.36–3.85) | .838 | 1.44 (0.42–4.5) | .539 | 1.19 (0.26–4.08) | .799 | 1.09 (0.16–4.23) | .917 |
American Indian or American Indian/Hispanic | 1.26 (0.56–3.02) | .584 | 0.85 (0.32–1.99) | .719 | 0.81 (0.23–2.15) | .698 | 0.9 (0.21–2.64) | .864 |
Unknown race | 1.01 (0.81–1.24) | .962 | 1.09 (0.87–1.35) | .454 | 1.03 (0.79–1.32) | .846 | 0.85 (0.62–1.15) | .297 |
Male sex | 1 (0.88–1.13) | .985 | 0.94 (0.82–1.07) | .352 | 0.96 (0.82–1.11) | .561 | 0.99 (0.83–1.18) | .949 |
Admit age, y | 1.01 (0.99–1.03) | .417 | 1.01 (0.99–1.03) | .270 | 1 (0.98–1.02) | .972 | 1.01 (0.98–1.04) | .601 |
. | >1 d/≤1 d . | >2 d/≤2 d . | >3 d/≤3 d . | >4 d/≤4 d . | ||||
---|---|---|---|---|---|---|---|---|
OR (95% CI) . | P . | OR (95% CI) . | P . | OR (95% CI) . | P . | OR (95% CI) . | P . | |
Weekend | 1.3 (1.13–1.5) | <.001 | 1.56 (1.36–1.8) | <.001 | 1.3 (1.1–1.53) | .002 | 1.16 (0.96–1.4) | .128 |
Chronic condition count | 1.16 (1.09–1.24) | <.001 | 1.15 (1.09–1.23) | <.001 | 1.24 (1.16–1.33) | <.001 | 1.23 (1.14–1.32) | <.001 |
Procedure count | 1.73 (1.59–1.89) | <.001 | 1.59 (1.49–1.7) | <.001 | 1.51 (1.42–1.61) | <.001 | 1.47 (1.38–1.57) | <.001 |
Hispanic or Latino, nonwhite | 1.33 (1.11–1.6) | .002 | 1.35 (1.12–1.61) | .001 | 1.23 (1–1.52) | .049 | 0.94 (0.73–1.21) | .652 |
Black or Black/Hispanic | 1.18 (1.02–1.36) | .025 | 1.16 (0.99–1.34) | .058 | 1.09 (0.91–1.3) | .332 | 1.02 (0.83–1.24) | .851 |
Asian or Asian/Hispanic | 0.72 (0.45–1.16) | .180 | 1.1 (0.65–1.79) | .717 | 0.83 (0.43–1.49) | .553 | 0.71 (0.31–1.43) | .375 |
Pacific Islander or Pacific Islander/Hispanic | 1.13 (0.36–3.85) | .838 | 1.44 (0.42–4.5) | .539 | 1.19 (0.26–4.08) | .799 | 1.09 (0.16–4.23) | .917 |
American Indian or American Indian/Hispanic | 1.26 (0.56–3.02) | .584 | 0.85 (0.32–1.99) | .719 | 0.81 (0.23–2.15) | .698 | 0.9 (0.21–2.64) | .864 |
Unknown race | 1.01 (0.81–1.24) | .962 | 1.09 (0.87–1.35) | .454 | 1.03 (0.79–1.32) | .846 | 0.85 (0.62–1.15) | .297 |
Male sex | 1 (0.88–1.13) | .985 | 0.94 (0.82–1.07) | .352 | 0.96 (0.82–1.11) | .561 | 0.99 (0.83–1.18) | .949 |
Admit age, y | 1.01 (0.99–1.03) | .417 | 1.01 (0.99–1.03) | .270 | 1 (0.98–1.02) | .972 | 1.01 (0.98–1.04) | .601 |
Adjusting for the same covariates, LOS >2 days was associated with weekend admission (P < .001; OR = 1.56, 95% CI: [1.36–1.80]), as well as LOS >3 days (P = .002; OR = 1.30, 95% CI: [1.10–1.53]), and LOS >4 days was positively associated with weekend admission (P = .128; OR = 1.16, 95% CI: [0.96–1.4]), although LOS > 4 days was found to not be statistically significant. Increased LOS was not associated with Asian, Pacific Islander, or American Indian ethnicity or with sex or age of the patients. Each model was assessed for goodness of fit by using the Hosmer-Lemeshow test and was determined to have good fit (P = .400, 0.135, 0.194, and 0.285 for LOS >1, >2, >3, and > 4 days, respectively).
Discussion
This study reveals that weekend admissions for patients with SSRDs were significantly associated with an increased LOS, even when adjusting for the number of chronic conditions and procedures performed. As SSRD is now a well-defined entity with interdisciplinary treatment pathways confirmed to ameliorate disease, and in some cases cure, the case for expanded 7-day services and improved outpatient mental health services is becoming more obvious. Currently, patients with SSRDs often arrive at the diagnosis after prolonged testing and medicalization, leading to delay in appropriate treatment, decreased family acceptance of the diagnosis, and prolonged recovery,6,7,28–30 making further adherence to inpatient SSRD national guidelines6 a necessity. This adherence should not be delayed or skipped because of a lack of multidisciplinary availability because doing so delays discharge and may further cause iatrogenic harm to the patient.6 This request for improved availability is made more relevant because somatization is the second leading reason for consultation requests received by child and adolescent psychiatrists in pediatric hospitals.31
Hospital systems also have a vested interest in preventing prolonged LOS given issues with capacity management and the increased cost attributed to each additional inpatient day. As previously mentioned, a recently published evidence-informed clinical pathway from an expert group of pediatric psychiatrists reinforces the need for relevant early interdisciplinary assessment by psychology, psychiatry, physical therapy and/or occupational therapy, and social work for all patients admitted to the hospital with SSRDs.6 A single-center study by Bujoreanu et al revealed that earlier psychiatry consultation by medical and surgical units was associated with shorter LOS in pediatric hospitalization, including patients with SSRDs, and more recent research reveals that early consultation with these services decreases the overall cost to the health system.8,14 Unfortunately, these important resources are often limited on the weekend.16,17,20 Similar to other 7-day service studies for different patient populations, including failure-to-thrive, pediatric leukemia, and adult heart failure, revealing increased LOS with weekend admission day,32–34 increased access to these consultation teams may provide expedited, improved quality of care and a subsequent decrease in the LOS for patients with SSRDs.21–25 Increasing bed availability by decreasing LOS has been particularly relevant during the past year as the COVID-19 pandemic led to an increase in pediatric bed demand because of increased admissions for pediatric psychiatric disorders, along with the need to accommodate adult patients with COVID-19 on inpatient pediatric units.35,36
However, increasing interdisciplinary services to 7 days a week may not be a feasible short-term solution given long-standing workforce shortages of child psychiatrists, pediatric psychologists, and other mental health professionals.37 Existing medical providers need to prioritize the rising psychiatric presentations with acute safety concerns for weekend admissions. As some patients with SSRDs are planned admissions or nonurgent, a recommendation to discourage such admissions from occurring on weekends will be an immediate remedy that could lead to a decreased LOS and improved patient and family experience.
Moreover, the increasing recognition and prevalence of SSRDs underscores our society’s great need for development of expansive collaboration among mental health and medical providers. To truly combat SSRD-related admissions, it is essential to provide today’s pediatric population with increased mental health resources easily accessible in the patient’s medical home, as well as provide generous support to our mental health professionals, to decrease overall stigma regarding mental health conditions and encourage the pursuit of prompt therapies.38,39
There are several limitations to our study. First, the nature of SSRDs makes diagnosis difficult. It is possible there is a larger population of patients who were discharged with another diagnosis but later found to have an SSRD. Although there is a set of International Classification of Diseases, 10th Revision, codes for SSRD, there is no validated study that reveals which codes are accurate for true SSRD admission via chart review, which also could have limited our study (Supplemental Table 3). In addition, our data set had 65 outliers with LOS >20 days. We used the Wilcoxon rank test in our analysis to take these outliers into account. Furthermore, the patients who are admitted on weekends may have other factors that could contribute to LOS that cannot be detected from this database, such as more severe mental health comorbidities or challenging family situations. We also recognize that in our study, we do not account for holidays that fall on weekdays but essentially function at the same level of resources as a weekend, which may decrease the strength of the significance of our results. In addition, by excluding patients with LOS <1 day, the number of patients admitted over a weekend but ultimately discharged because they could not receive multidisciplinary care will not be represented. Lastly, we were not able to identify which patients may have had sedated imaging, which may contribute to a longer LOS.
Conclusions
Similar to patients with other disease processes, patients with SSRD admitted on the weekend have an increased LOS compared with those admitted on a weekday, even when adjusting for the number of chronic conditions and procedures performed. This may be due to a decrease in multidisciplinary care available during weekends.16,17,20,22 In future studies, researchers should aim to better understand the specific factors that contribute to this disparity and test interventions that may close the gap in care, including expanding to 7-day services and working to decrease the need for inpatient admissions by improving mental health outpatient care and access.
Acknowledgments
We thank the University of Michigan Department of Pediatrics Charles Woodson Clinical Research Fund.
FUNDING: No external funding.
Dr Rappaport conceptualized and designed the study, supervised data collection, drafted the initial manuscript, and reviewed and revised the manuscript; Dr VanderVennen participated in the literature review, in addition to drafting the initial manuscript, and reviewed and revised the manuscript; Mr McCaffery conducted the initial analyses, drafted initial methods and results, and reviewed and revised the manuscript; Dr Monroe participated in critically reviewing the manuscript, editing and revising the manuscript, as well as approving the final manuscript; Dr Stewart conceptualized and designed the study and critically reviewed the manuscript for important intellectual content; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
References
Competing Interests
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
Comments