Transcript View Poster - pantherFILE - University of Wisconsin–Milwaukee

Impact of Psychiatric Diagnoses on Inpatient Health Care
Utilization in Sickle Cell Disease
Lisa M. Burks, BA1 & Matthew P Myrvik PhD2,
1 University
of Wisconsin-Milwaukee
2Medical College of Wisconsin Department of Pediatrics, Milwaukee WI
• Sickle cell disease (SCD) in children has been associated with
physical limitations, neurological complications, frequent school
absences, limited opportunities for socialization, and feelings of
helplessness and self-consciousness (Brown et al., 1993).
• Children with SCD are found to exhibit more psychiatric
symptoms than healthy peers and other chronic medical
conditions (Bennett, 1994; Cepeda et al., 1997).
• Psychiatric symptoms have been found to negatively impact the
course of a number of pediatric chronic illnesses (Brand et al.,
1986; von Weiss, 2002).
•Within SCD, psychiatric symptoms may impact a child’s ability
to manage pain, which may complicate admissions and increase
the frequency of admissions for SCD pain.
OBJECTIVES
Specific Aim 1: Determine the impact of psychiatric disorders on
the hospital length of stay (LOS) in pediatric patients diagnosed
with SCD admitted for a vaso-occlusive pain event.
Specific Aim 2: Determine the impact of psychiatric disorders on
inpatient healthcare utilization in pediatric patients diagnosed with
SCD admitted for a vaso-occlusive pain.
METHODS
Participant Selection
•Discharge data acquired through the nationally representative
Pediatric Health Information System (PHIS) from January 2005
through December 2011.
• Ages 5 to 18 years.
• Primary discharge diagnosis of sickle cell crisis (ICD-9 code
282.62) .
• Children with other sickle cell-related diagnoses as a primary
diagnoses were excluded from analysis.
Independent Variable
• Psychiatric Diagnosis: Patients with any psychiatric diagnosis
listed under secondary diagnoses.
• Psychiatric Diagnosis Subtypes: Mood disorder, anxiety
disorder, disruptive behavior disorder, and substance use disorder.
Dependent Variable
•Length of Stay: Measured in days from date of discharge.
•Frequency of Inpatient Admission: Number of inpatient
admissions for vaso-occlusive pain event.
Confounding Variables
• Patient Demographics: Patient gender, age, and primary payor for
hospital services.
• Medical Variables: Acute chest syndrome, asthma, and disease
severity.
METHODS
TABLE I (Mean LOS in days)
Data Analysis
• LOS and frequency of inpatient admission data was collected for psychiatric disorder
groups only after initial visit with documented psychiatric comorbidity.
•LOS was positively skewed so LOS was adjusted using log transformation.
• T-tests were used to examine differences in LOS by psychiatric diagnosis category.
• Linear regression analysis used to examine effect of psychiatric diagnoses on adjusted
LOS while controlling for significant covariates.
• Parameter estimates were performed and results presented as estimates of the relative
difference between groups.
RESULTS
• 9,533 pediatric patients with a primary diagnosis of SCD with crisis accounted for 34,855 of
recorded visits.
• Mean age was 11.75 years (±4.95 years) and 47.62% of the patients discharged were female.
• Mean unadjusted LOS was 4.09 days (±5.60 days).
• Roughly 5% (n=482) of the patients with SCD with crisis discharged had a psychiatric
diagnosis and accounted for 2,645 total visits.
• Psychiatric diagnosis subtype estimates include mood disorder (2%), anxiety disorder
(2%), disruptive behavior disorder (2%), and substance use disorder (1%).
Psychiatric Diagnosis and LOS: Mean LOS for sickle cell pain events increased by 1.83
days in patients with a psychiatric disorder relative to patients without a psychiatric
disorder (Table I). After controlling for potential covariates, LOS for patients with a
psychiatric diagnosis was 1.26 times longer (p<0.0001) than for patients without a
psychiatric disorder.
TABLE II (Mean inpatient visits per patient)
7
p<0.0001
p<0.0001
6
p<0.0001
p=0.02
Mean Number of Visits /patient
INTRODUCTION
5
p=0.29
4
Diagnosis Present
Diagnosis Absent
3
2
1
0
Psychiatric Subtype Diagnosis and LOS: Mean LOS for sickle cell pain events increased
by 3.38 days for patients with mood disorders, 2.69 days for patients with anxiety disorders,
0.59 days for patients with disruptive behavior disorders, and 2.02 days for patients with
substance disorder relative to patients without a psychiatric diagnoses (Table I). After
controlling for potential covariates, LOS was 1.38 times longer (p<0.0001) for patients with
mood disorders,1.34 times longer (p<0.0001) for patients with anxiety disorders, and 1.35
times longer (p=0.01) for patients with substance disorder relative to patients with no
psychiatric disorder after controlling covariates. Presence of a disruptive behavior disorder
were not associated with significant increases in LOS (p=0.62).
Psychiatric Diagnosis and Inpatient Healthcare Utilization: Mean inpatient healthcare
utilization for sickle cell pain events increased by 1.91 visits in patients with a psychiatric
disorder relative to patients without a psychiatric disorder (Table II). After controlling for
potential covariates, inpatient healthcare utilization for patients with a psychiatric diagnosis
was 1.36 times more (p<0.0001) than for patients without a psychiatric disorder.
Psychiatric Subtype Diagnosis and Inpatient Healthcare Utilization: Mean inpatient
healthcare utilization for sickle cell pain events increased by 2.41 visits for patients with
mood disorders, 2.47 visits for patients with anxiety disorders, 1.14 visits for patients with
disruptive behavior disorders, and 0.47 visits for patients with substance use disorders
relative to patients without a psychiatric diagnosis (Table II). After controlling for potential
covariates, inpatient healthcare utilization was 1.36 times more (p<0.0001) for patients with
mood disorders and 1.43 times more (p<0.0001) for patients with anxiety disorders relative
to patients with no psychiatric disorder. The presence of disruptive behavior disorders
(p=0.12) and substance use disorders (p=0.67) was not associated with significant increases
in inpatient healthcare utilization.
Mood Disorder
Anxiety Disorder
Disruptive Behavior
Disorder
Substance Disorder
Any Psychiatric Disorder
CONCLUSIONS
• Psychiatric diagnoses were found in approximately 5% of the SCD sample, with
mood disorders and anxiety disorders being the most frequent disorders. These
estimates were lower than previous findings (Bennett, 1994; Cepeda, 1997).
•Pediatric patients with SCD and a psychiatric diagnosis were found to have
significantly longer LOS relative to patients without a psychiatric diagnosis.
• Patients diagnosed with mood disorders, anxiety disorders, or substance
disorders had longer LOS relative to patients without these disorders. However,
disruptive behavior disorders were not associated with significantly longer LOS.
• Pediatric patients with SCD and a psychiatric diagnosis were found to have
significantly more inpatient admissions for sickle cell pain relative to patients
without a psychiatric diagnosis.
• Patients diagnosed with mood disorders or anxiety disorders had more inpatient
admissions for sickle cell pain than patients without these disorders. However,
disruptive behavior and substance disorders and were not associated with
significantly higher inpatient healthcare utilization for sickle cell pain.
•Findings highlight the need for psychological assessment and intervention in
children with SCD to facilitate adjustment with pain.
LIMITATIONS
• Actual presence of psychiatric diagnosis at each visit could not be verified given
existing data.
• Psychiatric disorders were diagnosed through ICD-9 codes and not formal
psychiatric screening.