Background/Purpose . Frequent visitors of the emergency department (ED) among the general population share several demographic, health system and disease characteristics, including older age, poverty, government-financed insurance and poorer health. The burden of depression is also high among frequent ED visitors. Depression strikes up to 75% of SLE patients and is more severe among blacks compared to whites. Previous reports among predominantly white SLE samples suggest that there is an association between depression and ED usage. We sought to examine whether the severity of depressive symptoms increases the risk of being a frequent user of the ED in a predominantly black SLE cohort in the Southeastern US.
Methods . Georgians Organized Against Lupus (GOAL) is a longitudinal cohort largely drawn from a population-based registry of people with SLE, which has been established in Atlanta, GA. Annual surveys furnish self-administered data on demographics, disease outcomes and healthcare utilization. Over 75% of participants are black, and 35% live under the poverty level. We used the 9-item Patient Health Questionnaire (PHQ-9) to assess severity of depressive symptoms. PHQ-9 can be assessed as a continuous (score range 0-27) or categorical (5 categories from minimal to severe depressive symptoms) variable. Individuals who visited the ED ≥ 3 times in the past year were considered frequent ED users. We conducted logistic regression analyses to test the effect of depression severity on being a frequent ED user, after controlling for potential confounders.
Results . Of 566 SLE participants, 96 (17%) visited the ED at least 3 times in the past year. Frequent usage of the ED was found in 10%, 17%, 26%, 24%, and 28% of patients with minimal, mild, moderate, moderately severe, and severe symptoms of depression, respectively. Severity of depressive symptoms, demographic factors, type of insurance, disease activity, and organ damage were associated with frequent ED utilization (Table 1). The PHQ-9 score remained positively associated with the outcome after controlling for major confounders. PHQ-9 and self-reported disease activity (SLAQ) scores were highly correlated (rho=0.65). The association between the PHQ-9 score and frequent ED usage was not longer significant when SLAQ was included in the model (OR=0.98; p=0.85).
Conclusion . Our data suggest that the severity of depressive symptoms may modulate healthcare-seeking behavior in SLE. However, other factors often disproportionately prevalent among socioeconomically disadvantaged subgroups with SLE, such as severe organ damage, greater disease activity and being on Medicaid showed stronger association with the outcome. Longitudinal studies are needed to tease out the complex pathways implicated in the usage of avoidable healthcare resources among minorities with SLE, particularly among those stricken by depressive symptoms.
|
Table 1. Association of Depressive Symptoms with Frequent Usage of the Emergency Department |
|||||
|
Characteristics |
Univariable LR
|
|
Multivariable LR (stepwise)
|
||
|
Odds Ratio
|
P Value
|
|
Odds Ratio
|
P Value
|
|
|
Depressive symptoms (5-unit increase in PHQ-9 score) |
1.43 (1.21-1.70) |
<0.0001 |
|
1.30 (1.07-1.59) |
0.009 |
|
Age at diagnosis (5 year increase) |
0.91 (0.83-1.00) |
0.051 |
|
0.85 (0.75-0.95) |
0.006 |
|
Gender (female) |
0.79 (0.35-1.79) |
0.58 |
|
|
|
|
Race (non-white) |
2.68 (1.31-5.52) |
0.0073 |
|
2.22 (0.95-5.19) |
0.07 |
|
Disease duration (1 year increase) |
0.99 (0.96-1.01) |
0.29 |
|
0.96 (0.93-0.99) |
0.009 |
|
Educational attainment (3 year increase) |
0.80 (0.63-1.00) |
0.051 |
|
|
|
|
Household income below poverty level |
2.90 (1.81-4.63) |
<0.0001 |
|
|
|
|
Insurance type (ref: Private) No Insurance
|
2.83 (1.17-6.84) |
0.021 |
|
1.83 (0.71-4.75) |
0.80 |
|
Medicare
|
3.29 (1.65-6.55) |
0.0007 |
|
2.30 (1.08-4.90) |
0.50 |
|
Medicaid
|
6.27 (3.22-12.23) |
<0.0001 |
|
3.63 (1.76-7.47) |
0.003 |
|
Married or living with partner |
0.59 (0.37-0.94) |
0.028 |
|
|
|
|
Disease activity (5-unit increase in SLAQ score) |
1.65 (1.44-1.88) |
<0.0001 |
|
Not included* |
|
|
Organ damage (SA-BILD score) Mild damage (1-2) vs. no damage (0)
|
2.07 (0.95-4.51) |
0.067 |
|
1.89 (0.82-4.34) |
0.76 |
|
Severe damage (³3) vs. no damage (0)
|
4.91 (2.34-10.33) |
<0.0001 |
|
4.26 (1.84-9.86) |
0.0001 |
|
Abbreviations: LR: logistic regression; PHQ-9: 9-item Patient Health Questionnaire; SA-BILD: Self-administered Brief Index Lupus Damage; SLAQ: Systemic Lupus Activity Questionnaire. * Because SLAQ was highly correlated with PHQ-9 (Pearson correlation coefficient =0.65; p<0.0001), it was not included in the multivariate model. |
|||||
Disclosure:
A. Aguirre,
None;
S. S. Lim,
NIH,
2,
GlaxoSmithKline,
2,
Emory University,
3;
G. Bao,
GlaxoSmithKline,
2,
Emory University,
3;
C. T. Molta,
GSK,
1,
GSK,
3;
H. Kan,
GSK,
1,
GSK,
3;
C. Drenkard,
NIH,
2,
Emory,
3,
GlaxoSmithKline ,
2.
« Back to 2014 ACR/ARHP Annual Meeting
ACR Meeting Abstracts - https://acrabstracts.org/abstract/association-between-depression-and-high-utilization-of-emergency-department-in-patients-with-systemic-lupus-erythematosus-from-the-southeastern-united-states-the-goal-cohort/