Depression Increases Stroke Hospitalization Cost: An Analysis of 17,010 Stroke Patients in 2008 by Race and Gender

Objective. This analysis focuses on the effect of depression on the cost of hospitalization of stroke patients. Methods. Data on 17,010 stroke patients (primary diagnosis) were extracted from 2008 Tennessee Hospital Discharge Data System. Three groups of patients were compared: (1) stroke only (SO, n = 7,850), (2) stroke + depression (S+D, n = 3,965), and (3) stroke + other mental health diagnoses (S+M, n = 5,195). Results. Of all adult patients, 4.3% were diagnosed with stroke. Stroke was more prevalent among blacks than whites (4.5% versus 4.2%, P < 0.001) and among males than females (5.1% versus 3.7%, P < 0.001). Nearly one-quarter of stroke patients (23.3%) were diagnosed with depression/anxiety. Hospital stroke cost was higher among depressed stroke patients (S+D) compared to stroke only (SO) patients ($77,864 versus $47,790, P < 0.001), and among S+D, cost was higher for black males compared to white depressed males ($97,196 versus $88,115, P < 0.001). Similar racial trends in cost emerged among S+D females. Conclusion. Depression in stroke patients is associated with increased hospitalization costs. Higher stroke cost among blacks may reflect the impact of comorbidities and the delay in care of serious health conditions. Attention to early detection of depression in stroke patients might reduce inpatient healthcare costs.


Depression and Healthcare Cost
Several studies have reported the effect of depression/anxiety on healthcare costs. For example, while female Medicare patients had a higher prevalence of depression and higher use of outpatient services, inpatient hospital costs for male patients were 47% higher compared to females ($15,060 versus $10,240, < 0.001) [30]. In another study, the medical cost of depressed patients was 54% higher compared to nondepressed patients [34]. While higher cost among stroke patients is associated with greater number of readmissions, longer hospitalizations, and greater number of outpatient visits compared to a control group without depression, evidence is sparse about whether these costs vary by race and gender.
In this study of Tennessee stroke patients ( = 17, 010), we examine two issues: (1) prevalence of depression among stroke patients by race and gender and (2) the effect of depression on total hospitalization cost in 2008 by race and gender.

Data.
We obtained inpatient discharge data from the 2008 Tennessee Hospital Discharge Data System (HDDS) compiled by The Tennessee Department of Health's (TDH) Division of Health Statistics. All hospitals licensed by the TDH are required by law to report patient-level discharge information. Data are reported on a uniform billing form developed by the National Uniform Billing Committee. Diagnoses in the administrative files are given by the attending physicians (according to the ICD-9 codes), and it is unclear what tests are used in arriving at those diagnoses. Further, these diagnoses appear only when the patient is treated for those conditions in the hospital. We extracted data on primary diagnosis of stroke (ICD-9 codes of 430-438) along with the secondary diagnoses of depression/anxiety (ICD-9 codes 296.2-major depressive disorder, single episode, 292.3-major depressive disorder, recurrent episode, 3000.4-neurotic depression, 309.0-brief depressive reaction, 309.1-prolonged depressive reaction, 311-depressive disorder, not elsewhere classified, and 300-anxiety states, hysteria, phobic disorders, and neurotic depression) for blacks and whites since they constitute 97% of Tennessee population. Since there is a high overlap in symptoms of depression and anxiety ranging from 48% to 74% [37,38], we combined the diagnoses for depression and anxiety as a single variable for our analysis. Data extraction on stroke patients included sex, age, race, days of hospitalization, number of re-admissions, and costs associated with stroke treatment as well as the total hospital charges for the entire year of 2008 when the patient was readmitted for illnesses other than stroke. Extracted data also included co-morbidities such as atrial fibrillation, hypertension, diabetes, cholesterol, and cardiovascular events such as heart attacks. The stroke sample included whites (82%) and females (55%), and the average age in the sample was 70 years. Stroke rates were age adjusted per 2000 US population.

Statistical Analysis.
Analysis of variance compared the average hospitalization costs [39] for three groups of stroke patients: (1) stroke only (S O , = 7, 850), (2) stroke + depression/anxiety (S +D , = 3, 965), and (3) stroke + other mental diagnoses (S +M , = 5, 195). The Fisher exact test was used for comparison of healthcare cost and prevalence of comorbidities by race and sex. Percentages of stroke diagnoses were compared using Pearson's Chi-squared test with Yates' correction for continuity, and odd ratios (ORs) were obtained through logistic regression analyses, which controlled for age, sex, hypertension, diabetes, cholesterol, and atrial fibrillation. A probability value of < 0.05 was the accepted threshold for statistical significance.

Effect of Depression on Hospital Cost for Stroke.
We examined cost and associated co-morbidities including Charlson Index of comorbidity for three stroke groups including: (1) patients with stroke only (S O ); (2) patients with stroke + depression (S +D ); (3) stroke patients with other mental diagnoses (S +M ). Within each stroke category, we compared cost and associated factors by race and gender. Table 2 shows that the average healthcare cost was nearly 63% higher for stroke patients with S +D compared to S O ($77,864 versus $47,790, < 0.001, a difference of 63%) or S +D compared to S +M ($77,864 versus $62,387, < 0.001, a difference of 24.8%). Clearly, these data show that depression among stroke patients is associated with higher hospital costs compared with stroke patients who have other mental illnesses. Table 2 provides costs and comorbidities data for three groups of stroke patients, further illustrating that both stroke prevalence and annual costs were higher among blacks, and the race-sex differences are made evident. Among depressed stroke patients (S +D ), black males had higher annual hospital charges compared to white males ($97,196 versus $88,115, < 0.001), in part due to longer hospital stays compared to white males (24.6 versus 20.2, < 0.001). Similarly, black S +D females had higher cost compared to white S +D females ($95,269 versus $68,184, < 0.001). For black males, the higher cost cannot be attributed to depression/anxiety as only 11% of black males had a diagnosis of depression; the higher cost here appears to reflect complexities (denoted by a higher Charlson comorbidity index) that develop from co-morbid conditions such as higher prevalence of hypertension and diabetes. Similar race and gender trends also existed for black males and females across S +M and S O groups of patients.

Comments
Previous studies on healthcare cost have reported substantially higher cost (54% higher) for patients with cardiovascular disease (CVD) and stroke in association with depression and anxiety [31][32][33][34]. Our analyses show that depression and anxiety among Tennessee stroke patients is associated with a 63% increase in the annual hospital care cost. Further, our findings of higher cost for depressed stroke patients, especially among women, are consistent with those reported previously [34]. Since depression can be considered as an independent risk factor for CVD [40,41] and since women outnumber men in the population (as well as in our S +D group-54% versus 46%), costs attributable to depression 4 Stroke Research and Treatment may be reduced by early diagnosis and treatment of depression. The stroke patients in our sample had higher prevalence of both hypertension (more than 80%) and diabetes (more than 35%). Addressing depression and reducing risk factors through preventive programs [42] could substantially reduce the morbidity, mortality, and healthcare costs associated with stroke [42,43]. The average healthcare cost among blacks compared to whites were higher regardless of whether the stroke was hemorrhagic or ischemic (hemorrhagic cost-$64,643 versus $48,246 < 0.001; ischemic cost-$41,120 versus $28,833, < 0.001). These higher costs remained intact when total stroke costs (combined ischemic+hemorrhagic+unspecified stroke + TIA) were compared between blacks and whites ($41,370 versus $30,215, < 0.001), particularly black males compared to white males ($41,586 versus $31,359, < 0.001). The same cost pattern emerges when the annual cost for the entire 2008 year was combined (blacks had higher annual cost compared to whites: $74,338 versus $55,884, < 0.001, Table 1) and when racial comparisons for nonstroke patients were made. These differences suggest that blacks with chronic conditions may seek medical services later in the progression of their disease and that this late entry to care [44,45] may require more services and longer hospitalization as is evident in our data (16 days for black patients compared to 12 days for white patients, Table 1). Further, the higher cost among black males may in part exist because previous studies suggest that they are more likely to drop out of behavioral and pharmacological therapies [46] which in turn leads to more complications and readmissions (re-admissions are higher among blacks-see Table 1).
The lower overall cost of care among women (particularly white women) compared to men may result from a number of factors including that women, in general, seek professional help earlier on in the development of their illness compared to men [47,48] and this alone may reduce complications and hence reduce length of hospitalization and cost [47][48][49]. In order to impact CVD end points among women, depression/anxiety must be treated both as independent risk factors for preventing CVD and for reducing cost in females with known CVD [50].
Finally, Our findings of higher hospitalization cost of stroke is associated with depression and anxiety that consistently appear as a co-morbid condition requiring greater attention in managing healthcare cost. Findings of higher cost and greater utilization of services, though scantly reported (see Table 3 below), nonetheless are supportive of monitoring ways to contain higher treatment cost associated with stroke and other major events.

Limitations
The administrative hospital discharge files do not provide clinical data regarding severity/duration of diseases, test results, or cost of pharmacological treatment provided. Further, these administrative files do not provide itemized cost, and hence it is impossible to determine the cost of pharmacological treatment for depression/anxiety for any patient. The administrative data only include the total cost for the entire hospital stay, number of admissions, and sometimes within the total cost per admission, the cost associated with major procedures such as CABG. In addition to the primary diagnosis, these administrative files provide data on secondary diagnoses (i.e., co-morbidities) only when treatment is provided for those conditions. These administrative files lack clinical details of diagnoses or co-morbid conditions which may shed additional light on racial and gender differences in healthcare cost. Our data are from a single state and for only one year (2008), and as such they may not reflect outcomes from other geographic areas/regions. Finally, based on this Stroke Research and Treatment 5 After adjusting for patient demographic and clinical factors, patients with stroke and poststroke depression had significantly < 0.0001, more hospitalization, outpatient visits, and longer length of stays, 12 months after stroke compared with patients with stroke but no poststroke depression cross-sectional data, we were unable to differentiate prestroke depression from poststroke depression. However, in either case, the association of stroke with depression in our study seems to contribute to increased hospital stay, greater comorbidities, and significantly greater cost of healthcare.

Conclusion
Stroke patients with depression/anxiety have significantly higher healthcare costs compared to those with stroke only (i.e., without depression/anxiety) or those with other mental health diagnoses. Based on the patterns reported here, greater attention to prevent comorbidities and early detection of depression in stroke patients are all promising interventions aimed at reducing inpatient healthcare costs while improving overall care, with the greatest opportunities for improved health and cost savings in the black male population. Analytic epidemiologic studies are needed to examine whether the higher healthcare costs among blacks exist due to delays in seeking treatment and/or poor access to services, leading to more complex problems and longer hospitalizations. Additionally, research is needed to determine whether aggressive treatment of depressed patients that have suffered stroke might reduce the overall costs of stroke care.