Change in 1-year hospitalization of overall and older patients with major depressive disorder after second-generation antipsychotics augmentation treatment

Journal of Affective Disorders 230 (2018) 118–124

Contents lists available at ScienceDirect

Journal of Affective Disorders journal homepage: www.elsevier.com/locate/jad

Research paper

Change in 1-year hospitalization of overall and older patients with major depressive disorder after second-generation antipsychotics augmentation treatment Chun-Yuan Lina,b,c, Te-Jen Laia,d, Yu-Hsin Wuc,e, Ping-Kun Chenf, Yuan-Fu Ling, I.-Chia Chienb,h,

T

⁎

a

School of Medicine, Chung Shan Medical University, Taichung, Taiwan Tsaotun Psychiatric Center, Ministry of Health and Welfare, Nantou, Taiwan c National Changhua University of Education, Changhua, Taiwan d Department of Psychiatry, Chung Shan Medical University Hospital, Taichung, Taiwan e Feng Yuan Hospital, Ministry of Health and Welfare, Taichung, Taiwan f School of Medicine, China Medical University, Taichung, Taiwan g College of Management, National Taiwan University, Taipei, Taiwan h Department of Public Health and Institute of Public Health, National Yang-Ming University, Taipei, Taiwan b

A R T I C L E I N F O

A B S T R A C T

Keywords: Geriatric/Aging/Elderly Depression Pharmacoepidemiology Pharmacotherapy Treatment resistance

Background: Studies on second-generation antipsychotics (SGA) augmentation treatment for older adults with major depressive disorder (MDD) remain limited. We aimed to investigate the effectiveness of SGA augmentation for overall and older patients with MDD inpatient history by assessing the change in 1-year hospitalization before and after SGA augmentation using the latest National Health Insurance Research Database (NHIRD) in Taiwan. Methods: The samples were MDD patients (ICD-9 CM code: 296.2 and 296.3) who had psychiatric inpatient history. A total of 2602 MDD patients including 430 elderly subjects (age ≥ 60 years) who received SGA augmentation for 8 weeks between January 1998 and December 2012 were included in this 1-year mirror-image study. Outcome measures included number and length of psychiatric and all-cause hospitalizations. Results: After 8-week continuous SGA augmentation in the study subjects, the total number and days of psychiatric hospitalizations among overall patients reduced by 33.57% (p < .0001) and 18.24% (p < .0001), respectively; the total number and days of psychiatric hospitalizations among older patients (age ≥ 60) reduced by 44.52% (p < .0001) and 27.95% (p < .0001), respectively. Similarly, the total number and days of all-cause hospitalizations were significantly reduced. Limitations: MDD patients without inpatient history were not included due to data limitation; hence, the results may not be generalized to all patients. Conclusions: The results support that SGA may be effective in reducing psychiatric and all-cause hospitalization among overall and elderly MDD patients. More studies focusing on the safety of SGA among older MDD patients is warranted.

1. Introduction Major depressive disorder (MDD), the most striking psychiatric disease with a high rate of treatment resistance, burdens the individuals, families, and society (Kasper and Montgomery, 2013, Kessler et al., 2003, Rush et al., 2006). Clinical trials (Bauer et al., 2009, Berman et al., 2009, 2007, Mahmoud et al., 2007, McIntyre et al., 2007, Rapaport et al., 2006) have revealed that antidepressants augmented with second-generation antipsychotics (SGA) result in better treatment responses for difficult-to-

⁎

treat MDD patients (Komossa et al., 2010, Nelson and Papakostas, 2009). Our previous population-based study (Lin et al., 2014) revealed that use of key psychiatric services due to MDD was reduced after SGA augmentation. The subjects were MDD patients with psychiatric inpatient history before 2008, and patients with new onset inpatient history after January 1, 2008 were not included due to data limitation (Lin et al., 2014). With SGA augmentation in MDD becoming an alternative treatment option, patients with inpatient history after 2008 merit investigation when the updated database is available.

Corresponding author at: Department of Public Health and Institute of Public Health, National Yang-Ming University, Taipei, Taiwan. E-mail address: [email protected] (I.-C. Chien).

https://doi.org/10.1016/j.jad.2018.01.011 Received 8 September 2017; Received in revised form 9 December 2017; Accepted 24 January 2018 Available online 31 January 2018 0165-0327/ © 2018 Elsevier B.V. All rights reserved.

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

not be observed for service utilization of an entire year before and after the index date were further excluded. Thus, the present study comprised a final sample of 2602 MDD patients, among which 430 were elderly subjects (age ≥ 60) (Fig. 1).

Importantly, MDD is common across an individual's life span. In older population with MDD, comorbid conditions account for a major portion of disability (Greenberg et al., 2015). Although treatment of MDD often includes use of multiple medications (Nelson and Papakostas, 2009, Rush et al., 2006), pharmacotherapy for older adults with MDD requires careful consideration (Steffens et al., 2011). While older patients bring unique challenges, evidence for effectiveness of SGA augmentation treatment is more limited among the elderly than in the general population (Alexopoulos et al., 2008, Lenze et al., 2015, MacQueen et al., 2016, Steffens et al., 2011). Thus, analysis on health service utilization among older patients would provide clinicians with more information on the benefits and risks of SGA augmentation in this population. Hospitalization can serve as a valid outcome indicator (Wingård et al., 2017) because it indicates relapse of severe symptoms and functional impairment of MDD patients, while shorter length of hospitalization represents better treatment response of the illness. In this 1year mirror image study, MDD patients with continuous SGA augmentation for 8 weeks were identified for assessing the change of hospitalization service utilization before and after SGA augmentation treatment.

2.3. Outcome measures This study used a mirror-image study design to compare number and length of psychiatric and all-cause hospitalizations before and after SGA augmentation treatment (pre- and post-SGA periods) among 2602 MDD patients. The number and length of hospitalizations were measured using inpatient records in NHIRD. If a patient was readmitted to the same hospital or transferred to another hospital within 3 days with the same diagnosis, the subsequent admission was counted as part of the previous hospitalization. In addition to measuring hospitalization service utilization, changes in concomitant treatment after SGA augmentation were compared. 2.4. Ascertainment of other variables The demographic and clinical information of the study subjects including age, gender, comorbidities that frequently influence the outcome of MDD, antidepressant treatments, and other augmentation treatments were analyzed (American Psychiatric Association, 2010, Kornstein and Schneider, 2001). Physical comorbidities included neurologic disorders, cardiac diseases, endocrine disorders, and chronic liver disorders. Psychiatric comorbidities included anxiety, dysthymia and alcohol/substance uses. The comorbidity was coded as one if it appeared in the inpatient register once or in the outpatient register 3 times prior to the index day. Uses of antidepressants, first-generation antipsychotics (FGA), lithium, anticonvulsants, stimulant, thyroid hormone, or electroconvulsive therapy (ECT) were defined on the basis of at least one dispensation for a year prior to the index date. Antidepressant dosages were converted into imipramine-equivalent milligrams (Bollini et al., 1999) and antipsychotics dosages were converted into chlorpromazine-equivalent milligrams (Gardner et al., 2010).

2. Methods 2.1. Data source In Taiwan, 99% of the entire population is enrolled in the National Health Insurance (NHI) program which covers all illness (National Health Insurance Adminstration, 2017). The National Health Insurance Research Database (NHIRD) contains healthcare data from the NHI program and has been successfully used for pharmaco-epidemiological research and studies on psychiatric disorders (Hsu et al., 2017, Lin et al., 2013, 2016). The sample in this study was derived from the NHIRD and included 266,328 patients who received inpatient psychiatric service from January 1, 1996 to December 31, 2013. Demographic data and both psychiatric and all-cause inpatient and outpatient health care utilization information including diagnostic codes, as well as details of procedures, surgeries and pharmacy records of the enrolled patients from January 1, 1996 to December 31, 2013 were collected and analyzed.

2.5. Statistical analysis The data were expressed as mean ( ± standard deviation) for continuous variables and frequency (%) for categorical data. Chi-square test or independent t-test was employed to test the differences between patients with continuous SGA treatment and those never treated with any SGA. Since data of inpatient service utilization are usually nonnormal distributed and skewed, Wilcoxon sign-rank test was utilized to assess the differences including number and length of hospitalizations between pre- and post-SGA period of the subjects in the 1-year mirrorimage study. McNemar's test was employed to compare the 1-year difference of concomitant treatments beyond SGA including FGA, lithium, anticonvulsants, stimulant, thyroid hormone and ECT (Nielsen et al., 2012). All analyses were performed using the SAS system (version 9.2; SAS Institute, Cary, NC) and the STATA 11th edition (Stata Corporation, College Station, TX, USA) and p < .05 was considered statistically significant.

2.2. Study subjects Ethical approval was obtained from the Institutional Review Board of Tsaotun Psychiatric Center for this study. The flowchart of study sample selection is shown in Fig. 1. The International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9 CM) codes of the diseases and the medicines analyzed in the study are shown in Supplementary Table 1 and Supplementary Table 2, respectively. The MDD subjects were assessed according to ICD-9 CM records. A total of 58,539 patients whose initial hospitalization to a psychiatric ward with the diagnosis of MDD (ICD9-CM code 296.2 or 296.3) occurred between January 1, 1996 and December 31, 2013 were identified. In addition, for ascertaining the diagnostic stability of unipolar depression (Lin et al., 2014, Sadock and Sadock, 2003), 19,393 patients with records of psychiatric inpatient stays with a diagnosis of MDD (ICD9-CM code 296.2 or 296.3) were identified. The claims for prescription drugs were employed to confirm the use of antidepressants, SGA, and other medicines. According to previous clinical trials (Komossa et al., 2010, Nelson and Papakostas, 2009) with a mean treatment duration of 8 weeks, the subjects were MDD patients in Taiwan who received antidepressants plus SGA, including amisulpride, aripiprazole, clozapine, olanzapine, paliperidone, quetiapine, risperidone, ziprasidone, or zotepine, continuously for at least 8 weeks. The index date was the day when the subjects started SGA augmentation. For accurate calculation of outcome measures, patients who could

3. Results 3.1. Patient characteristics For overall patients and those older than 60 y/o, the number of patients who received SGA had grown steadily by year (Nelson and Papakostas, 2009, Papakostas et al., 2007), revealing a trend of increasing usage of SGA (Supplementary Table 3). Quetiapine (n = 1461), risperidone (n = 350), aripiprazole (n = 264) and olanzapine (n = 237) are the leading SGA administered to MDD patients with 119

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

Fig. 1. Flow chart of sample selection Note: No patient received augmentation treatments of SGA continuously for at least 8 weeks before 1998. Abbreviations: ICD-9 CM: The International Classification of Diseases, Ninth Revision, Clinical Modification, MDD: Major Depressive Disorder, SGA: Second Generation Antipsychotics.

3.2. Pre-post differences of psychiatric and all-cause inpatient services and concomitant treatment uses

psychiatric inpatient history in Taiwan. The mean daily chlorpromazine-equivalent dosage of SGA was 94.0 ± 89.2 mg in overall patients and 78.5 ± 122.5 mg in older patients. To show the differences in characteristics between MDD patients with continuous SGA augmentation and those without any SGA treatment, 11,967 MDD patients (including 1374 older patients) never received any kind of SGA were selected from the database. Table 1 shows the comparison of characteristics between MDD patients (overall age and age ≥ 60 years) with continuous SGA augmentation and those without any SGA treatment. Similar to previous findings (Lin et al., 2014), significant differences were observed. Among older patients (age ≥ 60 years), those with SGA treatment were more likely females, comorbid with endocrine disorders, anxiety disorders and dysthymia and were treated with higher doses of antidepressants and various kinds of treatments within the year prior to SGA augmentation. Whether receiving SGA or not, older MDD patients (age ≥ 60 years) with psychiatric inpatient history had multiple comorbid diseases (Table 1).

In the year after SGA augmentation treatment, the total number of psychiatric admissions among overall MDD patients decreased by 33.57% (p < .0001) and the total annual days of psychiatric hospitalization decreased by 18.24% (p < .0001). The all-cause hospitalization analysis also reveals significant reduction in number and length of admissions among patients, by 29.44% (p < .0001) and 18.48% (p < .0001), respectively (Table 2). The average numbers of psychiatric and all-cause hospitalizations were decreased from 0.82 to 0.54 and 1.31 to 0.92, respectively. The average lengths of psychiatric and allcause hospitalizations were decreased from 21.94 days to 17.94 days and 26.42 days to 21.54 days, respectively. In addition, the proportion of overall patients with decreased numbers of psychiatric and all-cause hospitalizations were 78.82% (N = 2501) and 76.79% (N = 1998); the proportion of overall patients with decreased lengths of psychiatric and 120

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

Table 1 Demographics of study subjects VS patients not treated with any SGAa. p-value

All age Study subjects N = 2602 Age when SGA began (Mean, SD) 43.9 ± 16.4 Gender (N%) Female 1673 (64.3) Male 929 (35.7) Comorbidity (N%) Neurologic disorders 422 (16.2) Cardiac diseases 462 (17.8) Endocrine disorders 85 (3.3) Chronic liver disorders 664 (25.5) Anxiety/Dysthymia 1628 (62.6) Alcohol abuse/dependence 131 (5.0) Substance abuse/dependence 93 (3.6) Treatments prior to SGA augmentation Antidepressants uses Average daily dose (mg/day) (Mean, SD) 156.9 ± 99.5 Prescription days (Mean, SD) 175.8 ± 112.8 Kinds (N%) 0 0 (0.0) 1 kind 959 (36.9) ≧2 kinds 1643 (63.1) Augmentation beyond SGAs prior to SGA initiated (N%) FGA 1639 (63.0) Anticonvulsants 448 (17.2) Lithium 99 (3.8) Stimulant 58 (2.2) Thyroid hormone 4 (0.2) ECT 33 (1.3)

Patients not treated with any SGA N = 11,967 34.8 ± 17.5

Study subjects

< .001 < .001

3897 (32.6) 8070 (67.4)

p-value

Age≧60

N = 430

Patients not treated with any SGA N = 1374

69.5 ± 7.4

71.6 ± 8.4

283 (65.8) 147 (34.2)

693 (50.4) 681 (49.6)

< .001 < .001

1072 (9.0) 1170 (9.8) 120 (1.0) 1726 (14.4) 4389 (36.7) 280 (2.3) 121 (1.0)

< .001 < .001 < .001 < .001 < .001 < .001 < .001

185 (43.0) 212 (49.3) 26 (6.0) 157 (36.5) 250 (58.1) 7 (1.6) 7 (1.6)

572 (41.6) 696 (50.7) 30 (2.2) 472 (34.4) 582 (42.4) 30 (2.2) 7 (0.5)

0.649 0.664 < .001 0.446 < .001 0.607 0.046

74.7 ± 90.5 70.4 ± 94.8

< .001 < .001 < .00

158.4 ± 113.5 171.4 ± 113.2

65.8 ± 101.2 81.7 ± 117.4

< .001 < .001 < .001

0 (0.0) 153 (35.6) 277 (64.4)

695 (50.6) 438 (31.9) 241 (17.5)

277 (64.4) 36 (8.4) 11 (2.6) 10 (2.3) 2 (0.5) 5 (1.2)

341 (24.8) 53 (3.9) 7 (0.5) 6 (0.4) 0 (0.0) 1 (0.1)

4476 (37.4) 4921 (41.1) 2570 (21.5) 3198 (26.7) 1010 (8.4) 85 (0.7) 65 (0.5) 6 (0.1) 5 (0.0)

< .001 < .001 < .001 < .001 0.157 < .001

< .001 < .001 < .001 < .001 0.089 0.003

Note: Chi-square test or independent t-test was used to test the difference between groups. b The comorbidity was coded as one if it showed up in the inpatient record once or in the outpatient records 3 times prior to the index day. c Usages of antidepressants, FGA, anticonvulsants, lithium, stimulant, thyroid hormone, or ECT were defined with at least 1 dispensation prior 1 year to the index day. Antidepressants dosages were converted to imipramine-equivalent milligrams25 and antipsychotics dosages were converted to chlorpromazine-equivalent milligrams.26. Abbreviations:. SGA: Second generation antipsychotics. FGA: first generation antipsychotics. ECT: electroconvulsive therapy. a MDD patients who never received any kind of second generation antipsychotics those are available in Taiwan, includes: risperidone, olanzapine, quetiapine, aripiprazole, amisulpride, paliperidone, clozapine, zotepine, and ziprasidone. To define the index date of the non-SGA users, they were matched to the study subjects according to their first MDD admission dates and then their index days were assigned the same as the index days of the corresponding study subjects.

users (≥ 60). As compared with the 1-year psychiatric inpatient utilization before SGA augmentation treatment, the total number and days of psychiatric admissions reduced by 44.52% (p < .0001) and 27.95% (p < .0001), respectively. Similarly, the total number and days of allcause hospitalizations reduced by 32.57% (p < .0001) and 24.05% (p < .0001), respectively. The average numbers of psychiatric and allcause hospitalizations of older patients decreased from 0.70 to 0.39 and 1.42 to 0.96, respectively. The average lengths of psychiatric and allcause hospitalizations were decreased from 21.77 days to 15.68 days and 29.60 days to 22.48 days, respectively. The proportion of older patients with decreased numbers of psychiatric and all-cause hospitalizations were 83.49% (N = 359) and 79.07% (N = 340); the proportion of older patients with decreased lengths of psychiatric and all-cause hospitalizations were 80.23% (N = 345) and 73.49% (N = 316), respectively. After analyzing the outcomes by sub-grouping overall patients into each SGA, significant reduction in numbers and lengths of psychiatric and all-cause admissions was observed in patients receiving quetiapine, risperidone, aripiprazole, olanzapine, zotepine and amisulpride (p values not shown). With regard to older patients, reduction trends of psychiatric and all-cause inpatient service utilization were noted for all SGA (except clozapine which no older patient received). Significant decrease in number and length of psychiatric and all-cause hospitalizations were observed in patients receiving quetiapine, risperidone, and olanzapine (Tables 3 and 4).

Table 2 Differences on 1-year psychiatric and all-cause hospitalization service utilization of overall SGA receiversa between pre-SGA and post-SGA periods. Hospitalization service uses

Pre-SGA

Post-SGA

Difference

p-valueb

Number of psychiatric admissions Number of all-cause admissions Days of psychiatric hospitalizations Days of all-cause hospitalizations

2130 3404 57,096

1415 2402 46,680

−33.57% −29.44% −18.24%

< .0001 < .0001 < .0001

68,748

56,040

−18.48%

< .0001

Note:. Abbreviations:. SGA: second generation antipsychotics. a 2602 MDD patients who received antidepressants plus SGA in Taiwan including amisulpride, aripiprazole, clozapine, olanzapine, paliperidone, quetiapine, risperidone, ziprasidone, or zotepine continuously for at least 8 weeks. b Wilcoxon sign-rank test was used to assess the differences.

all-cause hospitalizations were 76.40% (N = 1988) and 71.91% (N = 1871), respectively. To test if the significances were overestimated by Wilcoxon rank-test, t-test was used to recheck the differences and the results indicated that reduction in hospital utilization among SGA subjects was not overestimated (all p values < .0001). Tables 3 and 4 show the pre-post differences in psychiatric and allcause inpatient service utilization, respectively for older continuous SGA 121

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

Table 3 Differences on 1-year psychiatric hospitalization service uses of older SGA receiversa between pre-SGA and post-SGA periods. Groups

Overall (n = 430) Quetiapine (n = 260) Risperidone (n = 65) Olanzapine (n = 58) Aripiprazole (n = 30) Amisulpride (n = 9) Zotepine (n = 6) Paliperidone (n = 2)

Number of psychiatric admissions

Length of psychiatric hospitalization (Days)

Pre-SGA

Post-SGA

Difference

p-valueb

Pre-SGA

Post-SGA

Difference

p-valueb

301 193 35 38 21 8 4 2

167 113 13 24 13 2 1 1

−44.52% −41.45% −62.86% −36.84% −38.10% −75.00% −75.00% −50.00%

< .0001 < .0001 0.0023 0.0332 0.1066 0.0477 0.1489 1

9359 6053 935 1271 783 188 71 58

6743 4665 408 1101 368 60 36 105

−27.95% −22.93% −56.36% −13.38% −53.00% −68.09% −49.30% 81.03%

< .0001 < .0001 0.0223 0.0131 0.0383 0.0759 0.5839 1

Note:. No older MDD patients (age≧60 when SGA began) received clozaril. Abbreviations:. SGA: second generation antipsychotics. a 430 older MDD patients (age≧60 when SGA began) who had continuous SGA augmentation treatment in Taiwan including amisulpride, aripiprazole, olanzapine, paliperidone, quetiapine, risperidone, ziprasidone, or zotepine for 8 weeks. b Wilcoxon sign-rank test was used to assess the differences.

Table 4 Differences on 1-year all-cause hospitalization service uses of older SGA receiversa between pre-SGA and post-SGA periods. Groups

Overall (n = 430) Quetiapine (n = 260) Risperidone (n = 65) Olanzapine (n = 58) Aripiprazole (n = 30) Amisulpride (n = 9) Zotepine (n = 6) Paliperidone (n = 2)

Number of all-cause admissions

Length of all-cause hospitalization (Days)

Pre-SGA

Post-SGA

Difference

p-value

611 396 98 60 33 10 12 2

412 273 70 31 27 4 6 1

−32.57% −31.06% −28.57% −48.33% −18.18% −60% −50% −50%

< .0001 < .0001 0.0199 0.0028 0.1166 0.089 0.2693 1

b

Pre-SGA

Post-SGA

Difference

p-valueb

12,729 8227 1441 1812 866 211 114 58

9668 6309 1099 1510 503 76 66 105

−24.05% −23.31% −23.73% −16.67% −41.92% −63.98% −42.11% 81.03%

< .0001 < .0001 0.0434 0.0088 0.0943 0.0759 0.4185 1

Note:. No elderly MDD subjects (age≧60 when SGA began) received Clozaril. Abbreviations:. SGA: second generation antipsychotics. a 430 elderly MDD subjects (age≧60 when SGA began) who had continuous SGA augmentation treatment in Taiwan including amisulpride, aripiprazole, olanzapine, paliperidone, quetiapine, risperidone, ziprasidone, or zotepine for 8 weeks. b Wilcoxon sign-rank test was used to assess the differences.

augmentation. After 8 weeks of SGA augmentation, both number and length of psychiatric hospitalizations were significantly reduced in overall and elderly MDD patients. These results were consistent with previous findings (Lin et al., 2014). Although this research included a patient population different from that in the prospective clinical trials, the results obtained using the updated NHIRD in Taiwan were in line with those from clinical trials (Nelson and Papakostas, 2009) suggesting that SGA augmentation therapy has a positive impact on MDD patients. The subjects showed worse clinical conditions and more comorbidities than patients who never received SGA; therefore, one concern is that admissions due to somatic diseases are possible events beyond psychiatric admission and might overestimate the reduction in psychiatric hospitalization. Similar to the trend of reduced use of psychiatric services, the all-cause hospitalization analysis showed significant decrease in number and length of hospitalizations in these patients. Depression may increase the risk for subsequent development of many health problems, and direct adverse physiological effects are associated with increased medical morbidity (Cronin-Stubbs et al., 2000). Although the association between depression and physical comorbidity is modulated by many intermediate factors and is not a simple cause-and-effect relationship, treating MDD could improve depressive symptoms, health-related functioning, and quality of life in patients (Cronin-Stubbs et al., 2000). The current findings indicate that SGA augmentation might render better outcome in terms of reducing hospitalization for difficult patients with MDD.

It is possible that the observed pre-post difference in inpatient utilization among SGA subjects were related to a general trend of inpatient utilization, thus overestimating or underestimating the effect of SGA. To rule out this possibility, the analysis was adjusted according to the overall trend in psychiatric and overall inpatient utilization for all MDD patients. The annual average number and days of psychiatric and allcause hospitalization were calculated as total counts and days of hospitalization divided by the total number of MDD inpatients in each year for the period 1997–2013. A linear regression analysis was performed to assess the changes in use of psychiatric and all-cause hospital services over time. After adjusting according to the slope (Nielsen et al., 2012), the reduction in number and length of psychiatric and all-cause hospitalization remained significant (all p values < 0.0001), indicating that reduction in hospital utilization among SGA subjects was not overestimated. In addition, after comparing the difference in concomitant treatments beyond SGA including FGA, anticonvulsants, lithium, stimulant, thyroid hormone, and ECT between pre- and post-SGA periods, the use of FGA was significantly reduced in overall and older patients by 21.45% (p < .0001) and 29.3% (p < .0001), respectively. 4. Discussion This population-based study assessed 1-year change in use of psychiatric and all-cause inpatient service among MDD patients with SGA 122

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

utilization did not seem to be driven by any one particular SGA drug. In this study, quetiapine was the most widely used SGA and the results supported findings from trials on MDD patients aged 18–65 years (Bauer et al., 2010, Weisler et al., 2013), a retrospective chart review focused on elderly inpatients (Tadger et al., 2011), and an open-label study on depressed elderly patients with cerebrovascular damage (Carta et al., 2007). A recent randomized, double-blind, flexible-dose study on elderly patients (age ≥ 66 years) demonstrated the acute efficacy of quetiapine XR against depressive symptoms in elderly patients (Katila et al., 2013, Weisler and McIntyre, 2013), more studies on the efficacy and tolerability of SAG monotherapy in patients with MDD are needed. This study has several limitations. First, MDD patients without inpatient history according to the NHIRD database were not included due to data limitation; hence, selection biases may exist. Although the results may not be generalized to all patients, knowledge of the role of SGA in elderly population with more severe psychopathology in MDD is enhanced. Second, outcome measures other than hospitalizations were not included and some MDD patients with less pathology may not use inpatient services. Nevertheless, significant reductions in revealed crises as indicated by hospitalizations were found, measures of other less severe outcomes may even reveal a larger impact of SGA augmentation. Third, the traditional mirror-image design using the starting date of SGA therapy as the index day may underestimate the true impact of therapy, because patients usually receive the new therapy during a crisis event that may utilize more inpatient resources and patients may not respond immediately during the treatment period (Gianfrancesco et al., 2002). Fourth, some psychosocial factors possibly related to outcomes of MDD could bias the outcomes observed. However, the pre-post mirror-image design could eliminate some bias caused by these confounders. Fifth, owing to the sample size and the mirror-image study design, no direct comparisons of each drug-drug combination were made; only the subgroup analysis was performed to examine the trend of outcome in each SGA. Finally, little is known about the longer-term safety of SGA augmentation treatment in MDD, especially in older individuals. To consider the risk-benefit ratio, more studies focusing on the adverse events related to SGA for longer duration need further investigation (Komossa et al., 2010, Lin et al., 2016).

It should be noted that regardless whether receiving SGA or not, older patients (age ≥ 60 years) with psychiatric inpatient history had multiple comorbid diseases according to our data (Table 1). Depression in late life tends to be a recurrent or persistent condition and impacts adversely both medical and psychiatric morbidity (Alexopoulos et al., 2011). In the United States, the growths of MDD prevalence and medical costs were unevenly distributed by age, with the ≥ 50-year age group increasing fastest (Greenberg et al., 2015). In older patients, the greater the number of depressive symptoms, the lower the likelihood of their recovering from a physical disability would be (Cronin-Stubbs et al., 2000). The health consequences of depression in late life will increase over time as the population continues aging and could be a great challenge. The present data also showed that older patients with SGA treatment were more likely to have endocrine disorders, anxiety disorders and dysthymia and had received various kinds of augmentation treatments prior to SGA use. Comorbidity is a particular problem in the elderly population with depression; comorbidity with other psychiatric illnesses may affect overall treatment response (Beekman et al., 2000) while comorbidity with physical illnesses may limit treatment strategies. For example, there have been concerns regarding drug-drug interaction and drug-disease interaction, especially in antipsychotics use with metabolic syndrome, prolongation of QTc, and Parkinson's disease (Lenze et al., 2015). The mean doses of SGA were lower than the mean final dose for MDD in previous clinical trials (Komossa et al., 2010) but similar to those in clinical settings (Cakir and Senkal, 2016, Tadger et al., 2011), implying that clinicians tend to use lower dose of SGA for augmentation treatment in MDD, especially in older population. In April 2005, the US Food and Drug Administration (FDA) issued warnings regarding the risks of both FGA and SGA use among elderly patients with dementia (Dorsey et al., 2010). Although patients with dementia and depression may experience adverse events differently, sedation, dizziness, orthostatic hypotension, akathisia and gastrointestinal adverse events were found in elderly patients receiving SGA (Weisler et al., 2013). Findings from the present natural population study suggest that SGA augmentation treatment of depression have beneficial effects on reducing hospitalization in older patients and partially substantiate the results of previous clinical trials (MacQueen et al., 2016, Tadger et al., 2011). Besides optimizing clinical efficacy for patients with MDD and an inadequate response to first-line therapy, it is noteworthy that differences in tolerability profiles should be considered when selecting treatment regimens for elderly patients with MDD (Katila et al., 2013, Lenze et al., 2015). More frequent visits and slow titration when using SGA augmentation therapy could be beneficial for patients’ safety. Older adults with depression may exhibit cognitive impairment that would be associated with increased risks of poor response to first-line antidepressant treatment and developing neurodegenerative process (Brailean et al., 2017, Cronin-Stubbs et al., 2000). To focus on unipolar depression, the current study excluded patients with major psychiatric disorders such as dementia. Although patients with minor cognitive impairments or psychiatric and behavioral problems might not be excluded, the current study could not provide information regarding the effect of SGA on MDD with cognitive impairments. It is important to conduct more trials to identify the efficacy of SGA on depression with cognitive dysfunction. The use of therapies including anticonvulsants, lithium and FGA prior to SGA augmentation in the study subjects reflected the psychiatrists’ clinical decisions on more augmentation treatments in the study subjects. After SGA augmentation, the use of FGA reduced significantly. Research results on the effects of FGA in MDD are controversial (Mortimer et al., 2003) and use of FGA has been discouraged due to the risk of tardive dyskinesia in MDD (Nelson and Papakostas, 2009). With subjects who had history of receiving FGA excluded, the overall outcome measures were similar, indicating that the decreased hospitalization was not related to FGA use. The subgroup analysis revealed that the reductions in overall service

5. Conclusions Improving treatment outcomes of MDD across the life span warrants more efforts and studies. The present findings provide support that SGA may be effective in reducing psychiatric and all-cause inpatient service utilization among MDD patients.

Funding This study was supported by Hospital and Social Welfare Organizations Administration Commission Grant (number 105-52) and Ministry of Science and Technology Grant (number MOST 106 - 2314 B - 894 - 001 -).

Material support The present study was based in part on data from the NHIRD provided by Ministry of Health and Welfare and managed by the National Health Research Institute.

Role of the funding The sponsors were not involved in the design and conduct of the study; collection, management, analysis, and interpretation of the data; and preparation, review, or approval of the manuscript. 123

Journal of Affective Disorders 230 (2018) 118–124

C.-Y. Lin et al.

Ethical standards

Kessler, R.C., Berglund, P., Demler, O., et al., 2003. The epidemiology of major depressive disorder. JAMA 289, 3095–3105. Komossa, K., Depping, A., Gaudchau, A., et al., 2010. Second-generation antipsychotics for major depressive disorder and dysthymia. Cochrane Database Syst. Rev. 8 (CD008121). Kornstein, S., Schneider, R., 2001. Clinical features of treatment-resistant depression. J. Clin. Psychiatry 62 (Suppl 16), S18–S25. Lenze, E.J., Mulsant, B.H., Blumberger, D.M., et al., 2015. Efficacy, safety, and tolerability of augmentation pharmacotherapy with aripiprazole for treatment-resistant depression in late life: a randomized placebo-controlled trial. Lancet 386, 2404–2412. Lin, C.Y., Lane, H.Y., Chen, T.T., et al., 2013. Inverse association between cancer risks and age in schizophrenic patients: a 12-year nationwide cohort study. Cancer Sci. 104, 383–390. Lin, C.Y., Tsai, G.E., Wang, H.S., et al., 2014. Effectiveness of aripiprazole, olanzapine, quetiapine, and risperidone augmentation treatment for major depressive disorder: a nationwide population-based study. J. Clin. Psychiatry 75, e924–e931. Lin, C.Y., Wu, Y.H., Wang, H.S., et al., 2016. Risk of new onset type II DM in MDD patients receiving second-generation antipsychotics treatment: a nationwide cohort study. Depress. Anxiety 33, 435–443. MacQueen, G.M., Frey, B.N., Ismail, Z., et al., 2016. Canadian Network for Mood and Anxiety Treatments (CANMAT) 2016 Clinical guidelines for the management of adults with major depressive disorder: section 6. special populations: youth, women, and the elderly. Can. J. Psychiatry 61, 588–603. Mahmoud, R.A., Pandina, G.J., Turkoz, I., et al., 2007. Risperidone for treatment-refractory major depressive disorder. Ann. Int. Med. 147, 593–602. McIntyre, A., Gendron, A., McIntyre, A., 2007. Quetiapine adjunct to selective serotonin reuptake inhibitors or venlafaxine in patients with major depression, comorbid anxiety, and residual depressive symptoms: a randomized, placebo-controlled pilot study. Depress. Anxiety 24, 487–494. Mortimer, A., Martin, M., Vega, J.W., et al., 2003. Conventional antipsychotic prescription in unipolar depression, II: withdrawing conventional antipsychotics in unipolar, nonpsychotic patients. J. Clin. Psychiatry 64, 668–672. Nelson, J.C., Papakostas, G.I., 2009. Atypical antipsychotic augmentation in major depressive disorder: a meta-analysis of placebo-controlled randomized trials. Am. J. Psychiatry 166, 980–991. National Health Insurance Adminstration, 2017. About National Health Insurance. Retrieved May 9, 2017, from 〈http://www.nhi.gov.tw/english/webdata/webdata. aspx?Menu=11&menu_id=290&webdata_id=2974&WD_ID=290〉. Nielsen, J., Kane, J.M., Correll, C.U., 2012. Real-world effectiveness of clozapine in patients with bipolar disorder: results from a 2-year mirror-image study. Bipolar Disord. 14, 863–869. Papakostas, G., Shelton, R., Smith, J., et al., 2007. Augmentation of antidepressants with atypical antipsychotic medications for treatment-resistant major depressive disorder: a meta-analysis. J. Clin. Psychiatry 68, 826–831. Rapaport, M.H., Gharabawi, G.M., Canuso, C.M., et al., 2006. Effects of risperidone augmentation in patients with treatment-resistant depression: results of open label treatment followed by double-blind continuation. Neuropsychopharmacology 31, 2505–2513. Rush, A.J., Trivedi, M.H., Wisniewski, S.R., et al., 2006. Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am. J. Psychiatry 163, 1905–1917. Sadock, B., Sadock, V., 2003. Kaplan & Sadock's Synopsis of Psychiatry. Williams & Wilkins, New York. Steffens, D.C., Nelson, J.C., Eudicone, J.M., et al., 2011. Efficacy and safety of adjunctive aripiprazole in major depressive disorder in older patients: a pooled subpopulation analysis. Int. J. Geriatr. Psychiatry 26, 564–572. Tadger, S., Paleacu, D., Barak, Y., 2011. Quetiapine augmentation of antidepressant treatment in elderly patients suffering from depressive symptoms: a retrospective chart review. Arch. Gerontol. Geriatr. 53, 104–105. Weisler, R., McIntyre, R., 2013. The role of extended-release quetiapine fumarate monotherapy in the treatment of patients with major depressive disorder. Expert Rev. Neurother. 13, 1161–1182. Weisler, R., McIntyre, R., Bauer, M., 2013. Extended-release quetiapine fumarate in the treatment of patients with major depressive disorder: adjunct therapy. Expert Rev. Neurother. 13, 1183–1200. Wingård, L., Bodén, R., Brandt, L., et al., 2017. Reducing the rehospitalization risk after a manic episode: a population based cohort study of lithium, valproate, olanzapine, quetiapine and aripiprazole in monotherapy and combinations. J. Affect. Disord. 217, 16–23.

This study has been approved by the Institutional Review Board of Tsaotun Psychiatric Center. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jad.2018.01.011. References Alexopoulos, G.S., Canuso, C.M., Gharabawi, G.M., et al., 2008. Placebo-controlled study of relapse prevention with risperidone augmentation in older patients with resistant depression. Am. J. Geriatr. Psychiatry 16, 21–30. Alexopoulos, G.S., Raue, P., Kiosses, D., et al., 2011. Problem-solving therapy and supportive therapy in older adults with major depression and executive dysfunction: effect on disability. Arch. Gen. Psychiatry 68, 33–41. American Psychiatric Association, 2010. Practice guideline for the treatment of patients with major depressive disorder. 3rd edition. Arlington, VA. Bauer, M., El-Khalili, N., Datto, C., et al., 2010. A pooled analysis of two randomised, placebo-controlled studies of extended release quetiapine fumarate adjunctive to antidepressant therapy in patients with major depressive disorder. J. Affect. Disord. 127, 19–30. Bauer, M., Pretorius, H.W., Constant, E.L., et al., 2009. Extended-release quetiapine as adjunct to an antidepressant in patients with major depressive disorder: results of a randomized, placebo-controlled, double-blind study. J. Clin. Psychiatry 70, 540–549. Beekman, A., de Beurs, E., van Balkom, A., et al., 2000. Anxiety and depression in later life: co-occurrence and communality of risk factors. Am. J. Psychiatry 157, 89–95. Berman, R.M., Fava, M., Thase, M.E., et al., 2009. Aripiprazole augmentation in major depressive disorder: a double-blind, placebo-controlled study in patients with inadequate response to antidepressants. CNS Spectr. 14, 197–206. Berman, R.M., Marcus, R.N., Swanink, R., et al., 2007. The efficacy and safety of aripiprazole as adjunctive therapy in major depressive disorder: a multicenter, randomized, double-blind, placebo-controlled study. J. Clin. Psychiatry 68, 843–853. Bollini, P., Pampallona, S., Tibaldi, G., et al., 1999. Effectiveness of antidepressants. Metaanalysis of dose-effect relationships in randomised clinical trials. Br. J. Psychiatry 174, 297–303. Brailean, A., Aartsen, M.J., Muniz-Terrera, G., et al., 2017. Longitudinal associations between late-life depression dimensions and cognitive functioning: a cross-domain latent growth curve analysis. Psychol. Med 47, 690–702. Cakir, S., Senkal, Z., 2016. Atypical antipsychotics as add-on treatment in late-life depression. Clin. Interv. Aging 11, 1193–1198. Carta, M.G., Zairo, F., Mellino, G., et al., 2007. Add-on quetiapine in the treatment of major depressive disorder in elderly patients with cerebrovascular damage. Clin. Pract. Epidemiol. Ment. Health 3, 28. Cronin-Stubbs, D., de Leon, C., Beckett, L., et al., 2000. Six-year effect of depressive symptoms on the course of physical disability in community-living older adults. Arch. Int. Med. 160, 3074–3080. Dorsey, E., Rabbani, A., Gallagher, S., et al., 2010. Impact of FDA black box advisory on antipsychotic medication use. Arch. Int. Med. 170, 96–103. Gardner, D., Murphy, A., O'Donnell, H., et al., 2010. International consensus study of antipsychotic dosing. Am. J. Psychiatry 167, 686–693. Gianfrancesco, F., Wang, R.H., Mahmoud, R., et al., 2002. Methods for claims-based pharmacoeconomic studies in psychosis. PharmacoEconomics 20, 499–511. Greenberg, P.E., Fournier, A.A., Sisitsky, T., et al., 2015. The economic burden of adults with major depressive disorder in the United States (2005 and 2010). J. Clin. Psychiatry 76, 155–162. Hsu, J.H., Chien, I.C., Lin, C.H., 2017. Increased risk of chronic obstructive pulmonary disease in patients with bipolar disorder: a population-based study. J. Affect. Disord. 220, 43–48. Kasper, S., Montgomery, S.A., 2013. Treatment-Resistant Depression. Wiley-Blackwell. Katila, H., Mezhebovsky, I., Mulroy, A., et al., 2013. Randomized, double-blind study of the efficacy and tolerability of extended release quetiapine fumarate (quetiapine XR) monotherapy in elderly patients with major depressive disorder. Am. J. Geriatr. Psychiatry 21, 769–784.

124