Long-Acting Injectables and Risk for Rehospitalization Among Patients Wi..

of 7
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  Long-Acting Injectables and Risk for Rehospitalization Among Patients With Schizophrenia in theHome Care Program in Taiwan  Po-Chung Ju, MD,* Þ  Frank Huang-Chih Chou, MD, MS, PhD, þ  § Te-Jen Lai, MD, PhD,* Þ  Po-Ya Chuang, MHA,|| Yung-Jung Lin, MS, ¶   Ching-Wen Wendy Yang, BS, #   and Chao-Hsiun Tang, PhD||  Abstract:  We aimed at evaluating the relationship between medicationand treatment effectiveness in a home care setting among patients withschizophrenia. Patients with schizophrenia hospitalized between 2004and 2009 with a primary  International Classification of Diseases, Ninth Revision, Clinical Modification  code of 295 were identified from Psy-chiatric Inpatient Medical Claims Data released by the National HealthResearch Institute in Taiwan. Patients who joined the home care pro-gram after discharge and were prescribed long-acting injection (LAI)(the LAI group) or oral antipsychotic medications (the oral group) wereincluded as study subjects. The final sample for the study included 810 participants in the LAI group and 945 in the oral group. Logistic regres-sion was performed to examine the independent effect of LAI medicationon the risk for rehospitalization within the 12-month observation windowafter controlling for patient and hospital characteristics and propensityscore quintile adjustment. The unadjusted odds ratio for rehospitalizationrisk was 0.80 (confidence interval, 0.65  Y  0.98) for the LAI group compared to the oral group. The adjusted odds ratio was further reduced to 0.78(confidence interval, 0.63  Y  0.97). Results remained unchanged when the propensity score quintiles were entered into the regression for further ad- justment. In a home care setting, patients treated with long-acting anti- psychotic agents are at a significantly lower risk for psychiatricrehospitalization than those treated with oral medication. Consequently,LAI home-based treatment for the prevention of schizophrenia relapsemay lead to substantial clinical and economic benefits. Key Words:  home care, logistic regression, long-acting injection,risperidone, schizophrenia(  J Clin Psychopharmacol   2014;34: 23  Y  29) S chizophrenia is a chronic and burdensome mental illnessthat can have a major impact on an affected person’s ca- pacity for social adaptation and occupational function. 1  Y  6 Relapse prevention and rehospitalization are important goals inthe treatment strategy for schizophrenia. 7 Many factors are in-volved in the treatment of schizophrenia; antipsychotic medications,for example, canbeusedtoeffectivelymanageschizophrenia, whentaken as recommended. 8 However, the high incidence of adverseeffects such as extrapyramidal symptoms induced by conventionalantipsychotics is a potential contributor to nonadherence. Problemswithintolerabilityassociatedwithantipsychoticmedicationsmaybeaddressed by using atypical formulations. Many studies demon-strate that patients administered atypical antipsychotics, which havea better treatment profile than conventional antipsychotics, 9 tend tohave better adherence patterns than those administered typical an-tipsychotics. 10,11  Nonetheless, contradictory findings indicate that  patients’ opinions related to tolerability may not always be inagreement with doctors’ opinions. 12  Y  14  Noncompliance is a common problem in mental illness; ap- proximately 55% of patients with schizophrenia either fail to attend clinics or are unable to complete treatment plans. 15  Y  17 This lack of compliance is related to the increased risk for relapse, high medicalcosts, and poor outcomes. 18 The recent development of long-actinginjection (LAI) of antipsychotics has helped to improve medicationcompliance and may potentially improve treatment outcomes and minimize economic burden. 19 Long-acting depot formulations of conventional antipsychotic drugs have shown to be effective for reducing positive symptoms and relapse rates. 17,20 The first long-acting atypical antipsychotic, risperidone, 21  Y  23 has been shown tooffer advantages compared to not only to conventional depot for-mulations but also oral formulations of atypical antipsychoticmedications. 8 A summary on the effects of LAIs on clinicalsymptoms is found in Supplementary Table 1 (Supplemental Dig-ital Content 1, viable treatment strategy that was developed re-cently for patients with mental illness is home care service,which has been shown to have a positive influence on the livesof patients with schizophrenia, including improvements in psychotic symptoms, social function, interpersonal interaction,families’ perceived stress, living skills, medication compliance,and lower rates of rehospitalization. 24  Y  26 Home care serviceshave also been shown to improve both treatment adherenceand effectiveness. 24  Y  26 However, it is not clear if LAI treatment could further improve treatment outcomes for patients receivinghome care services. This study uses a population-based healthinsurance database to evaluate the effectiveness of LAI use inreducing rehospitalization among patients with schizophreniawho enrolled in the home care program under the NationalHealth Insurance (NHI) in Taiwan. MATERIALS AND METHODSData Sources The data set used in this study was compiled from Psy-chiatric Inpatient Medical Claims Data (PIMC) released by the O RIGINAL  C ONTRIBUTION  Journal of Clinical Psychopharmacology   &  Volume 34, Number 1, February 2014  23 From the *Department of Psychiatry, Chung Shan Medical University Hos- pital;  † Institute of Medicine, Chung Shan Medical University, Taichung; ‡ Department of Community Psychiatry, Kaohsiung Municipal Kai-SyuanPsychiatric Hospital, Kaohsiung City; §Graduate Institute of Health Care, MeihoUniversity, Ping-Tong County; ||School of Health Care Administration, TaipeiMedical University, Taipei;  ¶  Division of Corporate Communications and Government Affairs, Janssen Taiwan, Taipei, Taiwan; and #School of HealthSciences, University of California  Y  Irvine, Irvine, CA.Received November 1, 2011; accepted after revision April 23, 2013.Reprints: Chao-Hsiun Tang, PhD, School of Health Care Administration,Taipei Medical University, 250 Wu-Xin St, Taipei, Taiwan 110(e < mail:,theauthorshave retained total independence in the preparation of this manuscript.Supplemental digital content is available for this article. Direct URL citationappears inthe printed textandis provided inthe HTML and PDFversionsof this article on the journal’s Web site (   *  2014 by Lippincott Williams & WilkinsISSN: 0271-0749DOI: 10.1097/JCP.0b013e3182a6a142 Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.   National Health Research Institute in Taiwan. Longitudinalclaims data covering the period between January 1, 1996, and December 31, 2009, were collected for patients who were ad-mitted to the hospital for psychiatric illness at least once withina specified time frame. 27 Two cohorts based on the period during which the patients were admitted were created. The first cohort (PIMC_1) consisted of 91,104 patients with mental illnesswhowere admitted to the hospitalwith  International Classificationof Diseases, Ninth Revision, Clinical Modification  (  ICD-9-CM  )codes 290 to 319 or A codes A210 to A219 at least once betweenJanuary 1, 1996, and December 31, 2001. The second cohort (PIMC_2) consisted of 96,013 patients with mental illness whowere admitted with  ICD-9-CM   codes 290 to 319 at least once between January 1, 2002, and December 31, 2007. Study Cases and Comparison Groups Thefollowinginclusion/exclusioncriteriawereusedtoselect study participants from the PIMC: First, patients who joined the psychiatric home care program within 60 days after hospital dis-charge between 2004 and 2009 were initially included in thisstudy (5245 patients). Second, the date of the first home care/ outpatient visit and initial psychiatric prescription was defined asthe index prescription date (hereafter the index date). Patientswere classified into the following treatment groups: patients whohad at least 1 home care visit with  ICD-9-CM   code 295 and were prescribed at least 75 mg of risperidone LAI (RLAI) or whowere prescribed conventional LAI (CLAI) antipsychotics twice within3 months after the index date were categorized under the LAIgroup. Patientswhohadatleast 1homecarevisitwith  ICD-9-CM  code 295 and were prescribed oral antipsychotics (identified inthe Anatomical Therapeutic Chemical Classification System as N05A) with a medication possession ratio (MPR) of greater than0.5 within 3 months of the index date were assigned to the oralgroup. Medication possession ratio is defined as follows: MPR =(total number of days for which drugs were prescribed duringhome care/outpatient visits within 3 months of the index date  j total number of overlapped days during which drugs were pre-scribed within 3 months of the index date) / (90  j  hospitalizationdays within 3 months of the index date)  100. In the nominator of MPR, thenumber of overlapped days of drugusewas subtracted to account for the fact that patients might engage in ‘‘doctor shop- ping’’ behavior in the NHI system. 28 At this point, patientswho did not comply with inclusion criteria for the LAI group and the oralgroup were excluded from the study (1890 patients).Third, patients whose index date was after October 1, 2008,were excluded because their medication use could not be traced for 3 months nor could they be monitored for rehospitalizationfor 12 months after their index date. Fourth, patients who werehospitalized for nonpsychiatric reasons within 12 months of their index date were excluded from the analysis because theywere automatically disenrolled from the home care programaccording to the rules set by the Bureau of NHI. Fifth, patientswho had been prescribed clozapine at any point between 2004and 2009 were excluded because they were considered part of a patient group with treatment resistance and, consequently, theworst therapeutic potential. Finally, patients withdrawn from the NHI program within 12 months after their index date were ex-cluded. The final sample available for analysis included 810 patients in the LAI group and 945 patients in the oral group.The selection process for identifying study participants isoutlined in Figure 1. Variable Definitions The main outcome variable assessed in this study was probability for psychiatric rehospitalization (with principal  ICD-9-CM   codes of 290-319) within a 12-month window of observation between the fourth and 15th months after a patient’sindex date. The 3-month period after the index date was defined as awashout period to minimize the effects of previous medications.Patients were divided into four 5-year age groups, rangingfrom 30 years of age and younger to 50 years of age and older.To account for patients’ psychiatric health status, they werefurther classified into the following groups according to their use of psychiatric medical services during the 12 months beforetheir index date: number of psychiatric hospitalizations (0, 1,and   Q 2), number of days spent in the hospital for psychiatricreasons ( e 30, 31  Y  60, and   9 60), and number of psychiatric emer-gencydepartmentvisits(0,1,and  Q 2).Toaccountforinstitutionaleffects, the patients were further divided into subgroups based onthe ownership (public, nonprofit, or private) and level of accred-itation (medical centers, regional hospitals, or district hospitals)of the institutions to which they were admitted. Statistical Analysis The characteristics of the patients, physicians, and hospi-tals included in this study were described. To control for selec-tion bias, a logistical regression model was used to estimate the propensity scores (PSs) of patients, which is defined as the prob-ability of receiving LAI versus oral treatment based on the char-acteristics of patients, physicians, and hospitals. 29,30 The resultsof logistic regression based on which the PSs is estimated wasreported in Supplementary Table 2 (Supplemental Digital Content 2, Patients in the LAI groupand in the oral group were divided into 5 strata using quintilesof the estimated PSs. To evaluate the PS, a logistic regression was performedonwhethertheoutcomebeingtheLAIgroup,withand without adjustment for PS strata, for each of the confoundingvariables considered in assessing the risk of rehospitalization.After controlling for patient, physician, and hospital char-acteristics and PS quintile adjustment, a logistic regression was performed to examine the independent effect of LAI medicationon the probability of rehospitalization within the 12-monthobservation window. All analyses were performed using theStatistical Analysis Software/Stat System for Windows, version9.01 (SAS Institute, Cary, NC). RESULTS Patients’ and providers’ characteristics at the index date before and after PS adjustments are reported by treatment groups in Table 1. After adjustment for PS, there were no sta-tistically significant differences in patients’ and providers’characteristics. During the 12-month observation window, 228 patients (28.2%) in the LAI group and 311 patients (32.9%) inthe oral group were hospitalized for psychiatric services.Table 2 shows the results of logistic regression on the probability of any psychiatric rehospitalization within the12-month observation window after the index date. In the simplelogistic regression model, the patients in the LAI group were20% less likely to be rehospitalized within the given timeframethan those in the oral group (odds ratio [OR], 0.80; confidenceinterval [CI], 0.65  Y  0.98). After controlling for patients’ and hospital characteristics, the OR for the probability of re-hospitalization for the patients in the LAI group reduced to 0.78(CI, 0.63  Y  0.97). The results remained unchanged when the PSquintiles were entered into the regression model for further adjustment. DISCUSSION This study involved the use of a population-based healthinsurance database to demonstrate that Taiwan patients with  Ju et al   Journal of Clinical Psychopharmacology   &  Volume 34, Number 1, February 2014 24  *  2014 Lippincott Williams & Wilkins Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.  schizophrenia who were under LAI treatment had a 22% lower risk for rehospitalization than those under oral antipsychotictreatment in a home care setting. The results remained un-changed when patients’ and providers’ characteristics and PSquintiles were entered as covariates into a regression model. Toour knowledge, this study is the first to evaluate this therapeuticstrategy in a home care setting to determine if LAI is compar-atively effective in reducing rehospitalization among patientswith schizophrenia in a home care program.Our findings are consistent with those of previous studiesthat compared the efficacy between conventional depot anti- psychotics or RLAI and oral antipsychotics in schizophrenia patient groups. Some studies have shown a significant benefit of conventional depot antipsychotics compared to short-actingoral medication, including global improvement in function, lower relapse rates, and a lower incidence of movement disorder. 20,31,32 Other studies showed that patients with schizophrenia who weretreated with RLAI had better symptom improvement, fewer hospitalizations, and a better adverse effect profile than thosetreated with oral risperidone. 33  Y  36 In agreement with thesefindings, our study shows that even among patients who had received home care services, LAI treatment had a significantlyindependent effect on reducing rehospitalization compared tooral treatment. Previous studies have indicated that long-actingagents can help patients avoid the variability associated withabsorption and first-pass metabolism and usually result in a better correlation between the administrated dose and plasmalevels achieved. 32,37 Once a steady state is achieved, plasmalevels remain relatively stable, and daily peaks and troughs that occur with oral antipsychotics do not occur. 38 The lack of effect across positive, negative, cognitive, and mood-related symptoms, as well as a substantial associationwith extrapyramidal adverse effects, continues to be cause for clinical concern surrounding CLAI use. In contrast, RLAI,which combines the benefits of an atypical antipsychotic withthose of a long-acting injectable formulation, has been shown to be efficacious in improving the mean Positive and NegativeSyndrome Scale total score and Extrapyramidal SymptomRating Scale score even for patients whose conditions have been stabilized using CLAI. 31 This finding had led us to further explore the comparative effectiveness of RLAI versus that of CLAI among patients with schizophrenia in a home caresetting. In a separate analysis (data not shown), we compared the risk for rehospitalization between patients in the CLAI and RLAI groups, but no significant difference was observed. Thismay be due to the fact that the risk for rehospitalization is FIGURE 1.  Patient selection process.  Journal of Clinical Psychopharmacology   &  Volume 34, Number 1, February 2014  Long-Acting Injectables and Rehospitalization *  2014 Lippincott Williams & Wilkins  25 Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.  TABLE 1.  Patients’ and Providers’ Profiles at the Index Visit Total LAI Oral  P   BeforeAdjustmentfor PSQuintiles  P  AfterAdjustmentfor PSQuintilesNo. Patients Percent No. Patients Percent No. Patients Percent Patients’ characteristicsSex 0.314 0.866Female 781 44.5 350 43.21 431 45.61Male 974 55.5 460 56.79 514 54.39Age, y 0.007 0.995 G 30 264 15.04 109 13.46 155 16.430  Y  39 578 32.93 278 34.32 300 31.7540  Y  49 553 31.51 278 34.32 275 29.1 Q 50 360 20.51 145 17.9 215 22.75Mean (SD) 41.48 (11.19) 41.13 (10.05) 41.78 (12.08)Medical use within 1 year before the index date No. psychiatric hospitalizations 0.024 0.9520 354 42.56 163 20.12 191 20.211 944 28.09 412 50.86 532 56.3 Q 2 457 29.34 235 29.01 222 23.49Mean (SD) 1.21 (1.07) 1.25 (1.11) 1.18 (1.02) No. psychiatric hospitalization days 0.006 0.922 e 30 747 42.56 323 39.88 424 44.8731  Y  60 493 28.09 219 27.04 274 28.99 9 60 515 29.34 268 33.09 247 26.14Mean (SD) 47.03 (45.49) 50.40 (48.45) 44.14 (42.6) No. psychiatric emergency department visits 0.067 0.9730 1366 77.83 611 75.43 755 79.891 274 15.61 143 17.65 131 13.86 Q 2 115 6.55 56 6.91 59 6.24Mean (SD) 0.33 (0.76) 0.35 (0.73) 0.31 (0.78)Physicians’ characteristicsSex 0.702 0.965Female 229 13.05 103 12.72 126 13.33Male 1526 86.95 707 87.28 819 86.67Age, y 0.823 0.999 G 35 528 30.09 249 30.74 279 29.5235  Y  39 483 27.52 226 27.9 257 27.240  Y  44 388 22.11 178 21.98 210 22.22 Q 45 356 20.28 157 19.38 199 21.06Mean (SD) 39.62 (7.64) 39.43 (7.55) 39.78 (7.72)Hospital characteristicsOwnership  G 0.0001 0.737Public 988 56.3 441 54.44 547 57.88 Nonprofit 522 29.74 299 36.91 223 23.6Private 245 13.96 70 8.64 175 18.52Level of hospital accreditation  G 0.0001 0.653Medical center 243 13.85 158 19.51 85 8.99Regional hospital 1080 61.54 493 60.86 587 62.12District hospital 432 24.62 159 19.63 273 28.89Level of urbanization  G 0.0001 0.9651 406 23.13 197 24.32 209 22.122 610 34.76 292 36.05 318 33.653 383 21.82 195 24.07 188 19.89 Q 4 356 20.28 126 15.56 230 24.34 ( Continued on next page  )  Ju et al   Journal of Clinical Psychopharmacology   &  Volume 34, Number 1, February 2014 26  *  2014 Lippincott Williams & Wilkins Copyright © 2013 Lippincott Williams & Wilkins. Unauthorized reproduction of this article is prohibited.
Similar documents
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!