High‐dose antipsychotic drug use as a predictor for readmission of inpatients with borderline personality disorder: A retrospective chart review in a Japanese psychiatric hospital

Abstract Objective This study aimed to determine predictors associated with readmission of inpatients with borderline personality disorder. Methods This observational study evaluated 83 inpatients with borderline personality disorder admitted to the National Center of Neurology and Psychiatry Hospital in Japan from January 2013 to January 2016. Data were retrospectively obtained from electronic medical records. Results There was no significant difference in the daily antipsychotic dose equivalent to chlorpromazine at admission between the readmitted and nonreadmitted groups, which indicated that there was no between‐group difference in the psychiatric disease severity at admission. Multivariate logistic regression analyses revealed that the use of antipsychotics equivalent to >400 mg of chlorpromazine at discharge was associated with readmission within 1 year. Conclusions In conclusion, high‐dose antipsychotic drug use at discharge may be a risk factor for readmission. The present findings may have important clinical implications since they alert physicians to a possible predictor for readmissions of patients with borderline personality disorder.

medication increased the length of hospitalization; contrastingly, mood stabilizer usage reduced it. 8 However, there is little available information regarding the predictors associated with readmission of patients with BPD. Therefore, this study aimed to evaluate predictors associated with readmission of inpatients with BPD.

| ME THODS
This study was approved by the Clinical Research Ethics Committee of the National Center of Neurology and Psychiatry Hospital (A2017-039).

| Study design and data sources
We evaluated 83 inpatients diagnosed with BPD based on the ICD-

Classification of Mental and Behavioral Disorders in the National
Center of Neurology and Psychiatry in Japan from January 2013 to January 2016 (see Table S1). The requirement for informed consent was waived, and an opt-out option was availed since the data were retrospectively obtained from electronic medical records.

| Study variables
The study variables included the following: age, sex, family history of psychiatric disorder, education level, job career, history of alcohol drinking, history of smoking, history of drug abuse, self-harm during hospitalization, Wechsler Adult Intelligence Scale scores (full scale, verbal, and performance IQ [intelligence quotient]), diagnosis on admission, psychiatric comorbidity, age at first-visit of a psychiatric hospital, hospitalization duration, duration of preceding hospitalization, readmission within 1 year, and prescribed drugs at admission and discharge. We classified prescribed drugs into the following four categories: antipsychotics, mood stabilizers, antidepressants, and benzodiazepines. Furthermore, the equivalent doses of chlorpromazine, diazepam, and imipramine at admission and discharge were calculated (see Table S2). 9  Table 1 summarizes the demographic features of the inpatients.

| RE SULTS
There were 110 and 83 inpatients with personality disorder (PD) and BPD, respectively (see Table 1 and Table S1). 10 Among the inpatients with BPD, 32 (38.6%) individuals were readmitted to hospital within 1 year (see Table 1). Moreover, there was no significant between-group difference in the daily antipsychotic doses equivalent to chlorpromazine at admission (readmitted group: max = 1200, median = 68; nonreadmitted group: max = 1132.7, median = 43), which indicated no between-group difference in the psychiatric disease severity at admission. Independent-sample t tests and Fisher's exact tests revealed significant between-group differences among patients with BPD in the duration of antecedent hospitalization (P = .001) and daily antipsychotic dose >400 mg equivalent to chlorpromazine at discharge (P = .005) ( Table 1). Contrastingly, there were no significant between-group differences among the patients with PDs included other than BPD (see Table 1). During hospitalization, benzodiazepines were the most frequently prescribed drugs, followed by antipsychotics, mood stabilizers, and antidepressants (see Table 2). Further, data about psychosocial interventions were not obtained from the electronic medical records, as it was difficult to quantitatively or qualitatively evaluate them.
Multivariate logistic regression analyses confirmed that the use of antipsychotics equivalent to >400 mg of chlorpromazine at discharge increased the odds of readmission within 1 year (see Table 3).

| D ISCUSS I ON
This study examined predictors associated with readmission of inpatients with BPD. Since the hospitalization duration should be as short as possible 4 due to its poor effectiveness 3,5 and economic burden, 6 our findings may have important implications for future efforts of reducing the hospitalization frequency and duration among patients with BPD.
To our knowledge, this is the first study to identify a predictor of increased readmission for patients with BPD. Using the

Key points
• There is limited available information regarding predictors associated with repetitive admission of patients with BPD. This study aimed to evaluate predictors associated with readmission of inpatients with BPD.
• Multivariate logistic regression analyses revealed that the use of antipsychotics equivalent to >400 mg of chlorpromazine at discharge was associated with readmission within 1 year.
• High-dose antipsychotic drug use at discharge may be a risk factor for readmissions. daily antipsychotic dose equivalent to chlorpromazine at admission, we estimated the approximate severity of psychiatric diseases at admission in the readmitted and nonreadmitted groups.
Multivariate logistic regression analyses revealed that the use of antipsychotics equivalent to >400 mg of chlorpromazine at discharge was associated with increased readmission within 1 year (see Table 3). to 227 mg of chlorpromazine was found to significantly reduce the severity of BPD symptoms. 12 Therefore, the present study tested the effectiveness of a higher antipsychotic dose equivalent to >400 mg of chlorpromazine.
Although the mechanism underlying BPD treatment using antipsychotics remains unclear, second-generation antipsychotics are commonly prescribed in BPD patients. 13  Note: Antipsychotics, antidepressants, and benzodiazepines were converted into equivalent chlorpromazine, imipramine, and diazepam doses, respectively.

TA B L E 3
Correlation between readmission and use of antipsychotics equivalent to >400 mg of chlorpromazine in the inpatients with borderline personality disorder consistent with our results. This could be attributed to several diffi- In conclusion, the use of high-dose antipsychotic drug use at discharge may be a risk factor for readmissions. Our findings may have important clinical implications since they alert physicians to possible predictors for readmissions of inpatients with BPD.

ACK N OWLED G M ENTS
We would like to thank all the participants who took part in the treatment and the staff at the National Center Hospital, National Center of Neurology and Psychiatry in Tokyo, Japan.

CO N FLI C T O F I NTE R E S T
The authors declare no competing interests.

AUTH O R CO NTR I B UTI O N S
Yuji Yamada and Yuma Yokoi planned and designed the study. Yuji Yamada collected the data and drafted the first manuscript. Yuma Yokoi and Zui Narita analyzed the data. Yuma Yokoi, Zui Narita, and Naotsugu Hirabayashi critically reviewed the draft and revised it.
All authors made substantial contributions and approved the final manuscript.

A PPROVA L O F TH E R E S E A RCH PROTO CO L BY A N I N S TITUTI O N A L R E V I E WER B OA R D
This study was approved by the Clinical Research Ethics Committee of the National Center of Neurology and Psychiatry Hospital (A2017-039).

I N FO R M E D CO N S E NT
The requirement for informed consent was waived, and an opt-out option was availed since the data were retrospectively obtained from electronic medical records.

DATA AVA I L A B I L I T Y S TAT E M E N T
The raw data belonged to the present study cannot be made publicly available, because the disclosure of personal data was not included in the research protocol of the present study.