Psychotropic Polypharmacy for the Treatment of Bipolar Disorder in Taiwan

We used the nationwide psychiatric inpatient medical claims in Taiwan’s National Health Insurance Research Database, 2000 to 2007, for all patients with a diagnosis of bipolar disorder (ICD-9-CM codes 296.0–296.16, 296.4–296.81, and 296.89) who had been hospitalized during the study period. Patients with a primary diagnosis of bipolar disorder at admission were enrolled, and patients with a concurrent diagnosis of schizophrenia (ICD-9-CM code 295) were excluded. We examined the patients’ first outpatient prescription after discharge from their first psychiatric hospitalization for mood stabilizers, antipsychotics, or antidepressants. [A complete list of medications identified in the claims data is available online as a data supplement to this report.] Consequently, for each patient, only one prescription (which could contain more than one medication) was documented. For each prescription, we determined if between-class polypharmacy and within-class polypharmacy were present. Between-class polypharmacy referred to the prescription of two or more classes of medication, and within-class polypharmacy referred to the prescription of two or more medications in the same class.

Multivariable regression analysis was applied to examine potential predictors of both between-class polypharmacy and within-class polypharmacy, including sex, age, number of psychiatric diagnoses, outpatient clinic setting, hospitalization length, and prescribed dose. Outpatient clinic setting was categorized into medical centers (>500 beds, N=18), regional hospitals (250–500 beds, N=74), and district hospitals (<250 beds, N=249). The daily dose of each medication was defined by dividing the prescribed daily dose (PDD) by the defined daily dose (DDD), which represented the average maintenance daily dose of a medication for an adult (13). We then obtained, for each prescription, an average daily dose of medication by adding up the PDD:DDD ratio for each drug and dividing by the number of drugs contained. Last, logistic regression was used to examine the one-year readmission rate as a function of polypharmacy. A p value of <.01 was considered to be statistically significant. Statistical analyses were performed with the SAS software, version 9.1 for Windows.

Our study sample consisted of 5,449 Chinese patients with bipolar disorder (51% women; mean±SD age=36.8±12.4). A total of 5,155 (95%) were diagnosed as having bipolar I disorder. Among them, 2,923 (57%) presented with a manic episode; 1,076 (21%), depressed episode, 786 (15%), mixed episode; and 370 (7%), unspecified episode. Of our study patients, 4,331 (80%) received mood stabilizers, 4,031 (74%) received antipsychotics, and 1,192 (22%) received antidepressants.

For 3,853 (71%) patients, prescriptions involved between-class polypharmacy, and for 941 (17%), prescriptions contained within-class polypharmacy. The combination of mood stabilizer and antipsychotic was the most common type of between-class polypharmacy (N=2,825, 52%), followed by mood stabilizer, antipsychotic, and antidepressant (N=456, 8%), mood stabilizer and antidepressant (N=324, 6%), and antidepressant and antipsychotic (N=248, 5%). Polypharmacy with mood stabilizers (N=623, 11%) was the most frequently prescribed type of within-class polypharmacy, followed by antipsychotics (N=296, 5%) and antidepressants (N=68, 1%). The daily dose for each medication was .62±.29 for mood stabilizers, .77±.02 for antipsychotics, and 1.24±.82 for antidepressants. These data indicated that doses of mood stabilizers and antipsychotics were, on average, lower than the recommended dose, and doses of antidepressants were more likely to be higher than recommended. The average daily dose per prescription was .7±.9.

Multivariable regression analysis showed that patients older than 50 years and those who were treated at medical centers had a lower likelihood of receiving between-class polypharmacy. Each unit increase of average daily dose halved the odds of both between-class polypharmacy (odds ratio [OR]=.50, p<.001) and within-class polypharmacy (OR=.57, p<.001) (Table 1). Finally, neither between-class polypharmacy nor within-class polypharmacy was associated with a higher rate of readmission within one year.

Our findings suggest that over 70% of psychotropic regimens among recently discharged patients with bipolar disorder involved between-class polypharmacy. Within-class polypharmacy was present in nearly 20% of prescriptions. Notably, a lower prescribed dose was strongly associated with the chance of polypharmacy.

The strength of our study lies in its illustration of various types of polypharmacy among a nationally representative sample consisting mostly of inpatients with bipolar I disorder. Hospitalized patients are likely to have a more solid diagnosis of bipolar disorder, but the psychotropic regimen is specifically used to relieve an acute mood episode. Meanwhile, our data require cautious interpretation, given that they may not be generalizable to patients with bipolar disorder other than bipolar I or to patients who have not been hospitalized.

In our findings, mood stabilizers accounted for a majority of polypharmacy among hospitalized patients, in contrast to outpatient regimens, in which antidepressants are the most frequently used medications for polypharmacy with another psychotropic (2,8). This finding is concordant with the clinical states of our study population. Over 70% of our patients suffered from an exacerbation of a manic or mixed episode, whereas many outpatients may be dealing with unresolved depression (2). Consistent with the report from Biancosino and colleagues (12), our data suggest that older patients are less likely to receive polypharmacy. One possible explanation is that the prescription of polypharmacy for the elderly is prevented by comorbid general medical illnesses or intolerable side effects. In addition, receiving treatment at a medical center is a protective factor for polypharmacy. As shown in the Systematic Treatment Enhancement Program for Bipolar Disorder studies, academic teaching hospitals are more likely to provide guideline-concordant care and may, therefore, be more reserved in prescribing polypharmacy (14).

Furthermore, receiving polypharmacy did not predict a higher risk of readmission within one year, indicating that patients treated with polypharmacy do not necessarily have a more severe course of disease. Future research is needed to determine if severity of disease predicts polypharmacy or if polypharmacy brings substantial long-term benefits.

The association between underdosing and polypharmacy calls for special notice. The most common mode of psychiatric polypharmacy is the “add-on” strategy to boost or accelerate therapeutic effect (11,15), the trade-off being potentially complex drug-drug interactions, intolerable side effects, higher cost, and decreased adherence (10). In our findings, polypharmacy was accompanied by the underdosing of mood stabilizers and antipsychotics, indicating the importance of considering whether to raise a single medication to its full therapeutic dose before proceeding to polypharmacy. Such an approach would also be more concordant with guideline recommendations. In contrast, higher-than-recommended prescribed doses of antidepressants warrant caution because of their undetermined efficacy for bipolar depression and the associated risk of triggering hypomania and escalation of cycling (2).

This study suggested that there is substantial use of various forms of polypharmacy in the treatment of inpatients with bipolar disorder. We suggest that randomized studies be used to compare the cost-effectiveness of common psychotropic combinations and monotherapy to treat bipolar disorder.

This research was supported by grants (99001-62-049) from Taipei City Hospital, Taipei, Taiwan. This study is based in part on data from the National Health Insurance Research Database provided by the Bureau of National Health Insurance, Department of Health, and managed by the National Health Research Institutes. The interpretation and conclusions contained herein do not represent those of the Bureau of National Health Insurance, Department of Health, or the National Health Research Institutes.

The authors report no competing interests.