Aim: The population-based National Health Insurance database was used to investigate the prevalence, correlates, and disease patterns of antipsychotic use in Taiwan.
Methods: The National Health Research Institutes provided a database of 200 000 random subjects for study. A random sample of 145 304 subjects was obtained, aged ≥18 years in 2004. Study subjects who had been given at least two antipsychotic drug prescriptions during this year were identified. The factors associated with any antipsychotic use were identified. The proportion of antipsychotic use for psychiatric and non-psychiatric disorders was also examined.
Results: The 1-year prevalence of antipsychotic use was 3.5%. Antipsychotic use was found to be more prevalent by age; for women; for individuals with a lower insurance amount; for individuals with disability; and among those subjects who lived in the central or southern area. Among subjects with antipsychotic use, higher proportions of psychiatric disorders were found for schizophrenia, anxiety state, major depressive disorder, neurotic depression, dementia, and bipolar disorder. With respect to medical disorder, higher proportions of antipsychotic use were found for diseases of the digestive system; symptoms, signs, and ill-defined conditions; diseases of the respiratory system, musculoskeletal system and connective tissue, circulatory system, nervous system and sense organs, and genitourinary system.
Conclusions: Higher proportions of antipsychotic use were found for schizophrenia, depressive disorder, anxiety disorder, dementia, and bipolar disorder. More than 60% of subjects used antipsychotics for non-psychiatric disorders in Taiwan, which deserves further study for the sake of patient safety.
THE USE OF antipsychotic medications has increased worldwide in past decades. The overall 1-year prevalence of antipsychotic prescription varies worldwide, ranging from 0.5% to 1.2%.1–4 Moreover, Hamann et al. conducted a retrospective analysis using a large outpatient database in Germany and reported that up to 6% of the population had been prescribed an antipsychotic between 1999 and 2001.5 Second-generation antipsychotics have been widely prescribed in the past 10 years due to better tolerance, resulting in fewer cognitive and neurological side-effects.6 Consequently second-generation antipsychotics have accounted for an increased proportion of total antipsychotic prescriptions.5,7,8
The antipsychotic use rate differs between sexes. Kaye et al. reported that the average annual antipsychotic use was higher in women than men.9 Antipsychotic drug prescription also varies by age group. In an Italian epidemiological study conducted in 2001, antipsychotic use progressively rose with age in both sexes, especially among the old and very old subjects.4 Similarly, older persons had the highest prevalence in a 4-year survey period based on the general population.10 Additionally, several studies reported race and socioeconomic status (SES) disparities with secondary-generation antipsychotic use.11–14
There are heterogeneous indications for antipsychotics between or even within countries, thus ‘off-label use’, defined as prescribing outside the licensed indications, varies greatly worldwide.15,16 For example, risperidone and olanzapine are both classified as second-generation antipsychotics but have different indications in the UK. Further, the licensed uses for quetiapine, risperidone, and olanzapine were not the same in the UK and Italy. Considering the off-label use, antipsychotic agents have been ordered as a tranquilizer (e.g. for sleep disturbances) or an anxiolytic, partly reflecting clinicians' concern about dependence on benzodiazepine.17
A cross-national community survey, the Taiwan Psychiatric Epidemiological Project, conducted from 1981 to 1986, found that the lifetime prevalence rate of any psychiatric disorder was 21.56%.18 No antipsychotic use data, however, were included in the Taiwan Psychiatric Epidemiological Project report. Taiwan implemented the National Health Insurance (NHI) program, a comprehensive, unified, universal health insurance program for all citizens of Taiwan in March 1995. Therefore, a certain amount of risk pooling for national health insurance should be expected. All citizens with at least 4 months of established legal residency in the Taiwan area may participate in the NHI program. The Bureau of NHI has contracted with 91% of medical institutions in Taiwan. As many as 96% of residents of Taiwan have joined the NHI program since 1996.
In the present study we first investigated the prevalence of antipsychotic use in 2004. Second, we identified the factors associated with any antipsychotic use. Third, we examined the proportion of antipsychotic use for psychiatric disorders. Fourth, we examined the proportion of antipsychotic use for medical disorders.
Data sources and sample
The National Health Research Institutes provided a database of medical claims from 200 000 random subjects, approximately 1% of the population, for use in health service studies. The data consisted of ambulatory care and inpatient care files, as well as registration records of the insured. There were no statistically significant differences in age, sex, and health-care costs between the sample group and all enrollees.
We conducted a population-based, random-sample study using antipsychotic drug utilization data from 1 January 1 through 31 December 2004. The study subjects were limited to those aged ≥18 years on 1 July, and we excluded individuals who died or were foreigners. Finally, 145 304 enrollees were studied in this cohort. The study was approved by the Jianan Mental Hospital Institutional Review Board.
Assessment of antipsychotic use
Those subjects who had received at least two antipsychotic drug prescriptions either in ambulatory care or in inpatient care in 2004 were identified. Antipsychotic drugs were recorded based on the Anatomical Therapeutic Chemical classification system.19 Data presented in this article include first-generation antipsychotics (butyrophenone, phenothiazine, thioxanthene, and miscellaneous antipsychotics such as pimozide, loxapine, and sulpyride) and second-generation antipsychotics (clozapine, risperidone, olanzapine, quetiapine, ziprasidone, amisulpride, and aripiprazole). In this study we examined the prevalence of antipsychotic drug use.
We obtained the characteristic data, including age, sex, race, insurance amount, disability, region, and urbanicity from the Bureau of NHI database. Age was classified into one of four categories: 18–24 years, 25–44 years, 45–64 years, and ≥65 years. Data on race included indigenous and non-indigenous ethnicity. The insurance amount was classified into one of four categories: fixed premium, <US$640 (<NTD 20 000), US$640–1280 (NTD 20 000–39 999), and ≥US$1281 (≥NTD 40 000). Data on disability included no disability and with disability. Study subjects were classified according to geographical distribution into one of four regions: northern, central, southern, and eastern. Urbanicity was divided into urban, suburban, and rural categories. We also identified the factors associated with any antipsychotic use.
Assessment of psychiatric disorder
Those subjects with antipsychotic use in the present study who had at least one service claim in the year 2004 for either ambulatory care or inpatient care, with a primary or secondary diagnosis of a psychiatric disorder, were identified.20 We divided psychiatric disorders into major psychiatric disorders, minor psychiatric disorders, and mental retardation.21 Major psychiatric disorders included ICD-9-CM codes 290–299; minor psychiatric disorders included ICD-9-CM codes 300–316; and mental retardation included ICD-9-CM codes 317–319. Additionally, we examined the proportion of psychiatric disorders in the use of antipsychotics.
Assessment of medical disorder
Subjects with antipsychotic use in the present study who had at least one service claim in the year 2004 for either ambulatory care or inpatient care, with the primary diagnosis of a medical disorder (excluding ICD-9-CM codes 290 through 319), were identified.20 In addition, we examined the proportion of medical disorders in the use of antipsychotics according to the medical disease system.
Multiple logistic regression was used to assess the factors associated with any antipsychotic use. The models were fitted to the data using SAS/ETS for Windows, version 9.1 (SAS Institute, Cary, NC, USA). The significance level was set at 0.05.
Table 1 lists the demographic characteristics of the study sample, including age, sex, race, insurance amount, disability, region, and urbanicity.
Table 1. Subject characteristics
|Age (years)|| || |
| 18–24||22 067||15.2|
| 25–44||63 039||43.4|
| 45–64||41 630||28.6|
| ≥65||18 568||12.8|
|Sex|| || |
| Male||73 013||50.2|
| Female||72 291||49.8|
|Race|| || |
| Non-indigenous||143 023||98.4|
| Indigenous||2 281||1.6|
|Insurance amount (US$)†|| || |
| Fixed premium||23 538||16.5|
| <640 (<20 000 NTD)||61 112||42.8|
| 640–1280 (20 000–39 999 NTD)||32 856||23.0|
| ≥1281 (≥40 000 NTD)||25 296||17.7|
|Disability|| || |
| No||139 784||96.2|
| Yes||5 520||3.8|
|Region‡|| || |
| Northern||66 064||46.0|
| Central||27 605||19.2|
| Southern||46 139||32.1|
| Eastern||3 844||2.7|
|Urbanization‡|| || |
| Urban||73 668||51.3|
| Suburban||23 609||16.4|
| Rural||46 375||32.3|
Table 2 shows the prevalence of antipsychotic use according to age, sex, race, insurance amount, disability, region, and urbanicity, respectively. Table 2 also lists the logistic regression of factors associated with the prevalence of antipsychotic use. Antipsychotic use was found to be more prevalent by age; for women; for individuals with a lower insurance amount; for subjects with disability; and among individuals who lived in the central region or the southern region. No significant difference was associated with race and urbanicity in the prevalence of antipsychotic use in Taiwan.
Table 2. Prevalence and correlates of antipsychotic drug use
|Age (years)|| || || |
|Sex|| || || |
|Race|| || || |
|Insurance amount (US$)|| || || |
| Fixed premium||5.1||1.85***||1.67–2.05|
| <640 (<20 000 NTD)||3.8||1.43***||1.30–1.58|
| 640–1280 (20 000–39 999 NTD)||2.7||1.18**||1.06–1.31|
| ≥1281 (≥40 000 NTD)||2.3||1.00||–|
|Disability|| || || |
|Region|| || || |
|Urbanicity|| || || |
|Total||3.5|| || |
Table 3 lists the proportion of psychiatric disorders among subjects with antipsychotic use, including any major psychiatric disorder, any minor psychiatric disorder, and any mental disorder. With respect to major psychiatric disorder, higher proportions of antipsychotic use were found for schizophrenia, major depressive disorder, dementia, and bipolar disorder. With respect to minor psychiatric disorder, higher proportions of antipsychotic use were found for anxiety state, neurotic depression, and special symptoms or syndromes not elsewhere. More than 60% of subjects with antipsychotic use, however, had no psychiatric disorder.
Table 3. Proportion of psychiatric disorder among subjects with antipsychotic use
|Without psychiatric disorder||3075||61.3|
|Any psychiatric disorder||1940||38.7|
| Any major psychiatric disorder||1403||28.0|
| Schizophrenic disorders (295)||664||13.2|
| Major depressive disorder (296.x)||267||5.3|
| Senile and pre-senile organic psychotic conditions (290)||214||4.3|
| Bipolar affective disorder (296.x)||200||4.0|
| Other organic psychotic conditions (294)||139||2.8|
| Other non-organic psychosis (298)||82||1.6|
| Paranoid states (297)||64||1.3|
| Any minor psychiatric disorder||923||18.4|
| Anxiety state (300.0)||311||6.2|
| Neurotic depression (300.4)||235||4.7|
| Special symptoms or syndromes not elsewhere (307)||196||3.9|
| Physiological malfunction arising from mental factors (306)||81||1.6|
| Depressive disorder, not elsewhere classified (311)||68||1.4|
| Unspecified neurotic disorder (300.9)||58||1.2|
| Any mental retardation||37||0.7|
Table 4 lists the proportion of medical disorders among subjects with antipsychotic use for non-psychiatric disorders. With respect to medical disease system, antipsychotic use for >10% of the subjects was for diseases of the digestive system (ICD-9-CM 520–579); symptoms, signs, and ill-defined conditions (ICD-9-CM 780–799); diseases of the respiratory system (ICD-9-CM 460–519); diseases of the musculoskeletal system and connective tissue (ICD-9-CM 710–739); diseases of the circulatory system (ICD-9-CM 390–459); diseases of the nervous system and sense organs (ICD-9-CM 320–389); and diseases of the genitourinary system (ICD-9-CM 580–629).
Table 4. Proportion of medical disorder among subjects with antipsychotic use
|Diseases of the digestive system (520–579)||1340||43.6|
| Gastritis and duodenitis (535)||441||14.3|
| Other non-infectious gastroenteritis and colitis (558)||441||14.3|
| Disorders of function of stomach (536)||172||5.6|
| Functional digestive disorders, not elsewhere classified (564)||170||5.5|
| Peptic ulcer, site unspecified (533)||154||5.0|
|Symptoms, signs, and ill-defined conditions (780–799)||1216||39.6|
| General symptoms (780)||686||22.3|
| Dizziness and giddiness (780.4)||539||17.5|
| Sleep disturbances (780.5)||138||4.5|
| Symptoms involving head and neck (784)||267||8.7|
| Headache (784.0)||261||8.5|
| Other symptoms involving abdomen and pelvis (789)||259||8.4|
| Abdominal pain (789.0)||256||8.3|
| Nausea and vomiting (787.0)||71||2.3|
|Diseases of the respiratory system (460–519)||914||29.8|
| Acute upper respiratory infections of multiple or unspecified site (465)||501||16.3|
| Acute bronchitis and bronchiolitis (466)||148||4.8|
| Acute nasopharyngitis (common cold) (460)||83||2.7|
|Diseases of the musculoskeletal system and connective tissue (710–739)||608||19.8|
| Other and unspecified disorders of back (724)||175||5.7|
| Backache, unspecified (724.5)||70||2.3|
| Lumbago (724.2)||68||2.2|
| Other disorders of soft tissues (729)||164||5.3|
| Myalgia and myositis, unspecified (729.1)||76||2.5|
| Neuralgia, neuritis and radiculitis, unspecified (729.2)||48||1.6|
| Osteoarthrosis and allied disorders (715)||107||3.5|
|Diseases of the circulatory system (390–459)||562||18.3|
| Essential hypertension (401)||224||7.3|
| Hypertensive heart disease (402)||104||3.4|
|Diseases of the nervous system and sense organs (320–389)||429||14.0|
| Vertiginous syndromes and other disorders of vestibular system (386)||220||7.2|
|Diseases of the genitourinary system (580–629)||374||12.2|
| Disorders of menstruation and other abnormal bleeding from female genital tract (626)||102||3.3|
To our knowledge the present study is the first to use NHI data to identify the factors associated with antipsychotic use in Taiwan. There has been no comprehensive report about the proportion of antipsychotic use with respect to psychiatric disorders and medical disorders in Taiwan.
In the present study the 1-year prevalence of antipsychotic use in Taiwan (3.9%) was higher than in Canada (0.5%),3 in the USA (0.7%),2 and in Europe (1.2%).1 As many as 96% of residents of Taiwan have joined the NHI program since 1996. The Bureau of NHI has also contracted with 91% of medical institutions in Taiwan. Thus, the high enrollment rate and accessibility of the NHI in Taiwan may make it easier for Taiwanese residents to receive antipsychotic treatment than those of other countries. In the present study the treated prevalence of major psychiatric disorders (including ICD-9-CM codes 290–298) in Taiwan was 1.90%, which was lower than the prevalence of antipsychotic use. Possibly there could be more non-psychotic treatment or off-label use of antipsychotics in Taiwan, including the treatment of non-psychotic depression, non-psychotic mania state, dementia without psychotic symptoms, aggressive behavior, severe neurotic disorder, severe insomnia, or other medical problems.17
With respect to age, the present study found that the prevalence rate of antipsychotic use increased with age, which was consistent with studies in other countries.4,22 There are several conditions that should be considered. First, older persons might have a higher percentage of major psychiatric disorder such as dementia, major depressive disorder, and organic mental disorder than younger persons. Second, older persons with dementia could suffer from cognitive problems, restlessness, disordered behavior, depression, or sleep disorder, so they had more causes for receiving antipsychotic treatment.22 When antipsychotics are prescribed to older persons, we must pay close attention to the side-effects, including extrapyramidal effects, anticholinergic effects, anti-adrenergic effects, metabolic syndrome, and sedation.23
The present study also found that women had a higher rate of antipsychotic use than did men, which was consistent with studies in other countries.4,9 As we know, there is no sex difference in the prevalence of schizophrenia. Thus, a large proportion of antipsychotic use in women was for non-psychotic conditions such as depression, anxiety, insomnia, or other off-label reasons.9 Some side-effects such as weight gain, passivity, hypotension, and hyperprolactinemia are reported to be particularly problematic for women.24 Thus, the sex difference in antipsychotic prescriptions is noteworthy.
There was no racial difference in antipsychotic use in the present study. Several reports found that African–American people were less likely than white people to receive second-generation antipsychotics.12,14 With regard to SES, persons with a lower SES had a higher proportion of antipsychotic use than those for higher SES. This suggests that persons with a lower SES had a higher prevalence of major psychiatric disorder. In an Italian national survey, second-generation antipsychotics were less likely to be prescribed to patients having a lower educational level or living in rural areas.13 These different antipsychotic prescription patterns may reflect the disparities in SES, such as race, educational level, and urbanicity, which result from financial barriers to health-care services.11–14 Thus, we should further investigate the race and SES associated with second-generation antipsychotic use.
In considering disability, persons with disabilities such as intellectual or chronic mental disability, severe dementia, and other medical disabilities, have a higher percentage of antipsychotic drug use for control of psychotic symptoms, behavior problems, and emotional disturbances.25 The present study produced similar results. There are still no definite data for comparison of treatment effects of first-generation antipsychotics and second-generation antipsychotics for persons with disability.
In the present study only 38.7% of the recipients with antipsychotic prescriptions were being treated for psychiatric disorders. More than 60% of the recipients received antipsychotic prescriptions for non-psychiatric disorders. A study on antipsychotic agent prescriptions in Italian inpatients found that 50% of second-generation antipsychotic use and <15% of first-generation antipsychotic use were for off-label indications.16 Among the recipients with major psychiatric disorders, the statistics show that antipsychotics were primarily prescribed for schizophrenia, major depression, dementia, bipolar disorder, and organic psychotic disorder. In general, persons with schizophrenia used a larger proportion of antipsychotics compared to persons with other major psychiatric disorders. Patients with a mood disorder, either major depression or bipolar disorder, combined with psychotic symptoms such as delusions or hallucinations should receive antipsychotic drug treatment. Additionally, patients with dementia may also have co-occurring psychotic symptoms, behavior problems, or other psychological problems that need treatment with antipsychotics.17,22,26
Among individuals with minor psychiatric disorder, the statistics show that antipsychotics were primarily prescribed for anxiety state, neurotic depression, and sleep disorder. Thus, they could be used for anxiolytic, sedative, hypnotic, antidepressant, and drive-inhibiting effects.22 Usually, antipsychotics are used as combined therapy with benzodiazepine for the treatment of severe sleep disorder.17 Antipsychotics are also sometime used as adjuvant therapy with benzodiazepine for the treatment of severe anxiety or with an antidepressant for the treatment of refractory depression. A pharmacoepidemiological database covering >80% of prescriptions of antipsychotics in Germany found that up to 63% were used for neurotic disorders, sleep disorders, or dementia.22 Thus to a considerable degree, the sedative, aggression-inhibiting, or anxiolytic effects rather than the antipsychotic effects of antipsychotics played an important role when managing the related symptoms clinically.22
In contrast, antipsychotics are also used for mental retardation. Persons with mental retardation may experience combined behavior problems and poor impulse control, which could benefit from low-dose antipsychotic treatment.25 Moreover, for patients using antipsychotics for mental retardation with psychotic symptoms, the relationship between mental retardation and schizophrenia should be considered.
With respect to medical disorder, the highest proportion of antipsychotic use was for diseases of the digestive system, including gastritis and duodenitis, other non-infectious gastroenteritis and colitis, disorders of stomach function, functional digestive disorders, not elsewhere classified, and peptic ulcer, site unspecified. Several kinds of antipsychotics (such as haloperidol, chlorpromazine, and fluphenazine) with anti-emetic effects were indicated in the use of nausea and vomiting symptoms in Taiwan. Sulpiride was especially indicated for the treatment of peptic ulcer. Another notable use of antipsychotics was for the symptoms, signs, and ill-defined conditions, including general symptoms (such as dizziness and giddiness), symptoms involving head and neck (such as headache), and other symptoms involving abdomen and pelvis (such as abdominal pain). Antipsychotics have also been effectively used in headache treatment in several reports. Antipsychotics given i.v. or i.m. were found to be more effective than agents such as acetaminophen and ibuprofen.27 In addition, the phenothiazines such as perphenazine and fluphenazine inhibit serotonin uptake and block dopamine receptors, making them effective analgesics in chronic pain.28
Antipsychotics were also used for diseases of the respiratory system, including acute upper respiratory infections of multiple or unspecified site, acute bronchitis and bronchiolitis, and acute nasopharyngitis (common cold). Perhaps the antipsychotics were prescribed as an alternative to benzodiazepine for persons with respiratory diseases; antipsychotics should cause less respiratory depression than benzodiazepine. Another use of antipsychotics was for the diseases of the musculoskeletal system and connective tissue, including other and unspecified disorders of back (such as backache and lumbago), other disorders of soft tissues (such as myalgia and neuralgia), and osteoarthrosis and allied disorders. Patients with spinal disease, back pain, myalgia, neuralgia, and osteoarthrosis can be treated with antipsychotic therapy combined with analgesic for the treatment of musculoskeletal diseases.28 To summarize, in the present study antipsychotics were found to be used to relieve pain such as headache, backache, myalgia, or neuralgia, and further investigation should be performed with regard to the dosage and frequency of antipsychotics.
Antipsychotics were also used for diseases of the circulatory system, including essential hypertension and hypertensive heart disease. Persons with hypertension may have a higher percentage of psychosomatic disorders, so some physicians prescribe antipsychotics for sedative effect. Nevertheless, individuals with hypertensive heart disease who are taking antipsychotics must watch out for the anti-adrenergic side-effects. Antipsychotic use was also observed for diseases of the nervous system and sense organs, including vertiginous syndromes and other disorders of the vestibular system. Several kinds of antipsychotics are used to treat vertigo and vomiting symptoms of vestibular disease. Finally, antipsychotics were also used for diseases of the genitourinary system.
The current study was performed using a random, population-based NHI sample for antipsychotic utilization analysis. Compared to previous studies, we investigated not only antipsychotic use for the treatment of psychiatric disorders, but also for the treatment of medical disorders. There were still some limitations, however: (i) off-label use of antipsychotics in persons with medical disorders is still not clear; (ii) different study designs, study subjects, and sampling methods limit comparisons with other antipsychotic use studies; (iii) we did not compare the differences between first-generation antipsychotics and second-generation antipsychotics; and (iv) we adopted stricter criteria for further investigation of disease patterns of antipsychotic use in Taiwan. We should also note the limitation that we excluded persons with only one antipsychotic prescription.