Application of a data-mining technique to analyze coprescription patterns for antacids in Taiwan

Improved diagnosis-medication association mining to reduce pseudo-associations

BACKGROUND AND OBJECTIVE

Association rule mining has been adopted to medical fields to discover prescribing patterns or relationships among diseases and/or medications; however, it has generated unreasonable associations among these entities. This study aims to identify the real-world profile of disease-medication (DM) associations using the modified mining algorithm and assess its performance in reducing DM pseudo-associations.

METHODS

We retrieved data from outpatient records between January 2011 and December 2015 in claims databases maintained by the Health and Welfare Data Science Center, Ministry of Health and Welfare, Taiwan. The association rule mining's lift (Q-value) was adopted to quantify DM associations, referred to as Q₁ for the original algorithm and as Q₂ for the modified algorithm. One thousand DM pairs with positive Q₁-values (Q₁⁺) and negative or no Q₂-values (Q₂^- or Q₂^∅) were selected as the validation dataset, in which two pharmacists assessed the DM associations.

RESULTS

A total of 3,120,449 unique DM pairs were identified, of which there were 333,347 Q₁⁺Q₂^- pairs and 429,931 Q₁⁺Q₂^∅ pairs. Q₁⁺Q₂^- rates were relatively high in ATC classes C (29.91%) and R (30.24%). Classes L (69.91%) and V (52.52%) demonstrated remarkably high Q₁⁺Q₂^∅ rates. For the 1000 pairs in the validation, 93.7% of the Q₁⁺Q₂^- or Q₁⁺Q₂^∅ DM pairs were assessed as pseudo-associations. However, classes M (5.3%), H (4.5%), and B (4.1%) showed the highest rates of plausible associations falsely given Q₂^- or Q₂^∅ by the modified algorithm.

CONCLUSIONS

The modified algorithm demonstrated high accuracy to identify pseudo-associations regarded as positive associations by the original algorithm and would potentially be applied to improve secondary databases to facilitate research on real-world prescribing patterns and further enhance drug safety.

Prevalence and Association of Pain Experiences, Medication Literacy, and Use of Medication among Children and Adolescents in Taiwan

OBJECTIVE

This study assessed the prevalence of children's and adolescents' pain experiences and use of medicine and examined the relationships between pain experiences, medication knowledge, literacy, and use of medicine.

METHOD

A probability-proportionate-to-size sampling method was used to systematically draw a random sample of schools. In 2014, a national representative sample of 2309 students from 35 primary schools (5th-6th grade), 2700 students from 30 middle schools, and 2013 students from 20 high schools completed the online survey.

RESULTS

Overall, 85.6% of children and adolescents reported experiencing pain during the past year that included headache (63.0%), throat ache (59.3%), muscle ache (58.3%), stomach pain (42.9%), menstrual pain (girls: 42.1%), and dental pain (38.5%). Children and adolescents had taken cold/cough medicine (48.1%), acetaminophen (15.0%), antacids (14.8%), and nonsteroidal anti-inflammatory drugs (NSAIDs) (10.5%) in the past year. Multivariate analysis results indicated that after controlling for pain experiences children and adolescents who had lower levels of medication knowledge and literacy were more likely to use pain medication and antacids more frequently. In addition, children and adolescents who had lower medication knowledge, lower literacy, asked doctors to prescribe antacids, and co-administered with antacids were more likely to report long-term use of antacids.

CONCLUSIONS

Lower levels of medication knowledge and literacy among children and adolescents were associated with more frequent use of pain medication and antacids.

Policy-driven revolution of prescription record in outpatient use of fluoroquinolones

OBJECTIVE

A policy initiated in 2001 by Taiwan's National Health Insurance (NHI) Administration has effectively reduced outpatient antibiotic use except fluoroquinolones (FQs). The influence of differential regulation policy of narrow-spectrum versus broad-spectrum FQs on the prescriptions is unknown.

METHODS

This study analyzed the claim records of oral FQs prescription at outpatient visits during 2000-2010 using the NHI Research Database and compared prescriptions for narrow-spectrum FQs, which are inactive against Streptococcus pneumoniae and lack formulary restriction, with those for broad-spectrum FQs.

RESULTS

Oral antibiotics were prescribed in 13.3% of visits and FQs accounted for 2.2% of them. During the study period the population-based rates of FQ prescription visits to children decreased, which was offset by increased use in the adult and geriatric populations (all p < 0.001). The most common encoded diagnoses for all FQs were urinary tract infection (19.2%) and sinusitis (10.9%), skin/bone/joint infections (7.9%), and lower respiratory tract infections (LRTI, 4.8%). Narrow-spectrum FQs accounted for 88.4% of all FQ prescriptions. Up to 95.4% of visits from patients with sinusitis and 34.3% of those with LRTI used narrow-spectrum FQs, while S. pneumoniae is an important etiology. Otorhinolaryngologists in non-hospital-based clinics prescribed most of narrow-spectrum FQs to patients with sinusitis or LRTI.

CONCLUSIONS

We found debatable prescription of narrow-spectrum FQ based on claim records, particularly for LRTI and sinusitis, possibly due to the lack of formulary restriction. Additional efforts are needed to improve the appropriate selection of optimal FQs.

Impact of grouping complications on mortality in traumatic brain injury: A nationwide population-based study

Traumatic brain injury (TBI) is an important health issue with high mortality. Various complications of physiological and cognitive impairment may result in disability or death after TBI. Grouping of these complications could be treated as integrated post-TBI syndromes. To improve risk estimation, grouping TBI complications should be investigated, to better predict TBI mortality. This study aimed to estimate mortality risk based on grouping of complications among TBI patients. Taiwan's National Health Insurance Research Database was used in this study. TBI was defined according to the International Classification of Diseases, Ninth Revision, Clinical Modification codes: 801-804 and 850-854. The association rule data mining method was used to analyze coexisting complications after TBI. The mortality risk of post-TBI complication sets with the potential risk factors was estimated using Cox regression. A total 139,254 TBI patients were enrolled in this study. Intracerebral hemorrhage was the most common complication among TBI patients. After frequent item set mining, the most common post-TBI grouping of complications comprised pneumonia caused by acute respiratory failure (ARF) and urinary tract infection, with mortality risk 1.55 (95% C.I.: 1.51-1.60), compared with those without the selected combinations. TBI patients with the combined combinations have high mortality risk, especially those aged <20 years with septicemia, pneumonia, and ARF (HR: 4.95, 95% C.I.: 3.55-6.88). We used post-TBI complication sets to estimate mortality risk among TBI patients. According to the combinations determined by mining, especially the combination of septicemia with pneumonia and ARF, TBI patients have a 1.73-fold increased mortality risk, after controlling for potential demographic and clinical confounders. TBI patients aged<20 years with each combination of complications also have increased mortality risk. These results could provide physicians and caregivers with important information to increase their awareness about sequences of clinical syndromes among TBI patients, to prevent possible deaths among these patients.

Associations between socio-demographic characteristics and chemical concentrations contributing to cumulative exposures in the United States

Association rule mining (ARM) has been widely used to identify associations between various entities in many fields. Although some studies have utilized it to analyze the relationship between chemicals and human health effects, fewer have used this technique to identify and quantify associations between environmental and social stressors. Socio-demographic variables were generated based on U.S. Census tract-level income, race/ethnicity population percentage, education level, and age information from the 2010-2014, 5-Year Summary files in the American Community Survey (ACS) database, and chemical variables were generated by utilizing the 2011 National-Scale Air Toxics Assessment (NATA) census tract-level air pollutant exposure concentration data. Six mobile- and industrial-source pollutants were chosen for analysis, including acetaldehyde, benzene, cyanide, particulate matter components of diesel engine emissions (namely, diesel PM), toluene, and 1,3-butadiene. ARM was then applied to quantify and visualize the associations between the chemical and socio-demographic variables. Census tracts with a high percentage of racial/ethnic minorities and populations with low income tended to have higher estimated chemical exposure concentrations (fourth quartile), especially for diesel PM, 1,3-butadiene, and toluene. In contrast, census tracts with an average population age of 40-50 years, a low percentage of racial/ethnic minorities, and moderate-income levels were more likely to have lower estimated chemical exposure concentrations (first quartile). Unsupervised data mining methods can be used to evaluate potential associations between environmental inequalities and social disparities, while providing support in public health decision-making contexts.

Predicting biomedical metadata in CEDAR: A study of Gene Expression Omnibus (GEO)

A crucial and limiting factor in data reuse is the lack of accurate, structured, and complete descriptions of data, known as metadata. Towards improving the quantity and quality of metadata, we propose a novel metadata prediction framework to learn associations from existing metadata that can be used to predict metadata values. We evaluate our framework in the context of experimental metadata from the Gene Expression Omnibus (GEO). We applied four rule mining algorithms to the most common structured metadata elements (sample type, molecular type, platform, label type and organism) from over 1.3million GEO records. We examined the quality of well supported rules from each algorithm and visualized the dependencies among metadata elements. Finally, we evaluated the performance of the algorithms in terms of accuracy, precision, recall, and F-measure. We found that PART is the best algorithm outperforming Apriori, Predictive Apriori, and Decision Table. All algorithms perform significantly better in predicting class values than the majority vote classifier. We found that the performance of the algorithms is related to the dimensionality of the GEO elements. The average performance of all algorithm increases due of the decreasing of dimensionality of the unique values of these elements (2697 platforms, 537 organisms, 454 labels, 9 molecules, and 5 types). Our work suggests that experimental metadata such as present in GEO can be accurately predicted using rule mining algorithms. Our work has implications for both prospective and retrospective augmentation of metadata quality, which are geared towards making data easier to find and reuse.

Evaluation of rational nonsteroidal anti-inflammatory drugs and gastro-protective agents use; association rule data mining using outpatient prescription patterns

Background

Nonsteroidal anti-inflammatory drugs (NSAIDs) and gastro-protective agents should be co-prescribed following a standard clinical practice guideline; however, adherence to this guideline in routine practice is unknown. This study applied an association rule model (ARM) to estimate rational NSAIDs and gastro-protective agents use in an outpatient prescriptions dataset.

Methods

A database of hospital outpatients from October 1st, 2013 to September 30th, 2015 was searched for any of following drugs: oral antacids (A02A), peptic ulcer and gastro-oesophageal reflux disease drugs (GORD, A02B), and anti-inflammatory and anti-rheumatic products, non-steroids or NSAIDs (M01A). Data including patient demographics, diagnoses, and drug utilization were also retrieved. An association rule model was used to analyze co-prescription of the same drug class (i.e., prescriptions within A02A-A02B, M01A) and between drug classes (A02A-A02B & M01A) using the Apriori algorithm in R. The lift value, was calculated by a ratio of confidence to expected confidence, which gave information about the association between drugs in the prescription.

Results

We identified a total of 404,273 patients with 2,575,331 outpatient visits in 2 fiscal years. Mean age was 48 years and 34% were male. Among A02A, A02B and M01A drug classes, 12 rules of associations were discovered with support and confidence thresholds of 1% and 50%. The highest lift was between Omeprazole and Ranitidine (340 visits); about one-third of these visits (118) were prescriptions to non-GORD patients, contrary to guidelines. Another finding was the concomitant use of COX-2 inhibitors (Etoricoxib or Celecoxib) and PPIs. 35.6% of these were for patients aged less than 60 years with no GI complication and no Aspirin, inconsistent with guidelines.

Conclusions

Around one-third of occasions where these medications were co-prescribed were inconsistent with guidelines. With the rapid growth of health datasets, data mining methods may help assess quality of care and concordance with guidelines and best evidence.

Comorbidity study of borderline personality disorder: applying association rule mining to the Taiwan national health insurance research database

Background

Borderline personality disorder (BPD) is a complex clinical state with highly polymorphic symptoms and signs. Studies have demonstrated that people with a BPD diagnosis are likely to have numerous co-occurring psychiatric disorders and physical comorbidities. The aim of our study was to obtain further insight about the associations among comorbidities of BPD and to demonstrate the practicality of using association rule mining (ARM) technique in clinical databases.

Methods

A retrospective case–control study was conducted on information of 1460 patients (292 BPD patients and 1168 control patients) selected from the Taiwan National Health Insurance Research Database. Information on physical and psychiatric comorbidities, which were diagnosed within 3 years before and after enrollment, was collected. A logistic regression model was used to calculate the odds ratios of comorbidities between patients with and without BPD. ARM technique was used to study the associations of BPD and two or more psychiatric comorbidities.

Results

We classified physical comorbidities into 13 categories according to the International Classification of Diseases, Ninth Revision, Clinical Modification system, and the results indicated that the 12 categories were more common in the BPD patients than in the control patients (except congenital anomalies). However, psychiatric comorbidities, including depressive disorder, bipolar disorder, anxiety disorder, sleep disorder, substance use disorder, and mental retardation were more common in the BPD patients than in the control patients. Furthermore, the associations of BPD and two or more comorbidities were evaluated.

Conclusion

Most physical and psychiatric disorders were more common in the BPD patients than in the control patients. Because the failure to remit from BPD is associated with suffering from chronic physical conditions and because psychiatric comorbidities may lead to delays in diagnosis of BPD, clinicians caring for people with BPD should be aware of possible comorbidities.

Doctor-Shopping Behaviors among Traditional Chinese Medicine Users in Taiwan

Doctor-shopping has caused an increase in medical expense, potential to receive duplicate medications, and suffer adverse drug reactions. We carried out a population-based retrospective study aimed at examining the user patterns of traditional Chinese medicine (TCM) ambulatory care in Taiwan. We retrieved complete TCM ambulatory visit datasets for the year 2007 from the National Health Insurance database in Taiwan. We defined the patients whose distribution of TCM physician numbers scored more than 97.5 percent (more than, or equal to, five TCM physicians) within one year as TCM doctor-shoppers. In total, 6,596,814 subjects (28.9%) paid TCM visits during that year. All 177,728 subjects (2.69%) who visited more than five (including) TCM physicians were classified as TCM shoppers. The most prevalent diagnostic grouping was upper respiratory infections (44.7%) and sprains and strains (44.0%). Men had a lower odds ratio (OR) among TCM shoppers than women (OR = 0.94, 95% confidence interval (CI) = 0.93-0.96). Younger people were less likely to be TCM shoppers than other people were. The ORs of TCM shoppers were higher among veterans and low-income patients (OR = 1.29 (1.23-1.35), and 1.33 (1.27-1.41)). In conclusion, health education on the potential of drug interactions and iatrogenic health risks incurred from doctor-shopping should be addressed to those high-risk patients.

Characteristics of ambulatory care visits to family medicine specialists in Taiwan: a nationwide analysis

Although family medicine (FM) is the most commonly practiced specialty among all the medical specialties, its practice patterns have seldom been analyzed. Looking at data from Taiwan's National Health Insurance Research Database, the current study analyzed ambulatory visits to FM specialists nationwide. From a sample dataset that randomly sampled one out of every 500 cases among a total of 309,880,000 visits in 2012, it was found that 18.8% (n = 116, 551) of the 619,760 visits in the dataset were made to FM specialists. Most of the FM services were performed by male FM physicians. Elderly patients above 80 years of age accounted for only 7.1% of FM visits. The most frequent diagnoses (22.8%) were associated acute upper respiratory infections (including ICD 460, 465 and 466). Anti-histamine agents were prescribed in 25.6% of FM visits. Hypertension, diabetes and dyslipidemia were the causes of 20.7% of the ambulatory visits made to FM specialists of all types, while those conditions accounted for only 10.6% of visits to FM clinics. The study demonstrated the relatively low proportion of chronic diseases that was managed in FM clinics in Taiwan, and our detailed results could contribute to evidence-based discussions on healthcare policymaking and residency training.

Frequency and co-prescription pattern of Chinese herbal products for hypertension in Taiwan: a Cohort study

BACKGROUND

Chinese herbal products (CHPs) have been frequently used among patients with chronic diseases including hypertension; however, the co-prescription pattern of herbal formulae and single herbs remain uncharacterized. Thus, this large-scale pharmacoepidemiological study evaluated the frequency and co-prescription pattern of CHPs for treating hypertension in Taiwan from 2003 to 2009.

METHODS

The database of traditional Chinese medicine (TCM) outpatient claims was obtained from the National Health Insurance in Taiwan. Patients with hypertension during study period were defined according to diagnostic codes in the International Classification of Disease Ninth Revision, Clinical Modification. The frequencies and percentages of herbal formula and single herb prescriptions for hypertension were analyzed. We also applied association rules to evaluate the CHPs co-prescription patterns.

RESULTS

The hypertension cohort included 154,083 patients, 123,240 patients of which (approximately 80 %) had used TCM at least once. In total, 81,582 visits involving CHP prescriptions were hypertension related; Tian-Ma-Gou-Teng-Yin and Dan Shen (Radix Salvia Miltiorrhizae) were the most frequently prescribed herbal formula and single herb, respectively, for treating hypertension.

CONCLUSIONS

This study elucidated the utilization pattern of CHPs for treating hypertension. Future studies on the efficacy and safety of these CHPs and on drug-herb interactions are warranted.

Identifying chinese herbal medicine network for eczema: implications from a nationwide prescription database

Eczema is a highly prevalent dermatological disease that can severely affect the patient's quality of life. Chinese herbal medicine (CHM) is commonly used in combination for eczema due to the complicated pathogenesis. This study aimed to identify a CHM network for the treatment of eczema by using a nationwide database. During 2011, 381,282 CHM prescriptions made for eczema (ICD-9-CM 692.x) were obtained from the National Health Insurance Research Database (NHIRD) in Taiwan and analyzed by using association rule mining and social network analysis. Among 661 available CHMs, 44 important combinations were identified. Among the CHM networks, seven clusters with the predominant traditional Chinese medicine (TCM) pattern were recognized. The largest CHM cluster was used to treat the wind-dampness-heat pattern, and Xiao-Feng-San (24.1% of all prescriptions) was the core of this cluster with anti-inflammation, antioxidation, and antiallergic effects. Lonicera japonica (11.0% of all prescriptions) with Forsythia suspense (17.0% of all prescriptions) was the most commonly used CHM combination and was also the core treatment for treating the heat pattern, in which an antimicrobial effect is found. CHM network analysis is helpful for TCM doctors or researchers to choose candidates for clinical practice or further studies.

Investigation on Chinese herbal medicine for primary dysmenorrhea: implication from a nationwide prescription database in Taiwan

OBJECTIVE

Primary dysmenorrhea is a common gynecological condition, for which Chinese herbal medicine (CHM) has been widely used in addition to western medicine. The aim of this study is to explore CHM commonly used to treat dysmenorrhea in young Chinese women.

DESIGN

Observational retrospective study.

SETTING

The National Health Insurance Research Database in Taiwan.

POPULATION

Women aged from 13 to 25 years with single diagnosis of primary dysmenorrhea.

METHODS

CHM prescriptions made for primary dysmenorrhea women during 1998-2008 were extracted to build up CHM prescription database. Association rule mining was used to explore the prevalent CHM combination patterns in treating primary dysmenorrhea.

MAIN OUTCOME MEASURES

Prevalence and mechanisms of CHM combinations.

RESULTS

Totally 57,315 prescriptions were analyzed and, on average, 5.3 CHM was used in one prescription. Dang-Gui-Shao-Yao-San (DGSYS) was the most commonly used herbal formula (27.2%), followed by Jia-Wei-Xiao-Yao-San (JWXYS) (20.7%) and Wen-Jing-Tang (WJT) (20.5%). Corydalis yanhusuo and Cyperus rotundus were the most commonly used single herb, found in 33.1% and 29.2% of all prescriptions. Additionally, C. yanhusuo with C. rotundus is the most commonly used two CHM in combination, accounting for 14.24% of all prescriptions, followed by DGSYS with C. yanhusuo (10.47%). Multi-target effects on primary dysmenorrhea, such as analgesia, mood modifying and hormone adjustment, were found among commonly prescribed CHM in this study.

CONCLUSIONS

This study discovered the potential importance of C. yanhusuo, C. rotundus and DGSYS in treating primary dysmenorrhea. Further clinical trials or bench studies are warranted based on the results.

Developing a data mining approach to investigate association between physician prescription and patient outcome - a study on re-hospitalization in Stevens-Johnson Syndrome

Stevens-Johnson syndrome (SJS) is a potentially life-threatening skin reaction. Drugs are the major causes for cases of SJS. While treating patients with SJS, the first and most important step is to identify and discontinue any possible responsible drugs. However, potential drugs that may lead to SJS are many and encompass various therapeutic areas. Very few physicians are familiar with the potential risk of all these drugs. If properly treated, most SJS cases are expected to recover without much sequelae. All drugs that have been associated with SJS should be avoided in these patients to prevent recurrence. If the physicians fail to identify and discontinue the drugs causing SJS, or even adding new drugs related to SJS, the patient may get worse or SJS may recur. These conditions can cause SJS patients to be re-hospitalized. Currently the reasons for re-hospitalization of SJS patients in Taiwan are not known. This study uses Taiwan National Health Insurance Research Database to analyze the causes of re-hospitalization for cases of SJS. First, we classified prescription history of re-hospitalized patients through the rule-based classification method. Secondly, by using the basic prescription actions, we identified drug association patterns. Then, by employing A-priori algorithm, pairs of drugs with relatively higher frequency of appearance were identified and their degrees of association were measured by using selected symmetric and asymmetric association mining methods. Finally, by listing and ranking up these pairs of drugs according to the value of support based on their degrees of association, we provide prescribing physicians with possible means of increasing the awareness and reducing re-hospitalization of SJS patients.

Frequency and pattern of Chinese herbal medicine prescriptions for urticaria in Taiwan during 2009: analysis of the national health insurance database

Background

Large-scale pharmaco-epidemiological studies of Chinese herbal medicine (CHM) for treatment of urticaria are few, even though clinical trials showed some CHM are effective. The purpose of this study was to explore the frequencies and patterns of CHM prescriptions for urticaria by analysing the population-based CHM database in Taiwan.

Methods

This study was linked to and processed through the complete traditional CHM database of the National Health Insurance Research Database in Taiwan during 2009. We calculated the frequencies and patterns of CHM prescriptions used for treatment of urticaria, of which the diagnosis was defined as the single ICD-9 Code of 708. Frequent itemset mining, as applied to data mining, was used to analyse co-prescription of CHM for patients with urticaria.

Results

There were 37,386 subjects who visited traditional Chinese Medicine clinics for urticaria in Taiwan during 2009 and received a total of 95,765 CHM prescriptions. Subjects between 18 and 35 years of age comprised the largest number of those treated (32.76%). In addition, women used CHM for urticaria more frequently than men (female:male = 1.94:1). There was an average of 5.54 items prescribed in the form of either individual Chinese herbs or a formula in a single CHM prescription for urticaria. Bai-Xian-Pi (Dictamnus dasycarpus Turcz) was the most commonly prescribed single Chinese herb while Xiao-Feng San was the most commonly prescribed Chinese herbal formula. The most commonly prescribed CHM drug combination was Xiao-Feng San plus Bai-Xian-Pi while the most commonly prescribed triple drug combination was Xiao-Feng San, Bai-Xian-Pi, and Di-Fu Zi (Kochia scoparia).

Conclusions

In view of the popularity of CHM such as Xiao-Feng San prescribed for the wind-heat pattern of urticaria in this study, a large-scale, randomized clinical trial is warranted to research their efficacy and safety.

A study on PubMed search tag usage pattern: association rule mining of a full-day PubMed query log

Background

The practice of evidence-based medicine requires efficient biomedical literature search such as PubMed/MEDLINE. Retrieval performance relies highly on the efficient use of search field tags. The purpose of this study was to analyze PubMed log data in order to understand the usage pattern of search tags by the end user in PubMed/MEDLINE search.

Methods

A PubMed query log file was obtained from the National Library of Medicine containing anonymous user identification, timestamp, and query text. Inconsistent records were removed from the dataset and the search tags were extracted from the query texts. A total of 2,917,159 queries were selected for this study issued by a total of 613,061 users. The analysis of frequent co-occurrences and usage patterns of the search tags was conducted using an association mining algorithm.

Results

The percentage of search tag usage was low (11.38% of the total queries) and only 2.95% of queries contained two or more tags. Three out of four users used no search tag and about two-third of them issued less than four queries. Among the queries containing at least one tagged search term, the average number of search tags was almost half of the number of total search terms. Navigational search tags are more frequently used than informational search tags. While no strong association was observed between informational and navigational tags, six (out of 19) informational tags and six (out of 29) navigational tags showed strong associations in PubMed searches.

Conclusions

The low percentage of search tag usage implies that PubMed/MEDLINE users do not utilize the features of PubMed/MEDLINE widely or they are not aware of such features or solely depend on the high recall focused query translation by the PubMed’s Automatic Term Mapping. The users need further education and interactive search application for effective use of the search tags in order to fulfill their biomedical information needs from PubMed/MEDLINE.

The exploration of disease pattern, zheng, for differentiation of allergic rhinitis in traditional chinese medicine practice

Pattern, or "zheng," differentiation is the essential guide to treatment with traditional Chinese medicine (TCM). However, the considerable variability between TCM patterns complicates evaluations of TCM treatment effectiveness. The aim of this study was to explore and characterize the relationship between patterns and the core patterns of allergic rhinitis. We summarized 23 clinical trials of allergic rhinitis with mention of pattern differentiation; association rule mining was used to analyze TCM patterns of allergic rhinitis. A total of 205 allergic rhinitis patients seen at Chang Gung Memorial Hospital from March to June 2005 were included for comparison. Among the 23 clinical trials evaluated, lung qi deficiency and spleen qi deficiencies were the core patterns of allergic rhinitis, accounting for 29.50% and 28.98% of all patterns, respectively. A higher prevalence of lung or spleen qi deficiency (93.7%) was found in Taiwan. Additionally, patients with lung or spleen qi deficiency were younger (27.99 ± 12.94 versus 58.54 ± 12.96 years) and the severity of nasal stuffiness was higher than among patients with kidney qi deficiency (1.35 ± 0.89 versus 0.62 ± 0.65; P < 0.05). Lung and spleen qi deficiencies are the core patterns of allergic rhinitis and determining the severity of nasal stuffiness is helpful in differentiating the TCM patterns.

Prescription patterns of Chinese herbal products for menopausal syndrome: analysis of a nationwide prescription database

ETHNO-PHARMACOLOGICAL RELEVANCE: Chinese herbal products (CHP) are commonly used for menopausal syndrome in traditional Chinese medicine (TCM).

AIM OF THE STUDY

Because menopausal syndrome is a common issue among women worldwide, this study analyzed the prescription patterns and frequencies of CHP in relieving menopausal syndrome in Taiwan.

MATERIALS AND METHODS

Prescriptions of CHP for menopausal syndrome were obtained from the nationwide prescription database (National Health Insurance Research Database) of Taiwan. Every prescription with a leading diagnosis of menopausal syndrome made in the year 2002 for women out-patients aged 45-55 years was analyzed. Descriptive statistics and data mining method (association rule mining) were applied to the pattern of co-prescription.

RESULTS

A total of 54456 prescriptions were made for 15486 subjects in clinical visits. Salvia miltiorrhiza and Dan-zhi-xiao-yao-san were the most commonly prescribed single herb (SH) and herbal formula (HF), respectively. For combinations of two CHP, Salvia miltiorrhiza with Dan-zhi-xiao-yao-san, Zhi-bai-di-huang-wan with Dan-zhi-xiao-yao-san, and Ligustrum lucidum with Eclipta prostrata were the most frequently co-prescribed CHP couplets as SH-HF, HF-HF, and SH-SH, respectively. The most commonly used combination of three CHP was Eclipta prostrata with Ligustrum lucidum and Dan-zhi-xiao-yao-san. The core formula for menopausal syndrome was Dan-zhi-xiao-yao-san.

CONCLUSIONS

Combinations of CHP are widely used for menopausal syndrome. Data mining analysis demonstrates the core formula and the commonly combined CHP in prescriptions. Further studies are needed to evaluate their efficacy for menopausal syndrome.

Application of frequent itemsets mining to analyze patterns of one-stop visits in Taiwan

Background

The free choice of health care facilities without limitations on frequency of visits within the National Health Insurance in Taiwan gives rise to not only a high number of annual ambulatory visits per capita but also a unique “one-stop shopping”phenomenon, which refers to a patient' visits to several specialties of the same healthcare facility in one day. The visits to multiple physicians would increase the potential risk of polypharmacy. The aim of this study was to analyze the frequency and patterns of one-stop visits in Taiwan.

Methodology/Principal Findings

The claims datasets of 1 million nationally representative people within Taiwan's National Health Insurance in 2005 were used to calculate the number of patients with one-stop visits. The frequent itemsets mining was applied to compute the combination patterns of specialties in the one-stop visits. Among the total 13,682,469 ambulatory care visits in 2005, one-stop visits occurred 144,132 times and involved 296,822 visits (2.2% of all visits) by 66,294 (6.6%) persons. People tended to have this behavior with age and the percentage reached 27.5% (5,662 in 20,579) in the age group ≥80 years. In general, women were more likely to have one-stop visits than men (7.2% vs. 6.0%). Internal medicine plus ophthalmology was the most frequent combination with a visited frequency of 3,552 times (2.5%), followed by cardiology plus neurology with 3,183 times (2.2%). The most frequent three-specialty combination, cardiology plus neurology and gastroenterology, occurred only 111 times.

Conclusions/Significance

Without the novel computational technique, it would be hardly possible to analyze the extremely diverse combination patterns of specialties in one-stop visits. The results of the study could provide useful information either for the hospital manager to set up integrated services or for the policymaker to rebuild the health care system.

Data mining in healthcare and biomedicine: a survey of the literature

As a new concept that emerged in the middle of 1990's, data mining can help researchers gain both novel and deep insights and can facilitate unprecedented understanding of large biomedical datasets. Data mining can uncover new biomedical and healthcare knowledge for clinical and administrative decision making as well as generate scientific hypotheses from large experimental data, clinical databases, and/or biomedical literature. This review first introduces data mining in general (e.g., the background, definition, and process of data mining), discusses the major differences between statistics and data mining and then speaks to the uniqueness of data mining in the biomedical and healthcare fields. A brief summarization of various data mining algorithms used for classification, clustering, and association as well as their respective advantages and drawbacks is also presented. Suggested guidelines on how to use data mining algorithms in each area of classification, clustering, and association are offered along with three examples of how data mining has been used in the healthcare industry. Given the successful application of data mining by health related organizations that has helped to predict health insurance fraud and under-diagnosed patients, and identify and classify at-risk people in terms of health with the goal of reducing healthcare cost, we introduce how data mining technologies (in each area of classification, clustering, and association) have been used for a multitude of purposes, including research in the biomedical and healthcare fields. A discussion of the technologies available to enable the prediction of healthcare costs (including length of hospital stay), disease diagnosis and prognosis, and the discovery of hidden biomedical and healthcare patterns from related databases is offered along with a discussion of the use of data mining to discover such relationships as those between health conditions and a disease, relationships among diseases, and relationships among drugs. The article concludes with a discussion of the problems that hamper the clinical use of data mining by health professionals.

Prescriptions of Chinese Herbal Medicines for Insomnia in Taiwan during 2002

Chinese herbal medicine (CHM) has been commonly used for treating insomnia in Asian countries for centuries. The aim of this study was to conduct a large-scale pharmaco-epidemiologic study and evaluate the frequency and patterns of CHM use in treating insomnia. We obtained the traditional Chinese medicine (TCM) outpatient claims from the National Health Insurance in Taiwan for the year 2002. Patients with insomnia were identified from the diagnostic code of International Classification of Disease among claimed visiting files. Corresponding prescription files were analyzed, and an association rule was applied to evaluate the co-prescription of CHM. Results showed that there were 16 134 subjects who visited TCM clinics for insomnia in Taiwan during 2002 and received a total of 29 801 CHM prescriptions. Subjects between 40 and 49 years of age comprised the largest number of those treated (25.3%). In addition, female subjects used CHMs for insomnia more frequently than male subjects (female : male = 1.94 : 1). There was an average of 4.8 items prescribed in the form of either an individual Chinese herb or formula in a single CHM prescription for insomnia. Shou-wu-teng (Polygonum multiflorum) was the most commonly prescribed single Chinese herb, while Suan-zao-ren-tang was the most commonly prescribed Chinese herbal formula. According to the association rule, the most commonly prescribed CHM drug combination was Suan-zao-ren-tang plus Long-dan-xie-gan-tang, while the most commonly prescribed triple drug combination was Suan-zao-ren-tang, Albizia julibrissin, and P. multiflorum. Nevertheless, further clinical trials are needed to evaluate the efficacy and safety of these CHMs for treating insomnia.

Home healthcare services in Taiwan: a nationwide study among the older population

Background

Home healthcare services are important in aging societies worldwide. The present nationwide study of health insurance data examined the utilization and delivery patterns, including diagnostic indications, for home healthcare services used by seniors in Taiwan.

Methods

Patients ≥65 years of age who received home healthcare services during 2004 under the Taiwanese National Health Insurance Program were identified and reimbursement claims were analyzed. Age, gender, disease diagnoses, distribution of facilities providing home healthcare services, and patterns of professional visits, including physician and skilled nursing visits, were also explored.

Results

Among 2,104,978 beneficiaries ≥65 years of age, 19,483 (0.9%) patients received 127,753 home healthcare visits during 2004 with a mean number of 6.0 ± 4.8 visits per person. The highest prevalence of home healthcare services was in the 75-84 year age group in both sexes. Females received more home healthcare services than males in all age groups. Cerebrovascular disease was the most frequent diagnosis in these patients (50.7%). More than half of home healthcare visits and around half of the professional home visits were provided by community home nursing care institutions. The majority of the home skilled nursing services were tube replacements, including nasogastric tubes, Foley catheter, tracheostomy, nephrostomy or cystostomy tubes (95%).

Conclusions

Nine out of 1,000 older patients in Taiwan received home healthcare services during 2004, which was much lower than the rate of disabled older people in Taiwan. Females used home healthcare services more frequently than males and the majority of skilled nursing services were tube replacements. The rate of tube replacement of home healthcare patients in Taiwan deserves to be paid more attention.

Prescriptions of Chinese herbal medicine for constipation under the national health insurance in Taiwan

BACKGROUND

Constipation is a common gastrointestinal problem worldwide. The aim of this study was to determine the frequency of use and prescriptive patterns of Chinese herbal medicine (CHM) in treating constipation by analyzing the claims data of traditional Chinese medicine (TCM) from the National Health Insurance (NHI) in Taiwan.

METHODS

The computerized claims dataset of the TCM office visits and the corresponding prescription files in 2004 compiled by the NHI Research Institute in Taiwan were linked and processed. Visit files with the single diagnostic coding of constipation (ICD-9-CM code 564.0) were extracted to analyze the frequency and pattern of corresponding CHM prescriptions. The association rule was applied to analyze the co-prescription of CHM in treating constipation.

RESULTS

There were 152,564 subjects who visited TCM clinics only for constipation in Taiwan during 2004 and received a total of 387,268 CHM prescriptions. Subjects between 20 and 29 years of age comprised the largest number of those treated (25.5%). Female subjects used CHM for constipation more frequently than male subjects (female:male = 3.31:1). There was an average of 4.6 items of single Chinese herbs or formula in a single prescription for constipation. Ma-zi-renwan was the most commonly prescribed herbal formula, while Da-huang (Rheum palmatum) was the most commonly used single Chinese herb. According to the association rule, the most common prescribed pattern of 2-drug combination of CHM for treating constipation was Ban-xia-xie-xin-tang plus Ma-zi-ren-wan, while the 3-drug combination of CHM was Fang-feng-tong-sheng-san, Rheum palmatum and Ma-zi-ren-wan.

CONCLUSION

This study showed the pattern of single Chinese herbs or herbal formulae used in treating constipation in Taiwan. Further clinical trials are needed to evaluate the efficacy and safety of these CHMs in treating constipation.

Prevalence of the prescribing of potentially inappropriate medications at ambulatory care visits by elderly patients covered by the Taiwanese National Health Insurance program

BACKGROUND

The use of potentially inappropriate medications can have profound medical consequences for elderly patients and place a substantial burden on the health care system.

OBJECTIVES

This study was conducted to determine the prevalence of potentially inappropriate medication prescribing at ambulatory care visits by patients aged > or =65 years covered by the Taiwanese National Health Insurance program, to examine the characteristics of and risk factors for such prescribing, and to investigate its influence on health care resource utilization.

METHODS

Ambulatory care visits by patients aged > or =65 years in 2001-2004 were identified from the National Health Insurance claims database. The 2003 Beers criteria for drugs to be avoided in the elderly were used to identify potentially inappropriate medications prescribed at these visits. Only drugs with the potential to lead to higher-severity adverse events were included. Multivariate logistic regression was used to determine predictors of the prescribing of potentially inappropriate medications at ambulatory care visits. Independent variables in the regression model included patient characteristics (eg, sex, age), physician characteristics (sex, age, and specialty), and visit characteristics (site and prescribed drug number). The dependent variable was visits that included a prescription for a potentially inappropriate medication.

RESULTS

Overall, 176,661,994 ambulatory care visits by patients aged > or =65 years were identified in 2001-2004. Of these, 19.1% involved a prescription for a potentially inappropriate medication. Although the frequency of potentially inappropriate medication prescribing declined over the study period, 62.5% of elderly patients were exposed to such medications in 2004. The only patient characteristic associated with an increased likelihood of the prescribing of potentially inappropriate medications was female sex (male sex: odds ratio [OR] = 0.982 [95% CI, 0.980-0.983], P < 0.001). Physician characteristics associated with a greater likelihood of the prescribing of potentially inappropriate medications was male sex (OR = 1.206 [95% CI, 1.202-1.210], P < 0.001); older age (43-50 years: OR = 1.021 [95% CI, 1.018-1.025], P < 0.001; >/=51 years: OR = 1.238 [95% CI, 1.235-1.242], P < 0.001); and family medicine/general practice (OR = 1.267 [95% CI, 1.265-1.269], P < 0.001). For visit characteristics, significant associations were found with visits to a primary care clinic (OR = 1.887 [95% CI, 1.881-1.892], P < 0.001) and the number of drugs prescribed (4-6 drugs: OR = 2.701 [95% CI, 2.696-2.706], P < 0.001; > or =7 drugs: OR = 4.528 [95% CI, 4.517-4.538], P < 0.001). The most commonly prescribed types of potentially inappropriate medications were antihistamines (4.8% of all prescriptions in 48.3% of elderly patients), muscle relaxants/antispasmodics (4.0% and 40.3%, respectively), and long-acting benzodiazepines (2.4% and 21.4%). In 2004, the mean number of ambulatory care visits per patient was significantly higher among those who received potentially inappropriate medications compared with those who did not (30.78 vs 16.57, respectively; P < 0.001). Patients who received potentially inappropriate medications also had significantly more emergency department visits (0.27 vs 0.15; P < 0.001) and hospital admissions (0.46 vs 0.27; P < 0.001).

CONCLUSION

There was a high prevalence of the prescribing of potentially inappropriate medications at ambulatory care visits by elderly patients in Taiwan in 2001-2004.

Comorbidity study of ADHD: applying association rule mining (ARM) to National Health Insurance Database of Taiwan

OBJECTIVE

This paper intends to apply association rule mining (ARM) to explore the labyrinthian network of ADHD comorbidity, and to examine the practicality of ARM in comorbidity studies using clinic databases.

METHODS

From clinic records of enrollees of Taiwan National Health Insurance (NHI), 18,321 youngsters aged 18 or less with diagnosis of ADHD in 2001 were recruited as case group in this study. And all their clinic diagnoses made from 2000 to 2002, as comorbidity, were categorized according to "The International Classification of Disease, 9th Revision, Clinical Modification" (ICD-9-CM) diagnosis system. For comparison, fourfold non-ADHD controls were recruited from 2001s NHI enrollees on a random base but matched gender and age of cases. ARM was done with Apriori algorithm to examine the strengths of associations among those diagnoses. The support and confidence values of ARM results were examined. Comorbidity rates and relative risk (RR) ratios of both groups of each diagnosis were compared one another.

RESULTS

ADHD case group has apparently higher risk of comorbidity with psychiatric comorbidity than with other physical illnesses. From results of ARM, developmental delay (DD) appears as an important node between ADHD and anxiety disorder (support: 5.12%, confidence: 97.42%), mild mental retardation (support: 4.42%, confidence: 92.09%) and autism (support: 6.49%, confidence: 94.93%).

CONCLUSIONS

The finding of this study, an important role of DD between ADHD and other psychiatric comorbidity, supports neurological findings in developmental delay of ADHD children's front cortex, as well as some epidemiology findings. This study also demonstrated the practicality of ARM in comorbidity studies using enormous clinic databases like NHIRD.

Potentially inappropriate medication for emergency department visits by elderly patients in Taiwan

PURPOSE

The potential for adverse drug events caused by potentially inappropriate medication (PIM) use in elderly patients at emergency department (ED) visits is a growing concern. The objects of this study were to determine the prevalence, characteristics and risk factors of PIM use among elderly ED visits in Taiwan.

METHODS

The nationwide computerized claims database of elderly ED visits under the National Health Insurance (NHI) in Taiwan during 2001-2004 was accessed. PIM, independent of diseases diagnoses or conditions and should be generally be avoided in elderly people, was evaluated using the updated 2003 Beers criteria.

RESULTS

Between 2001 and 2004, 14.7% of total 1 429 463 elderly ED visits with prescriptions had PIM, and 19.3% of elderly people who visited ED received at least one PIM annually. Odds ratio for PIM prescriptions to ED elderly was higher for visits at which more drugs were prescribed, visits at local community hospital, female and older physicians, patients aged 65-69 years and female patients. Common PIM categories were short acting nifedipine, muscle relaxants and anti-spasmodics, antihistamines and ketorolac. When health care resource utilization was compared in 2004, subjects receiving PIM at ED visit had significantly more mean ambulatory care visits, ED visits and hospital admissions than subjects who did not receive PIM.

CONCLUSIONS

About one fifth of elderly people who visited ED received PIM annually in Taiwan. The public and physicians should be educated, and a computerized drug surveillance system might be needed to avoid PIM prescriptions to the ED elderly patients.

Frequency and pattern of Chinese herbal medicine prescriptions for chronic hepatitis in Taiwan

ETHNOPHARMACOLOGICAL RELEVANCE

Chinese herbal medicine (CHM) has been commonly used in treating liver diseases in Asian countries.

AIM OF STUDY

To conduct a large-scale pharmacoepidemiological study and evaluate the frequency and pattern of CHM prescriptions in treating chronic hepatitis.

MATERIALS AND METHODS

We obtained the database of traditional Chinese medicine outpatient claims from the national health insurance in Taiwan for the whole 2002. Patients with chronic hepatitis were identified by the corresponding diagnosis of International Classification of Disease among claimed visiting files. Corresponding prescription files were analyzed, and association rule were applied to evaluate the co-prescription of CHM in treating chronic hepatitis.

RESULTS

Among the 91,080 subjects treated by CHM for chronic hepatitis, the peak age was in the 40 s, followed by 30 s and 50 s. Male/female ratio was 2.07:1. Long-dan-xie-gan-tang and Saliva miltiorrhiza (Dan-shen) were the most commonly prescribed Chinese herbal formula and single herbal drug, respectively. The most common two-drug prescription was Jia-wei-xia-yao-san plus Saliva miltiorrhiza, and the most common three-drug prescription was Jia-wei-xia-yao-san plus Saliva miltiorrhiza and Artemisia capillaries (Yin-chen-hao).

CONCLUSIONS

This study showed the utilization pattern of Chinese herbal drugs or formulae in treating chronic hepatitis. Further researches and clinical trials are needed to evaluate the efficacy of these Chinese herbs or its ingredients in treating chronic hepatitis.

An introduction to association rule mining: an application in counseling and help-seeking behavior of adolescents

Association rule mining (ARM) is a technique used to discover relationships among a large set of variables in a data set. It has been applied to a variety of industry settings and disciplines but has, to date, not been widely used in the social sciences, especially in education, counseling, and associated disciplines. This article thus introduces ARM and presents aspects of existing work that will be relevant and useful to researchers practitioners in the social sciences. Definitions and concepts are presented, and examples of ARM applications are highlighted to strengthen these ideas. We also discuss an example from our existing research to show that ARM can be used to investigate help-seeking behavior in a sample of secondary school students in Singapore. We also present some guidelines and recommendations for using ARM.

Use frequency of traditional Chinese medicine in Taiwan

Background

Use of Traditional Chinese medicine (TCM), an important category of complementary and alternative medicine (CAM), has increased substantially in Western countries during the past decade. Use of TCM is also widespread in the Chinese population. However, few informative data have been obtained to date by large-scale investigations of TCM use in the Chinese population. This study was aimed at elucidating the demographics and patterns of TCM use in Taiwan.

Methods

We employed the complete datasets of TCM outpatient reimbursement claims from 1996 to 2001, including the use of Chinese herbal remedies, acupuncture and traumatology manipulative therapy, to analyse use frequencies, the characteristics of TCM users, and the disease categories that were treated by TCM in Taiwan.

Results

At the end of 2001, 6,142,829 (28.4%) among the 21,653,555 valid beneficiaries of the National Health Insurance in Taiwan had used TCM during the year. However, 13,536,266 subjects (62.5%) had used TCM at least once during the whole 6-year period from 1996 to 2001, with a total of 156,224,266 visits (mean 11.5 visits per user). The mean number of TCM users per annum was 5,733,602, with a mean increment of 1,671,476 (29.2%) of new users yearly. Among TCM users, female was higher than male (female:male = 1.13:1), and the age distribution displayed a peak at around the 30s, followed by the 20s and 40s. Chinese herbal remedies (85.9%) were the most common TCM modality used by this population, followed by acupuncture (11.0%) and traumatology manipulative therapies (3.1%). Private TCM clinics provided most of the TCM care (82.6%), followed by private TCM hospitals (12.0%). The top ten major disease categories for TCM visits were diseases of the respiratory system, musculoskeletal system and connective tissue; symptoms, signs and ill-defined conditions; injury and poisoning; diseases of the digestive system, genitourinary system, skin and subcutaneous tissue, nervous system and sense organs, circulatory and endocrine system; nutritional and metabolic diseases; and immunological disorders.

Conclusion

TCM was popular among the Chinese population in Taiwan during the period studied. More than 60% of all subjects had used TCM during the 6-year interval. TCM was widely used by the Chinese population to treat problems and diseases of major human organ systems recognised by western medicine. This study provides information about the use frequencies of TCM and the disease categories treated by TCM, which should be useful for health policy makers and for those considering the integration of TCM and Western medicine.

The prescriptions frequencies and patterns of Chinese herbal medicine for allergic rhinitis in Taiwan

BACKGROUND

The evaluation of Chinese herbal medicines (CHM) to allergic rhinitis (AR) by large-scale pharmaco-epidemiological study is not ease found, although CHM had been reported to have potential effect for AR in some clinical trials.

AIMS OF THE STUDY

To explore the frequency and pattern of CHM prescriptions on AR, we have the study by analysing the population-based CHM database in Taiwan.

METHODS

The way for this study was linked and processed the complete traditional Chinese medicine database for Taiwanese recorded in the year 2002. The diagnosis of AR was extracted with the only single ICD-9 Code of 477 to calculate the frequency and pattern of prescriptions. Association rule was applied to analyse co-prescription of CHM for patients with AR.

RESULTS

In the year 2002, among the 22 520 776 valid beneficiaries of the National Health Insurance, Taiwan, 914 612 subjects (3.8% of the total valid beneficiaries) have diagnosed AR. There were 35.6% of AR patients been treated by CHM. The peak age of AR patients treated by CHM was at the first decade (0-10). For the AR patients, the most common Chinese herbal formula prescription was Shin-yi-qing-fei-tang, or Angelica dahurica (Bai-zhi) for the single Chinese herb. While for the combination treatments the most common prescription was the two formulae, Xiao-qing-long-tang and Shin-yi-san.

CONCLUSIONS

Because of the high utilization rate of the CHM treatment for AR, a large-scale randomized trial warrants further research for its efficacy and safety.

Demographics and patterns of acupuncture use in the Chinese population: the Taiwan experience

BACKGROUND

The use of complementary and alternative medicine (CAM), including Traditional Chinese Medicine (TCM) and acupuncture, has substantially increased in western countries in the past decade. However, informative data concerning large-scale investigations of acupuncture used in the Chinese society remain rare so far. DESIGN AND OUTCOME MEASURE: The complete datasets of acupuncture outpatient reimbursement claims from 1996 to 2002 were supplied by the National Health Insurance Research Database, Taiwan, and the usage frequencies and characteristics of the acupuncture users, as well as the disease categories that were treated by acupuncture in Taiwan were analyzed.

RESULTS

At the end of 2002, among the 21,869,478 total valid beneficiaries of National Health Insurance (NHI), 1,362,351 subjects (6.2%) used acupuncture during this year, but 4,948,464 subjects (22.6%) had used it for the whole 7-year period since 1996. A mean increment of 1,191,164 (53.6%) new users had been involved yearly. Among all those acupuncture users, a female predominance was observed (female:male = 1.12:1), and the age distribution displayed a peak at around the 40s, followed by the 30s and 50s. Private TCM clinics provided more acupuncture usage (82.0%) than did private TCM hospitals (13.7%). The disease categories treated mostly by acupuncture were diseases of the musculoskeletal system (46.2%); injury (41.8%); diseases of the nervous system (3.5%); and symptoms, signs, and ill-defined conditions (2.7%).

CONCLUSIONS

Twenty-three percent (23%) of people in Taiwan had used acupuncture during this 7-year period. Musculoskeletoal and neurologic disorders were two major categories commonly treated with acupuncture.

Association rule discovery with the train and test approach for heart disease prediction

Association rules represent a promising technique to improve heart disease prediction. Unfortunately, when association rules are applied on a medical data set, they produce an extremely large number of rules. Most of such rules are medically irrelevant and the time required to find them can be impractical. A more important issue is that, in general, association rules are mined on the entire data set without validation on an independent sample. To solve these limitations, we introduce an algorithm that uses search constraints to reduce the number of rules, searches for association rules on a training set, and finally validates them on an independent test set. The medical significance of discovered rules is evaluated with support, confidence, and lift. Association rules are applied on a real data set containing medical records of patients with heart disease. In medical terms, association rules relate heart perfusion measurements and risk factors to the degree of disease in four specific arteries. Search constraints and test set validation significantly reduce the number of association rules and produce a set of rules with high predictive accuracy. We exhibit important rules with high confidence, high lift, or both, that remain valid on the test set on several runs. These rules represent valuable medical knowledge.

Patterns of ambulatory care utilization in Taiwan

BACKGROUND

We used the insurance claims of a representative cohort to quantify the patterns of ambulatory care visits, especially the doctor-shopping phenomenon, in Taiwan.

METHODS

The ambulatory visit files of the 200,000-person cohort datasets from the National Health Insurance Research Database in 2002 were analyzed. Only a visit with physician consultation would be considered. We computed the visit patterns both by visit count and by patient count.

RESULTS

In 2002, there were 182,474 eligible people with 2,443,003 physician consultations. During the year, 87.4% of the cohort had visited physician clinics and 57.5% had visited hospital-based outpatient or emergency departments. On average, a person had 13.4 physician consultations and consulted 3.4 specialties, 5.2 physicians, and 3.9 healthcare facilities in a year. In 2002, 17.3% of the cohort had ever visited different healthcare facilities on the same day; 23.5% had ever visited physicians of the same specialty at different healthcare facilities within 7 days and the percentage of second visits was 3.8% of all visits. Besides, 7.6% of the cohort had visited two or more specialties at the same facility on the same day, and such visits make up 2.5% of all visits.

CONCLUSION

The people in Taiwan did visit the physicians and outpatient departments frequently. Many patients not only consulted several physicians of different specialties and at different healthcare facilities during the year, but also switched the physicians and facilities quickly. An effective referral system with efficient data exchange between facilities might be the solution.

Survey of medication knowledge and behaviors among college students in Taiwan

OBJECTIVES

To survey the knowledge, attitudes, and practices towards safe medication use of first-year college students in Taiwan.

METHODS

One hundred forty-seven departments in 27 universities were sampled by stratified randomization. Three sections of the questionnaire were developed, including 10 true/false questions to measure knowledge, 4 questions for attitude in a 5-point scale, and 10 questions for practice in a 5-point scale.

RESULTS

A total of 6270 subjects completed the survey instrument for a valid response rate of 91.9%. Students who were health science majors had safer medication practices than non-health science majors. However, students in both groups had low scores on appropriate antibiotics use, with a correctness rate of 45.8%. Knowledge relating to antacid use had the lowest number of correct responses among the 10 items included in the survey. Overall, the students showed positive attitudes toward (12.6 +/- 2.2) and trusted (3.6 +/- 0.7) pharmacists' consultation.

CONCLUSIONS

College students lack appropriate knowledge, attitudes, and practices related to the safe use of medications, and efforts are warranted in Taiwan.

Patterns and costs of antipsychotic drug use in Taiwan: 1997 to 2001

The advent of atypical antipsychotics greatly changed the treatment pattern for mental illnesses worldwide. This study was designed to determine the trend in prevalence, prescribing pattern, and cost of antipsychotic agents in Taiwan from 1997 to 2001. Data were obtained from claims completed for a random sample of 200,000 people registered with the National Health Insurance program. The antipsychotics monitored included all group N05A drugs in the Anatomical Therapeutic Chemical classification system (version 2000). Conventional and atypical antipsychotics were handled separately. Of the 195,971 eligible registrants, 37,441 (19.1%) received any kind of antipsychotic during this 5-year period, but only 713 (0.4%) used atypical antipsychotics. The prevalence of conventional antipsychotic use during each successive year of this study was 5.2%, 5.7%, 6.6%, 6.2%, and 6.1% and 0.1%, 0.1%, 0.1%, 0.2%, and 0.3% for atypical agents. Although far fewer registrants used them, atypicals comprised 19.1% of all prescribed amounts measured in defined daily doses and 56.1% of the cost for all antipsychotics in 2001. During the 5-year study period, atypical antipsychotics were prescribed for 405 (57%) patients with schizophrenia, 132 (19%) with affective disorder, 128 (18%) with other psychiatric disorders, and 48 (7%) with a nonpsychiatric disorder. With the loosening of reimbursement restrictions in 2002, continued growth of atypical antipsychotic use in Taiwan might be expected.