Over the last several years, a number of health insurance claims have been electronically submitted and entered into the national database(NDB) by the Ministry of Health, Labor & Welfare(MHLW). A trial was initiated where the NDB would be evaluated for its utility in quality improvement of healthcare services. I had the opportunity to perform the drug utilization study using the NDB, and in this article, I review the procedures by which MHLW make the NDB available and report on issues in analyzing the NDB and in using the NDB for pharmacoepidemiological studies. The NDB is much larger and more encompassing than other medical databases and its findings are expected to havea great influence on a pharmacoepidemiological studies in Japan in providing valuable information on the rational use of drugs. (Jpn J Pharmacoepidemiol 2012; 17(2): 155-162)

To evaluate and promote the rational use of antiarrhythmic drugs, a series of pharmacoepidemiological studies were performed. First, studies on hypoglycemia induced by cibenzoline were performed. The mechanism of the hypoglycemic effect of cibenzoline is related to an increase in insulin secretion. A significantly increased risk of hypoglycemia was observed in patients treated with cibenzoline in a case-controlled study. In particular, close attention should be paid to the occurrence of cibenzoline-induced hypoglycemia in elderly patients, those receiving high doses and in those with reduced renal function. After the introduction of TDM, the risk of hypoglycemia associated with cibenzoline use decreased together with an increase of the percentage of patients whose serum concentrations of cibenzoline had been measured. Dose adjustment based on TDM was beneficial for patients treated with cibenzoline in order to prevent hypoglycemia. In general, drugs are used in accordance with an approved dosage regimen in the expectation of an appropriate balance between efficacy and toxicity. However, a difference was seen between the approved dosage and the actual dose in cibenzoline therapy. Secondly, prescription research of several antiarrhythmic drugs was performed at five national hospitals. Antiarrhythmic drugs were used in lower doses than the approved dosage in clinical practice in Japan. Differences were seen between the approved dosage and the actual dose, and remarkable variations were seen in the dose distribution among the hospitals. The discrepancy between the approved dosage and practical dosage suggests that there is doubt as to whether the approved dosing regimens for antiarrhythmic drugs are appropriate.

Objective: To quantitatively investigate the history of medical pharmacy research by pharmacists in Japan, original article titles from the Japanese Journal of Pharmaceutical Health Care and Sciences (Jpn J Pharm Health Care Sci) and the Journal of Japanese Society of Hospital Pharmacists (JJSHP) were analyzed by text-mining.
Method: The titles of all original articles (2,611 and 2,260 articles) published in Jpn J Pharm Health Care Sci and JJSHP between 1975 and 2009 were collected from article databases and analyzed using KH Coder, the free software for quantitative text analysis of the Japanese language.  KH Coder extracts basic text information data by counting the occurrence rate of certain words.  Article titles were assigned to nine research categories according to coding rules, and the categorization results were analyzed quantitatively.
Results: Between 1975 and 1989, “pharmaceutical investigation” was the major area of research in the Jpn J Pharm Health Care Sci.  Articles assigned to the category “drug therapy” gradually increased through the 1990s and, since 2000, “drug therapy” has dominated medical pharmacy research.  In the JJSHP between 1975 and 2004, no characteristic research area was found, and mainly research articles directly related to pharmacist practice were published.  However, from 2005 to 2009, articles assigned to the “drug therapy” category accounted for 34% of all published articles making “drug therapy” the major research area in the JJSHP.  Thus, in recent years, there is no obvious difference in research areas between the two journals.
Conclusion: Our analyses suggest that drug therapy research is now at the center of medical pharmacy research by pharmacists in Japan.

Objective: To examine the signal of gastrointestinal tract injury induced by aspirin and other drugs, we analyzed the US FDA Adverse Event Reporting System (FAERS).
Methods: After deleting duplicate submissions, we analyzed the reports involving gastrointestinal tract injury associated with aspirin, H2-receptor antagonists (H2RAs), proton pump inhibitors (PPIs), ACE inhibitors, angiotensin II receptor blockers (ARBs), and antiplatelet and antithrombotic drugs.  The reporting odds ratio (ROR), a recognized pharmacovigilance tool, was used for the quantitative detection of signals.
Results: Based on 29,017,485 co-occurrences, i.e., drug-adverse event pairs, found in 1,645,605 reports from 2004 to 2009, the ROR-associated gastrointestinal tract injury for aspirin alone, aspirin with H2RAs, aspirin with PPIs, aspirin with ACE inhibitors, aspirin with ARBs, and aspirin with antiplatelet and antithrombotic drugs were 2.88, 1.42, 1.46, 1.00, 1.05, and 2.98-8.26, respectively.  The following summarizes the types of listed reports: 86 reports described the daily aspirin doses, and 36/86 were between 75 and 100 mg; 343 reports described the periods between the start-date for aspirin and the date when gastrointestinal tract injury occurred, of which 128/343 were within one month while 215/343 were over one month; additionally, 78 reports described the total cumulative doses of aspirin, and 17/78 were between 1 and 5 g.
Conclusion: The data suggest that H2RAs, PPIs, ACE inhibitors, and ARBs may reduce gastrointestinal tract injury associated with aspirin in possibility.

Objective: To examine the association between statin use and the risk of sleep disturbances, data mining was performed on a claims database.
Methods: Symmetry analysis was carried out to identify the risk of sleep disturbances after statin use during the period from January 2005 to December 2011.  Statin use in combination with hypnotic drugs was examined by prescription sequence symmetry analysis.  In this study, hypnotic drugs that are commonly prescribed for the treatment of insomnia were used as markers of sleep disturbances produced by statins. Likewise, event sequence symmetry analysis was undertaken to evaluate the association between statin use and the diagnosis of sleep disturbances.
Results: Significant associations of statin use with short-acting hypnotic drugs were found, with an adjusted SR (sequence ratio) of 1.23 (95%CI: 1.04-1.45) at an interval of 12 months.  Otherwise, significant associations between individual statin use and hypnotic drug use were not found.  Significant associations between use of statins and the diagnosis of sleep disturbances were not also found in this study.
Conclusions: Analysis of the claim database demonstrated that statin therapy might be associated with an emergence of sleep disturbances.  Therefore, individuals prescribed statins should be considered as having an increased risk of sleep disturbances.

Objective: Signal detection by analyzing adverse event spontaneous report databases is used to monitor drug safety.  One of the major spontaneous report databases is the FDA Adverse Event Reporting System (FAERS).  Recently, the Japanese Adverse Drug Event Report database (JADER) was released.  To compare FAERS and JADER, we calculated the signals of adverse events by new quinolones (NQs).
Methods: We extracted reports of adverse events by NQs from FAERS and JADER, and analyzed them using the ROR data mining algorithm.  Thirteen kinds of NQs were extracted, and the terms of adverse events extracted were defined by MedDRA.
Results: There were 35,990,645 reports in FAERS and 1,643,404 reports in JADER.  Significant RORs were found for hypersensitivity (FAERS: 1.78, JADER: 1.47), arrhythmia (1.07, 0.68), hypoglycemia (1.80, 2.03), hyperglycemia (0.72, 0.78), rhabdomyolysis (1.01, 0.78), tendon disorders (15.18, 6.59), psychiatric symptoms (1.12, 0.45) and convulsion (0.99, 1.31).  We identified 4 types of adverse events by comparing FAERS and JADER: 1) Signal detection in both, 2) No signal detection in either, 3) Signal detection only in FAERS, 4) Signal detection only in JADER.
Conclusion: Analyzing spontaneous report databases has several limitations, but is still a valuable tool for identifying potential associations between drugs and adverse events.  Spontaneous report databases may also be useful for detecting differences in adverse events between different races, countries and regions.

Objective: To examine the association between atypical and typical antipsychotics and extrapyramidal symptoms (EPS), we analyzed the US Food and Drug Administration Adverse Event Reporting System (FAERS) and the Japanese Adverse Drug Event Report database (JADER) from the Pharmaceuticals and Medical Devices Agency (PMDA).
Methods: A reporting odds ratio was calculated and used to detect spontaneous report signals, with detection defined as a lower limit >1 in a 95% confidence interval.  In addition, time to onset and age at onset of EPS were investigated.
Results: Drug-reaction pairs were identified in both FAERS (n=29,017,485) and JADER (n=2,079,653).  In analyses of both databases, significant associations were found between atypical and typical antipsychotics and EPS.  Atypical antipsychotics cause EPS with a longer duration of therapy compared to typical ones.  EPS in patients treated with atypical antipsychotics was observed at a broad range of ages compared to the patients treated with typical ones.
Conclusion: Atypical antipsychotics, like typical ones, may increase the risk of EPS.  Because of the longer latency of onset, it may be difficult to find EPS associated with atypical antipsychotics.  Therefore, the severe symptom may be developed in patients treated with atypical antipsychotics.  The attention should be paid to the EPS in patients of all ages treated with atypical antipsychotics.

Objective: To compare effects of the fluoropyrimidines S-1 and capecitabine on prothrombin time international normalized ratios (PT-INR) of warfarin following coadministration and after discontinuation of each fluoropyrimidine treatment.
Methods: Medical records of patients receiving warfarin with either S-1 (6 patients) or capecitabine (7 patients) were obtained from four hospitals.
Results: Increased PT-INR was observed until peak levels of warfarin were achieved in all patients in S-1 and capecitabine treatment groups.　Moreover, PT-INR significantly changed after coadministration within each group (p＜0.05).　Specifically, ratios of peak PT-INR after coadministration of each fluoropyrimidine and those following administration of warfarin alone (PT-INR elevation ratio) were 3.31 and 3.29 in S-1 and capecitabine coadministration groups, respectively.　Moreover, numbers of days to peak PT-INR were 38.3 and 31.3 days, respectively, and did not significantly differ between the treatment groups.　Furthermore, PT-INR returned to pretreatment levels by 17.5 and 15.1 days after discontinuation of S-1 and capecitabine, respectively, and did not significantly differ between the treatment groups.
Conclusion: Coadministration of S-1 and capecitabine similarly prolongs PT-INR by approximately 3-fold compared with administration of warfarin alone; therefore, these drug-drug interactions were clinically suggested to be of high risk for episodes of bleeding and remarkable alterations in coagulation parameters.　Therefore, blood coagulation ability should be more carefully monitored with regard to PT-INRs in patients receiving warfarin with S-1 or capecitabine not only during coadministration but also after discontinuation of fluoropyrimidine treatments.