Adverse Drug Reaction Signals Research Articles

The validity of sequence symmetry analysis (SSA) for adverse drug reaction signal detection

To determine the validity of sequence symmetry analysis (SSA) method to detect adverse drug reactions from an administrative claims database. Published randomised controlled trials (RCTs) of 19 medicines were identified through search databases, product information (PI) or the US Food and Drug Administration Web site. All adverse events (AEs) in the RCTs and the PI for the medicines were extracted. AEs were considered 'gold standard positive events' if they were reported as being statistically significant events in adequately powered RCTs. The remaining AEs were considered 'gold standard negative events' if the event was not listed as an AE in the PI for that medicine or any other medicine in its class. Indicators of AEs were identified by consensus from two clinical researchers. SSA was run for each medicine-indicator pair using four different time windows: 3, 6, 9 and 12 months. A total of 120 randomised placebo controlled trials were reviewed for the 19 tested medicines. A total of 165 medicine-indicator pairs (44 positive and 121 negative events) were identified and tested by SSA. At the 12-month time window, the sensitivity, specificity, positive and negative predictive values of SSA were 61% (95%CI 0.46-0.74), 93% (95%CI 0.87-0.96), 77% (95%CI 0.61-0.88) and 87% (95%CI 0.80-0.92), respectively. Using a 3-month time window, the SSA had a lower sensitivity (52%). The SSA technique was found to have moderate sensitivity and high specificity for detecting ADRs. These results suggest that SSA is a potential tool for detecting ADRs using administrative claims data that could complement existing pharmacosurveillance methods.

Comparison of Algorithms That Detect Drug Side Effects Using Electronic Healthcare Databases

The electronic healthcare databases are starting to become more readily available and are thought to have excellent potential for generating adverse drug reaction signals. The Health Improvement Network (THIN) database is an electronic healthcare database containing medical information on over 11 million patients that has excellent potential for detecting ADRs. In this paper we apply four existing electronic healthcare database signal detecting algorithms (MUTARA, HUNT, Temporal Pattern Discovery and modified ROR) on the THIN database for a selection of drugs from six chosen drug families. This is the first comparison of ADR signalling algorithms that includes MUTARA and HUNT and enabled us to set a benchmark for the adverse drug reaction signalling ability of the THIN database. The drugs were selectively chosen to enable a comparison with previous work and for variety. It was found that no algorithm was generally superior and the algorithms’ natural thresholds act at variable stringencies. Furthermore, none of the algorithms perform well at detecting rare ADRs.

A Signal Detection Method to Detect Adverse Drug Reactions Using a Parametric Time-to-Event Model in Simulated Cohort Data

Current quantitative signal detection methods have been primarily developed for the purpose of detecting signals from spontaneous reports. These methods are not always appropriate for cohort data. More recently, parametric time-to-event models have been proposed to model hazard functions with the ultimate aim of detecting adverse drug reactions (ADRs). The rate of occurrence of ADRs after starting a drug will depend upon the causal mechanism and therefore will often vary with time, in contrast to events not associated with the drug, which will tend to occur at a constant background rate. After starting treatment, the onset of ADRs will be rapid for some but delayed for others. A non-constant rate over time may indicate a drug-event relationship. The aim of this study was to propose a simple test to detect signals of ADRs in cohort data and to investigate the power of this test using simulated data. A signal detection tool using the proposed test to improve the power of detection is also described. In order to test for a non-constant hazard (rate of occurrence), the hazard function was estimated using the model shape parameter for the Weibull function. If the shape parameter was found to be significantly different (p < 0.05) from the value one (the value for a constant hazard) a signal was raised. Simulation of background event rates used were 1%, 5% and 10% of the cohort size. The ADR rate was varied in proportion to the background rate; a 10%, 20% and 50% increase in the background rate was explored. The time of occurrence of the ADR will dictate the shape of the hazard function, therefore the ability of the model to detect a signal depending when the highest risk for ADR was also explored. The power of the test was investigated by simulation. The Weibull Shape Parameter (WSP) test was most powerful at detecting signals that occur shortly after starting treatment. These preliminary simulations had low power when the underlying hazard function was symmetrical (e.g. when ADRs occurred in the middle of the study period). The power of the test was improved by censoring the data as this broke the symmetry of the hazard function. A tool that censored the data at regular intervals and repeated the WSP test was found to correctly detect ADR or no ADR around 90% of the time when the sample size was at least 5000. The WSP test is simple to implement using standard statistical software, and can be used to detect non-constant hazards over time in order to raise signals of time-dependent ADRs. When there is no pre-specified event of interest or the time of the ADR is uncertain, the WSP tool should be used instead of the WSP test. These methods do not require any external data for comparative purposes and thus can be implemented in a single cohort of participants exposed to a drug.

Translational Bioinformatics: Data-driven Drug Discovery and Development

Internet-accessible computing power and data-sharing mandates now enable researchers to interrogate thousands of publicly available databases containing molecular, clinical, and epidemiological data. With emerging new approaches, translational bioinformatics can now provide answers to previously untouchable questions, ranging from detecting population signals of adverse drug reactions to clinical interpretation of the whole genome. There are challenges, including lack of access to some data sources and software, but there are also overwhelming doses of hopes and expectations.

Surveillance of contrast-media-induced hypersensitivity reactions using signals from an electronic medical recording system

BackgroundContrast-media (CM) hypersensitivity is a well-known adverse drug reaction. Surveillance of adverse drug reactions usually depends on spontaneous reports. However, the rate of spontaneous reports is low. Recent progress in information technology enables the electronic search on signals of adverse drug reactions from electronic medical recording (EMR) systems. ObjectivesTo analyze the incidence and clinical characteristics of CM hypersensitivity using an EMR-based surveillance system. MethodsThe surveillance system used signals from standardized terms within the international classification of nursing practice terms that can indicate symptoms of CM hypersensitivity and from the order codes for procedures that used contrast media, antihistamine, and epinephrine. The search strategy was validated by allergists comparing the electronic search strategy versus manually reviewing medical charts over one month. The main study covered for one year period. ResultsDetection rate of the electronic search method was 0.9% (7/759), while that of the manual search method was 0.8% (6/759). EMR-based electronic search method was highly efficient: reduced the charts that needed to be reviewed by 96% (28/759). The sensitivity of electronic screening was 66.7%, specificity was 99.6%, and the negative predictive value was 99.7%. CM hypersensitivity reactions were noted in 266 among 12,483 cases (2.1%). Urticaria was the most frequent symptom (74.4%). CT was the most frequent procedure (3.6%) that induced CM hypersensitivity. ConclusionA surveillance system using EMR may be a useful tool in the study of drug hypersensitivity epidemiology and may be used in an adverse drug reaction alarm system and as a clinical, decision making support system.

Detection of Adverse Drug Reaction Signals Using an Electronic Health Records Database: Comparison of the Laboratory Extreme Abnormality Ratio (CLEAR) Algorithm

Electronic health records (EHRs) are expected to be a good source of data for pharmacovigilance. However, current quantitative methods are not applicable to EHR data. We propose a novel quantitative postmarketing surveillance algorithm, the Comparison of Laboratory Extreme Abnormality Ratio (CLEAR), for detecting adverse drug reaction (ADR) signals from EHR data. The methodology involves calculating the odds ratio of laboratory abnormalities between a specific drug-exposed group and a matched unexposed group. Using a 10-year EHR data set, we applied the algorithm to test 470 randomly selected drug-event pairs. It was found possible to analyze a single drug-event pair in just 109 ± 159 seconds. In total, 120 of the 150 detected signals corresponded with previously reported ADRs (positive predictive value (PPV) = 0.837 ± 0.113, negative predictive value (NPV) = 0.659 ± 0.180). By quickly and efficiently identifying ADR signals from EHR data, the CLEAR algorithm can significantly contribute to the utilization of EHR data for pharmacovigilance.

Statistical Analysis System of Spontaneous Adverse Drug Reaction Reports

Background: Spontaneous adverse drug reaction (ADR) reporting data has been used for safety of post-market drug surveillance. A system has been required that is able to detect signals associated with drugs by analyzing the collected ADR data. Methods: We developed the web-based automated analysis system (ADR-detector). We used the data which reported ADR spontaneously between March 2009 and December 2010 to Korean Food and Drug Administration. We used 3 statistical indicators for evaluating ADR signals: proportional reporting ratio (PRR), reporting odds ratio (ROR), and information component (IC). The ADR reports which were detected as significant signals based on the indicators have been reviewed. Results: Among 153,774 reports, 9,955 cases were related to 4 analgesics which were most frequently reported analgesic drugs during the study period. The numbers of ADR reports associated with each drug are as follow: 5,623 reports in tramadol (56.5 %), 1,720 reports in fentanyl (17.3 %), 1,463 reports in tramadol-combination (14.7 %), and 1,149 reports in ketorolac (11.5 %). Top 5 ADR were nausea (3,351 reports – 33.7 %), vomiting (1,755 reports – 17.6 %), dizziness (1,130 – 11.4 %), rash (412 reports – 4.1 %), and pruritus (354 reports – 3.6 %). 6,674 ADR reports were significant based on PRR and ROR, and 336 reports were significant based on IC. Conclusion: By using the automated analysis system, not only statisticians but also general researchers are able to analyze ADR signals in real-time. Also ADR-detector would provide rapid review and cross-check of ADR.

Comparison of Adverse Drug Reactions for Patients Treated with Tyrosine Kinase Inhibitors: Data Mining of the Public Version of the FDA Adverse Event Reporting System (AERS)

Abstract 1698 Backgrounds:The consecutive approvals of tyrosine kinase inhibitors (TKIs) have been changing the landscape of treatment strategy for patients with chronic myeloid leukemia (CML). Currently, three TKIs are available worldwide, including imatinib (Glivec/Gleevec; Novartis Pharmaceuticals, East hanover, NJ), nilotinib (Tasigna; Novartis Pharmaceuticals) and dasatinib (Sprycel; Bristol-Myers Squibb, Princeton, NJ). Although second generation TKIs (nilotinib and dasatinib) have shown their efficacy and safety in recent clinical trials, additional data are needed for better understanding and differences in their safety profiles may be helpful when choosing a TKI. We compared the adverse drug reactions (ADRs) for patients treated with three TKIs using spontaneous adverse event reporting after approval to investigate the characteristics of safety profiles. Method:To compare adverse events characteristics among three TKIs, the case/noncase adverse events reports associated with TKIs use were retrieved from the U.S. Food and Drug Administration Adverse Event Reporting System (AERS) between 2004 and 2010. We calculated the reporting odds ratio (ROR), which is known as one of data mining algorithms for signal detection techniques of ADRs, characterized by providing a fast and cost-efficient way of detecting possible ADR signals. All events in the AERS have been coded for data entry in accordance with the standardized terminology, known as Preferred Terms, in the Medical Dictionary for Regulatory Activities. The ROR is similar to the idea of odds ratio, calculating the odds of exposure of the suspected drug in patients who had events divided by the odds of exposure of the suspected drug in those without events. The ROR -1.96 standard error greater than 1 with at least 4 ADR reports was used as a signal criterion in this study. Results:We identified 18,979 ADRs for imatinib, 5,388 ADRs for nilotinib, and 2,482 ADRs for dasatinib. The number of ADRs flagged by our signal criterion was 91 for imatinib, 82 for nilotinib, and 109 for dasatinib. Top 10 lists of ADRs with higher ROR are shown in Table for each TKI. The safety profiles were almost different among TKIs. ADRs related to skin and hepatic function were noted for imatinib, whereas ADRs related to cardiac events were prominent for nilotinib, and ADRs related to lymphocytosis, edema and effusion were noticeable for dasatinib. The different dosing requirements of dasatinib and nilotinib may be an additional factor of ADRs.Table:Top 10 Lists of ADRs for TKIs, Imatinib, Nilotinib and DasatinibType of ADRs ranked by RORRORReported Number of Pts with ADRsImatinibImatinibNilotinibDasatinibImatinibNilotinibDasatinibTotal1HEPATITIS14.80.00.024252AMNESIA11.70.00.6191203INTERNATIONAL NORMALISED RATIO INCREASED9.90.30.0161174LICHEN PLANUS9.20.00.015165HEPATIC CIRRHOSIS8.60.30.0141156CHOLESTASIS8.00.00.9131147RASH MACULAR8.00.00.013148HIP ARTHROPLASTY7.40.40.0121139INJURY6.80.01.01111210DERMATITIS BULLOUS6.80.40.011112NilotinibImatinibNilotinibDasatinibImatinibNilotinibDasatinibTotal1CORONARY ARTERY STENOSIS0.141.30.24441492ARTERIAL STENT INSERTION0.137.50.01893HEPATITIS VIRAL0.137.50.01894INTERMITTENT CLAUDICATION0.137.50.01895LABELLED DRUG-FOOD INTERACTION MEDICATION ERROR0.032.80.00786CORONARY ARTERY BYPASS0.125.00.62161197ANGINA UNSTABLE0.123.50.01568DRUG ADMINISTRATION ERROR0.023.52.305169ELECTROCARDIOGRAM ST SEGMENT DEPRESSION0.218.80.03121510TROPONIN T INCREASED0.118.81.318110DasatinibImatinibNilotinibDasatinibImatinibNilotinibDasatinibTotal1LYMPHOCYTOSIS0.20.045.428102ACUTE PULMONARY OEDEMA0.20.59.5315113PLEURAL EFFUSION0.40.77.4130471403544OVERDOSE0.40.07.155135FULL BLOOD COUNT DECREASED1.00.06.85386PULMONARY EMBOLISM0.60.86.713410277ACUTE FEBRILE NEUTROPHILIC DERMATOSIS0.80.65.751398COLITIS1.00.05.2105169MIGRAINE0.41.65.26451610PULMONARY HAEMORRHAGE0.70.05.07413 Conclusions:ADRs reported in the AERS for each TKI were relatively consistent with known characteristics of ADRs reported in previous clinical trials. Our information would be supportive data for choosing a TKI for CML patients based on comorbidities and drug safety profiles. The choice of therapy in a given patient with CML may depend on age, past history and comorbidities as well as disease risk score and mutational analysis. Disclosures:Oshima:Sanofi Aventis: Employment.

Benefits and strengths of the disproportionality analysis for identification of adverse drug reactions in a pharmacovigilance database

Adverse drug reactions (ADRs) represent an important medical issue: they result in 3–7% of all hospital admissions and are associated with a substantial increase in morbidity and mortality. Numerous methods can be used to investigate ADRs. Each has its strengths and weakness. The insufficiencies of basic (experimental) pharmacology as well as clinical trials for studying ADRs are well known. Animal physiology often differs from that of humans. Clinical trials, although necessary, do not allow definite conclusions, because they are built to evaluate efficacy more than safety. Thus, spontaneous notifications remain the cornerstone for ADRs despite their mandatory limitations (under-reporting, selective reporting, lack of denominator etc.). Intensive studies could allow quantification of a specific problem of drug safety (i.e. drug admission into hospital for ADRs). For some years, several pharmacoepidemiological methods have been used to identify and also to quantify ADRs. Among thesse methods, case–control or cohort studies are most widely used. However, their setting up requires specific organization, delay, money and also use of large databases, which are not often specifically built for evaluation of ADRs. In the present issue of the journal, clinical pharmacologists and pharmacoepidemiologists from Poitiers University Hospital (France) have used another method, working from the French PharmacoVigilance database. They used disproportionality analysis for identification of memory disorders associated with drugs [1]. The present paper discusses the benefits and strengths of this method.

Safety research in Chinese medicines based on application of data mining

The current status of the application of data mining in Chinese medicine (TCM) safety research are analyzed in this paper, and the future development trend are discussed, which include: to establish ADR (adverse drug reaction) signal detection and automatic warning system based on traditional Chinese medicine (TCM) adverse reaction database, to explore the characteristics and influencing factors of TCM safety problems, to devise appropriate new algorithm, and to develop TCM adverse reaction literature mining.

A novel algorithm for detection of adverse drug reaction signals using a hospital electronic medical record database

Quantitative analytic methods are being increasingly used in postmarketing surveillance. However, currently existing methods are limited to spontaneous reporting data and are inapplicable to hospital electronic medical record (EMR) data. The principal objectives of this study were to propose a novel algorithm for detecting the signals of adverse drug reactions using EMR data focused on laboratory abnormalities after treatment with medication, and to evaluate the potential use of this method as a signal detection tool. We developed an algorithm referred to as the Comparison on Extreme Laboratory Test results, which takes an extreme representative value pair according to the types of laboratory abnormalities on the basis of each patient's medication point. We used 10 years' EMR data from a tertiary teaching hospital, containing 32,033,710 prescriptions and 115,241,147 laboratory tests for 530,829 individual patients. Ten drugs were selected randomly for analysis, and 51 laboratory values were matched. The sensitivity, specificity, positive predictive value, and negative predictive value of the algorithm were calculated. The mean number of detected laboratory abnormality signals for each drug was 27 (±7.5). The sensitivity, specificity, positive predictive value, and negative predictive value of the algorithm were 64-100%, 22-76%, 22-75%, and 54-100%, respectively. The results of this study demonstrated that the Comparison on Extreme Laboratory Test results algorithm described herein was extremely effective in detecting the signals characteristic of adverse drug reactions. This algorithm can be regarded as a useful signal detection tool, which can be routinely applied to EMR data.

A Potential Causal Association Mining Algorithm for Screening Adverse Drug Reactions in Postmarketing Surveillance

Early detection of unknown adverse drug reactions (ADRs) in postmarketing surveillance saves lives and prevents harmful consequences. We propose a novel data mining approach to signaling potential ADRs from electronic health databases. More specifically, we introduce potential causal association rules (PCARs) to represent the potential causal relationship between a drug and ICD-9 (CDC. (2010). International Classification of Diseases, Ninth Revision (ICD-9). [Online]. Available: http://www.cdc.gov/nchs/icd/icd9.html) coded signs or symptoms representing potential ADRs. Due to the infrequent nature of ADRs, the existing frequency-based data mining methods cannot effectively discover PCARs. We introduce a new interestingness measure, potential causal leverage, to quantify the degree of association of a PCAR. This measure is based on the computational, experience-based fuzzy recognition-primed decision (RPD) model that we developed previously (Y. Ji, R. M. Massanari, J. Ager, J. Yen, R. E. Miller, and H. Ying, "A fuzzy logic-based computational recognition-primed decision model," Inf. Sci., vol. 177, pp. 4338-4353, 2007) on the basis of the well-known, psychology-originated qualitative RPD model (G. A. Klein, "A recognition-primed decision making model of rapid decision making," in Decision Making in Action: Models and Methods, 1993, pp. 138-147). The potential causal leverage assesses the strength of the association of a drug-symptom pair given a collection of patient cases. To test our data mining approach, we retrieved electronic medical data for 16,206 patients treated by one or more than eight drugs of our interest at the Veterans Affairs Medical Center in Detroit between 2007 and 2009. We selected enalapril as the target drug for this ADR signal generation study. We used our algorithm to preliminarily evaluate the associations between enalapril and all the ICD-9 codes associated with it. The experimental results indicate that our approach has a potential to better signal potential ADRs than risk ratio and leverage, two traditional frequency-based measures. Among the top 50 signal pairs (i.e., enalapril versus symptoms) ranked by the potential causal-leverage measure, the physicians on the project determined that eight of them probably represent true causal associations.

The proportion of patient reports of suspected ADRs to signal detection in the Netherlands: case–control study

To determine the contribution of patients' adverse drug reaction (ADR) reports to signals detection, through a study of the signals sent by Lareb to the Dutch Medicines Evaluation Board. The percentage of patient's ADR reports contributing to generate signals of adverse drug reactions was determined. A case-control design was used to study if the proportion of patient reports in associations that had been selected as 'signals' differed from non-signals. A logistic regression analysis was used to calculate the Odds Ratio with 95%CI for patient reports in the cases and controls. The number of patient reports which contributed to a signal has increased from 0 reports in 2003 to 31 reports in 2008 (9% of total). Since 2005 patient reports have triggered particular associations to be selected as a signal. In 2007, 28% of all trigger reports were reported by a patient. The case-control analysis showed that patient reports were equally present in the reports used in signal formation (cases) as in the controls, reports not contributing to a signal. Odds Ratio (OR) was 1.10 (95%CI 0.81-1.49) and OR = 0.96 (95%CI 0.50-1.87) for the 'trigger reports'. The proportion of patient reports contributing to generate signals was equal to the proportion of patient reports in the database. Patient's reports of adverse drug reactions can provide a valuable contribution to the detection of signals in addition to healthcare professionals' reports. In the Netherlands, direct patient reports have added to the signals of adverse drug reactions sent to the Medicines Evaluation Board.

Improving the Reporting of Adverse Drug Reactions in the Hospital Setting

The US Food and Drug Administration (FDA) is perceived by the public as having a substantial responsibility to ensure drug safety; however, the FDA has limited resources for active surveillance and relies on voluntary reporting of adverse events and potential adverse drug reactions. Studies have shown that underreporting of adverse events and adverse drug reactions is widespread. Furthermore, a review of several studies demonstrates that most adverse drug reactions are reported by pharmacists and nurses, with physicians reporting the fewest. The hospital setting, with its clearly defined patient population observed around the clock, is an ideal setting in which to identify potential adverse drug reaction signals and to report them to either the drug manufacturer or the FDA. In this article we describe the present system for addressing adverse events, obstacles to reporting them, and the important role any hospital physician could play in reporting adverse events and potential adverse drug reactions.

A novel method for signal detection of adverse drug reactions based on proportional reporting ratios

The proportional reporting ratio (PRR) is a statistical method for signal detection of adverse drug reactions (ADRs) based on unbalanced proportions. Although effective, this method only takes into account the proportional relation based on target adverse reactions and ignores the relation between a given ADR and the others for the same drug. Therefore, it is necessary to improve the calculation deviation in PRR. In this study, we developed a novel PRR (NPRR) method and compared it with the original PRR method for the purpose of a combined application of these two methods for ADR signal detection. NPRR is also based on unbalanced proportions, in which the proportion for a given ADR is linked to a specific drug (or all other drugs), and then divided by the corresponding proportion for all other ADRs. Applying this method to the ADR data of Jiangsu Province, China in 2008 and 2009, we detected 3,021 signals. Compared with the PRR method, the sensitivity of our method is 0.99, the specificity is 0.97, and the Youden index is 0.96. NPRR is an excellent method supplementary to PRR. The combination of these two methods can reduce calculation deviation and detect ADRs more effectively.

Detection of Adverse Drug Reaction Signals in the Thai FDA Database: Comparison between Reporting Odds Ratio and Bayesian Confidence Propagation Neural Network Methods

The study aimed to compare performance between the reporting odds ratio (ROR) and the Bayesian confidence propagation neural network (BCPNN) methods in identifying serious adverse drug reactions (ADRs) using the Thai FDA spontaneous database. The two methods were retrospectively applied to identify new, serious ADRs reported with antiretroviral therapy (ART) drugs using the data set between 1990 and 2006. We plotted the ROR and the information component against time to compare the differential timing of signal detection and the pattern of signaling over time between these methods. The ROR and the BCPNN methods identified the associations between ART drugs and serious ADRs at the same time. Both methods were similar in detecting the first signal of a potential ADR. However, the pattern of signaling seems relatively different with each method. Additional analyses of different drugs, ADRs, and databases will contribute to increase understanding of methods for postmarketing surveillance using spontaneous reporting system.

Signal detection of rosuvastatin compared to other statins: data‐mining study using national health insurance claims database

To detect adverse drug reaction (ADR) signals of rosuvastatin compared to other statins with a novel data-mining approach based on relative risk (RR) using the national health insurance claims database, and to evaluate the usefulness of this method as a tool for signal detection. We used the Health Insurance Review & Assessment Service (HIRA) claims database (Seoul, Korea). Serious adverse events (SAE) were defined as any diagnostic code at the time of hospitalization within 12 weeks from a statin prescription date, regardless of causality. Among statin users, RRs were calculated to compare the proportion of rosuvastatin-specific SAE pairs for rosuvastatin users with the corresponding proportion of drug-SAE pairs for users of other statins. Any SAE for which the lower limit of the RR's 95% confidence interval was greater than 1 was defined as a signal. All detected signals were reviewed to determine whether the signals corresponded with published adverse events (AEs) exclusive to rosuvastatin. Among 96 236 elderly outpatients who received rosuvastatin, or other statins, from January 2005 to September 2005, 40 304 drug-SAE pairs and 376 SAEs were observed. Twenty-five (6.6%) SAEs were detected as signals by the RR-based data-mining approach. Among the 13 references AEs published to be exclusive to rosuvastatin, 8 (61.5%) were found to correspond with the detected signals with a positive predictive value (PPV) of 32%. The RR-based data-mining approach successfully detected signals for rosuvastatin using a national health insurance claims database. This approach could be useful for safety surveillance of marketed products.

A computerized system for detecting signals due to drug–drug interactions in spontaneous reporting systems

WHAT IS ALREADY KNOWN ABOUT THE SUBJECT: * Concomitant use of different drugs may yield excessive risk for adverse drug reactions and it is a challenging task to do surveillance on the safety profile of the interaction between different drugs. * Currently, several methods are used by pharmacoepidemiologists and statisticians to detect possible drug-drug interactions in spontaneous reporting systems. * However, with the increasing number of reports in the system, there is a growing need for a computerized system that could facilitate the process of data arrangement and detection of drug interaction. * We had already developed a computerized system to detect adverse drug reaction signals due to single drugs. * After the development of this system, interaction between different drugs could also be detected automatically and intelligently. In spontaneous reporting systems (SRS), there is a growing need for the automated detection of adverse drug reactions (ADRs) resulting from drug-drug interactions. In addition, special attention is also needed for systems facilitating automated data preprocessing. In our study, we set up a computerized system to signal possible drug-drug interactions by which data acquisition and signal detection could be carried out automatically and the process of data preprocessing could also be facilitated. This system was developed with Microsoft Visual Basic 6.0 and Microsoft Access was used as the database. Crude ADR reports submitted to Shanghai SRS from January 2007 to December 2008 were included in this study. The logistic regression method, the Omega shrinkage measure method, an additive model and a multiplicative model were used for automatic detection of drug-drug interactions where two drugs were used concomitantly. A total of 33 897 crude ADR reports were acquired from the SRS automatically. The 10 drug combinations most frequently reported were found and the 10 most suspicious drug-drug ADR combinations for each method were detected automatically after the performance of the system. Since the detection of drug-drug interaction depends upon the skills and memory of the professionals involved, is time consuming and the number of reports is increasing, this system might be a promising tool for the automated detection of possible drug-drug interactions in SRS.

A Distributed, Collaborative Intelligent Agent System Approach for Proactive Postmarketing Drug Safety Surveillance

Discovering unknown adverse drug reactions (ADRs) in postmarketing surveillance as early as possible is of great importance. The current approach to postmarketing surveillance primarily relies on spontaneous reporting. It is a passive surveillance system and limited by gross underreporting (<10% reporting rate), latency, and inconsistent reporting. We propose a novel team-based intelligent agent software system approach for proactively monitoring and detecting potential ADRs of interest using electronic patient records. We designed such a system and named it ADRMonitor. The intelligent agents, operating on computers located in different places, are capable of continuously and autonomously collaborating with each other and assisting the human users (e.g., the food and drug administration (FDA), drug safety professionals, and physicians). The agents should enhance current systems and accelerate early ADR identification. To evaluate the performance of the ADRMonitor with respect to the current spontaneous reporting approach, we conducted simulation experiments on identification of ADR signal pairs (i.e., potential links between drugs and apparent adverse reactions) under various conditions. The experiments involved over 275,000 simulated patients created on the basis of more than 1000 real patients treated by the drug cisapride that was on the market for seven years until its withdrawal by the FDA in 2000 due to serious ADRs. Healthcare professionals utilizing the spontaneous reporting approach and the ADRMonitor were separately simulated by decision-making models derived from a general cognitive decision model called fuzzy recognition-primed decision (RPD) model that we recently developed. The quantitative simulation results show that 1) the number of true ADR signal pairs detected by the ADRMonitor is 6.6 times higher than that by the spontaneous reporting strategy; 2) the ADR detection rate of the ADRMonitor agents with even moderate decision-making skills is five times higher than that of spontaneous reporting; and 3) as the number of patient cases increases, ADRs could be detected significantly earlier by the ADRMonitor.

Consumers' reports of suspected adverse drug reactions volunteered to a consumer magazine

In some countries, reporting of adverse drug reactions (ADRs) by patients has been incorporated into spontaneous reporting systems, primarily to increase the number of reports and detect new ADR signals earlier [1]. To date only seven countries have allowed patients to report ADRs: Sweden (since 1978 via KILEN), USA (since 1993), Australia, Canada, Denmark and the Netherlands (since 2003) and the UK (since 2005) [1]. An analysis of ADR reports to the Danish Medicines Agency (DKMA) showed that patients are more likely to report ADRs from the nervous and psychiatric system than are health professionals, that patients' share of reports on serious ADRs was comparable to that of physicians, and that patients provided new and unknown information about ADRs [2]. The use of the internet as a source of information and discussion forum for health matters is rapidly increasing. Therefore we decided to investigate the information about possible ADRs provided by consumers who had easy reporting access via the internet. A consumer magazine provided the opportunity for their readers and others to report ADRs via an open access website. This letter examines whether these reports contributed with information about ADRs not previously described in the product information. Data were collected from a questionnaire on the website of a consumer magazine during a 2-month period in autumn 2008 (September–October). If patients had experienced ADRs from the use of medicines, they were requested to answer the following questions: To which medicine did you have an ADR? What were you being treated for? What sort of ADRs did you experience? Did you tell anyone about the ADRs you experienced either at the time or later? Did you report your ADRs either via http://www.borger.dk or to the Danish Medicines Agency? We analysed the reports with respect to therapeutic groups [Anatomical Therapeutic Chemical (ATC) system][3] and types of reported ADRs (system organ classes) [4]. The reported ADRs were checked against the official summary of product information (SPC). ADRs not mentioned in the SPC were classified as unlabelled. Table 1 displays the number of reported ADRs distributed on therapeutic groups and the number of unlabelled ADRs (ATC level 1). Only 11 of the reports (3%) had been reported to the DKMA. In 42% of reports, patients had shared the information about ADRs with a healthcare professional, a family member or a relative. Forty percent of all ADRs were reported for medicines belonging to ATC group N (nervous system). In total, 15% of reported ADRs (n= 188) were not mentioned in the official product information. Of these ADRs, the highest share, 34%, was reported for medicines belonging to ATC group N (nervous system disorders). The majority of reported ADRs were of the types ‘nervous system disorders’ and ‘gastrointestinal disorders’. Table 1 Reports and ADRs of suspected medicines (ATC level 1) by therapeutic group Open access to other patients' reports as well as the simple structure of the questionnaire may have motivated many patients to report ADRs. Such a data collection method could be used as a rapid data collection instrument in the event of suspected serious or rare ADRs not previously documented. The large number of ADRs reported for psychotropic medicines (ATC group N) could reflect that people with central nervous system problems are more likely to report these symptoms than other people. Although the types of reported ADRs and suspected medicines are in line with results from other countries [1, 2, 5], patients also reported information about previously unlabelled ADRs. This study has confirmed that patients report rather unspecific symptoms, e.g. indisposition, dizziness and insomnia, as they use lay terminology to describe symptoms differently from terminology used by healthcare professionals. Patients also reported several ADRs such as drowsiness, weight gain and sexual problems, which prescribers may not consider serious but are troublesome to patients and limit the full enjoyment of daily life [1] and which patients find worthy of reporting in a questionnaire. The content and quality of the reported data are inappropriate for causality analysis as the reports contained no information about age, sex, diagnosis and concomitant medicines. Hence, the collected ADR data may actually consist of ADRs that patients believe to be possibly drug-related reactions rather than confirmed ADRs, and the value of these data in drug surveillance is limited. Consumer ADR reports might act as whistleblowers of new and previously undetected ADRs, but if the quality of the reports is questionable they bring too much noise rather than valuable information to the pharmacovigilance systems. Further studies to explore the quality, validity and impact of consumer reports in the pharmacovigilance systems are needed and appropriate systems for patient reports of ADRs should be explored. The unlabelled ADRs should lead to further investigations and possible detection of new ADRs.