Untitled

Real-time Prescription Surveillance and its Application toMonitoring Seasonal Influenza Activity in Japan Tamie Sugawara1, PhD; Yasushi Ohkusa1, PhD; Yoko Ibuka2, PhD; Hirokazu Kawanohara3, BS; Kiyosu Taniguchi1, 1National Institute of Infectious Diseases, Infectious Disease Surveillance Center, Tokyo, Japan2Hitotsubashi University, Tokyo, Japan3EM Systems Co., Ltd, Tokyo, Japan Corresponding Author:
Tamie Sugawara, PhD
National Institute of Infectious Diseases
Infectious Disease Surveillance Center
1-23-1Toyama
Shinjuku
Tokyo, 162-8640
Japan
Phone: 81 3 5285 1111
Fax: 81 3 5285 1129
Email:
Background: Real-time surveillance is fundamental for effective control of disease outbreaks, but the official sentinel surveillance
in Japan collects information related to disease activity only weekly and updates it with a 1-week time lag.
Objective: To report on a prescription surveillance system using electronic records related to prescription drugs that was started
in 2008 in Japan, and to evaluate the surveillance system for monitoring influenza activity during the 2009-2010 and 2010-2011
influenza seasons.
Methods: We developed an automatic surveillance system using electronic records of prescription drug purchases collected
from 5275 pharmacies through the application service provider's medical claims service. We then applied the system to monitoring
influenza activity during the 2009-2010 and 2010-2011 influenza seasons. The surveillance system collected information related
to drugs and patients directly and automatically from the electronic prescription record system, and estimated the number of
influenza cases based on the number of prescriptions of anti-influenza virus medication. Then it shared the information related
to influenza activity through the Internet with the public on a daily basis.
Results: During the 2009-2010 influenza season, the number of influenza patients estimated by the prescription surveillance
system between the 28th week of 2009 and the 12th week of 2010 was 9,234,289. In the 2010-2011 influenza season, the number
of influenza patients between the 36th week of 2010 and the 12th week of 2011 was 7,153,437. The estimated number of influenza
cases was highly correlated with that predicted by the official sentinel surveillance (r =.992, P <.001 for 2009-2010; r =.972,
P <.001 for 2010-2011), indicating that the prescription surveillance system produced a good approximation of activity patterns.
Conclusions: Our prescription surveillance system presents great potential for monitoring influenza activity and for providing
early detection of infectious disease outbreaks.
(J Med Internet Res 2012;14(1):e14)   
KEYWORDS
Surveillance; influenza; real-time surveillance; prescriptions; pharmacy; anti-influenza virus; automatic surveillance; earlyresponse J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.1 (page number not for citation purposes) Web-based surveillance, tracks the rate of influenza using querylogs In addition to monitoring disease activities, syndromic In Japan, the official sentinel surveillance reports the number surveillance helps monitor bioterrorism-related disease or of influenza patients per health care provider after collecting health consequences of natural events ].
information from approximately 5000 clinics and hospitals. The Real-time information related to influenza activity is intensity of influenza activity is assessed according to the number of influenza patients per clinic or hospital. Influenza is countermeasures against a sudden increase of influenza activity.
regarded as highly active if the ratio exceeds 1. In 2009, the Therefore, daily updates of influenza activity are indispensable number of patients per clinic or hospital approached 1 in the for improved understanding and control of an influenza 32nd week, earlier than in any of the preceding 10 years, mainly epidemic. We developed an automatic real-time prescription because of the influenza pandemic A (H1N1), which started in surveillance system with the collaboration of EM Systems Co.
April 2009 Accordingly, the vast majority of the reported Ltd. (Tokyo, Japan) to provide timely information related to a cases were H1N1 novel influenza []. The number of influenza disease outbreak. We applied the surveillance system to monitor patients per health care provider declined below 1 in the 13th influenza activity during the 2009-2010 and 2010-2011 week of 2010. The total number of weeks during which influenza seasons to examine the magnitude and trajectory of influenza was highly active was 29, a longer active period than an outbreak more closely and to share that information with in any of the prior 10 years. In 2010, the reported number of public health authorities, as well as participating pharmacies.
influenza patients per clinic or hospital exceeded 1 in the 50thweek []; a second peak week was detected in March 2011.
We used prescription drug purchase data for surveillance of Because of these irregular patterns of influenza activity, it is influenza activity for three reasons. First, prescribing necessary that both policy makers and clinicians follow anti-influenza drugs such as oseltamivir or zanamivir is a influenza activity closely to implement effective control of an common clinical practice for diagnosed influenza cases in Japan.
influenza outbreak throughout the year.
Japan has the highest annual level of oseltamivir usage in theworld [Therefore, prescription drugs can serve as a good Syndromic surveillance is a useful tool for seasonal influenza indicator of the overall number of influenza patients. Physicians monitoring []. In Japan, the official sentinel surveillance of often perform rapid influenza diagnostic tests on patients who infectious diseases is implemented by the National Institute of have a fever or report other influenza-like symptoms. If the test Infectious Diseases. It reports the estimated number of influenza result is positive or, alternatively, if the physician clinically patients weekly as the Infectious Diseases Weekly Report diagnoses influenza even when the test result is negative, then The official sentinel surveillance collects the number of anti-influenza drugs are often prescribed. This contrasts to influenza cases from approximately 5000 hospitals and clinics practices in some other developed countries, where all over the country and then estimates the number of influenza anti-influenza drugs are recommended for those who are at high patients based on the reported cases []. The entire process of risk [or who have severe conditions from influenza collecting information from health care providers, estimating infections []. In such circumstances, surveillance of the number of clinical influenza cases, and reporting them to prescriptions of anti-influenza drugs would trace influenza the public usually takes 7-10 days. Furthermore, the cases are reported by health care providers as a weekly aggregate number.
Some diseases spread rapidly, and the weekly aggregates might Second, many pharmacies have adopted the electronic not provide sufficiently detailed information reflecting the prescription record system (EPRS), which enables automatic, complete character of disease activity. In addition, the official continuous, and constant information collection, and real-time sentinel surveillance updates influenza activity less frequently analysis of prescriptions and patients. In Japan, the utilization during major holidays. In Japan, seasonal influenza activity rate of the EPRS among pharmacies was 99.0% in 2009 ].
usually starts to become active during the New Year holidays.
Japan also has a high rate of outpatient or office-based clinician Constant monitoring and reporting of activity during that period visits in cases where people feel ill [partly because of the universal health insurance system. Therefore, one might inferthat the number of influenza patients collected through the EPRS Syndromic surveillance is in widespread use for monitoring would closely approximate the number of symptomatic influenza diseases, but usage of prescription drug sales as a source of information is fairly limited. In the United States, the mostcommon source of syndromic surveillance reported by health Third, in contrast to the United States or Taiwan ], in Japan officials is emergency department visits (84%), followed by electronic medical record (EMR) systems are not yet well outpatient clinic visits (49%) and over-the-counter medication established. In the United States, surveillance for influenza sales (44%); less than 10% of health departments reported activity is based on data on outpatient visits along with data prescription medications as a source In the context of related to sales of over-the-counter drugs, school absenteeism, influenza, emergency department surveillance is used to monitor and ambulatory care encounters ,Surveillance for the impact of influenza by age For more rapid feedback, influenza activity using the EMR has been intensively discussed the Web recently has become a powerful tool for syndromic and widely applied ]. By contrast, the Survey of Medical surveillance For example, health surveillance using a Institutions by the Ministry of Health, Labour and Welfare in Web-based self-reporting daily questionnaire is applied to Japan showed that the share of health care providers using EMRs monitor influenza activities ]. Google Flu Trends, a J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.2 (page number not for citation purposes) was just over 10% in 2008, or 948 hospitals (10.8% of all service provider, data related to prescriptions from all hospitals) and 12,939 clinics (13.1% of all clinics) [ participating pharmacies were collected and deposited in a singleserver, making the data collection secure, efficient, and nearly We developed the surveillance system to collect the number of cost-free. Medications covered by the surveillance system prescriptions together with patients' characteristics from the included drugs for relief of fever and pain, drugs for common EPRS automatically, to analyze the data simultaneously to colds, antibiotics, and antiviral drugs including anti-influenza estimate the number of influenza cases, and then to provide virus drugs and antivaricella-zoster virus drugs. The current real-time information of influenza activity to health care study specifically addressed prescriptions for anti-influenza providers and policy makers. The system was tested for a limited virus medication. The neuraminidase inhibitors oseltamivir, time at the G8 Summit meeting in Toyako, Hokkaido in July zanamivir, and laninamivir were included, but amantadine was 2008 for 1 month ]. The present report summarizes details excluded because it is not commonly prescribed for influenza of our prescription surveillance system and presents an evaluation of its performance in the first two influenza seasons,those of 2009-2010 and 2010-2011, since the start of the The original prescriptions contain information related to nationwide operation of the system. The evaluation of patients' sociodemographic and social security information, as surveillance performance, particularly outbreak detection well as the health care providers' information. The automatic performance, is challenging and few studies conduct such surveillance system aggregated the number of prescriptions for analyses ]. A study showed that weekly variation in visits each type of drug and provided tabulations by age and by for lower respiratory tract infections approximated the national geography at both national and prefectural levels. The number mortality data for pneumonia and influenza [Similarly, our of influenza patients was then estimated from the aggregated retrospective evaluation analyzed how closely the estimates of number of prescriptions for anti-influenza drugs by adjusting influenza cases followed the trajectory of influenza epidemics the number of prescriptions for anti-influenza drugs with the proportion of participating pharmacies and of prescriptionspurchased through pharmacies. The analysis and estimation were conducted overnight and the report of the analysis wassent automatically at 7:00 AM on the next day to the registered Prescription Surveillance
recipients, including participating pharmacies and public health We started collecting and analyzing data related to prescriptions authorities. In addition, figures showing the number of automatically through the application service provider of the prescriptions for each type of drug and of the estimated number EPRS in April 2009 [As of March 2011, the of patients were created and posted on the website for public number of participating pharmacies was 5275. In the application J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.3 (page number not for citation purposes) Figure 1. Prescription surveillance. Pharmacies A-D use the application service provider's (ASP) medical claims service. All data are stored in a
central database. The surveillance system automatically counts oseltamivir, zanamivir, and laninamivir prescriptions at the data center. The information
is analyzed using multiple regression models. The results are presented as figures and tables and feedback to participating pharmacies as well as public
health authorities.
prefecture was calculated and reported publicly ]. The Performance Evaluation
number of influenza patients in Gifu Prefecture was surveyed We evaluated our surveillance system from two perspectives during November 16-22, 2009 by the local public health for the 2009-2010 and 2010-2011 influenza seasons. First, we authority as a response to the A/H1N1 influenza pandemic. A compared the estimated number of influenza patients with the survey questionnaire asking for the number of influenza patients estimates provided by the official sentinel surveillance []. The who visited health care providers was sent to all hospitals and official sentinel surveillance estimates the number of influenza clinics located within the prefecture (total of 1677 health patients based on the number of influenza patients reported by providers); 1033 providers responded to the survey (response 5000 health care providers, including 3000 pediatricians, in Japan. We chose the evaluation period to include the periodwhen influenza activity was high for the 2009-2010 influenza The Internal Review Board at the National Institute of Infectious season. The epidemiological threshold of seasonal influenza Diseases approved the current study (approval number 57, activity is determined by the number of influenza patients per "Development and application of real-time surveillance system hospital or clinic. If the ratio is equal to or greater than 1 based to monitor syndromic and symptomatic cases using electronic on the official sentinel surveillance, activity is high by the definition that is accepted and widely used throughout Japan []. This corresponds to the period between the 28th week of2009 (the week starting on July 6, 2009) and the 12th week of For the 2009-2010 influenza season, the total number of 2010 (the week starting on March 21, 2010) for the 2009-2010 influenza patients estimated by the prescription surveillance influenza season. For the 2010-2011 season, the performance system between the 28th week of 2009 and the 12th week of was evaluated between the 36th week of 2010 (the week starting 2010 was 9,234,289 ). The largest number of influenza on September 6, 2010) and the 12th week of 2011 (the week patients, 234,519, was reported on November 24, 2009. For the starting on March 21, 2011). Second, for the 2009-2010 2010-2011 influenza season, the number of influenza patients influenza season, we also compared our estimates with the between the 36th week of 2010 and the 12th week of 2011 was number of influenza patients estimated by the Gifu Medical 7,153,437 The largest number of influenza patients, Association, where the total number of influenza patients in the 230,288, was reported on January 24, 2011. The official sentinel J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.4 (page number not for citation purposes) surveillance estimated the total number of patients for the same (95% confidence interval 13,350,000-14,010,000) for the 2010-2011 influenza seasons [indicating that the sentinel 20,460,000-20,860,000) for the 2009-2010 and 13,680,000 estimates were approximately double our estimates.
Table 1. Number of influenza cases estimated by the prescription surveillance, the official sentinel surveillance, and the Gifu Medical Association in
Gifu Prefecture, 2009-2010 and 2010-2011 influenza seasonsa Estimate by the prescription surveillance Estimate by the official sentinel surveillance Adjusted estimation by the survey in Gifu Prefecture a Sources: the official sentinel surveillance []; Kawai et al b Adjusted estimation by the survey in Gifu Prefecture is shown only for the 2009-2010 influenza season because the data are available only for thatyear.
Pearson correlation coefficient (r) of time-series data on The estimated number of influenza cases in the 2009-2010 influenza patients between our estimates and the official sentinel influenza season was also compared with that ascertained from estimate was.992 (P <.001) for the 2009-2010 influenza the survey of the number of influenza patients at all clinics and season, and.972 (P <.001) for the 2010-2011 influenza season hospitals conducted in Gifu Prefecture. The estimated number (see ). A similar analysis was conducted at the from the survey collection in the prefecture based on the prefecture level. The correlation was.950 or greater in 33 prescription surveillance was 127,568, whereas the number of prefectures,.900-.949 in 5 prefectures, and.770-.899 in 8 influenza cases reported by the survey conducted by Gifu prefectures. The correlation was the lowest in Akita Prefecture Medical Association was 132,474. The official sentinel surveillance estimated the number as 277,890.
Figure 2. Number of influenza cases, 2009-2011, estimated by the prescription surveillance and reported by the official sentinel surveillance. The
estimated number of influenza cases by prescription surveillance was calculated based on the number of oseltamivir, zanamivir, and laninamivir
prescriptions adjusted by the proportion of participating pharmacies and extramural dispensing percentage. See text for details. The reported number
by the official sentinel surveillance shows the number of influenza patients per clinic or hospital, calculated with the reported number of influenza
patients from 5000 sentinel clinics and hospitals.
J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.5 (page number not for citation purposes) is covered by the sentinel surveillance, continuous monitoringof influenza activity is necessary to detect outbreaks early in Our analyses showed that the time-series pattern of influenza their course. Our automatic prescription surveillance system activity reported by the prescription surveillance system in the uses the same standard for detection of a disease outbreak and first two influenza seasons was highly correlated with the pattern runs continuously, providing an important complementary role reported by the official sentinel surveillance, showing that in support of existing surveillance systems in Japan.
pharmacy surveillance can be a good indicator of influenza If EMRs were widely kept, then information related to influenza activity in Japan. Although the estimated number of influenza patients could be collected even faster and possibly more patients was double that of the official sentinel surveillance, it accurately. However, the share of health care providers that was close to the estimate by Gifu Prefecture, where the total have adopted the EMR system was slightly above 10%. Under number of influenza patients was collected in a survey.
such circumstances, purchases of anti-influenza drugs can serve The significance of our prescription surveillance is threefold.
as an alternative indicator of influenza activity.
First, the syndromic surveillance system collected, analyzed, Limitations to this study exist. First, the total number of and reported data related to influenza patients simultaneously.
influenza cases was estimated as almost half of the estimate Therefore, clinicians and policy makers were able to obtain the based on the official sentinel surveillance, although it estimated number of influenza patients of the previous day. This approximated estimates based on a survey collecting the total meant that the estimates were available 1 week ahead of those number of influenza cases in Gifu Prefecture. One reason for reported by the official sentinel surveillance, enabling this gap might lie in the choice of health care providers predictions of influenza activity for the immediately following participating in the official sentinel surveillance. The sentinel week. This was particularly important at the outset of a seasonal health care providers have, on average, a larger number of epidemic, when the trajectory of a quickly spreading disease patients than others, potentially resulting in an overestimation would have changed. Though the Google Flu Trends tool, of the overall number of influenza patients. Second, another real-time surveillance, has been shown to perform well anti-influenza drugs are also prescribed for prophylaxis in in the United States and European countries ], the results addition to treatment, which might engender overestimation of may be sensitive to variations in patients' behavior across the total number of influenza cases. However, in Japan the preventive usage of oseltamivir is limited to household members Second, our prescription surveillance was national and observed of influenza patients who are 65 years or older or who are regional variations in influenza activity at the prefecture level, high-risk individuals ]. Third, the prophylactic usage of although the precision of surveillance varied somewhat between anti-influenza drugs for health care providers and for the public prefectures. This provided helpful information to public health was most intensive at the beginning of the H1N1 pandemic services to plan for the allocation of medical, pharmaceutical, outbreak. We did not include those prescriptions in our and human resources for influenza control, shifting limited surveillance data because they were not prescribed through resources to the most affected regions.
health care providers. Fourth, 60% of the prescriptions werepurchased through pharmacies in 2008. The other prescriptions Third, our surveillance runs constantly, maintaining the method were purchased directly through health care providers and were of counting and estimating influenza cases at all times, and thus not included in our surveillance [This is still much higher we were able to obtain the complete trajectory of the influenza than the rate of adoption of the EMR system in hospitals and pandemic in the 2009-2010 season. Initially during the clinics. Fifth, the participation rate of pharmacies is low, pandemic, the law required hospitals and clinics to report all particularly in certain areas. If the number of participating influenza cases, but that practice was terminated on July 24, pharmacies were increased, then estimating influenza cases 2009, after which activity was tracked only by the official would be possible even for smaller geographical units.
Despite these limitations, pharmacy surveillance provided an Our surveillance system also promises great potential for future approximation of the trend of influenza activity in the first two application to the early detection of an infectious disease influenza seasons after the start of its nationwide operation. It outbreak or bioterrorism attack, which could happen potentially provided both clinicians and policy makers with helpful anywhere at any time. When we started operating a prescription real-time information related to influenza activity. Our pharmacy surveillance system in 2009, all other surveillance systems surveillance system has great potential for detection as well as running in Japan covered only specific regions of the country for monitoring of infectious disease outbreaks in the population for practical reasons ]. Furthermore, because influenza and in cases of significant political or cultural events.
outbreaks do not necessarily occur during winter, the time that Acknowledgments
This research was financially supported by a Health and Labour Sciences Research Grant from the Ministry of Health, Labourand Wealth, "Research for practical application of early detection system for health risk," headed by the second author, Dr YasushiOhkusa. Dr Yoko Ibuka received a grant from the Abe Fellowship Program administered by the Social Science Research Council J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.6 (page number not for citation purposes) and the American Council of Learned Societies in cooperation with and with funds provided by the Japan Foundation Center forGlobal Partnership.
Conflicts of Interest
References
1.
Shimada T, Gu Y, Kamiya H, Komiya N, Odaira F, Sunagawa T, et al. Epidemiology of influenza A(H1N1)v virus infectionin Japan, May-June 2009. Euro Surveill 2009 Jun 18;14(24) [Medline: National Institute of Infectious Diseases. 2011. Infectious Diseases Weekly Report: Trend Graph [in Japanese] [accessed 2011-06-30] Buehler JW, Sonricker A, Paladini M, Sope P, Mostashar F. Syndromic surveillance practice in the United States: findingsfrom a survey of state, territorial, and selected local health departments. Adv Dis Surveill 2008;6(3).
Taniguchi K, Hashimoto S, Kawado M, Murakami Y, Izumida M, Ohta A, et al. Overview of infectious disease surveillancesystem in Japan, 1999-2005. J Epidemiol 2007 Dec;17 Suppl:S3-13 ] Olson DR, Heffernan RT, Paladini M, Konty K, Weiss D, Mostashari F. Monitoring the impact of influenza by age:emergency department fever and respiratory complaint surveillance in New York City. PLoS Med 2007 Aug;4(8):e247[] [Medline: Brownstein JS, Freifeld CC, Reis BY, Mandl KD. Surveillance Sans Frontières: Internet-based emerging infectious diseaseintelligence and the HealthMap project. PLoS Med 2008 Jul 8;5(7):e151 ] [doi:] Sugiura H, Ohkusa Y, Akahane M, Sano T, Okabe N, Imamura T. Development of a web-based survey for monitoringdaily health and its application in an epidemiological survey. J Med Internet Res 2011;13(3):e66 [doi:] Ginsberg J, Mohebbi MH, Patel RS, Brammer L, Smolinski MS, Brilliant L. Detecting influenza epidemics using searchengine query data. Nature 2009 Feb 19;457(7232):1012-1014. [doi: ] Buehler JW, Berkelman RL, Hartley DM, Peters CJ. Syndromic surveillance and bioterrorism-related epidemics. EmergInfect Dis 2003 Oct;9(10):1197-1204 ] Elliot AJ, Singh N, Loveridge P, Harcourt S, Smith S, Pnaiser R, et al. Syndromic surveillance to assess the potential publichealth impact of the Icelandic volcanic ash plume across the United Kingdom, April 2010. Euro Surveill 2010;15(23)[] ] Ujike M, Shimabukuro K, Mochizuki K, Obuchi M, Kageyama T, Shirakura M, Working Group for Influenza VirusSurveillance in Japan. Oseltamivir-resistant influenza viruses A (H1N1) during 2007-2009 influenza seasons, Japan. EmergInfect Dis 2010 Jun;16(6):926-935 [doi: ] National Institute for Health and Clinical Excellence. 2009 Feb. Amantadine, Oseltamivir and Zanamivir for the Treatmentof influenza: Review of NICE Technology Appraisal Guidance 58 (NICE Technology Appraisal Guidance 168) URL:[accessed 2011-06-30] Centers for Disease Control and Prevention. 2011 Jan 31. Interim Guidance on the Use of Influenza Antiviral Agents Duringthe 2010-2011 Influenza Season [accessed 2011-06-30][] Harper SA, Bradley JS, Englund JA, File TM, Gravenstein S, Hayden FG, Expert Panel of the Infectious Diseases Societyof America. Seasonal influenza in adults and children--diagnosis, treatment, chemoprophylaxis, and institutional outbreakmanagement: clinical practice guidelines of the Infectious Diseases Society of America. Clin Infect Dis 2009 Apr15;48(8):1003-1032 [doi: ] Sugawara T, Ohkusa Y, Kawanohara H, Taniguchi K, Okabe N. [The real-time pharmacy surveillance and its estimationof patients in 2009 influenza A (H1N1)]. Kansenshogaku Zasshi 2011 Jan;85(1):8-15. [Medline: Ministry of Health, Labour and Welfare, Japan. 2010. Constructing an Appropriate, Stable and Efficient Healthcare InsuranceSystem [in Japanese] [accessed 2011-06-30][] Sugawara T, Ohkusa Y, Kondo M, Honda Y, Okubo I. [Research on choices of people with mild symptoms of commoncold between consulting physicians and taking OTC (over-the-counter) medicine using a hypothetical question method].
Nihon Koshu Eisei Zasshi 2005 Jul;52(7):618-626. [Medline: Wu TS, Shih FY, Yen MY, Wu JS, Lu SW, Chang KC, et al. Establishing a nationwide emergency department-basedsyndromic surveillance system for better public health responses in Taiwan. BMC Public Health 2008;8:18 ] [doi: ] [Medline: Lazarus R, Kleinman K, Dashevsky I, Adams C, Kludt P, DeMaria A, et al. Use of automated ambulatory-care encounterrecords for detection of acute illness clusters, including potential bioterrorism events. Emerg Infect Dis 2002 Aug;8(8):753-760[] ] J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.7 (page number not for citation purposes) Henning KJ. What is syndromic surveillance? MMWR Morb Mortal Wkly Rep 2004 Sep 24;53 Suppl:5-11 [] [Medline: Ohkusa Y, Sugiura H, Sugawara T, Taniguchi K, Okabe N. [Symptoms of outpatients as data for syndromic surveillance].
Kansenshogaku Zasshi 2006 Jul;80(4):366-376. [Medline: South BR, South BR, Chapman WW, Chapman W, Delisle S, Shen S, et al. Optimizing A syndromic surveillance textclassifier for influenza-like illness: Does document source matter? AMIA Annu Symp Proc 2008:692-696. [Medline: Gundlapalli AV, Olson J, Smith SP, Baza M, Hausam RR, Eutropius LJ, et al. Hospital electronic medical record-basedpublic health surveillance system deployed during the 2002 Winter Olympic Games. Am J Infect Control 2007Apr;35(3):163-171. [doi: ] Lewis MD, Pavlin JA, Mansfield JL, O'Brien S, Boomsma LG, Elbert Y, et al. Disease outbreak detection system usingsyndromic data in the greater Washington DC area. Am J Prev Med 2002 Oct;23(3):180-186. ] Ministry of Health, Labour and Welfare, Japan. 2011 Jun 27. Survey of Medical Institutions [accessed 2011-06-30] Ohkusa Y, Yamaguchi R, Sugiura H, Sugawara T, Yoshida M, Shimada C, et al. [2008 G8 Hokkaido Toyako SummitMeeting Syndrome Surveillance]. Kansenshogaku Zasshi 2009 May;83(3):236-244. ] Mandl KD, Overhage JM, Wagner MM, Lober WB, Sebastiani P, Mostashari F, et al. Implementing syndromic surveillance:a practical guide informed by the early experience. J Am Med Inform Assoc 2004 Apr;11(2):141-150 [][doi: ] [Medline: Lazarus R, Kleinman K, Dashevsky I, DeMaria A, Platt R. Using automated medical records for rapid identification ofillness syndromes (syndromic surveillance): the example of lower respiratory infection. BMC Public Health 2001;1:9[] ] Kawai N, Kawade Y, Kobayashi H, Okada S, Higuchi Y, Kawaji H, et al. Analysis of influenza activity during 2009influenza pandemic A (H1N1) using real-time infectious disease surveillance in Gifu prefecture [in Japanese]. SyukanNihon Iji Shinpou 2010;4487:58-64.
Valdivia A, Lopez-Alcalde J, Vicente M, Pichiule M, Ruiz M, Ordobas M. Monitoring influenza activity in Europe withGoogle Flu Trends: comparison with the findings of sentinel physician networks - results for 2009-10. Euro Surveill2010;15(29) [] [Medline: Ohkusa Y, Shigematsu M, Taniguchi K, Okabe N. Experimental surveillance using data on sales of over-the-countermedications--Japan, November 2003-April 2004. MMWR Morb Mortal Wkly Rep 2005 Aug 26;54 Suppl:47-52 [[Medline: Chugai Pharmaceutical Co. Ltd. 2011 Jun 30. Tamiflu 75 [in Japanese] [accessed 2011-06-30] [ Ministry of Health, Labour and Welfare, Japan. 2008. Survey of Medical Care Activities in Public Health Insurance [inJapanese] URL: [accessed 2011-06-30] [] Abbreviations
EMR: electronic medical record
EPRS: electronic prescription record system
Edited by G Eysenbach; submitted 30.06.11; peer-reviewed by D Zeng; comments to author 26.09.11; revised version received25.10.11; accepted 01.11.11; published 16.01.12 Please cite as:Sugawara T, Ohkusa Y, Ibuka Y, Kawanohara H, Taniguchi K, Okabe NReal-time Prescription Surveillance and its Application to Monitoring Seasonal Influenza Activity in JapanJ Med Internet Res 2012;14(1):e14PMID: Tamie Sugawara, Yasushi Ohkusa, Yoko Ibuka, Hirokazu Kawanohara, Kiyosu Taniguchi, Nobuhiko Okabe. Originallypublished in the Journal of Medical Internet Research (http://www.jmir.org), 16.01.2012. This is an open-access article distributedunder the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0/), which permitsunrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.8 (page number not for citation purposes) Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication onhttp://www.jmir.org/, as well as this copyright and license information must be included.
J Med Internet Res 2012 | vol. 14 | iss. 1 | e14 | p.9 (page number not for citation purposes)

Source: http://www.syndromic-surveillance.net/ronbun_download/pdf/Sugawara%20T_2012_J%20Med%20Internet%20Res_1-9.pdf