ABSTRACT
Objective To evaluate the timeliness of the three sets of influenza surveillance data (influenza reported cases from Nationwide Notifiable Infectious Diseases Reporting Information System (NIDRIS), influenza-like illness consultation rate (ILI%) and influenza virus positive rate from Chinese Influenza Surveillance Information System) in mainland China. Methods The three sets of influenza surveillance data of North and South China from 2017 to 2018 were compared using peak comparison, cross correlation and Early Aberration Reporting System C3 method. Results The influenza epidemic trends reflected by the three sets of influenza surveillance weekly data from 2017 to 2018 were generally consistent and significantly correlated. However, the three sets of data had different timeliness. From 2017 to 2018, ILI% in the North was not timely at alarming the first epidemic peak, which was 6 weeks and 9 weeks later than influenza cases from NIDRIS and positive rate of influenza virus respectively. While in the South, ILI% was the most sensitive indicator, which was 4 weeks and 7 weeks earlier than influenza cases from NIDRIS and positive rate of influenza virus respectively. However, the three sets of data had little difference in the timeliness of the second epidemic peak both in the North and South. Conclusions The three sets of influenza surveillance data in mainland China could all roughly reflected the epidemic trend of influenza. After comparing the timeliness, a combination of influenza reported cases from NIDRIS together with ILI% and influenza virus positive rate could improve timeliness and accuracy for early warning of influenza.
ABSTRACT
<p><b>OBJECTIVE</b>To identify patterns of hand, foot and mouth disease (HFMD) incidence in China during declining incidence periods of 2008, 2009, and 2010.</p><p><b>METHODS</b>Reported HFMD cases over a period of 25 months were extracted from the National Disease Reporting System (NDRS) and analyzed. An interrupted time series (ITS) technique was used to detect changes in HFMD incidence rates in terms of level and slope between declining incidence periods of the three years.</p><p><b>RESULTS</b>Over 3.58 million HFMD cases younger than 5 years were reported to the NDRS between May 1, 2008, and May 31, 2011. Males comprised 63.4% of the cases. ITS analyses demonstrated a significant increase in incidence rate level (P<0.0001) when comparing the current period with the previous period. There were significant changes in declining slopes when comparing 2010 to 2009, and 2010 to 2008 (all P<0.005), but not 2009 to 2008.</p><p><b>CONCLUSION</b>Incremental changes in incidence rate level during the declining incidence periods of 2009 and 2010 can potentially be attributed to a few factors. The more steeply declining slope in 2010 compared with previous years could be ascribed to the implementation of more effective interventions and preventive strategies in 2010. Further investigation is required to examine this possibility.</p>
Subject(s)
Female , Humans , Male , China , Epidemiology , Hand, Foot and Mouth Disease , Epidemiology , Incidence , Models, Statistical , Population Surveillance , Time FactorsABSTRACT
Objective The purpose of this study was to explore the spatial clustering,specific clustering areas,as well as changing trend of clustering areas of hand-foot-mouth disease (HFMD).Methods Exploratory spatial data analysis (ESDA) was used to conduct spatial statistical analyses for the HFMD using 2008-2011 data at both provincial and county/district levels.Results The Global Moran' s I coefficients appeared to be 0.3336,0.6074,0.3372,0.4620 and 0.4367 for 2008-2011and for the combined 4 years,respectively.The corresponding P-values were 0.002,0.001,0.004,0.001 and 0.001 respectively,when using the Monte Carlo tests with all the P-values less than 0.05.Moran' s Ⅰ coefficients ranged between 0.3 and 0.7,showing the appearance of moderate or higher clustering nature.Based on the results from nationwide analyses on clustering areas at the county/district levels between 2008 and 2011 (Moran' s I=0.5198,P=0.001),it appeared a moderate clustering nature.When local autocorrelation analysis was applied at the provincial level,3 hot spot areas in Beijing,Tianjin and Shanghai cities in 2008；7 hot spot areas in Beijing,Tianjin,Hebei,Shanxi,Shanghai,Jiangsu and Shandong in 2009; four hot spot areas:Beijing,Tianjin,Guangdong and Guangxi; five hot spot areas:Fujian,Jiangxi,Hunan,Guangdong and Guangxi in 2011,were discovered.390 hot-spot counties/districts were found through local autocorrelation analyses using the three-year data of 2008 to 2010.Conclusion Spatial clustering nature of HFMD incidence between 2008 and 2011 in China appeared to be moderate or high,with the clustered areas a north to south shifting trend.However,further investigation was in need to address this changing trend.
ABSTRACT
Based on data related to human brucellosis which was collected from the national notifiable infectious disease reporting system in the 6 provinces(Inner Mongolia,Shanxi,Heilongjiang,Shaanxi,Jilin and Liaoning)of north China from 2004 to 2007,at the county scale.Data would include age and gender standardized mortality ratios(SMRs)while ESDA was including histograms,box plots and box maps,global and local Moran' s I statistics,etc.The global Moran' s I values from 2004 to 2007 were 0.2581,0.4574,0.4457,0.4841,respectively and all with statistically significant differences.Most of local Moran' s I values were significant positive statistically.High-high counties were mainly in the northeast,most of which were pastoral areas,but the farming-pastoral areas and agricultural areas/town had an increasing trend over time.Low-low counties were mainly in the western and southern areas and most of which were agricultural areas/towns.Low-high counties appeared to be rare,mainly around the counties with high incidence,mainly belonged to agricultural areas/towns.The incidence rates of brucellosis in the six provinces of north China had a trend of increase from 2004 to 2007,namely spreading from east to west,from south to north,and from pastoral areas to farming-pastoral areas and agricultural areas/towns.ESDA could be used to develop effective measures for prevention and control of brucellosis.
ABSTRACT
Objective To analyze the results of application on China Infectious Diseases Automated-alert and Response System(CIDARS)and for further improving the system. Methods Amount of signal, proportion of signal responded, time to signal response, manner of signal verification and the outcome of each signal in CIDARS were descriptively analyzed from July 1,2008to June 30, 2010. Results A total of 533 829 signals were generated nationwide on 28 kinds of infectious diseases in the system. 97.13% of the signals had been responded and the median time to response was 1.1 hours. Among them, 2472 signals were generated by the fixed-value detection method which involved 9 kinds of diseases after the preliminary verification, field investigation and laboratory tests. 2202 signals were excluded, and finally 246 cholera cases, 15 plague cases and 9H5N1 cases as well as 39 outbreaks of cholera were confirmed. 531 357 signals were generated by the other method - the 'moving percentile method' which involved 19 kinds of diseases. The average amount of signal per county per week was 1.65, with 6603 signals(1.24%)preliminarily verified as suspected outbreaks and 1594 outbreaks were finally confirmed by further field investigation. For diseases in CIDARS, the proportion of signals related to suspected outbreaks to all triggered signals showed a positive correlation with the proportion of cases related to outbreaks of all the reported cases (r=0.963, P＜0.01). Conclusion The signals of CIDARS were responded timely, and the signal could act as a clue for potential outbreaks, which helped enhancing the ability on outbreaks detection for local public health departments.
ABSTRACT
Objective To analyze the pilot results of both temporal and temporal-spatial models in outbreaks detection in China Infectious Diseases Automated-alert and Response System (CIDARS)to further improve the system. Methods The amount of signal, sensitivity, false alarm rate and time to detection regarding these two models of CIDARS, were analyzed from December 6,2009 to December 5,2010 in 221 pilot counties of 20 provinces. Results The sensitivity of these two models was equal(both 98.15%). However, when comparing to the temporal model, the temporal-spatial model had a 59.86% reduction on the signals(15 702)while the false alarm rate of the temporal-spatial model(0.73%)was lower than the temporal model(1.79%), and the time to detection of the temporal-spatial model(0 day)was also 1 day shorter than the temporal model.Conclusion Comparing to the temporal model, the temporal-spatial model of CIDARS seemed to be better performed on outbreak detection.
ABSTRACT
Objective To analyze the determinants of malaria in Anhui province during the year of 2004-2006.Methods A GIS-based database of malaria at the township scale of Anhui province was established,using remote sensing and spatial analysis technologies.Methods on statistical analysis,principal component analysis combined with logistic regression analysis were synthetically used to analyze the association between malaria and environmental factors.Results Malaria epidemics in Anhui province during 2004-2006 mainly occurred in the northern districts of Huai River,and the epidemics had become more serious yearly.The determinants of malaria at the township scale mainly included factors as temperature,rainfall,normalized difference vegetable index and elevation.If the lowest temperature in a year,which appeared the fourth principal component of the temperature index series,increased by one unit,the probability of incidence of malaria would decrease by 33%.If the total annual rainfall,which was the first principal component of the rainfall index series,increased by one unit,the probability would decrease by 27%.If the elevation increased by 10 meters,the probability would decrease by 2%.However,the relationship between Normalized Difference Vegetable Index (NDVI) and the probability of incidence of malaria was different.If the NDVI had a one unit increase,the probability would increase 3.28 times.Conclusion The northern districts of Huai River during 2004-2006 appeared to be a new spatio-temporal cluster when reemergence of malaria epidemics had occurred in Auhui province since 2000.Terrain and physiognomy,nature and circumstances factors,such as temperature and rainfall had affected the incidence rates of malaria.Our research data from Auhui province would provide some important references to the discovery of main reasons on the reemergence of malaria epidemics since 2000,especially in the central geographic areas of China.
ABSTRACT
<p><b>OBJECTIVE</b>To investigate the application of prospective space-time scan statistic in the early stage of detecting infectious disease outbreaks.</p><p><b>METHODS</b>The prospective space-time scan statistic was tested by mimicking daily prospective analyses of bacillary dysentery data of Chengdu city in 2005 (3212 cases in 102 towns and villages). And the results were compared with that of purely temporal scan statistic.</p><p><b>RESULTS</b>The prospective space-time scan statistic could give specific messages both in spatial and temporal. The results of June indicated that the prospective space-time scan statistic could timely detect the outbreaks that started from the local site, and the early warning message was powerful (P = 0.007). When the merely temporal scan statistic for detecting the outbreak was sent two days later, and the signal was less powerful (P = 0.039).</p><p><b>CONCLUSION</b>The prospective space-time scan statistic could make full use of the spatial and temporal information in infectious disease data and could timely and effectively detect the outbreaks that start from the local sites. The prospective space-time scan statistic could be an important tool for local and national CDC to set up early detection surveillance systems.</p>
Subject(s)
Humans , China , Communicable Disease Control , Methods , Communicable Diseases , Epidemiology , Models, Statistical , Population Surveillance , Methods , Prospective StudiesABSTRACT
Application of Rogerson Spatial Pattern Surveillance Method was discussed in real-time surveillance and early monitoring for infectious diseases. Theory of Rogerson Spatial Pattern Surveillance Method was introduced in detail. Data from the program on 'Monitoring epidemic cerebrospinal meningitis with Rogerson Spatial Pattern Surveillance Method' by the Chinese Center for Disease Control and Prevention provided evidence as how to choose crucial parameters such as K, H, n0 and tau, how to simulate surveillance process and how to interpret results. Rogerson Spatial Pattern Surveillance Method seemed to be able to continuously monitor the changes of infectious diseases. New cases could be added to this model at any moment and this model could provide early warning information in time when the excess of cases occurred. Simultaneously it could provide the time and site of the extra cases. Rogerson Spatial Pattern Surveillance Method could implement dynamic and real-time surveillance for infectious diseases. Simultaneously it could avoid the problem of multiple tests that causing large type I error. It greatly helped in providing reliable information for controlling infectious diseases in early epidemic period. Hence,it had important value of application in early warning on infectious diseases.
Subject(s)
Humans , Communicable Diseases , Epidemiology , Computer Simulation , Data Interpretation, Statistical , Population Surveillance , MethodsABSTRACT
<p><b>OBJECTIVE</b>To understand the underreporting on death cases through web-based reporting system from medical institutions at county level and above as well as to evaluate the quality of death cases reporting through the system.</p><p><b>METHODS</b>A large-scale survey was conducted at 130 medical institutions based on stratified random sampling and to evaluate the underreporting and the quality of death cases reporting from medical institutions through data from survey and reporting system.</p><p><b>RESULTS</b>In 2005, the total reporting rates were 78.25% at the county level and 37.93% at the institutes. Comparing with the results of 2004, these rates were going down slightly. The eligibility rate of reporting was 79.62%, increased when comparing with results of 2004. The percentage of obvious coding errors among deaths reported by county level and above medical institutions was 24.68%. A total of 5226 death cases were recorded from medical (outpatient and inpatient) sources. An average underreporting rate of 33.07% was found at the selected medical institutions. Statistical difference of underreporting rate was not found at medical institutions at different levels.</p><p><b>CONCLUSION</b>Since the initiation of the web-based reporting system of death cases at medical institutes from county level and above, the timeliness of data reporting had been increasing remarkably. The system showed irreplaceable advantages. However, there still existed some problems such as the underreporting of death cases,the poor timeliness of reporting, and the poor accuracy of coding. In the meantime, it was noticed that repetitive work existed among medical institutions due to multi-systems, suggesting that it was necessary to establish a national life registration in China.</p>
Subject(s)
Humans , China , Epidemiology , Death Certificates , Forms and Records Control , Reference Standards , Health Facilities , Reference Standards , Local Government , National Health Programs , Reference Standards , Public Health Informatics , Reference Standards , Quality ControlABSTRACT
<p><b>OBJECTIVE</b>Application of Knox method was discussed in exploring temporal-spatial cluster for infectious diseases.</p><p><b>METHODS</b>The theory of Knox method was introduced in detail. As an example, the temporal-spatial cluster of epidemic cerebrospinal meningitis cases from April to May in 2005 was explored in Beijing and the statistical process of Knox method was clearly demonstrated with related results well interpreted.</p><p><b>RESULTS</b>Only when time and distance threshold values were set as 13 days and 29.67 km, potential temporal-spatial cluster could be detected in epidemic cerebrospinal meningitis cases.</p><p><b>CONCLUSION</b>According to the characteristics of the epidemics, under Knox method, if appropriate threshold values of time and distance were chosen, information on temporal-spatial cluster of infectious diseases could be obtained. According to the records of different time and distance threshold values, the nature of an infectious disease might be explored. Thus Knox method could help us to strengthen the early warning system on infectious disease control strategy.</p>
Subject(s)
Humans , China , Epidemiology , Cluster Analysis , Communicable Diseases , Epidemiology , Disease Outbreaks , Meningitis , Epidemiology , Time FactorsABSTRACT
<p><b>OBJECTIVE</b>To establish a conceptual model of automatic early warning of infectious diseases based on internet reporting surveillance system, with a view to realizing an automated warning system on a daily basis and timely identifying potential outbreaks of infectious diseases.</p><p><b>METHODS</b>The statistic conceptual model was established using historic surveillance data with movable percentile method.</p><p><b>RESULTS</b>Based on the infectious disease surveillance information platform, the conceptual model for early warning was established. The parameter, threshold, and revised sensitivity and specificity of early warning value were changed to realize dynamic alert of infectious diseases on a daily basis.</p><p><b>CONCLUSION</b>The instructive conceptual model of dynamic alert can be used as a validating tool in institutions of infectious disease surveillance in different districts.</p>
Subject(s)
Humans , Communicable Diseases , Diagnosis , Epidemiology , Disease Outbreaks , Information Systems , Internet , Models, Statistical , Population Surveillance , Methods , Sensitivity and Specificity , Time FactorsABSTRACT
<p><b>OBJECTIVE</b>To study the quality of reporting network system on death cases among county and above levels' medical institutions.</p><p><b>METHODS</b>Data on variables related to county reporting rate, unit reporting rate, timeliness of reporting, eligibility rate of reporting, auditing rate, timeliness of auditing, eligibility rate of auditing, percentage of reporting deaths of medical institutes to deaths among total population, percentage of reporting deaths of county and above levels' medical institutes to deaths among estimated deaths at these institutes were collected and distribution of common coding errors was applied to the assessment of reporting deaths.</p><p><b>RESULTS</b>The total reporting rates were: 82.58% at the county level, 42.79% at the units with auditing rate as 96.96%. The eligibility rate of reporting was 69.10% with eligibility rate of auditing as 73.58%. The percentage of reporting deaths from medical institutes to deaths among total population was 8.91%, and the percentage of reporting deaths of county and above levels' medical institutes to deaths among estimated deaths of these institutes was 30.76%. The percentage of obvious coding errors among deaths reported by county and above levels' medical institutes was as high as 22.87%.</p><p><b>CONCLUSION</b>Network reporting system of death cases among county and above levels' medical institutes had remarkably increased the timeliness of data reporting system. Network reporting of data on death was the best opportunity to expand the coverage and to improve the quality of data reporting. Based on network reporting of death cases among county and above levels' medical institutes as well as deaths accrued at the communities should also be reported via this network in the eligible areas. The quality of coding on death causes among medical institutes were commonly poor, indicating that the training on ascertainment and coding of underlying death causes were quite essential.</p>