School Nurse Visit Surveillance System vs ED Visits
School Nurse Visit Surveillance System vs ED Visits
The NYC school nurse syndromic surveillance system detected increases in ILI consistent with positive influenza laboratory isolates and ED syndromic surveillance. School nurse visit signals were not consistent with ED signals. Increases in ILI and fever–flu visits to the school nurse sometimes signaled before and sometimes after ED signals depending on the timing of the start of influenza season. In NYC, influenza epidemic period ED peaks occurred earliest among school-aged children each season regardless of circulating influenza viral type, subtype, or strain. At the beginning of the 2009 pandemic H1N1 influenza outbreak, the school nurse syndromic surveillance system did not detect citywide increases in the ILI syndrome until a week and a half after the ED syndromic surveillance system. However, school nurse visit data were used during the pandemic as part of the criteria for school closure decisions, because it was more informative about illness patterns at individual schools.
Compared with the ED system, the school nurse syndromic surveillance system consistently identified signals in allergy and asthma syndromes during peak allergy season and a number of diarrhea and vomiting signals suggestive of winter gastrointestinal illness. The associations reported here are limited to correlations with the ED system, which serves as the current gold standard in syndromic surveillance in NYC. Therefore, this analysis cannot be used directly to assess the utility or accuracy of the NYC school nurse syndromic surveillance system for early event detection.
The use of school nurse visit data for disease surveillance is appealing for several reasons—data are population-based, represent a large proportion of all public school nurse visits, are confidential, and are available in near real time. However, there are limitations of the data that inhibit the system from detecting outbreaks and monitoring illness among schoolchildren. First, data are not captured throughout the year because of weekends and school breaks. Such gaps limit the ability of school-based data to monitor trends and detect signals when school is not in session. It is interesting that, in 3 out of 5 influenza seasons, a series of signals in ED visits for ILI or fever–flu occurred during winter break when children were not in school. Because positive influenza laboratory isolates did not significantly increase during this time— and, in fact, there are often decreases in overall ED visits among school-aged children during school breaks—these ED signals might not reflect an increase in ILI, but instead further reflect the challenge of understanding health care utilization patterns among school-aged children.
Second, school nurse staffing availability and scheduling can affect data quality, completeness, and timeliness. Inconsistent and delayed reporting by school nurses could result in missed or delayed identification of signals and possible outbreaks. Inconsistent reporting is one possible explanation for the delayed detection of increases in ILI visits to the school nurse during the beginning of the 2009 pandemic H1N1 influenza outbreak. Anecdotal reports suggested that at least some school nurses may have stopped recording visit information during this time because of the large number of ill students. Another possible explanation for this difference between ED and school nurse ILI trends was the heightened awareness of pandemic H1N1 2009 influenza during the beginning of the outbreak, such that parents kept their children out of school. Alternatively, the intense media coverage at the beginning of the outbreak might have encouraged mildly ill or non-ill patients (i.e., "worried well" persons) to seek care at emergency rooms, resulting in a false ED signal of increased influenza activity.
There are further limitations to the use of school nurse visit data for syndromic surveillance that might have affected our ability to make valid comparisons with ED visit data. School nurse visits may provide insight into patterns of illness that are less severe than visits to the ED among children. Unfortunately, this hypothesis cannot be tested because information about the severity of illness is not available from school nurse or ED visit data. Although similar keywords were used in both systems, ED syndrome definitions are based on categorizations of chief complaints that are either free-text or from a drop-down menu, whereas the school nurse syndrome definitions are based on a universally applied predefined list. As such, complaints available to school nurses are based on pediatric categorizations, whereas complaints available from the ED may not be specific to children. These differences in syndrome definitions for diarrhea and vomiting, in addition to the substantial daily variation in the percentage of ED visits for diarrhea and vomiting, limited our ability to make meaningful direct comparisons between ED and school nurse data.
Â
Explore product updates from manufacturers, available now.
Access Now
Information from Industry
Several authors have reported that school absenteeism has been a useful tool for influenza surveillance. Short et al. evaluated an influenza sentinel school monitoring system in Pennsylvania with data from 2009 to 2010. They compared absenteeism and ILI data with statewide reports of laboratory-confirmed influenza and found that trends in ILI correlated significantly with isolate data whereas absenteeism did not. The DOHMH previously evaluated all-cause absenteeism data and found little justification for its use in syndromic surveillance. There are many reasons why students are absent from school that are unrelated to illness and schools are usually not able to collect data about the specific reason for absence. Integrating disease-specific absenteeism data along with illness information from electronic health records could result in a more meaningful analysis of illness patterns in schools.
All-cause or illness-specific school absenteeism data were frequently used during the 2009 pandemic H1N1 influenza outbreak outside NYC, even though the perceived usefulness of the data for monitoring and decision-making was limited. By contrast, in NYC, school nurse visit data were used extensively during the pandemic to obtain a direct measure of ILI at individual schools. School closure has been promoted as a nonpharmaceutical intervention that can mitigate the impact of an influenza pandemic. However, the social and economic costs of such an intervention can be significant. A school nurse syndromic surveillance system similar to that in NYC can provide school-specific, near-real-time data to policymakers that may have to make school closure decisions in the future.
In NYC, the school nurse syndromic surveillance system takes advantage of existing electronic health records with minimal additional effort on the part of school nurses. However, a jurisdiction-wide school-based electronic health record system might be prohibitive for other local or state governments. Public health practitioners without a school-based surveillance system may be able to use school-age specific analyses of ED syndromic surveillance data to monitor illness in schoolchildren.
Â
Explore product updates from manufacturers, available now.
Access Now
Information from Industry
Discussion
The NYC school nurse syndromic surveillance system detected increases in ILI consistent with positive influenza laboratory isolates and ED syndromic surveillance. School nurse visit signals were not consistent with ED signals. Increases in ILI and fever–flu visits to the school nurse sometimes signaled before and sometimes after ED signals depending on the timing of the start of influenza season. In NYC, influenza epidemic period ED peaks occurred earliest among school-aged children each season regardless of circulating influenza viral type, subtype, or strain. At the beginning of the 2009 pandemic H1N1 influenza outbreak, the school nurse syndromic surveillance system did not detect citywide increases in the ILI syndrome until a week and a half after the ED syndromic surveillance system. However, school nurse visit data were used during the pandemic as part of the criteria for school closure decisions, because it was more informative about illness patterns at individual schools.
Compared with the ED system, the school nurse syndromic surveillance system consistently identified signals in allergy and asthma syndromes during peak allergy season and a number of diarrhea and vomiting signals suggestive of winter gastrointestinal illness. The associations reported here are limited to correlations with the ED system, which serves as the current gold standard in syndromic surveillance in NYC. Therefore, this analysis cannot be used directly to assess the utility or accuracy of the NYC school nurse syndromic surveillance system for early event detection.
The use of school nurse visit data for disease surveillance is appealing for several reasons—data are population-based, represent a large proportion of all public school nurse visits, are confidential, and are available in near real time. However, there are limitations of the data that inhibit the system from detecting outbreaks and monitoring illness among schoolchildren. First, data are not captured throughout the year because of weekends and school breaks. Such gaps limit the ability of school-based data to monitor trends and detect signals when school is not in session. It is interesting that, in 3 out of 5 influenza seasons, a series of signals in ED visits for ILI or fever–flu occurred during winter break when children were not in school. Because positive influenza laboratory isolates did not significantly increase during this time— and, in fact, there are often decreases in overall ED visits among school-aged children during school breaks—these ED signals might not reflect an increase in ILI, but instead further reflect the challenge of understanding health care utilization patterns among school-aged children.
Second, school nurse staffing availability and scheduling can affect data quality, completeness, and timeliness. Inconsistent and delayed reporting by school nurses could result in missed or delayed identification of signals and possible outbreaks. Inconsistent reporting is one possible explanation for the delayed detection of increases in ILI visits to the school nurse during the beginning of the 2009 pandemic H1N1 influenza outbreak. Anecdotal reports suggested that at least some school nurses may have stopped recording visit information during this time because of the large number of ill students. Another possible explanation for this difference between ED and school nurse ILI trends was the heightened awareness of pandemic H1N1 2009 influenza during the beginning of the outbreak, such that parents kept their children out of school. Alternatively, the intense media coverage at the beginning of the outbreak might have encouraged mildly ill or non-ill patients (i.e., "worried well" persons) to seek care at emergency rooms, resulting in a false ED signal of increased influenza activity.
There are further limitations to the use of school nurse visit data for syndromic surveillance that might have affected our ability to make valid comparisons with ED visit data. School nurse visits may provide insight into patterns of illness that are less severe than visits to the ED among children. Unfortunately, this hypothesis cannot be tested because information about the severity of illness is not available from school nurse or ED visit data. Although similar keywords were used in both systems, ED syndrome definitions are based on categorizations of chief complaints that are either free-text or from a drop-down menu, whereas the school nurse syndrome definitions are based on a universally applied predefined list. As such, complaints available to school nurses are based on pediatric categorizations, whereas complaints available from the ED may not be specific to children. These differences in syndrome definitions for diarrhea and vomiting, in addition to the substantial daily variation in the percentage of ED visits for diarrhea and vomiting, limited our ability to make meaningful direct comparisons between ED and school nurse data.
Â
Explore product updates from manufacturers, available now.
Access Now
Information from Industry
Several authors have reported that school absenteeism has been a useful tool for influenza surveillance. Short et al. evaluated an influenza sentinel school monitoring system in Pennsylvania with data from 2009 to 2010. They compared absenteeism and ILI data with statewide reports of laboratory-confirmed influenza and found that trends in ILI correlated significantly with isolate data whereas absenteeism did not. The DOHMH previously evaluated all-cause absenteeism data and found little justification for its use in syndromic surveillance. There are many reasons why students are absent from school that are unrelated to illness and schools are usually not able to collect data about the specific reason for absence. Integrating disease-specific absenteeism data along with illness information from electronic health records could result in a more meaningful analysis of illness patterns in schools.
All-cause or illness-specific school absenteeism data were frequently used during the 2009 pandemic H1N1 influenza outbreak outside NYC, even though the perceived usefulness of the data for monitoring and decision-making was limited. By contrast, in NYC, school nurse visit data were used extensively during the pandemic to obtain a direct measure of ILI at individual schools. School closure has been promoted as a nonpharmaceutical intervention that can mitigate the impact of an influenza pandemic. However, the social and economic costs of such an intervention can be significant. A school nurse syndromic surveillance system similar to that in NYC can provide school-specific, near-real-time data to policymakers that may have to make school closure decisions in the future.
In NYC, the school nurse syndromic surveillance system takes advantage of existing electronic health records with minimal additional effort on the part of school nurses. However, a jurisdiction-wide school-based electronic health record system might be prohibitive for other local or state governments. Public health practitioners without a school-based surveillance system may be able to use school-age specific analyses of ED syndromic surveillance data to monitor illness in schoolchildren.
Â
Explore product updates from manufacturers, available now.
Access Now
Information from Industry
Source...