CORONAVIRUS (COVID-19) RESOURCE CENTER Read More

Implementation, Enhancement, and Rapid Adoption of a School-Based Online ILI Surveillance Tool

State: TX Type: Promising Practice Year: 2016

Harris County Public Health & Environmental Services (HCPHES) is the health department for Harris County, Texas – the third most populous county in the United States – providing comprehensive public health services to the community through an annual budget of over $80 million and a workforce of over 700 public health professionals. The HCPHES jurisdiction includes approximately 2.2 million people within the unincorporated areas and over 30 other municipalities located in Harris County, Texas (excluding the city of Houston).  Certain services, such as mosquito/vector control, Ryan White/Part A HIV funding and refugee health screening, HCPHES provides for the entirety of the county, ultimately covering over 4 million people. According to the 2010 U.S. Census, the population of Harris County population is younger and more ethnically diverse than either the Texas or U.S. populations. Public Health Issue Tracking flu is no easy task; there are millions of residents, dozens of acute care hospitals, countless physicians’ offices, and many public health collaboration partners. Adding to this complexity, only pediatric fatal flu and novel influenza are notifiable in Texas. The solution for HCPHES was to approach schools for voluntary reporting through an easily accessed web-based portal.  This innovative approach utilized voluntary reports come from nurses affiliated with the local independent school districts in the form of daily absenteeism and absenteeism due to influenza-like illness (ILI) counts for both students and staff. These data, combined with a sample of rapid flu test results from acute care hospitals, provide a way for Epidemiologists at HCPHES to examine trends in ILI activity during the school year. Daily absenteeism counts to HCPHES are made online using a secure, credentialed reporting system called the “Flu Portal.” The Flu Portal was developed using a free, publicly available software platform. This basic software is then custom programmed to conform to a specific set of needs and preferences. Goal:  Build, implement and sustain a system that provides near real-time reporting of Influenza like Illness (ILI) and other mandatory notifiable conditions in Harris County, Texas public schools and hospitals. Objectives: Exceed 60% of independent school districts (ISD) within HCPHES jurisdiction participating in HCPHES secure web-based Flu Portal. Foster working relationships with school nurses and school district health coordinators. At least 50 % of schools reporting ILI data via the HCPHES secured web-based Flu Portal are consistent reporters, defined as reporting at least 3 of 5 days each week on at least 50% of weeks during Flu Season. Expand web-based practice to include reporting of notifiable conditions in an encrypted secured environment. Practice Implemented Since at least 2008, our department followed developments for electronic tracking of Flu cases. The development and implementation of the Flu Portal model practice began in 2009, when our department reached out to the University of Texas School of Public Health Informatics program to help in the development of a simple surveillance software tool.  The initial surveillance software tool was developed in an accelerated fashion in response to the 2009 H1N1 Influenza Pandemic. Data for the Flu season following the 2009 H1N1 Influenza Pandemic was also collected using this system. Immediately thereafter, a robust and flexible system was developed and fully maintained within the department.  The Flu Portal as exists today was implemented in the fall of 2010. Impact As school funding is in part determined by daily attendance counts, comprehensive records are kept regarding absenteeism at the individual school level. Access to these data, however, is often restricted and does not detail why the student was absent.  By leveraging the strong working relationships HCPHES has developed with the school districts in its jurisdiction, the department successfully implemented an electronic reporting system that allows for secure entry of information that can be exported, summarized, and reported in a rapid and efficient manner.  This allows for a more targeted and rapid response in the event of increases in ILI in the community. HCPHES is responsible for disease surveillance activities in 18 of 21 ISDs within Harris County.  The percentage of ISDs participating in the web-based Flu Portal has exceeded 60% each Flu Season, ranging from 67% to 89% with a five-year average of 74%.  The close collaboration between public health and public schools has allowed for a significant percentage of schools that are willing to actively participate in this voluntary reporting effort.  The Flu Portal allows nurses to communicate with HCPHES by submitting information regarding other disease trends that may be occurring.  The functionally of the Flu Portal was later expanded to accept reports of all notifiable conditions  
Problem Statement Influenza/School based Surveillance Influenza, more commonly referred to simply as “the flu,” is a viral respiratory illness that occurs seasonally. The disease spreads via respiratory droplets generated when an infected individual coughs or sneezes. Transmission to others may occur before the infected person even knows that they are ill. The illness can range in severity from mild to very severe, and may result in fatalities. In the United States, the influenza season generally ranges from October to May. Laboratories and healthcare providers in the United States are not required to report to public health departments. As a result, public health departments only have a partial picture of the impact of the disease. In order to begin to estimate influenza activity, voluntary reporting of influenza-like illness (ILI) occurs at some health departments throughout the United States from a variety of sources, including hospitals, private physician’s offices, and public health clinics; HCPHES is one such health department. This passive surveillance activity, accomplished via the reporting of individuals who exhibit a specific clinical syndrome, is a simple form of syndromic surveillance. The Centers for Disease Control and Prevention reports that: Each year, there more than 200,000 hospitalizations related to influenza, including an average of 20,000 children younger than 5 years hospitalized due to flu-related complications; Children with chronic health conditions are at higher risk of developing serious complications (http://www.cdc.gov/flu/protect/children.htm ); School-aged children generally have the highest attack rates during community outbreaks, and serve as an important source of transmission within the community (http://www.cdc.gov/vaccines/pubs/pinkbook/flu.html ). Literature examining the use of school-based influenza surveillance began in the late 1970’s (Peterson et al., 1979).  Initial studies like the Minnesota Department of Health’s study in 1979 examined the use of school-based absenteeism data, discussed the usefulness and practicality of collecting school-based surveillance data and suggested the use of school-based surveillance data for other health departments across the country (Peterson et al., 1979).  However, it was not until 1995 that school-based surveillance was evaluated as a public health measure (Lenaway & Ambler, 1995).  In 1995, the Boulder County Health Department evaluated attributes of their school-based surveillance system such as sensitivity, timeliness and simplicity and found that the use of such a system enabled local health departments to detect and track the extent of influenza outbreaks, was a simpler type of surveillance and tended to follow the trends seen with sentinel flu surveillance data (Lenaway & Ambler, 1995).  It has also been found that school-based epidemic peaks generally occurred earlier than community-wide sentinel peaks and could be used to alert early changes in influenza activity (Zhao et al., 2007, Lenaway & Ambler, 1995).  Later studies noted the limitations of school-based surveillance data which include gaps in the data particularly on weekends and during school holidays (Schmidt et al., 2010).  Another limitation of school-based absenteeism data is that the data are non-specific (Besculides et al., 2005).  Thus, the reason for absence is unknown in most cases.  Some later studies have also argued that school absenteeism data can only detect early outbreaks during mild and high levels of influenza activity and can be difficult to interpret (Faustino et al., 2010).  More recently, school-based influenza surveillance has moved from being paper-based to a web-based system.  Harris County is one of several health departments that are currently using a web-based system to collect such data.  However, current literature analyzing electronic school absenteeism data is limited.  Web-based influenza surveillance systems have been found to lead to stronger partnerships with health departments and schools and allow for rapid collection, dissemination and analysis of school health data (Mann et al., 2011, Tarrant County APC, 2011).  Currently, the Arizona Department of Health Services utilizes a web-based school influenza surveillance system and has demonstrated that such electronic systems have the potential to be expanded to collect other data such as absences due to ILI and considers it to be a “near real-time surveillance” tool for health status within schools (Trujillo et al., 2007).  Yet, web-based reporting does have its limitations such as manual entry on the part of the school, which may lead to data entry errors (Mann et al., 2011).  Similar to paper-based reporting, web-based reporting of absenteeism data still faces difficulties in obtaining reasons for absences in order to obtain absenteeism due to ILI data (Faustino et al., 2010). Academic Success It is well documented that academic success is an indicator of the overall health outcome of an individual.  The Texas Education Agency (TEA) distributes funding to the school districts, based in part on average daily student attendance, a fact that ensures a reliable and timely source of data on absenteeism.  This partnership between HCPHES and the school nurses allows for the collection of near real-time reports to identify increased levels of student and staff absenteeism; in the event of a respiratory outbreak, this system would be very valuable. HCPHES Strategic Planning The HCPHES 2013-2018 Strategic Plan identifies and prioritizes key issues affecting the health of Harris County populations and guides the planning and implementation of HCPHES activities. The Strategic Plan serves as a tool to help HCPHES ensure a healthier Harris County providing a foundation which allows policy priorities to drive organizational structure, program design and resource allocation. Three of the 5 Strategic directions aligned with the Flu Portal practice: Strategic Direction 1: Enhance capacity to assess and proactively respond to current and emerging public health issues in a manner that optimizes population health and achieves health equity. Strategic Direction 4: Strengthen partnerships and develop robust networks that leverage organizational and community resources across stakeholders and advance HCPHES and the public health mission. Strategic Direction 5: Emphasize acquisition and optimal use of tools and technologies that can be effectively deployed to advance our strategic priorities Target Population The goal of the Flu Portal is to present a picture of ILI activity in Harris County residents each Flu Season, utilizing school-age children as a proxy population for the community.  The national average for school-age children is 19%; this age group accounts for 24% of the HCPHES jurisdiction, thus indicating that a significant proportion of the population attends school.  Schools can be thought of as a microcosm of the community at large.  Children congregate in schools and after school activities, facilitating transmission of illness to other students and their families.  While the population being monitored through the Flu Portal is school-age children, the real target population is the community at large.  The population of Harris County excluding Houston approximates 2.2 million people with over 600,000 being between the ages of 5 and 19 years of age.  The Flu Portal monitors ILI activities in the 580,000 students enrolled within the 632 public schools which are part of these 18 ISD. Unlike adult populations, careful records are kept on whether children attend school each day. By partnering with school nurses, we can gain access to not only absenteeism data, but the nurses can collect information on how many students are absent specifically due to ILI. The Flu Portal has proven to be a good tool to monitor this population.  Since its inception, an annual average of 59% of schools reporting to the Flu Portal are considered consistent reporters (reporting 3 of 5 days, at least 50% of weeks during the Flu season). During the 2014-2015 Flu Season, 167 schools from 13 ISD were rated consistent reporters. These 167 consistent reporters were geographically diverse and had a combined enrollment of ~176,000 students, representing 33% of the school-age population for those 13 ISD.  When evaluating the representativeness of the Flu Portal it was found that at the ISD level, median absenteeism was seldom more than one standard deviation away from the county-level median. The mean, median, and standard deviation of student enrollment in the good reporter schools were nearly identical to the average enrollment of Harris County schools, suggesting that these schools are representative of typical Harris County school enrollments. Previous Practice Monitoring of student absenteeism has been carried out previously at HCPHES.  Initially, this was accomplished via faxed paper reporting; school nurses would keep daily/weekly tallies which would be faxed to HCPHES.  This reporting required time-consuming data entry on the part of the LHD.  Historically, HCPHES has depended on direct reports from school nurses and school health coordinators from each independent school district.  These nurses monitor their own schools and school districts and report unusual events to the epidemiologists.  Although a close relationship exists between schools and the health department this method was informal and subject to delays in reporting.  In addition to school reports, the health department depended on the reports of hospitals and healthcare providers who would fax laboratory reports of rapid flu tests and other influenza testing.  This practice still exist and the number of laboratory reports are included with the information now collected via the Flu Portal to have a more robust estimate of Flu in the Harris County area.  It should be mentioned that it is well known that most individuals do not seek medical care for the flu, therefore estimates originating from the healthcare system are lower than the true incidence of illness. Clearly, in a crisis such as the 2009 H1N1 Influenza Pandemic, there’s limited time to gather and analyze data and share data efficiently with all of the stakeholders and partners who might be able to help a LHD extend its reach and effectiveness. For situations such as that, a fully automated, Web-based portal that accommodates rapid collection, analysis and dissemination of school and child health data is useful, and having the system hosted in-house reduces the time required to access the data.  Additionally, it enhances the estimates originating in the healthcare system as stated above. Innovation/Best Practice Our surveillance system was adapted from The Southwest Center for Advanced Public Health Practice located at Tarrant County Public Health, but we developed it in a way that quickly reached the majority of schools in our jurisdiction. In the first year of implementation, nearly half of the schools in our jurisdiction, representing 5 of the 18 ISD, participated in the surveillance program, with over 100 schools reporting consistently throughout flu season. By the second year, 60% of schools from 15 of our 18 ISD participated. Participation has been consistent since then, providing more than sufficient amounts of data on absenteeism and ILI incidence and prevalence in Harris County. Part of the reason for such enthusiastic participation by the school districts is the issuance of weekly reports during flu season to all the participating school nurses and district health coordinators. By getting regular feedback from HCPHES, the participating schools can immediately recognize their part in the county-wide flu picture. The approach at HCPHES is innovative in that the majority of these voluntary reports come from nurses affiliated with the local independent school districts in the form of daily absenteeism and absenteeism due to ILI counts for both students and staff. These data, combined with a sample of rapid flu test results from acute care hospitals, provide a way for Epidemiologists at HCPHES to examine trends in ILI activity during the school year.  
Goal:  Build, implement and sustain a system that provides near real time reporting of Influenza like Illness (ILI) and other mandatory notifiable conditions in Harris County, Texas public schools and hospitals. Objectives: 1.       Exceed 60% of independent school districts (ISD) within HCPHES jurisdiction participating in HCPHES secure web-based Flu Portal. 2.       Foster working relationships with school nurses and school district health coordinators. 3.       At least 50 % of schools reporting ILI data via the HCPHES secured web-based Flu Portal are consistent reporters, defined as reporting at least 3 of 5 days each week on at least 50% of weeks during Flu Season. 4.       Expand web-based practice to include reporting of notifiable conditions in an encrypted secured environment. The approach at HCPHES of using a web-based portal is innovative in that these reports are voluntarily sent from nurses affiliated with the local independent school districts in the form of daily absenteeism and absenteeism due to ILI counts for both students and staff.  These data provide a way for epidemiologists at HCPHES to examine trends in ILI activity during the school year. Those who report daily absenteeism counts to HCPHES do so online using a secure, credentialed reporting system called the Flu Portal. The Flu Portal was developed using freeware, a software platform that is available free of charge, but usually requires a programmer to customize the basic application to conform to a specific set of needs and preferences. The Epidemiology Program at HCPHES spent several months working with a contract programmer to customize the forms needed for ILI data collection from school nurses and hospital infection prevention staff. Each week, an Epidemiologist exports the data into Excel for analysis. The system requires a user account and password for access, and is secure and encrypted. The “story” of the Flu Portal began in 2008, when HCPHES attended a conference and attended a presentation that described an electronic reporting system by The Southwest Center for Advanced Public Health Practice located at Tarrant County Public Health.  The development and implementation of the practice began in earnest during the summer of 2009, when the University of Texas School of Public Health (UTSPH) was contacted regarding development of a syndromic surveillance system. UTSPH was approached due to their use of mobile technology to conduct cot-to-cot surveys in the Astrodome during the 2005 Katrina Hurricane response.  A system was quickly developed that, while based on the Tarrant County model, utilizes a different technological solution.  After the 2009-2010 flu season, planning for an enhanced online ILI surveillance system with additional capabilities and flexibility that would be housed at HCPHES began, including meetings with Tarrant County officials and their technology contractor, Ingen Systems.  During the fall 2010, HCPHES launched the Flu Portal, a web-based system for ILI surveillance, for local ISDs and acute care hospitals.  Two years later, planning for expansion of capabilities to accommodate online reporting of all Texas Notifiable Conditional was initiated.  In the fall of 2013, HCPHES pilot tested this new capability with one local ISD, and launched for all participating ISDs a year later (Fall 2014). Development of the Flu Portal The Flu Portal was developed using freeware, a software platform (DotNetNuke) that is available free of charge, but usually requires a programmer to customize the basic application to conform to a specific set of needs and preferences.  Annual maintenance fees are paid to a contractor in order to handle software updates, security certificates, and troubleshooting and repair of bugs and other system issues.  The application includes a credentialed, secure web portal containing a data entry form where the school nurses enter their daily reports.  The system requires a user account and password for access, and is secure and encrypted.  HCPHES created over 575 user accounts for schools in the Harris County jurisdiction.  Field-level validation was added to ensure integrity of data being entered into data entry form.  Additionally, the portal offers report generating and data export capabilities, thus simplifying the process of data acquisition and analysis.  Security certificates are maintained by the contract programmer and updated as required.  Finally, training materials and other technical documentation was also developed.  Promotion/Relationship with Independent School Districts (ISD) The Epidemiology Program has established excellent working relationships with the public schools within its jurisdiction over the years conducting routine communicable disease surveillance activities.  For this practice, all public schools located in the 18 ISDs with the HCPHES jurisdiction were approached for participation in the Flu Portal.  District level nurse coordinators were approached from each school district to participate in ILI reporting activities.  At the start of the project, HCPHES provided training and promotional materials to the schools and issued login credentials for nurses at each school as well as the district nurse coordinator.  Security certificates were incorporated into the site in subsequent years.  At the beginning of each school year, a recruitment email is sent to the district nurse coordinator for each school district to welcome back existing, and solicit participation from new, school nurses.  As an incentive for participation, the district nurse coordinator is sent a summary report each week of the reporting statistics for each school, as well as overall ILI activity for their school district compared to Harris County (excluding the city of Houston) as a whole. Distribution of Reports: A variety of reports have been developed over time that use the data obtained from the Flu Portal.  A weekly summary report using the ILI data is made available to the general public on the HCPHES Website; this report contains aggregate data from the local ISDs as well as measures of state and national ILI activity.  A separate weekly report, specifically tailored to each ISD, is sent to the district nurse coordinator that compares ILI activity at the ISD level to that of Harris County, as well as reporting metrics for each reporting school.  Weekly summary statistics are also reported to the Texas Department of State Health Services (DSHS) Regional Office. Data Methods At the end of each flu season, daily absenteeism and ILI counts reported for each school are used to calculate weekly means and medians for each CDC Epi Week.  The methodology used for data cleaning and analysis was developed and documented in 2012.  Each year the stated methodology is used.  Refinements to this methodology have been done to incorporate additional analysis and any necessary revisions to the original methodology.  While the data does not provide the total impact of influenza on the residents of Harris County, HCPHES is able to compare the weekly ILI activity with data reported by the school nurses for previous school years to determine if a given week in the current season is comparatively more or less active.  Cleaning is conducted in Excel to ensure validity of the data used in the calculations.  So far, data for 2009-2010 thru 2014-2015 Flu Seasons have been exported from the Flu Portal into Microsoft Excel 2013 and analyzed in STATA 13.0.  Analysis is conducted with data collected through the Flu Portal then overlaid to the Texas ILINet and the ILINet data (national data).  Calculations include median percent absent, mean percent absent due to ILI, and determination of “Good Reporters” – these are defined as those schools who reported daily absenteeism counts for at least three days a week for at least half of the flu season. Stakeholders In order for a system to be sustained over time, stakeholders must continue to be engaged.  A number of internal stakeholders were involved with or benefit from surveillance activities conducted via the Flu Portal.  Staff from the Epidemiology Program, including the Epidemiology Program Manager and others, provided leadership in developing project proposal, coordinated and facilitated the development and implementation of system, provided expertise and guidance in the needs, requirements and functionality of platform use, and maintain and develop relationships with schools and hospitals.  Other internal stakeholders include: the Project and Business Technology (PBT) Office, who assist with technical aspects of practice; the Office of Public Health Preparedness and Response (OPHR), who provided input on system/process to maintain Harris County, Texas awareness of potential public health emergencies; and the Office of Public Information (OPI), who use information included in the weekly reports to respond to media requests regarding flu (only useful for reporting trends in activity). External stakeholders include the Harris County Independent School Districts, who provided input during pilot phase, submit daily ILI data to the Department via the Flu Portal system, and use the weekly summary reports provided to them from HCPHES for situational awareness regarding flu and other related topics in their school districts; each year we invite all schools to participate in the school-based surveillance system, utilizing the Flu Portal.  Acute care hospitals who are within the HCPHES jurisdiction have reporting capabilities via the portal as well.  There are also some technical stakeholders, including the contract programmer who customized the software and the University of Texas, School of Public Health – Informatics who assisted in the development of the initial customized software solution. Start-up costs and budget breakdown of Current Flu Portal Initial expense for contracting services to set up and customize the Flu Portal was $2,800.  Additional cost was incurred by HCPHES for technical requirements including server space, security services and IT staff support time.  HCPHES pays an annual maintenance fee of $4,900.  In order to enhance the system to accept notifiable condition reports additional contract fees of $9,000 we paid to developed the forms and link the Flu Portal to the program database to insert new reports.  Additional a maintenance fee of $4,200 is paid for contract services necessary to support the information exchange aspects of the system.  Throughout the process consultation was provided in-kind by various Epidemiology Program staff. Each year the analysis is done by student interns/volunteers who devote about 100 hours supervised by epidemiology staff.  
Meeting an Important Public Health Need The timing of the implementation of the HCPHES Flu Portal was paramount to its success.  The 2009 H1N1 Influenza Pandemic helped to refocus interest on influenza as an important infectious disease.  For years the public health community reminded the public of the potential of a novel flu virus emergence, and the 2009 H1N1 Influenza Pandemic underlined the seriousness of such an event.  It was under these conditions (the emergence of a novel influenza virus outbreak) that the Flu Portal was introduced to schools in Harris County.  Additionally, the majority of observed morbidity and mortality during the 2009 H1N1 Influenza Pandemic occurred in younger age groups rather than the typical group of adults 65 and older observed during seasonal flu.  These factors may have added to the interest in active participation in ILI surveillance via the Flu Portal.  Product Discussion and Evaluation Daily absenteeism data are entered into the HCPHES Flu Portal by school nurses utilizing a secure web-based system. Each day, nurses are asked to enter the following indicators: total student enrollment for the school, the number of students who are present, total number of students absent and those absent due to ILI, the number of students seen in the clinic (for any reason) and those seen due to ILI, number of faculty and staff employed, and the number of faculty and staff absent and those absent due to ILI.  The Flu Portal initiative is considered a hybrid approach that involves both primary and secondary data collection methods.  Total absenteeism data are routinely collected by schools for other purposes, such as funding, making this a secondary data source.  In this application, the information is entered by the school nurse directly into Flu Portal for reasons other than its originally intended purpose.  Other data points collected via the Flu Portal, such as staff absenteeism and absences due to ILI, are newly collected data points only intended for ILI surveillance.  Additional secondary sources of data include the ILINet data for both Texas and the U.S., which are used for performance measures. As a result of the 2009 H1N1 Influenza Pandemic, HCPHES was able to leverage existing long-term relationships with local public school officials to implement this new initiative.  ILI surveillance is voluntary in Texas. Until this time, all efforts to acquire this information were done manually, utilizing fax machines and telephones to report ILI activity in the school setting.  The Flu Portal made this process more efficient for health department employees, offering a streamlined method for data entry and subsequent analysis.  With the new system in place, data were directly exported from the Flu Portal for analysis, eliminating the need for time-intensive data entry by health department staff.  School nurses reporting ILI data to the health department had long voiced frustration over the need to fax daily reports and readily welcomed the introduction of the flu portal.  The implementation of the Flu Portal not only freed up health department fax machines already overburdened by numerous data reports, it also acted to separate the reporting streams of routine notifiable conditions from ILI reports, effectively “decongesting” the department’s surveillance system.  In a routine Flu season, HCPHES receives 30,000 to 40,000 ILI reports.  Separating the reporting streams minimizes the risk of lost or delayed reports, enhancing the department’s ability to respond in a timely manner to public health emergency reports. Flu Portal Goal The main goal of this initiative was to build, implement and sustain a system that provides near real time reporting of Influenza-Like Illness (ILI) and other notifiable conditions in Harris County, Texas public schools and hospitals.  By building and later enhancing the Flu Portal with the capacity of receiving encrypted individual reports of notifiable conditions, we achieved this goal.  The practice was evaluated using qualitative and quantitative performance measures, including participation outcomes, estimated population representativeness, and comparison of Flu Portal data with ILI data collected at the state and national levels. Objective 1: Exceed 60% of HCPHES jurisdiction public schools participating in HCPHES secure web-based Flu Portal Key Findings: Data analysis revealed that this outcome was met.  The percentage of ISDs participating in the secured web-based Flu Portal has exceeded 60% each Flu Season, ranging from 67% to 89% with a five-year average of 74%, through the continued commitment of the school districts in Harris County.  Harris County excluding Houston is responsible for disease surveillance activities for 18 ISDs.  These include two ISDs with a majority of schools in counties outside of Harris County, therefore its participation in this initiative has been inconsistent, thus affecting our percentages.  The frequency of school districts participating each Flu Season is regularly monitored and reported annually.  Objective 2: Foster working relationships with school nurses and school district health coordinators. Key Findings: HCPHES has a long standing excellent relationship with local school districts.  School nurses are one of the most reliable groups of reporters of notifiable conditions.  HCPHES routinely provides support to school officials when confronted with infectious disease outbreaks.  It bears mentioning here that of 81 institutional gastrointestinal outbreaks reported and investigated by HCPHES between 2005 and 2013 41% occurred in school settings.  The majority of these outbreaks were caused by shigellosis and norovirus infections, common organisms affecting school-age children.  HCPHES works closely with school officials during these investigations to provide mitigation strategies, as well as information for students, parents and staff alike.  The Flu Portal enhanced the already strong relationship with school districts, as it is a daily reminder of the role of public health and the accessibility that schools have to the public health department.  Through the Flu Portal, school nurses can send messages to the epidemiologists by entering a comment in their daily reports.  This comment field is regularly monitored by HCPHES staff and messages are send to epidemiologist for follow up. Objective 3: At least 50% of schools reporting ILI data via the HCPHES secured web-based Flu Portal are consistent reporters, defined as reporting at least 3 of 5 days each week on at least 50% of weeks during Flu Season. Key Findings: Data analysis revealed that this outcome was met.  The percentage of schools reporting consistently has remained high, ranging from 58% to 66% with a five-year average of 62%, through the continued commitment of the school districts in Harris County. Good reporting was defined as schools that reported at least three days for at least 16 weeks out of the Flu Season.  The consistent participation of schools has shown a sustained pattern of participation above 50 percent in the most recent five Flu Seasons. Objective 4: Expand web-based practice to include reporting of notifiable conditions in an encrypted secure environment. Key Findings: In 2013, HCPHES expanded the Flu Portal to include the ability to submit electronic reports for all notifiable conditions in an encrypted secure environment, thus successfully meeting this outcome.  This feature has been rolled out to previously credentialed individuals already reporting flu data such as school nurses.  Other public health partners such as private healthcare providers will be given access to this feature of the Flu Portal in the next phase of the project as permitted by resources. Additional evaluation of the data collected via the Flu Portal includes monitoring trends of flu activity in Harris County excluding Houston, at the following levels: 1) Harris County excluding Houston at large; 2) school district level; and 3) individual school level.  During Flu Season each school district receives a weekly report with cumulative ILI data at the Harris County level, its own ISD level and for each school reporting in their own school district.  School districts are granted access to the data of their own school district and Harris County at large, but they are not provided information on any other school district.  The report also includes information on each school reporting that week and the number of reports received by each school on a given week.  This information allows them to evaluate how consistent their schools are reporting ILI data each week during Flu Season. Data collected via the Flu Portal have been used the compare trends in ILI activity during Flu Seasons to the National and Texas data collected from outpatient settings (ILINet and Texas ILINet).  Although these data originate from different settings, school-base absenteeism data exhibit similar trends, with peaks in student absenteeism preceding the corresponding peak in outpatient data by an average of two weeks.  These data have also been used to compare individual Flu Seasons’ data to each other with the purpose of evaluating trends in ILI activity across time.  
Lessons Learned Several key lessons were learned by HCPHES staff related to the development, implementation and maintenance of this innovative approach to influenza-like-illness reporting, related to a variety of issues.  These are outlined below:  Product Design and Creation First and foremost HCPHES learned a valuable lesson related to building the actual product.  Freeware platforms are a reasonable solution for health departments wishing to develop online reporting capabilities.  However, it is important to know going into such a project that these products are not ready to use “out of the box.”  A programmer is required to initially customize the application according to the specific needs and preference of the department, and while additional functionality can be built out later as necessary, the desired customizations are limited to the constraints of the individual freeware platform being utilized.  Also, while an additional expense, it is beneficial to maintain a maintenance agreement with the contract programmer to ensure that future programmatic needs are able to be accommodated.  The Flu Portal developed by HCPHES was later expanded to include data collection for other notifiable conditions, a practice that provides for rapid notification and initiation of disease investigations.  Having a dedicated contract programmer for ongoing maintenance and upgrades of the flu portal has been useful. Furthermore, having that contractor also be the creator of the system has proven to be quite advantageous. Data Collection The second set of lessons learned relates to generalizability of the collected data.  Using data from the first six influenza seasons confirmed that these data collected from the school nurses provide a good estimate of influenza activity in Harris County.  Our analysis revealed that median estimates were found to be a better measure to use than means, as long as enough data points were available to reliably calculate them. School Partnerships and Collaboration HCPHES learned an important lesson regarding the strength of the relationship and level of commitment that must exist between the health department and school districts providing data to the portal, especially given the voluntary nature of reporting.  Strong, established relationships with school districts and their school nurses is a big asset to disease surveillance.  In order for a voluntary reporting system to be sustainable, it must be considered acceptable to use by those proving the data.  Ease of use, time commitment required, and user “return on investment” must be considered during the development of such a system.  A strong relationship between HCPHES and the reporting school districts was vital to the success of this project.  With so many other competing priorities, only a strong relationship with perceived vested interest could compel an already-overburdened school district to provide such data on a consistent basis to the flu portal.   HCPHES is proud of its strong relationship with local school districts, and applauds the commitment from the school nurses who provide the data.  Sustainability Plans HCPHES plans to continue to use the flu portal for ILI reporting.  This has become a vital means of tracking flu in the community, and providing public access to such information on a weekly basis.  The flu portal also has served to strengthen HCPHES’s already close ties with local school districts. Funding the Portal HCPHES maintains a contracted relationship with the programmer who developed the portal.  This agreement costs the health department $4900 per year, which includes the cost of any needed software upgrades.  No hardships are anticipated in maintaining payment of this annual fee. Staffing and Administrative Support Maintaining the flu portal requires ongoing staffing sufficient to support product administrative needs, and routine data analysis.  The HCPHES Epidemiology Program has integrated all flu portal activities into the basic surveillance operations of the program.  There are no anticipated staffing shortages that would preclude its operation.   In sum, the HCPHES Flu Portal is an integral part of disease reporting and surveillance within the Epidemiology Program and will continue to be funded as such.  It has streamlined the reporting of a disease that is otherwise very difficult and cumbersome to track.   The success of the HCPHES flu portal serves as a model for the entire health department as it seeks to embrace technological advances that help protect public health, both locally and nationwide.    
I am a previous Model Practices applicant|Colleague in my LHD|Conference|E-Mail from NACCHO|NACCHO website