HANDI Using a Mobile App for Mass Intervention Data Collection

State: CO Type: Model Practice Year: 2012

Denver Public Health (DPH) has developed the Hand-held Automated Notification for Drugs and Immunizations (HANDI) application as a data collection tool for mass immunization and prophylaxis events. HANDI facilitates data collection by healthcare workers; functioning as a team to register individuals, monitor contraindications and track prophylaxis/immunizations administered, compiled information is forwarded to standardized data repositories. Using commonly available mobile devices (e.g., iPhone, iPod touch), HANDI demonstrates the utility of mobile technology in public health and emergency preparedness. After many years of emergency response planning and experience conducting mass H1N1 immunization clinics during the 2009/2010 flu season, DPH recognized the need for an improved mass intervention process. Mandated to serve the City and County of Denver’s population of 600,158, any reduction in time required to administer vaccine or chemoprophylaxis could potentially go far in protecting the public’s health. From prior experiences, two time consuming steps were patient registration and subsequent data entry. A mobile application addresses both issues by utilizing scanning technology to capture patient data and collecting standardized data for direct database and registry entry. Rapid electronic data storage allows for real-time data analysis (e.g., total people served, demographic characteristics, risk factors and geo-locatable information). The primary objective of the project is to support efficient public health immunization and prophylaxis activities through rapid collection and transfer of standardized data. To achieve this objective, DPH invested approximately $83,000 and developed HANDI, a software application (“app”) that resides on a mobile device (i.e., iPhone and iPod touch). During a mass intervention, event specific information is downloaded onto the device. There are three different workflows or “stations” which may be combined or separated based on personnel and throughput for patients during the intervention process: Station 1 – REGISTRATION: the healthcare worker scans each patient’s driver’s license (using magnetic stripe or 2-D barcode) and collects additional data such as phone number, race, and health insurance information. A barcode with the patient’s unique identifier is printed and given to the patient for presentation at the next stations. Station 2 – INTERVIEW/CONTRAINDICATION: the patient’s barcode is scanned, and the patient is asked customizable health questions specific to the intervention. Station 3 – DOCUMENTATION: the patient’s barcode is scanned, and intervention specific data (e.g., manufacturer, lot number) are recorded, the dosage and site are determined if applicable, and the vaccine or drug is given. A card with vaccine or drug information is provided to the patient. After the station-specific workflow is completed, data are saved on the device and transferred to the application server either real-time or when a connection is established. The server aggregates the data from each station using the unique id, and the complete patient record is transferred to designated databases or registries (e.g., state immunization registry). In collaboration with the Larimer County Department of Public Health and Environment, HANDI interfaces with Larimer County’s Public Health Event Web Registration (PHEWR) product. Patients can use PHEWR to register online in advance of an event, and PHEWR prints a barcoded ticket for scanning by HANDI. Registering online allows the patient to skip the demographic data collection step in the HANDI application. HANDI was first tested during the Colorado State University (CSU) meningococcal vaccination campaign in November, 2010. DPH partnered with the Larimer County Department of Public Health and Environment and shadowed 42 patients. The average time from when the patient signed consent to when the injection was complete was 5 (sd 2) minutes. HANDI users reported that data entry was easy, straightforward, intuitive, and fast. HANDI was next deployed at the Denver Health mandatory employee flu campaign conducted during the fall of 2011. HANDI was used during five flu clinics to vaccinate 242 employees. The average time from when the employee signed consent to when the injection was complete was 4 (sd 2) minutes. HANDI users strongly agreed that HANDI is a good way to enter data and would use HANDI again to collect data. Given the quick vaccination times, user acceptance and positive feedback, DPH deemed the pilot tests successful and our objectives met. Many lessons were learned including the complexity of mobile device management and security, and the importance of a good relationship with the software vendor. We realized that additional experience and testing with mobile printers and network topologies such as a self contained HANDI network would be advantageous. A HANDI wish list was developed for future iterations including workflow flexibility, more intuitive user interface changes, and a rules engine (e.g., recommend specific vaccines based on age or responses to health questions).
Health Issues In 2009-2010, many health departments were tasked with planning for, distributing and monitoring the results of a major H1N1 vaccine campaign. Tracking on high risk target groups and assuring immunization coverage was difficult given the enormous workload involved with conducting mass immunization clinics and distributing vaccine to healthcare organizations. The goal of these events is to quickly serve a large number of people. To effectively monitor, there must be a rapid method to capture large volumes of data in near real-time. Most health departments found the most time consuming steps to be patient registration (typically the patient completed a paper form) and subsequent manual data entry of demographics, interview question results and vaccine delivery (i.e., vaccine manufacturer, lot number, and injection site). Few health departments were able to keep up with this demand and even fewer were able to rapidly transfer the information to a registry. Data entry was resource intensive, often was backlogged and lacked accuracy. In an era of limited vaccine availability, assuring vaccine coverage to those at highest risk was dependent on rapid analysis. Incomplete data meant limited capacity to assure delivery to specific individuals or target groups. This informatics application helps health departments rapidly register, interview and document the range and monitor the effectiveness of a public health intervention. Mandatory monitoring of 2009 H1N1 vaccine delivery to target risk groups was identified through after action reports as a serious deficit in our DPH response. A paper optical character recognition approach failed miserably as data could not be accurately read from the paper forms. A tool to assist with mass public health intervention (e.g., data collection and reporting) was needed. Consistent with the Cities Readiness Initiative funding deliverables, DPH Public Health Preparedness is responsible for planning, conducting and dispensing mass vaccination and/or chemoprophylaxis to the entire population of the City and County of Denver within 48 hours. A mobile application such as HANDI, adaptable to a variety of clinic formats and environments, would significantly reduce the time necessary to serve the population and considerably enhance capacity to monitor these efforts to protect the public’s health. The HANDI project defined three separate and sequential steps to vaccine administration: registration, interview and documentation. While separate workflows, they may be integrated depending on staffing and patient load or throughput. The business processes of mass vaccination are clearly identified and efficiencies maximized by using innovative technologies to replace paper ridden processes. Scanning (e.g., magnetic stripe and 2-D barcode) of driver’s licenses or state ID cards rapidly enhances the registration process. Integration with web-based pre-registration makes this step even more efficient. Providing a barcode from a portable printer allows the individual to be rapidly identified at the next work station. Children can be easily added to the parent’s demographic record to rapidly register those without a driver’s license or state ID card. Data entry pathways are still available for those without any valid ID card. By creating a hand-held device enabled application, HANDI addresses inefficient mass intervention processes. It improves data collection speed, accuracy and reporting timeliness. Patients may leave the clinic with a card that confirms exactly what vaccine was given and the burden of manually entering data is significantly reduced for public health personnel. Innovation To our knowledge there are no commercial applications that rapidly configure for an intervention campaign, operate in a wireless environment, leverage the smartphone adoption curve, utilize scanning (i.e., magnetic stripe and 2-D barcode) technologies, capture intervention (e.g., contraindication and immunization) data, and provide a summary card for individuals leaving the intervention site. DPH wanted to develop a tool not solely for immunization campaigns. This tool may be extended to other public health applications such as disease and outbreak investigations and/or community outreach activities. Similar types of flexible public health data collection tools, such as HANDI, do not exist. The vaccination use case was chosen as a starting point due to our previous experience with mass immunization clinics and upcoming mandatory flu vaccination campaigns. HANDI differs from the typical process of having patients manually complete a consent form, receive the intervention and data entry personnel entering a significant amount of data following an event. HANDI takes advantage of preexisting standardized driver’s license data, scanning technology, intrinsic standardized data collection features of a mobile device or smartphone and prepares an HL7 message consistent with CDC guidelines for immunization registry consumption. Scanning technologies involve the use of an innovative “sled” that couples with the smart phone to rapidly read the driver’s license or state ID.  
Primary Stakeholders The primary stakeholders in the HANDI project included DPH Public Health Preparedness, Center for Occupational Safety and Health and eHealth Services of Denver Health. Role of Stakeholders/Partners The stakeholders worked closely together to ensure the success of the project. Stakeholders communicated regularly through meetings, conference calls and email to define objectives and requirements and to coordinate clinic logistics. Starting in spring 2011, these communications were conducted as needed to accomplish the goal of utilizing HANDI during the Denver Health employee flu campaign in October, 2011. LHD Role Denver Public Health is responsible for all aspects of this project except for the technical specification and implementation of the software which was developed by a software vendor. DPH provided the resources, funding and project management necessary to bring this project to fruition and use during the Denver Health employee flu campaign in October, 2011. DPH works with community and regional partners to support greater preparedness and response capacity. DPH Public Health Preparedness collaborates on a regular basis with community agencies responsible for emergency response planning (e.g., Denver Environmental Health, Denver Police Department, Denver Health Paramedics, Denver Mayor’s Office of Emergency Management). In response to H1N1 in 2009, planning for mass vaccination clinics required multiple partnerships, and DPH looked to experts in their related fields for guidance. Multi-agency presentations at community hearings were held to provide information and address public concerns. Established relationships with our partners were crucial to the success of the overall H1N1 response. HANDI eases the burden on all groups involved with a mass intervention (e.g., public health, law enforcement, health care providers, administration) by improving patient flow, increasing throughput and streamlining the data collection process. During the H1N1 response DPH was also responsible for allocation and distribution of vaccine. The Advisory Committee on Immunization Practices issued guidelines for vaccine allocation based on defined risk groups. Tracking allocation and inventory proved to be an extremely challenging task as initial plans for vaccine allocation had to be frequently adjusted due to significantly reduced supply. At the time, DPH lacked an efficient way to track distribution to individuals. Tracking could only be accomplished by manually entering thousands of consent forms into a vaccination registry. DPH recognized that a more efficient process was needed to meet the needs of future emerging infectious disease events. HANDI provides a way to electronically track specific interventions (e.g., vaccine or antidotes) and facilitates informed decision making based on accurate, up to date information. Lessons Learned To conduct the Denver Health employee flu clinics, DPH Public Health Preparedness collaborated with other groups within the Denver Health agency including the Center for Occupational Safety and Health, eHealth Services, DPH Epidemiology and Surveillance, DPH Administration, Denver STD/HIV Prevention Training Center, and the Denver Metro Health Clinic. These groups provided clinic oversight, staff and technical assistance. Lessons learned through these collaborations include the importance of involving partners and establishing relationships early in the project. Familiarity with the project objectives and the personnel involved helped the clinics operate smoothly. Timely, clear and consistent messages to the employee groups being vaccinated were also essential as the HANDI vaccination process was different from normal vaccination procedures and required the employees to follow different instructions such as presenting their driver’s license. Just in time training was provided to HANDI users working at the flu clinics, and while all HANDI users reported that HANDI is intuitive and easy to use, we learned that those users without previous hand-held device experience required more time to feel comfortable using HANDI and specifically the Station 1/registration workflow. These users would benefit from more hands on time with HANDI in order to feel confident in a mass clinic setting. Implementation Strategy Once a strategic decision was made to pursue a mobile data collection tool, a software vendor was selected based on previous experience in mobile technology. The locally located vendor benefited development by reducing travel-related costs and providing more on-site meetings and vendor representatives during testing. A comprehensive requirements gathering was conducted which included analysis of existing business processes used during the H1N1 immunization clinics, examination of forms and databases, and stakeholder interviews. Once the requirements gathering was complete, a detailed software specification document was written and reviewed by HANDI team members and the Denver Health eHealth Services Department. The software specification document provided a technical “road map” for the application and guided the development process. The application was then implemented, and reviewed and tested by the project team as functionality was added and milestones were reached. DPH and the software vendor collaborated with the Larimer County Department of Public Health and Environment to design and build an interface between Larimer County’s Public Health Event Web Registration (PHEWR) product and HANDI. Larimer County enhanced PHEWR to generate a unique patient barcode on the registration ticket to be scanned by HANDI which eliminated the need for the patient to visit HANDI’s Station 1. Establishing this relationship with Larimer County provided the opportunity for DPH to attend the CSU mass meningococcal vaccination clinics. HANDI team members first observed the clinic to assess clinic flow and the best method to unobtrusively test HANDI and then HANDI was tested during the second clinic. Piloting HANDI at a mass vaccination event provided valuable feedback and led to several software refinements and improvements. Due to the success of the pilot, HANDI returned to a development phase in order to extend the data model for application in a broader range of public health applications. In late spring of 2011, the opportunity arose to apply HANDI to Denver Health’s employee flu vaccination campaign, and HANDI was updated and used during five clinics in October, 2011. The timeframe for the HANDI project was approximately a year from initial requirements gathering to the pilot test at CSU. Requirements gathering took place during the winter of 2010 followed by the development of the software specification during the spring. Software implementation occurred during the summer of 2010 followed by testing and quality assurance culminating in the CSU pilot test in November, 2010. HANDI then returned to a development phase and was tested again at mass flu clinics in the fall of 2011.  
Process & Outcome To support efficient public health immunization and prophylaxis activities through rapid collection and transfer of standardized data using a mobile application. Objective 1: The performance measures used to evaluate HANDI during the CSU and Denver Health mass vaccination clinics included the number of people served, the length of time required to use HANDI and administer the intervention, the number of successfully scanned driver’s licenses and the number of complete data records. HANDI users also completed an evaluation about using the application. The number of people served was obtained by counting the records in the HANDI database and reconciling that number with the signed consent forms. The length of time needed to complete the vaccination was performed by members of the HANDI team who recorded the time the person presented at the clinic and the time the vaccination was complete. At CSU, this measure was recorded by a HANDI user, and at Denver Health a designated person was assigned this task. Successful driver’s license scans were either noted on the consent form by the HANDI user, or could be identified by incomplete data in the HANDI database. For example, during the Denver Health clinic, address was not a required data field, but if the driver’s license scanned correctly that information would be present and if it did not, the address field would be blank. Data from the mass vaccination clinics at CSU and Denver Health showed that the average time from when the patient signed consent to when the injection was complete was 5 (sd 2) and 4 (sd 2) minutes respectively. HANDI users reported that data entry was easy, straightforward, intuitive, and fast and strongly agreed that HANDI is a good way to enter data and would use HANDI again to collect data. At CSU, 17.2% of the driver’s licenses could not be scanned and at Denver Health the percentage rose slightly to 18.7%. Scanning failure was frequently a result of a damaged license (e.g., bent, scratched or worn license). To overcome this, HANDI provides the option of scanning either the license’s magnetic stripe or 2-D barcode and also allows the data to be entered on the keypad if scanning is unsuccessful. Several licenses and ID cards from states other than Colorado were successfully scanned. These outcomes supported our objective and encouraged us to further develop and enhance HANDI. The evaluation results were reviewed by the HANDI project team comprised of DPH Public Health Preparedness and Informatics Group members. Many lessons were learned including the complexity of mobile device management and security, and the importance of a good relationship with project partners and stakeholders. We realized that additional experience and testing with mobile printers and network topologies such as a self contained HANDI network would be advantageous. A HANDI wish list was developed for future iterations including workflow flexibility, more intuitive user interface changes, and a rules engine (e.g., recommend specific vaccines based on age or responses to health questions). We also intend to expand our scanning capability to include other forms of patient identification such as employee badges and student ID cards. These modifications will be incorporated as the HANDI project moves forward to utilize an expanded data model that can be applied to additional public health use cases.  
Health care reform will require public health to play a significantly different role. Patient activities outside of the clinic have a major impact on health. The public health team will include many outreach workers who will be collecting information to support improved health. Our management team recognizes the value of a hand-held device to more accurately record activities that happen in the community and link the results of that effort with other data sources (e.g., electronic health records). DPH has invested in this hand-held technology and will extend this work to address other public health efforts such as community health workers and their interventions with obese children and their families. DPH continues to explore this paradigm and how we may continue to expand this tool in other public health business processes. Although HANDI was first developed for mass immunization events, development is already underway to expand its utility. DPH is currently completing the second phase of development which addresses a number of issues identified regarding connectivity and flexibility for greater utility and more rapid deployment to several field conditions (e.g., using wireless guest networks, temporary wireless servers and non-wireless environments). HANDI is being modified to incorporate an expanded data model that will accommodate a wide range of public health services beyond immunization and chemoprophylaxis. We are also looking to expand to other platforms including iPad and Android. Through these modifications, it is our hope that HANDI will become a versatile and invaluable tool that can be sustained through grant funding or operational investments in many areas of public health.
Colleague in my LHD|Conference|NACCHO website|NACCHO Connect