Applying Online Training to Optimize Mobilization for Public Safety and Readiness

Concurrent Session 10

Brief Abstract

A pilot study shows advantages of applying online training to optimize mobilization for public safety by analyzing synchronous and asynchronous training outcomes.   With partnership between academia and government, we conduct a point of distribution exercise testing equivalency and effectiveness of training outcomes and processing rates following a public health emergency.  

Extended Abstract

We structured a training activity to validate effectiveness of asynchronous training compared to synchronous training for a health response to disaster using a novel a partnership among academia and local government to engage first responders while teaching university students.  Prior to 11 September 2001 or 9/11, much of the public safety readiness responsibility was limited to small groups of government officials.  Today, public safety is more widely managed and government must mobilize resources from a wider audience to succeed with the myriad of challenges.  For this training and research effort, we developed a strategic partnership among state and local government and Purdue University that provides an improved environment for learning and for public health and safety. 

By using an exercise deployment Strategic National Stockpile (SNS) in a Point of Distribution (POD) exercise, our collaboration efforts between government and the university provide benefits and opportunities to each.  Simultaneously, we tested a full-scale POD mass prophylaxis response to an anthrax attack by teaching and training university students who also gain valuable internship-like experiences.  The ongoing relationship between government and the university’s student talent can benefit all in developing paths for future research and data analyses expected of academia and of improving public safety and responsiveness of government.   We developed this research seeking an optimal resource allocation and training procedures for mobilizing government partners for a public safety health emergency such as a POD.

Previous work for preparing for a POD included motivation for developing the government and university partnership, developing a community sample for representative testing of the POD, speed of processing community members to better allocate staffing and allocating resources, and reduction of medication errors distributed by the POD.   The processing rate data led to the development of computer models to better plan on the needs of specific volunteer stations in the POD like reception, collection of medical information, distribution of medications, and collecting processing data and medical errors during a POD implemented by local and county government for a county size of approximately 250,000 individuals. 

Our data included the time needed by POD volunteers to process a representative family seeking medications for the family unit.   The POD volunteers were trained by the county emergency health officer either by video recorded training or by training provided in a classroom.   Both training methods were provided immediately before the POD was administered.   This research study was created to explore the possible optimization of staff time for the emergency health officer since this position in most small and medium sized counties occupied by   a single staff person who has a myriad of tasks to accomplish during a health emergency such as managing the emergency, acquisition of logistics, securing volunteers, providing direction and control, coordinating with local government and providing impact to state government.   The training, if conducted with similar effectiveness between synchronous and asynchronous procedures, could allow the emergency health officer to record training to better allocate time during the emergency.    The asynchronous training also has advantages in being immediately available to volunteers upon arrival, being highly reproducible in content, and minimizing the time needed by more highly trained individuals during the emergency.    The number of volunteer necessary also can be reduced by ensuring a minimal wait for training thereby, improving the use of volunteers.   Using every resource effectively or optimally is vitally important when responding to natural and technology disasters.  This optimization should help to better utilize volunteers, reduce costs, and speed assistance to those needing help in a time of need.

Either training regardless of method required approximately 30 minutes to delivery to the POD volunteer staff.   The POC was assembled in a suitable space similar in form and function as explained in the training.   Each volunteer staff received an orientation to the POD and we documented the locations of individuals to track processing rates and effectiveness based on the training received.    Data collection researchers were provided at one volunteer per student. The researchers were provided with a stopwatch and they timed from beginning of the exercise until they had completed the dispensing station. At the same time, they would rate the effectiveness of each participant that received the training on a scale from best outcome to worst outcome. The four categories were screener correct and dispenser correct, screener correct and dispenser incorrect, screener incorrect and dispenser incorrect, and screener incorrect and dispenser correct decision.

Our POD exercise produced 159 data points for family distribution of medications.  Family sizes were distributed from 1 to 5 members of various ages that represented the county’s 2010 U.S. Census.   Each dataset was separated by group, player, health background screening station and medication dispensing, total time, training type used for volunteers, and medication count. Each of these data helped to ascertain which was superior, synchronous or asynchronous training. Medication count also was very important because each volunteer could be responsible for up to five patients in their family.          

In addition to processing time, dispensing the proper medication from the four available antibiotics was important.  To access proper medication, two binary decisions were addressed:  proper screening medical recommendation and proper dispensing outcome.    Four decisions pairings are possible: (a) the best outcome is proper screener decision followed by proper dispenser decision. This outcome was rated the best of the four, because both volunteers correctly evaluated the patient; (b)the next level was when the screener made the right decision, but the prescriber made the wrong one. This was label as an incorrect decision due to the prescription being wrong for the patient, but was not the worst of the outcomes; (c) The third level is when the screener made the wrong decision, but the prescriber made the wrong decision. This level was labeled as the worst because both volunteers failed to accurately assess the patient; (d) the final level was when the screener made the wrong decision, but the prescriber made the right decision. This outcome was rated under best because even though the patient received the correct medication, it was due to the prescriber catching the error the screener made.

There were similar processing times observed regardless of the volunteer training method, synchronous and asynchronous.  The final counts of these four varying levels were recorded and totaled for each type of training to measure how they compared.  Our data compiled final counts of decisions observed, asynchronous had 87 correct decisions and a total of 11 incorrect decisions. This gives us an 11.22% error rate for the asynchronous training.  If we then look at the synchronous training, which had 91 correct decisions and 11 incorrect decisions, the error rate is 10.78%. This leaves us with a .44% difference in the error rates based on each type of training.  No statistical difference was found in effectiveness and retention based on the type of training received.

Though this was a pilot study, it helps to show the possible advantages of applying online training to optimize mobilization for public safety and readiness.   Since we have developed a partnership between academia and local government that has modest resource needs, we can repeat the training process each semester to increase our understanding of the POD operation needed.   This study provides initial understanding of the potential for asynchronous or online training.  The process rates provide insight into the proper level of mobilization of volunteers and training to operate a POD.     The error rates, while high, are being addressed in subsequent exercises to improve understanding of the type and amount of volunteer training needed to effectively perform during a POD.    

The results analyzed showed savings in money and time that could help save lives by allowing the emergency planning experts to focus on other duties of importance or urgency.  This type of training would save not only the trainer’s time, but the volunteer’s time who normally would arrive and wait on training.  Use of online training means volunteers could be trained as they arrive and learn at their own pace, and spend more time on materials they are unfamiliar with to increase their proficiency and efficiency.

Prior to 9/11, the federal government determined the need for the Strategic National Stockpile (SNS) to be used in case of any public health emergency and we have shown the importance of developing training to accompany the medical supplies.  Our partnering efforts among local government and the universities provide mutual benefits to sustain the exploration.  Both government and academia benefit from this partnership.  Our data and repetition in exercise conduct provide an opportunity for testing various e allocation of   duties based on volunteer availability and population of the community.

Future research may offer more accurate reviews of optimal resource allocation, improve data and resource predictions, and further aid in developing and evaluating modeling of the POD with AnyLogic® as a means of testing optimal levels of staffing.