The issue this practice addresses is the coordination of staff during a large-scale public health emergency. One of the central functions of emergency preparedness is to plan for medical countermeasure dispensing, a capability that has long been prioritized by the Centers for Disease Control (CDC). One key component of this function focuses on the use of Points of Dispensing (PODs) to distribute life-saving medications to the public. Public PODs are an efficient method of providing mass prophylaxis but are also resource intensive, requiring a large staff complement. In the event of an aerosolized anthrax attack, for example, jurisdictions may require hundreds or thousands of staff to be identified, assigned, and deployed to different venues to provide medications to those affected. Nor are the challenges associated with emergency staffing restricted to mass prophylaxis operations. As PDPH learned during its mass care responses to Hurricanes Irene (2011) and Sandy (2012), coordinating as few as 50-100 personnel to staff medical field operations is also a formidable challenge. Given the lack of federal guidance on how best to conduct emergency staffing during a public health surge event, in conjunction with anecdotal experiences from other jurisdictions, PDPH posits that staffing during a large-scale emergency is a major, yet largely unrecognized gap.
The impact of this practice is to ensure a more efficient field response, which means saving more lives in the event of a public health emergency.
Philadelphia Department of Public Health
The Unrecognized Gap: Staffing During a Public Health Emergency
1)The Philadelphia Department of Public Health (PDPH) is located in Philadelphia, Pennsylvania.
2)The issue this practice addresses is the coordination of staff during a large-scale public health emergency. One of the central functions of emergency preparedness is to plan for medical countermeasure dispensing, a capability that has long been prioritized by the Centers for Disease Control (CDC). One key component of this function focuses on the use of Points of Dispensing (PODs) to distribute life-saving medications to the public. Public PODs are an efficient method of providing mass prophylaxis but are also resource intensive, requiring a large staff complement. In the event of an aerosolized anthrax attack, for example, jurisdictions may require hundreds or thousands of staff to be identified, assigned, and deployed to different venues to provide medications to those affected. Nor are the challenges associated with emergency staffing restricted to mass prophylaxis operations. As PDPH learned during its mass care responses to Hurricanes Irene (2011) and Sandy (2012), coordinating as few as 50-100 personnel to staff medical field operations is also a formidable challenge. Given the lack of federal guidance on how best to conduct emergency staffing during a public health surge event, in conjunction with anecdotal experiences from other jurisdictions, PDPH posits that staffing during a large-scale emergency is a major, yet largely unrecognized gap.
3) To PDPH’s knowledge, there is no federal or state system or non-proprietary software for assigning staff, and historically local approaches have proven tedious and time consuming, allowing many opportunities for errors. In a public health emergency, the inability to effectively coordinate staff could result in widespread morbidity and mortality. In 2011, the Bioterrorism and Public Health Preparedness Program at PDPH sought to address the complexities of emergency staffing by developing a custom-designed Microsoft Access-based database, which is populated with contact and credentialing information for all PDPH staff and MRC volunteers. In summary, the Emergency Staffing Database and corresponding staffing protocols were designed to address the following objectives:
To enable PDPH to assign large numbers of staff to field operations based on staff responses to alert messages, credentials, and staffing needs;To increase speed, accuracy, and efficiency of the staffing processTo combine all potential staff into one centralized system which contains up-to-date contact and credentialing information.
4) The first iteration of the database was tested in March 2012 as part of a larger functional exercise that simulated a mass medication response. In March 2013, during a full-scale exercise, a second iteration of the database was tested, which addressed user feedback collected after the 2012 test. The database has also been adapted to accommodate mass care responses, a function that will be tested during an exercise in 2014. At this time, no real event has occurred to warrant use of the database; however, it is updated and ready for use at a moment’s notice.
5) With each subsequent test, PDPH has demonstrated continued improvements to the staffing process. During the Hurricane Irene activation, prior to the development of the database, the MRC Coordinator acted as a lone Staffing Specialist, relying on an email and Excel spreadsheet method that was time consuming and tedious. It took 3.3 hours (195 minutes) to assign 22 volunteers to 5 shifts across 3 locations. During the 2012 exercise, which tested the first iterationof the database, the staffing rate increased to 88.2 assignments per Staffing Specialist per hour. During the 2013 exercise, the staffing rate increased again to 115.2 assignments per hour, increasing emergency staffing efficiency by 30.6% compared to the 2012 exercise, and 753% compared to the previous method of email/spreadsheets.
6) The initial objectives have been satisfied, although the database is an evolving tool that will be tested again in March 2014 during a full-scale exercise that utilizes a mass care scenario.
7) Several factors have facilitated the success of this practice. First, the practice recognizes that the burden of coordinating staff during an emergency will likely be on the Bioterrorism and Public Health Preparedness Program at PDPH, not on a Human Resources Department or other entity. Second, the practice was developed with respect to the unique limitations (technological, financial, etc.) of this LHD and functions within those strictures. Finally, the database was designed to dovetail with existing systems, such as the broadcast communication systems that are used to issue alerts and collect responses.
8) The impact of this practice is to ensure a more efficient field response, which means saving more lives in the event of a public health emergency.
1) The City of Philadelphia is a sprawling urban area with a culturally diverse population of 1.5 million residents. Approximately 16%-18% of the population lack health insurance and 28% live in poverty.
2) Efficient coordination of staff during a public health emergency is a major gap documented by this local health department. The Emergency Staffing Database and corresponding protocols developed by PDPH address public health preparedness, and specifically, readiness to facilitate mass medication dispensing/medical field operations in the event of an emergency or public health surge event. If the population of Philadelphia experiences either a naturally occurring disease outbreak that triggers a public health emergency or is exposed to an intentional release of a hazardous bioagent, such as anthrax or smallpox, the appropriate medical countermeasures (pills and/or vaccine) would be mass dispensed to the affected population by PDPH as soon as possible. The disease agents that could potentially trigger a mass medication response are extremely hazardous. The covert aerosol release of anthrax spores, for instance, has the potential to cause mass fatalities as the mortality rate of inhalational anthrax is estimated at 75% (Centers for Disease Control and Prevention). Although much of the work of PDPH’s Bioterrorism and Public Health Preparedness Program is oriented to CDC’s Category A biological agents (e.g., anthrax, smallpox, plague) because these agents have the potential to cause major public health impacts and social disruption, there are many naturally occurring diseases and disasters that PDPH also plans for which could cause similarly deleterious impacts, including pandemic (ranked as the number one public health threat in Philadelphia by a 2012 Hazard Vulnerability Analysis) and severe weather events/natural disasters that result in mass evacuations/displacements. One of the most challenging aspects of any large-scale public health emergency is staffing. For example, assuming a city-wide activation of POD sites with multiple shifts over 48 hours, 6000-8800 response personnel will be required to adequately staff the response. Timing is critical given the virulence and short incubation periods of certain disease agents or the short notice that may precede a severe weather event; therefore, response staff must be alerted, assigned, and deployed within a very short span of time. PDPH draws on both internal personnel, as well as Medical Reserve Corps volunteers, to staff responses and has found that coordinating the deployment of staff to multiple shifts and sites is extremely challenging, especially in a large, populous urban area that could be simultaneously running 50-100 PODs over a period of several days (or weeks) to support a mass medication response.
3) It is very difficult to predict when or how sudden disasters will transpire or who they will affect; therefore, PDPH’s planning is both all-hazards and scalable and is inclusive of the entire population of Philadelphia (1.5 millions residents). Response operations can be scaled up or down depending upon the size of the population affected and their specific needs.
4) As a result of PDPH’s extensive efforts to plan, test, and validate its mass prophylaxis plans, it became clear that the staffing process was a major gap in PDPH’s ability to effectively activate and operate PODs and could potentially derail a response, which could result in widespread illness, death, and public panic. PDPH’s response to Hurricane Irene further highlighted this gap as the MRC Coordinator struggled to assign clinical staff to three Red Cross Shelters using only email and an Excel spreadsheet. This process was extremely tedious and took over 1.5 hours to assign only 22 staff. During a large public health emergency, this method would be untenable. To PDPH’s knowledge, there are no state, federal, or non-proprietary databases or systems that could be used to efficiently facilitate staff and volunteer work assignments during a major public health disaster.
5) PDPH’s current emergency staffing protocol improves upon previous systems because all necessary information (venues, staff roles needed, personnel contact information, credentials, and availability to work, etc.) is centralized in a single database and allows for multiple, simultaneous users, thereby greatly increasing the efficiency and accuracy of the staffing process, as evidenced by the staffing rates displayed in the Staffing Methodologies Comparative Tables. The database also allows for users to pull a variety of reports based on different criteria (e.g., assignments by location, unassigned staff, clinical staff), which facilitates improved quality control and completeness of rosters. Finally, the current system was created to accommodate other pools of staff in addition to PDPH staff and volunteers. For instance, if another City agency were to offer staff to support a public health response, minor tweaks could be made to the messaging process, SAS code, and database design in order to import responses from a new pool of potential staff and make staffing assignments accordingly.
6) To the best of our knowledge, this practice is new to the field of public health preparedness. A literature review was conducted in Medline (1996-October 2012) using various keyword searches, including the following: public health emergency staffing, medical countermeasure dispensing, mass medication and staffing, staffing database, Points of Dispensing and staffing, mass prophylaxis and staffing, mass medication clinic and staff. These searches produced no relevant results. In addition, the Bioterrorism and Public Health Preparedness Program Manager has inquired with state and federal liaisons, as well as other relief agencies, about how other agencies have addressed staffing during an emergency. PDPH has not learned of any comparable approaches to address the deployment of thousands of personnel to mass medication clinics, leading this health department to believe that the prioritization of pharmaceuticals as the crucial factor in a mass medication response has overshadowed any planning around how personnel will be effectively managed. Anecdotal evidence following Hurricane Sandy further supports the supposition that the issue of public health emergency staffing has not been sufficiently evaluated or addressed.
7) Several public health emergency responses that PDPH has either led or participated in were instrumental in documenting gaps in staffing capability and in leading to strategies that best address those gaps. Several responses—Lebanon repatriation at the Philadelphia International Airport in 2006, the H1N1 pandemic in 2009, and Hurricanes Irene (2011) and Sandy (2012)—have clearly demonstrated the need for PDPH to be prepared to deploy highly skilled personnel to respond to public health emergencies with limited notice, for extended periods of time, and under challenging circumstances. In 2011, PDPH’s involvement in the Hurricane Irene response yielded several important lessons learned that were consulted when developing the Emergency Staffing Database and related protocols. The evidence from this response (as documented in the PDPH Hurricane Irene After Action Report) clearly indicated the following:
One person cannot be solely responsible for managing the staffing of any mass care/mass medication response;
Efficiently managing staff assignments requires the integration of key information about response personnel into one system; and
The best system for assigning large numbers of staff within a short period of time is one that allows for multiple, simultaneous users;
Staffing protocols must be succinct, standardized, well-documented, and easy to follow. These basic principles were all considered by PDPH during database development. In addition, improvements were made to the early iterations of the database following PDPH's 2012 functional exercise and again following the 2013 full-scale exercise.
Objectives of the practice are as follows:
To enable PDPH to effectively assign Health Department staff and volunteers to mass prophylaxis operations/medical field clinics based on staff responses to alert messages, credentials, and staffing needs;
To increase speed, accuracy, and efficiency of the staffing process;
To combine all potential staff into one centralized system and to maintain active records containing up-to-date contact and credentialing information;
To develop protocols and best practices that address the demands of a public health emergency response requiring the deployment of thousands of personnel; 5. To expand the functions of the database to enable staffing based on geographic proximity to response sites, staff preferences, date of hire criteria, and other special parameters, as needed.
In order to achieve the objectives, the Bioterrorism and Public Health Preparedness Program at PDPH partnered with the Epidemiology Unit to develop a database that would facilitate the rapid, accurate assignment of staff during an emergency. The current database, which has evolved significantly from the first iteration, is tested regularly and is maintained in a state of readiness through bi-annual updates and field-tests that address any changes to either the system or corresponding protocols. In order to implement the practice, approximately 25 PDPH staff have been trained to use the system, have participated in simulations, and have offered feedback for improvements.
3) Not applicable—during an emergency response, the practice would benefit all affected City residents and visitors.
4) Emergency staffing is a work in progress, so the time frame is ongoing with two scheduled updates per year and at least one annual field test conducted under simulated emergency conditions. The current database has more functionality than the original 2011 prototype, and we anticipate that the 2014 update will demonstrate even greater functionality and flexibility.
5) Stakeholders include staff from other departments within PDPH who are likely to be involved in emergency staffing, as well as all just-in-time emergency response personnel who could be called upon to respond. Staff from other programs at PDPH’s Division of Disease Control who are likely to function as response staff during an emergency have been cross-trained to perform both emergency staffing and communications functions. For example, in 2013, twelve PDPH staff members who were new to the emergency staffing process were trained to operate the Emergency Staffing Database and apply corresponding protocols and procedures. After each training/field-test, participants are surveyed to elicit their feedback, and suggestions for improvement are considered and implemented accordingly. As this project continues to evolve, there is potential for other groups, such as the Philadelphia Medical Reserve Corps and/or other City agencies, to become more actively engaged in the ongoing development of emergency staffing policies and procedures.
5a) Not applicable—this is an internal practice that is not appropriate to share with external community stakeholders.
6) All work on the database was conducted in-house using existing software. Aside from staff time, there were no costs associated with the project.
1) As stated in the methodology section, the objectives of the practice are as follows:
To enable PDPH to effectively assign Health Department staff and volunteers to mass prophylaxis/medical field clinics based on staff responses to alert messages, credentials, and staffing needs;
To increase speed, accuracy, and efficiency of the staffing process;
To combine all potential staff into one centralized system and to maintain active records containing up-to-date contact and credentialing information;
To develop protocols and best practices that address the demands of a public health emergency response requiring the deployment of thousands of personnel;
To expand the functions of the database to enable staffing based on geographic proximity to response sites, staff preferences, date of hire criteria, and other special parameters, as needed.
The database was designed to handle call-down responses from two separate emergency notification systems: Roam Secure Alert Network (RSAN) and SERVPA. RSAN includes all PDPH employees and SERVPA contains all Pennsylvania Medical Reserve Corps (MRC) volunteers. After a call-down is completed, responses are extracted from these two systems and saved as Excel spreadsheets in a shared network location. A member of the Epidemiology unit then uses SAS coding to assimilate the data from both systems into matching categories, simplify longer responses, and automate the data entry into the Emergency Staffing Database.
Once response data are populated into the Emergency Staffing Database, users can open individual records and assign those responders to positions at PODs or other response locations. Data included in the Emergency Staffing Database includes, but is not limited to: name, contact information, medical licensure information if applicable, ability to prescribe medication or administer vaccine, any languages the responder speaks other than English, and any PDPH Leadership training the responder has completed. Staffing Specialists can quickly analyze the responder’s record and assign him/her to a shift date and time, location, and role (leadership, medical, or other). This data then enables the generation of reports that show the full staffing complement at each POD. Once an adequate staffing complement for a shift is completed, as determined by the Staffing Supervisor, a report can be generated that exports the assignment data into an Excel spreadsheet. This spreadsheet is then shared with the Communications Arm for their use in activating response staff through RSAN and SERVPA.
Although emergency staffing will likely always be an evolving system to accommodate different disaster scenarios and changes staff resources, the objectives as stated above have largely been achieved.
2) Since its development, the Emergency Staffing Database has been tested during several exercises and has demonstrated continued improvements to the emergency staffing process. Evaluation of the practice has been robust, including both quantitative measures and qualitative feedback. During both the 2012 and 2013 exercises, data were collected in the database and analyzed later for speed and accuracy of staff assignments. In addition, during both exercises, evaluators who observed the staffing process and participants who played the role of Staffing Specialists provided qualitative feedback. Periodically, staff who have been involved in the development process have examined process outcomes as well. For instance, reviewer feedback from last year's Model Practices application suggested that regular maintenance of the database would prove to be prohibitively time-consuming. The project lead considered this comment and subsequently met with the Communications Specialist and Epidemiology Program Manager (the two staff members who are primarily responsible for updates and maintenance) to assess the time commitment.
As updates to staff and volunteer registries occur regularly to satisfy other program requirements, this first step in the process occurs regardless of database maintenance. To determine the time commitment associated with the second step and the corresponding burden on the Epidemiology Unit, the Epidemiology Program Manager timed an update (which pulls in all contact information from both RSAN and SERVPA using pre-existing SAS code) and determined that the process typically takes under 60 minutes. Hence, it was determined that although the contact registries are likely to expand or include additional pools of response staff, the process of incorporating that information is negligible because it is largely automated and relies on already written SAS code. Project staff continue to evaluate process measures in order to ensure that the procedures that support use of the database are streamlined and do not cause undue burdens.
Outcome evaluation has been largely comprised of quantitative data collected during real events and exercise simulations. During the Hurricane Irene activation, prior to the development of the Emergency Staffing Database, the MRC Coordinator acted as a lone Staffing Specialist. Without the Emergency Staffing Database, assignments relied on an email and manual data entry-Excel spreadsheet method that was time consuming and tedious. It took 3.3 hours (195 minutes) to assign 22 volunteers to 5 shifts across 3 locations; this time frame includes sending the activation messages in addition to making the staff assignments. Because staffing and communications activities were concurrent and performed by one person during the Hurricane Irene response, the time measurement was divided in half for analysis of an hourly staffing rate, yielding an hourly rate of 13.5 assignments per Staffing Specialist per hour for this method.
During the 2012 exercise, the ability of Staffing Specialists to make assignments remained very consistent over time. During the morning session, the assignment rate was 86.8 staff per Staffing Specialist per hour. During the afternoon session, the assignment rate was 89.6 staff per Staffing Specialist per hour. The overall rate for the entire 2012 exercise was 88.2 assignments per Staffing Specialist per hour. During the 2012 exercise, use of the Emergency Staffing Database increased emergency staffing assignment efficiency by 553% compared to the previous method of manual data entry using email and excel spreadsheets.
During the 2013 exercise, staffing occurred at a faster rate than in 2012. Throughout the day, metrics were collected on the following: 1. Total number of PODs staffed; 2. Total number of staff assigned; 3. Hourly rate for Staffing Specialists to staff PODs. Without any special parameters, Staffing Specialists assigned 115.2 staff per Specialist per hour using the 2013 version of the Emergency Staffing Database compared to the previous year’s rate of 88.2 staff assigned per Specialist per hour using the first iteration of the database and the previous rate of 13.5 staff per hour using the manual data entry and email method.
During the morning session (150 minutes), four Staffing Specialists filled roles for Shift 1 at two Mega-PODs, eight Tier 1 PODs in full, and partially for one Tier 1 POD, assigning a total of 896 staff. Once these assignments were made, activation messages were sent to all assigned staff by the Communications Arm. During the afternoon shift (137 minutes), a different set of four Staffing Specialists filled roles for Shift 2 at ten Tier 2 PODs in full, assigning a total of 750 staff. Throughout the exercise, a total of 1,646 PDPH staff and MRC volunteers were assigned to POD roles. Use of the Emergency Staffing Database increased emergency staffing assignment efficiency by 30.6% compared to the 2012 exercise, and 753% compared to the previous method of manual data entry using email and Excel spreadsheets.
Building upon the 2012 exercise, the 2013 exercise challenged Staffing Specialists by adding additional layers of complexity to the staffing process. During the morning session, Staffing Specialists were instructed to assign employees in a hierarchical order based on date-of-hire parameters. During the afternoon session, Staffing Specialists were instructed to assign available personnel to POD locations according to geographic proximity. In addition to speed of assignments, accuracy of role assignments was also analyzed during the 2013 exercise. When assigning staff to PODs, Staffing Specialists must consider staff qualifications and credentials for leadership and medical roles. In a standard public POD, there are five leadership roles (Manager, Assistant Manager, Medical Operations Lead, Line Lead, and Logistics Lead). The prerequisite for filling one of these roles is previous POD Leadership Training (participation is tracked in the Emergency Staffing Database). The Medical Operations Lead must also have a clinical license (e.g., medical, nursing, pharmacist). There are also four additional roles in the POD that mandate a clinical license. All licenses and credentials are recorded in the Emergency Staffing Database.
When assigning staff to the aforementioned roles, Staffing Specialists must consult the database records to match properly qualified staff to these roles. During the 2013 exercise, all leadership and medical staff were assigned appropriately: 106 management positions and 99 medical positions were filled correctly. Of the total staff assigned, 1.5% (25/1642) were inactive or unavailable, meaning that they should not have been assigned. When working with special parameters, results were mixed: 87% of people with geographic priority PODS were assigned to one of them (647/745) and 81.2% of union workers were assigned in the correct order (203/250). Although the error rate increased when special parameters were introduced, accuracy rates were still reasonably high and future trainings for Staffing Specialists could be targeted to focus on the protocols involved in applying special parameters in order to increase accuracy rates.
The major modification that has occurred is the development of a database that supports medical field clinic operations. During Hurricane Sandy, the Emergency Staffing Database was only applicable to POD operations, and PDPH staff again found the process of assigning staff to shelters using Excel spreadsheets and email to be burdensome. Therefore, another version of the database was developed that includes the same pools of response staff but is customized for shelter operations in Philadelphia with designated clinic roles and shelter sites. This database has been field-tested with very good results and will be tested during a full-scale exercise with simulated emergency conditions in Spring 2014.
1) Several key lessons have been learned during simulations. First, multiple tests by multiple users are critical for finding and correcting problem areas. Users have also been instrumental in providing suggestions to streamline and improve the staffing process. Second, documentation must be clear, concise, and accessible to all users. Documentation must also be brief so that users can receive it during a just-in-time training and be prepared to operate the system on short notice with minimal assistance. Third, secondary and tertiary systems must be maintained in the event of technology failures that preclude use of a database. Finally, systems and practices must be tested and updated regularly to ensure that they are working properly and that sufficient staff have had at least some experience using them. An overarching lesson is that a practice should be both flexible and scalable in order to be useful given different scenarios.
2) Not applicable.
3) For Philadelphia, this practice far exceeds previous methods or approaches to emergency staffing in the public health sector. Given the lack of federal and state guidance on the issue, this LHD questions how much work has been done to address the complexities of emergency staffing and would be very interested in hearing about other local approaches, although previous inquiries have yielded sparse dialogue. Given the dearth of information on the issue, PDPH would assert that we have developed a very useful tool that could be useful to other health departments, especially those with large numbers of staff serving comparably large populations.
4) All work on the database was conducted in-house at no cost (aside from staff time) through a successful partnership between the Bioterrorism and Public Health Preparedness Program and the Epidemiology Unit. Projecting the benefits in terms of lives saved is very complex. Use of this tool following an aerosolized anthrax attack would absolutely facilitate rapid opening of PODs, which would likely increase the number of residents who would be prophylaxed before disease onset. Similarly, use of this tool during a weather emergency requiring the operation of medical field clinics at shelters will help to ensure coverage of all shifts, as well as better coordination with response partners, including the City of Philadelphia Managing Director's Office of Emergency Management and the Southeastern Pennsylvania Chapter of the American Red Cross. With no real costs aside from staff time, the benefits far outweigh the cost.
5) At this time, primary stakeholders are internal, and the stakeholders who have been involved in the development and implementation process are committed to both sustaining the practice and making improvements to it as necessary. Locally, stakeholders recognize the importance of ensuring a smooth and efficient staffing process during a public health emergency, as the success of the response hinges upon the ability of PDPH to deploy hundreds or thousands of response staff to many different roles across a variety of locations. A federal framework also exists that would indicate support of local emergency staffing capacity. In March 2011, the Centers for Disease Control issued 15 public health preparedness capabilities to guide local health departments in their strategic planning efforts. Several of these capabilities--Emergency Operations Coordination, Mass Care, Medical Countermeasure Dispensing, Medical Surge, and Volunteer Management--relate directly to a local health department's ability to coordinate staff effectively during an emergency response.
Specific functions and performance measures for each capability have been outlined by CDC and will impact local health departments' annual assessments and perhaps funding opportunities as well. PDPH recognizes that the effective coordination of staff is an important element of any successful response, and given real events and recent federal guidance, emergency staffing is also a cross-cutting capability with broad applications. In addition to continuing scheduled updates and maintenance, PDPH will test the database again during a full-scale exercise in Spring 2014, which will likely yield new areas for improvement and shape revisions and/or the addition of new features. In addition, PDPH is currently exploring pandemic preparedness with local healthcare partners and has proposed a model for a Triage Center that can be co-located on hospital grounds during a severe pandemic to alleviate some of the burden on Emergency Departments. As this model moves forward, a corresponding staffing strategy and database module will be developed. Given these factors, sustainability is certain.
I am a previous Model Practices applicant