Skip Navigation U.S. Department of Health and Human Services
Agency for Healthcare Research Quality
Archive print banner

Adapting Community Call Centers for Crisis Support

Public Health Emergency Preparedness

This resource was part of AHRQ's Public Health Emergency Preparedness program, which was discontinued on June 30, 2011, in a realignment of Federal efforts.

This information is for reference purposes only. It was current when produced and may now be outdated. Archive material is no longer maintained, and some links may not work. Persons with disabilities having difficulty accessing this information should contact us at: Let us know the nature of the problem, the Web address of what you want, and your contact information.

Please go to for current information.

Quarantine/Isolation Support

Significance. Health call centers can assist with monitoring and supporting patients in quarantine and isolation. Several reports are published of managing and monitoring by telephone those placed in quarantine for SARS.2, 21-23 Other research has indicated public support of the use of quarantine to control disease and for monitoring the status of those in quarantine by telephone.24 The Seattle/King County Advanced Practice Center offers a toolkit for "Planning & Managing for Isolation & Quarantine" ( to assist in proactively planning for and managing the implementation of large-scale isolation and quarantine. They suggest establishing a public health hotline and call center for public inquiries and caution:

"In an emergency situation, focus on your expertise. Do not try to reinvent yourself or your organization by taking on something outside of what you know. Admit where your expertise ends and find the community organizations with whom you can partner to acquire their expertise for the response."

Community partners such as health call centers are well suited to assist with monitoring or contacting those in quarantine and isolation, especially if they had appropriate guidance and resources.

Scenarios. Of the 15 National Planning Scenarios, only the two scenarios involving infectious diseases, Pandemic Influenza Outbreak and Plague Outbreak, would potentially require the use of quarantine and isolation as disease control measures and could benefit from using call centers. Though SARS outbreak is not one of the DHS scenarios, SARs and other infectious diseases would require planning and response capabilities similar to those for influenza and plague scenarios.

Current Examples. Telephones were used to monitor those in quarantine in the 2003 SARS outbreaks in Toronto2 and Taiwan.23 Public health departments around the country currently use quarantine and isolation measures with tuberculosis patients, and in Nassau, NY, they use videophones to monitor patient status and compliance.22 The Centers for Disease Control and Prevention (CDC) Division of Global Migration and Quarantine also has field stations at certain airports when suspected infectious persons are identified on airplanes inbound to the United States.29

Technology. Basic telephone technology is required to initiate periodic calls to those assigned to home quarantine by public health agencies. Simply having a staff person place calls on a single phone line can attain this response capability. As larger numbers of people are placed in quarantine, additional personnel and phone lines are needed to manage them. As the number of those in quarantine reaches into the thousands (as in the Toronto and Taiwan experiences), an IR system would be beneficial in automating those periodic calls, recording those who answered and their reported health status, transferring people who needed further attention to staff, and reporting those that did not answer within the specified attempts for followup.

Staffing. Trained information providers could manage quarantine monitoring and provide assistance and referrals according to established public health agency protocols. Agencies like the American Red Cross could be used to help support those in quarantine by delivering food or other supplies.

Proposed Resource. We have developed the Quarantine/Isolation Monitoring Application to support quarantine and isolation monitoring strategies, which would use an IR system. The application would free staff to handle only callers needing further attention, such as those developing symptoms or those who could not be reached (potentially noncompliant with quarantine). In the Taiwan SARS experience, 113,132 people were quarantined with only 133 (0.1 percent) having suspect or probable disease diagnosis. Of the 108 who were tested, only 21 were SARS positive via polymerase chain reaction. Only 286 persons (0.2 percent) were fined for violation of quarantine. This demonstrates that less than 1 percent of those in quarantine needed much more than periodic monitoring. Automating much of that monitoring with a tool such as the QI Monitoring Application could be very useful. This application is summarized here and fully described in Appendix 4, Developing an Interactive Response Tool.

QI Monitoring Application. The QI Monitoring Application has the capability to automatically place outbound calls to individuals in home quarantine/isolation to assess their current health status. The application calls the quarantined person at specified times, provides messages in English or Spanish identifying the purpose of the call, requires the person to select an option that reflects his or her current health status (using temperature as a decision point), transfers him or her to an information provider for assistance if indicated, and provides information about who to call if he or she needs assistance before the next monitoring call.

Though this application was developed for use with an IR system, it could be used without such technology. The call flows, decision trees, and message scripts could guide call center staff in how to handle calls and what information to provide. The application could be adapted to other scenarios that might require telephoning individuals, such as sheltering in place strategies or followups to vaccinations. Public health agencies may want to develop referral protocols and guidance for call center staff in handling situations in which a quarantined individual needs food, medication, or financial assistance.

Outpatient Drug Information/Adverse Event Reporting

Significance. The CDC's Cities Readiness Initiative program requires that participating cities prepare plans for mass prophylaxis with Strategic National Stockpile assets. Depending on the potential for exposure, this could result in thousands or millions of people being dispensed antibiotic medications. Health call centers can support these efforts by providing information about the incident and the supplied medications, as well as collecting any potential adverse event reports.

Scenarios. This response capability primarily is applicable for the two National Planning Scenarios involving agents treated with antibiotics: Aerosolized Anthrax and Plague Outbreak. Other scenarios that involve either mass vaccinations (Pandemic Influenza Outbreak) or wide-scale use of medications for treating radiation exposure (Nuclear Detonation, Radiological Dispersal Devices) or potentially Food Contamination may also require this response capability.

Current Examples. The HELP program has collected adverse event information regarding smallpox using the Vaccine Adverse Event Reporting System (VAERS) form fields ( In addition, drug information centers collect information for the Food and Drug Administration's (FDA) MedWatch reporting program (, as well as handling drug information and identification calls. Poison control centers have much experience in handling drug identification calls, which comprise 61 percent of their non-exposure calls.9

Technology. In addition to the call center technology and data collection systems mentioned previously, staff would require drug identification resources such as online searchable databases such as ( and ( or commercial database IDENTIDEX System ( MedWatch and VAERS form data fields could be incorporated into data collection systems or paper forms could be completed.

Staffing. Trained information providers could manage providing drug identification assistance to the public. Clinicians would be more appropriate for collection of adverse drug or vaccine event reports.

Proposed Resource. We have developed the Drug Information (DI) Application to support mass prophylaxis with antibiotic drug strategies using an IR system. The IR system application would assist callers in the identification of the drug they were given and provide them with information on how to take it and its potential adverse reactions. The application would allow public health agencies to concentrate on operating mass dispensing sites and allow health care providers to care for those who are ill. This application is summarized here and fully described in Appendix 4.

DI Application. The DI Application allows callers to identify medications based on the appearance of the antibiotic drugs that are being dispensed at POD locations during a public health emergency. The callers are given clear directions from menu messages and can repeat messages or drug descriptions. This application accommodates one language selection (English), but it could be modified for additional language selections.

Though this application was developed for use with an IR system, it could be used without such technology. The call flows, decision trees, and message scripts could be used without technology to guide call center staff in how to handle calls and what information to provide. The application could be adapted to other scenarios that might require mass administration of medications or vaccinations and to provide relevant information.

Mental Health Assistance/Referral

Significance. Call centers providing health information and support will help to relieve anxiety and stress among the public, but some callers may need further assistance. Call center staff can assist these callers by referring them to community mental health resources.

Scenarios. Most of the 15 National Planning Scenarios will result in varying degrees of community fear, panic, anxiety, and even depression.

Current Examples. Countless suicide prevention and counseling hotlines currently exist and are run by trained mental health staff. The National Suicide Prevention Lifeline provides a 24-hour toll-free service that routes callers to crisis centers across the country ( Additionally, nurse advice lines are capable of handling patients with depression. Poison control centers regularly receive suicide and intentional harm calls.

Technology. Besides the call center technology and the data collection systems mentioned previously, staff would require protocols and referral resources to access agencies providing counseling and mental health services.

Staffing. Information providers would not necessarily be trained in assessing mental health concerns but could adequately provide referral assistance with training. Clinicians trained in identifying mental health warning signals would be an integral resource for assisting such callers.

Proposed Resources. No specific resources are proposed. However, it is recommended that good risk communication principles be used for handling callers who are anxious or under stress.4 Call center staff should be trained on how to handle calls from those under stress. The Kansas Department of Health and Education developed phone bank operator training that includes communications protocols, techniques, and role playing for a variety of health emergencies.30 The CDC also offers good disaster mental health resources (

It is recommended that call centers review the content of their recordings and FAQs to determine if they can be improved to reduce caller anxiety. Using a voice for recordings that is pleasant and that mentions appropriate reassurances such as "there will be adequate supplies of medications for everyone" will help to alleviate caller anxiety. Following unpleasant information with positive statements also can help in many situations. The Center for Risk Communication ( provides information on communication methods for high concern, high stress, or emotionally charged issues based on behavioral-science research and practice.

4. Develop a mechanism to test and evaluate the model with a local exercise

HELP Model Testing

The HELP model has been tested in more than 3 years of daily operations and response to several major health events. The HELP model has made it possible for us to provide consistent, accurate, and up-to-date information during bioterrorism exercises and public health emergencies in partnership with the Colorado Department of Public Health and Environment. HELP served as a proving ground for the implementation of some HEALTH concepts and strategies. Since its launch, the HELP program has continually developed as it responded to three major health events in Colorado: the deadliest WNV outbreak in the United States (2003), an influenza outbreak with early increased pediatric deaths (2003/2004), and an influenza outbreak during a vaccine shortage (2004/2005).26,28 The HELP program provides a model for disseminating and collecting information that, to date, has involved handling more than 75,000 calls related to several health events and outbreaks.

The first test of the HELP model occurred on the day the program began daily operations to support a WNV outbreak in Colorado. The service began answering telephone calls at 7:00 a.m., and a press release announcing the first human case of WNV in the State occurred within the next few hours. State health department staff handled up to 1,000 calls regarding WNV the previous year (first year of the outbreak in the State) and expected several times that number during the second year of the outbreak because of its greater potential to result in human disease. However, the 12,500 calls over the next 3 months were much more than expected.

During the same period, human WNV cases surpassed 2,500, and human deaths totaled 47. Figure 5 depicts the average call volumes to HELP by hour of the day for the initial 7 weeks. During that time, call volumes averaged greater than 1,500 calls weekly and 220 calls daily, with peak call volumes of 2,229 in 1 week, 524 in 1 day, and 178 in 1 hour. These call volumes accurately reflected the public's demand for information, since all 96 channels (individual phone lines) dedicated to HELP were never all used at once. If that were to occur, additional callers would have received a busy signal, and we would have no means to determine the number of callers that could not get through.

The drivers for these call volumes involved both the status of the outbreak in the news each day (number of human cases and deaths) and the time of day that such news was disseminated by the media. As Figure 5 shows, hourly spikes in call volume were related to times of television newscasts, usually featuring the HELP toll-free phone number as part of the news crawler during updates. Therefore, we began to staff up for certain hours to better accommodate those volume surges. Staffing was usually limited to no more than four information providers at a time (the number that our funding from the State health department could support).

Therefore, we used our initial announcement to relay the most requested information to alleviate a caller's need to speak with staff. On average, 60 percent of callers listened to the recording and then terminated the call; the remaining 40 percent chose to remain on the line to speak with a staff person. This indicated to us that most callers were having their concern addressed with recorded information; otherwise they would have waited in queue to speak with a staff person (those waiting at least 6 seconds past end of recording were counted as those needing to speak with staff). It is important to keep announcements and recordings reasonable in length—less then 30 seconds and only sparingly up to 1 minute. The average of 60 percent of callers having their concern addressed by the initial announcement has remained fairly constant over the last 3 years and for a range of health events.

A strategy to assist those waiting in queue (which has at peak times reached up to 20 callers with some waiting up to 30 minutes) is to cycle recordings of other frequently requested information in hopes of answering their questions while they are waiting. Many callers may get the information they require from those messages and no longer need to wait for assistance. This ensures that staff is assisting those who could not be helped easily by other means. The recordings can also refer callers to other information sources (such as the Internet) that they may opt to explore instead of waiting in queue or investigate first before calling back.

We have found this most effective for inquiries about finding flu vaccination sites that could be easily located via a Web site. We typically cycle a recording stating, "We are experiencing high call volumes at this time; please consider calling back at another time," for those in queue for more than a few minutes. Callers appreciate being kept informed about wait times, and technology that estimates queue times for callers can be used.

Assuming that messaging will work for a majority of callers, the number of staff and phone lines required to deliver information to the same number of callers can be decreased. Here is the example that we used for calculating the resources needed to handle 28,000 callers depending on an event lasting from 1 to 5 days.

Event Duration (Days) 1 2 3 4 5
Average Calls/Day 28,000 14,000 9,333 7,000 5,600
Staffing (FTE) 354 181 123 93 76
Phone Lines 401 213 148 116 95

The more callers who can be handled effectively with messaging strategies, the fewer staff and phone lines will be needed (good recorded messages can deliver information consistently and at a constant rate).

Event Duration (Days) 1 2 3 4 5
Agent Calls/Day 11,200 5,600 3,733 2,800 5,600
Message Calls/Day 16,800 8,400 5,600 4,200 3,360
Staffing (FTE) 146 76 52 40 33
Phone Lines 174 95 68 54 45

These calculations assume that only 40 percent of callers speak with an agent for an average call length of 293 seconds, including after call activities; the remainder receive only messages/recordings.

Furthermore, by reviewing the concerns of callers speaking with staff, managers can determine if additional information should be added to the initial announcement, added to queue messages, or disseminated by the media or other sources in hopes of meeting demand without call center staff involvement.

All of these strategies for providing information to the greatest number of callers with limited resources greatly assisted us in handling the call volumes related to the 2003/2004 influenza season. From November 17, 2003 to January 31, 2004, HELP received almost 24,000 calls with peak call volumes of 7,145 weekly, 2,565 daily, and 345 hourly. During influenza seasons, many callers are trying to locate vaccination sites. We have successfully referred callers with Internet access to a Web site that they can easily use to find these locations in Colorado. We provided the direct URL address in our initial announcement to off-load many callers to this resource.

Our information providers use the same Web site to assist callers in finding vaccination locations. The announcement content seemed to be effective, since those who waited to speak to staff for that reason prefaced it by saying "Sorry, I know I could use the Internet, but I do not have access." A reporter doing an influenza story showed viewers how to locate vaccination sites using a computer on the air and assisted in reducing the overall HELP call volume almost immediately. Partnering with media can greatly assist in providing the most requested information and reducing the demand on hotlines during events.

IR Applications Evaluation

It is the challenges we encountered with surges in demand to HELP that led us to develop and test the four IR applications. Appendix 4 provides the full details of the two exercises used to evaluate the applications. This is a brief summary of those exercises and user feedback about improving each application:

  • Quarantine/Isolation (QI) Monitoring.
  • Point of Dispensing (POD) Locations.
  • Drug Identification (DI).
  • Frequently Asked Question (FAQ) Library.

Quarantine/Isolation Monitoring. A prototype version of the QI Monitoring Application was tested in 12 rural volunteers who served as "isolation cases" in conjunction with an influenza vaccination exercise in October 2005. Data were collected in the volunteers and entered into the State health department's Outbreak Management System.

The application was revised to reflect many of the user suggestions. In an exercise in May 2006, the revised and more fully developed QI Monitoring Application and the other three applications were evaluated in an urban user group consisting primarily of local health personnel from 10 counties. The goal of the second exercise was to test the ability of the four IR applications either to initiate contact and determine health status of those in quarantine or to effectively communicate key information to users calling in to the applications.

An issue realized from this exercise was that the application could not guarantee quarantine compliance. Even if someone answered the application during every calling period, there was no way to prevent another person from answering on behalf of the person in quarantine. Requiring entry of data to specific prompts like "last 4 digits of social security number" would not guarantee the identity of the person who answered, as almost any identifying information could be shared with another. Adding confirmatory prompts would make the application more complicated and could result in more people finding it difficult to use while being in compliance.

Agencies using home quarantine strategies could consider certain qualifications for individuals to reduce undetected circumvention, such as requiring a land line (and not a cell phone) for contact and agreeing to have call forwarding features disabled. Therefore, it will be important to develop effective risk communication messages to the public and adequate support for those in quarantine to assure good overall compliance. Public health agencies will find it difficult to monitor individuals in home quarantine without strategies to reduce the need for staff.

This application monitored up to 70 percent of quarantined persons demonstrating compliance with few personnel resources. The QI Monitoring Application (or some similar monitoring strategy) will permit limited staff resources to concentrate on obviously noncompliant individuals and those with additional needs or to manage the myriad of other response actions required in a health emergency.

The second version of the QI Monitoring Application required the person answering the call to indicate their most current temperature reading as an objective means to monitor their health status. The public health departments helping to design this exercise believed that providing a thermometer to everyone in quarantine would be realistic and would help to identify those potentially developing signs of illness. Those selecting the option for a temperature reading of less than 100oF were considered well but also were given an option to select if they needed to speak to someone. Those selecting the option for a temperature reading equal to or greater than 100oF or who indicated difficulty in taking their temperature were transferred to the HELP service for assistance. For testing results of the second version, go to Appendix 4.

Concern was raised about whether this application could work for everyone, including the elderly and those with special needs. It was never our intention that the application could work for everyone, rather that it could work for most. It would be at the discretion of public health agencies coordinating quarantines to decide which individuals this application could assist in monitoring, thereby freeing resources to monitor those with additional or special needs.

Drug Identification (DI). The DI Application was tested to determine how effective such an application would be to assist the public in identifying antibiotic drugs that may be dispensed during certain public health events. The underlying challenge is that more than one brand of the same medication will be distributed to the same household during an emergency, and each may have a different appearance. For example, there are several manufacturers of doxycycline. Figure 6 contains the five different appearances of 100 mg doxycycline preparations that are contained in local and national stockpiles. This IR application offers a self-service alternative for callers to correctly identify drugs by type (capsule or tablet), shape, color, and imprints rather than calling their doctor or pharmacist. The application can identify, in addition to doxycycline, ciprofloxacin and Levaquin, which are other antibiotic drugs in many local stockpiles.

Point of Dispensing (POD) Locations. The POD application was tested to determine if callers entering their 5-digit zip code could get correct POD locations in a self-service manner. This application could be modified to provide any zip-code-specific information and to ensure:

  • Consistent, accurate information based on zip code.
  • Collection of zip code data to characterize events (situational awareness—caller locations and the potential need for more media messaging).
  • Expanded capacity for handling surges since calls are handled without personnel.
  • Support for mass prophylaxis/immunizations, evacuations, or sheltering in place information.

Volunteers were assigned to evaluate this application by calling a toll-free number, entering a 5 digit zip code, and recording that zip code and the location they were given on an evaluation form. We received all evaluations back (100 percent return rate), and all recorded the correct POD location for their entered zip code.

Frequently Asked Question (FAQ) Library. The FAQ Library Application was tested to evaluate the ability of users to navigate a library of messages and to obtain desired information. Our HELP program uses this library for handling callers after hours with great success by allowing self-service information delivery that is consistent and accurate. The application collects entered data to characterize the information needs of the public (the entered zip code for situational awareness—identifying public information needs and where to target them). The application is able to expand capacity for handling surges and is capable of adapting to different events.

We met our overall exercise objectives and obtained excellent feedback to improve the tested applications. We also obtained important information on user acceptance for these IR applications. Although evaluations were mostly favorable for all four applications, the FAQ Application seemed more acceptable than the DI Application (perhaps because the latter concerned medications to be taken). The comments and evaluations of these applications should help public information officers in determining which ones may be acceptable for different events and in developing messaging strategies. These results also suggest areas for potential community outreach efforts for public health agencies to create a more informed public. One lesson learned is that the tools will be only as good as the information that is developed for them and how it is provided to the public.

Return to Contents
Proceed to Next Section


The information on this page is archived and provided for reference purposes only.


AHRQ Advancing Excellence in Health Care