BACKGROUND

After a mass casualty incident (MCI), rapid mobilization of hospital personnel is required because of an expected surge of victims. Risk assessment of our department's manual phone tree recall system revealed multiple weaknesses that would limit an effective response. Because cell phone use is widespread within the department, we developed and tested a staff recall system, based in our anesthesia information management system (AIMS), using Short Message Service (SMS) text messaging.

METHODS

We sent test text messages to anesthesia staff members' cell phone numbers, determined the distance from their home to the hospital, and stored this information in our AIMS. Latency testing for the time from transmission of SMS test messages from the server to return of an e-mail reply was determined at 2 different times on 2 different dates, 1 of which was a busy holiday weekend, using volunteers within the department. Two unannounced simulated disaster recall drills were conducted, with text messages sent asking for the anticipated time to return to the hospital. A timeline of available staff on site was determined. Reasons for failure to respond to the disaster notification message were tabulated.

RESULTS

Latency data were fit by a log-normal distribution with an average of 82 seconds from message transmission to e-mail reply. Replies to the simulated disaster alert were received from approximately 50% of staff, with 16 projecting that they would have been able to be back at the hospital within 30 minutes on both dates. There would have been 21 and 23 staff in-house at 30 minutes, and 32 and 37 staff in-house at 60 minutes on the first and second test date, respectively, including in-house staff. Of the nonresponders to the alert, 48% indicated that their cell phone was not with them or was turned off, whereas 22% missed the message.

CONCLUSIONS

Our SMS staff recall system is likely to be able to rapidly mobilize sufficient numbers of anesthesia personnel in response to an MCI, but actual performance cannot be predicted with confidence. Using our AIMS as the source for contact information and from which to send messages was simple, inexpensive, and easy to implement. Updating contact information, periodic testing, and analysis of responses to simulated disaster alerts are essential for the effective functioning of such a system. However, maintenance of alternative methods of communication is recommended, because there may be more significant message transmission delays and failures during an actual MCI, and not all staff will receive the text message in a timely fashion.