Identifying Locations for Public Access Defibrillators using Mathematical Optimization

T. C. Y. Chan, H. Li, G. Lebovic, S. K. Tang, J. Y. T. Chan, H. C. K. Cheng, L. J. Morrison, S. C. Brooks
2013 Circulation  
Me ec ch chanical and Industrial Engine eer er erin n ng; 4 Div of Eme mergen en ncy cy cy Medicine, Dept of M Me Medicine, Un Un Univ iv ive e ersi si ity ty ty of of of T T Tor or ron on onto to to, To To Toro o ront nt to; o; o 2 2 A App pl pli ied d He He Heal al alth th th R R Res es e ea ea earc rc ch h h C Ce C nt nt tre re re; ; ; 3 3 3 Re Re Resc sc cu u, u, K K Kee ee eena na nan Re Re Research C Cent tr re e, Li i K K Ka S Sh h hin ng ng K Kno now w wle e edge e e In n nstit t tu ut
more » ... In n nstit t tu ut te, e, S S St. t. t. M M Mic c chae el's Ho o osp p pit ital al, , To T ro o on nt to; o; 5 D D Dep pt o o of Em Em Emer erge ge genc nc ncy y y Me Me Medi di d ci ine ne ne, , , Qu Qu Quee een' n' n's s s Un Un Univ iv iver er ersi si sity ty ty a a at t t Ki Ki Abstract: Background-Geo-spatial methods using mathematical optimization to identify clusters of cardiac arrests and prioritize public locations for defibrillator deployment have not been studied. Our objective was to develop such a method and test its performance against a population-guided approach. Methods and Results -All public location cardiac arrests in Toronto, Canada from December 16, 2005 to July 15, 2010, and all automated external defibrillator (AED) locations registered with Toronto Emergency Medical Services as of September 2009, were plotted geographically. Current AED coverage was quantified by determining the number of cardiac arrests occurring within 100 meters of a registered AED. Clusters of cardiac arrests without a registered AED within 100 meters were identified. Using mathematical optimization techniques, cardiac arrest coverage improvements were computed and shown to be superior to results from a populationguided deployment method. There were 1310 eligible public location cardiac arrests and 1669 registered AEDs. Of the eligible cardiac arrests, 304 were within 100 meters of at least one registered AED (23% coverage). The average distance from a cardiac arrest to the closest AED was 281 meters. With AEDs deployed in the top 30 locations, an additional 112 historical cardiac arrests would be covered (32% total coverage) and the average distance to the closest AED would be 262 meters. Conclusions-Geographical clusters of cardiac arrests can be easily identified and prioritized using mathematical modeling. Optimized AED deployment can increase cardiac arrest coverage and decrease the distance to the closest AED. Mathematical modeling can augment public AED deployment programs. g p q , coverage improvements were computed and shown to be superior to results from m m a p p pop op pul ul ulat at atio io ion-nguided deployment method. There were 1310 eligible public location cardiac arrests and 1669 egistered AEDs. Of O the eligible cardiac arrests, , 304 were within 100 meters of at least one e egi gi gist st ster ered d A A AE E ED ( (23 23% % co cove vera r ge ge). ) Th T e av ver e ag ge e di dis s sta a ance f fro rom m a ca ca ard rd r ia ac c ar arre r st to o th t e cl los oses st AED w was s s 28 2 1 meters rs. . W W Wit ith h h AE AE AEDs Ds Ds d d dep eplo lo l y ye yed d d i in th h he top 30 0 lo oc ocat atio io on ns s, a a an a add dd dit it i io o ona na al l l 1 11 112 2 hi hi hist st s or or ric c cal a al ca ca ard rd rdia ia i c c arrest st sts s s w wou ul uld d be be e cov v ver er red ed e (3 (3 32% 2% 2% t tot otal al cov over er ra ag ge) e) ) a a and nd th he he a ave ve erage ge e d d dis ista ta anc nc ce e to to t th h he c cl lo ose ses s st AE ED D wo w ul uld d be e 2 262 2 m meter rs. s Conclusion ns s s-G -G Geo eo e gr g grap ap aph hica ca cal l l cl clus ust te ters r rs o o of f f ca card rd rdia ia iac c c ar arre r res st sts ca a can n be be be e e eas a il il ly y y id id iden en enti ti tifi fi f ed ed d a and nd nd p pri r r oritized by guest
doi:10.1161/circulationaha.113.001953 pmid:23553657 fatcat:ulyggnlczrd4ja3br76ssx4coy