Characterization of Call Prioritization Time in a Police Priority Dispatch System™
David Warner, Shawn Messinger, Chris Knight, Greg Scott, MBA, EMD-QI, Jeff J. Clawson, MD, Isabel Gardett, PhD, Tracey Barron, BS, Mark Rector, Brett Patterson, Lazaro Guerra, Angela VanDyke, and Christopher Olola, PhD
Aug 10, 2014|AEDR 2014 Vol. 2 Issue 2|Original Research
Introduction: Time-to-dispatch in a 911 center continues to be a topic of much discussion in public safety. This study represents a first attempt to classify a subset of time-todispatch, call prioritization time, the time required to gather critical information prior to dispatching the call. The study characterizes call prioritization time in two Police dispatching agencies by determining overall median call prioritization time for all Chief Complaints (CCs) in the agencies studied, then by determining specific call prioritization times for the top five most commonly-used CCs, both overall and by agency.
Objectives: The primary objective in this study was to describe the distribution of call prioritization time in a PPDS™ agency for specific emergency dispatch CCs and priority levels, within the PPDS.
Methods: This was a retrospective and non-controlled descriptive study involving two police dispatch agencies with International Academies of Emergency Dispatch (IAED)-Accredited Centers of Excellence designation.
Results: Nearly half a million (n=456,804) emergency call cases were received at the two 911 communication centers during the study period. The overall median call prioritization time for all protocols was 45 seconds, but the times varied by call type and priority level. Overall, the higher-priority DELTA and ECHO calls—which represent the majority of in-progress events—were prioritized significantly faster than lower-acuity ALPHA and BRAVO-level calls.
Conclusions: This study provides a first published description of call prioritization time and measurement of call prioritization time in two communication centers. Broader adoption and continued study of this new measure may help set the stage for improvements in evidence-grounded calltaking time standards.
Time-to-dispatch in a 911 center (i.e., the time it takes to gather critical incident demographic information, formulate a response priority, and notify the emergency responders, continues to be a topic of much discussion in public safety. There is no blanket approach to establishing a time standard for dispatch, unless defined at the level of a specific type of event (Chief Complaint (CC)) and when it occurred.1 Additionally, because most public safety standards define total response time to a 911 event as beginning in the communication center (when the calltaker picks up the phone, for example, or when the call "hits the switch" and is recognized),2 time-to-dispatch is viewed as a critical sub-component of total response time to an emergency scene. This places the 911 center in the awkward position of getting constant pressure from response agencies to shorten dispatch times to conform to standards that were set without evaluating the efficacy of the 911 calltaking process or determining how much calltaking time is actually necessary to collect and transmit critical case information.
To the best of investigators' knowledge, there is no empirical evidence to suggest that incrementally shorter response times actually improve case outcomes—i.e., save lives or protect people and property—in the vast majority of public safety responses.3, 4, 5, 6, 7, 8 In EMS, it has been demonstrated that rapid response may result in beneficial outcomes in a few specific types of cases such as cardiac arrest, extreme difficulty breathing, and acute cardiac conditions.9, 10, 11, 12, 13 While no similar evaluation has been performed in police dispatch, it is often taken for granted that speed is equally critical when notifying responding police units, particularly in the most dangerous and life/property threatening cases, so as to facilitate a fast response to the scene and increase the likelihood of apprehending a perpetrator or protecting the public from harm.
Given the dearth of good quantitative information on this topic, there is a pressing need to assess dispatch times for 911 police calls. A thorough evaluation of the distribution patterns of calls by CC type and priority level has already been conducted.14 However, time-to-dispatch research remains to be done. This study investigated a critical subcomponent of time-to-dispatch, a time interval we are calling call prioritization time. Call prioritization time is the time it takes for the calltaker to gather the specific information required to determine the call type, call severity, time of event occurrence (In Progress/Just Occurred versus Past events), and any scene safety issues such as weapons or other hazards. This time interval begins when a calltaker opens the emergency dispatch software after call demographics (location and call back number) have been verified in the computer-aided dispatch (CAD) system and ends when a dispatch code (the code that identifies the call type and priority and is used to determine appropriate response) is selected.
Using the Police Priority Dispatch System (PPDS™), certified 911 Emergency Police Dispatchers (EPDs) are able to gather incident information according to an organized, standardized method, then categorize each incident within a specific CC protocol (Fig. 2)called Determinant Descriptors. The priority level defines the relative urgency and type of response needed for a given event. The lowest priority, the OMEGA (Ω) level, indicates a call that typically does not require the response of a police officer and can be handled strictly as a telephone report or inquiry. However, the actual response sent (or not sent, in this case) is determined by each agency, as defined by its local response plan and based on the acuity/urgency of these codes compared to available response and mitigation resources. The ALPHA (A) level indicates a non-urgent case that involves routine information gathering, reporting, or follow-up tasks undertaken by a responding officer. BRAVO (B)- and CHARLIE (C)-level cases require a more urgent officer response and often involve past crimes, minor hazards, or unwitnessed, potentially dangerous situations and circumstances. DELTA (D)-level cases are those for which an officers are needed immediately, including "in-progress" and "just occurred" crimes, violent or potentially violent suspects or events, confirmed-injury traffic accidents, and possession, use, sale, or manufacture of illicit substances.
Finally, the ECHO level allows the earliest dispatch and pre-arrival instructions for specific immediate dangers and involve trained personnel who would not routinely respond to such events (e.g., detectives, public information officers, and crime prevention officers). These may be dispatched to ECHO situations when they are the closest units to the scene, when safe to do so.
Therefore, the hypothesis in this study was that there are significant variations in the call prioritization time requirements among the different dispatch protocols, and priority levels studied.
The primary objective in this study was to describe the distribution of call prioritization time in a PPDS agency for specific emergency dispatch CCs and priority levels, within the PPDS (version 3.0, 2008 release). This study classifies a component of time-to-dispatch in two Police dispatching agencies; it does so by determining overall median call prioritization time for all CCs in the agencies studied, then by determining specific call prioritization times for the top five most commonly used CCs, both overall and by agency.
Design and setting
The retrospective and non-controlled descriptive study involved two International Academies of Emergency Dispatch (IAED)-Accredited Centers of Excellence (ACE) police dispatch agencies: City of Hialeah Police Department, Florida, USA, and Prince George's County (PGC) Public Safety Communications, Maryland, USA. The two centers provide useful counterpoints. Hialeah is smaller in area but significantly higher in population density, with a largely Hispanic population, while PGC has a population three times the size of Hialeah spread over a an area nearly 20 times larger, and the population of PGC is predominantly black. Further details about the centers were previously described elsewhere.14
Data management and analysis
Police dispatch data was collected between 2010 and 2011 at Hialeah, and between 2011 and 2012 at Prince George's County. The study sample comprised all the data that was available at these agencies during the study period. For this study, call prioritization time was defined as the time from when the calltaker opened ProQA (which starts at "Tell me exactly what happened") to the time of initial dispatch of the incident to officers (selection of initial ProQA Determinant Code).
The frequency distributions of the usage status of all CC protocols were graphed on a bar chart. The percentage of calls and call prioritization time for each priority level was also tabulated and graphed (on box plots) respectively. Call prioritization time was characterized for the top 5 most frequently used protocols and for all other protocols in general. Call prioritization time was also categorized by priority level within each of these protocols.
All calls in which the call prioritization time was less than 5 seconds or greater than 5 minutes were excluded from analysis, because per the investigators' experiences, these time values were unrealistic in a police dispatching environment. Even in the case of ECHO calls, reaching a dispatch point in under 5 seconds is extremely rare because the calltaker must, at minimum, determine the nature of the event. Similarly, it would be at least unusual, if not impossible, for a legitimate call handled correctly in ProQA to require more than five minutes to reach the dispatch point. This is especially true given that many of the calls with call prioritization times of greater than five minutes actually had times of several hours, up to several days, generally not containing or reaching a dispatch code at all. These outliers almost certainly reflect user error of some kind (such as leaving ProQA open when one leaves work), rather than legitimate calls.
STATA for Windows® software (STATA Statistical Software: Release 11.2 ©2009, StataCorp, College Station, TX, USA) was used for data analysis. Descriptive statistics such as median, minimum, and maximum times-to-dispatch and percentages were used to describe non-normally distributed (i.e., skewed) time measures. The non-parametric median test or Mann–Whitney U test (or Wilcoxon rank-sum test) was used to assess call prioritization time differences between study groups, and the significance of the differences were evaluated at the 0.05 significance level. The analysis was classified by CC and priority level.
The primary endpoint was the call prioritization time classified by CC and priority level.
Nearly half a million (n=456,804) emergency call cases were received at the two 911 communication centers during the study period (Figure 4). Of these cases, 0.5% (n=2,147) were excluded due to call prioritization time equal to or less than 5 seconds (n=104) or call prioritization time greater than 5 minutes (n=2,043). Therefore, 454,657 (99.5%) cases were included in the study sample.
Overall (n=454,657), the top 5 most frequently used dispatch protocols were Protocol 113 (Disturbance/ Nuisance) (22.6%), Protocol 131 (Traffic/Transportation Incident [Crash]) (12.7%), Protocol 130 (Theft [Larceny]) (12.5%), Protocol 114 (Domestic Disturbance/Violence) (7.2%), and Protocol 129 (Suspicious/Wanted [Person, Circumstances, Vehicle]) (7.0%), in the order of most to the less frequently used Protocol (Fig. 5). These top 5 protocols constituted a majority (62.0%) of the total call volume., However, the order and type of top 5 most frequently used protocols differed by agency. In Hialeah (n=72,989), the top 5 were protocol 131 (Traffic/ Transportation Incident [Crash]) (21.0%), protocol 104 (Alarms) (17.5%), protocol 113 (Disturbance/Nuisance) (15.3%), protocol 130 (Theft [Larceny]) (9.0%), and protocol 114 (Domestic Disturbance/Violence) (4.6%). In PGC (n=381,668), the top five were protocol 113 (Disturbance/Nuisance) (24.0%), protocol 130 (Theft [Larceny]) (13.2%), protocol 131 (Traffic/Transportation Incident [Crash]) (11.2%), protocol 114 (Domestic Disturbance/ Violence) (7.7%), and protocol 129 (Suspicious/Wanted [Person, Circumstances, Vehicle]) (7.5%). See Table 3 for a breakdown of call prioritization time for the top five protocols by agency and overall.
The DELTA-priority level consistently had the highest percentage of emergency cases (in all CCs) in each agency (Hialeah: 47.1%, PGC: 45.8%) and overall (46.0%) (Table 1). BRAVO-priority level had the second highest percentage of cases (Hialeah: 36.7%, PGC: 30.6%, Overall: 31.6%).
The overall median call prioritization time for all protocols was 45 seconds: 51 seconds for the City of Hialeah and 44 seconds for PGC. The specific median times (miniummaximum) were: 12 (6-275) seconds for ECHO, 41 (6-300) seconds for DELTA, 47 (6-300) seconds for CHARLIE, 53 (6-300) seconds for BRAVO, 53 (6-300) seconds for ALPHA, and 23 (6-300) seconds for OMEGA-priority level calls.
In addition, the overall median call prioritization time an ALPHA-level call (reference measure) in the city of Hialeah was significantly longer than the call prioritization time in any other level except for the BRAVO priority level (p=0.144). This comparison held true in PGC as well, where the call prioritization time for ALPHA-level calls was significantly higher than the call prioritization time for the other priority levels. However, in the City of Hialeah, the call prioritization time for CHARLIE-level calls was 54 (8-300) seconds, which was not significantly different from the 55 (12-294) second call prioritization time for an ALPHA-priority level call (p=0.475), most likely because both CHARLIE and ALPHA-level calls are the most likely to be prioritized at the end of Key Questions, with all Key Questions asked (as opposed to DELTAs, with their early send points, and BRAVOS, with their shortened "cold call" pathways). The sample size for ECHO calls was too small to make meaningful significance calculations.
Altogether, 93.3% of all calls had a call prioritization time of less than 120 seconds. Specifically, 89.7% of ECHOs, 94% of DELTAs, 95.8% of CHARLIEs, 90.4% of BRAVOs, 89.9% of ALPHAs, and 98.4% of OMEGAs had a call prioritization time of less than 120 seconds (Table 2). At the DELTA and ECHO levels, 72.7% and 79.5% of calls were prioritized in 60 seconds or less. Overall, approximately 86% of calls had a call prioritization time of less than 90 seconds, and 67.5% had a call prioritization time of less than 60 seconds.
Overall, among the top 5 most frequently used dispatch protocols, Protocol 113 (Disturbance/Nuisance) had the shortest call prioritization time (37 seconds) (Table 3). The other protocols ranged from 42-49 seconds. Similarly, in PGC protocol 113 had the shortest call prioritization time (35 seconds). However, in Hialeah, protocol 131 (Traffic/ Transportation Incident [Crash]), registered the shortest call prioritization time (18 seconds), among the top 5 most frequently used protocols. In general, Hialeah had longer call prioritization time values, compared to PGC, among the top 5 protocols.
Overall, considering all CC protocols, Protocol 107 (Assist Other Agencies) had the longest (122 seconds) and Protocol 104 (Alarms) had the shortest (19 seconds) median call prioritization time (Table 4: online supplement). Otherwise, Protocols 128 (Supplemental) and 131 (Traffic/Transportation Incident [Crash]) had the longest (87 seconds) and shortest (25.8 seconds) median call prioritization times for the ALPHA-priority level calls, respectively. Protocols 127 (Suicidal Person/Attempted Suicide) and 111 (Damage/ Vandalism/Mischief) had the longest (159.5 seconds) and shortest (30 seconds) median call prioritization times for the BRAVO-priority level calls, respectively. For the CHARLIEpriority level calls, Protocols 107 (Assist Other Agencies) and 130 (Theft [Larceny]) had the longest (129.5 seconds) and shortest (31 seconds) median call prioritization times, respectively. Protocols 123 (Missing/Runaway/Found Person) and 131 (Traffic/Transportation Incident [Crash]) had the longest (193 seconds) and shortest (104 seconds) median call prioritization times for the DELTA-priority level calls, respectively. Finally, Protocol 131 (Traffic/Transportation Incident [Crash]) was the only protocol with ECHO-priority level calls, and the median call prioritization time for these calls was 12 seconds; the single 275-second outlier was almost certainly the result of human error.
Overall, the higher-priority DELTA and ECHO calls— which represent the majority of in-progress events—were prioritized significantly faster than lower-acuity ALPHA and BRAVO-level calls. At the CC level, there was significant variation among the different dispatch protocols, with call prioritization time influenced by both call type and protocol structure. For example, Protocol 123 (Missing/ Runaway/Found Person) had an unusually high call prioritization time, but only at the DELTA level. This is logical given that, in missing/runaway calls, the description of the person is the most important information to provide to the responding officer (it is in fact the reason for the call). Thus, gathering a complete/thorough description ("description essentials") of the person in these types of calls may take longer than other calls-for-service that may not require such and in-depth persons description.
In general, the study findings demonstrate a great deal of consistency across what might be termed "crimes-againstpeople" protocols, which represent the majority of highacuity police events. These include Protocols 110 (Burglary/ Home Invasion), 113 (Disturbance/Nuisance), 114 (Domestic Disturbance/Nuisance), 126 (Robbery/Carjacking), and 130 (Theft/Larceny). Not surprisingly, these protocols are very similarly structured, with only the most critical information (number of people involved, physical or verbal in nature, and weapons involved or mentioned, as appropriate) gathered before an early DELTA-level send point.
In this study, very little difference appeared to exist in call prioritization time between the two agencies studied, except in that PGC does not handle Alarm calls through the dispatch system studied here (and thus times for these calls could not be compared with those of Hialeah). In general, in both agencies, although higher-acuity and in-progress events require the gathering of more information—including suspect and/or victim description information—call prioritization time for these events was nonetheless faster than for lower-acuity or past events because of early send points built into the system (early opportunities to select a dispatch code and dispatch units in high-priority cases).
Although many US states, agencies, or communication centers have developed their own standards for appropriate call-taking time (including, in many cases, standards for call prioritization time), little information exists to provide evidential support for how long it should ideally take to collect sufficient information to appropriately dispatch a police call. This study provides a first attempt to develop such support, offering a first glimpse of how long it takes to gather the amount and type of information both dispatcher and responder must have to send or execute the correct response for the event. Understanding the realities of call prioritization time in two communication centers serving rather different populations, in other words, offers a baseline for the creation of more realistic, and evidence-based, standards for dispatching times.
This study provides a snapshot view, a descriptive look at call prioritization time in centers that are not only already using a PPDS structured call-taking process, but are accredited at an ACE-level of compliance. Future studies should also attempt to compare call prioritization time (among other metrics) in communication centers not using a structured call-taking process, centers that have not received ACE accreditation, and ACE centers like those studied here. The investigators in this study have planned future studies which will review additional time metrics, such as the amount of time calltakers spend gathering Description Essentials (i.e., suspect and vehicle information) and total call time.
Finally, as in any study of time measure in the dispatch environment, it is important to remember that while time metrics are clearly important and are used by most dispatch agencies in one form or another, the gathering of quality key information and the sending of the right response are at least equally important. This study did not specifically examine which types of information are (or should be) gathered before, as opposed to after, dispatch points (especially the early dispatch points used for high-priority DELTA-level calls). Future studies should do so to determine which information is, in fact, vital to the determination of the problem type, the safety of the scene, and the most appropriate response to send—the three key pieces of information necessary to an effective response. The existence of every question used in police call-taking is based on its defined objective (i.e., what it is looking for and why) and its ability to actually gather that information.
There are two limitations in this study that suggest that future research is required to validate the study findings. First, the sample size—representing only two police dispatch agencies—is relatively small. Given that differences in dispatching procedure and local policy exist among all agencies, a wider sampling would help provide a more broadly-applicable picture of median call prioritization times. Second, and related, is the fact that one of the agencies involved in this study handles Alarm calls outside of police dispatch system studied, making those calls impossible to compare between agencies. Further research would help determine whether this agency's method of handling Alarms (by externalizing them to a special Alarms center) is the more common, or whether most agencies using the PPDS in fact handle alarms using Protocol 104 (Alarms).
The study findings demonstrated that the large majority of all calls are prioritized in 60 seconds or less, especially the high-acuity DELTA and ECHO calls. Moreover, nearly 95% of all calls were prioritized in under 120 seconds. Although considerable variation in call prioritization time exists among various CCs based on event type, protocol structure, and location of early dispatch points, there is less variation within the core group of "crimes-against-people" that make up the majority of high-acuity calls. This study provides a first glimpse of call prioritization time in two centers, setting the stage for more evidence-grounded standards for call-taking time. It also provides a foundation upon which to develop more definitive police dispatch studies.
TABLES AND GRAPHS
- Clawson JJ, Sinclair B. Creating time standards that work. The National EMD Journal. 1999; 1(1):14-15.p
- Technical Committee on Public Emergency Service Communication. NFPA 1221: Standard for the Installation, Maintenance, and Use of Emergency Services Communications Systems. 2010 Edition. Quincy, Massachusetts: National Fire Protection Association (NFPA); 2009. Chapter 7, Operations; p. 14-16. http://www.nfpa.org/catalog/product.asp?pid=122110&icid=B484
- Blackwell TH, Kaufman JS. Response time effectiveness: comparison of response time and survival in an urban emergency medical services system. Acad Emerg Med. 2002; 9(4):288-95.
- Blackwell TH, Kline JA, Willis JJ, Hicks GM. Lack of association between prehospital response times and patient outcomes. Prehosp Emerg Care. 2009; 13(4):444-50.
- Newgard CD, Schmicker RH, Hedges JR, Trickett JP, Davis DP, Bulger EM, Aufderheide TP, Minei JP, Hata JS, Gubler KD, Brown TB, Yelle JD, Bardarson B, Nichol G. Emergency medical services intervals and survival in trauma: assessment of the "golden hour" in a North American prospective cohort. Ann Emerg Med. 2010; 55(3):235-246.e4.
- Pons PT, Haukoos JS, Bludworth W, Cribley T, Pons KA, Markovchick VJ. Paramedic response time: does it affect patient survival? Acad Emerg Med. 2005; 12(7):594-600.
- Blanchard IE, Doig CJ, Hagel BE, Anton AR, Zygun DA, Kortbeek JB, Powell DG, Williamson TS, Fick GH, Innes GD. Emergency medical services response time and mortality in an urban setting. Prehosp Emerg Care. 2012; 16(1):142-51.
- Salvucci A, Kuehl A, Clawson JJ, Martin RL. The response time myth does time matter in responding to emergencies? Topics in Emergency Medicine. 2004; 26(2):86-92.
- Neukamm J, Graesner JT, Schewe JC, Breil M, Bahr J, Heister U, Wnent J, Bohn A, Heller G, Strickmann B, Fischer H, Kill C, Messelken M, Bein B, Lukas R, Meybohm P, Scholz J, Fischer M. The impact of response time reliability on CPR incidence and resuscitation success - a benchmark study from the German Resuscitation Registry. Crit Care. 2011; 15(6):R282.
- Van Vleet LM, Hubble MW. Time to first compression using medical priority dispatch system compression-first dispatcher-assisted cardiopulmonary resuscitation protocols. Prehosp Emerg Care. 2012; 16(2):242-50.
- Freimark D, Matetzky S, Leor J, Boyko V, Barbash IM, Behar S, Hod H. Timing of aspirin administration as a determinant of survival of patients with acute myocardial infarction treated with thrombolysis. Am J Cardiol. 2002; 89(4):381-5.
- Weaver WD, Hill D, Fahrenbruch CE, Copass MK, Martin JS, Cobb LA, Hallstrom AP. Use of the automatic external defibrillator in the management of out-of-hospital cardiac arrest. N Engl J Med. 1988; 319(11):661-6.
- Shuster M, Keller JL. Effect of fire department first-responder automated defibrillation. Ann Emerg Med. 1993; 22(4):721-7.
- Messinger S, Warner D, Knight C, Scott G, Rector M, Barron T, Van Dyke A, Guerra L, Gardett I, Ph.D, Patterson B, Clawson J, MD, Olola C, Ph.D. The Distribution Of Emergency Police Dispatch Call Incident Types and Priority Levels Within the Police Priority Dispatch System. Annals of Emergency Dispatch & Response 2013; 1(2):12-17
Citation: Warner D, Messinger S, Knight C, Scott G, Clawson J, Gardett I, Barron T, Rector M, Patterson B, Guerra L, VanDyke A, Olola C. Characterization of call prioritization time in a Police Priority Dispatch System. Annals of Emergency Dispatch & Response. 2014;2(2):16-22.