{"id":20630,"date":"2016-08-01T12:00:00","date_gmt":"2016-08-01T16:00:00","guid":{"rendered":"http:\/\/iacpmag.wp.matrixdev.net\/building-better-early-intervention-systems\/"},"modified":"2024-10-07T13:01:04","modified_gmt":"2024-10-07T17:01:04","slug":"building-better-early-intervention-systems","status":"publish","type":"post","link":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/","title":{"rendered":"Building Better Early Intervention Systems"},"content":{"rendered":"<p><a name=\"eis\"><\/a>A few years ago, the Charlotte-Mecklenburg, North Carolina, Police Department\u2019s (CMPD\u2019s) Early Intervention System (EIS) flagged an officer who had been involved in an unusually high number of incidents resulting in injuries. A supervisor reviewed the officer\u2019s records and determined that, while the officer was using force at appropriate times, he was executing takedowns incorrectly. The supervisor sent the officer to the academy for a refresher course. As a result, injuries from the officer\u2019s arrests plummeted, the officer avoided punishment and learned to properly execute takedowns, and the department reduced its legal liability. The officer, the department, and the community benefited from the early notification.<\/p>\n<p>EISs are a staple in U.S. police departments\u2014a 2007 survey showed that 65 percent of surveyed police departments with 250 or more officers had an EIS.<a href=\"#1\"><sup>1<\/sup><\/a> The U.S. Commission on Civil Rights, the Commission on Accreditation of Law Enforcement Agencies, the U.S. Department of Justice, the International Association of Chiefs of Police, and the Police Foundation have all recommended departments use these systems.<a href=\"#2\"><sup>2<\/sup><\/a> Most federal consent decrees require a department to use an EIS.<a href=\"#3\"><sup>3<\/sup><\/a><\/p>\n<p>Despite the widespread use of EISs, empirical studies on their effectiveness have been limited, and the findings give mixed conclusions. Case studies have shown that using an EIS reduces citizen complaints, but it is unclear whether this decrease arises from a reduction in problematic behavior or from discouraging officers from proactive policing.<a href=\"#4\"><sup>4<\/sup><\/a> A large-scale study of emerging EISs concludes that effectiveness depends on departmental characteristics and details of implementation, such as which indicators are tracked, what thresholds are assigned, and how supervisors handle the system\u2019s flags.<a href=\"#5\"><sup>5<\/sup><\/a> In addition, many departments have expressed frustration with EIS accuracy.<\/p>\n<p>With these issues in mind, CMPD and the University of Chicago\u2019s Center for Data Science and Public Policy (DSaPP) joined forces to build a better EIS. CMPD has been a technology leader in the public safety sector, building a centralized data warehouse and much of its own software. DSaPP has expertise in building predictive systems for nonprofits and governments, working with dozens of schools, social service agencies, hospitals, departments of public health, and more.<a href=\"#6\"><sup>6<\/sup><\/a> Together, the researchers analyzed the performance of CMPD\u2019s current EIS, then used advanced analytics to build a new EIS prototype. The prototype EIS appears to be a great start in changing how CMPD addresses employee issues. Preliminary results suggest the new system increases the odds of flagging officers who will be involved in an adverse incident in the next 12 months. Equally important, it reduces the number of officers who were flagged but did not go on to have an adverse incident in the next year. CMPD and DSaPP are conducting a field trial of this system and building tools to ready it for use.<\/p>\n<p><big><b>Background<\/b><\/big><\/p>\n<p>CMPD has more than 1,800 sworn personnel patrolling Charlotte, North Carolina, and the unincorporated areas of Mecklenburg County. CMPD\u2019s jurisdiction covers more than 400 square miles with a population of 900,000. To use its resources efficiently, CMPD has invested heavily in technology and even built its own EIS in 2005. While building the EIS, CMPD\u2019s leaders established the following set of thresholds for flagging officers:<\/p>\n<ul>\n<li>2 accidents in 180 days<\/li>\n<li>2 injuries in 180 days<\/li>\n<li>2 pursuits in 180 days<\/li>\n<li>3 complaints in 180 days<\/li>\n<li>3 uses of force in 90 days<\/li>\n<li>5 of any of these incidents in 180 days<\/li>\n<\/ul>\n<p>If an officer surpasses any one of these thresholds, the system flags the officer and notifies the commanding officers. A CMPD supervisor then reviews the officer\u2019s records to identify whether the department can help the officer avoid adverse incidents through training, counseling, or other departmental<br \/>\nsupport.<\/p>\n<p>Unfortunately, threshold systems such as these tend to be inaccurate. In a typical year, the CMPD EIS incorrectly flags hundreds of officers who don\u2019t experience adverse incidents, while failing to flag dozens of officers who do, which leads the department to spend resources reviewing and providing interventions for the wrong officers. More sophisticated systems adjust the thresholds by one or two factors (such as the officer\u2019s beat) or use relative thresholds instead (such as flagging officers in the top 5 percent), but even these higher-end EISs still ignore important information, such as an officer\u2019s long-term history.<\/p>\n<p>Threshold-based systems pose other challenges, too. First, threshold-based systems are difficult to customize. At least one vendor built hard-coded thresholds into their EIS, making changes difficult and costly\u2014which is good for the vendor, but bad for the department. Ideally, the system should improve as the department collects more data, but most threshold systems rely on many rules of thumb, making such changes unlikely. Second, threshold systems cannot identify officers at low risk of adverse incidents. Departments choose thresholds to identify the highest-risk officers, but identifying the lowest-risk officers can help the department better assign officers or look for best practices to incorporate into its training programs. Third, multiple departments have suggested that threshold systems are easy to \u201cgame\u201d because the thresholds are so easy to see and understand.<\/p>\n<p>Machine learning can help address these issues. <i>Machine learning<\/i> is the ability of a computer to learn patterns in data and to use those patterns to make predictions. It has been used over the past 30 years to solve thousands of problems, from driving cars to providing search results on Google. The next section describes the CMPD-DSaPP efforts to build a machine-learning EIS.<\/p>\n<table style=\"float: right;\" width=\"420\" bgcolor=\"ffffff\">\n<tbody>\n<tr>\n<th colspan=\"3\" align=\"left\"><b>Table 1. Records and time coverage for CMPD data sources.<\/b><\/th>\n<\/tr>\n<tr>\n<td>\n<table border=\"1\" width=\"420\" cellpadding=\"5\" bgcolor=\"669999\">\n<tbody>\n<tr>\n<td><b>Database<\/b><\/td>\n<td><b>Number of Records<\/b><\/td>\n<td><b>Time Window<\/b><\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Training<\/b><\/td>\n<td align=\"center\">1.4M<\/td>\n<td align=\"center\">2001\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td><b>Internal Affairs<\/b><\/td>\n<td align=\"center\">\n<p>20K<\/p>\n<p>&nbsp;<\/td>\n<td align=\"center\">2002\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Employee Records<\/b><\/td>\n<td align=\"center\">20K<\/td>\n<td align=\"center\">2002\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td><b>Traffic Stops<\/b><\/td>\n<td align=\"center\">1.6M<\/td>\n<td align=\"center\">2002\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Dispatch Incidents<\/b><\/td>\n<td align=\"center\">14M<\/td>\n<td align=\"center\">2003\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td><b>Field Interviews<\/b><\/td>\n<td align=\"center\">180K<\/td>\n<td align=\"center\">2003\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Arrests<\/b><\/td>\n<td align=\"center\">350K<\/td>\n<td align=\"center\">2005\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td><b>Crime Reports<\/b><\/td>\n<td align=\"center\">959K<\/td>\n<td align=\"center\">2005\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Existing EIS<\/b><\/td>\n<td align=\"center\">14K<\/td>\n<td align=\"center\">2005\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td><b>Citations<\/b><\/td>\n<td align=\"center\">946K<\/td>\n<td align=\"center\">2006\u20132015<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td><b>Secondary Employment<\/b><\/td>\n<td align=\"center\">651K<\/td>\n<td align=\"center\">2009\u20132015<\/td>\n<\/tr>\n<tr>\n<td colspan=\"3\" align=\"right\">* M = million; K = thousand<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p><big><b>CMPD\u2019s Prototype System<\/b><\/big><\/p>\n<p>CMPD defines an adverse incident as a sustained complaint; a preventable accident or injury; or an unjustified use of force, pursuit, raid and search, discharge of firearm, or tire deflation deployment. DSaPP built a system to predict whether CMPD\u2019s internal affairs (IA) process would find that a given officer was involved in an adverse incident in the next year, excluding internal complaints that DSaPP considered less egregious, such as misuse of sick leave.<\/p>\n<p>CMPD possesses event-level data extending more than a decade into the past. CMPD shared almost all of it with DSaPP researchers, except for officers\u2019 names and badge numbers, officers\u2019 military veteran status, officers\u2019 home addresses, citizens\u2019 names, and text fields. The data were anonymized. Table 1 outlines the number of records and time coverage for each CMPD source. DSaPP also included GIS shapefiles from Mecklenburg County and American Community Survey data from the U.S. Census.<\/p>\n<p>A <i>feature<\/i> is a value used to describe an officer\u2014such as years of experience, number of complaints, and education\u2014that may predict which officers experience adverse incidents. DSaPP created hundreds of features from these data sources, which roughly fell into four categories:<\/p>\n<ul>\n<li><i>Officer demographics,<\/i> such as race or ethnicity, years on force, and training<\/li>\n<li><i>Officer behaviors,<\/i> such as 1-, 3-, and 15-year counts of arrests, citations, uses of force, and secondary employment<\/li>\n<li><i>IA records,<\/i> such as EIS flags and investigations<\/li>\n<li><i>Neighborhood features,<\/i> such as crime rate and the average voter rate in the last 15 years (for the neighborhood in which each incident occurred)<\/li>\n<\/ul>\n<p>To choose the best machine-learning method and estimate its predictive accuracy, DSaPP simulated history. For example, DSaPP used data available on December 31, 2009, to make predictions for 2010 and compared those predictions to actual results; repeated the process by using data available on December 31, 2010, to make predictions for 2011 and compared them to actual results; and so forth, through 2014. DSaPP and CMPD then conducted a field trial. CMPD shared data only through May 31, 2015. DSaPP then predicted which officers would have an adverse incident between June 1, 2015, and May 31, 2016.<\/p>\n<p><big><b>Results<\/b><\/big><\/p>\n<table style=\"float: right;\" width=\"300\" bgcolor=\"ffffff\">\n<tbody>\n<tr>\n<th colspan=\"2\" align=\"left\"><b>Table 2: DSaPP\u2019s system<br \/>\ncompared to CMPD\u2019s system.<\/b><\/th>\n<\/tr>\n<tr>\n<td>\n<table border=\"1\" cellpadding=\"5\" bgcolor=\"669999\">\n<tbody>\n<tr>\n<td><b>Outcome<\/b><\/td>\n<td align=\"center\"><b>Percent Change<\/b><\/td>\n<\/tr>\n<tr bgcolor=\"a3c2c2\">\n<td>Correct flags<\/td>\n<td align=\"center\">+12%<\/td>\n<\/tr>\n<tr bgcolor=\"f0f5f5\">\n<td>Incorrect flags<\/td>\n<td align=\"center\">-32%<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Table 2 shows how DSaPP\u2019s system compares to CMPD\u2019s current EIS on correct flags (true positives) and incorrect flags (false positives).<\/p>\n<p>The results show that moving beyond the current threshold system and using a broader set of data with machine learning improves accuracy. DSaPP\u2019s system flags 12 percent more officers who go on to have adverse incidents, while flagging 32 percent fewer officers who do not. As a result, dozens more officers are correctly flagged and can receive the appropriate training, counseling, and intervention, while departments can avoid using these resources with hundreds of officers who were inaccurately flagged by the existing system.<\/p>\n<p>Figure 1 shows the most important features. There are three types: (1) past behaviors, (2) stress, and (3) patrol neighborhood. First, officers who are routinely found to have been engaged in an adverse incident are likely to engage in another. This is consistent with numerous studies that find past behaviors best predict future behaviors. Second, as members of CMPD\u2019s officer focus group predicted, officers who respond to suicide calls are at higher risk of adverse incidents. The stress caused by responding to suicides and similar incidents (such as domestic violence calls involving children) may put officers at risk. Third, the 3-1-1 (non-emergency government services) call rate for an officer\u2019s patrol area also predicts adverse incidents, which suggests that an officer\u2019s patrol neighborhood is an important feature in determining risk.<br \/>\nWhy each individual officer is flagged can vary. Figure 2 shows five officers the prototype flagged as being at high risk, plus the four most important reasons the system flagged them. Supervisors can use this information to provide officers with the appropriate support. For example, a supervisor might refer Officer 5, who has dealt with many suicide incidents, to an employee assistance program, or a supervisor may look for recent changes in Officer 4\u2019s life, who has multiple risk factors from the last year, to assign an appropriate intervention.<\/p>\n<p><big><b>Other Lessons Learned<\/b><\/big><\/p>\n<p>Humans are almost always the biggest challenge to implementing technical solutions. The White House press release that launched this project as part of the Police Data Initiative (PDI) did not make clear enough that the PDI\u2019s open-data projects and EIS projects were separate. Many officers were understandably upset because they thought DSaPP would make their most sensitive data, including home addresses, available to the public. They were also concerned about CMPD sending data off site without them being made anonymous. CMPD and DSaPP resolved these issues, but not before causing unnecessary anxiety.<\/p>\n<p>To ensure success, DSaPP and CMPD have collaborated closely. DSaPP personnel traveled to Charlotte, went on ride-alongs with officers, and met with numerous CMPD officers and staff, including the department leadership, IA investigators, and mid- and low-level supervisors. CMPD also convened an officer focus group, which included officers who were concerned about the project. DSaPP and the focus group discussed how the system is built and monitored and how it can be improved. Participating officers suggested features that proved to be among the best predictors. Both sides participate in weekly or biweekly phone calls to ask questions, make suggestions, and monitor progress.<\/p>\n<p><big><b>Next Steps<\/b><\/big><\/p>\n<p>DSaPP is extending the system to predict whether an officer will experience specific types of adverse incidents in the next year, such as officer injuries. This information will help CMPD identify which officers are at high risk for which incidents and which risk factors are associated with which incidents.<\/p>\n<p>CMPD and DSaPP are working to implement the system on CMPD\u2019s servers. They are also designing a dashboard that helps everyone in the department consume and use the system\u2019s results.<\/p>\n<p>To help supervisors use the new system, CMPD has convened a supervisor focus group to look at the system and provide feedback. Based on the first round of reviews, DSaPP is working to clarify features\u2019 names and allow supervisors to review results in the system. When supervisors review a flagged officer\u2019s record, they will be able to rate the quality of the flag and the reasons for the flag. The system will use that information to increase its accuracy and to more heavily weigh the features that supervisors think are most useful.<\/p>\n<p>DSaPP is also building a system to predict whether dispatches will result in an adverse incident. CMPD may be able to lower the risk of adverse incidents through better dispatching. Figure 3 shows the most important features for this exploration. The clear outlier is travel time, which appears to have a major impact in predicting adverse incidents. Other significant features include the JST-OCC dispatch typecode, indications that an incident has just occurred, and the officer\u2019s discretionary and overall career arrest rates.<\/p>\n<p><big><b>Conclusion<\/b><\/big><\/p>\n<p>The Los Angeles, California, Sheriff\u2019s Department; the Knoxville, Tennessee, Police Department; and the Metropolitan Nashville, Tennessee, Department have joined this project. DSaPP will adapt the prototype for them and, in the process, learn if it is easy to use this system in other departments, which patterns are consistent across departments, and what data all departments should collect. DSaPP has made its code available for departments to use and invites other departments to join the project.<a href=\"#7\"><sup>7<\/sup><\/a><\/p>\n<p>CMPD wants to make sure they are taking affirmative steps to build stronger relationships with the community, and identifying problem officers is a critical piece of this goal. To achieve this, CMPD partnered with DSaPP to develop a better EIS for flagging police officers who may be at high risk of an adverse incident. Preliminary results suggest the prototype significantly outperforms CMPD\u2019s existing system, allowing the department to more accurately target training, counseling, and other interventions toward officers who are at the highest risk of having an adverse incident. This will allow the department to better allocate resources, reduce the burden on supervisors, and reduce unnecessary administrative work for officers who are not high risk.<\/p>\n<p>In addition, the prototype provides insight into what factors (features) best predict whether an officer is likely to have an adverse incident. Experts identified many of the best officer-, incident-, and neighborhood-level features, but other features the department had not yet considered also correlate with adverse incidents. This information will hopefully allow police departments to develop more effective early interventions. DSaPP also explored the factors associated with adverse dispatches. The results suggest that the characteristics of a particular dispatch are far better predictors than officer features for adverse incidents. &#x2666;<\/p>\n<p><b>Notes:<\/b><br \/>\n<a name=\"1\"><\/a><sup>1<\/sup>John A. Shjarback, \u201cEmerging Early Intervention Systems: An Agency-Specific Pre-Post Comparison of Formal Citizen Complaints of Use of Force,\u201d <i>Policing<\/i> 9, no. 4 (2015): 314\u2013325.<br \/>\n<a name=\"2\"><\/a><sup>2<\/sup>United States Commission on Civil Rights, <i>Who Is Guarding the Guardians?<\/i> (Washington, D.C.: United States Commission on Civil Rights, 1981); Commission on Accreditation of Law Enforcement Agencies, \u201cPersonnel Early Warning System Standard 35.1.15,\u201d in <i>Standards for Law Enforcement Agencies,<\/i> 5th ed. (Fairfax, VA: CALEA, 2001); U.S. Department of Justice, <i>Principles for Promoting Police Integrity<\/i> (Washington, D.C.: U.S. Department of Justice, 2001); IACP National Law Enforcement Policy Center, <i>Early Warning System, Concepts and Issues Paper<\/i> (Alexandria, VA: International Association of Chiefs of Police, 2002).<br \/>\n<a name=\"3\"><\/a><sup>3<\/sup>Samuel Walker, \u201cThe New Paradigm of Police Accountability: The U.S. Justice Department \u2018Pattern or Practice\u2019 Suits in Context,\u201d <i>Saint Louis University Public Law Review<\/i> 22 (2003): 3\u201352.<br \/>\n<a name=\"4\"><\/a><sup>4<\/sup>Shjarback, \u201cEmerging Early Intervention Systems\u201d; Samuel Walker, Geoffrey P. Alpert, and Dennis J. Kenney, <i>Early Warning Systems: Responding to the Problem Police Officer<\/i> (Washington, D.C.: National Institute of Justice, 2001); Robert E. Worden et al., \u201cIntervention with Problem Officers: An Outcome Evaluation of an EIS Intervention,\u201d <i>Criminal Justice and Behavior<\/i> 40, no. 4 (2013): 409\u2013437.<br \/>\n<a name=\"5\"><\/a><sup>5<\/sup>Shjarback, \u201cEmerging Early Intervention Systems.\u201d<br \/>\n<a name=\"6\"><\/a><sup>6<\/sup>Center for Data Science and Public Policy, \u201cProjects\u201d (accessed June 22, 2016).<br \/>\n<a name=\"7\"><\/a><sup>7<\/sup>DSaPP, \u201c<a href=\"https:\/\/github.com\/dssg\/police-eis\">DSaPP Police Early Intervention System: Using Machine Learning to Predict Adverse Incidents<\/a>\u201d (accessed June 21, 2016).<\/p>\n<table cellpadding=\"6\" bgcolor=\"d7e6f4\">\n<tbody>\n<tr>\n<td><b>Crystal Cody<\/b> is the Computer Technology Solutions Manager for the Charlotte-Mecklenburg, North Carolina, Police Department (CMPD). Captain <b>Estella Patterson<\/b> leads CMPD\u2019s Professional Standards Unit. <b>Jennifer Helsby<\/b> is a post-doctoral fellow at the University of Chicago\u2019s (UChicago\u2019s) Center for Data Science and Public Policy (DSaPP). <b>Samuel Carton<\/b> is a PhD student at The University of Michigan\u2019s School of Information. <b>Kenneth Joseph <\/b>is a PhD student at Carnegie Mellon\u2019s School of Computer Science (Pittsburgh, Pennsylvania). <b>Ayesha Mahmud<\/b> is a PhD student at Princeton University\u2019s Office of Population Research (New Jersey). <b>Youngsoo Park<\/b> is a post-doctoral research fellow at the University of Arizona. <b>Joe Walsh<\/b> is a data scientist, <b>Lauren Haynes<\/b> is the senior project manager, and <b>Rayid Ghani<\/b> is the director at DSaPP. <b>Kerr Putney<\/b> is the chief of the Charlotte-Mecklenburg Police Department.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<table width=\"370\" cellpadding=\"9\" align=\"center\" bgcolor=\"669999\">\n<tbody>\n<tr>\n<td>\n<p><b>ACKNOWLEDGEMENTS<\/b><\/p>\n<p>The authors would like to thank the men and women of the Charlotte-Mecklenburg Police Department who have devoted their time and energy to this project. They also thank the Eric and Wendy Schmidt Family Foundation for supporting this work through DSaPP\u2019s Data Science for Social Good program.<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<p>Please cite as<\/p>\n<p style=\"margin-left: .5in;\">Crystal Cody et al., \u201cBuilding Better Early Intervention Systems,\u201d <i>The Police Chief<\/i> 83 (August 2016): 20\u201325.<\/p>\n<p align=\"center\">\n","protected":false},"excerpt":{"rendered":"<p>A few years ago, the Charlotte-Mecklenburg, North Carolina, Police Department?s (CMPD?s) Early Intervention System (EIS) flagged an officer who had been involved in an unusually high number of incidents resulting in injuries. A supervisor reviewed the officer?s records and determined that, while the officer was using force at appropriate times, he was executing takedowns incorrectly. The supervisor sent the officer to the academy for a refresher course. As a result, injuries from the officer?s arrests plummeted, the officer avoided punishment and learned to properly execute takedowns, and the department reduced its legal liability. The officer, the department, and the community benefited from the early notification.<\/p>\n<p>EISs are a staple in U.S. police departments?a 2007 survey showed that 65 percent of surveyed police departments with 250 or more officers had an EIS.1 The U.S. Commission on Civil Rights, the Commission on Accreditation of Law Enforcement Agencies, the U.S. Department of Justice, the International Association of Chiefs of Police, and the Police Foundation have all recommended departments use these systems.2 Most federal consent decrees require a department to use an EIS.3<\/p>\n","protected":false},"author":1,"featured_media":20898,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[144,145,146],"tags":[],"class_list":["post-20630","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-recruitment-personnel","category-research-evidence-based-policing","category-technology"],"acf":{"post_author":"<strong>Crystal Cody, Estella Patterson, and Kerr Putney, Charlotte-Mecklenburg, North Carolina, Police Department<\/strong>\r\n\r\n<strong> Jennifer Helsby, Joe Walsh, Lauren Haynes, and Rayid Ghani, University of Chicago, Illinois<\/strong>\r\n\r\n<strong> Samuel Carton, University of Michigan<\/strong>\r\n\r\n<strong> Kenneth Joseph, Carnegie Mellon University, Pittsburgh, Pennsylvnani<\/strong>\r\n\r\n<strong> Ayesha Mahmud, Princeton University, New Jersey<\/strong>\r\n\r\n<strong>Youngsoo Park, University of Arizona<\/strong>\r\n\r\n&nbsp;\r\n\r\n&nbsp;","legacy_article_id":"4228","legacy_issue_id":"82016","main_category":"Research & Evidence-Based Policing"},"yoast_head":"<!-- This site is optimized with the Yoast SEO Premium plugin v24.4 (Yoast SEO v24.5) - https:\/\/yoast.com\/wordpress\/plugins\/seo\/ -->\n<title>Building Better Early Intervention Systems - Police Chief Magazine<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\" \/>\n<meta property=\"og:locale\" content=\"en_US\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Building Better Early Intervention Systems\" \/>\n<meta property=\"og:description\" content=\"A few years ago, the Charlotte-Mecklenburg, North Carolina, Police Department?s (CMPD?s) Early Intervention System (EIS) flagged an officer who had been involved in an unusually high number of incidents resulting in injuries. A supervisor reviewed the officer?s records and determined that, while the officer was using force at appropriate times, he was executing takedowns incorrectly. The supervisor sent the officer to the academy for a refresher course. As a result, injuries from the officer?s arrests plummeted, the officer avoided punishment and learned to properly execute takedowns, and the department reduced its legal liability. The officer, the department, and the community benefited from the early notification.  EISs are a staple in U.S. police departments?a 2007 survey showed that 65 percent of surveyed police departments with 250 or more officers had an EIS.1 The U.S. Commission on Civil Rights, the Commission on Accreditation of Law Enforcement Agencies, the U.S. Department of Justice, the International Association of Chiefs of Police, and the Police Foundation have all recommended departments use these systems.2 Most federal consent decrees require a department to use an EIS.3\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\" \/>\n<meta property=\"og:site_name\" content=\"Police Chief Magazine\" \/>\n<meta property=\"article:publisher\" content=\"https:\/\/www.facebook.com\/TheIACP\" \/>\n<meta property=\"article:published_time\" content=\"2016-08-01T16:00:00+00:00\" \/>\n<meta property=\"article:modified_time\" content=\"2024-10-07T17:01:04+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1650\" \/>\n\t<meta property=\"og:image:height\" content=\"1293\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"author\" content=\"matrixsuperadmin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:creator\" content=\"@TheIACP\" \/>\n<meta name=\"twitter:site\" content=\"@TheIACP\" \/>\n<meta name=\"twitter:label1\" content=\"Written by\" \/>\n\t<meta name=\"twitter:data1\" content=\"matrixsuperadmin\" \/>\n\t<meta name=\"twitter:label2\" content=\"Est. reading time\" \/>\n\t<meta name=\"twitter:data2\" content=\"14 minutes\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\/\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#article\",\"isPartOf\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\"},\"author\":{\"name\":\"matrixsuperadmin\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/845991022f6502e521826e97f251a3f0\"},\"headline\":\"Building Better Early Intervention Systems\",\"datePublished\":\"2016-08-01T16:00:00+00:00\",\"dateModified\":\"2024-10-07T17:01:04+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\"},\"wordCount\":2768,\"commentCount\":0,\"publisher\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/#organization\"},\"image\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg\",\"articleSection\":[\"Recruitment &amp; Personnel\",\"Research &amp; Evidence-Based Policing\",\"Technology\"],\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\",\"url\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\",\"name\":\"Building Better Early Intervention Systems - Police Chief Magazine\",\"isPartOf\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage\"},\"image\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage\"},\"thumbnailUrl\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg\",\"datePublished\":\"2016-08-01T16:00:00+00:00\",\"dateModified\":\"2024-10-07T17:01:04+00:00\",\"breadcrumb\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#breadcrumb\"},\"inLanguage\":\"en-US\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage\",\"url\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg\",\"contentUrl\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg\",\"width\":1650,\"height\":1293},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\/\/www.policechiefmagazine.org\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Building Better Early Intervention Systems\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#website\",\"url\":\"https:\/\/www.policechiefmagazine.org\/\",\"name\":\"Police Chief Magazine\",\"description\":\"Police Chief Magazine\",\"publisher\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\/\/www.policechiefmagazine.org\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"en-US\"},{\"@type\":\"Organization\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#organization\",\"name\":\"Police Chief Magazine\",\"url\":\"https:\/\/www.policechiefmagazine.org\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#\/schema\/logo\/image\/\",\"url\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/logo-name-blue.png\",\"contentUrl\":\"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/logo-name-blue.png\",\"width\":500,\"height\":246,\"caption\":\"Police Chief Magazine\"},\"image\":{\"@id\":\"https:\/\/www.policechiefmagazine.org\/#\/schema\/logo\/image\/\"},\"sameAs\":[\"https:\/\/www.facebook.com\/TheIACP\",\"https:\/\/x.com\/TheIACP\",\"https:\/\/www.instagram.com\/theiacp\/\",\"https:\/\/www.linkedin.com\/company\/international-association-of-chiefs-of-police\/\"]},{\"@type\":\"Person\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/845991022f6502e521826e97f251a3f0\",\"name\":\"matrixsuperadmin\",\"image\":{\"@type\":\"ImageObject\",\"inLanguage\":\"en-US\",\"@id\":\"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/image\/\",\"url\":\"https:\/\/secure.gravatar.com\/avatar\/25da67e1201b43a6eb59a7b51ba51cfa?s=96&d=mm&r=g\",\"contentUrl\":\"https:\/\/secure.gravatar.com\/avatar\/25da67e1201b43a6eb59a7b51ba51cfa?s=96&d=mm&r=g\",\"caption\":\"matrixsuperadmin\"},\"url\":\"https:\/\/www.policechiefmagazine.org\/author\/matrixsuperadmin\/\"}]}<\/script>\n<!-- \/ Yoast SEO Premium plugin. -->","yoast_head_json":{"title":"Building Better Early Intervention Systems - Police Chief Magazine","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/","og_locale":"en_US","og_type":"article","og_title":"Building Better Early Intervention Systems","og_description":"A few years ago, the Charlotte-Mecklenburg, North Carolina, Police Department?s (CMPD?s) Early Intervention System (EIS) flagged an officer who had been involved in an unusually high number of incidents resulting in injuries. A supervisor reviewed the officer?s records and determined that, while the officer was using force at appropriate times, he was executing takedowns incorrectly. The supervisor sent the officer to the academy for a refresher course. As a result, injuries from the officer?s arrests plummeted, the officer avoided punishment and learned to properly execute takedowns, and the department reduced its legal liability. The officer, the department, and the community benefited from the early notification.  EISs are a staple in U.S. police departments?a 2007 survey showed that 65 percent of surveyed police departments with 250 or more officers had an EIS.1 The U.S. Commission on Civil Rights, the Commission on Accreditation of Law Enforcement Agencies, the U.S. Department of Justice, the International Association of Chiefs of Police, and the Police Foundation have all recommended departments use these systems.2 Most federal consent decrees require a department to use an EIS.3","og_url":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/","og_site_name":"Police Chief Magazine","article_publisher":"https:\/\/www.facebook.com\/TheIACP","article_published_time":"2016-08-01T16:00:00+00:00","article_modified_time":"2024-10-07T17:01:04+00:00","og_image":[{"width":1650,"height":1293,"url":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","type":"image\/jpeg"}],"author":"matrixsuperadmin","twitter_card":"summary_large_image","twitter_creator":"@TheIACP","twitter_site":"@TheIACP","twitter_misc":{"Written by":"matrixsuperadmin","Est. reading time":"14 minutes"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#article","isPartOf":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/"},"author":{"name":"matrixsuperadmin","@id":"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/845991022f6502e521826e97f251a3f0"},"headline":"Building Better Early Intervention Systems","datePublished":"2016-08-01T16:00:00+00:00","dateModified":"2024-10-07T17:01:04+00:00","mainEntityOfPage":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/"},"wordCount":2768,"commentCount":0,"publisher":{"@id":"https:\/\/www.policechiefmagazine.org\/#organization"},"image":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage"},"thumbnailUrl":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","articleSection":["Recruitment &amp; Personnel","Research &amp; Evidence-Based Policing","Technology"],"inLanguage":"en-US","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/","url":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/","name":"Building Better Early Intervention Systems - Police Chief Magazine","isPartOf":{"@id":"https:\/\/www.policechiefmagazine.org\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage"},"image":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage"},"thumbnailUrl":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","datePublished":"2016-08-01T16:00:00+00:00","dateModified":"2024-10-07T17:01:04+00:00","breadcrumb":{"@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#breadcrumb"},"inLanguage":"en-US","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/"]}]},{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#primaryimage","url":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","contentUrl":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","width":1650,"height":1293},{"@type":"BreadcrumbList","@id":"https:\/\/www.policechiefmagazine.org\/building-better-early-intervention-systems\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/www.policechiefmagazine.org\/"},{"@type":"ListItem","position":2,"name":"Building Better Early Intervention Systems"}]},{"@type":"WebSite","@id":"https:\/\/www.policechiefmagazine.org\/#website","url":"https:\/\/www.policechiefmagazine.org\/","name":"Police Chief Magazine","description":"Police Chief Magazine","publisher":{"@id":"https:\/\/www.policechiefmagazine.org\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.policechiefmagazine.org\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"en-US"},{"@type":"Organization","@id":"https:\/\/www.policechiefmagazine.org\/#organization","name":"Police Chief Magazine","url":"https:\/\/www.policechiefmagazine.org\/","logo":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.policechiefmagazine.org\/#\/schema\/logo\/image\/","url":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/logo-name-blue.png","contentUrl":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/logo-name-blue.png","width":500,"height":246,"caption":"Police Chief Magazine"},"image":{"@id":"https:\/\/www.policechiefmagazine.org\/#\/schema\/logo\/image\/"},"sameAs":["https:\/\/www.facebook.com\/TheIACP","https:\/\/x.com\/TheIACP","https:\/\/www.instagram.com\/theiacp\/","https:\/\/www.linkedin.com\/company\/international-association-of-chiefs-of-police\/"]},{"@type":"Person","@id":"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/845991022f6502e521826e97f251a3f0","name":"matrixsuperadmin","image":{"@type":"ImageObject","inLanguage":"en-US","@id":"https:\/\/www.policechiefmagazine.org\/#\/schema\/person\/image\/","url":"https:\/\/secure.gravatar.com\/avatar\/25da67e1201b43a6eb59a7b51ba51cfa?s=96&d=mm&r=g","contentUrl":"https:\/\/secure.gravatar.com\/avatar\/25da67e1201b43a6eb59a7b51ba51cfa?s=96&d=mm&r=g","caption":"matrixsuperadmin"},"url":"https:\/\/www.policechiefmagazine.org\/author\/matrixsuperadmin\/"}]}},"category_objects":[{"term_id":144,"name":"Recruitment &amp; Personnel","slug":"recruitment-personnel","term_group":0,"term_taxonomy_id":144,"taxonomy":"category","description":"","parent":126,"count":210,"filter":"raw"},{"term_id":145,"name":"Research &amp; Evidence-Based Policing","slug":"research-evidence-based-policing","term_group":0,"term_taxonomy_id":145,"taxonomy":"category","description":"","parent":126,"count":131,"filter":"raw"},{"term_id":146,"name":"Technology","slug":"technology","term_group":0,"term_taxonomy_id":146,"taxonomy":"category","description":"","parent":126,"count":393,"filter":"raw"}],"thumbnail_url":"https:\/\/www.policechiefmagazine.org\/wp-content\/uploads\/Graphs-Data.jpg","_links":{"self":[{"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/posts\/20630","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/comments?post=20630"}],"version-history":[{"count":15,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/posts\/20630\/revisions"}],"predecessor-version":[{"id":82353,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/posts\/20630\/revisions\/82353"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/media\/20898"}],"wp:attachment":[{"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/media?parent=20630"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/categories?post=20630"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.policechiefmagazine.org\/wp-json\/wp\/v2\/tags?post=20630"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}