Search

Background Timely, precise, and localized surveillance of nonfatal events is needed to improve response and prevention of opioid-related problems in an evolving opioid crisis in the United States. Records of naloxone administration found in prehospital emergency medical services (EMS) data have helped estimate opioid overdose incidence, including nonhospital, field-treated cases. However, as naloxone is often used by EMS personnel in unconsciousness of unknown cause, attributing naloxone administration to opioid misuse and heroin use (OM) may misclassify events. Better methods are needed to identify OM. Objective This study aimed to develop and test a natural language processing method that would improve identification of potential OM from paramedic documentation. Methods First, we searched Denver Health paramedic trip reports from August 2017 to April 2018 for keywords naloxone, heroin, and both combined, and we reviewed narratives of identified reports to determine whether they constituted true cases of OM. Then, we used this human classification as reference standard and trained 4 machine learning models (random forest, k-nearest neighbors, support vector machines, and L1-regularized logistic regression). We selected the algorithm that produced the highest area under the receiver operating curve (AUC) for model assessment. Finally, we compared positive predictive value (PPV) of the highest performing machine learning algorithm with PPV of searches of keywords naloxone, heroin, and combination of both in the binary classification of OM in unseen September 2018 data. Results In total, 54,359 trip reports were filed from August 2017 to April 2018. Approximately 1.09% (594/54,359) indicated naloxone administration. Among trip reports with reviewer agreement regarding OM in the narrative, 57.6% (292/516) were considered to include information revealing OM. Approximately 1.63% (884/54,359) of all trip reports mentioned heroin in the narrative. Among trip reports with reviewer agreement, 95.5% (784/821) were considered to include information revealing OM. Combined results accounted for 2.39% (1298/54,359) of trip reports. Among trip reports with reviewer agreement, 77.79% (907/1166) were considered to include information consistent with OM. The reference standard used to train and test machine learning models included details of 1166 trip reports. L1-regularized logistic regression was the highest performing algorithm (AUC=0.94; 95% CI 0.91-0.97) in identifying OM. Tested on 5983 unseen reports from September 2018, the keyword naloxone inaccurately identified and underestimated probable OM trip report cases (63 cases; PPV=0.68). The keyword heroin yielded more cases with improved performance (129 cases; PPV=0.99). Combined keyword and L1-regularized logistic regression classifier further improved performance (146 cases; PPV=0.99). Conclusions A machine learning application enhanced the effectiveness of finding OM among documented paramedic field responses. This approach to refining OM surveillance may lead to improved first-responder and public health responses toward prevention of overdoses and other opioid-related problems in US communities.

Study objective We demonstrate the feasibility and utility of emergency department (ED) syndromic surveillance using a regional emergency medicine Internet application to minimize impact on ED and public health staffing. Methods Regional (multi-ED) surveillance was established for 2 periods, one characterized by a high-profile national sports event and the other during an international disease outbreak. Counts of patient visits meeting syndrome criteria and total patient visits were reported daily on the secure regional emergency medicine Internet site and downloaded by public health staff. Trends were analyzed and displayed on the secure Web site. ED participants were surveyed about the acceptability and time cost of the project. Results In the first ("All Star Game") project, 8 departments reported daily counts for 4 weeks, covering more than 26,000 patient visits. In the second ("severe acute respiratory syndrome" [SARS]) project, an average of 11 departments in the same region reported daily data on febrile respiratory illnesses, travel, and contacts for 10 weeks. Experience with the first project allowed for rapid implementation of the second project during a 3-day period. In both instances, the surveillance efforts were undertaken without the need for extraordinary ED or public health staffing requirements. Conclusion A regional emergency medicine Internet approach permitted rapid implementation of multisite syndromic surveillance without additional staff. Some problems were identified with the first project, related to clinician checklist completion and manual data tabulation and entry. The SARS project addressed these by simplifying data collection and restricting it to triage.

Three Hundred Sixty Degree Feedback systems, while popular in business, have been less commonly implemented in local public health agencies. At the same time, they are effective methods of improving employee morale, work performance, organizational culture, and attainment of desired organizational outcomes. These systems can be purchased "off-the-shelf," or custom applications can be developed for a better fit with unique organizational needs. We describe the City of Milwaukee Health Department's successful experience customizing and implementing a 360-degree feedback system in the context of its ongoing total quality improvement efforts.

Hospital overcrowding and diversion of ambulances from emergency departments are being recognized as increasing problems in the health care system. This article, a descriptive narrative, examines the various factors contributing to the problem and describes how collaborative approaches to public health issues can be applied. It describes Milwaukee's experience with a collaborative approach. The use of a technological tool to assist with tracking and reporting on ambulance diversion and emergency department overload is explained, and data are provided to show the impact of various methods to blunt the impact of the flu season on diversion frequency. The article encourages use of similar collaborative approaches and Internet-based technology to address other public health problems.

Objectives. This study examined whether differences in heat alone, as opposed to public health interventions or other factors, accounted for the reduction in heat-related deaths and paramedic emergency medical service (EMS) runs between 1995 and 1999 during 2 heat waves occurring in Milwaukee, Wis. Methods. Two previously described prediction models were adapted to compare expected and observed heat-related morbidity and mortality in 1999 based on the city's 1995 experience. Results. Both models showed that heat-related deaths and EMS runs in 1999 were at least 49% lower than levels predicted by the 1995 relation between heat and heat-related deaths or EMS runs. Conclusions. Reductions in heat-related morbidity and mortality in 1999 were not attributable to differences in heat levels alone. Changes in public health preparedness and response may also have contributed to these reductions.

We have proposed needed information management capabilities for future US health departments predicated on trends in health care reform and health information technology. Regardless of whether health departments provide direct clinical services (and many will), they will manage unprecedented quantities of sensitive information for the public health core functions of assurance and assessment, including population-level health surveillance and metrics. Absent improved capabilities, health departments risk vestigial status, with consequences for vulnerable populations. Developments in electronic health records, interoperability and information exchange, public information sharing, decision support, and cloud technologies can support information management if health departments have appropriate capabilities. The need for national engagement in and consensus on these capabilities and their importance to health department sustainability make them appropriate for consideration in the context of accreditation.

Medicine is increasingly practiced through the application of information sciences. Medical informatics deals with optimal information use within bioinformatics, imaging, clinical, and population health domains. Population health informatics plays an important role in that it critically informs practice in each of the other domains. Proper functioning of health care systems requires an advanced health information network that supports clinical care, personal health management, population health, and research. But this infrastructure does not yet exist in the United States. A number of federal initiatives are underway to address this problem, including the development of a framework for a national health information network and funding for implementation. This network will be facilitated by federal leadership, but public and private partnerships, and state, regional, and local implementation and policy development will play a critical role. In this article, we describe several Wisconsin initiatives that are keys to developing a strategic framework and building the state's electronic health information infrastructure.

The City of Milwaukee Health Department piloted a short-term, near real-time syndromic surveillance and communication tool by using an existing secure regional Internet infrastructure. Voluntary, active syndromic case reporting by hospital Emergency Departments was combined with other data streams, including clinical laboratory reports of communicable disease, hospital emergency room diversions, ambulance runs, medical examiner reports of unusual or suspicious deaths, poison control and nursing hotline call volumes, and pharmacy over-the-counter sales. These data were aggregated into a "Surveillance Dashboard" format that was used to communicate community syndromic health trends to hospitals, Emergency Departments, and other providers using a secure Internet technology. Emergency Departments at 8 area hospitals reported a total of 314 cases meeting syndromic criteria from 26,888 patient encounters. Participants were satisfied with data entry and communications. All participating Emergency Departments received e-mail and text pager alerts sent by the Milwaukee Health Department. No unexplained findings or suggestions of an early outbreak were reported through syndrome surveillance for the 4-week duration of the project. Similar surveillance and communications systems could provide multiple benefits to Emergency Department workflow and management, as well as to public health and emergency response.

S i127 The nation’s emergency departments (EDs) are a potential source of surveillance information. The Frontlines of Medicine Project is a collaborative effort of emergency medicine, public health, emergency government, law enforcement, and informatics to develop nonproprietary, standardized methods for reporting emergency department patient data. An initial proposal, published in April 2002, proposed a standardized message structure based on XML (Extensible Markup Language) for reporting triage information from emergency departments to regional surveillance centers and called for reader comments. Subsequently, a consensus conference, with attendees chosen through a modified nominal consensus process, was held to discuss the initial Frontlines proposal and provide recommendations for next steps. Since the consensus conference, an Internet-based Delphi survey technique has been used to refine further the Frontlines recommendations. The technique was utilized for two rounds to yield a consensus exceeding 75% acceptance of the proposed data elements and preferred International Classification of Diseases, 9th Revision (ICD-9)–coded chief complaint values. The data elements for the triage surveillance report include provider facility ID, patient ID, encounter ID, patient age, age unit, gender, date/time first documented in ED, date/time symptom onset, chief complaint, first ED responsiveness assessment, first ED systolic blood pressure, first ED diastolic blood pressure, first ED heart rate, first ED temperature, ED temperature unit, and ZIP codes for home, work, and incident site. The preferred chief complaint categories include 159 complaints arranged in 16 hierarchical categories that are expected to describe the reason for visit in greater than 99% of ED encounters. Further details are available at www.frontlinesmed.org. Foodborne Outbreak Early Detection System (FOEDS) Paul C. Bartlett, Holly Wethington, Bryan DeZeeuw, Sally Bidol, John Tilden, Theresa Bernardo, Lixin Zhang, Dean Sienko, and Mary Grace Stobierski National Center for Food Safety and Toxicology, Michigan State University, Ingham County Health Department, Department of Epidemiology, University of Michigan, Food and Dairy Division. Michigan Department of Agriculture, Bureau of Epidemiology, Michigan Department of Community Health, Information Technology, Michigan State University College of Veterinary Medicine The FOEDS (Foodborne Outbreak Early Detection System) Forum (www.RUsick2.msu .edu) is a structured, Web-based forum that collects and shares data regarding a 4-day food history, food sources, animal contact, and other risk factors that are helpful in establishing the existence of a time-space cluster of possible foodborne origin. It is a syndromic surveillance system that allows users to search the database to evaluate the possibility that a group of people became sick with the same symptoms at about the same time after eating the same food from the same source. The FOEDS Forum is designed to identify suspicious time-space disease clusters that may, at the local health department’s discretion, be worthy of further investigation. As such, it can be viewed as a “front end” to our current national system for identifying and investigating foodborne outbreaks. Data collection was scheduled to begin in October 2002 in the three-county area of Greater Lansing, Michigan. Clinic-based and population-based advertisements were to encourage people with suspected foodborne disease to visit the Web site to determine if they ate the same food that others ate before becoming ill with similar symptoms. Input screens and output reports will be presented, as will program implementation in the three-county pilot area. The FOEDS Forum was developed by epidemiologists from state and local governmental agencies and academic departments working under the umbrella of the National Food Safety and Toxicology Center at Michigan State University.

Lenert and Sundwall identify opportunities and challenges of the Meaningful Use (MUse) incentive programs that advance standardized electronic reporting to health departments at a time when there is limited funding to upgrade systems. We concur that cloud-based Platform as a Service (PaaS) is a possible remedy. However, we disagree with their conclusion that “the security risks inherent in BioSense 2.0’s public cloud implementation may make this effort better suited to a demonstration project than a national level biodefense system.”1 The Afghan version of Conflict Kitchen is Bolani Pazi (www.conflictkitchen.org). Graphic design by Brett Yasko. Architectural design by Pablo Garcia of POiNT. Printed with permission. BioSense 2.02 is a functional public health surveillance system, not a demonstration project, hosted in a highly secure Internet government cloud. Before going live in November 2011, the system was tested, certified, and accredited to meet requirements of the Federal Information Security Management Act and independent security stewards. BioSense 2.0 is a PaaS; offers Software as Service (SaaS) to public health departments; and leverages Public Health Information Network (PHIN) standards and services such as PHIN Messaging, as well as new protocols such as Direct. BioSense 2.0 offers a low-barrier and low-cost approach that health departments can elect to offer to health care providers for reporting syndromic surveillance information. It accommodates secure data sharing among jurisdictions or with the Centers for Disease Control and Prevention (CDC), data exchange standards, jurisdictional information ownership, data sharing agreements, security and privacy rules, and data retention. BioSense 2.0 is user designed and governed in partnership with state and local health agencies, health care providers at the local and federal level, and national associations representing public health officials.3-6 The data sharing model and community-driven governance create a sustainable surveillance ecosystem to maintain situation awareness from local to national levels. The replacement of expensive technological services with a cloud-computing platform has resulted in substantial savings, allowing increased funding for health departments. (More information is available at http://biosense2.org.) CDC continues to work with health departments to enhance their capacity to receive and use information from MUse-prompted expansion in reporting of immunization services, laboratory test results for conditions of public health concern, and syndromic surveillance.7 BioSense 2.0 provides a foundation of services suitable for a future public health platform while simultaneously supporting current MUse requirements.

The sudden emergence of 2009 H1N1 influenza in the spring of that year sparked a surge in visits to emergency departments in New York City and other communities. A larger, second wave of cases was anticipated the following autumn. To reduce a potential surge of health system utilization without denying needed care, we enlisted the input of experts from medicine, public health, nursing, information technology, and other disciplines to design, test, and deploy clinical algorithms to help minimally trained health care workers and laypeople make informed decisions about care-seeking for influenza-like illness. The product of this collaboration, named Strategy for Off-Site Rapid Triage (SORT) was disseminated in 2 forms. Static algorithms, posted on the Centers for Disease Control and Prevention's Web site, offered guidance to clinicians and telephone call centers on how to manage adults and children with influenza-like illness. In addition, 2 interactive Web sites, http://www.Flu.gov and http://www.H1N1ResponseCenter.com, were created to help adults self-assess their condition and make an informed decision about their need for treatment. Although SORT was anchored in a previously validated clinical decision rule, incorporated the input of expert clinicians, and was subject to small-scale formative evaluations during rapid standup, prospective evaluation is lacking. If its utility and safety are confirmed, SORT may prove to be a useful tool to blunt health system surge and rapidly collect epidemiologic data on future disease outbreaks.