EMS systems exist to provide high-quality care to the patients we serve. Therefore, one of the major goals of a state-of-the art EMS data system is to help ensure that the care you provide is meeting your standards; another goal is to help improve that care over time. In this article, we look at what’s on the horizon for clinical quality assurance and improvement and how EMS systems are using data to integrate with the rest of with the rest of the healthcare system.
Standardizing Process & Outcomes Measures
Electronic patient care report (ePCR) systems have been shown to improve billing and collections, facilitate mandatory data submissions to state and national EMS data repositories, and allow for instantaneous answers to a wide variety of on-the-fly questions that previously took several hours or even months to complete, depending on the number of reports that had to be accessed in file drawers full of paper patient care reports.
However, the ePCR systems in most EMS organizations fall far short of their full potential to facilitate improvements in patient care. The weak link has not been so much in the design of the ePCR data collection software or hardware, but in the reporting. In the past, most of the reporting features provided by software suppliers have tended to focused on billing and activity reports. The activity reports on the clinical side typically include pre-configured reports on metrics such as the numbers of responses, number of medical procedures or medication administrations, success rate percentage for medical procedures, percentage of responses with transports, and response-time intervals.
These reports, although good, don’t directly address the goal of using data for clinical quality assurance. To be clear, clinical quality assurance in EMS is the overall process used by an EMS provider organization and/or regulatory agency to ensure that clinical care meets applicable standards.
One of the major obstacles is configuring the reporting features of the ePCR software in a way that compares care delivered to the applicable standard(s). Because every EMS system can have different standards, developing such reports has been a huge barrier.
Progress is being made, albeit slowly, on developing standardized measures of process and outcomes performance. Protocols are typically built around the ideal for how a process should be carried out (e.g., cardiac arrest, STEMI), but the goals of the process remain the same even if the standards differ.
Example: In cardiac arrest, the outcome goal is patient survival without long-term neurological or other functional damage. One of several process goals, addressed in the design of the protocol, is to minimize interruptions in chest compressions. The applicable standardized outcome measures for cardiac arrest are found in the Utstein criteria for out-of-hospital resuscitation.
The standardized process measure for chest compression interruptions is the compression fraction—the percentage of time that compressions were actually performed during the time the patient was pulseless. The local standard in one EMS system may be a compression fraction of more than 70%, but it may be more than 90% in another system. By standardizing the process measures, each system can set its own thresholds and use the same metrics to monitor its performance on individual cases and in aggregate for the entire EMS provider organization or EMS system.
As the EMS profession matures in its use of quality management, and as the Centers for Medicare and Medicaid (CMS) increases the use of pay-for-measurement and pay-for-performance strategies, more and more process metrics will become standardized and required nationwide. How much an EMS provider organization gets paid by CMS may vary with its use of applicable clinical process performance metrics.
We see this in hospitals with “core measures.” Hospital that do not track and report their core measures get paid less. Among those hospitals that track and report their core measures, the ones that show better outcomes on their core measures get paid more. CMS has stated that core measures will be coming to all aspects of healthcare it pays for—including EMS. The only question is when. Accountable Care Organizations (ACOs) will also be looking for such metrics and the accountability they bring from their EMS system partners.
If we have processes in place to measure our clinical performance against our standards for quality assurance purposes, we have much of what we need to facilitate quality improvement.
In this context, quality improvement in EMS is the overall process used by an EMS provider organization and/or regulatory agency to change processes with the intent of getting better outcomes. This may be in terms of process outcomes (e.g., better compression fractions); better patient outcomes (e.g., higher survival rates); better operational outcomes (e.g., shorter task times); or better efficiency outcomes (e.g., lower cost to get the same process, patient or operational outcomes).
The current level of process performance is assessed using applicable process and outcomes performance metrics. Here is where standardized process and outcomes measures come to the rescue again. Statistical tools are used to compare the control group results with the experimental gtoup results and determine the likelihood that the difference was not the result of simple chance.
This raises the bar for our software tools, the people designing them, the people using them and the people interpreting the results. Our EMS software tools will need the ability to add in modules or updates that utilize standardized process and outcomes metrics. As more metrics are developed and as existing metrics are refined, the software will need to keep up.
The vocabularies of best-practice EMS organizations will be expanding to include terms like dependent, independent, extraneous and confounding variables; statistical significance; statistical power; statistical process control and control limits; and user specification limits. These terms are central to the methods used by top-performing improvement programs and must be among the skill set of whoever you purchase data collection software from.
Obtaining Outcomes Data from Hospitals
Even if we have standardized patient outcome measures, they do little good if we cannot access the outcomes information from the hospital we brought our patients to. This has been a huge barrier in many EMS systems. The good news is that hospitals are being asked by payers for process and outcomes data and to show improvements over time to maximize their revenues (i.e., pay for performance). This is now limited to a handful of core measures, but will be growing very quickly with the implementation of ACOs and associated incentives.
The typical pushback from hospitals when EMS agencies ask for outcomes data relates to privacy concerns. This is a bogus argument. The real issues seem to be more related to the time and effort that hospitals have to expend to give EMS the data it wants. Now that the hospitals have some motivation (requirement) from the payers and hospital accreditation bodies to look at the whole continuum of care and need EMS data to help their own improvement efforts, their resistance has been diminishing.
One method for obtaining outcomes data is based on manual queries. For example, an EMS agency might ask a hospital for the outcome and some event time data for cardiac arrest, STEMI or stroke patients. In the request, the agency would have to include some information to allow the hospital staff to look up the patient record in their hospital medical record system.
The problem: Protected health information (PHI) is usually needed, such as patient names, dates of birth, addresses, etc. Sending PHI back and forth via email creates security risks. Encryption or other mechanisms may be needed to address those concerns and keep the whole effort compliant with HIPAA and other regulations. Common workarounds include crews recording a hospital chart number on the EMS report, or doing the queries over the telephone or in person to avoid having PHI exchanges in writing.
So long as the volume of such queries is small, manual methods can work. For larger EMS systems and hospitals, however, this approach may require dedicated staff time—and expense.
A complicating factor to consider is when the EMS data set needed for quality assurance or improvement purposes involves more than one data system or more than one organization. Inside a single EMS provider organization, clinical process metrics may require information from the ambulance communications center, which may not be integrated with the ePCR data and therefore requires a process to pull data from both sources and place it into a separate file.
Alternatively, the data from multiple sources can be linked using relational database systems from which queries and reports can be generated. If more than one organization is involved, such as a 9-1-1 center, a medical first-response provider and an ambulance provider, the process becomes complicated.
The databases you may want to query or link to are often out of any one organization’s control. Like the hospitals, this requires organizational cooperation. Also like the hospitals, that’s much more likely to happen if all organizations with needed data have an interest in the performance metrics or share accountability for quality assurance and improvement efforts.
To minimize the labor on the hospital side, some EMS systems rely on getting copies of data the hospital has already collected for its own internal purposes. For example, if the hospital’s cardiac catheterization laboratory participates in the ACTION Registry for acute coronary syndrome cases, the data that EMS is interested in—and that the hospital typically needs from EMS—are all included in the ACTION registry data set (www.ncdr.com/webncdr/action).
Because the hospital has already made the queries to enter their information into the ACTION registry, EMS agencies can help provide their data, and then both entities can share the completed data sets for patients they mutually cared for. Similar approaches are possible for stroke and other cases where registries are utilized.
Ideally, all of the organizations involved in the continuum of care participate together on standing quality assurance/improvement committees and collaborate on ad hoc quality improvement project teams.
Electronic Data Exchange
The downsides of getting outcomes and process data from multiple organizations are obvious in terms of time, expense and hassle. The newest approaches use software to help the process and outcomes data move between organizations more efficiently. There are two primary methods to accomplish this: data file interoperability and health information exchanges.
Data file interoperability uses a standardized data format to send and receive information between provider organizations. Health Level 7, a Standards Developing Organization, developed the HL7 Clinical Document Architecture (CDA) standard (www.HL7.org). This standardized format is recognized worldwide for interoperability of health information technology.
The National EMS Information System (NEMSIS) has worked with HL7 to develop a NEMSIS Version 3-specific CDA to promote the exchange of EMS data with hospitals and the rest of the healthcare industry.
Using this approach, EMS provider organizations send a file of data elements in the specified format to the hospital. The information is used to place a copy of the EMS PCR into the hospital’s medical records system.
In return, software on the hospital side can be programmed to send a compatible file back to the EMS agency containing the outcomes information for the patient after they leave the emergency department (ED) or are discharged from the hospital. The NEMSIS Version 3 HL7 CDA has been successfully implemented in two hospitals through a collaboration between ZOLL Data Systems and the NEMSIS Technical Assistance Center.
A more robust approach uses an intermediary computer system called a health information exchange (HIE). Think of an HIE as a data “middleman” that attaches to the computer systems in one or multiple organizations, providing a portal for each organization to search and retrieve information on patients across all of the participating organizations. Multiple hospitals and groups of individual physician offices or clinics that participate in an HIE can share records on the same patient—to obvious advantage. If EMS is included in the HIE, sharing data back and forth for continuity of care and quality assurance/improvement is readily accomplished.
The HIE infrastructure includes appropriate security controls so individuals and organizations can only access the information they’re authorized to (e.g., ambulance services are only able to search and retrieve data on their patients).
Imagine a process that queries the 9-1-1 communications center, the medical first-response provider, the ambulance provider and the hospital to assemble all of the data needed on an episode of care and then generates specified reports and pushes them out to designated recipients—automatically. That’s the power of an HIE.
Healthcare System Integration
Almost every effort to improve the efficiency and effectiveness of healthcare delivery involves healthcare integration—smoothing out the rough spots when a patient transitions from one organization, or care unit within the same organization, to the next. The many rough spots in those transitions are the source of untold numbers of problems and enormous costs.
Most healthcare system integration issues are related to data. Example: The 9-1-1 communications center gets a call. Somehow, it needs to pass the caller and whatever information it collected to the medical first-response agency dispatcher and/or the ambulance dispatcher. One of those dispatch centers may be the provider of emergency medical dispatch services to perform medical triage and give pre-arrival instructions to the caller. Some of that information must be passed along to the responding crew(s). That’s a transition point with several potential rough spots. If the medical first responders arrive first, they will gather information as they make their initial assessment and begin care. When the ambulance arrives, the information must be passed along to them—another transition point and lots of other potential rough spots.
And so it goes, from the ambulance to the ED to the specialty-care units (e.g., cardiac cath labs, trauma center, stroke center, ICU, etc.) to the general wards to the attending physician(s) and to any rehabilitation services.
The software tools to manage data across the transition points are just one element of healthcare integration. Healthcare system integration also involves the same patient in multiple care settings on multiple occasions.
A well-integrated healthcare system would allow each provider at each transition point to have access to pertinent information in real time. Example: A call is made to 9-1-1 and transferred to medical first responders and the ambulance service. They both can access pertinent information on prior calls to the same patient. They might also see alerts (“do not resuscitate” orders, or other advance directives) that are important to know before making patient contact.
In this model, the crews would also be able to access the patient’s past history, medications, allergies and emergency contacts. When considering the right destination, they would consult information on the in-network hospital and its ability to care for the patient’s current condition. If the patient is low severity, perhaps they don’t go the ED. It may be best to make an appointment for follow-up with their family physician or specialist. This starts to engage the processes of care commonly discussed in community paramedicine programs. The value and utility of an HIE to facilitate healthcare integration is what’s likely to make them popular as the economic benefits and regulatory mandates to do so increase.
The Possibilities Ahead
The gathering of healthcare data can make it possible to do things we could only imagine in the past. Consider the HIE being able to look at data on a diabetic patient from 9-1-1, medical first response, physician offices, EDs, hospital specialty-care units, general wards, and rehabilitation facilities. It would be possible to spot dangerous trends that would never be detected by any other means and alert the patient and appropriate care providers to allow intervention sooner—with better outcomes and at a lower total cost.
And consider looking not just at one patient, but an entire community. Trends may be spotted to detect the emergence of geographic or time-associated patterns of disease or injury.
We see the very basic elements of this type of trend detection capability in syndromic surveillance software systems used with public health data and emergency medical dispatch data.
The bottom line: Technology can and will do amazing things for the treatment of individual patients and the advancement of mobile integrated healthcare. We have a long way to go before the systems that need to communicate seamlessly can do so, but the promise of better care at lower costs should compel each of us in EMS to strive for standardization of process measures, sharing of data with all organizations within the healthcare continuum, and the development of reporting tools that help us use this data to its greatest potential.
The data is out there—it is up to us to figure out how to share and use it.