Data helps make better decisions, but for this to be true, that data has to be reliable.
We all appreciate the need for accurate data. Usually we focus on the entry aspect of this and pester our people to please, please, please enter all information they have on a response in a timely and accurate manner.
This is a chief officer’s plea to all companies many times a year. Once the plea is made, we expect the company officers to go forth and do good things or at least ensure their people enter it accurately.
The other aspect of accurate data is its extraction from our databases. This component is less reliant on entry accuracy, but nonetheless, it’s as critical with regard to accuracy.
Data extraction depends on two primary variables:
- The strength of software analytics
- The awareness of staff who are performing the analysis
It’s important that we recognize and acknowledge some points of concern.
Software analytics have developed tremendously over the last five years, and they continue to improve. However, there has been a simultaneous shift in patient-care reporting that has moved data collection from being primarily incident-based to patient-based, and rightfully so.
The complication is that many of us still want, need and expect to review data from the perspective of calls for service. That is, what, where, when and how often?
This shift from an incident perspective to a clinical perspective in data processing requires a shift in our perspective as well. The benefit in this shift of perspective has been most noticeable in my agency’s quality-assurance program. We have gradually evolved from analyzing response data about general district-resource utilization into a program that intently tracks time-critical diagnosis datasets and skill-utilization rates. These metrics range from a full-range agency perspective down to individual provider considerations.
Data Extraction and Reporting Processes
Staff involvement in data extraction is closely related to the agency’s records system, and reporting accuracy is directly relative to staff’s awareness of datasets. There are many avenues to minimize errors in data collection, but the familiarity of staff with the data source has a direct impact on their awareness of erroneous reporting and recognition of general typographical errors that cause these issues.
Although individual agencies have reporting requirements to state and national systems, we’re still afforded the ability to collect data specific to our own system’s design and performance measurements. It’s in this regard that I examined the workflow-process demand placed on our staff. This is specific to EMS data collection and analysis.
Our system uses routine collection of 170 data points on a routine patient encounter. This is the source of about 15% of in-depth assurance reviews that consume approximately 15 minutes per chart; when aberrant interventions or performance is identified, these progress to consume another one to three hours of staff time, depending on many variables.
Data relative to the performance of our entire system is aggregated monthly through 154 datasets that provide 1,991 points of analysis and are made digestible to the general population in 154 charts and graphs. This cumulative effort of data extraction and analysis consumes about 20 hours per month, the equivalent 12.5% of a fulltime employee.
The question becomes, “Can we make more efficient use of our staffs’ time to provide what should be routine and readily available data?”
The primary points discussed to this point are specific to routine training and awareness that have been integrated into my agency’s data-digestion process. The learning point for us began years ago when we began to use data for decision making in place of using anecdotal reporting.
Data extraction from our records was far from intuitive; I believe the only reason I was somewhat competent in this was that I had a programming background. As I began to ask for more information, it became apparent we had to attend specialized training. As an agency, we invested considerable time and funds toward this effort. The fruit of this labor has been accurate reports and ultimately the best decisions.
Why have I brought forward such seemingly intuitive points here? Every one of us reading to this point already understands that we have undergone a shift from solely collecting response data, such as what our busiest ambulance from date X to date Y was, and we also certainly understand the importance of accurate data extraction.
My overarching point here is, yes, we do understand these points, but it’s essential that we ensure our staff understands them as well as and even better than we do. Unless we’re fortunate enough to have a powerhouse software that extrapolates all pertinent data points to guide us in making programmatic and system-wide decisions, we must have an educated staff to provide us with this data to make such important decisions.