Vol. 8 Issue 10
Patient Classification Systems
Using data to develop ways of providing better patient care while improving a healthcare organization's bottom line
One surefire way to elicit spirited discussion among nurses is to bring up the topic of patient classification (PC) systems. Some despair over attempting to capture the essence of nursing practice with an objective measurement system. But most defend the importance of measuring performance with objective data.
"For decades, patient classification was the unidimensional weighting of tasks. As a result, patient acuity and nurse staffing, to a large degree, became synonymous," said Ann Van Slyck, MSN, RN, CNAA, FAAN, president and CEO of Van Slyck & Associates, Phoenix, and patient classification pioneer. "If [nurses] want patient-focused staffing, who better to sort patients on variables of risk, complexity, severity, time and other indicators?"
Sally Millar, MBA, RN, director of patient care services information systems, Massachusetts General Hospital, Boston, agreed.
"A lot of people in nursing think of patient classification the way it was 20 years ago and think of it as a tool for shift-to-shift staffing," she said. "Instead, we need to use the data to develop ways to provide better patient care and ensure better patient outcomes."
History of PC Systems
According to Van Slyck, the first known patient classification initiative occurred in the 1930s in Manhattan, NY, at what is now Lenox Hill Hospital, when two staff nurses realized that although they had the same number of patients on consecutive days, the patients on the second day were far easier to manage. The pair deduced it was the differences in the acuity of the patients from one day to the next that had altered their workloads.
At that time, nursing did not have the body of knowledge to define, quantify or measure its practice, Van Slyck said.
"Patient classification went through a journey over many, many years," she said. "Nursing was pretty much task-oriented in the 1930s, '40s and '50s, so we turned to our industrial engineering colleagues whose focus was timing tasks" to measure productivity.
By the 1970s and '80s, nursing had matured as a profession and began scientifically distinguishing patients, Van Slyck continued.
"Nurses began to recognize time is not the only variable of differentiating patients, and the process of sensitizing volume by defined nursing care intensity began. So what emerged was a different way of sorting patients, incorporating elements such as risk É complexity of services É and skill level," she said. "The unidimensional timing of tasks appropriately gave way to a multidimensional way of differentiating patients. All of a sudden, weighting of services emerged as the professionally sound, critical-thinking model for concurrent patient differentiation."
By the 1990s, PC systems held thousands of critical indicators, making it easier than ever to classify patients.
"Sometimes when nurses look at classification systems, they say they don't measure everything nurses do," said Millar. "But we don't need to look at everything. In our med/surg tool, which is used on the adult and pediatric general units, as well as critical care, we have 30 discriminating indicators intended to place a patient into the correct patient type so we can measure the nursing workload for that individual."
Nurses at Mass General have been classifying patients for the past 21 years, generating acuity and workload data, Millar noted. "[But] we have never used our data for shift-to-shift staffing decisions. We use it for trending, tracking and budgeting."
While acknowledging patient acuity should be used as one of the variables to influence staffing, Van Slyck emphasized other factors, such as volume, lengths of stay, the judgment of nurse managers, staff competencies, unit geography, the skill mix of staff and dollars also should be part of the staffing equation.
Using Nurse-Authored Data
Once a facility has a PC system in place, the next question is what to do with the data it generates, Van Slyck said. She recommends sharing the information with all staff as a good place to start.
"We should be sharing acuity data with the healthcare team in care conferences, as part of the discharge data, and as part of patient-occurrent reporting," she said. "It should be used in change-of-shift report, in care planning, in physician rounds, and incorporated within the patient chart as well. We can use data for resource deployment, and we also should be using that same data for risk reduction, patient placement and recidivism rates, to mention a few applications."
Linda Bark, PhD, RN, Alameda County Medical Center, Oakland, CA, rejoiced when she started receiving information about patient needs after the facility deployed a PC system. Bark is the hospital's GRASP coordinator. GRASP is the acronym for the workload management system based on the Grace Reynolds Application and Study of PETO, which stands for Poland, English, Thornton and Owens, the researchers who conducted the original time management study of nursing activities.
"The thing I'm most excited about is we can really target acuity increases or decreases within a particular unit," she said. "For example, we can identify whether it's a greater need for coordination of care and consultation with other disciplines, an increase in the number of IVs, or an influx of patients who need more help with ADLs. GRASP fits with the way we think about nursing and helps us explain nursing to others outside of our profession."
Bark anticipates using the data to plan changes in nursing care and hospital systems.
"For example, our nursing consultant told us about a hospital that was considering a new, less expensive feeding tube," she explained. "[But] when they looked at what the nurses would have to do, the change actually would be more expensive. The hospital had the data to look at the change holistically, and they ultimately decided not to buy the product."
Prospective & Retrospective Data
Applications analyst Ellen Walsh, MPA, RN, St. Joseph's Regional Medical Center, Paterson, NJ, talked about the combination of prospective and retrospective acuity data available to nurse managers.
"We created hybrid grids to track by census, with some workload measures behind the scenes, and loaded them into the automated system," she said. "All staffing sheets print-out the day before, and then they're completed manually when we're doing staffing. We then look at the actual and targeted staffing the next day and have the data to explain variances."
Joyce Walls, MSN, RN, associate director for clinical informatics, Mount Nittany Medical Center, State College, PA, uses data collected by a suite of workload analysis software to help determine staffing. But that's not all. "We also are using it retrospectively to validate our budgeting process to make sure we have adequate staffing," she said. "We integrate the data from our patient care and time-and-attendance systems to create the information we need."
Depending on how the data collected by PC systems is utilized, it can directly impact a facility's bottom line, Van Slyck pointed out. "If we can generate data that improves patient care outcomes as well as financial outcomes, it's a win-win situation," she said. "Nursing and finance really do not make strange bedfellows."
Christina M. Graf, PhD, RN, director of patient care services management systems at Mass General, said nurses need to translate the work they do into a language hospital leaders can understand and then weave that data into the fabric of the organization.
"From the day we started classifying patients, we began educating everybody in the organization about this system and about nursing workload," Graf said. "Our senior executives and our finance department actually talk about patient acuity and what proposed organizational changes might mean to nursing workload."
The PC system used by Mass General is different from some others in that it does not look at what nurses have done for the patient; it looks at what the patient needs, Graf said.
"We then look at productivity retrospectively for variance analysis, tracking and trending," she explained. "We use the data for budgeting and projecting, we report unit-based length of stay, describe changes in a patient population, project and analyze variable nonsalary expenses, and use it for evaluating staffing effectiveness to meet JCAHO requirements."
Cathy Turner, MBA, BSN, RN,C, director of implementation and nursing informatics programs for Meditech, has a different point of view about how PC data can be used to improve a healthcare organization's bottom line.
"I don't think it serves hospitals well to measure nursing care after the fact," she said. "What we've done is imbed the patient classification tool into the nursing care plan and the documentation system. We gather retrospective data based on actual care delivered to patients, and that data becomes a way to identify and quantify variances in workload. It allows us to plan staffing for the next shift, to explain overtime and to look at nursing resource consumption."
Because the hospital's PC system provides nurse managers with statistics compiled in a graphic way, they can see how staffing decisions for a particular shift impacted patient care and use that knowledge to plan accordingly, Turner added.
"We can say, 'These units ran into overtime this month because the acuity was high, because we were overstaffed' or whatever the reason was. The data are available for planning the next year [and] identifying patterns and trends over the course of the year."
Sandy Keefe is a regular contributor to ADVANCE.
Patient Classification Systems & Nurse-Patient Ratios
Cathy Turner, MBA, BSN, RN,C, director of implementation and nursing informatics programs for Medical Information Technology Inc. (Meditech), has been in the patient classification field for many years, so she knew what was coming when states like Massachusetts and California started talking about mandated nurse-patient ratios.
"We saw quite a lull in the use of acuity data for awhile in the '90s, and I think the interest heated up once again when the issue of ratios came to the forefront," she explained. "Nurses started looking at acuity data once again because they were well-aware that ratios don't tell the whole story, and that we need more data to effectively manage our resources."
Linda Bark, PhD, RN, GRASP coordinator, Alameda County Medical Center, Oakland, CA, has found nursing workload data archived in the automated medical record demonstrates the hospital's commitment to meeting state-mandated nurse-patient ratios.
"The system gives us solid evidence we're meeting the ratios, since we'll be able to show that a nurse has been assigned to five patients, or two patients, in this system," she said. "That's really important in terms of compliance with ratios."
Sally Millar, MBA, RN, director of patient care services information systems, Massachusetts General Hospital, Boston, said there is a push for mandated ratios in her state; however, the facility does not use ratios as the basis for staffing decisions.
"Ratios only look at the patient in the bed and do not discriminate between patient care needs and how ill a patient is," she said. "Nurses need to look at patient care needs, and we need to factor in the experience and education of the staff, the support services supplied to a particular unit, and the geographical layout of that unit."
Christina M. Graf, PhD, RN, director of patient care services management systems at Mass General, made the case for nursing judgment as the final factor in unit staffing, rather than ratios or patient classification data.
"The nurse managers have the best sense of what's going on and what's needed shift-to-shift," she said. "The manager may use classification data as one piece of the information for staffing decisions, but it is not the driver of these decisions."