To take guesswork from decisions, check out activity-based costing

Manipulating correct data gives you a real-world view

The resources of ICU managers are shrinking, the staff mix is in fluctuation, and even the types of cases you handle are as permanent as your hospital's current contracts with managed care companies. So how do you bring any semblance of staffing and administrative organization to such chaos? By basing your decisions on the numbers, experts say.

The day is gone when staffing and other decisions can be based simply on subjective experience. And even most accounting and decision-support systems can't itemize resource requirements and costs at the patient level, says Joyce Howe, RN, project manager of the team that reorganized the way administrative decisions are made at the Sentara Health System in Norfolk, VA. Her duties included six months of research of activity based information systems.

"Accounting systems, even those declaring to be activity-based, did not capture cost information at the right level and typically used cost and charge data synonymously," she says. "Other products fell short in terms of providing clinical managers with the necessary information for budget development and for making day-to-day staffing decisions."

Sentara's solution was to move to activity-based costing, and it selected a computer tool called JMT BASIS designed by Joule Manage-ment Technologies in Atlanta. The system contains a database customized from Sentara's clinical paths that can analyze costs based on task requirements, rather than hospital charges.

"It allows us to file costs by tracking activities across the care continuum, within a care delivery setting, by care delivery category and, by staff position, including RNs, secretaries, housekeepers, and lab technicians," Howe says.

For example, all 11 intensive care units in the four-hospital system spent about six months establishing baseline data that "reflect work as it is done today," Howe explains.

Gathering data on ICU performance

Each ICU, which Sentara organizes according to service lines called operating centers, collected demographics from the previous year such as DRGs, patient volume, and lengths of stay. ICU managers grouped patients into the types of treatment needed and used the computer tool to document care requirements for each patient group, based on clinical pathways previous developed. (See list of neuro-intensive care patient groups, at right.)

They developed more than 150 patient summary templates that tracked the frequency, duration, and number of staff required for various categories of ICU care. activities. (See copy of a patient summary for neuro-intensive care, inserted in this issue.) In addition to critical care, areas documented included:

· admissions/pre-admissions;

· assessments;

· care team communication;

· clinical documentation;

· diet;

· nutrition services;

· personal hygiene;

· education/teaching;

· general secretarial;

· laboratory processing;

· laboratory specimen collection;

· tests;

· medications;

· respiratory therapy treatments;

· vital signs;

· laundry and linens;

· materials management;

· environmental services.

The computer tool also allowed managers to track these activities at the following four stages of the intensive care continuum:

1. admission;

2. intubation;

3. extubation;

4. transfer or discharge.

Each patient template also noted the length of stay in each stage as well as the total number of minutes, the percentage of total minutes, and the work minutes per hour for each stage and activity.

That baseline data then allowed intensive care managers to base their decisions regarding staffing and other issues on hard evidence rather than guesswork or subjectivity, Howe notes. "This tool gives us a completely different way to think about delivering patient care."

Historically, she explains, hospitals have costed out nursing care by attaching it to a room and bed. "The typical workload standard - hours per patient day - makes the assumption that all patients require the same care. But the amount of care varies as to the patients' diagnosis as well as where they are in their stay," Howe explains.

The computer tool provides a level of detail that allows managers to forecast accurate budgets and staffing needs because resources are linked to patient type and stage - even the more unusual ones.

"Although a neuro-intensive care unit doesn't get on a regular basis a patient with severe head trauma who is a candidate for organ donation, when we do, [such a case] consumes a huge amount of nursing resources," she explains.

By knowing how many such candidates the unit has admitted in the past year and the amount of resources needed, managers can incorporate those figures in the budget, Howe points out.

Redesigning patient assessments

One of the first practices Sentara questioned was patient assessments. The computer tool showed that patient assessments accounted for 20% of total annual labor dollars. Scrutiny showed that assessments occurred with the same frequency and comprehensiveness across the continuum of care, regardless of acuity, even though an improving patient needs less-comprehensive and less-frequent assessments.

Research showed that changing that practice alone could bring a workload reduction of 8,000 hours and a savings of $150,000 annually on the step-down unit. It would also cut documentation time by 2,900 hours, resulting in a $54,000 savings, Howe says.

When Sentara's redesign team used the program to analyze what tasks occurred, it discovered that some tasks were scattered throughout the day unnecessarily. To increase staff efficiency and reduce the number of patient interruptions, Sentara bundled activities and reduced their frequency if deemed redundant. For example, catheters and chest tubes were checked, dressings changed, and cough and debridement activities were all done at different times of the day at four- or 12-hour intervals. Now, they are all performed twice a day at the same time, once during the day and once during the evening shift.

The change resulted in a 50% reduction in labor hours. Howe says changes of this type could result in an overall reduction of up to 30% in care delivery labor hours.

Other processes the tool has brought into question include the transferring of patients to different care levels as their status changes. "You can start seeing costs that have little or no value, Howe says. "If a patient is only in the hospital for three or four days, it may be cheaper not to transfer. There are hidden costs such as transfer orders, and there are new caregivers who have to familiarize themselves with the patient."

The system also allows Howe to play "what if" with various scenarios. "We can see what happens with costs and workload on paper before we implement them in reality," she says.

In September, the burn/trauma and neurosurgical intensive care units will go live with a redesign modeled on the computer tool, says Pam Streetman, RN, operating center manager.

"Our hours per patient day haven't been re-verified in about 10 years, so we needed to look at our functions and tasks," she says. "

As Streetman concentrated on identifying what RNs were doing - bedside tasks - and what they should be doing - assessments and care plans - she found that some functions could be bundled with the tasks of "other" to better organize time.

"It was a real eye-opener to see the time we RNs were allotting to each task. We're always saying,'we're so busy, we need more help,' when part of the problem is that we need to delegate some tasks to others," Streetman says.

For example, in the new neurosurgical ICU plan, respiratory therapists will be doing lab work, setting up pressure lines, and handling cardiac outputs.

Howe's team is also working to smooth out the peaks and valleys of workflow. "We're in the analytic phase now, examining what an average day looks like," she explains. "Traditionally, hospitals schedule too much work between 7 a.m. and 11 a.m., so we want to find out what time of day the work really needs to be done so we can schedule more efficient staffing."