Simple variance tracking programs yield better data and MD relations

Scannable forms are the trend

The key to designing an effective variance tracking tool that provides useful data and influences physicians lies in simplicity. That's the advice Larry Strassner, MS, RN, CNA, director of critical paths at The Johns Hopkins Hospital in Baltimore, and Derek Spellman, MSN, RN, CS, chief nursing officer at TENET's Brookwood Medical Center in Birmingham, AL, presented to attendees at the First Annual Hospital Case Management Conference held last month in Atlanta.

Staff nurses at 1,000-bed Johns Hopkins document variances once daily during the night shift using a simple optical scanning form, says Strassner. The nurse answers a series of questions by filling in the appropriate response, completes a daily summary, and then scans the information for concurrent data analysis.

Information from the variance sheet is used to compile a daily progress form for each patient. Each variance's yes or no answer is included along with a comment for nursing and medical staff. The report also is included with the pathway in the patient's medical record, with the exception of the comments section, which is omitted for risk management reasons.

Each pathway team at Johns Hopkins, rather than administration, determines which variances will be collected for the specified patient group, says Strassner. Variances are phrased in a yes/no format with the type of variance coded for each response. Each pathway usually has no more than eight questions for each day, Strassner notes.

Variance types are divided into patient/family, practitioner, system, and community categories. "Daily, rapid response to deviations maintains the integrity of the data and allows physicians and clinicians to address problems as they occur," Strassner explains.

Pathway development teams can revise the variance questions and have updated scannable forms at the bedside within 72 hours of any changes, notes Strassner. Scanning equipment, manufactured by OptiMark in New York City, has been installed on each unit at an average cost of $16,000 per unit.

The 586-bed Brookwood facility incorporates several computer information systems into its physician data dissemination process, but the No. 1 source for useful, accurate information is critical paths, he says.

Brookwood's case managers provide hands-on care for patients and collect variance and other clinical information, while outcomes managers analyze and disseminate data to physicians. "The outcomes manager can immediately look at a report and tell if the data are wrong, because they are familiar with the physician's practice," he explains.

Physicians already realize the need for accurate data because consumer, regulatory, and managed care interests are requesting such data, notes Spellman. Once credible data are reported to physicians, they will begin requesting how to provide it. "Before long, you'll be trying to keep up with all the requests for reports."

In fact, Spellman recommends case managers provide physicians with an overview of their practice over time rather than breaking the data down into diagnosis-specific areas. "Then have the clinical nurse specialist most familiar with their practice talk with them about the data," says Spellman.

Individual areas where a physician's practice pattern indicates possible improvements are reported to the physician in a summary report. Each item is listed on the left side of the page, and a corresponding suggestion is listed under an "opportunity" column. Physicians who order 24-hour urine collection for creatinine clearance, for example, are provided with two suggestions:

1. Eliminate the test.

2. Collect the sample in an outpatient setting.

Physician data at Brookwood are shared only with the individual physician. Medical department chairpersons do not receive copies of individual physicians' reports, notes Spellman. Data should be presented individually with the outcomes manager or clinical nurse specialist and should never be presented in a threatening manner, he advises.

Spellman also suggests that small sample sizes are acceptable to disseminate to physicians because timely data are important. And provide internal and external benchmark data that don't identify individual physicians by name to give the physicians points of reference and comparison. *