The path to improved results: a Colorado hospital uses clinical decision support tools to lower LOS and its rate of unplanned readmissions and patient complications.

In the quest to improve quality of care and control costs, hospitals continuously incorporate the latest medical knowledge, healthcare standards and best practices into their treatment protocols. Reliable, detailed, easy to-use clinical data is required to guide these actions and measure and evaluate their outcomes.

Valley View Hospital, a Colorado community hospital nestled between Aspen and Vail, understood the importance of reliable, actionable information as it began creating clinical pathways (standardized treatment plans) to bolster the quality of care and improve operational efficiency, However, the hospital's manual process for collecting and analyzing clinical data was cumbersome and slow, says Cathy Desautels, R.N., director of patient information at Valley View Hospital Worse, the hospital didn't have complete confidence in the consistency and validity of its data oil unplanned readmissions and complications of care, because of limitations inherent in the manual review of patient charts and processing of data.

Manual Monitoring

To evaluate whether pathways were delivering the anticipated results, the Valley View team wanted to quantify, monitor and compare length of stay (LOS), charges, unplanned and avoidable readmissions, complications of care, mortality, utilization of resources and other indicators. This required staff to manually enter each patient's demographic data, clinical indicators and other information into a spreadsheet which the hospital's IT department, could use to build a database for generating charts illustrating medians and trends, But neither reliable comparisons nor severity adjustment methodology were available, according to Desautels.

Valley View specifically looked for unplanned, avoidable readmissions that occurred within 30 days of the previous admission. However, what defined "unplanned and unavoidable" was quite subjective during the manual review, Desautels says, "and topping that with cross-referencing 30-day increments for each patient was time-consuming, even with automation." As a result, it was difficult to evaluate the clinical pathway initiative, to compare outcomes for patients following alternate treatment plans or to identify areas with the potential for clinical improvements and cost savings.

The manual-based system challenged staff to the fullest when the hospital monitored only one clinical pathway, Desautels says. As Valley View added more--for a current total of 26--manually monitoring and evaluating pathways became virtually unmanageable. For instance, if manually abstracting and entering the data from one medical record required 15 minutes, that figure is compounded by the number of patients on the pathway and the number of pathways being implemented.

Unless it found another option, the hospital would have to hire more staff or scale back efforts to monitor clinical pathways; both options stood to compromise quality of care and efforts to contain costs. "It was unrealistic to continue to monitor pathways using a manual approach," Desautels says.

Smooth Automation

Valley View joined other independent hospitals affiliated with Quorum Health Resources to find an outcomes data management and analysis solution. After a review of the products available, the team selected three vendors to evaluate.

Valley View chose Medstat's Outcomes Analyst application, which is now part of its Performance Expert solutions, because it was fully automated and flexible, uploaded data directly from a hospital's mainframe and accommodated many of the clinical indicators hospitals needed to monitor. The system allows data to be analyzed by DRG, DRG product line, ICD-9 diagnosis and procedure groups, severity levels and disease stages.

Clinical outcomes information generated by the new system is detailed, consistent and validated. Reports are easier to understand and present to administration, physicians, the hospital board and various clinical committees. Physicians accept the validity of the outcomes measures, because they were adjusted to reflect the severity of a patient's condition and comorbidities.

Automating the evaluation of clinical pathway monitoring has benefited patients, caregivers and the organization as a whole, Desautels says. For instance, Valley View has realized a significant decrease in average LOS as a result of many pathways having been implemented has occurred. Patient benefits from that reduction include reduced costs, reduced exposure to possible hospital-acquired infections, and increased satisfaction due to knowing in advance the predictability and timeframe of their hospital course. Physician benefits include confidence in knowing that the standard of care is being provided in a convenient, mapped-out format. The hospital benefits from a financial perspective because it can more effectively predict charges and more appropriately deploy and utilize resources.

Branching Out

Valley View had outcomes information with the depth and breadth needed to guide clinical decisions that directly affected patient care and operational efficiency. Building on the success of the clinical pathways project, the hospital used Medstat information to support other key initiatives:

Augmenting Training. When physicians and clinical staff are trained to implement a new clinical pathway, pre-pathway outcomes data and benchmarks are shared to establish expectations, standards and goals. Physicians and clinical staff are presented with pre-pathway data on charges, LOS, complications of care (specific complications), unplanned/avoidable readmissions, and mortality that is provided by Outcomes Analyst, according to Clinical Pathway Coordinator and Clinical Data Analyst Linn Kight, R.N. This process helps staff to understand why a specific disease or procedure is chosen for a clinical pathway. Clinicians are also educated on the hospital's current practice patterns, and on changes in care based on care standards or practice guidelines included in the pathway development.

Guiding Peer Review. Quarterly, Valley View clinical analysts review complications of care to identify outliers. When they occur, an administrative review is conducted to determine whether peer review by the hospital's medical care evaluation committee is warranted. A list of indicators is identified and approved by the medical staff, and on a quarterly basis, Valley View runs reports for those indicators by individual physician.

Confidence levels are established by comparison with external databases within the Outcomes Analyst system. If an individual physician unfavorably exceeds that confidence level, all of those cases are reviewed for patterns. For instance, if the upper confidence level for bladder perforations during hysterectomies in a six-month period is two, and an individual surgeon experiences one perforation, then no chart review takes place. If three occur, however, each of those cases is reviewed in a focused fashion for trends, patterns and improvement opportunities.

Supporting Negotiations. Benchmarking Valley View's severity-adjusted charges against other hospitals has provided valuable insights for negotiations with managed care organizations over fees for specific conditions. For instance, Valley View can clearly show its charges and LOS for all patients compared with other hospitals' patients for the same conditions. "The hospital's data more often than not," says Desautels, "is on the favorable side of other organizations with which we are comparing, and allows for latitude within which to negotiate costs and quality issues."

Meeting State and Federal Requirements. Data from Medstat assist the hospital in complying with reporting requirements from the Joint Commission on Accreditation of Healthcare Organizations (JCAHO), Centers for Medicare and Medicaid Services (CMS) and the state of Colorado. With the ability of detailed patient type in the Outcomes Analyst application, Valley View is able to identify patients on specific clinical pathways and especially those areas required by CMS and JCAHO which include specific clinical indicators for acute myocardial infarction, heart failure, community-acquired pneumonia, and pregnancy and related conditions, according to Kight.

Because of its success with Medstat, Valley View chose the company as its vendor for CMS and JCAHO reporting. Clinical pathways for these DRGs and clinical conditions were already in place, and Valley View had already been tracking this data before reporting was required. The hospital is also required to monitor complications in trauma patients for state reporting. Valley View uses Outcomes Analyst to identify these patients and the complications of care component to drill down to specific cases and monitor for trends that may be potential areas for improvement.

Realizing Results

Quality of care improved and costs declined after pathways and the Performance Expert Outcomes Analyst application were implemented. For example, results for the community-acquired pneumonia pathway include the following improvements:

* Unplanned readmission rates declined from 36 percent in 1996 to approximately 7 percent by the end of 2002. This marked improvement accompanied equally impressive declines in charges and length of stay--findings which confirmed that high quality and cost-effectiveness were being achieved together.

* Complications of care declined from 15 percent in 1996 to about 5 percent currently. This is an important indicator, because patient complications and errors cost an average of $4,700 per inpatient stay.

* LOS decreased from four to six days in 1996 to two to three days in 2002. Valley View's results are well below the national average for this indicator.

* Charges have declined an average of $2,000 per admission. This helps the hospital's revenue picture, because it creates a more favorable reimbursement rate.

At the end of 2002, using clinical pathways that apply to more than 50 percent of Valley View's patient population, the hospital has seen an overall decrease in charges of $3,721 per pathed patient, a decrease in the length of stay by 1.5 days per pathed patient, an 11.5 percent decrease in the number of patients with complications and a 2 percent decrease in unplanned/ avoidable readmissions.

For more information about Performance Expert from Medstat, www.rsleads.com/312ht-205

SOURCES

Cathy Desautels, R.N Director of Patient InformaLion Linn Kight, R.N. Clinical Pathway Coordinator and Clinical Data Analyst Valley View Hospital Glenwood Springs, Colo.

PRODUCT/COMPANY

Performance Expert Medstat Ann Arbor, Mich. www.medstat.com