William Mohlenbrock, MD, founder and CMO of Verras and a practicing orthopedic surgeon at Scripps Memorial Hospital in La Jolla, Calif., discussed physician variation in a recent webinar presented by Becker's Hospital Review. The webinar, titled "Reducing Physician Variation: How to Maximize Cost Savings by Integrating Each Physician's Best Practices with Hospital Case Management," included four clinical case management focus areas to reduce physician variation, improve quality and save costs. For example, Dr. Mohlenbrock said for a 200-bed hospital with a goal of reducing variation by 8 percent in the hospital's major service lines, Verras projected cost savings of roughly $2.4 million.
1. Hospital and physicians' clinical and operational outcomes variations. Hospitals should start by looking at their top 10 diagnosis-related groups and the top five associated service lines. Hospitals can then divide the DRGs into five levels of acuity based on age and comorbidities to compare the expected mortality and morbidity rates with the actual rates. If the actual outcomes are unacceptable, hospitals should first focus on improving mortality and morbidity and then move on to efficiency. If the hospital has acceptable morbidity and mortality rates for each clinical service, they can then chart the charges and length of stay associated with each patient.
Dr. Mohlenbrock showed a chart Verras uses in which the upper-right quadrant represents patients who had below-average charges and below-average lengths of stay. The bottom-left cases on the chart have the opposite characteristics. By analyzing data on this chart, hospitals can calculate variation between the outcomes of the more efficiently managed patients and less efficiently managed cases and between physicians' outcomes. In one of the medical examples, a hospital had an $80,000 per case variance for pneumonia between the patient groups. In the surgical DRG examples the charge variations were $150,000 per case. These examples are average variations noted over the vast majority of hospitals.
2. Documentation integrity issues. Hospitals need to ensure physicians are documenting cases properly so mortality rates and other metrics are not misrepresented. For example, when showing DRGs by five levels of acuity, with one being the least severe and five the most severe, data often show that there are fewer fives than would be expected compared to the number of ones, twos, threes and fours. "It may mean they're treating people who aren't as sick as the doctors think they are," Dr. Mohlenbrock says. "But that's usually not the case; usually there is a paucity of acuity fives because doctors aren't documenting adequately to accurately assess the acuity of the patients. The negative consequences for physicians are that their mortality rates may look like they are too high." Proper documentation is essential for accurate outcomes reporting and hospital reimbursements.
3. Individual physicians' clinical practice pattern variations. After looking at variation between patients and physicians for cases within a DRG, hospitals can break down the data further to analyze an individual physician's own variation. In fact, data can be broken down to see an individual physician's variation in ordering specific tests, treatments and medicines. Dr. Mohlenbrock suggests having one-on-one meetings with physicians to discuss the causes of variation — whether physician- or hospital-related — and ways to reduce this variation. "It's very empowering to doctors to have objective, data-driven conversations with their chief medical officer and with the operational personnel of the hospital," Dr. Mohlenbrock says.
Physicians can then compare their practices with those of their hospital colleagues and identify common practices for cases in the upper-right corner of the chart (i.e., cases with lower-than-expected costs and LOS); these practices become a clinical pathway. Dr. Mohlenbrock says hospitals should create different clinical pathways for different acuity DRG levels due to their heterogeneity. "Pick patients from the most efficient patient cohort whose acuity scores are fours and fives because they're relatively homogeneous and create a clinical pathway around those. Then use patients with acuities of ones, twos and threes for a separate clinical pathway. But don't try to create a clinical care map for patients of an entire DRG — the patients are too heterogeneous." he says.
4. Cost and quality performance scoring. After implementing clinical pathways derived from a combination of individual physicians' best practices, hospitals can create a Verras Index of Quality Improvement over a three-year period of time that shows improvements in outcomes such as National Hospital Quality Measures, mortalities and morbidities patient satisfaction, reductions in variation and resource consumption (charges of costs). These data can be used to demonstrate to the physicians, hospital administrators and hospital boards a method for an objective allocation of dollars between the hospital and its physicians under a bundled payment model. Dr. Mohlenbrock says the more physicians reduce variation and improve care, the higher percentage of the bundled payment they should receive because their performance will improve Medicare and Medicaid profits for the hospital. Further breakdown of the data can help physicians decide what percentage of payments goes to each specialty. "When you have objective data, you can answer those questions in a non-confrontational manner," Dr. Mohlenbrock says. "That's why having solid data is key."
View or download the Webinar by clicking here (wmv). We suggest you download the video to your computer before viewing to ensure better quality. If you have problems viewing the video, which is in Windows Media Video format, you can use a program like VLC media player, free for download by clicking here.
Download a copy of the presentation by clicking here (pdf).
Reducing Clinical Variation When Treating Sepsis
Reducing Variation in the Use of MRIs
1. Hospital and physicians' clinical and operational outcomes variations. Hospitals should start by looking at their top 10 diagnosis-related groups and the top five associated service lines. Hospitals can then divide the DRGs into five levels of acuity based on age and comorbidities to compare the expected mortality and morbidity rates with the actual rates. If the actual outcomes are unacceptable, hospitals should first focus on improving mortality and morbidity and then move on to efficiency. If the hospital has acceptable morbidity and mortality rates for each clinical service, they can then chart the charges and length of stay associated with each patient.
Dr. Mohlenbrock showed a chart Verras uses in which the upper-right quadrant represents patients who had below-average charges and below-average lengths of stay. The bottom-left cases on the chart have the opposite characteristics. By analyzing data on this chart, hospitals can calculate variation between the outcomes of the more efficiently managed patients and less efficiently managed cases and between physicians' outcomes. In one of the medical examples, a hospital had an $80,000 per case variance for pneumonia between the patient groups. In the surgical DRG examples the charge variations were $150,000 per case. These examples are average variations noted over the vast majority of hospitals.
2. Documentation integrity issues. Hospitals need to ensure physicians are documenting cases properly so mortality rates and other metrics are not misrepresented. For example, when showing DRGs by five levels of acuity, with one being the least severe and five the most severe, data often show that there are fewer fives than would be expected compared to the number of ones, twos, threes and fours. "It may mean they're treating people who aren't as sick as the doctors think they are," Dr. Mohlenbrock says. "But that's usually not the case; usually there is a paucity of acuity fives because doctors aren't documenting adequately to accurately assess the acuity of the patients. The negative consequences for physicians are that their mortality rates may look like they are too high." Proper documentation is essential for accurate outcomes reporting and hospital reimbursements.
3. Individual physicians' clinical practice pattern variations. After looking at variation between patients and physicians for cases within a DRG, hospitals can break down the data further to analyze an individual physician's own variation. In fact, data can be broken down to see an individual physician's variation in ordering specific tests, treatments and medicines. Dr. Mohlenbrock suggests having one-on-one meetings with physicians to discuss the causes of variation — whether physician- or hospital-related — and ways to reduce this variation. "It's very empowering to doctors to have objective, data-driven conversations with their chief medical officer and with the operational personnel of the hospital," Dr. Mohlenbrock says.
Physicians can then compare their practices with those of their hospital colleagues and identify common practices for cases in the upper-right corner of the chart (i.e., cases with lower-than-expected costs and LOS); these practices become a clinical pathway. Dr. Mohlenbrock says hospitals should create different clinical pathways for different acuity DRG levels due to their heterogeneity. "Pick patients from the most efficient patient cohort whose acuity scores are fours and fives because they're relatively homogeneous and create a clinical pathway around those. Then use patients with acuities of ones, twos and threes for a separate clinical pathway. But don't try to create a clinical care map for patients of an entire DRG — the patients are too heterogeneous." he says.
4. Cost and quality performance scoring. After implementing clinical pathways derived from a combination of individual physicians' best practices, hospitals can create a Verras Index of Quality Improvement over a three-year period of time that shows improvements in outcomes such as National Hospital Quality Measures, mortalities and morbidities patient satisfaction, reductions in variation and resource consumption (charges of costs). These data can be used to demonstrate to the physicians, hospital administrators and hospital boards a method for an objective allocation of dollars between the hospital and its physicians under a bundled payment model. Dr. Mohlenbrock says the more physicians reduce variation and improve care, the higher percentage of the bundled payment they should receive because their performance will improve Medicare and Medicaid profits for the hospital. Further breakdown of the data can help physicians decide what percentage of payments goes to each specialty. "When you have objective data, you can answer those questions in a non-confrontational manner," Dr. Mohlenbrock says. "That's why having solid data is key."
View or download the Webinar by clicking here (wmv). We suggest you download the video to your computer before viewing to ensure better quality. If you have problems viewing the video, which is in Windows Media Video format, you can use a program like VLC media player, free for download by clicking here.
Download a copy of the presentation by clicking here (pdf).
More Articles on Verras:
Improving Clinical Variation & Quality: The Role of the CMOReducing Clinical Variation When Treating Sepsis
Reducing Variation in the Use of MRIs