The Marshfield (Wis.) Clinic, with 800 physicians at 50 sites, was one of the highest achievers in the model program for accountable care organizations, the Medicare Physician Group Practice Demonstration, which lasted from 2005 to March 2010. Here Theodore A. Praxel, MD, medical director for quality improvement & care management at Marshfield, describes what the demonstration required, how Marshfield responded and how the experience might compare to ACOs.
Shared savings. The 10 clinics participating in the demonstration shared savings they achieved with CMS. To determine how much was saved, Marshfield's costs were compared with costs of caring for Medicare beneficiaries in the same region who did not receive any care from Marshfield. Organizations saving 2-7 percent of the control group's costs were paid 80 percent of the savings. They were not rewarded for savings under or above that percentage. Marshfield was one of only two participating groups that earned shared savings in all three years. The other was the University of Michigan Faculty Group Practice. In years 1-3 Marshfield won a little over $23 million in shared savings. The money it earned went to support infrastructure and the needs of patients, Dr. Praxel said.
Quality measures. Participants had to achieve certain quality levels. To measure quality, the study looked at a mixture of efficiency and quality for patients in a number of chronic care categories that expanded over time. The quality measures started in year one with 10 diabetes measures. In year two, 17 coronary artery disease and heart failure measures were added, for a total of 27. In year three, five hypertension and preventive care measures were added, for a total of 32 and it then stayed at that level for years four and five. The mix changed over time from 70 percent efficiency and 30 percent quality in year one, to 60 percent efficiency and 40 percent quality in year two to 50 percent quality and 50 percent efficiency in years three to five.
Every patient included. The changes Marshfield made for the demonstration were applied to all patients, not just Medicare patients whose care was measured in the demonstration. Even among Medicare patients, Marshfield clinicians did not know which ones were being measured until 18 months later, when results were released. Medicare patients who might have been in the demonstration were not included if part of their care was delivered outside of Marshfield during the examined period.
Steps taken. The clinic took the following steps to make care more efficient and reduce costs.
1. Installed "convertible computers," like laptops, in clinical settings and enhanced applications to support caregivers. One enhanced application displays in red letters services the clinician still needs to complete for the patient. It also reminds staff when the patient is about to be due for an appointment so they can reach out to the patient.
2. Added a nurse telephone triage line at more Marshfield sites. The phone line now receives more than 75,000 calls a year. Nurses use evidence-based triage guidelines. This service was going to be expanded anyway, Dr. Praxel says.
3. Expanded a program to make sure anticoagulants such as Warfarin were given to the right patients, using a single set of protocols. This program was expanded from a very small pilot to more than 6,500 people.
4. Started a cholesterol control clinic, which is still in the pilot phase.
5. Developed treatment guidelines for physicians on all major conditions, which can be modified when more evidence comes in.
6. Developed regional teams to visit Marshfield sites. A team made up of a practicing physician and a clinical nurse specialist visits sites to share information on how to deal with patients and to get feedback on how processes could be improved.
Some concerns about the process. Dr. Praxel identified two ways the demonstration could have been improved.
1. More timely feedback. CMS still has not reported on results from year 4, which ended in March 2009. The delay makes it difficult to identify problems and make improvements.
2. Feedback needs to be more granular. There is little information on who individual patients were, which also makes it difficult to identify problems and make improvements.
How the demonstration relates to upcoming ACOs. Not much will be known about ACOs until proposed regulations are released in December, but it appears they will use the shared savings model. However, instead of comparing performance to a regional group, as occurred in the demonstration, the ACOs will be compared to a national norm. Also, it appears the new model will favor ACOs in high-cost regions that show great improvements, rather than ACOs in areas that already save substantial costs.
Will Marshfield start an ACO? "We need to see the regulations first," Dr. Prxel says. "There are a large number of legal hurdles to set up an ACO," he added, referring to federal anti-kickback and Stark laws. He added, "We have been committed to shared savings approach for seven years, ever since we first applied to the demonstration program."
Will practices or hospitals run ACOs? "It’s too early to tell," he says.
Learn more about the Marshfield Clinic.
Shared savings. The 10 clinics participating in the demonstration shared savings they achieved with CMS. To determine how much was saved, Marshfield's costs were compared with costs of caring for Medicare beneficiaries in the same region who did not receive any care from Marshfield. Organizations saving 2-7 percent of the control group's costs were paid 80 percent of the savings. They were not rewarded for savings under or above that percentage. Marshfield was one of only two participating groups that earned shared savings in all three years. The other was the University of Michigan Faculty Group Practice. In years 1-3 Marshfield won a little over $23 million in shared savings. The money it earned went to support infrastructure and the needs of patients, Dr. Praxel said.
Quality measures. Participants had to achieve certain quality levels. To measure quality, the study looked at a mixture of efficiency and quality for patients in a number of chronic care categories that expanded over time. The quality measures started in year one with 10 diabetes measures. In year two, 17 coronary artery disease and heart failure measures were added, for a total of 27. In year three, five hypertension and preventive care measures were added, for a total of 32 and it then stayed at that level for years four and five. The mix changed over time from 70 percent efficiency and 30 percent quality in year one, to 60 percent efficiency and 40 percent quality in year two to 50 percent quality and 50 percent efficiency in years three to five.
Every patient included. The changes Marshfield made for the demonstration were applied to all patients, not just Medicare patients whose care was measured in the demonstration. Even among Medicare patients, Marshfield clinicians did not know which ones were being measured until 18 months later, when results were released. Medicare patients who might have been in the demonstration were not included if part of their care was delivered outside of Marshfield during the examined period.
Steps taken. The clinic took the following steps to make care more efficient and reduce costs.
1. Installed "convertible computers," like laptops, in clinical settings and enhanced applications to support caregivers. One enhanced application displays in red letters services the clinician still needs to complete for the patient. It also reminds staff when the patient is about to be due for an appointment so they can reach out to the patient.
2. Added a nurse telephone triage line at more Marshfield sites. The phone line now receives more than 75,000 calls a year. Nurses use evidence-based triage guidelines. This service was going to be expanded anyway, Dr. Praxel says.
3. Expanded a program to make sure anticoagulants such as Warfarin were given to the right patients, using a single set of protocols. This program was expanded from a very small pilot to more than 6,500 people.
4. Started a cholesterol control clinic, which is still in the pilot phase.
5. Developed treatment guidelines for physicians on all major conditions, which can be modified when more evidence comes in.
6. Developed regional teams to visit Marshfield sites. A team made up of a practicing physician and a clinical nurse specialist visits sites to share information on how to deal with patients and to get feedback on how processes could be improved.
Some concerns about the process. Dr. Praxel identified two ways the demonstration could have been improved.
1. More timely feedback. CMS still has not reported on results from year 4, which ended in March 2009. The delay makes it difficult to identify problems and make improvements.
2. Feedback needs to be more granular. There is little information on who individual patients were, which also makes it difficult to identify problems and make improvements.
How the demonstration relates to upcoming ACOs. Not much will be known about ACOs until proposed regulations are released in December, but it appears they will use the shared savings model. However, instead of comparing performance to a regional group, as occurred in the demonstration, the ACOs will be compared to a national norm. Also, it appears the new model will favor ACOs in high-cost regions that show great improvements, rather than ACOs in areas that already save substantial costs.
Will Marshfield start an ACO? "We need to see the regulations first," Dr. Prxel says. "There are a large number of legal hurdles to set up an ACO," he added, referring to federal anti-kickback and Stark laws. He added, "We have been committed to shared savings approach for seven years, ever since we first applied to the demonstration program."
Will practices or hospitals run ACOs? "It’s too early to tell," he says.
Learn more about the Marshfield Clinic.