The Bundled Payment Physician Attitude Survey

The Bundled Payment Physician Attitude Survey

I. The Bundled Payment Initiative Proposal

"The Bundled Payments for Care Improvement initiative seeks to improve patient care

through payment innovation. The program goal is to foster improved coordination and

quality through a patient-centered approach. The CMS Innovation Center will begin

studying four broadly defined models of care in 2011. Three models involve a retrospective

bundled payment arrangement, and one model would pay providers prospectively.

-CMS Innovation Request for Application for Bundled Payment Initiative July 2011[1]

Hospital systems and health care providers will work together to develop care delivery

systems and decide how payments will be divided among the various providers involved in

the patient's care. Applicants would propose the target price based on total costs for a similar

episode of care as determined from historical data. Participants in these models would be paid

for their services under the traditional fee-for-service (FFS) system. After the conclusion of the

episode, the total payments would be compared with the target price. Participating providers

may share in those savings if the total cost of care is lower than the set target price.

Applicants for the model payment programs would also decide whether to define

the episode of care as the acute care hospital stay only (Model 1), the acute care hospital

stay plus post-acute care associated with the stay (Model 2), or just the post-acute care,

measuring services given after discharge from an acute inpatient stay (Model 3). Under the

last model (Model 4), CMS would make a single, prospective bundled payment. This payment

will encompass all care delivered during inpatient stay by the hospital, physicians and other

practitioners. [1]

II. How Payment Bundling Will Influence Physicians

Payment bundling will encourage physicians to streamline their practice habits by having them

take on financial risk in patient management. According to a recent study in 2009 overuse of

12 common diagnostic tests cost over $6.8 billion annually.[2] The study was limited to primary

care physicians, so if specialists were included in the study group the amount of money wasted

would be even higher.

Hospital costs per day in the United States range from $985 in South Dakota to $2514 in the

District of Columbia [3]. Thus length of stay in hospital would clearly affect the amount of

money spent per hospital episode. Part B drug use is also a significant expense for Medicare.

Expenditures for Medicare Part B drugs have risen drastically over the past several years,

increasing from approximately $3.3 billion in 1998 to over $10 billion in 2004.[4] There should

be significant cost savings if physician behavior in these areas is modified toward decreasing

cost.

III. Survey of Physician Attitudes Toward The CMS Bundled Payment Initiative

The original research portion of our project was designed to see if physicians would modify

possibly wasteful clinical behaviors under a bundled payment system. The physicians were

asked a series of questions on each model. The questions covered perceived ease of payment,

changes in management such as ordering diagnostic tests, and length of stay. The goal of these

questions is to show whether payment bundling will lead physicians to change current practice

strategies that are costly to the Medicare system. The survey also asks physicians which model

they would find most palatable as a working environment. This information would help guide

hospitals with their choice of which Model to launch if the hospital system is considering

participation in one of the CMS Bundled Payment Model Programs.

The survey was distributed to 64 physicians. Physicians selected to receive the survey

had current or past experience with billing Medicare and commercial insurers. There were

two methods of distribution- and online version was distributed to physicians across the

United States using the surveymonkey.com website and a hard copy version was distributed to

physicians in the Los Angeles area. Twenty two physicians responded to the survey.

IV. Survey Results

The overwhelming majority of responding physicians were specialists, only 13.6% reported

a primary care specialty. The specialists included General Surgeons (27%) Obstetrician/

Gynecologists(13.6%), Orthopedic Surgeons(13.6%), Pulmonologists (9%) Medical Oncologists

(9%), Radiologists (4%), Cardiologists (4%) Of the 22 physicians 68% were familiar with the

CMS Bundled Payment Initiative. Over 64% had been in clinical practice from 6 to 15 years.

For each of the four models, over 50% of the physicians responding felt they would receive

less money under the bundled payment program. They did not feel the opportunity to share in

savings would offset the potential decreased in compensation.

Physicians also felt payments would be more difficult to obtain under the payment bundling

system. Well over 50% of responders selected either “Disagree” or “Strongly Disagree” for each

Model. This was true for all four models, but was most significant for model 4, the prospective

payment model.

Table 4: Perceived Ease of Payment Under Model 4 (Prosepective Payment, Acute Care)

Of interest, Model 4 also garnered the most “agree” responses to the question of

ease of payment. A subset of physicians, those employed by managed care organizations

or academic medical centers, felt payment would be easiest with Model 4. One physician

who chose Model 4 as the most appealing stated “By getting paid prospectively I would have

less concerns about cash flow”. This sentiment about payment up front was echoed by most

physicians who chose the prospective payment model as the most favorable.

Model 2 which included both acute care and post-acute care showed the strongest

ability to change physicians’ outpatient management. Half of physicians stated they would

change their outpatient management under this model. This was not seen in Model 3, which

includes only the post-acute care period. One physician comment on Model 2 states “There

are too many non-physician providers involved. Their management would affect the total cost

of care and would likely make it harder to stay below the target price” This was a common

thread among providers who least favored model 2. This concern may be valid, since part of

the Geisinger Health model showed the post discharge period incurred high cost of care.[5]

Table 5: Effects of Model 3 (Retrospective Post-Acute Care) on Outpatient Management

Table 6: Effects of Model 2 (Retrospective Acute and Post-acute Care) on Outpatient

Management

Medicare Part B medications were included in the payment bundle for Model 3, the

retrospective post-acute care program. The cost of medications covered under part B has risen

significantly in the past few years. [4] The use of brand name drug rather than generics has

contributed significantly to the cost increase. [4] While physicians were not asked directly if

they would change to generic drug under Model 3 they were asked if they would modify their

drug choices in any way. Fifty four percent of responders stated they would make changes.

The “agree” response was more common among the medical specialists. Surgical specialists

made up all of the “neither agree nor disagree” responders, likely because they rarely if ever

use part B drugs.

Table 7: Modification of Medicare Part B Drug Choices Under Model 3 (Retrospective

Post-Acute Care)

Physicians were asked which models were most and least appealing. Overall Model

1 was the most appealing choice, though several physicians described it as “the least of

four evils” in the comments section. Model 1 was more favored by physicians in solo or

independent group practices. The acute care retrospective model was not as highly favored

by physicians employed by academic centers, managed care organizations, or public health

systems such as county hospitals or the Veterans’ Administration

Table 8: Most Appealing Bundled Payment Model

Model 4, the prospective payment for acute care only program, was the least appealing

Model overall. In contrast to Model 1, Model 4 was least popular among independent group

practitioners and solo practitioners. Most physicians who disliked Model 4 commented that

the hospital controlling the payment was a big concern. There was alos concern that the

risk of exceeding the target price was too high. One respondent saw this as “a physician

penalty” with “no physician bonus opportunity” to balance the penalty. Model 1 was far

less favorable among employed physicians and the non-surgical specialists. One physician,

an employed medical oncology specialist expressed concern that the retrospective acute

care only model “covers inpatient care only without consideration to the outpatient related

manifestations and complications of disease entities”.

Table 9: Least Appealing Bundled Payment Model

V. Recommendations Based on Survey Findings

Though the sample size of our survey was relatively small, some results were consistent and

useful. Certain physician behaviors are very expensive and a burden on the Medicare system

[2,3,4]. Our data strongly suggest that wasteful practices would likely be curtailed once the

physician takes on a portion of the risk.

The difference in preference of model based on practice setting will also be of use to

hospitals or health care systems considering participation in the payment bundling model

program. Hospitals with a high degree of physician alignment would likely fare best under

Model 4, the prospective payment for acute care program, since this model was most favored

by employed physicians. Hospitals staffing solo or group practitioners would likely fare best

with Model 1. Respondents perceived more independence with the retrospective acute care

only model. They also felt there was less risk involved since the global period is shorted (30

days) and there are no non-physician providers (eg skilled nursing facilities, durable medical

equipment providers) involved in the payment sharing

Overall all four models will cause physicians to reassess and scale back their clinical

practices[6] Incurring part of the risk in patient care will encourage physicians to make

streamlined evidence based decisions in partnership with their patients. The data obtained

in the physician survey compounds the evidence that financial incentives (or in this case

disincentives) are an effective, though not pleasant, way to change physician behavior Peer

feedback should be added to discourage overuse, as this has been proven to be a palpable

method for changing physician behavior [7] Thus physician leadership is desperately needed to

make the changes to Medicare successful.

Conclusion

While the bundled payment reimbursement mechanism has shown promise in early

demonstrations, there is no guarantee that bundled payments will become a dominant payer

mechanism in the future, and we expect 2012 will bring more insights. It is also difficult to

prognosticate whether the current pilot models will become the future bundled payment

methods. Other than being a great experiment, the ACE demonstration will also have the

intended consequence of making physicians more cost-conscious than they have ever had to

be under the fee for service model. Effective service line management can take out additional

costs out of the system in the form of effective LOS management, decreased pathway

variability, decreased clinical complications and perhaps most importantly, in understanding

post-acute care relationships and efficiencies since these will become part of the episode of

care. Hospitals and physicians will also need to continue to increase their collaboration in this

model especially if gain-sharing incentives are part of the final implementation.

References

1. Center for Medicare and Medicaid Services Bundled payments for Care Improvement

Initiative Request for Application http://www.innovations.cms.gov/areas-of-focus/

patient-care-models/Bundled-Payments-%20Care-Improvement-Application.html

2. Top 5" Lists Top $5 Billion

Minal S. Kale; Tara F. Bishop; Alex D. Federman; Salomeh Keyhani

Arch Intern Med. 2011;171(20):1856-1858. Published online October 1, 2011

3. Kaiser Family Foundation State Health Facts. Hospital Adjusted Expenses per Inpatient

Day 2009

4. Department of Health and Human Services Office of the Inspector General Medicare

Payments for Newly Available Generic Drugs January 2011 http://oig.hhs.gov/oei/

reports/oei-03-09-00510.pdf

5. Health Affairs 11/7/11 Univ of MI, Ann Arbor

6. Achieving Health Care Reform — How Physicians Can Help

Elliott S. Fisher, M.D., M.P.H., Donald M. Berwick, M.D., M.P.P., and Karen Davis, Ph.D.

N Engl J Med 2009; 360:2495-2497June 11, 2009

7. Regional collaboration to improve radiographic staging practices among men with early

stage prostate cancer.

Miller DC, Murtagh DS, Suh RS, Knapp PM, Schuster TG, D