letterhead by cuiliqing


									 13034 Ballantyne Corporate
 Charlotte, NC 28277

                          August 25, 2011
   T   704 357 0022
   F   704 357 6611

 444 N Capitol Street NW Donald Berwick,   MD
 Suite 625
 Washington, DC 20001-1511

   T   202 393 0860
                          Centers for Medicare & Medicaid Services
   F   202 393 6499      Hubert H. Humphrey Building
                          200 Independence Avenue, SW, Room 445-G
   premierinc.com
                          Washington, DC 20201

                          Re: CMS–1525-P, Medicare and Medicaid Programs: Hospital Outpatient
                          Prospective Payment; Ambulatory Surgical Center Payment; Hospital
                          Value-Based Purchasing Program; Physician Self-Referral; and Provider
                          Agreement Regulations on Patient Notification Requirements

                          Dear Dr. Berwick:

                          On behalf of the Premier healthcare alliance serving more than 2,500 leading
                          hospitals and health systems and 76,000-plus other healthcare sites, we
                          appreciate the opportunity to comment on the Centers for Medicare & Medicaid
                          Services’ (CMS) proposed rule for the calendar year (CY) 2012 Medicare
                          Outpatient Prospective Payment and Ambulatory Surgical Center (ASC)
                          Payment Systems and the fiscal year (FY) 2014 Hospital Value-Based
                          Purchasing Program. Premier, a 2006 Malcolm Baldrige National Quality Award
                          recipient, maintains the nation's most comprehensive repository of hospital
                          clinical, financial and operational information and operates one of the leading
                          healthcare purchasing networks. Premier also partnered with CMS for six years
                          in the successful operation of the Hospital Quality Incentive Demonstration with
                          more than 250 of our member hospitals participating, which paved the way for
                          this national program. Our comments primarily reflect the concerns of our owner
                          hospitals and health systems. As service providers, the members of our alliance
                          have a vested interest in the effective operation of the outpatient prospective
                          payment system (PPS) and the hospital outpatient quality reporting (OQR)

                                        SPECIFIED COVERED OUTPATIENT DRUGS

                          Setting the payment rate for specified covered outpatient drugs (SCODs) has
                          presented challenges since the beginning of the OPPS, and special payment rules
Donald Berwick, MD
August 25, 2011
Page 2 of 17

have been applied either by legislation or regulation in most years. Premier appreciates CMS’
recognition of the various factors affecting its ASP + X payment rate methodology and the
changes it adopted beginning in 2010 to address some of the problems. These changes, especially
reallocating a portion of overhead costs from packaged drugs to unpackaged drugs, achieve
greater equity between packaged and separately payable drugs in the rate-setting methodology.

For 2012, CMS proposes to contin/ue the CY 2010 and CY 2011 overhead adjustment
methodology. Further, to account for inflation that has occurred since the overhead redistribution
amount of $200 million was established, CMS proposes to adjust this amount using the PPI for
Prescription Drugs. With the inflation adjustment, the $150 million previously redistributed from
coded packaged drugs and biologicals with reported ASP data would be $161 million, and the
$50 million previously redistributed from the cost of uncoded packaged drugs and biologicals
without an ASP would be $54 million. Thus, CMS proposes to reallocate a total of $215 million
from the costs of coded and uncoded packaged drugs and biologicals to separately payable drugs
and biologicals; this is equivalent to the $200 million redistributed in 2011 adjusted for inflation.

The Premier alliance continues to support the reallocation of a portion of overhead costs from
packaged drugs to separately payable drugs, and we also support the inflation adjustment applied
in the proposed rule. We are concerned, however, with the proposed payment rate of ASP + 4
percent and the prospect that CMS envisions that the payment rate could fall to ASP + 3 percent
or lower in the final rule. We believe that CMS should take these additional steps to ensure an
adequate and appropriate payment rate for separately payable drugs:

   1) In calculating the ASP + X rate, CMS should exclude the claims data for 340B hospitals.
      Claims and cost report data from these hospitals are not representative of drug acquisition
      costs from the majority of hospitals. Premier agrees, however, that 340B hospitals should
      continue to be paid the same amounts for separately payable drugs and biologicals as are
      paid to hospitals that do not participate in the 340B program.

   2) CMS should reallocate a larger portion of overhead costs from packaged drugs to
      separately payable drugs. The proposed rule for 2012 would redistribute an amount of
      $161 million in overhead cost from coded packaged drugs and biologicals with an ASP or
      approximately 35 percent, which falls well within the one-third to one-half of the
      estimated pharmacy overhead cost that CMS indicates would be a reasonable range for
      the amount to be reallocated. Premier urges CMS to increase the amount to be
      redistributed to maintain the payment rate at the current ASP + 5 percent level.

   3) Alternatively or in combination with number 2 above, CMS could increase the amount
      redistributed from uncoded packaged drugs and biologicals. The proposed rule would
      redistribute an amount of $54 million in overhead cost from uncoded packaged drugs and
      biologicals, or approximately 11 percent. The 11 percent compares with the 23 percent
      reallocated from coded packaged drugs, as shown in the table below. Premier believes
      that the amount redistributed for uncoded packaged drugs could be increased
Donald Berwick, MD
August 25, 2011
Page 3 of 17

       substantially provided it does not exceed the percentage reallocated from coded packaged
       drugs. We urge CMS to consider making these two percentages the same.

     Comparison of Proposed Reallocated Overhead Amounts as a Share of Total Costs

                                    Reallocated        Reallocated Overhead/
                         Total Cost Overhead           Total Cost
         packaged                     $54 million      11%
         packaged                     $161 million     23%

The Premier alliance supports the CMS methodology for determining the payment rate for
separately payable drugs and biologicals but we urge CMS to make the additional changes
necessary to continue the payment rate at ASP + 5 percent.


Premier opposes the CMS proposal to cap the payment rate for both the newly proposed
composite APC for CRT-D and the existing APC for Implantable Cardioverter Defibrillator
(ICD) based on the payment amount for MS-DRG 227. We are concerned about the precedent of
linking an OPPS payment level to payments external to the OPPS, such as an MS-DRG payment
rate under the inpatient prospective payment system. For years, CMS has upheld the principle of
setting APC payment rates based on the outpatient claims data, and the agency has worked very
hard to improve the overall quality of the claims data used to set the rates. In previous years,
Premier and other stakeholders concerned about the differential payment for injected or infused
drugs in a hospital outpatient department compared to a physician’s office have urged CMS to
adjust the OPPS rate to ASP+6 percent, the rate paid in the physician setting. CMS, however, did
not agree that the site of service differential was an issue. Premier believes that the OPPS
cost-estimation and rate-setting principles carefully developed by CMS over the years should
continue to apply here as well – and that CMS should be consistent in its application of those

Composite APCs were created by CMS to provide payment for two or more procedures when
they are commonly provided in the same encounter. The methodology enables CMS to use more
claims data and to use claims data that more accurately represent the full cost of the services
when they are furnished in the same encounter. The composite payment methodology inherently
results in more accurate payment for the services paid under a composite APC. Applying a cap to
Donald Berwick, MD
August 25, 2011
Page 4 of 17

the calculated composite APC median cost is contrary to the CMS stated purpose of using the
composite payment methodology. For CRT-D, the cap proposed by CMS for the composite APC
is below the median cost for one of the two procedures included in the composite. CMS is
effectively proposing to use the composite APC methodology, which was developed to reflect
the true cost of procedures, to pay less than the median cost of one of the procedures in the
composite. In Table 24 of the preamble, CMS indicates that the device-related portion of these
procedures is 87 percent of the median cost. CMS calculated the median cost for these two
procedures based on 2010 claims data at $38,854 of which $33,803 would represent the device
portion of the procedure. According to CMS’ calculations, the proposed payment rate does not
even cover the cost of the devices used to perform the procedure.

Premier also notes that the alleged site of service payment differential cited by CMS for CRT-D
between the inpatient and outpatient settings was not calculated correctly. CMS compared the
inpatient CRT-D payment amount for only the operating portion of the payment to the total
OPPS payment amount which covers both operating and capital costs. CMS should have
included the capital payment portion in the inpatient side of the comparison as well. In addition,
CMS must account for the indirect medical education and disproportionate share payments that
hospitals receive. For the comparison, CMS should use the weighted average (weighted by OPPS
volume) indirect medical education and disproportionate share adjustment factors for hospitals
billing for CRT-D under the OPPS. An appropriate comparison such as this would show that an
exception to its well-established OPPS rate-setting methodology is not warranted. Premier
strongly urges CMS not to finalize the MS-DRG 227 payment cap in the final rule.


The Notice of Proposed Rule Making includes several proposals related to the inpatient hospital
Value-Based Purchasing (VBP) program. These build on the final rule issued in April 2011,
which implemented the measures and scoring for FY 2013 and some measures and scoring for
FY 2014. The proposed changes in this rule include the addition of one clinical process measure,
setting minimums for cases and measures in the outcomes domain, performance standards and
scoring methodology for the hospital-acquired conditions (HAC) measures, and the weighting of
performance measure domains.

In a number of places, CMS makes reference to and cites results from various analyses that
support its proposals for specific VBP policies. It is impossible to provide informed comments in
response to these proposals without benefit of the underlying analysis. Premier urges CMS to
make public the data analyses used in developing the proposed policies, including the
studies by Brandeis and Mathematica regarding the minimum number of cases and CMS
study on which the seven-month performance period for the outcomes domain were based.
Donald Berwick, MD
August 25, 2011
Page 5 of 17

SCIP –INF-9 Postoperative Urinary Catheter Removal on Postoperative Day 1 or 2

In this rule, CMS proposes to add Postoperative Urinary Catheter Removal on Postoperative Day
1 or 2 to the set of clinical process of care measures included in the VBP program for FY 2014.
CMS states that information about this NQF-endorsed measure first appeared on Hospital
Compare in December 2010 and is among the NQF SCIP infection prevention measures included
in the Health and Human Services Action Plan to Prevent Healthcare Associated Infections.
Premier agrees this measure meets the criteria for VBP, and supports inclusion of this

Minimum Cases and Measures for the Outcomes Domain for FY 2014

CMS proposes the minimum number of cases and measures required for calculation of an
outcomes domain score for a hospital. As noted earlier, the specific proposals cite analyses by
Brandeis University and Mathematica that have not been publicly available, so Premier is not
able to adequately assess these proposals. Ideally the appropriate minimum case count would fall
directly out of the risk model, which we have argued elsewhere should be part of the
measurement of all adverse outcomes, including HACs and other patient safety indicators.
Sufficient statistics from those models would guide the determination of minimum case counts.
Certainly we do not agree that a single Medicare claim is sufficient basis for calculating a score
on hospital-acquired conditions, which CMS itself notes are extremely rare events. Additionally,
the three-case minimum proposed for calculation of a score on the AHRQ composite measures
also appears quite low. Moreover, by proposing that hospitals need only have a score on the
HACs, which every hospital will meet as proposed, and three of the other outcome measures, it
appears that a hospital with only 10 cases might be able to have an outcome score if those cases
also met the requirements for calculating the AHRQ composite measures as well. If not, 16 cases
would be required. Given the importance of patient outcomes and a weight of 30 percent, this
seems a very low threshold for fairly assessing differences in hospital performance in the first
year. Because most hospitals have many more cases and will far exceed all these thresholds, we
fail to see the value of setting them so low that they will include hospitals for which performance
measurement may not be accurate. In light of the insufficient data offered in making these
proposals, Premier does not support the low thresholds for minimum cases and minimum
measures proposed for the outcome domain.

Performance Periods

CMS proposes that the performance periods for HACs and the AHRQ composite measures be
March 3, 2012, through September 12, 2012. We have several concerns regarding this proposal.
First, the performance period does not meet the statutory requirement that measures be included
in Hospital Compare for one year prior to the start of the performance period. Performance data
on these measures were posted on Hospital Compare by providing a link to downloadable files
available on the CMS website. While the AHRQ file was posted on April 8, 2010, it was later
removed, and the HAC file was not posted on March 31, 2011. CMS is just now preparing to
Donald Berwick, MD
August 25, 2011
Page 6 of 17

post such data on Hospital Compare in October 2011. In the final FY 2012 IPPS/LTCH rule,
CMS clarifies its view that presenting a “Spotlight” notice discussing future addition of the
HACs as including the HACs on the Hospital Compare website. CMS further finalizes a
performance period of May 15, 2012 to February 14, 2013 for the efficiency domain, when the
measure is not fully specified let alone posted on Hospital Compare. Premier believes Congress
clearly intended that posting data on hospital performance to Hospital Compare is the
relevant action, not defining the measure or discussing its future addition to the website,
and thus questions the legality of these policies.

Second, we agree with CMS that a 12-month performance period is the ideal, and the proposed
performance periods are shorter than that. In particular, a period of less than seven months is too
short to fairly distinguish performance among hospitals on the HACs, especially given that CMS
points out the rarity of HACs. In fact, the existing measures on Hospital Compare use three years
of data to fairly compare HAC rates. We note that the proposed performance period on the
Medicare spending per beneficiary measure is also less than 12 months, and the proposed rule
does not discuss the implications of different performance periods for this measure, which is not
NQF endorsed and for which no data has ever been made public or shared with hospitals. For
these reasons, Premier opposes the proposed performance period for the HACs, AHRQ
composite measures, and the Medicare spending per beneficiary measure.

Performance Standards for HACs

CMS proposes a different performance standards methodology for the eight HAC measures that
were adopted in the final rule for the FY 2014 VBP Program than the methodology used for the
other measures. Citing data on the rarity of HAC occurrences, CMS reports that if the
methodology that was finalized for the mortality measures were used for the HACs, the
achievement threshold for each of the measures would be zero. According to CMS, data for
October 1, 2008-June 30, 2010, show that one-quarter of hospitals had no reportable occurrences
of any of the HACs, and at least one-half had no occurrences on six of the eight HACs.

Specifically, CMS proposes that for the single combined HAC measure, achievement and
improvement thresholds and benchmarks would be established based on hospital combined
performance on seven or eight measures as applicable. While we generally support a composite
measure due to the low frequency of the events, we are concerned that the methodology to
calculate a pair of requirements for those who perform surgeries and those who do not is
unnecessarily complicated and exacerbate the concern over a shortened performance period and
low frequency of events.

The Premier healthcare alliance has expressed its concerns with the CMS inclusion of rare
events, such as blood incompatibility and object retained after surgery, in the pay for reporting
and performance programs in past comments on rulemaking. We believe that serious reportable
events should continue to be included in the existing payment policy whereby a HAC cannot
Donald Berwick, MD
August 25, 2011
Page 7 of 17

qualify a case for higher diagnosis-related group payments as the low frequency does not lend
itself to risk-adjusted rate-based measurement that should be the basis for pay for reporting.

Furthermore, comingling infection rates with serious reportable events in a single composite
measure for VBP complicates the meaning of the measure and may be misleading to consumers,
especially if claims-based versions of measures such as CAUTIs will continue to appear
separately on the Hospital Compare website. We were pleased to see in the inpatient PPS final
rule that CMS is considering retirement of claims-based version of CAUTIs stating,

       “because the topic of HAIs is of great importance, and a large quantity of data for
       the NHSN version of the measure will not be available to CMS for some time, we
       will continue to utilize the claims-based measure until such time as the NHSN
       version is available to CMS. We will seek an appropriate time to retire the
       claims-based version of the measure taking into account the needs of and impact
       on other programs, such as the Hospital VBP Program.”

We believe that CMS should routinely remove the measures based on administrative data once
an outcomes-based, chart-abstracted measure is integrated into the program, this avoiding
“double counting” of certain events or infections.

Premier urges CMS to only include risk-adjusted rate-based measures in VBP and to
remove the serious reportable events, which would also obviate the need for a pair of
benchmarks and thresholds.

HAC Validation Process

CMS is proposing to add a targeted validation process for ensuring the accuracy of the HAC
data. This targeted process would be in addition to the current Medicare Administrative
Contractor claims review. The specific proposal is to target hospitals that “report zero or an
aberrantly low percentage of HACs on the Medicare-fee-for-service inpatient PPS claims relative
to the overall national average of HACs.” CMS also noted that they will take appropriate action
if systematic under-reporting of HAC or adverse event information is discovered.

Premier concurs that data used for quality reporting and improvement needs to be accurate,
especially when used for payment. Premier is generally supportive of a targeted HAC
validation process as long as it does not cause undue burden to the hospitals. Additionally,
CMS should make public all analysis for determining the “low percentage of HACs” value
for use in this proposed target validation along with the specific validation process. Premier
also believes that the validation process should provide hospitals with an opportunity to
review and appeal the validation findings.

Domain Weighting
Donald Berwick, MD
August 25, 2011
Page 8 of 17

For FY 2014, CMS proposes major changes to the domain weighting used to calculate a
hospital’s VBP program total performance score. For FY 2013, the clinical process of care
domain is weight at 70 percent and the HCAHPS is weighted at 30 percent. For 2014, HCAHPS
would remain unchanged, but the clinical process of care would be dropped from 70 percent to
only 20 percent, and the new outcomes domain would receive a weight of 30 percent and the
efficiency domain, consisting only of the Medicare spending per beneficiary measure, would
receive a weight of 20 percent.

Premier opposes the drastic shift in domain weighting for the VBP program. FY 2014 is only the
second year that the program is in place, and we generally believe that changes should be made
more gradually, especially since hospitals are newly adapting to the program. More specifically,
however, we believe the new Medicare spending per beneficiary measure should not be included
at all in 2014. That is, the efficiency domain should be given a weight of zero for FY 2014. The
specifications for the measure are not yet complete, no data have been shared with hospitals, and
results have not been posted on Hospital Compare for the requisite year as required by law. CMS
is proposing to give this incomplete, untested measure the same weight as all the clinical process
of care measures, which have been tested and in wide use.

We agree that the outcome domain should eventually be given the highest weight among all the
domains, but the initial weight for this category when it is first introduced in FY 2014 should be
lower and it should be phased in. As noted earlier, we do not believe that the HACs and AHRQ
measures should be included in the scores for this domain for FY 2014 because the performance
periods for these measures are too short and do not meet the statutory requirement for prior
inclusion on Hospital Compare for one year. Thus, for FY 2014, the outcome domain should
consist only of the mortality measures, and we continue to have concerns about the length of the
performance period and the sufficiency of risk adjustment for these measures. Likewise, Premier
continues to believe that a 30 percent weight for the HCAHPS domain is too high given that the
HCAHPS composite lacks sufficient risk-adjustment. Research shows that high-acuity patients
score their patient experience at a lower level systematically disadvantaging hospitals that take
on complex and sicker patients. Furthermore, the new consistency score CMS implemented has
not been thoroughly tested. Premier proposes the following alternative weighting scheme for
calculation of the total performance score for FY 2014: clinical process of care weight of 65
percent, HCAHPS weight of 20 percent, outcome weight of 15 percent, and efficiency
weight of 0 percent.
Donald Berwick, MD
August 25, 2011
Page 9 of 17

HCAPHS Data Review and Correction Process

CMS is proposing to use a “two-phased” approach for the review and correction of HCAHPS
data. The first phase would reduce the submission deadline by one week to provide a review and
correction period. The hospitals would have the opportunity to provide missing data or replace
incorrect data for the existing data records but not submit new data records. In the second phase,
hospitals would have the opportunity to review their HCAHPS scores for the Hospital VBP
Program. Hospitals would be allowed to submit questions relating to their VBP calculations to
CMS. If there was an error CMS would recalculate the score.

Premier supports the two-phased review and correction process. We recommend that CMS
provides clear guidance and definition for missing and incorrect data correction in phase
one. Also, CMS notes that hospitals will not be allowed to add new data records during a
one-week correction period for phase one. We would encourage CMS to reconsider the
submission of “new records” during the correction period. The data may be invalid due to
missing patient records and there should be an opportunity to submit the missing records.


Regarding the hospital OQR program, the proposed rule includes changes for the payment
determinations in FY 2014 and FY 2015. For FY 2014, data collection procedures would be
modified for one previously finalized emergency department (ED) throughput measure and
measures would be added on surgical site infection, diabetes, cardiac rehabilitation referrals, use
of a safe surgery checklist and procedure volume. For 2015, a measure on influenza vaccination
for healthcare personnel is proposed for addition.

OP-22 Left Without Being Seen

We support the proposed change to annual aggregate reporting, because it is a less burdensome
way for hospitals to provide the information than the original approach of quarterly chart
abstraction. We remain concerned, however, that some urban hospitals with emergency
departments in high demand for non-emergency services will show poor performance on this
measure for reasons beyond their control. In addition, counting patients who left the ED without
being seen is not a clear-cut indicator of quality of care. Moreover, CMS should ensure that
effective triage procedures to lower levels of care are not inadvertently captured in this measure
resulting in higher rates for hospitals that appropriately redirect non-emergent cases.

Premier remains concerned that access to healthcare providers is not uniform across
communities. In some poor urban centers, access to care for a significant portion of the
community may be primarily through the emergency department, whereas in more affluent
communities where access to care is usually not an issue, the emergency department may not be
a significant source of primary care. Factors such as healthcare literacy also impact utilization of
the emergency department and such factors vary by socioeconomic status. While discovering and
Donald Berwick, MD
August 25, 2011
Page 10 of 17

documenting problems with access to care is a worthy endeavor, Premier urges CMS not to
conflate access to care issues with throughput as part of a payment system. At the very least
CMS should stratify hospitals by setting and by socioeconomic status so that hospitals can be
held accountable for what they can control, and not be penalized for deeply rooted and
long-standing problems of access to care and healthcare literacy, which are difficult for any one
provider to address in the long term, and certainly impossible for any one provider to impact in
the short term.

We support the changes to the reporting of this measure, but remain concerned about its
inclusion in pay for reporting and suggest CMS study the results for systematically higher
rates for certain types of hospitals such as safety net providers so that appropriate
adjustments can be made.

Surgical Site Infection

This proposed measure is among those collected by the Centers for Disease Control and
Prevention (CDC) National Healthcare Safety Network (NHSN) and assesses the percentage of
surgical site infections (SSIs) occurring within 30 days after an NHSN-defined operative
procedure if no implant is left in place or within one year if an implant is left in place and the
infection appears to be related to the procedure.1

CMS proposes data submission for this measure would relate to events occurring between
January 1, 2013, and June 30, 2013, for the CY2014 payment. It would be reported using the
NHSN infrastructure and protocols and references. Further, CMS proposed that because this SSI
measure was adopted for the Hospital Inpatient Quality Reporting (IQR) Program for the
FY 2014 payment determination, CMS views the proposed addition of this measure to the
hospital OQR program as meeting the goal of aligning measures across programs where feasible.
As admirable as this goal is, Premier would like to note that there are a considerable number of
issues to consider.

In the final rule for the FY 2012 inpatient PPS, published on August 18, 2011, CMS announced
that for the IQR program, reporting of the SSI measure will be limited to the colon and
abdominal hysterectomy procedures for the FY 2014 payment determination. Data collection for
the IQR program will start on January 1, 2012. We do not believe these procedures are an
appropriate choice for hospital outpatient departments (HOPDs) or ambulatory surgical
centers (ASCs), and it is therefore not feasible to align the SSI measure by applying it to the
same procedures across all three settings. Given the same day discharges, the HOPD and ASC
setting provide a different mix of procedures than inpatient hospitals. Although the approved list

 SSI NQF #0299: Numerator: Deep incisional or organ/space infections occurring within 30 days after an operative
procedure if no implant is in place or within one year if an implant is in place. Denominator: Number of operative
procedures. Procedures in scope for the measure are coronary artery bypass graft and other cardiac surgery, hip or
knee arthroplasty, colon surgery, hysterectomy (abdominal or vaginal) and vascular surgery.
Donald Berwick, MD
August 25, 2011
Page 11 of 17

of NHSN procedures includes some that apply to both inpatient and outpatient settings, the
proposed rule recognizes that many of these are not carried out in outpatient settings and,
therefore, would not be applicable.

There is urgent need for more experience in using the NHSN infrastructure for regular reporting
to CMS, not just by providers but also by vendors, and the CDC’s experience in collecting and
transmitting the data to CMS. Premier is a proponent of NHSN, a free, publicly available system,
and supports participation in NHSN through the use of data mining systems that permit hospitals
to electronically transfer NHSN data to the CDC, and then to QualityNet/CMS. However, there
is only pilot testing of this interface between various vendors and the CDC at present. Without
electronic capabilities to transfer data, participation in NHSN is labor-intensive. Even with such
systems, but without the direct connectivity to CDC, it is time-consuming to generate electronic
reports that can then be submitted separately to NHSN.

Additionally, we note that CMS specifically proposed that HOPDs report on the SSI measure for
the hospital OQR program using the existing NHSN infrastructure and protocols. However,
HOPDs use Healthcare Common Procedure Coding System (HCPCS) codes for procedures,
meaning that the current SSI modules in NHSN that rely on International Classification of
Disease (ICD) coding would not be easily applicable. We are aware of efforts by CDC to
develop a module that would fit this setting, including both ICD and outpatient PPS coding for
surgical procedures, as well as one that meets the needs of ASCs to be addressed later. However,
this revised module is neither currently available nor expected before CY 2013.

If the SSI measure were added to the hospital OQR program as proposed, hospitals would need
time to plan, enroll in the NHSN, and train and educate appropriate staff for reporting this
measure. Surveillance practices in outpatient settings differ significantly from those used with
inpatients. This adds additional challenges to ensure the follow-up of each outpatient to
determine whether an infection developed, and if so, whether it meets the NHSN definition for

Finally, although we note discussion in the proposed rule for other measures’ validation
procedures, we noted no mention of NHSN validation processes. We would appreciate a
discussion of CMS’ plans for validation of the SSI data. Would this be similar to the approach
taken with respect to the IQR program measure for Central Line-Associated Bloodstream
Infection (CLABSI)? That process appears to involve time-consuming manual record
identification. CMS should consider working with CDC on a validation process that draws a
sample directly from the SSI measure information submitted to NHSN.

The Premier healthcare alliance urges CMS to begin with one or two of the NQF-endorsed
CDC/NHSN outpatient surgical procedures, based on volume and SSI risk. This will ensure that
the process for collection, transmission and, importantly, validation have been analyzed and
assessed carefully before broader application of this measure. To ensure that NHSN modules are
adapted to the needs of hospital OQR program reporting, (e.g., use of outpatient surgical
Donald Berwick, MD
August 25, 2011
Page 12 of 17

procedure coding), and to provide sufficient time for outpatient facilities to enroll, educate and
train staff, and develop an IT vendor interface with NHSN in this adapted module for HOPD, we
urge delay of the CY 2014 payment determination to CY 2015, with data collection for
events between January 1 and June 30, 2014.


CMS proposes to add to the hospital OQR program five chart-abstracted measures relating to
diabetes care of adults. All the measures are NQF-endorsed and assess levels of HgA1c
(NQF#0059) and LCL-C (NQF #0064), diagnosis of hypertension (NQF #0061), annual eye
exams (NQF # 0055) and urine protein screening (NQF #0062).

In commenting on the proposed rule regarding Medicare Shared Savings Program, Premier was
supportive of including these proposed measures as a means of assessing how well an
Accountable Care Organization (ACO) manages the care of Medicare beneficiaries with
diabetes. However, adoption of these measures to the hospital outpatient department (HOPD)
setting more broadly will be challenging and will require identification of appropriate exclusions,
which is not discussed in the proposed rule. Patients are seen in the HOPD for a wide range of
services. Does CMS intend that the diabetes measures apply to any HOPD patient, or only those
receiving evaluation and management services in primary care clinic settings? The measures
assess ongoing diabetes management. Would they apply to a patient seen in the HOPD only
once, or would there be a minimum number of visits required before these measures were

In our view it would be inappropriate for CMS to effectively treat an HOPD as if it were an ACO
by applying the diabetes measures to any patient seen in the HOPD. For example, an HOPD
treating a patient once for a surgical procedure or for imaging will not have an opportunity for
ongoing care management and will not have a record of whether the patient has received the
specific annual screenings that the measures assess (eye exam, urine protein screening).

In addition to these conceptual and definitional concerns, the proposed diabetes measures rely on
use of the CPT II codes, which are used in physician billing but not used in the HOPD. CMS
should not implement the Diabetes measures in the hospital OQR program, but rather
focus such efforts within PQRS and the MSSP programs.

Cardiac Rehabilitation

CMS proposes to add this measure (NQF #0643) on whether certain patients in the HOPD are
referred to a cardiac rehabilitation or secondary prevention program. The measure assesses the
percentage of patients meeting clinical criteria related to heart disease in the last 12 months
(heart attack, chronic angina, or specific cardiac procedures) who have not participated in a
cardiac rehabilitation program who are referred for these services.
Donald Berwick, MD
August 25, 2011
Page 13 of 17

The measure seems better targeted to patients being discharged from a cardiac-related inpatient
stay than to the HOPD, where patients may be seen for a wide variety of unrelated conditions or
procedures. Although a parallel inpatient measure is also NQF-endorsed, CMS has not proposed
it for addition to the hospital IQR program. Does CMS have evidence that applying this measure
in the hospital outpatient setting would improve the percentage of patients who appropriately
receive cardiac rehabilitation services? This measure appears to be merely reporting whether a
referral was made without regard to whether the patient ultimately received appropriate cardiac
rehabilitation services. Measures in the hospital OQR program, or any quality reporting program,
should be clearly tied to quality improvement. It is not clear that requiring hospitals to devote
resources to changing procedures to perform well on this measure would have any effect on
improving the quality of patient care. We also note that Medicare payment policies implemented
in recent years have had the effect of limiting the availability of cardiac rehabilitation services,
so a requirement for providing a referral may not guarantee a patient access to these services.
CMS should not implement the Cardiac Rehabilitation measure at this time, but rather
evaluate the inpatient measure as part of the IQR program.

Safe Surgery Checklist

CMS proposes that the HOPD annually report on whether or not it uses a safe surgery checklist
meeting certain requirements. Premier believes that the use of a safe surgery checklist is linked
to improved quality outcomes and is a useful quality improvement tool. We note, however, that
if this measure is adopted, it is important that hospitals not be required to use a specific checklist
to meet the requirement. As proposed, hospitals should be allowed to use any checklist tool that
includes safe surgery practices during three perioperative periods: prior to administration of
anesthesia, prior to skin incision, and closure of incision prior to patient leaving the operating
room. Premier supports the adoption of the sage surgery checklist in the hospital OQR
program if a particular checklist is not mandated.

Procedure Volume

CMS proposes that hospitals submit all-patient volume data for eight procedure categories that it
estimates account for 99 percent of all outpatient procedures. In proposing this measure, CMS
states that peer-reviewed literature links volume of surgical procedures to better patient

Premier opposes the addition of this measure. As we have stated in the past regarding similar
proposals, we do not believe that the posting of volume data is a valuable quality metric for
addition to the hospital OQR program. There may be literature linking the volume of specific
high-risk procedures provided by an individual practitioner to patient outcomes for those
procedures, but these high-risk procedures are generally not those performed in the HOPD. The
proposed rule does not relate to those high-risk procedures, but proposes to collect and report
aggregated volume by category of procedure.
Donald Berwick, MD
August 25, 2011
Page 14 of 17

Furthermore, newer literature indicates that volume of procedure is insufficient in itself to serve
as a quality measure. What is more useful is volume of procedure performed by a given provider
or team. CMS should consider the perverse incentive created when volume is used as a marker of
quality which might potentially be tied to payment. Indeed, one can only note with skepticism
the inherent contradiction of including a measure of volume in a system of payment that is
designed precisely to move away from paying for volume and rather toward paying for value.
CMS should not adopt procedure volume as a quality metric in the hospital OQR program.

Healthcare Personnel (HCP) Influenza Vaccination

CMS proposes to add the measure Influenza Vaccination Coverage Among Healthcare Personnel
(NQF#0431) to the hospital OQR program for the CY 2015 payment determination, with
reporting through the NHSN component for HCP. Premier agrees that HCP can unintentionally
expose patients to seasonal influenza if the HCP have not been vaccinated, and such exposure
can be dangerous to vulnerable patients. The Premier alliance supports mandatory HCP
vaccination, as well as public reporting of HCP vaccination rates.

Although we support the public reporting of HCP vaccination rates as an effective way to
capture regional trends and bring more attention to this public health issue, we must also
acknowledge the complexity and labor intensity for the collection of this information when one
considers the NQF specifications. In particular, collecting and reporting data for this measure can
require involvement of multiple HOPD departments such as credentialed medical staff and
resident departments, human resources, and reporting of vaccinations obtained outside the
hospital. Recent CDC studies demonstrate that hospitals, Ambulatory Surgery Care Centers
(ASCCs) and other outpatient care sites are very challenged in collecting numerator and
denominator data for any groups beyond employees such as credentialed non-employees or other
non-employees including students and volunteers, given the current NQF specifications.

Recognizing the reporting burden issues, in the FY2012 IPPS final rule published August 18,
2011, CMS indicates that the NHSN-HCP module is being modified to accept aggregate data.
We welcome this modification, but we believe it is critical to test how well it works for inpatient
settings before requiring its use in HOPD, where the challenges are even greater. Further, all the
comments noted earlier for SSIs remain true here as well, regarding implementing NHSN,
training staff, etc., as hospitals test out the modified NHSN-HCP module. We recommend that
implementation be delayed by one year, to coincide with the proposed addition of this measure to
the ASC quality reporting program. That is, the measure would be added to the hospital OQR
program for the CY 2016 payment determination, with data collection occurring for the flu
season from October 2014 to March 2015. This will provide additional time for improving
methods to collect information based on experience gained from reporting for the inpatient
setting using the revised NHSN module. Premier recommends the adoption of this measure
for payment determination in CY 2016.
Donald Berwick, MD
August 25, 2011
Page 15 of 17

Data Validation

CMS proposes several changes to the data validation procedures under the hospital OQR
program which parallel the changes adopted for IQR program data validation. Specifically, CMS
proposes to reduce the time frame for hospitals to submit the requested records to the CDAC
contractor from 45 calendar days to 30 calendar days. CMS notes that this will align the time
frame requirements with the current time frame for Quality Improvement Organization (QIO)
requests and will reduce the time needed to complete the validation process and provide
feedback to hospitals. Timely feedback will identify data abstraction errors which hospitals can
address and improve prior to future data submissions, thereby enhancing the quality of data in
the CMS QIO Clinical Data Warehouse. Premier supports reducing the time frame for
record submission to the CDAC contractor if it will improve the timeliness of feedback to
the hospitals.

Validation Frequency
CMS proposes to add hospitals to the validation sample if the hospital was open under the
current CCNs in FY 2012 and not selected for validation in the three previous IQR validation
samples. This would ensure that all eligible hospitals are selected for validation at least once
every 4 years. As CMS intends to use the results of Hospital IQR chart validation for Value
Based Purchasing, all eligible hospitals should be included in the chart validation process.
Premier supports a requirement that eligible hospitals are selected at least once every 4
years for validation.

Validation of NHSN measures
In the FY 2012 rulemaking for the IQR program, CMS proposed separate and specific
procedures for data validation of another NHSN measure related to central line-associated
bloodstream infection. As discussed above, Premier requests that CMS discuss its plans for
validation of data submitted through the NHSN for the proposed SSI measure.

                        ASC QUALITY REPORTING PROGRAM

We appreciate that CMS is requiring a quality program for ASCs for the first time, beginning
with 2014 payment determinations. We agree that ASC facilities are similar to HOPDs in that
both provide many of the same surgical procedures, and therefore similar standards and
guidelines with respect to surgical care improvement can be applied.

Claims-Based Measures Requiring Submission of Quality Data Codes

CMS is proposing to adopt seven NQF endorsed claims-based measures, and six were developed
by the ASC QC and the seventh measure is from the AMA’s Physician Consortium for
Performance Improvement. The proposed measures would be collected by Quality Data Codes
(QDCs) that would be submitted with the claim. The data collection/submission would begin
with January 1, 2012 encounters. The proposed measures are:
Donald Berwick, MD
August 25, 2011
Page 16 of 17

ASC-1: Patient Burn
ASC-2: Patient Fall
ASC-3: Wrong Site, Wrong Side, Wrong Patient, Wrong Procedure, Wrong Implant
ASC-4: Hospital Transfer/Admission
ASC-5: Prophylactic Intravenous (IV) Antibiotic Timing
ASC-6: Ambulatory Surgery Patients with Appropriate Method of Hair Removal
ASC-7: Selection of Prophylactic Antibiotic First OR Second Generation Cephalosporin

We have several concerns with this proposal. First, is the data collection methodology of using
QDCs. CMS notes that the ASC Quality Reporting Program final specifications will be posted
after publication of the CY 2012 OPPS/ASC final rule with comment period. If the final rule is
published in November 2011, and ASCs must start using the QDCs with January 2012 claims
there is insufficient time for education and training of staff to accurately assign the QDC.

Second, CMS is including measures of antibiotic timing and selection that are already in use by
the Hospital IQR and Hospital OQR programs. Although the measure specifications are similar,
the IQR and OQR programs rely on chart-abstraction not claims data. Due to different modes of
data collection, the measure results should not be publicly reported together.

Third CMS is proposing an Appropriate Hair Removal measure for ASC. However, the
comparable measure SCIP-Infection-6 Appropriate Hair Removal for the Hospital IQR program
was just suspended from data collection. The ASC quality measures should align with the other
CMS quality reporting programs. Thus we recommend that the ASC-7 measure not be
finalized for the ASC Quality Program.

Additionally Premier recommends that the ASC Quality Reporting be delayed until July 1,
2012 to allow hospitals the opportunity to experience the data collection using the new
QDCs process.

Surgical Site Infection

CMS proposes to require that like HOPDs, ASCs report on the Surgical Site Infection Rate (NQF
#0299) measure for the 2014 payment determination for events occurring on or after January 1,
2013, through June 30, 2013, using CDC’s NHSN as the reporting mechanism.

Our comments above regarding the application of the measure to HOPDs are also relevant here.
We note that ASCs are even less knowledgeable about NHSN procedures and capabilities, and
clearly need a system that is based on reporting using coding unique to HOPD/ASC.

We recommend a delay in adoption of the SSI measure for the ASC quality reporting. ASCs
would face the same issues as HOPDs with respect to reporting on the SSI measure and have no
experience with quality reporting in general or the NHSN system in particular. Moreover, when
Donald Berwick, MD
August 25, 2011
Page 17 of 17

both HOPDs and ASCs begin reporting on SSIs, CDC will face a substantial increase in volume
of NSHN reporting which it may not yet be prepared to handle. We recommend that SSI
measure be delayed to CY 2015 payment determinations with data collection beginning on
January 1, 2014, through June 30, 2014.

HCP Influenza Vaccination

In addition to retaining the 10 measures proposed for the 2015 payment determination, CMS
proposes to add one more measure in CY 2016, the NHSN HAI Influenza Vaccination Coverage
Among Healthcare Personnel (HCP) measure (NQF#0431). CMS notes its proposal includes the
addition of this measure to the hospital OQR Program for the FY 2015 payment determination,
using the NHSN component for HCP and refers to its discussion of the rationale for using this
measure. As noted earlier, Premier’s supports HCP influenza vaccination, but remain concerned
about the challenges to implementing a tracking measure to report the HCP coverage using

For HOPDs, we recommend delay in reporting until the CY 2016 payment determination, with
the hope that the feasibility of using this NHSN module may be more realistic by CY 2016,
particularly for availability of vendor support. This is the time frame CMS proposes to apply to
the ASC quality reporting, with data collection applying to immunizations from October 1, 2013,
to March 31, 2014. However, in light of the challenges under the new ASC OQR program,
Premier recommends that the ASC 2016 payment determination use data collected
between October 1, 2014, and March 31, 2015, for payment in CY 2016.

In closing, Premier appreciates the opportunity to submit these comments on the CY 2012
outpatient PPS proposed rule. Please do not hesitate to contact Danielle Lloyd, senior director for
reimbursement policy, at 202.879.8002 or Danielle_Lloyd@PremierInc.com if you would like to
discuss further.


Blair Childs
Senior vice president, Public Affairs

To top