PQRI: 2007 Results Prompt Changes
First-year results prompt changes to the heart of the program.
By Julie Orton Van, CPC, CPC-P
he first year’s results for the Physician Quality Reporting Initiative (PQRI)—the initiative to incent the use of and track standard quality initiatives—reflect the influence of a complicated new recipe. Not all chose to try it, and of those who did, only half found success. After such a start, how does a practice make PQRI participation worth the effort?
According to Centers for Medicare & Medicaid Services (CMS), data from 2007 showed approximately 16 percent of eligible professionals participated in the program. Of those who participated, just over half were successful in meeting the program and reporting requirements and, as a result, qualified to receive an incentive payment. Nationally, 109,349 National Provider Identifier/Taxpayer Identification Number (NPI/TIN) combinations (15.8 percent of eligible NPI/TIN combinations) submitted at least one quality data code (QDC).
The CMS collected quality information for services furnished from July 1, 2007 through Dec. 31, 2007. CMS paid $36 million in incentive payments to health care professionals who met the criteria for satisfactorily reporting data during the 2007 PQRI. These incentive payments were sent out starting in July 2008. So far, the average individual eligible professional incentive amount is more than $630. The largest incentive payment for a group practice is $205,795, and the average incentive is $4,713. To be eligible for an incentive payment, health care professionals are generally obligated to report at least three quality measures for at least 80 percent of the cases in which the measure was applicable. The 2007 budget for the PQRI was $150 million, meaning more than $114 million was not reimbursed.
CMS is not only looking closely at the mixed results of the program, they’re sharing information that will help us succeed.
For a valid claim submission, the QDC reported must apply to the patient according to the measure specifications (age, gender, diagnosis, and procedure). The claim must also include the rendering professional’s NPI, and otherwise comply with PQRI QDC submission business requirements. Of those 109,349 NPI/TIN combinations
92.5 percent validly submitted at least one QDC;
64 percent validly reported quality data on 80 percent of eligible cases for at least one measure; however, many didn’t earn an incentive because they didn’t meet the criteria for satisfactory reporting;
52 percent earned an incentive payment (met satisfactory reporting criteria by reporting data on one to three applicable measures for 80 percent of applicable cases); and
1 percent was subject to the cap while the rest qualified for the full 1.5 percent incentive payments.
A total of 14,089,837 QDCs were reported:
51.6 percent were submitted validly
48.4 percent were submitted invalidly
In the Dec. 3, 2008 CMS publication Physician Quality Reporting Initiative (PQRI) 2007 Reporting Experience, CMS said that since it began accepting the quality data in July 2007 for the 2007 PQRI, it identified and began to remedy issues and questions raised about the 2007 PQRI results and feedback. CMS result analysis of the completed first cycle of reporting has identified unanticipated issues they believe may have impacted physicians and other professionals’ success in meeting reporting quality data program requirements. These issues, which are outlined in more detail in the report, include claims-based reporting mechanisms issues, NPI numbers not included on the claims forms, incorrect quality reporting data or claims submission errors, and the feedback reports’ content.
Valid quality data reporting could only be determined if the measure specifications were adhered to. Based on the CMS review, the following invalid quality data submissions were caused by eligible participants (EPs) failing to adhere to measure specifications when submitting QDCs and denominator codes, which were established through a board consensus process that included the physician community.
Incorrect HCPCS denominator code: 18.9 percent. A QDC was submitted and a HCPCS (CPT® or HCPCS Level II) denominator code was submitted, but the HCPCS code submitted was not one that was appropriate for the measure. Each measure requires submission of a HCPCS (procedure or service code). A measure that had a high reporting error due to this reason was measure No. 30 (Perioperative Care: Timing of Prophylactic Antibiotic—Administering Physician). This was the only measure requiring the submission of a CPT® Category II code for the denominator rather than a HCPCS Level II procedure or service code.
Incorrect diagnosis code: 13.9 percent A QDC and a diagnosis code were submitted, but the diagnosis code was not the diagnosis appropriate for the reported measure. Incorrect diagnosis reporting rates were high for measures requiring multiple diagnoses, eg, No. 5 (Heart failure: Angiotensin-Converting Enzyme (ACE) Inhibitor or Angiotensin Receptor Blocker (ARB) Therapy for Left Ventricular Systolic Dysfunction), and no. 8 (Heart Failure: Beta-Blocker Therapy for Left Ventricular Systolic Dysfunction).
Incorrect HCPCS and diagnosis code: 7.2 percent. A QDC was submitted and a HCPCS code and diagnosis code were submitted, but neither the HCPCS nor the diagnosis code is in the reporting denominator for the applicable measure’s QDC. A measure with a high reporting error due to this reason was measure No. 7 (Beta Blocker Therapy for Coronary Artery Disease Patients with Prior Myocardial Infarction). This measure required the reporting of an ICD-9-CM diagnosis code to identify patients with a coronary artery disease diagnosis and a MI diagnosis and a CPT® service code for the denominator.
Incorrect age: 6 percent. The QDC was submitted for a patient outside of the measure’s age parameters. Many measures are limited by age parameters. Those that only apply to the pediatric or younger adult age population experience very high error rates, eg, measures no. 53 (Asthma; Pharmacologic Therapy for Ages 5 to 40), and no. 64 (Asthma Assessment for Ages 5 to 40). Even those with upper age ranges of age 75, such as the diabetes measures No. 1 (Hemoglobin A1c Poor Control in Type 1 or 2 Diabetes Mellitus for Ages 18 to 75), No. 2 (Low Density Lipoprotein Control in Type 1 or 2 Diabetes Mellitus for Ages 18 to 75), and No. 3 (High Blood Pressure Control in Type 1 or 2 Diabetes Mellitus for Ages 18 to 75), experienced large error rates.
Incorrect gender: 2 percent. The wrong patient gender was submitted for the measure’s QDC. Only a few measures are limited by gender, eg, measure no. 39 (Screening or Therapy of Osteoporosis for Women Aged 65 Years and Older).
Claims Submission/Split-Claims Errors
Valid reporting could only be determined if the QDC was submitted on the same claim as the associated measure’s billing and diagnosis code(s). In searching for technical issues to explain a portion of the invalid QDC submissions, we separated QDCs that were submitted invalidly because of missing HCPCS codes, not just an incorrect HCPCS code as described above. This was done because a missing HCPCS code could be an indicator that the claim was split (for example, separated into smaller claims) prior to submission or during the carrier or MAC claim processing; the initial single claim would have arrived in the claims data file used for the PQRI determinations as multiple claims.
CMS found that 6.3 percent of QDCs had no HCPCS codes on the claim. When claims are split prior to submission to the carrier or MAC by the EP’s billing software or clearinghouse software, there isn’t an identifier on these claims indicating that they were originally part of a single claim. For analysis purposes, it was assumed these split claims were separate, unrelated claims. To identify these circumstances they reviewed all claims containing only QDCs to determine if there were other claims with HCPCS codes for the same beneficiary, service date, TIN, and NPI. The 6.3 percent of QDC submissions without HCPCS on the claim can be broken out even further below:
Only QDC on claim: 5 percent. A QDC was submitted on a claim, but there wasn’t a HCPCS code on the claim to determine the measure’s appropriate denominator population. All measures must contain at least one HCPCS code in the reporting denominator.
QDC and incorrect diagnosis code on claim: 1.3 percent. A QDC was submitted on a claim, but there wasn’t a HCPCS code on the claim and the diagnosis code reported wasn’t the required diagnosis code for the measure’s denominator population. All measures must contain at least one HCPCS Level II code in the reporting denominator. However, not all measures require a diagnosis code to determine the denominator population.
CMS found that for the 32.8 percent of claims with only QDCs, they could match the claim with another claim containing a HCPCS code and diagnosis code with the same patient identifying information and service date. CMS estimates this technical issue’s impact as 2 percent overall. This could affect professionals who otherwise would have qualified for satisfactory reporting for 2007.
CMS was aware when developing 2007 analytics that in some cases carriers/MACs split claims for processing. To account for this, their contractor established a routine for 2007 analysis to reconnect these split claims; however, the routine was limited to claims containing 13 or more line items. In conducting a recent review, they found 2.4 percent of claims with only QDCs present had been split by the carrier/MACs. This amounts to only 0.2 percent of QDC submissions. Now that they are aware of claims with fewer than 13 lines being split by carriers/MACs, they will modify analytics for the 2008 PQRI program and future years’ programs to reconnect any claims split by carriers/MACs.
No NPI on the Claim
In some cases, the business rule requiring the NPI for valid quality data submission and analysis had an impact on determining whether an EP satisfactorily reported. A total of 12.2 percent reported did not include an NPI on the line where the QDC was reported. Lack of an NPI on the claim or line may have been caused by the EP’s failure to include it in their billing software or clearinghouse software.
When reviewing the missing NPI issue, CMS tracked representative cases from the initial claim all the way to the claims warehouse, the database used for the satisfactory reporting determination, and the incentive payment calculation. They have not encountered NPI mishandling once the NPI was received by the carrier/MAC in the appropriate place on the claim (ie, as the rendering NPI for the HCPCS and QDC line item codes). They did, however, encounter situations where EPs’ electronic data interface software or clearinghouse processes led to not submitting or incorrectly submitting NPIs (such as transposing the NPI from the line item to the referring NPI field). In these circumstances, the requirement of the NPI appearing on the line item for the HCPCS Level II and QDC was not met.
Other Unsatisfactory Reporting Issues
When conducting additional reviews of the data submitted for the 2007 reporting period, CMS found a data issue involving claim types for certain durable medical equipment (DME) items included in the 2007 analysis. In the preliminary review, particular claims which were submitted correctly by EPs to the carrier/MAC and sent on to the National Claims History (NCH) system also included data elements (HCPCS codes, diagnosis codes and the EP’s NPI), giving the appearance of claims that should have a QDC included for a measure. CMS did not adjust for these claims in the analysis for 2007, so they were included. Consequently, this may have caused a falsely inflated measure for an affected EP’s denominator population, potentially resulting in the EP’s unsatisfactory reporting a 2007 measure and not being eligible for an incentive payment.
CMS received concerns about difficult-to-access or complicated feedback reports. Others have asked for more frequent reports. They are investigating other avenues to help EPs access the reports, but they do not expect to make any changes impacting physicians and professionals who have already established security accounts.
As for report substance, some professionals found the reports contained too much information. In response to these comments, CMS will simplify the reports and plans to list only measures the EP actually submits. They will also provide more detailed information to help submitters understand the submitted data analysis.
CMS agrees that using more frequent and up-to-date reports would be helpful; however, they face certain practical limitations that make it difficult to achieve that goal. After a reporting period closes, there is a time lag to allow for completion of submitted claims. Where possible, they will provide information about reporting errors.
Keep Moving Forward
CMS has put together an education and outreach plan allowing EPs and their staffs to easily obtain information about the PQRI program. For the 2009 program, CMS will do the following:
• Conduct monthly National Provider Calls focusing on important topics for reporting 2009 programs’ PQRI measures.
• Update and expand the frequently asked questions available on CMS’ PQRI Web site at www.cms.hhs.gov/PQRI.
• Post the updated measure specifications used in the 2009 program.
• Develop and post updated tip sheets and fact sheets for the 2009 program to assist EPs in reporting quality data to CMS by the claims-based system or through qualified registries.
• Conduct more educational sessions with the carriers/MAC contractors.
• Continue to provide speakers at local, regional, and national conferences on PQRI topics. CMS regional office staff will also continue to be a resource to EPs in providing assistance, education, and outreach activities at the local level.
• Develop a Web-based education course on PQRI and the E-Prescribing Incentive, offering continuing medical education credit to EPs (a new activity for 2009).
• Actively partner with the American College of Physicians and the American Academy of Family Physicians to conduct and develop additional education and outreach materials/activities to increase participation in PQRI.
• Widely share information learned from the 2007 program experience.
You can find more information about the 2007 PQRI experience by going to the CMS Web sit.
Don’t Wait For CMS’ PQRI Feedback
As noted in a recent article, “New AMA Survey Highlights Need to Improve Medicare’s PQRI,” posted on the American Medical Association’s (AMA) Web site (www.ama-assn.org/ama/pub/category/20208.html), “Key elements of Medicare’s Physician Quality Reporting Initiative (PQRI) must be improved so that physicians can successfully participate and use the information to increase the quality of patient care. This is the main take-away from the AMA new survey of physicians who participated in the PQRI during its first year of implementation.”
According to AMA board member Ardis Hoven, MD, more than six out of 10 physicians surveyed rated the program difficult and only 22 percent were able to download the PQRI feedback report for their practices. Hoven said, “Physicians who began reporting in July 2007 did not receive a feedback report until 12 months later, halfway through the program’s second year, making it impossible to fix any reporting problems. This may have contributed to the fact that nearly half of all PQRI participants did not receive any bonus payment. If reforms are not made to the program, physicians who participate in 2008 will not receive feedback reports until 18 months after initial reporting.”
Advice for Successful Reporting
According to successful PQRI participants, you should never wait for the Centers for Medicare & Medicaid Services (CMS) to report back to you. All successful practice members whom I spoke with stressed the need to institute a closed-loop process for reporting PQRI quality data codes (QDCs). What this means is for every claim you submit with PQRI data, you need to track whether or not CMS acknowledged the data and validated it toward your successful reporting percentage.
For every quality measure submitted on a claim, CMS will reject that line with rejection code N365 if it was accepted into the PQRI data base. You’ll know a measure was not successfully accepted by CMS if you do not see rejection code N365 on your remittance advice for claims you thought you reported QDCs. Each instance of the “missing” rejection code needs to be examined and researched. This is the only way you can identify what you are doing wrong in a timely manner. Once you determine what you were doing incorrectly, immediately make the necessary changes in your reporting process. Validating in this manner is really no different than reconciling your accounts to contractual terms for private payer contracts. You definitely do not want to wait for CMS or any other payer to report this to you 18 months after the fact.