By Rod Baird w/ clinical review by Bob Wells, MD CMD & Enrique Varela, MD CMD

PQRI is shorthand for The Physicians Quality Reporting Initiative, which was established by law and implemented by CMS in 2007. Clinicians who participate in the program, by reporting on selected ‘quality measures’, can earn an incentive payment. Currently that payment is set at 2% of allowable Medicare Charges. Each quality measure applies to a defined set of individuals. The criteria for the set include patient demographics (age, sex, and diagnosis) and Clinical parameters (type of treating provider and CPT code).

ECP had one physician (Dr. Varela) participate in PQRI in 2008. He successfully used the Diabetes Measures Group as his method of participation. We engineered a strategy where Enrique was able to document PQRI measures fo a group of sequential visits to Diabetic Patients under 76 years of age (the eligibility criteria for this Measure group). Results were submitted via the NC based Registry DocSite. Selection of the registry was based on the ability to submit data with no charge for physicians who are members of the NC Medical Society. Normal submission fees at DocSite are $350.

In 2009, an additional eight clinicians are participating in PQRI via use of the Diabetes Measures Group. We plan to perform 8 submissions via a registry, and one via the Claims Based method of submission. The final success of either method of submission is indeterminable until ~ October, 2010 – it takes CMS that much time to assemble and evaluate data.

There are two overarching methods of PQRI participation – by use of individual measures, or use of a multiple related measures which are assembled into a Measures Group. There are distinct advantages to use of the Measures Group participation strategy – the clinician only has to submit data on a maximum of 30 unique patients during the year. If individual measures are used, the threshold is 80% of eligible events (an event may be either per encounter or per patient, based on the measure’s definition).

In 2007-09 there were 7 measures groups available; the Diabetes Measures Group was the only one which applied to the CPT code family (99304 – 99318) used by the majority of ECP clinicians.

For ECP, using the Diabetes Measures Group was a challenge because:

• Our elderly patient population has a relatively small number of individuals who are both <76 y/o and have a diabetes diagnosis.
• Several of the measures in the Diabetes Measures Group require the completion of external events (consultations or laboratory tests) as a prerequisite for achieving the “Meets Performance” scoring. For some other measures, the ‘Meets Performance’ standard require levels of medication of dietary control (i.e. Hemoglobin A1c, B/P, LDL-C)  which may be incompatible with the best treatment choices for the LTC population. Those measures don’t have a reporting option that exempts the patient for clinical reasons.

These limitations restricted participation to clinicians with a high volume of patients – sufficient to have the required sample size of 30 active patients. Many clinicians who participated elected to report results which indicated ‘Performance Not Met’ instead of subjecting patients to added medications or dietary restrictions inconsistent with the desires of the patient and/or family.

For the period of 2009 and 2010, there will be public reporting by CMS about participating clinicians – it is limited to reporting on who participated in PQRI, but no reporting on their frequency of ‘Meeting Performance’. However, based on proposals published by CMS, we believe that individual clinicians will have statistics published about the degree to which their PQRI reports show they met performance benchmarks. This may start as early as 2011.

Based on2009 PQRI program rules, ECP anticipated having to elect the use Individual Measures for many of our staff in 2010 (because they didn’t have enough patients to readily satisfy the 30 patient group for Diabetes). The next year (2011) we expected to have to use Individual Measures for all clinicians. That was because of the difficulty in reporting results that met performance benchmarks for Diabetes in our population.

However, CMS just published the new PQRI Measure definitions for 2010 and some significant changes occurred. In 2010 there are 13 Measures Groups, three of which apply to the CPT code family 99304 – 99318. In addition to Diabetes, there are Measures Groups for Coronary Artery Disease and Heart Failure. Both of these measures groups apply to all patients over 17 years of age. Every ECP clinician has enough patients in each of these groups, whose care reflects the targeted benchmarks, to make successful reports. The patients who are chosen for the reports aren’t required to be ‘sequential’, which eliminates a significant complexity managing the reporting strategy and monitoring submitted claims.

Here are key strategy points we are considering as we move into 2010:

• PQRI participation via Measures Groups is less onerous – patients in the sample can be ‘nonsequential’ now. The sample size is the lesser of 30 or 80% of eligible patients (but at least 15 individual patients are required for the full 2% incentive).

• Assume your reports will be graded in the future (2011?). Therefore you should select measures which correspond with ‘best practices’ in your specific clinical setting – don’t manage the patient to the measure, select the measure to fit the patient.

• Avoid measures that require external resources to ‘Meet Performance’. A primary care physician can’t satisfactorily report on Measure #117 (Dilated Eye Exam in Diabetic Patient) until an Ophthalmologist or Optometrist performs and reports back on an exam. This measure is part of the Diabetes Measures Group.

• Consider the possible benefit of using a registry – some measures are ‘registry only’ – e.g. Heart Failure, and CAD, others – e.g. Diabetes, are reportable via Claims or Registry. You can select your 30 patient group for registry reporting after you’ve recorded pertinent data in your records, and submitted claims for payment. You can pick the group months after the actual encounter. Once your PQRI data is entered into a registry, you get immediate feedback – Registries ‘scrub’ your data and report back any apparent errors. While they don’t validate this against your previously submitted claims, at least you know if your PQRI data set meets standard reporting criteria.
  • When submitting data via claims, you only know that it was recognized by verification on a Medicare Remittance Advice – but this doesn’t prove the data conforms or that it will reach the PQRI data repository.
  • A downside to registries is the cost - $200-$400 per clinician per year to process and submit your data. As the number of registries grows, expect the price to drop. If you have a larger number of reporting clinicians, you might be able to negotiate a discount.

• Your PQRI payment is limited to 2% of you Medicare Allowable charges, there is no partial payment for incomplete performance, and no bonus for extra reports.

• Consider the amount of work required to participate. Picking three widely applicable individual measures may simplify your record keeping, but add time to every encounter, add lines to you dictation and notes, and add time to your billing staff (unless you are 100% automated).
  • If you report on 100% of your patients, the added compensation per report is 2% of your average Medicare payment per encounter (if you are 100% Medicare). If Medicare is a lesser percentage, your added payment per encounter is proportionally less. Assume you generate $120,000 of Medicare Allowable Charges a year. If you average Medicare allowable per encounter is $60, and 50% of you patients are Medicare, then you are paid $60*2%*50% = $0.60 per encounter for PQRI.
    • This revenue can be enhanced (per documented encounter) by picking measures that apply to subsets of patients – but this requires careful monitoring of records to assure you meet the requirement of reporting on 80% of eligible patients (as determined by your submitted claims).

• Conversely, picking a Measures Group which represents a high frequency condition in you practice can prove quite cost effective. Measure Groups are more difficult to report (even via a Registry), because multiple reporting codes are required of each encounter. However, the need to only report on 30 patients makes it less onerous. If your total Medicare Allowable charges are $120,000 then your 2% PQRI incentive is $2,400. Even if you choose a Registry that charges $400, that’s the same as getting $67 extra payment for each of the 30 encounters you reported.  If your Medicare allowable charges are higher than $120,000 in the example, you payment per report will increase accordingly.

• Some Measures Groups can be reported via Claims. Election of one of these Groups can save the expense of using a Registry. However, the problem inherent with all claims based submissions persists – no concurrent review of the quality of your data.

In a nutshell, here are our recommendations:

• Participate in PQRI via Measures Groups. Search this document for your common CPT codes to narrow down you choices, then see which also contain patient demographics that correspond to your population:
  • http://www.cms.hhs.gov/PQRI/Downloads/2010_GettingStartedwithPQRIReportingofMeasuresGroups_111309.pdf
• Note which reporting options apply to your preferred measure group, claims only, or claims and registry.
• Then look at the 2010 Measures Group Specification Manual (http://www.cms.hhs.gov/PQRI/Downloads/2010PQRIMeasuresGroupsSpecsManualandReleaseNotes_121809.zip ). This describes the codes used to report each measure. (note – any applicable code corresponds to ‘reporting satisfactorily’, while only a subset of codes qualify for the ‘Meets Performance’ designation. The difference is between attending class and getting a good grade.
• Determine which codes correspond to ‘Meets Performance’ by looking up each of your Measure Group’s measures in the QDC Categories table:
  • http://www.cms.hhs.gov/PQRI/Downloads/2010_QDC_Categories_111309_Print8_5x14(2).pdf
  • Verify that the parameters associated with ‘meets performance’ in each Measure correspond with your practice’s standards of care.

• Verify that your practice has sufficient patients eligible to satisfy the requirements of the chosen Measures Group.
• Educate you clinical staff about the standards of the Measures Group – the information available from CMS is of little help. The AMA, and some registries, have created data gathering tools for specific measures. This link takes you to the 2009 AMA participation tools:
  • http://www.ama-assn.org/ama/pub/physician-resources/clinical-practice-improvement/clinical-quality/participation-tools-measures.shtml

• Our practice is planning on using either the CAD or Heart Failure measures group. Since these are new, no data gathering worksheets are currently available. We know that we can perform all of our required data gathering in a month, so we’ll wait to start PQRI until later this year. We will perform initial education of staff during the winter so they can begin thinking about integrating the needed measures into their clinical routine.
• While staff treat patients, have them document PQRI measures in the clinical record.
  • Make sure the claim for the particular encounter has a valid ICD-9 diagnosis (one associated with the chosen PQRI measures group). We believe it safest to make this the #1 diagnosis on the claim.
• Devise a system where you flag the encounters which contain PQRI measures data in the patient record.
  • Track the number of encounters with the required data, and alert clinicians about their progress.
    • Verify they will reach 30 records by year’s end.
• Once your report data is gathered, organize it, and begin reporting via your chosen registry.

Copyright© Dec. 2009 by Geriatric Practice Management and Rod Baird