Clinical Quality Measures and EHRs
It’s tempting to look at the Clinical Quality Measures portion of Meaningful Use as just another hoop to jump through, with little actual clinical impact on day-to-day care. But, upon a little reflection, embedding Clinical Quality Measures (CQM) into Electronic Health Records (EHR) systems is actually quite revolutionary.
What are CQMs? They are any of a number of measures developed through a deliberative, evidence-based process, under the auspices of the National Quality Forum (NQF), which measure various aspects of clinical care. Some of them are process measures – things like “what percentage of my diabetic patients had an LDL-cholesterol level drawn?” Some of them are more endpoint measures – “what percentage of my diabetics have LDL-cholesterol levels less that 100?”
The NQF process for evaluating CQMs makes sure that the items (1) can be reliably measured, and (2) make a difference in health outcomes. They are based on data, on evidence, and serve as a basis for standard-of-care.
A little historical perspective
CQMs have been around for quite a long time – they have been used extensively since the mid 1990s, and in more limited settings before that. They have been used by payers – private insurance companies and CMS (Medicare) – in order to create measures of community health, as well as to build “report cards” for health plans, hospitals, medical groups, and individual physicians.
Traditionally, quality measures have been based on administrative data – information gleaned from bills submitted to Medicare and to insurance plans. Therefore, the kinds of measures that have been used have been based on those things that appear on bills – diagnoses (from the ICD9 codes on bills), dates of service, office visits and procedures performed (from the CPT codes on bills). They also are drawn from medications (from billing codes submitted by pharmacies – doctors do not include medications on their bills to insurers, but pharmacies do), and from lab tests performed (from billing codes submitted by laboratories – therefore the lab tests done, but not the results of those tests, are collected by the payers).
CQMs generally have a numerator and a denominator. Denominators (“how many of my patients are diabetics?”) can be gleaned from billing data pretty easily. Some numerators can be gathered from basic billing information (“how many of my diabetics have had an LDL-cholesterol done?” can be determined from laboratory billing data). However, some other kinds of numerators cannot be easily gotten from bills alone (“how many of my diabetics had LDL-cholesterol levels less than 100?”).
To try to capture CQM numerator information that is not included in bills, CMS (in conjunction with the AMA, who created the CPT billing system) has developed a set of zero-dollar reporting-only billing codes (the CPT-II codes) to try to capture information not ordinarily found on bills. Additionally, CMS has extended the CPT billing-code system with their own HCPCS codes, to more specifically code certain kinds of procedures (using the G-codes) for this kind of measurement.
The only problem is that doctors have not really paid systematic attention to routinely including CPT-II codes with their bills, since no money is attached to those codes. The benefits of using them is thus limited.
Pay for performance
Since 2006, Medicare has had a quality-incentive program in place, the Physician Quality Reporting System (PQRS, formerly PQRI) that collects CQM information (based on billing, using CPT-II codes to capture some numerator data). Physicians who successfully report on a minimum number of measures receive a bonus from CMS in addition to their ordinary Medicare payments.
Being optional, and inconsistently taken up across the country, the PQRS effort, though attempting to promote a business case for improved quality of care, has been criticized as merely being a “pay for reporting” program, rather than true pay-for-performance. Nevertheless, PQRS has been a model for federal-level quality-payments, and the program continues to evolve.
At the same time, in the private sector, in places in the country (such as California) where Managed Care has continued to work, group-level CQM measurement has been commonplace for over a decade. Medical groups and Independent Physician Associations (IPAs) that assume risk from HMO payers (they receive a fixed payment, per-member-per-month, from the health plan, and then are responsible for paying their member physicians according to internal formulas within the group) can also get an additional Pay-For-Performance (P4P) bonus if they demonstrate good performance at the group level in CQMs. This is a competitive arena, where different groups vie for dollars, based on how well they do in their CQMs – groups that do better get more money. This is a “business model for quality” that has been the focus of the Integrated Health Association (which brings together medical groups, health plans, and hospitals to all agree on a common set of measures that everyone will report on).
Moving CQM into EHRs
The history of CQMs, therefore, has been that payers – private health insurance plans, and Medicare – have collected these measures based on billing data. Managed-care medical groups have also collected these measures, based on billings from individual participant doctors sent to the group, and supplemented with (often) manual data entry to capture numerator data not found on bills.
The bottom line – rank-and-file physicians have not been directly involved in measuring their own CQMs. Outside parties (payers and medical groups) have been where the measurement has occurred, and has been done on billing data that has been submitted. Little extra work has been asked of physicians.
So the notion of embedding CQM measurement and reporting within the EHRs that physicians on the front-lines actually use is quite revolutionary. Never before has the observation of CQM data been so direct, so immediate, and so modifiable based on changes in behavior at the point of care.
Further, by being based on direct EHR data, rather than second-hand from billing data, CQM measurement from within an EHR is arguably more accurate. It also involves innovation on the part of EHR developers to create ways to capture “numerator” and “denominator” information systematically within the product, directly from clinical data (rather than from billings).
CQMs and Meaningful Use
The threshold of CQM reporting for Meaningful Use is actually a very low bar. It does not matter what the actual results are – all the numerators can be zero, in fact. It merely establishes a precedent for reporting, and the Meaningful Use CQM criterion is simply a yes/no (“did you report CQMs to CMS?”)
Of course, this is only a preliminary step – one that gets physicians used to measuring CQMs themselves, directly (rather than having some third party report on them, for them).
What comes next? Two areas come to mind.
(1) Practices and groups (such as ACOs) may end up competing for higher payments from payers, based on demonstrated CQM-based quality. This is the model seen in Managed Care in the California IHA experience. It is a business case for increased quality, rather than simply a pay-for-quantity, which has been the Achilles heel of the current fee-for-service approach to health care delivery.
(2) Clinical Decision Support can be based on these CQM items. This is the most likely next-step for Decision Support – if a CQM measure is “due”, then a prompt at the point of care can be built into the EHR. “This patient is a diabetic, and is due for an annual diabetic eye exam” (process prompt), or “this diabetic patient’s most recent LDL-cholesterol was above target – consider making changes to the therapeutic regimen” (endpoint prompt).
When this kind of immediate feedback is available, and physicians get used to capturing and self-measuring the CQMs appropriate to their practice, an actual change in the standard of care can be seen. This has been the experience with IHA-based P4P in California – the CQM curve (not just the cost curve) has been bent as a result of focusing attention (and dollars) on quality measurement. And modern EHRs, with embedded CQMs, are a necessary foundation for such change.