Performance

The HCP proposed standard that is of least interest to URAC, at least if you look at the weight it would be assigned, is HCP 7, which reads:

Organizations have a written plan to aggregate data from various sources using an appropriate methodology for the purposes of health care performance reporting on physicians and other providers. [1]

Even the interpretive information that URAC provides is sparse:

Collecting and aggregating data, including data collected across various payers, helps to achieve a more representative and valid sample.

There is little reason to think that an organization would need to do more than to submit such a written plan at the AccreditNet phase of the process.  It seems unlikely that this would be the subject of any activity during the onsite review. 

Of all the new standards proposed for the revised Health Plan/Health Network v. 5.1 standards, this is the most significant, if only for its length and therefore for the relative weight in the new scoring system.  It reads:

If the organization creates programs for physician or other provider performance measurement, reporting and tiering (including rating, ranking or measurement designations): [--]
(a) Accurate, reliable, and valid measures must be used, and [--]
(i) Endorsed by the National Quality Forum where available; and [3]
(ii) If not available, then the Ambulatory Quality Alliance, the AMA-convened PCPI, or national accreditor-endorsed measures are used; and [3]
(iii) If none of the above are available, then measures based upon nationally recognized practice guidelines are used. [3]
(b) Measures used must be: [--]
(i) Clearly delineated; and [3]
(ii) Appropriately risk adjusted to account for characteristics of the physician’s or other provider’s patient population such as case mix, severity of patient’s condition, comorbidities, and outlier populations. [4]
(c) Categories of measurement cannot exclusively include cost efficiency, but must also include quality of performance. [M]
(i) Both types of measures are calculated and reported separately; and [M]
(ii) If combined, the individual component scores and their proportion of the combined score must be disclosed separately. [M]
(d) When comparing or ranking physicians and other providers: [--]
(i) The relative weight of each measure and each type of measure must be publicly available; [4]
(ii) Cost efficiency measures must compare physicians and other providers by specialty area or peer group and geography. [4]
(e) Data: [--]
(i) Must describe statistical basis for number of patients (data) used; and [3]
(ii) Must use current data consistent with the proper timeframe to attain adequate sample sizes. [3]

As you can see, it has 3 mandatory subsections and 30 other "scoring points", making this about as consequential a standard as you'll find anywhere in the revised standards.  

This standard does five basic things:

1. Establishes a minimum threshold for the source of the measures to be used the performance measurement system;
2. Requires risk adjustment;
3. Insists on the inclusion of quality measures and thereby prohibits cost-efficiency-only performance measurement;
4. Creates standards of fairness in the use of performance measures to compare providers; and
5. Establishes minimum requirements -- albeit vague ones -- for the statistical validity of the measures being used, with particular attention on sampling methodology.

One can assume that the documentation for this standard will need to be extensive.  Most of it, I'm imagining, will be submitted during the AccreditNet phase of the application process -- extensive P&Ps, program descriptions, and documentation supporting the choices made about measurement selection and sampling methodology.  The onsite review, I'm guessing, will hold few surprises, assuming the initial documentation submission is sufficient to get past the desktop review.

The proposed new standard, HCP 5, establishes minimum requirements for the due process that must be allowed to providers who choose to appeal an organization's performance measurement.  It reads:

If the physician or other provider makes an appeal within 30 days of receiving notice from the organization, then no change will be made to the information or ranking and it will remain unpublished until such time as the dispute is resolved. [4]

By requiring the health plan or health network to resolve a timely-filed provider appeal before it can implement a performance measurement publication or use in tiering, this standard places a premium on handling appeals promptly.  Think of it this way -- if you publish or use in network tiering a new or newly revised performance measurement system when there is an outstanding appeal that was timely filed, you have violated this standard.

This suggests what will be useful information for reviewers.  Certainly, at the AccreditNet phase you'll need to submit a P&P that mirrors the standard.  The onsite review likely will look like a two-fold request -- one for the a list of the dates of any launch or revision to your performance measurement system and the other for a list of provider appeals.  The reviewer will then cross check the two lists, looking for publication before appeal resolution.

HCP 4, a proposed new standard, would read:

The organization shall have a reasonable, prompt, and transparent appeals process to address physician and provider performance measurement issues that provides: [4]
(a) A written response back to the physician or other provider regarding the outcome of the appeal that includes: [--]
(i) A reason for the appeal decision; and [4]
(ii) An indication of the next actions that the organization will take as a result of the appeal decision. [4]

This standard creates an appeal mechanism for providers who are aggrieved by an organization's performance measurement with the right to appeal. That right includes a right to a written response that explains the decision on appeal and what response, if any, the organization has to the appeal.

Documentation requirements at the AccreditNet phase of the application process, while not yet published, are likely to be the submission of a policy and procedure describing the appeal mechanism. The onsite review process is likely to involve providing the reviewer with a list of all appeals submitted, from which the reviewer will select a sample in order to look at the documentation of the appeal, and particularly the written response to the provider.

Because this standard has a couple of technical issues with it, I today submitted the following comment to URAC about this standard:

First, I'd like to make a comment that applies to all the HCP standards. "Physicians" are a subset of "provider", so all references to "physicians and other providers" or, as in the case of HCP 4, "physicians and providers", offer superfluous language. "Provider" should suffice.

Second, a basic rule of outlining is to not create numbered or lettered lists out of only one item. In this case, the standard has an "(a)" without a "(b)".

I recommend, therefore, that this standard read as follows:

The organization shall have a reasonable, prompt, and transparent appeals process to address provider performance measurement issues that provides a written response back to the provider regarding the outcome of the appeal that includes: [4]

(a) The reason for the appeal decision; and [4]

(b) An indication of the next actions that the organization will take as a result of the appeal decision. [4]

This new standard would read:

The organization must give physicians and other providers 45 days notice regarding material changes to its measurement programs. [4]

Like HCP 1, this is a reflection of URAC's value of transparency: your processes should be open to scrutiny, particularly by the people and institutions that are likely to be directly impacted by those processes. However, by requiring advanced notice, we start to get into the realm of due process, by making sure that affected providers have enough time to object to the organization's proposed changes before they go into effect.

The scope of the standard is fairly broad, as a number of types of changes in the performance measurement system are covered: "Examples of “material changes” to an organization’s measurement program may include, but are not limited to, a change in measures, units of measure, sampling methodology, or use of the information."

The impact of this standard is to take away some of the organization's flexibility in changing measurement processes "on the fly." Affected HMOs and PPOs will have to build this lag time into their implementation plans.

How will URAC reviewers assess compliance? We have a hint from P-MR 4, which requires that the organization notify consumers before making a change in covered benefits. For that standard, the reviewers request a list of all benefit changes, and then ask for a dated copy of a letter or other notification to consumers for one or more of those benefit changes. One can expect a similar approach here -- the reviewer will request a listing of all changes to methodology, then dated, written copies of notifications to affected providers of those changes.

So, as you can see, the premium will be on documentation, both of your changes in methodology and notification of providers. Remember the URAC mantra: "If it isn't documented, it didn't happen."

For the most part, the proposed changes to Health Plan and Health Network in v. 5.1 are modest -- too modest for my taste in the case of provider dispute resolution as I discussed in an earlier post.  However, what ought to get the attention of HMOs and PPOs that publicly report physician or other provider performance and/or establish tiered networks or differentiating physicians based upon quality and cost data is that URAC proposes eight new standards called "Health Care Performance."  I will devote this and the next several posts to explaining these standards, which are the first significant modification to the Health Plan and Health Network Standards since the late-1990s and v. 2.0.

The proposed HCP 1, which would be in both Health Plan and Health Network accreditation programs, would read as follows:

If the organization creates programs for physician or other provider performance measurement, reporting and tiering (including rating, ranking or measurement designations), the organization will display prominently in its communications to consumers: [--]
(a) Where its physician or other provider performance ratings are found; [4]
(b) A statement that performance ratings are only a guide to choosing a physician or other provider and such ratings have a risk of error and should not be the sole basis for selecting a physician or other provider; [4]
(c) Information explaining the rating system; [4]
(d) Any limitations of the data the organization uses; [4]
(e) How physicians or other providers are selected for inclusion or exclusion in a tier; [4]
(f) Details on the factors and criteria used in the organization’s rating systems; [4]
(g) How consumer and purchasers’ perspectives were incorporated in the development of the physician or provider reporting program; and [4]
(h) How a consumer may register a complaint or comment on the ranking or quality information of a physician or other provider to the organization. [4]

What jumps out with this standard is that it has eight subsections, each weighted 4, for a total of 32 weight-points for this one standard.  None of the subsections is mandatory, but you get a sense of how important transparency to consumers is to URAC from these choices about scoring.

URAC has not fleshed out its interpretive information completely (still no statement about what kind of documentation would be required).  However, one comment that URAC does provide are worth repeating here:

The unit of measure for performance measures may be an individual physician, a group practice or practice location, as well as other types of health care practitioners and delivery systems (e.g., clinics, IPA, PHO, etc.).

As we'll see in subsequent posts, these new HCP standards are not only about transparency.  They also provide due process for aggrieved providers and specific requirements regarding what kind of measures may be used in these reporting or tiering systems.