CSSC finally uploaded the long awaited guides for Encounter Data Submission to their website last Friday.

Here are links to the files (pdf):

Encounter Data FAQs

Encounter Data Slides

ED Participant Guide

EDCompanionGuide_837P – Professional

EDCompanionGuide_837I – Institutional

Or, if you want to grab them from the CSSC website: Use this link:

Encounter Page

For those of you who were not able to attend the 10.29.2010 national meeting on the CMS Plan to have MA plans submit Encounter Data, here is the presentation (pdf):

EncounterData National Meeting 10.29.2010 [click to view OR right-click to "save as" to computer]

The goal is to have MA plans ready to submit full encounter data  for professional and institutional claims by January 3, 2012.

Here are the top Medicare Advantage plans by Part C enrollment.

Data from cms website: 07/2010 enrollment data: http://www.cms.gov/MCRAdvPartDEnrolData/EP/itemdetail.asp?filterType=none&filterByDID=-99&sortByDID=2&sortOrder=descending&itemID=CMS1237234&intNumPerPage=10

Did not count Part D plans, sorted by “Parent Organization”.

**ADDENDUM – 9.21.2010 – add Universal American to this list – 291, 814- Thanks, Matt

Rick Bowles, Principal, Ingenix Consulting, and I will be co-hosting an HCC Webinar sponsored by Franklin Research Associates on the topic of HCC Coding Red Flags: Strategies to Avoid Common Errors Losing Revenue.  The webinar is scheduled at 10:30 ET on Tuesday, July, 21, 2009.  This is the first of a 3-part webinar series, and more details can be found:here.  I will be speaking to common HCC documentation and coding errors that your providers should be made aware of, and Rick will be speaking to the incorporation of Health Risk Assessment data in the risk adjustment process.

The target audience is especially HCC Program Managers, Provider Education/Provider Relations, HCC Auditors.

Here is an interesting research article from the Center on Budget and Policy Priorities about HCC upcoding- pdf. I do not particularly agree with their conclusions, but it is an interesting summary of a common (although, I feel erroneous) viewpoint that Medicare Advantage plans are somehow systemically gaming the system by upcoding patients without justification only to inflate their payments from CMS. Part of the confusion stems from how folks use the word “upcoding”. I think it is often erroneously applied to situations where risk scores have improved due to actually better documentation and ICD-9 code submission. The CBPP’s assumption is that risk scores have increased solely due to an upcoding phenomenon, discounting the possibility that risk scores are up due to closing the gap between diagnoses in claims data and diagnoses that are in the charts but otherwise never made it to the claims data to begin with.

So, risk scores are improving due to better code capture, not, I believe, due to a conspiracy to submit codes that are not medically justified. That being said, there are probably subsets of providers who share capitated risk with a plan or who are engaged in poorly constructed incentive payments from MA plans who may be under the unfortunate influence of having financial incentives be tied to patient’s diagnoses or risk score. The ideal provider incentives are those where the amount of incentive is disconnected from the risk score or number of diagnoses submitted. For example, a flat fee per chart review, independent of how many HCCs are resubmitted. Or a flat fee per chart pull to let outside auditors review charts. Or a flat fee to review an annual or semi-annual problem list during a patient visit and document those conditions that are active, or are inactive but influence medical decision making.

So, the biggest concern for plans should be minimizing risk aversion by ensuring proper HCC coding and documentation, i.e. that ICD-9 codes being submitted to the risk adjustment database are supported by proper medical documentation, so this is less an issue of overcoding, rather an issue of proper documentation.

When plans are constructing their bids, disease management costs can be categorized as either medical expenses or administrative expenses depending on where the DM activity occurred.  If the DM occurred in a clinical setting by a provider, then the DM is treated as a medical expense.  Other disease management programs by the plan, and their costs to advertise and enroll members are treated as administrative expenses.   Durable Medical Equipment related to disease management activity is treated as a medical expense.

MedPAC just released their March 2009 report on payments to MA plans. It contains an excellent discussion of why CMS pays MA plans an average of 114% of the cost of taking care of FFS patients. Simply, this is based on statutory benchmarks in each county that were historically high in order to attract companies to offer private Medicare Advantage products in all parts of the country. Interestingly, MA plans in total are bidding to take care of medicare patients at around 101-102% of the cost of FFS patients. MA plan HMOs typically bid under the FFS rate, actually bidding at around 98% FFS costs. 75% of the difference between the MA plans bids and the higher 114% benchmarks is passed on to plans to either offer more services like vision or dental, or to reduce Part B premiums or to reduce beneficiary cost sharing. So, CMS is paying for these extra benefits.

Here is the MedPAC March 2009 pdf report.

Here is a link to ACAP’s (Association for Community Affiliated Plans) download page for SNP or Special Needs Plans related articles.  ACAP’s mission is to support not-for-profit plans like SNPs.

CMS just released the 2010 Advance Notice for MA payment.  Download pdf here. For the 2010 payment year there will be a projected downward adjustment of 3.74% to account for the different risk factor score averages between FFS and private Medicare beneficiaries. This factor will be finalized for the April, 2009 Announcement.  Interestingly, the downward adjustment is not solely due to increased coding intensity among MA plans, rather, CMS projects that 50% of the difference is due to enrollment patterns. Specifically, FFS attracts more newly-eligible-to Medicare benes who are healthier and have lower scores – this depresses the average FFS score. The other half of the FFS – private difference is due to increased coding intensity.

Patients under MA plans, for the 2006-2007 year, had on average a higher disease risk score by 0.025 compared to similar FFS cohorts.  CMS specifically looked at “stayers”, benes who had stayed within Medicare Advantage for at least 2 years to find these coding patterns.

This risk score enrollment/coding intensity adjuster is just a temporary phenomenon until CMS gathers enough MA plan data to base the CMS-HCC model on MA patients only. The current and previous iterations of the CMS model (i.e. risk score values, annual denominator value) are based on historical FFS data. Newer versions of the model will build-in the enrollment/coding intensity assumptions into the risk scores themselves and as such there will not be a need to always have a separate coding/enrollment adjuster per se.

Interesting idea from Al Lewis as it appears on thehealthcareblog.com. One way to help fix Medicare Advantage (i.e. reduce the up to 112% of costs compared to FFS in some areas) is to make enrollment automatic in areas that have MA plans available – which is everywhere – so that the patient would have to opt-in to FFS. This would bypass the $500 or so commission that brokers get – quite a nice savings per member.