NCPA is touting the results of a new study that purports to provide “incontrovertible evidence of the slow reimbursement pharmacies have endured with Medicare Part D prescription drug claims.” (See the press release.) NCPA is quite forthright in their desire to influence two bills pending before Congress: S.1954 (Pharmacy Access Improvement Act of 2007) and H.R.1474 (The Fair and Speedy Treatment of Medicare Prescription Drug Claims Act of 2007).
I strongly disapprove of the manner in which this study’s results are being presented to the public and to Congress. It would be foolish to base health care policy based on the partially-disclosed results of this incomplete and biased study.
Bikini Research
The full study has not been released so we can only rely on the skimpy Executive Summary. And like a skimpy bikini, what the summary reveals is interesting -- but what it conceals is essential.
Consider this example of selective disclosure.
As we all remember, the Part D launch in January 2006 was fraught with operational difficulties, many of which were subsequently resolved. According to IMS Health, Part D scripts grew from 0% to 17% of the retail market during the first six months of 2006 before stabilizing at 17% for the second half of 2006.
Yet the Executive Summary conveniently averages all of 2006 together, thereby artificially inflating the magnitude of the "prompt payment problem."
Here are the implied median number of days between submission and adjudication as reported in the summary:
January 2006: 106 days
February 2006: 93 days (-12% vs. January)
March/April 2006: 54 days (-49% vs. January)
May through December: No disclosure
The trend is clear, but where are the data from the rest of the year? If the Executive Summary was intellectually honest, then it would have provided more information about changes over time. At a minimum, the summary should show the results for the first half versus second half of 2006.
These omitted data are crucial for interpreting the overall 2006 results. Figure 1 implies that the annualized median is 30 days for independent pharmacies and much lower for chains. Thus, the trend shown for the first four months must have continued. The median may have even dropped below 30 days for certain months in 2006!
Real-World Impact?
I published some Heretical Questions about the AMP War last month. Let me ask another heretical question: Is the magnitude of the allegedly slow payment worthy of a legislative fix?
The average independent pharmacy in this study filled 4,138 Part D scripts in 2006. The average retail script generated $60 in revenue at an independent pharmacy in 2006 (Source: NACDS). Therefore, Part D represented about $250,000 in annual revenue (about $21,000 per month) for a typical independent pharmacy in 2006. If half of these reimbursements come after 30 days, then the pharmacy is floating an additional $10,500 for a few weeks throughout the year.
Is this estimated float amount financially significant enough to require an Act of Congress? The study is silent on this matter, as well as other relevant questions. Are Part D claims being paid faster or slower than the other 90% of scripts? How are Days of Sales Outstanding (DSO) changing? What else should pharmacies be doing to better manage their balance sheet? What explains the performance gap between independents and chains? How sensitive are the results to the choice of break points (15 days, 30 days, etc.)?
The Full Story, Please
In my opinion, NCPA and the researchers will ultimately damage their credibility by peddling these partial results, which do nothing more than reinforce NCPA's longstanding preconceptions about PBMs. In the meantime, I call on the researchers and the NCPA to immediately release the raw data underlying the Executive Summary so that it can be subject to independent analysis and scrutiny.
I’d love to hear from you about this post. Email me or post a comment (anonymously, if you choose).