Data Mining: Use Caution

With 78 million Baby Boomers retiring at the rate of 10,000 per day, and with the life expectancy of Boomers continuing to increase, the demand for pharmacy services will continue to increase exponentially. As they meet this demand, pharmacies are nevertheless facing a “Perfect Storm” of challenges. These include (i) lower reimbursement; (ii) termination by PBMs of pharmacy contracts because pharmacies are engaged in compounding and/or mail-order; and (iii) aggressive audits. To counter these challenges, pharmacies are having to be innovative in how they market to customers, work with physicians and other referral sources, and deal with third party payors. An example of an innovative approach is “data mining.” While data mining is not wrong in and of itself, pharmacies need to be aware of the potential pitfalls attendant to certain data mining activities.

In one type of data mining arrangement, a company (“ABC”) assists the pharmacy in researching alternative drug options that result in much larger reimbursement. The pharmacy then approaches physicians and suggests that they switch their prescriptions from the drug with lower reimbursement to the drug with higher reimbursement. The pharmacy will educate the physicians regarding the clinical benefits of the more expensive drug. If the physicians agree and change prescriptions, then the pharmacy makes significantly more money, but the physicians do not financially benefit from the arrangement.

At a pharmacy conference I spoke at, a pharmacist approached me and told me that he was involved in a data mining program such as the one described above. He said that he was netting a substantial amount of money per script for the replacement drug. The pharmacist admitted that while his discussions with physicians focused on the enhanced clinical efficacy of the new drug, in reality the motivation behind the pharmacist approaching physicians is for the pharmacy to make more money. The pharmacist told me that this data mining program made him feel uneasy. He also stated that PBMs are beginning to shut this arrangement down by taking the higher reimbursed replacement drugs off the PBMs’ formulary.

Percentage Compensation

With some data mining arrangements, the pharmacy pays ABC a percentage of the net revenue generated by the data mining program. If some of the patients are covered by a federal government health care program (“FHCP Patients”), then this type of percentage arrangement may violate the federal anti-kickback statute (“Federal AKS”) which states that the pharmacy cannot pay anything to ABC for (i) referring an FHCP Patient to the pharmacy, (ii) arranging for the referral of an FHCP Patient to the pharmacy, or (iii) recommending the purchase of a drug that a federal government program pays for. In addition to the Federal AKS, each state has its own anti-kickback statute (“State AKS”). Some State AKSs apply only if the payor is the state’s Medicaid program. Other State AKSs apply even if the payor is a commercial insurer or a cash-paying customer. Each state also has a set of laws that are specific to pharmacies. Some of the pharmacy-specific state laws prohibit kickbacks, fee splitting, and similar arrangements.

Let us assume that a federal government health care program (Medicare Part D, TRICARE, Medicaid, etc.) pays for the replacement drug. If the pharmacy is paying percentage compensation to ABC, the question becomes: “Is ABC arranging for the referral of FHCP Patients to the pharmacy and/or is ABC recommending the purchase of drugs that are reimbursable by a government health care program?” I can argue both sides of the equation. On the one hand, I can argue that because ABC is not having any contact with the physicians (i.e., ABC is only working with the pharmacy), then ABC cannot be construed to be “arranging for the referral” of patients nor “recommending the purchase of drugs.” I can also argue the opposite. I can argue that by allowing the pharmacy to use ABC's software platform and by showing the pharmacy how to find similar drugs with higher reimbursement, then such acts rise to the level of “arranging for the referral” of patients and “recommending the purchase of drugs.” This is where the “smell test” comes in. Governmental agencies have a great deal of discretion in deciding whether or not to bring an enforcement action. Likewise, third party payors (such as PBMs) have a great deal of contractual discretion to (i) terminate a pharmacy from the network and/or (ii) audit the pharmacy’s claim submissions to determine if a recoupment action is appropriate. If an arrangement falls within a “gray area,” but it is not otherwise abusive or offensive, then the governmental agency or third party payor will likely leave the arrangement alone. On the other hand, if it looks like the parties to the arrangement are “gaming the system” to substantially increase their revenue, then the governmental agency (and/or a third party payor) will likely be motivated to shut the arrangement down.

Now let us switch gears and assume that no government program is involved. Assume that the only payors are commercial insurers. If the pharmacy is operating in a state in which there is (i) a state anti-kickback statute that applies to all payors and/or (ii) there are pharmacy-specific laws addressing kickbacks/fee splitting, then depending on the wording of the statute, paying percentage compensation to ABC may constitute a kickback.

Fixed Annual Compensation

Now let us change the facts and assume that the compensation paid by the pharmacy to ABC is fixed one year in advance and is the fair market value equivalent of ABC’s services. Fixed annual (fair market value) compensation is an important element of the Personal Services and Management Contracts safe harbor (“PSMC Safe Harbor”) to the Federal AKS. Because of the breadth of the Federal AKS, the Office of Inspector General (“OIG”) has issued a number of “safe harbors” such as the PSMC Safe Harbor. If an arrangement meets all of the elements of a safe harbor, then as a matter of law the arrangement does not violate the Federal AKS. If an arrangement does not comply with a safe harbor, it does not mean that the arrangement violates the Federal AKS. Rather, it means that the arrangement will have to be closely scrutinized in light of the language of the Federal AKS, court decisions, and published OIG guidance.

It will likely be difficult for a fixed annual fee data mining arrangement to comply with all of the elements of the PSMC Safe Harbor. However, if the arrangement substantially complies with the PSMC Safe Harbor, then the risk is low that (i) a federal government enforcement agency will assert a violation of the Federal AKS and (ii) a state government enforcement agency will assert a violation of a State AKS. Nevertheless, PBMs will likely take steps to neutralize the arrangement by (i) removing the replacement drugs with higher reimbursement from the formulary or (ii) reducing the reimbursement for the replacement drugs. Additionally, most (if not all) PBM contracts give the PBM the right to terminate the pharmacy from the contract “without cause.” A PBM may take the position that by engaging in the data mining arrangement, the pharmacy is a “bad player” and terminate it from the contract. Also, there is a possibility that the contract between the pharmacy and the PBM contains restrictions that prohibit the data mining arrangement.

There is a risk that PBMs will take the position that these types of data mining arrangements are similar to the past practice by compounding pharmacies of submitting huge claims to TRICARE and PBMs for compounded pain creams. The pharmacies did this because TRICARE and the PBMs paid the claims the pharmacies submitted. TRICARE and the PBMs brought this “pain cream phenomenon” to a screeching halt.

Conclusion

As stated at the beginning of the article, a data mining arrangement is not, in and of itself, wrong. However, if a governmental agency and/or a PBM nevertheless concludes that the arrangement is “gaming the system,” then the agency/PBM will likely be motivated to bring the arrangement to an end. If a pharmacy engages in a data mining arrangement that will result in significantly increased reimbursement (and significantly increased costs to the payors), it needs to have its “eyes wide open” and be aware of the possible risks.

______________________________________________________________________________

Jeffrey S. Baird, Esq. is Chairman of the Health Care Group at Brown & Fortunato, P.C., a law firm based in Amarillo, Texas. He represents pharmacies, home medical equipment companies, and other health care providers throughout the United States. Mr. Baird is Board Certified in Health Law by the Texas Board of Legal Specialization. He can be reached at (806) 345-6320 or jbaird@bf-law.com.