[Ip-health] NH Court constituionalizes phrma datamining

[Sean Flynn]

This is a multi-part message in MIME format.
--
[ Picked text/plain from multipart/alternative ]
Statement in response to decision in IMS v. Ayotte,

NH District Court, No. CV-06-280-PB (April 30, 2007)

Sean Flynn

Associate Director, Program on Information Justice and Intellectual Propert=
y

May 1, 2007



Yesterday, a New Hampshire District Judge stated that our constitution prev=
ents New Hampshire from implementing the same common sense data privacy law=
s that restrict the trade in prescription records as currently exist in all=
 of Europe and several Canadian provinces.  Those laws prohibit datamining =
companies from selling information to pharmaceutical companies that contain=
 doctors' names and other identifying information.  Instead, the companies =
can only compile and use information in aggregated "blocks" so as to protec=
t the identity of specific doctors and their prescribing habits.



Judge Barbadoro's opinion striking down New Hampshire's ban on the trading =
of prescriber-identified data is sweeping and disturbing in many respects. =
 The decision sweeps far beyond established First Amendment precedent and i=
s sure to be appealed by the State of New Hampshire.  The decision does con=
tain some reasoning that can be of guidance to states seeking policy soluti=
ons to the excesses of detailing and datamining in the pharmaceutical indus=
try.  On the whole, however, the opinion is unwelcome news to states and he=
alth care policy advocates.



The New Hampshire Act was passed against the background of evidence showing=
 that nearly a third of the five-fold increase in U.S. spending on drugs ov=
er the last decade can be attributed to the increased efficacy of pharmaceu=
tical marketing efforts that shift doctors' prescribing from existing, effe=
ctive, and lower cost (often generic) therapies to new and more expensive t=
reatments, which often have little or no increased therapeutic value.  One =
key change has been the recent ability of large datamining companies to pur=
chase computerized prescription records from insurers and pharmacies and th=
en sell that information to pharmaceutical companies to track every prescri=
ption a doctor writes.  Coincident with the rise of physician identity data=
 mining, the pharmaceutical industry increased its spending on direct marke=
ting to doctors by over 275 percent  and doubled its sales force to over 90=
,000 drug reps.   There is now a pharmaceutical sales representative for ev=
ery five office-based physicians in the U.S.   In 2004, the industry spent =
$27 billion on drug marketing (more than any other sector in the U.S. on it=
s sales force or media advertising),  over 85 percent of which was targeted=
 at doctors.



Judge Baradoro begins his reasoning with a rejection of New Hampshire's arg=
ument that the Prescription Information Law does not restrict speech becaus=
e it regulates "uses" (i.e. sale and trade in) of prescriber-identifiable i=
nformation, rather than First Amendment protected speech.  He specifically =
rejected New Hampshire's comparison of the law with the Act upheld by the S=
upreme Court in Bartnicki v. Vopper, 532 U.S. 514 (2001).



In Bartnicki, the Court struck down a section of the statute that constitut=
ed a "naked prohibition against disclosures" of information obtained throug=
h a wiretap, but approved of the section of the law that penalized any pers=
on who "uses . . . the contents of" a wiretapped communication.   The court=
 recognized cases holding that the use prohibition made it unlawful to "use=
 an illegally intercepted communication . . . to create a competing product=
," "in trading in securities," "to prepare strategy for contract negotiatio=
ns," or "to discipline a subordinate."   These prohibitions did not implica=
te the First Amendment, the Court explained, because "the prohibition again=
st the 'use' of the contents of an illegal intraception" is "a regulation o=
f conduct."



Like the Act in Barnicki, the New Hampshire act allows a broad range of dis=
closures of prescription information for non-commercial uses, but bans the =
trade of such information for marketing and other commercial purposes.  But=
 the judge rejected this distinction, holding that "The law is [  ] a speec=
h restriction because it limits both the use and disclosure of prescriber-i=
dentifiable data for commercial purposes."



The court further held that the Act would subject to First Amendment scruti=
ny even it was a restriction on conduct because it "restricts speech by pre=
venting pharmaceutical companies from using prescriber-identifiable informa=
tion both to identify a specific audience for their marketing efforts and t=
o refine their marketing messages.  Such laws are subject to First Amendmen=
t interests of their pharmaceutical company customers."



            Importantly, the court did hold that the act was a commercial r=
egulation subject to the commercial speech doctrine, rather than the strict=
 standard applied to political and other non-commercial speech.  It explain=
ed:



In understanding why this is so, it is important to bear in mind that the c=
hallenged law only restricts the transmission or use of prescriber-identifi=
able information for certain commercial purposes. It does not prevent anyon=
e from transmitting or using the information for law enforcement purposes, =
research purposes, educational purposes, compliance review purposes, or for=
 any non-commercial purpose. In short, the law is a commercial speech restr=
iction under Central Hudson because it restricts only speech that is "solel=
y in the individual interest of the speaker and its specific business audie=
nce," Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 762=
 (1985) (plurality opinion).



            Other courts, including the courts in U.S. West, Inc. v. Fed.

Commc'n Comm'n, 182 F.3d 1224, 1232 (10th Cir. 1999) and Trans Union Corp. =
v. Fed. Trade Comm'n, 245 F.3d 809, 818 (D.C. Cir. 2001) (applying intermed=
iate scrutiny to ban on sale of targeted marketing lists), have held that r=
egulations restricting use of customer information for marketing purposes r=
egulate speech protected by the First Amendment.  But the New Hampshire Dis=
trict Court went far beyond the holdings of these cases by striking down th=
e act as being based on an insufficient government interest in protecting d=
ata privacy rather than on the existence of other more narrowly tailored me=
ans to address the interest. Those previous cases explicitly affirmed opt i=
n or opt out programs that require a company to abide by the wishes of cons=
umers in trading their identifying data.  The New Hampshire court did not l=
eave this same policy avenue open in its opinion because it held that New H=
amshire lacks a sufficient interest to even legislate in this area.  This a=
spect of the opinion is extremely broad and is likely to be subject to a vi=
gorous appeal.



            The so-called commercial speech doctrine requires that truthful=
 commercial speech that does not promote unlawful activity can be limited o=
nly if it (1) is in support of a substantial government interest, (2) 'dire=
ctly advances the government interest asserted,' and (3) 'is not more exten=
sive than is necessary to serve that interest.



            The Court first explained that it found an insufficient record =
supporting the Act.  It stated:



There is nothing in the record, however, to support a conclusion that the l=
egislature had established expertise in the regulation of prescriber-identi=
fiable data. Moreover, it acted quickly after the bill was introduced, rece=
ived hearing testimony by numerous individuals who had yet to review propos=
ed amendments, made no express findings either on the record or incorporate=
d into the statute, failed to discuss alternative measures that

would not restrict speech, and cited no evidence as to how effective the re=
striction might prove to be.



            It then rejected the government's professed interest in protect=
ing the privacy interests of doctors in their records.  Distinguishing away=
 a host of laws that protect consumer's privacy in their identity to avoid =
unwanted and harassing sales calls, e.g. Health

Insurance Portability and Accountability Act of 1996, Pub. L. No. 104-191, =
110 Stat. 1936 (codified in scattered sections of 18 U.S.C., 26 U.S.C., 29 =
U.S.C., and 42 U.S.C.) (patient medical information); Fair Credit Reporting=
 Act, 15 U.S.C. =A7 1681 et seq.(2000) (credit reporting information); Fami=
ly Educational Rights and Privacy Act of 1974, 20 U.S.C. =A7 1232g (2000 & =
Supp. III 2003) (educational information); Video Privacy Protection Act

of 1988, 18 U.S.C. 2710 (2000) (video rental information); Cable Communicat=
ions Policy Act of 1984, Pub. L. No. 98-549, 98 Stat. 2779 (subscriber info=
rmation), the court held that doctors do not have a similar valid interest =
in avoiding the harassing sales calls of detailers that are augmented by th=
e trade of their prescribing patterns to pharmaceutical companies.  The cou=
rt specifically mentioned the lack of evidence offered at trial proving tha=
t the information is "pressed with such frequency or vehemence as to intimi=
date, vex, or harass the recipient," Edenfield, 507 U.S. at 769.



            This finding by the court is particularly disturbing and far re=
aching.  In other instances noted above, legislatures and courts have found=
 valid interests of consumers in having their identity and purchasing habit=
s kept private without their consent to avoid harassing sales tactics.  Yet=
 with medicines, the doctor is the one making the consuming choices - the d=
octor literally "prescribes" the medicine that the ultimate consumer purcha=
ses.  Thus it is the doctor who is the target of over 80% of the nearly $30=
 billion spent on pharmaceutical marketing every year.  This appears to be =
an area where a fuller record by state legislatures could help build a bett=
er factual background for subsequent legislation.



            The judge further found that the law does not directly control =
health costs and promote public health because, despite evidence showing th=
at billions of dollars every year are spend on needless drug shifting towar=
d more expensive brand name drugs driven by marketing, not all marketing dr=
iven shifts are bad and the "ban on the use of prescriber-identifiable data=
 affects both helpful and harmful brand-name prescribing practices in the s=
ame way." The court held:



Accordingly, the State simply does not have a substantial interest in shiel=
ding them from sales techniques that enhance the effectiveness of truthful =
and non-misleading marketing information. Instead, if the State is concerne=
d that truthful detailing is causing health care providers to make inadvisa=
ble prescribing decisions, "the remedy to be applied is more speech, not en=
forced silence." Whitney v. California, 274 U.S. 357, 377 (1927) (Brandeis,=
 J. concurring).



            Finally, the court found that there are other effective measure=
s to regulate the ill effects of detailing.  Specifically, the court stated=
 that



if legislators are concerned that pharmaceutical companies are improperly u=
sing samples, gifts, meals, and other inducements to promote inadvisable pr=
escribing practices, they can address this perceived problem by following o=
ther states that have adopted laws that limit such practices. See, e.g., Mi=
nn. Stat. Ann. =A7 151.461 (2007); Cal. Health and Safety Code =A7 119402(d=
)(1) (2007). Second, if legislators fear that pharmaceutical detailing is s=
imply too effective to go unrebutted, they can require the State to enter t=
he intellectual marketplace in several different ways with competing inform=
ation that will help health care providers balance and place in context the=
 sales messages that detailers deliver. . . . see, e.g., W. Va. Code Ann. =
=A7 5-16C-9(5) (2006) (authorizing state to develop counter-detailing progr=
ams); or they can require health care providers to regularly participate in=
 continuing medical education programs that are specifically designed to pr=
ovide practitioners with the best available information concerning the adva=
ntages and disadvantages of prescribing generic drugs rather than brand-nam=
e drugs. Finally, if legislators are concerned that pharmaceutical companie=
s are using prescriber-identifiable data to drive up Medicaid drug costs, t=
hey can address the issue directly by properly implementing a Medicaid Phar=
macy Program (preferred drug list) that takes into account the cost-effecti=
veness of brand-name drugs when compared with non-bioequivalent generic alt=
ernatives.



            At bottom, the opinion suggests several steps that states inter=
ested in regulating the ill effects of detailing might pursue:



*           states interested in regulating the trade in prescriber-identif=
ied prescription information need to create a fuller record explaining how =
frequently detailing based on datamining has become harassing to doctors in=
 ways that prohibiting datamining without their consent will halt;

*           states may consider more narrowly tailored mans to stop data tr=
ading, for example permitting the information to be released to pharmaceuti=
cal companies if the doctor has explicitly consented to the release;

*           states should include specific findings on the adequacy of alte=
rnative measures, including licensing detailers, prohibiting false and misl=
eading detailing, use of Medicaid formularies, bans on gifts, and counter-d=
etailing programs;

*           states may consider bolstering an expectation of privacy in pre=
scription records by including statutory findings and inviting testimony th=
at doctors do not and should not expect that their prescriptions will be us=
ed for purposes other than to fill and process prescriptions;

*           target the most harmful and biased uses of datamining, e.g. by =
allowing their use for evidence-based marketing, but not use for marketing =
information that isn't supported by independent studies & sources (this may=
 make the statute closer to the Barnicki "use" regulation);

*           states may consider alternative regulations, such as licensing =
and regulating detailers to cut down on the most abusive practices includin=
g gifts, pushing off-label uses, making misleading statements, etc.



Ultimately, the New Hampshire decision only binds New Hampshire and is like=
ly to be headed for appeal.  Other states may legislate in this area and cr=
eate conflicting precedents that will have to be unified by the First Circu=
it and ultimately the Supreme Court.