Characteristics and Impact of Drug Detailing for Gabapentin
Sales visits by pharmaceutical representatives ("drug detailing") are common, but little is known about the content of these visits or about the impact of visit characteristics on prescribing behavior. In this study, we evaluated the content and impact of detail visits for gabapentin by analyzing market research forms completed by physicians after receiving a detail visit for this drug. Market research forms that describe detail visits for gabapentin became available through litigation that alleged that gabapentin was promoted for "off-label" uses. Forms were available for 97 physicians reporting on 116 detail visits between 1995 and 1999. Three-quarters of recorded visits (91/116) occurred in 1996. Two-thirds of visits (72/107) were 5 minutes or less in duration, 65% (73/113) were rated of high informational value, and 39% (42/107) were accompanied by the delivery or promise of samples. During the period of this study, gabapentin was approved by the US Food and Drug Administration only for the adjunctive treatment of partial seizures, but in 38% of visits (44/115) the "main message" of the visit involved at least one off-label use. After receiving the detail visit, 46% (50/108) of physicians reported the intention to increase their prescribing or recommending of gabapentin in the future. In multivariable analysis, intent to increase future use or recommendation of gabapentin was associated with receiving the detail in a small group (versus one-on-one) setting and with low or absent baseline use of the drug, but not with other factors such as visit duration, discussion of "on-label" versus "off-label" content, and the perceived informational value of the presentation. Detail visits for gabapentin were of high perceived informational value and often involved messages about unapproved uses. Despite their short duration, detail visits were frequently followed by physician intentions to increase their future recommending or prescribing of the drug.
Full-text (PDF)Available from: Charles Seth Landefeld, Feb 20, 2015
Other full-text sources
Characteristics and Impact of Drug Detailing
Michael A. Steinman
, G. Michael Harper
, Mary-Margaret Chren
, C. Seth Landefeld
, Lisa A. Bero
1 San Francisco Veterans Affairs Medical Center, San Francisco, California, United States of America, 2 Division of Geriatrics, University of California, San Francisco, California,
United States of America, 3 Department of Dermatology, University of California, San Francisco, California, United States of America, 4 Department of Clinical Pharmacy,
University of California, San Francisco, California, United States of America, 5 Institute for Health Policy Studies, University of California, San Francisco, California, United States
Funding: This work was supported
by a Veterans Affairs Health Services
Research & Development Research
Career Development Award (MAS);
by grants from the National Institute
on Arthritis and Musculoskeletal and
Skin Disease (K24 AR052667) (MMC),
the National Institute on Aging
(AG00912) and the John A. Hartford
Foundation (2003–0244) (CSL), and
the California Tobacco-Related
Disease Research Programs (13RT-
0108) (LAB); and by the Health
Services Research Enhancement
Award Program (REAP) at the San
Francisco Veterans Affairs Medical
Center. The fund ers had no role in
study design, data collection and
analysis, decision to publish, or
preparation of the manuscript.
Competing Interests: See section at
end of manuscript.
Academic Editor: Steven E. Hyman,
Harvard University, United States of
Citation: Steinman MA, Harper GM,
Chren MM, Landefeld CS, Bero LA
(2007) Characteristics and impact of
drug detailing for gabapentin. PLoS
Med 4(4): e134 doi:10.1371/journal.
Received: May 9, 2006
Accepted: January 19, 2007
Published: April 24, 2007
This is an open-access article
distributed under the terms of the
Creative Commons Public Domain
declaration which stipulates that,
once placed in the public domai n,
this work may be freely reproduced,
distributed, transmitted, modified,
built upon, or otherwise used by
anyone for any lawful purpose.
* To whom correspondence should
be addressed. E-mail: mike.
Sales visits by pharmaceutical representatives (‘‘ drug detailing’’ ) are common, but little is
known about the content of these visits or about the impact of visit characteristics on
prescribing behavior. In this study, we evaluated the content and impact of detail visits for
gabapentin by analyzing market research forms completed by physicians after receiving a
detail visit for this drug.
Methods and Findings
Market research forms that describe detail visits for gabapentin became available through
litigation that alleged that gabapentin was promoted for ‘‘ off-label’’ uses. Forms were available
for 97 physicians reporting on 116 detail visits between 1995 and 1999. Three-quarters of
recorded visits (91/116) occurred in 1996. Two-thirds of visits (72/107) were 5 minutes or less in
duration, 65% (73/113) were rated of high informational value, and 39% (42/107) were
accompanied by the delivery or promise of samples. During the period of this study,
gabapentin was approved by the US Food and Drug Administration only for the adjunctive
treatment of partial seizures, but in 38% of visits (44/115) the ‘‘ main message’’ of the visit
involved at least one off-label use. After receiving the detail visit, 46% (50/108) of physicians
reported the intention to increase their prescribing or recommending of gabapentin in the
future. In multivariable analysis, intent to increase future use or recommendation of gabapentin
was associated with receiving the detail in a small group (versus one-on-one) setting and with
low or absent baseline use of the drug, but not with other factors such as visit duration,
discussion of ‘‘ on-label’’ versus ‘‘ off-label’’ content, and the perceived informational value of
Detail visits for gabapentin were of high perceived informational value and often involved
messages about unapproved uses. Despite their short duration, detail visits were frequently
followed by physician intentions to increase their future recommending or prescribing of the
The Editors’ Summary of this article follows the references.
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340743
Visits to physician ofﬁces by pharmaceutical sales repre-
sentatives are among the most visible and effective forms of
drug industry promotion . As such, their role has generated
substantial debate in the medical community [2–4]. Some
commentators have argued that these ‘‘ details’’ provide
unbalanced information and thus negatively impact prescrib-
ing quality, a problem compounded by the inﬂuence of gifts
that commonly accompany these interactio ns [5–10]. In
addition, negative public perceptions of these relationships
may harm the standing of the medical profession. Others
have argued that sales representatives provide important
information about drugs to physicians, and thus may improve
prescribing quality [11–13].
While a number of studies have investigated the frequency
of detail contacts and the association between these contacts
and prescribing behavior [1,14–21], relatively little has been
published in the medical literature on what actually happens
during detail visits [22,23]. Studies using physician self-
reports suggest a range of attitudes about the educational
content of these interactions and the perceived effect of
detail visits on prescribing [1,12,15,24–28]. However, these
studies may be subject to recall bias, due to reliance on self-
reports long after the conclusion of detail visits and a focus
on respondents’ general impressions rather than on speciﬁc
interactions. In addition, nearly all data on this topic
published in the medical literature have originated from
academic settings. These settings may subtly bias survey
responses, since physicians may believe industry contacts are
judged negatively in the academic community and thus may
tailor their answers to be more socially acceptable.
To learn more about the content and potential impact of
detail visits, we used data collected from physicians by a
market research company about detail visits for gabapentin
(trade name Neurontin). These data were obtained from
documents subpoenaed in a qui tam whistleblower lawsuit
alleging that drug manufacturer Parke-Davis violated federal
regulations in the mid- to late-1990s by promoting gabapen-
tin for uses not approved by the US Food and Drug
Administration . Over this period, the only approved
use of gabapentin was for the adjunctive treatment of partial
seizures in persons over 12 years of age at doses up to 1,800
milligrams per day. However, interest in the drug for the
management of pain syndromes, psychiatric diseases, and
other conditions presented an opportunity for a much wider
market, and Parke-Davis use d a variety of coordinated
strategies to promote gabapentin for both on- and off-label
uses . In 2004, this litigation concluded with a US$430
million out-of-court settlement by Warner-Lambert, Parke-
Davis’ parent corporation .
Using data from documents subpoenaed in this litigation,
we describe the characteristics and content of detail visits for
gabapentin and the self-reported impact of these details on
future prescribing of and attitudes toward the drug. Next, we
examine which aspects of details were associated with
physicians’ intention to increase their future prescribing or
recommendation of gabapentin.
Data Source and Inclusion Crite ria
All data used in this study were subpoenaed from Verispan,
a health care information company whose specialties include
market research for pharmaceutical companies, for use in
United States of America ex rel. David Franklin v. Pﬁzer, Inc,
and Parke-Davis, Division of Warner-Lambert Company .
Data on drug detail visits were collected by Verispan, which
recruited physicians in certain specialties throughout the
United States to serve as ‘‘ panelists’’ . At the conclusion of
each detail visit by a pharmaceutical sales representative,
panelists were asked to complete a brief standardized form
(known as a ‘‘ verbatim’’ or ‘‘detailing reporting form’’ ) that
described characteristics of the visit. Physicians were com-
pensated by Verispan for their participation in this program,
but had no other contact with the company, and were given
guarantees of conﬁdentiality . Data collected in this way
are typically purchased by pharmaceutical ﬁrms for market
research purposes .
Forms produced by Verispan in response to subpoena
(which requested all data pertaining to gabapentin from 1994
through 2002) were obtained from the US District Court in
Massachusetts and from the law ﬁrm that represented the
whistleblower plaintiff, and form the basis for this report. In
this study, we sought to focus on the ‘‘ classic’’ type of
pharmaceutical detail, i.e., a visit from a sales representative
to a single physician or a small group of physicians. In the
absence of a commonly accepted deﬁnition, we deﬁned a
classic detail as a visit by an industry sales representative to
an individual physician or group of up to three physicians in
the workplace setting. Among 142 forms, we excluded 26 that
appeared to describe other types of encounters, including 16
visits in which more than three physicians participated, four
visits in which more than one company was involved, three
visits that occurred in a nontraditional location or format
(‘‘inservice/[continuing education] program’’ ), and three visits
whose main message suggested that the encounter occurred
via teleconference. The remaining 116 forms composed our
analytic sample, and included three variations of the same
data collection instrument.
The 116 forms were contributed by 97 physicians, includ-
ing 87 physicians who contributed one form, six physicians
who contributed two forms each, and four physicians who
completed th ree to seven forms each (with each form
representing a separate visit). As our primary interest was
in the detail visit itself rather than the individual physician,
our primary analyses evaluated all 116 forms. Among the 116
forms, 91 (78%) were from 1996; eight (7%) were from other
years (1995 and 1997–1999), and the remaining 17 (15%)
forms did not provide complete date information. It is
unclear why the majority of forms provided were from a
single year, although the absence of other years may be
related to company practice of destroying forms more than
several years old . A spreadsheet provided in response to
the subpoena shows that an additional several hundred forms
were completed for encounters between 1995 and 1998, but
these data were not usable due to the limited amount of
information provided and the uncertain context of these
Information provided by the physician on verbatim forms
(Figure 1) included the date and location of the visit, whether
it was a one-on-one or group vis it, the main mess age
associated with each product (entered as free text), and the
informational value of the detail for each product as well as
the overall quality of pre sentation for the entire visit
(recorded on a ﬁve-point Likert scale with anchors of ‘‘ poor’’
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340744
Drug Detailing for Gabapentin
and ‘‘ excellent’’ ). In general, physicians did not provide any
other comments in the free text space. Physicians were also
asked whether they prescribed or recommended the product
and at what level (‘‘ low,’’ ‘‘ medium,’’ or ‘‘ high’’ ), with separate
responses for current and expected future practice. (On six
forms, similar information was collected using a different set
of questions). Physicians were asked to send in their reports
monthly, and an estimated 90% of reports were ﬁled within
one to two months after the visit .
Using free text entered in the ‘‘Main Message’’ ﬁeld of the
market research form (Figure 1), two of the authors (MAS,
GMH) independently classiﬁed the main message for each
detail visit into one or more clinical indications (see Box 1).
Disagreements were resolved by consensus, with additional
adjudication by the senior author (LAB) when necessary. For
comparative analyses, visits were categorized into three
mutually exclusive categories based on the type of messages
communicated. These included (1) messages that mentioned
the approved indication; (2) messages that mentioned uses
not approved by the FDA; and (3) messages that mentioned a
mix of approved and unapproved uses.
While physicians are allowed to prescribe medications for
any purpose they see ﬁt, drug manufacturers are allowed to
promote drug use only for indications that have been FDA
approved . Thus, we considered any mention of seizure or
epilepsy to be an approved message, except messages which
speciﬁcally mentioned gabapentin as monotherapy for
epilepsy, which was unapproved. Messages containing other
content (e.g., comments about pharmacokinetics, drug–drug
interactions, or nonspeciﬁc comments) could also be present
in any of the three categories. However, if the only message(s)
present was this other content, we classiﬁed the visit message
as approved. Of note, in 2002 gabapentin received FDA
approval as treatment for post-herpetic neuralgia, but
because the study period was well before this, we considered
this use to be ‘‘ off-label’’ in our analyses.
To evaluate which characteristics of the detail visit were
associated with the intent to increas e prescribing or
recommendation of gabapentin in the future, we constructed
a change measure for each physician. Physicians who
intended to change from nonusers to users or who intended
to increase their current level of use were classiﬁed as
intending to increase their future use. Corresponding logic
was used to identify physicians who intended to maintain or
decrease their future use. As no physicians reported an
intended decrease in their future use of gabapentin, the
change measure was dichotomized into increased versus
stable use. Seventeen physicians reported ‘‘ high’’ levels of
gabapentin prescribing or recommending at baseline and
were excluded from some of our analyses because there was
no reliable way to detect a future increase (i.e., a ‘‘ ceiling
effect’’ ). All 17 of these physicians reported the intention to
continue prescribing or recommending the drug at ‘‘ high’’
levels in the future.
We conducted bivariate analyses using Chi-square and
Fisher exact tests to evaluate characteristics associated with
an intention to increase future prescribing or recommending
of gabapentin. We then constructed a multivariable model
using backward stepwise selection of these variables, with p ,
0.20 to stay in the model. After stepwise selection, we used the
covariance matrix of the parameter estimates to identify and
remove highly collinear variables. To preserve power in this
analysis, missing data were coded as a separate category for
each variable. These categories of missing data are not
Figure 1. Detailing Reporting Form
Excerpt from a detailing reporting form providing information on a detail visit for gabapentin. Some markings (e.g., ‘‘ 032417 Neuro’’ and ‘‘ 12233’’ next
to ‘‘ Product: Neurontin’’ ) were annotations entered by Verispan employees to assist with coding and data entry of completed forms.
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340745
Drug Detailing for Gabapentin
reported. On sensitivity analysis of the ﬁnal model, results
were not substantively different when these ‘‘missing’’
categories were reclassiﬁed as each of the possible levels of
the corresponding variable, or recoded as missing data with
loss of the observation.
We conducted two additional sensitivity analyses. First, to
restrict our sample more speciﬁcally to one-on-one encoun-
ters, we repeated our analyses after excluding 11 visits in
which two or three physicians participated, and 25 visits in
which data were not provided on the individual versus group
nature of the encounter. Second, we repeated our analyses
after excluding multiple visits to the same physician (each
recorded on a separate form), keeping only the form with the
earliest date for a given physician.
Analyses were conducted using Stata 8.2 (http://www.stata.
com). This research was approved by Research and Develop-
ment Committee of the San Francisco Veterans Affairs
Medical Center and the Committee on Human Research of
the University of California, San Francisco.
Characteristics of Details and Events
Characteristics of the 116 visits in our analytic sample are
shown in Table 1. Over half of detail visits (56%) were to
Table 1. Characteristics of 116 Detail Visits
Category Characteristic Response n (%)
Characteristics of physician
and detai l visit
Physician specialty (n ¼ 112) Neurologist
First detailing on this product (n ¼ 90) Yes
Individual or small group setting? (n ¼ 91) Individual
Small group (2–3 physicians)
Time spent on detail (minutes) (n ¼ 107) 5min
Visit content Main messages of detail
(n ¼ 115): Approved use
Epilepsy—Adjunctive or not specified 53 (46%)
Main messages of detail
(n ¼ 115): Unapproved uses
Pain or neuropathy
Other specific unapproved uses
General mention of unapproved uses
Doses .1,800 mg/day
Main messages of detail
(n ¼ 115): Other messages
Was product compared to other products?
(n ¼ 96)
Support materials used (n ¼ 111) Yes
Value of support materials (if used) (n ¼ 77)
1–2 (wors t)
Informational value of detail (n ¼ 113) 1–2 (worst)
Overall quality of presentation (n ¼ 97) 1–2 (worst)
Product samples (n ¼ 107) Delivered or to be mailed
Prescribing or recommendation
Currently prescribe or recommend (n ¼ 115) Yes
Current level of prescribing or recommendation
(n ¼ 78)
Intended future change in level
of prescribing or recommendation (n ¼ 108)
Data are from 116 detail forms. Number of non-missing responses for each question is represented by n-values.
Among respondents who had responded ‘‘ yes’’ to the corresponding question (and excluding an additional six forms that did not ask these questions).
Intended future change included transition from nonusers to users, and transition between low, medium, and high levels of use.
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340746
Drug Detailing for Gabapentin
physicians in ﬁelds other than neurology, the majority of
which (46/63) comprised visits to doctors of internal
medicine, family or general practice, or osteopathic medi-
cine. The median visit duration was 5 min. Among a subset of
33 forms in which physicians were asked about formulary
status for gabapentin, 30 physicians (91%) responded that the
drug was on their hospital’s formulary and the remaining
three (9%) did not know.
Physicians recorded in their own words a number of
distinct ‘‘ main messages’’ associated with the detail visit for
gabapentin. Unapproved messages were present in 44 (38%)
of visits, including 26 visits (23%) whose ‘‘ main message ’’
cited only unapproved uses, and 18 visits (16%) that
mentioned both approved and unapproved uses. Neurologists
and non-neurologists reported a similar proportion of
approved messages, unapproved messages, and mess ages
containing both approved and unapproved content (Chi-
square statistic 2.02, p ¼ 0.36 for difference in message type
Physicians rated the informational value and over all
quality of the detail visits highly (Table 1). There was a
nonsigniﬁcantly lower perceived informational value for
detail visits that focused on unapproved messages. On a
ﬁve-point Likert scale (with 5 being best), physicians rated
the informational value as 4 or 5 in 46% (12/26) of details
with only unapproved messages, 71% (12/17) of details with a
mixture of approved and unapproved messages, and 70% (49/
70) of details with approved messages (Chi-square statistic
5.03, p ¼ 0.08 for difference in value rating between message
Factors Associated with Intention to Increase Future
Prescriptions or Recommendations for Gabapentin
Overall, 46% (50/108) of physicians stated that their
prescribing or recommending of gabapentin would increase
in the future. No physicians reported the intention to
decrease their future use or recommendation of the drug.
On bi variate analyses, physicia ns’ intention to increase
prescribing or recommending of gabapentin was higher
among non-neurologists, physicians with a lower current
level of prescribing, and for visits that occurred in small
group (versus one-on-one) settings (Table 2). There was also a
nonsigniﬁcantly higher intention to prescribe among physi-
cians who received visits that lasted longer than 5 min (p ¼
0.08) and among physicians who perceived the visit to be of
high informational value (p ¼ 0.07). On multivariable analysis,
receiving the detail visit in a small group setting and low
baseline use or recommendation of gabapentin were the only
variables associated with the intention to increase future
activity with the drug (Table 2).
When the sample was restricted to the 80 physicians with
known one-on-one details, there were no substantive differ-
ences in our results. Similarly, when we restricted the sample
to include only one form per physician (n ¼ 97), our analyses
were not substantively changed.
In this study of drug detail visits for gabapentin, we found
that most detail visits were brief and of high perceived
quality, and often discussed unapproved uses of this drug.
Moreover, after receiving the detail visit, half of participants
reported the intention to increase their future prescribing or
recommendation of gabapentin.
The high perceived quality of these presentations is
consistent with the ﬁndings of many (but not all) previous
studies of physician attitudes toward pharmaceutical repre-
sentatives [1,25,33–40]. However, the perceived quality of
these presentations may not reﬂect the validity of their
educational content. For example, some of the off-label uses
for which gabapentin was promoted are not well supported
by high-quality clinical trial evidence . Other uses such as
neuropathic pain and migraine prophylaxis have been found
to be effective in controlled clinical trials, yet these trials had
not been published during the period that we studied (aside
from two trials published in the ﬁnal month of 1998, near the
end of our study period) [42–44]. More generally, studies
suggest that incorrect or misleading information about drugs
may frequently be conveyed in promotional settings [5–
7,10,23,45,46], that physicians do not consistently distinguish
between correct and incorrect information [5,7,47,48], and
that it is the perceived (rather th an actual) qualit y of
information that produces behavior change [49,50]. In our
study, the association between perceived informational value
of details and self-reported behavior change was in the
expected direction, but did not meet statistical signiﬁcance.
In addition to the credibility of the information provided
at detail visits, the int erpersonal aspect of interactions
between pharmaceutical representatives and physicians has
been hypothesized to be a major driver of behavioral change,
and has been emulated with good success by ‘‘ academic
detailing’’ [1,8–10,22,51–54]. Unfortunately, the forms used
for this analysis provide little information about this inter-
personal content and its relationship to future behavioral
change. Among the visit characteristics that we were able to
measure, physicians who received details in small group
settings were more likely to increase their future use or
recommendation of gabapentin than physicians receiving
one-on-one details. This may reﬂect a ‘‘ group-think’’ men-
tality whereby the perception that one’s colleagues are
receptive to the detail visit increases one’s own receptivity
The variables that were not associated with future behavior
change may be as interesting as those that were. In particular,
physicians reported similar increases in future prescribing or
recommending of gabapentin after exposure to approved or
unapproved messages, suggesting potential susceptibility to
discussion of new and relatively unproven uses. Also, visit
duration was not associated with the frequency of future
behavior change on bivariate analyses. While the failure to
detect an association may be due to our limited sample size,
even so 50% of these brief visits were associated with reports
of increased future use or recommendation of gabapentin.
This suggests that even brief encounters may have substantial
impact. This may reﬂect the success of sophisticated market-
ing techniques that enable pharmaceutical representatives to
track the prescribing patterns of individual physicians and to
succinctly deliver tailored marketing messages, employ
inﬂuence techniques, and regulate the content of the visit
all the way down to the ﬁrmness and duration of the
introductory handshake [10,22,58,59].
While we are unaware of other studies published in the
medical literature that use methods similar to ours, other
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340747
Drug Detailing for Gabapentin
reports have found a range of real or perceived impacts of
pharmaceutical detailing on prescribing . Several studies
have found positive associations between the frequency of
interactions with sales representatives and prescrib ing
behavior, and in physician self-report studies approximately
one-third to two-thirds of respondents perceive themselves to
be inﬂuenced by pharmaceutical sales representatives,
although more believe that ‘‘ other physicians’’ are inﬂuenced
[15,16,60–62]. Studies using objective data on prescriptions
have found a consistently positive impact of detailing on
prescribing, often with a greater impact than other forms of
marketing such as direct-to-consumer advertisements [1,20].
For example, major effects were observed in a small study of a
resident-run psychiatr y clinic [19 ]. Other stu dies have
observed more modest effects, with substantial variation in
the effectiveness of detailing for different drugs . The
strong effects observed in our study may in part reﬂect the
market position of gabapentin, which at the time of the study
was relatively new and of substantial interest for a growing
number of conditions, a situation in which the effectiveness
of detailing may be greatest .
Certain characteristics of detailing for gabapentin may not
be universally applicable to detailing for other drugs. Most
notably, the promotional campaign for gabapentin involved
multifaceted efforts to encourage prescribing for uses not
approved by the FDA . As a result, detailing for this drug
may have involved an atypically high number of unapproved
messages (although a study fr om Franc e suggests tha t
discussion of unapproved uses may be common for other
drugs as well) . In addition, excitement in the medical
community about novel uses of gabapentin may have made
physicians particularly receptive to these messages.
Other visit characteristics, such as the duration and
perceived informational value of the visit, may be more
applicable to other drugs. However, our data do not allow us
to test this hypothesis, and our results should thus be
interpreted as a case study of detailing for one drug.
Nonetheless, the focused nature of our data also represents
a unique strength, as attention to a single product and
individual visits helps avoid the generalizations and recall
biases that may affect most existing surveys of pharmaceutical
Our study has several limitations. First, the perceived
acceptability of providing certain answers to the market
research ﬁrm could have affected physician responses to this
self-reported survey or affected their conduct in interactions
with pharmaceutical sales representatives. Second, we have
limited information about the content of detailing inter-
actions other than a brief ‘‘ main message’’ recorded by the
physician. Discussion of unapproved uses may have been
initiated by the physician. Third, the self-reported intention
to increase future prescribing or recommending of gabapen-
tin might have been affected by factors other than the detail.
Thus, we cannot prove a causal relationship between the
detail and self-reported behavior change, and we do not know
whether future intentions became future actions. Fourth, in
our multivariable model the presence of prior visits and the
distribution of samples were both collinear with the current
level of prescribing, which impacted our model construction.
Thus, we cannot reliably disentangle the effects of each of
these factors on intentions to increase future prescribing. In
addition, the small sample size limits the precision of the
observed point estimates and our ability to detect all but the
strongest associations. Fifth, although we did our best to focus
Table 2. Factors Associated with Intention to Increase Future Prescribing or Recommending of Gabapentin
Characteristic Response Bivariate Analysis Multivariable Analysis
n/N (%) p-Value Adjusted Odds Ratio (95% CI)
Physician specialty Neurologist
First detailing on this product Yes
Individual or small group detail Individual
Small group (2–3 physicians)
0.03 Reference group
Time spent on detail 5min
Main message of detail Approved uses or other (only)
Mixed approved and unapproved uses
Unapproved uses only
Informational value of detail Score of 1–3
Score of 4–5 (better)
Overall quality of presentation Score of 1–3
Score of 4–5 (better)
Product samples Delivered or to be mailed
Current level of prescribing or recommending None
,0.001 15.1 (3.9–58.2)
8.6 (2.4–31 .4)
Data are from 91 detail visits with information on intended changes in future prescribing or recommendation. To avoid a potential ‘‘ ceiling effect,’’ this analysis excluded 17 visits to
physicians who reported ‘‘ high’’ rates of prescribing or recommending gabapentin at baseline (each of these physicians reported the intention to continue their ‘‘ high’’ rates into the
On multivariable analyses, first time detailing and provision of samples were strongly collinear with the current level of prescribing, and were dropped from the final model.
N/A, variables not significant in the multivariable model (p . 0.20).
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340748
Drug Detailing for Gabapentin
on the classic form of detailing, it is possible that our sample
inadvertently included contact reports from other types of
interactions. Finally, as described earlier it is unclear why
most of the forms provided in request to the subpoena were
from 1996, when the period covered under the subpoena was
While new forms of marketing such as direct-to-consumer
advertising have captured the attention of the medical
community, drug detailing remains a major form of
pharmac eutical promot ion, with expenditures of US$6.7
billion per year helping to pay tens of thousands of sales
representatives (estimated at one representative for every 4.7
ofﬁce-based physicians) [63,64]. We found that detail visits for
gabapentin often involved messages about unapproved uses
and were perceived as having high informational value.
Despite the brevity of these visits, half of the respondents
indicated their intention to increase their future prescribing
or recommendation for the drug. In light of the potential for
transmission of unbalanced information and the impact of
detail visits on prescribing, the most prudent course of action
may be for physicians to curtail or abstain from detail visits,
and instead turn to the variety of free and low-cost resources
that provide a more objective perspective on the merits and
pitfalls of new drugs (http://www.ohsu.edu/drugeffectiveness)
The authors thank Miriam Schwarz for her assistance with data
coding and entry. This work was presented as a poster at the annual
meeting of the Society of General Internal Medicine in April, 2006
(Los A ngeles). This work is not under review for publication
Author contributions. MAS developed the research question,
obtained and coded the data, conducted the analyses, and wrote
the paper. GMH, MMC, CSL, and LAB assisted in the development of
the analytic plan and interpretation of the findings, and critically
reviewed successive drafts of the manuscript. GMH and LAB also
assisted in coding of the data.
Competing Interests. MAS, MMC, and CSL served as unpaid expert
witnesses for the plaintiff in the litigation from which source data for
this report was obtained (United States of America ex rel. David
Franklin v. Pfizer, Inc. and Parke-Davis, Division of Warner-Lambert
Company). GMH and LAB were not involved in this litigation. MAS is
a member of the board of No Free Lunch. He does not receive any
compensation for this position and No Free Lunch had no role in the
preparation or submission of this manuscript. Seed funding for an
online search able archive of documents from the gabapentin
litigation (http: //dida.library.ucsf.edu) was provided by a gift from
Thomas Greene, lawyer for the whistleblower p laintiff in this
litigation, to the University of California Board of Regents. Two of
the authors (MAS, CSL) participated in the creation and development
of the archive, including solicitation of start-up funding from Mr.
Greene. The cost of obtaining and photocopying some of the
documents used in this research was paid by the archive, which
incorporated these documents into the collection. The authors have
no financial conflict of interest with any product discussed in this
1. Manchanda P, Honka E (2005) The effects and role of direct-to-physician
marketing in the pharmaceutical industry: An integrative review. Yale J
Health Policy Law Ethics 5: 785–822.
2. Angell M (2004) The truth about the drug companies: How they deceive us
and what to do about it. 1st Ed. New York: Random House. 336 p.
3. Kassirer J (2005) On the take: How America’s complicity with big business
can endanger your health. New York: Oxford University Press. 272 p.
4. Avorn J (2004) Powerful medicines: The beneﬁts, risks, and costs of
prescription drugs. 1st Ed. New York: Knopf. 480 p.
5. Ziegler MG, Lew P, Singer BC (1995) The accuracy of drug information
from pharmaceutical sales representatives. JAMA 273: 1296–1298.
6. Hemminki E (1977) Content analysis of drug-detailing by pharmaceutical
representatives. Med Educ 11: 210–215.
7. Stryer D, Bero LA (1996) Characteristics of materials distributed by drug
companies. An evaluation of appropriateness. J Gen Intern Med 11: 575–
8. Chren MM, Landefeld CS, Murray TH (1989) Doctors, drug companies, and
gifts. Jama 262: 3448–3451.
9. Dana J, Loewenstein G (2003) A social science perspective on gifts to
physicians from indust ry. JAMA 290: 252–255.
10. Waxman HA (2005 May 5) Memorandum to Democratic Members of
House Government Reform Committee: The Marketing of Vioxx to
Physicians. Available: http://www.democrats.reform.house.gov/documents/
20050505114932–4127 2.pdf. Accessed 5 May 2005.
11. Tiner R (1999) Reasons for not seeing drug representatives. They should be
seen because they are a good resource. BMJ 319: 1002.
12. Thomson AN, Craig BJ, Barham PM (1994) Attitudes of general
practitioners in New Zealand to pharmaceutical representatives. Br J Gen
Pract 44: 220–223 .
13. Benbow AG (1999) Reasons for not seeing drug representatives. Drug
representatives have much to offer. BMJ 319: 1003.
14. Lexchin J (1993) Interactions between physicians and the pharmaceutical
industry: What does the literature say? CMAJ 149: 1401–1407.
15. Wazana A (2000) Physicians and the pharmaceutical industry: Is a gift ever
just a gift? JAMA 283: 373–380.
16. Steinman MA, Shlipak MG, McPhee SJ (2001) Of principles and pens:
Attitudes and practices of medicine housestaff toward pharmaceutical
industry promotions. Am J Med 110: 551–557.
17. Donohue JM, Berndt ER, Rosenthal M, Epstein AM, Frank RG (2004) Effects
of pharmaceutical promotion on adherence to the treatment guidelines for
depression. Med Care 42: 1176–1185.
18. Mizik N, Jacobson R (2003) Are physicians ‘‘ easy marks’’?: Quantifying the
effects of detailing and sampling on new prescriptions. ISBM Report 9–
2003. University Park (Pennsylvania): Institute for the Study of Business
Markets. Available: http://www.smeal.psu.edu/isbm/database/docume nts/
wp03/09–2003.pdf. Accessed 21 March 2007.
19. Schwartz TL, Kuhles DJ 2nd, Wade M, Masand PS (2001) Newly admitted
psychiatric patient prescriptions and pharmaceutical sales visits. Ann Clin
Psychiatry 13: 159–162.
20. Manchanda P, Chintagunta PK (2004) Responsiveness of physician
prescription behavior to salesforce effort: An individual level analysis.
Marketing Lett 15: 129–145.
21. Parsons LJ, Abeele PV (1981) Analysis of sales call effectiveness. J Marketing
Research 18: 107–113.
22. Roughead EE, Harvey KJ, Gilbert AL (1998) Commerci al detailing
techniques used by pharmaceutical representatives to inﬂuence prescrib-
ing. Aust N Z J Med 28: 306–310.
23. (2003) Performance of sales representatives in France: Still bad. Prescrire
International 12: 153–154.
24. Shaughnessy AF, Slawson DC, Bennett JH (1995) Teaching information
Box 1: Classification of Main Message Content
Seizures, either specified as adjunctive therapy or not specified
Nonapproved indications or doses
Monotherapy for seizures
Pain or neuropathic syndromes
Migraine or headache
Other specified nonapproved indications (e.g., restless leg syndrome)
Dose over 1,800 mg / day
Nonapproved indications not otherwise specified (e.g., message was
‘‘ off-label uses,’’ ‘‘ non anti-convulsant use,’’ or ‘‘ growing uses’’ )
Other message content
Other comments that do not reference a specific indication or dose
(e.g., side effects, pharmacokinetics, general comments on dosing)
*At the time of the visits, the only FDA-approved indication for gabapentin was for the
adjunctive treatment of partial seizures in patients over age 12 y in doses 1,800 mg/day
Other specified off-label uses included reflex sympathetic dystrophy, restless leg syndrome,
essential tremor, and ‘‘ neuritis’’
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340749
Drug Detailing for Gabapentin
mastery: Evaluating information provided by pharmaceutical representa-
tives. Fam Med 27: 581–585.
25. Brotzman GL, Mark DH (1993) The effect on resident attitudes of
regulatory policies regarding pharmaceutical representative activities. J
Gen Intern Med 8: 130–134.
26. McKinney WP, Schiedermayer DL, Lurie N, Simpson DE, Goodman JL, et
al. (1990) Attitudes of internal medicine faculty and residents toward
professional interaction with pharmaceutical sales representatives. JAMA
27. Hodges B (1995) Interactions with the pharmaceutical industry: Experi-
ences and attitudes of psychiatry residents, interns and clerks. CMAJ 153:
28. Monaghan MS, Galt KA, Turner PD, Houghton BL, Rich EC, et al. (2003)
Student understanding of the relationship between the health professions
and the pharmaceutical industry. Teach Learn Med 15: 14–20.
29. US Department of Justice (2004 May 13) Warner-Lambert to pay $430
million to resolve criminal & civil health care liability relating to off-label
promotion. Available: http://www.usdoj.gov/opa/pr/2004/May/04_civ_322.
htm. Acces sed 3 February 2005.
30. Steinman MA, Bero LA, Chren MM, Landefeld CS (2006) The promotion of
gabapentin: an analysis of internal industry documents. Ann Intern Med
31. US District Court for the District of Massachusetts (2003 May 15) Afﬁdavit
of Debra Farese with Exhibits A–E in United States of America ex. rel David
Franklin vs. Pﬁzer, Inc., and Parke-Davis, Division of Warner-Lambert
Company. Civil Action No. 96–11651-PBS. Available: http://dida.library.
ucsf.edu/pdf/qha00a10. Accessed 21 March 2007.
32. US District Court for the District of Massachusetts (2003 February 19)
Exhibit 146 to the Afﬁdavit of Thomas M. Greene in Support of Relator’s
Opposition to Defendant’s Motion for Summary Judgment [Deposition of
Debra Farese], in United States of America ex. rel David Franklin vs. Pﬁzer,
Inc., and Parke-Davis, Division of Warner-Lambert Company. pp. 10,11,
25–32, 34, 37–48, 76–84, 87, 93–97 . Civil Action No. 96–11651-PBS.
Available: http://dida.library.ucsf.edu/pdf/eza00a10. Accesse d 21 March
33. Sergeant MD, Hodgetts PG, Godwin M, Walker DM, McHenry P (1996)
Interactions with the pharmaceutical industry: A survey of family medicine
residents in Ontario. CMAJ 155: 1243–1248.
34. Caudill TS, Johnson MS, Rich EC, McKinney WP (1996) Physician s,
pharmaceutical sales representatives, and the cost of prescribing. Arch
Fam Med 5: 201–206.
35. Lurie N, Rich EC, Simpson DE, Meyer J, Schiedemayer DL, et al. (1990)
Pharmaceutical representatives in academic medical centers: Interaction
with faculty and housestaff. J Gen Intern Med 5: 240–243.
36. Bower AD, Burkett GL (1987) Family physicians and generic drugs: A study
of recognition, information sources, prescribing attitudes, and practices. J
Fam Pract 24: 612–616.
37. Peay MY, Peay ER (1988) The role of commercial sources in the adoption of
a new drug. Soc Sci Med 26: 1183–1189.
38. Anadaleeb SS, Tallman RF (1996) Relationships of physicians with
pharmaceutical sales representatives and pharmaceutical companies: An
exploratory study. Health Mark Q 13: 79–89.
39. Peay MY, Peay ER (1990) Patterns of preference for information sources in
the adop tion of new drugs by specialists. Soc Sci Med 31: 467–476.
40. Hayes TM, Allery LA, Harding KG, Owen PA (1990) Continuing education
for general practice and the role of the pharmaceutical industry. Br J Gen
Pract 40: 510–512.
41. Yatham LN, Kusumakar V, Calabrese JR, Rao R, Scarrow G, et al. (2002)
Third generation anticonvulsants in bipolar disorder: A review of efﬁcacy
and summary of clinical recommendations. J Clin Psychiatry 63: 275–283.
42. Chronicle E, Mulleners W (2004) Anticonvulsant drugs for migraine
prophylaxis. Cochrane Database Syst Rev: CD003226.
43. Wiffen PJ, McQuay HJ, Edwards JE, Moore RA (2005) G abapentin for acute
and chronic pain. Cochrane Database Syst Rev: CD005452.
44. Steinman MA, Landefeld CS, Chren MM (2003) Expert consultant’s report:
United States ex rel. Franklin vs. Pﬁzer, Inc et al. Available at: http://dida.
library.ucsf.edu/pdf/mha00a10. Accessed 12 January 2007.
45. Roughead EE, Gilbert AL, Harvey KJ (1998) Self-regulatory codes of
conduct: Are they effective in controlling pharmaceutical representatives’
presentations to general medical practitioners? Int J Heal th Serv 28: 269–
46. Lexchin J (1997) What information do physicians receive from pharma-
ceutical representatives? Can Fam Physician 43: 941–945.
47. Shaughnessy AF, Slawson DC, Bennett JH (1994) Separating the wheat from
the chaff: Identifying fallacies in pharmaceutical promotion. J Gen Intern
Med 9: 563–568.
48. Molloy W, Strang D, Guyatt G, Lexchin J, Bedard M, et al. (2002) Assessing
the quality of drug detailing. J Clin Epidemiol 55: 825–832.
49. Hovland CI, Weiss W (1951) The inﬂuence of source credibility on
communication effectiveness. Public Opin Q 15: 635–650.
50. Calfee JE, Ringold DJ (1994) The 70% majority: Enduring consumer beliefs
about advertising. J Public Policy Mark 13: 228–238.
51. Bero LA, Grilli R, Grimshaw JM, Harvey E, Oxman AD, et al. (1998) Closing
the gap between research and practice: An overview of systematic reviews
of interventions to promote the implementation of research ﬁndings. The
Cochrane Effective Practice and Organization of Care Review Group. BMJ
52. Avorn J, Soumerai SB (1983) Improving drug-therapy decisions through
educational outreach. A randomized controlled trial of academically based
‘‘detailing.’’ N Engl J Med 308: 1457–1463.
53. Soumerai SB, Avorn J (1990) Principles of educational outreach (‘‘academic
detailing’’ ) to improve clinical decision making. JAMA 263: 549–556.
54. Figueiras A, Sastre I, Tato F, Rodriguez C, Lado E, et al. (2001) One-to-one
versus group sessions to improve prescription in primary care: A pragmatic
randomized controlled trial. Med Care 39: 158–167.
55. Landefeld SM, Cheung AV (2004) Corporate governance: Improving board
decisionmaking in post-Sarbanes-Oxley. Insights 18: 9–14.
56. van Eijk ME, Avorn J, Porsius AJ, de Boer A (2001) Reducing prescribing of
highly anticholinergic antidepressants for elderly people: Randomised trial
of group versus individual academic detailing. BMJ 322: 654–657.
57. Simon SR, Majumdar SR, Prosser LA, Salem-Schatz S, Warner C, et al.
(2005) Group versus individual academic detailing to improve the use of
antihypertensive medications in primary care: A cluster-randomized
controlled trial. Am J Med 118: 521–528.
58. Steinbrook R (2006) For sale: Physicians’ prescribing data. N Engl J Med
59. Greene JA (2004) Attention to ‘‘ details’’ : Etiquette and the pharmaceutical
salesman in postwar America. Soc Stud Sci 34: 271–292.
60. Zipkin DA, Steinman MA (2005) Interactions between pharmaceutical
representatives and doctors in training. A thematic review. J Gen Intern
Med 20: 777–786 .
61. Madhavan S, Amonkar MM, Elliott D, Burke K, Gore P (1997) The gift
relationship between pharmaceutical companies and physicians: An
exploratory survey of physicians. J Clin Pharm Ther 22: 207–215.
62. Hopper JA, Speece MW, Musial JL (1997) Effects of an educational
intervention on residents’ knowledge and attitudes toward interactions
with pharmaceutical representatives. J Gen Intern Med 12: 639–642.
63. Rosenthal MB, Berndt ER, Donohue JM, Frank RG, Epstein AM (2002)
Promotion of prescription drugs to consumers. N Engl J Med 346: 498–505.
64. Blumenthal D (2004) Doctors and drug companies. N Engl J Med 351: 1885–
65. No Free Lunch (2007) Sources of drug info. Available: http://nofreelunch.
org/sources_druginfo.htm. Accessed 27 April 2005.
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340750
Drug Detailing for Gabapentin
Background. In the US, before a pharmaceutical company can market a
drug to doctors for use in a specific ‘‘ indication’’ (meaning the treatment
for a particular disease and group of patients), the drug has to be
approved as safe and effective for that use by a government agency, the
Food and Drug Administration. Once approved, doctors are allowed to
use a drug for whatever nonapproved indications they think are
appropriate, but the drug company cannot actively promote the drug
for anything other than its approved use. However, many people are
concerned that drug companies indirectly try to promote use of drugs
for indications that are not approved. Such illegal activity would help a
drug company increase its market share and sell more drugs. One tactic
that drug companies use to sell drugs is ‘‘ detailing.’’ Detailing involves
direct visits from drug company representatives to individual doctors,
during which the representative would provide information about their
company’s drugs. However, not a great deal is known about detail visits
and the effect that they have on doctors’ attitudes towards the drugs
that are being promoted.
Why Was This Study Done? The researchers carrying out this study
wanted to learn more about what happens during detail visits and what
impact these visits have on prescribing behavior. An opportunity for
researching this came about as a result of a lawsuit during which drug
company documents were subpoenaed (i.e., required by the court to be
made available). In that lawsuit, it was alleged that a drug company,
Parke-Davis, had promoted a drug, gabapentin, for many nonapproved
uses. The company that subsequently took over Parke-Davis eventually
made an out-of-court settlement. During the relevant time period, the
only approved use of gabapentin was for treatment of partial seizures in
adults with epilepsy, in combination with other drugs. However,
gabapentin was used for many other conditions such as treatment of
psychiat ric disorders and management of pain. Thes e researchers
therefore used the documents available as a result of the lawsuit to
research detailing and what impact detailing had on doctors’ attitudes
towards the drug being promoted.
What Did the Researchers Do and Find? The documents analyzed in
this study were produced by Verispan, a market research company.
Verispan asked doctors who had been visited by Parke-Davis sales
representatives to fill out a standard form after each detail visit. These
forms were then subpoenaed as part of the lawsuit against Parke-Davis.
The researchers here focused specifically on data relating to visits made
by a single sales representative to a doctor or small group of doctors,
and collected 116 forms. The data available from these forms included
the doctors’ ratings and comments regarding the main message
associated with the products; the informational value of the visit; the
quality of the presentation; and whether the doctor currently prescribed
or planned to prescribe the product. The researchers classified the
information available from the forms, identifying whether the ‘‘ main
message’’ related to approved uses of the drug or not; and extracting
data relating to whether doctors planned to increase, maintain, or
decrease their use of the drug. The majority of the visits studied were to
doctors who were not neurologists, and would therefore be unlikely to
prescribe gabapentin for its approved use. Doctors reported that a
substantial proportion of the detail visits contained messages relating to
nonapproved uses of gabapentin. Nearly half the doctors stated in the
forms that their use of gabapentin would increase in the future, and no
doctors said that their use would decrease following the visit. Doctors’
intention to increase their use of gabapentin in the future was similar
regardless of whether the message of the visit involved an approved or
What Do These Findings Mean? This study shows that in the case of
gabapentin, detail visits by drug company representatives frequently
promoted nonapproved uses of the drug; these visits often resulted in
doctors planning to increase their use of gabapentin. However, it is not
clear whether these findings are also true for other drugs and drug
companies, in part because these data came about as a result of a
unique opportunity granted by the lawsuit against Parke-Davis.
Additional Information. Please access these Web sites via the online
version of thi s summary at http://dx.doi.org/10.1371/journal.pmed.
Medline Plus (provided by the US National Library of Medicine) has an
entry about gabapentin
Introductory information is available from the US FDA Center for Drug
Evaluation and Research about the drug approvals process in the USA
Wikipedia has an entry on pharmaceutical marketing (Note that
Wikipedia is an internet encyclopedia anyone can edit)
The Drug Industry Document Archive is available at University of
California, San Francisco; this internet archive holds documents
relating to the lawsuit against Parke-Davis and from which the data
presented in this paper derives
Guidance is available from the International Federation of Pharma-
ceutical Manufacturers and Associations regarding ethical promotion
PLoS Medicine | www.plosmedicine.org April 2007 | Volume 4 | Issue 4 | e1340751
Drug Detailing for Gabapentin