ArticlePDF AvailableLiterature Review

As-needed Use of Benzodiazepines in Managing Clinical Anxiety: Incidence and Implications

Authors:

Abstract

Benzodiazepines (BZs) have been widely investigated in terms of clinical efficacy, factors underlying dependence, associated cognitive impairments, and interactions with psychotherapy for anxiety control. However, few studies have systematically considered manner of BZ administration in relation to these variables. Studies of chronic BZ users indicate that as-needed or p.r.n. use is a very common practice, increases with chronicity of BZ use, and is preferred compared to regularly scheduled BZ administration. Moreover, a recent study of physician prescription practices indicated that p.r.n. BZ use is a commonly recommended BZ use regimen for anxiety disorder management. Physician advocates of p.r.n. BZ prescriptions for anxiety disorders cite enhanced patient control over symptoms, facilitation of exposure to fear-provoking situations, and reduced frequency of use as rationales supporting this practice. Available data however, do not consistently support these hypothesized advantages of p.r.n. BZ use. And in general, findings from different investigations relevant to this question suggest that p.r.n. BZ administration may be associated with increased patient preference for BZs over placebo, continued use, and greater impairment on cognitive factors associated with positive long-term anxiety management. Ironically, p.r.n. BZ administration may also be associated with reduced anxiolytic efficacy over time. These suggestive findings argue for greater systematic investigation of manner of BZ administration as an important medication use parameter. Such investigations may also yield practical guidelines for navigating BZ discontinuation and promoting more successful long-term management of anxiety.
Current Pharmaceutical Design, 2002, 8, 59-74 59
As-needed Use of Benzodiazepines in Managing Clinical Anxiety: Incidence
and Implications
H. A. Westra
1*
and S. H. Stewart
2
1
Anxiety & Affective Disorders Service, London Health Sciences Centre, London, Ontario,
Canada
2
Departments of Psychology & Psychiatry, Dalhousie University, Halifax, Nova Scotia,
Canada
Abstract: Benzodiazepines (BZs) have been widely investigated in terms of clinical
efficacy, factors underlying dependence, associated cognitive impairments, and
interactions with psychotherapy for anxiety control. However, few studies have systema-
tically considered manner of BZ administration in relation to these variables. Studies of chronic BZ users
indicate that as-needed or p.r.n. use is a very common practice, increases with chronicity of BZ use, and is
preferred compared to regularly scheduled BZ administration. Moreover, a recent study of physician
prescription practices indicated that p.r.n. BZ use is a commonly recommended BZ use regimen for anxiety
disorder management. Physician advocates of p.r.n. BZ prescriptions for anxiety disorders cite enhanced patient
control over symptoms, facilitation of exposure to fear-provoking situations, and reduced frequency of use as
rationales supporting this practice. Available data however, do not consistently support these hypothesized
advantages of p.r.n. BZ use. And in general, findings from different investigations relevant to this question
suggest that p.r.n. BZ administration may be associated with increased patient preference for BZs over placebo,
continued use, and greater impairment on cognitive factors associated with positive long-term anxiety
management. Ironically, p.r.n. BZ administration may also be associated with reduced anxiolytic efficacy over
time. These suggestive findings argue for greater systematic investigation of manner of BZ administration as an
important medication use parameter. Such investigations may also yield practical guidelines for navigating BZ
discontinuation and promoting more successful long-term management of anxiety.
INTRODUCTION Pharmaceuticals and Specialties [8]). BZs are unique among
psychotropic medications in that they can be administered in
a regularly scheduled fashion (e.g. t.i.d., b.i.d.), but also on
an “as-needed” or p.r.n. basis. BZs may be utilized in this
latter manner given that the onset of drug effect is relatively
rapid, in comparison to other psychotropic medications such
as antidepressants which typically require several weeks of
regular use in order to achieve therapeutic efficacy [9]. This
means that patients can exert considerable control over the
timing and dose of BZ administered and may use them in
response to fluctuating daily symptoms of anxiety. For
example, BZs may be used to circumvent an oncoming
panic attack, to dampen generally heightened anxiety, or to
approach a feared or avoided situation [10].
Although prescription rates are changing, BZs continue
to be prescribed at high, albeit declining rates [1,2]. They
continue to be sanctioned by general physicians [1,3] and
psychiatrists [4,5] for long-term use in clinical anxiety
management. Moreover, there is a large group of existing
chronic BZ users suffering from clinical anxiety who have
difficulty discontinuing BZs [6,7]. Each of these factors
supports the importance of continued investigation of BZs
along a number of dimensions. BZs have been extensively
investigated and many variables have been considered in
relation to BZ use including clinical efficacy, abuse liability,
cognitive effects, and pharmacological variables underlying
dependence and withdrawal. These factors have been
considered across various types of BZs and different patient
populations.
Despite the potential of BZs to be administered in very
different ways (i.e., regularly or p.r.n.), little is known
regarding actual physician prescribing practices and/or the
manner in which BZs are self-administered by patients. In
those who are not chronically anxious but experiencing acute
anxiety, p.r.n. BZ use is common. Examples include
patients with transient sleep or adjustment problems [3] or
those undergoing brief aversive medical procedures [8].
However, in terms of prescription guidelines for management
of more chronic anxiety, few recommendations on manner of
BZ administration are offered. Moreover, despite intuitive
suggestions that as-needed use may have negative
consequences (e.g. [11]), little is known regarding the
Although BZs are widely investigated, manner of BZ
administration has not been routinely considered in relation
to the effects of BZ use and guidelines for BZ prescription are
relatively silent on this issue (e.g., Compendium of
*Address for correspondence to that author at the Anxiety and Affective
Disorders Service, W818 Victoria Campus, 375 South Street, London
Health Sciences Centre, London, Ontario, Canada, N6A 4G5, E-mail:
westrah@lhsc.on.ca, Tel: (519) 685-8500 ext. 75602, Fax: (519) 667-6864.
1381-6128/02 $35.00+.00 © 2002 Bentham Science Publishers Ltd.
60 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
consequences of p.r.n. BZ administration in relation to
treatment efficacy or dependence liability. Given the
continued high frequency of long-term BZ use in anxiety
disorders, coupled with commonly encountered
discontinuation difficulties (see reviews by [12,13], this
issue) and poor long-term treatment efficacy [14-16],
delineation of any significance of manner of BZ
administration is clearly warranted.
[7] and in a group of 133 chronic users of either lorazepam or
alprazolam [6]. These investigators reported that, over time,
BZ users tended to shift their pattern of use from regularly
scheduled to a pattern of as-needed use or a combination of
regular and as-needed use. Individuals with anxiety disorders
seem to be well-represented in this chronic user group. For
example Romach et al. [6], found that on diagnostic
interviewing, almost 50% of chronic BZ users had a current
anxiety disorder and an additional 18% had a lifetime history
of anxiety disorder.
This review has two principle goals. First, we consider
the frequency of p.r.n. BZ administration in terms of both
patient use and physician prescribing practices for the
treatment of anxiety disorders. That is, can we assume that
BZs are prescribed and/or taken on a regularly scheduled
basis for control of clinical anxiety? Since no published data
are available on actual prescription practices for manner of
BZ use, we present recent data from an author-conducted
survey of family physicians and psychiatrists [17]. This
review will consider whether the base rate of p.r.n.
administration is high enough to merit further consideration
of manner of use in future investigations of BZs in clinical
anxiety management. This review also considers the
rationales or motivations underlying p.r.n. BZ
administration among patients and prescribers.
We have analyzed data on manner of BZ use in our own
sample of chronic (duration of use of 3 months or more;
median = 30 months) and daily BZ-using patients
presenting for psychotherapy for panic disorder (n=40; [19]).
These patients were using alprazolam (37%), clonazepam
(32%), lorazepam (25%), or diazepam (6%). Among this
group, only 5% reported using BZs exclusively on a
regularly scheduled basis. Ninety-five percent of chronic
users reported administration of BZs in an as-needed fashion
at least sometimes, and 12.5% reported exclusive p.r.n. use.
Motivations for p.r.n. BZ use were examined among those
indicating frequent as-needed BZ use (i.e. scores greater than
or equal to 4 on a 1-7 scale of p.r.n. use with anchors of
“never” and “always”; n = 23). The most common reasons
endorsed for as-needed BZ use was to facilitate exposure to
feared situations and to dampen heightened general anxiety
(87% of frequent p.r.n. users indicated often using BZs for
each of these purposes). In addition, 78% of high frequency
p.r.n. BZ users reported frequent as-needed BZ use for
dampening an oncoming panic attack. These figures from a
group of panic disorder patients with clinical anxiety
requiring treatment, are largely consistent with data from
community samples reflecting a high rate of p.r.n. BZ use
among chronic BZ users (e.g., [18]).
The second goal of this review is to consider available
data on the implications of regular and p.r.n. BZ use in order
to assess whether one method of administration is preferable
to the other. Four areas of investigation potentially relevant
to this question are reviewed: (1) manner of BZ use and
clinical efficacy or control over anxiety symptoms; (2)
manner of BZ administration and rates of BZ use; (3) manner
of BZ use and dependence liability and discontinuation
difficulties; and (4) the impact of manner of BZ use on
psychological variables associated with long-term effective
anxiety control such as self-efficacy and fear reduction.
The use of retrospective self-reports is a limitation of
these studies [6,7,18,19]. Self-reports assume patient
awareness of the motives underlying their medication use
behavior – an assumption that has been questioned (e.g.,
[20]). Retrospective reports are also potentially subject to
memory errors [21]. It remains unknown to what degree
patients who report using BZs in a regularly scheduled
fashion actually take their medications “by-the-clock”, and
to what degree p.r.n. users’ reports of their reasons for as
needed use reflect reality. Thus, future studies would be
enhanced by the use of prospective self-monitoring of anxiety
patients’ BZ use patterns and their relations to fluctuating
daily symptoms of anxiety.
FREQUENCY OF REGULAR AND AS-NEEDED BZ
ADMINISTRATION SCHEDULES
Patient Use
Several epidemiological investigations of long-term BZ
users have yielded data on rates of regular and as-needed BZ
administration. Collectively, these studies find that BZs are
commonly utilized on a p.r.n. basis among chronic users. In
a survey of 25 long-term alprazolam users (median duration
of use = 104 weeks), Romach et al. [18] found that the large
majority of users administered alprazolam on a p.r.n. basis
either exclusively (36%) or in combination with regularly
scheduled use (36%). Only 28% of users reported utilizing
alprazolam solely on a regularly scheduled basis. Moreover,
Romach et al. [18] reported that use of alprazolam tended to
shift to greater patient-controlled, p.r.n. administration over
time. In this study, as well as reporting on their current
manner of BZ administration, BZ users also retrospectively
reported on their initial manner of administration at the time
of original prescription. It was found that the number of users
taking alprazolam exclusively on a p.r.n. basis had more
than doubled from the time of original prescription. This
pattern toward greater p.r.n. use over time was also
replicated in another survey of 188 chronic alprazolam users
In a prospective study Dammen et al. [22], assessed
patient preference for on-demand BZ-taking in a group of 12
diagnosed generalized anxiety disorder patients taking
different BZs on a daily basis (n = 7 diazepam users, n = 4
oxazepam users, n = 1 chlorodiazepoxide user). In this
group, on-demand use was found to be the most common,
and preferred, method of taking BZs. Moreover, 66% of
patients expressed a continued strong preference for p.r.n. BZ
use even when exposed to a fixed schedule of use for one-
week.
In summary, these previous studies reveal that p.r.n. use
of BZs is very common among chronic BZ users, and
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 61
among those with a major anxiety disorder. A clear majority
of patients report using BZs as-needed at least sometimes,
either in combination with regularly scheduled use or
exclusively on a p.r.n. basis. As such, it would seem that
many anxiety disorder patients exert significant control over
the timing and administration of BZ medication, and do so
increasingly over time. Finally, it would also seem that
anxiety disorder patients exhibit a strong preference to use
BZs in a p.r.n. fashion, citing facilitation of exposure to fear-
provoking situations and anxiety symptom control as
reasons for their preference for p.r.n. use.
When psychiatrists were compared with family
physicians in their manner of BZ prescription, a clear
divergence in prescription practices was observed.
Psychiatrists favoured prescribing BZs to be taken in a
regularly scheduled fashion (Mean = 55.5%) for anxiety
disorder cases where a BZ is prescribed. And in only an
average of 21.5% cases were BZs reportedly prescribed
exclusively on a p.r.n. basis for anxiety disorder treatment
by psychiatrists. Moreover, self-reported prescrip-tion rates
among psychiatrists for regular BZ use were significantly
higher than prescription rates for p.r.n. administration (X
2
(1)
= 15.02, p < 0.05). In contrast, 54.6% of BZ prescriptions
for anxiety disorder control by family physicians were for
exclusive p.r.n. administration. Family physicians reported
prescribing BZs to be taken on an exclusively fixed
administration schedule in only 32.6% of BZ prescriptions.
P.r.n. prescription rates of BZs among family physicians
were significantly higher than the self-reported prescription
rates of regularly scheduled BZ use (X
2
(1) =96.62, p <
0.05). Respondents were also asked to indicate their opinion
on how advantageous p.r.n. administration of BZs is on a 4-
point scale ranging from “It is disadvantageous or counter-
therapeutic for anxiety patients to use BZs p.r.n.” to “It is
very advantageous for anxiety patients to use BZs p.r.n.”.
Consistent with these differences in prescribing patterns,
compared to psychiatrists, GPs rated p.r.n. BZ use as
significantly more advantageous [t (122) = 2.66, p < 0.05].
Physician Prescription Practices
Although patients may use BZs commonly on an as-
needed basis, a complementary question concerns actual
prescription practices among physicians. Many physician
surveys have yielded data on BZ prescription rates [1], the
clinical indications for BZ prescription [3], and the types of
practices yielding high BZ prescription rates (e.g. higher
incidence of BZ prescriptions in inner city versus rural
practices, [3]). Yet, no study has systematically investigated
physicians’ actual manner of BZ prescription. How are
patients instructed to use BZs? Is p.r.n. use initiated by the
patient or the prescriber?
We surveyed family physicians and psychiatrists in Nova
Scotia via mail regarding recommended manner of BZ
administration for patients presenting with an anxiety
disorder in the past year [17]. A total of 132 physicians
completed the survey, representing a response rate of 26.4%.
The response rates for psychiatrists was 34% (n = 38) and for
family physicians was 25.2% (n = 94). Although the
response rate is relatively low, this response rate is
comparable to other physician surveys on mental health
issues (e.g., [23]).
A relatively low incidence of prescribing both regularly
scheduled and p.r.n. use was observed in both groups
(psychiatrists: 23.0%, family physicians: 12.8%).
Psychiatrists preference for regularly scheduled BZ
prescriptions was significant compared to their prescription
rates for both manners of BZ use (X
2
(1) =13.46, p < 0.05).
In a similar manner, family physician’s prescription
recommendations for p.r.n. BZ use were significantly higher
than their self-reported rate of prescription for both regular
and p.r.n. BZ use (X
2
(1) = 25.92, p < 0.05).
The results of the survey indicated that prescription
practices are sharply divided regarding the recommended
manner of BZ administration. Of the cases where they
prescribed a BZ for the treatment of anxiety disorder,
physicians were asked to indicate the percentage of cases they
prescribed BZs for regularly scheduled use only, p.r.n. use
only, or both. For the total sample of Nova Scotia
physicians, BZs were reportedly prescribed exclusively on a
regularly scheduled basis (i.e. b.i.d., t.i.d. etc.) in an average
of 38.4% of anxiety disorder cases. A relatively high
percentage (Mean = 46.3%) of anxiety disorder cases were
reportedly prescribed BZs for exclusive p.r.n. use. The
difference between self-reported prescription rates of p.r.n. and
regularly scheduled BZ use was not significant (X
2
(1) =
0.74, n.s.), indicating that both are commonly recommended
manners of BZ prescription among physicians. Moreover,
simultaneous regularly scheduled and p.r.n. use was
reportedly recommended in only a small percentage (Mean =
15.3%) of cases of BZ prescription for anxiety disorder
management. Either regularly scheduled BZ administration
or p.r.n. BZ administration was recommended significantly
more often than both manners of BZ administration
simultaneously (regular use compared to both: X
2
(1) =
9.94, p < 0.05; p.r.n. use compared to both: X
2
(1) = 15.60,
p < 0.05).
The results of the survey suggest a number of interesting
conclusions. First, it cannot be assumed that BZs are
routinely prescribed in a regularly scheduled fashion for
anxiety disorder management. BZs are reportedly prescribed
for exclusive use in response to fluctuating anxiety
symptoms at equivalent rates when compared to regularly
scheduled use. In addition to highlighting a high rate of
p.r.n. prescription, this finding also suggests that
prescription practices for BZs are highly variable with both
exclusive fixed administration and exclusive p.r.n.
administration being commonly recommended. And in the
vast majority of cases, physicians seem to prefer either one or
the other manner of administration, as opposed to
recommending both simultaneously. Finally, psychiatrist
and family physician prescription practices clearly diverge. In
the majority of cases of BZ prescription for the treatment of
anxiety disorder, psychiatrists clearly favour regularly
scheduled BZ use, whereas family physicians clearly favour
p.r.n. BZ prescriptions.
The differing BZ prescription practices between
psychiatrists and family physicians could be secondary to
any number of differences between these two groups:
populations served, patient and practice demands, familiarity
62 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
with diagnosis and treatment of anxiety disorders, access to
alternative psychological treatments, concerns about
discontinuation, etc. One possibility may be that family
physicians and psychiatrists see anxiety patients at different
points in their disorder. Family physicians may tend to see
anxiety patients early on in the course of their disorder. In
these cases, it may very well be that p.r.n. prescription of a
BZ is an effective practice for relief of possibly transient
anxiety symptoms. In contrast, psychiatrists may see these
patients when the diagnosis is well established and the
symptoms more chronic. In these cases, favouring more
regular BZ administration may well be indicated to avoid
inadvertent precipitation of withdrawal symptoms, for
example. However, the actual reasons for this discrepancy in
the BZ prescribing practices of family physicians versus
psychiatrists remains unknown.
guiding this practice seem to converge between patients and
physicians and include facilitating function (i.e. enhancing
exposure and mobility) and increasing symptom control.
Moreover, many physicians also feel that p.r.n. BZ use may
minimize use, reduce addictive potential, and facilitate
ultimate discontinuation. The validity of these assumptions
of increased symptom control, improved function, etc. await
empirical confirmation. Although physicians and patients
have clear preferences for manner of BZ administration, there
is little empirical data regarding which administration
practice is preferable. In the remainder of this review, we turn
our attention to this question. We review available empirical
evidence relevant to the issue of manner of BZ use. We
attempt to answer the question of whether there are
advantages to either as-needed or regularly scheduled BZ
use.
IMPLICATIONS OF MANNER OF BZ
ADMINISTRATION
The results of this survey raise questions regarding the
rationale underlying these divergent prescription practices for
BZs. The inconsistency in practice may highlight the lack of
adequate guidelines regarding the optimal regimen for BZ
prescription and also the lack of empirical data relevant to
the question of the utility or efficacy of one or the other
practice. In the absence of data or guidelines, what are
physicians using to guide them toward one or the other
prescription regimen? Our physician survey also explored
possible factors guiding the prescription of BZs for p.r.n. use
[17]. To examine this issue, we limited the sample to those
physicians indicating that they prescribe BZs in a p.r.n.
fashion, either exclusively or simultaneously with regular
use (n = 108). Physicians were asked to indicate the reasons
underlying p.r.n. BZ prescriptions by endorsing one or more
of five possible rationales. The most common reason
endorsed for p.r.n. BZ prescription was to enhance perceived
patient control (“to give the patient a greater sense of direct
control over anxiety symptoms”) and 70.8% of respondents
endorsed this item. The second most common rationale for
recommending use of BZs p.r.n. was to facilitate exposure
(“to help the person go into situations they might otherwise
avoid”) and 69.8% of respondents endorsed this item. A
high percentage (64.6%) of p.r.n. prescribers also endorsed
“to minimize use” as a rationale for as-needed use of BZs.
Finally, 45.8% of respondents endorsed “to facilitate
ultimate discontinuation” and 44.8% endorsed “to reduce
addictive potential” as reasons for p.r.n. BZ prescription.
In the following section we review available data relevant
to manner of BZ use along a number of dimensions. This
review largely investigates the validity of assumptions,
noted previously, underlying p.r.n. BZ prescription and use.
A number of questions relevant to manner of BZ use are
proposed and evaluated. First, does p.r.n. BZ use result in
more effective anxiety control and enhanced function?
Second, is as-needed BZ use associated with minimizing
overall levels of BZ use? Third, does p.r.n. BZ use reduce
addictive potential and enhance discontinuation ability?
Finally, does manner of BZ use have any impact on
psychological factors empirically associated with effective
long-term anxiety control? It should be noted that any
conclusions based on this review are tentative given that
manner of use has largely been neglected as an important
parameter of medication use in studies of BZs. Findings
from seemingly disparate literatures on aspects of BZs (BZ
reinforcement, conditioned drug effects, BZ dependence), and
more broadly on the correlates of p.r.n. administration of
other fast-acting medications (i.e. narcotics), are integrated.
The primary goal of this section is to determine whether or
not manner of BZ use merits further empirical investigation.
(1) Does p.r.n. BZ Use Provide Better Control Over
Anxiety Symptoms and Enhanced Function?
This physician survey is limited however by sample size
and regional representation. It may also be that certain BZs
are more likely to be prescribed in a p.r.n. fashion than
others and this was not assessed in the survey. For example,
considering that many physicians reported prescribing BZs
p.r.n. to minimize BZ use by patients and to reduce
addictive potential, it is possible that they may be
prescribing faster acting BZs, or those with greater addictive
potential, on a p.r.n. basis more often than slower acting
BZs, or those with lesser addictive potential. The
possibility of manner of BZ prescription varying across BZs
with different pharmacokinetic properties remains to be
investigated in future research.
Perceived Control and Self-efficacy
As noted earlier, physician advocates of p.r.n. BZ
prescriptions cited providing patients with enhanced control
over anxiety symptoms as a common rationale underlying
as-needed BZ use [17]. Similarly, many panic disorder
patients self-report using BZs in a p.r.n. fashion to control
fluctuating symptoms of anxiety and panic [10]. In general,
available data seem to support the assumption of an
enhanced perception of control with p.r.n. BZ
administration.
In a group of generalized anxiety disorder patients,
Dammen et al. [22] found that existing BZ user’s preference
for p.r.n. BZ administration was subjectively motivated by a
perception of “good symptom control” (41% of p.r.n. users)
In summary, available survey data seem to indicate that
p.r.n. use of BZs is common in terms of both patient use
and family physician prescription practices. The rationales
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 63
and “gives a feeling of security” (17% of p.r.n. users).
Conversely, when this group of p.r.n. users was exposed to
regularly scheduled administration for one week, Dammen et
al. noted increased anxiety and poorer symptom control
among 57% of users [22].
on-demand narcotic administration, there is less agreement
as to whether pain is more effectively controlled with patient-
controlled analgesia when compared to fixed interval
administration. Gust et al. reported that post-operative
cardiac patients achieved better pain control with patient-
controlled analgesia versus conventional fixed administration
analgesia [34]. In contrast, a number of studies find either
equivalent pain control across the two methods of analgesic
administration, or inferior pain control with on-demand
dosing. Deconno et al. [35] reported that analgesics
delivered post-operatively at fixed intervals were associated
with significantly better pain control, better sleep, and
improved function when compared to on-demand analgesic.
Similar findings were reported by Berntzen and Gotestam
[36] who found that a time-contingent schedule was more
effective in reducing pain ratings in chronic pain patients
than an on-demand schedule. Moreover, Tsang and Brush
[37] reported that despite no group differences in overall
dose, quality of post-operative pain control was comparable
between fixed-interval and as-needed administration of
morphine. The authors concluded that there is no significant
advantage to using patient-controlled analgesia routinely in
the early post-operative period, in terms of effects on
measures of pain.
In further support of an enhanced perception of symptom
control with as-needed use, anxiety patients are commonly
found to gain reassurance from merely possessing medication
[24]. In our survey of physicians (psychiatrists and family
physicians combined), 39% reported recommending in at a
least some cases, that patients carry BZs with them for
increased security [17]. Barlow et al. (1998) reported the
most common safety signal among panic disorder with
agoraphobia patients is carrying anxiolytic medication [25].
An additional literature that may be relevant to questions
of the effects of as-needed administration of rapidly acting
medications concerns studies of narcotic administration on a
p.r.n. basis. Although the population and the
pharmacotherapeutic agents involved are clearly different,
many of the concerns in the area of opiate analgesics appear
to parallel concerns with BZs, such as addictive potential,
perceived control, withdrawal upon medication
discontinuation, minimizing use, and efficacy. The impact of
as-needed narcotic administration has been extensively and
systematically studied in pain populations [26,27] and
consequently these investigations may provide a parallel
model for considering the implications of self-control of fast-
acting anxiolytic medications such as BZs. A relatively
recent development in this area has been patient-controlled
analgesia [28]. Here, patients are allowed to self-administer
predetermined small analgesic doses as-needed. This
contrasts with conventional analgesic administration, which
occurs on a fixed-interval basis.
In large-scale reviews of studies on post-operative pain,
earlier reviews concluded that patient-controlled analgesia
achieved equivalent to improved pain control [27] when
compared to conventional fixed-interval-administered
analgesic. However, a more recent review by Lehmann
concluded that although patients generally prefer self-control
over analgesic administration [26], pain relief is not
necessarily better than with conventional techniques.
Moreover, Etches [28] argues that there is no compelling
evidence that patient-controlled analgesia is associated with a
reduction in morbidity or a more rapid recovery.
A very robust finding in experimental studies of patient-
controlled analgesia is increased patient satisfaction with as-
needed administration, when compared to conventional
analgesic administration [26,27,29-32]. Among the
subjective advantages identified by patients are improved
pain relief with fewer side effects [29], and greater self-efficacy
and feelings of control with the ability to take a more active
role in their own care [33]. Improved patient satisfaction is
routinely cited in studies of patient-controlled analgesia,
despite wide variability in degree of pain control reported
across studies [26]. Taken together, these findings support
the assumption that patients derive an increased sense of
control with p.r.n. administration of fast-acting medications
such as BZs and opiate analgesics. The fact that anxiety
patients have a perception of increased anxiety control and
improved medication efficacy with as-needed BZ
administration in response to fluctuating anxiety symptoms,
may account for the common practice of carrying medication
(e.g. [25]) to facilitate a greater feeling of security and
control.
Similarly with BZs, subjective preference and perception
of decreased anxiety with p.r.n. BZ administration may not
be associated with objectively better anxiety control relative
to fixed-interval administration. For example, although
safety signals such as medication availability may produce
initial fear reduction, the important contribution of safety
signals to longer-term fear and avoidance perpetuation in
various anxiety disorders has been elaborated by Rachman
and his colleagues [38,39]. We now turn our attention to the
possible discordance between subjective and objective
symptom control with p.r.n. BZ administration.
Is Objectively Better Symptom Control Achieved with
p.r.n. BZ Administration?
Numerous studies have investigated the efficacy of fixed-
interval BZ administration in the treatment of clinical
anxiety. Effective panic control with BZs, particularly high
potency medications (e.g. alprazolam, lorazepam), have been
reported (e.g. [40,41]). Similar findings supporting the
efficacy of BZs in treating a broad range of anxiety disorders
have been accumulated [42-45]. However, when initial short-
term gains are re-examined at longer-term follow-up
intervals, treatment benefits with BZs clearly decline
[14,15,46,47]. This pattern of poor long-term consolidation
of BZ treatment gains is a fairly robust finding across a
As-needed Narcotic Administration and Efficacy
Again, data from controlled investigations on the efficacy
of p.r.n. and fixed administration of narcotics may be
relevant when considering the impact of manner of use on the
anxiolytic efficacy of BZs. Despite clear patient preference for
64 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
variety of different anxiety disorders and BZ compounds (see
review [48]).
as a result of its previous administration [51]. It is well
established that tolerance to BZ’s sedative and
anticonvulsant effects develops relatively rapidly. Tolerance
to BZ’s anxiolytic effects is more controversial [52], but
does appear to occur (at least in some cases and under some
conditions) in both animals and humans, albeit relatively
slowly. Physiological mechanisms underlying BZ tolerance
are reviewed by Bateson [52]. At another level of analysis,
learning mechanisms may be involved in BZ tolerance (see
reviews [51,53]).
In one of the only controlled studies investigating p.r.n.
BZ administration, Schweizer et al. [49] investigated the
impact of a rapidly acting BZ (i.e. midazolam) administered
intra-nasally on a p.r.n. basis for panic control. The results
indicated that small doses were highly effective in reducing
weekly panic frequency. Interestingly, these investigators
reported strong reductions in panic frequency despite very
infrequent patient use of midazolam. They conclude that this
effect may possibly be explained by “the illusion of control”
conveyed by simply possessing the drug and that much of
the therapeutic benefit reported may be attributable to non-
pharmacological factors such as the “safety signal” value of
carrying midazolam. This study was limited by a very small
sample size (n=5), short study period (6 weeks), and the
failure to include a fixed-interval BZ administration group for
comparison. Nevertheless, the results suggest the possibility
of effective panic control in the short-term with p.r.n.
administration of BZs and that this effect may be, in part,
mediated by non-pharmacological (i.e., expectancy) factors.
Further investigation of this effect would necessitate
examination of longer-term efficacy of p.r.n. administration,
given findings of poor long-term anxiety control with BZ
treatments [47].
Although little data are available on humans, conditioned
drug tolerance may be another factor suggestive of poorer
anxiolytic efficacy with p.r.n. BZ administration. The
phenomenon of conditioned drug tolerance suggests that
contextual or internal cues that are repeatedly associated with
drug intake can establish conditioned responses that alter the
drug’s unconditioned pharmacological effects [53,54]. These
conditioned responses are often compensatory and are
reported to be in the opposite direction of the drug’s original
unconditioned effect. For example, if decreased heart rate is
consistently achieved with drug intake, a conditioned
compensatory response could involve increased heart rate
prior to drug intake. With repeated exposure to drug taking,
clinical drug effects may be a balance of compensatory
conditioned responses and unconditioned drug effects. To
extrapolate to BZs, the development of compensatory
conditioned responses potentially altering the anxiolytic
effects of BZs, may be more likely under conditions of
symptom-contingent p.r.n. use [22]. In the case of the
anxious p.r.n. BZ user, bodily sensations or subjective
feelings of anxiety and/or anxiety-provoking external
contexts are cues that consistently precede BZ use. The body
may respond to these cues with a further compensatory
response of arousal, to prepare for the arousal-dampening
effect of the BZ. This increased arousal compensatory
response could come to interfere with the drug’s anxiolytic
effects in p.r.n. users over time. With regularly scheduled
use, on the other hand, there are no such consistent cues
signaling oncoming BZ since medications are administered
independent of anxiety symptoms. Under these
circumstances, conditioned compensatory responses would
be less likely.
It is not possible to examine findings across these studies
in order to make conclusions regarding whether fixed or
p.r.n. administration is preferable for objective symptom
control. No well-controlled investigation has directly
compared p.r.n. and fixed BZ administration for anxiety
control within the same study. Using a naturalistic design,
Westra et al. [10] compared the performance of high
frequency p.r.n. BZ users with regularly scheduled BZ users
in a cognitive-behavioural treatment (CBT) for panic
disorder. We found that there were no significant differences
between regular and p.r.n. BZ users in anxiety reduction or
panic frequency at the end of treatment and both groups
showed significant improvement on these two measures of
panic disorder symptomatology. However, at post-treatment
assessment, p.r.n. BZ users were significantly more avoidant
than regularly scheduled BZ users and evidenced fewer
reductions in fear of body sensations (i.e., "anxiety
sensitivity" [50]). These naturalistic findings are suggestive
that p.r.n. BZ use may be less desirable throughout
psychotherapy in terms of its association with overall poorer
CBT outcome in the treatment of panic disorder. However,
since the Westra et al. study did not involve random
assignment of patients to different BZ administration groups
[10], it is possible that some pre-existing difference between
p.r.n. and regular BZ users could account for findings of
differential response to CBT for panic disorder. No well-
controlled investigation however has compared the impact of
BZ use alone with fixed or p.r.n. administration, or the
impact of these two manners of BZ use on CBT outcome.
Such studies would appear important in determining the
possible causal status of p.r.n. BZ administration in relation
to poorer CBT outcome [10].
There is a small body of evidence in the animal literature
to support the importance of classical conditioning of
compensatory responses in explaining BZ tolerance (see
review [51]). Specifically, tolerance to BZ’s behavioral
effects has been observed in experiments where BZ
administration was consistently associated with test
apparatus cues (i.e., the analogue of p.r.n. BZ administration
conditions [55,56]). In contrast, no tolerance in the test
situation has been observed, if BZs were administered in the
home-cage (i.e., the analogue of regularly scheduled BZ
administration; [57]). There has been considerably less
research on the role of conditioned compensatory
mechanisms in explaining BZ tolerance in humans.
Nonetheless, the supposition that the development of
tolerance to anxiolytic effects of BZs is more likely under
p.r.n. administration conditions is supported through
examination of the literature on BZ reinforcement in humans.
In these studies, measures of anxiety are routinely included
in order to establish the presence of drug effects.
BZs and Conditioned Drug Effects
Tolerance to a drug’s effect is said to have developed
when a given dose has significantly less of a behavioral effect
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 65
Interestingly, across trials, significant anxiolytic effects of
BZs relative to placebo have been observed only under
regularly-scheduled administration conditions (e.g., [58]) but
not when BZs have been administered under conditions
more closely approximating p.r.n. use in vivo (e.g.,
[59,60]). A possible explanation for this pattern is that the
consistent cues preceding BZ administration under the p.r.n.
sampling situation provides contextual cues that signal the
onset of drug effects. Thus, participants in these latter
experiments had the opportunity to learn compensatory
responses to internal and/or external cues, which
counteracted BZ’s anxiolytic effects.
of possible deleterious effects of on-demand dosing relative to
fixed interval dosing for anxiety control. Moreover, studies
of conditioned drug effects suggest that the effects of BZs
may vary depending on the conditions consistently
preceding drug administration. These suggestive findings
argue for greater systematic investigation of the efficacy of
p.r.n. BZ administration in relation to variables such as type
of BZ, patient population, and chronicity of BZ use.
(2) Is p.r.n. BZ Use Associated With Minimizing BZ
Use?
Clearly, long-term therapeutic BZ use, in those without a
history of drug abuse, is not associated with dose escalation
or loss of control over BZ use, in comparison to other
addictive drugs [63-65]. Beyond this however, minimization
of BZ use is clearly a desirable goal of prescribers and
patients alike. In our physician survey mentioned
previously, many physicians endorsed a belief that p.r.n. use
of BZs may minimize use relative to fixed administration
[17]. A number of investigators concerned with BZ
dependence issues have articulated a similar argument
[60,66]. BZ-using patients also tend to either maintain their
original BZ dose or reduce it over time [67,6]. However
intuitive, the validity of the assumption that as-needed use
produces more minimal use relative to fixed administration,
is a matter of empirical evaluation.
A large body of data support acute anxiolytic effects of
BZs in a variety of controlled clinical trials with anxiety
disorder populations [40,61,62]. Findings from human BZ
reinforcement studies suggest that such anxiolytic effects may
be subject to tolerance with on-demand BZ administration.
However, these studies have been conducted with subjects
with little to no previous BZ experience. Investigations of
the impact of p.r.n. administration on tolerance to BZ
anxiolytic effects would be useful to conduct with chronic
users -- a more typical presenting population in the clinical
context. Additionally, all of the studies reviewed above
(with the exception of [60] which was conducted in vivo)
were conducted in the laboratory when participants were in a
non-stressed state (i.e., no stress induction procedures were
utilized). This raises questions about the generalizability and
ecological validity of the findings for the use of BZs by
anxious individuals in the real world. Moreover, direct
comparison of fixed and p.r.n. BZ use within the same study
would be helpful in elucidating these findings further.
In an uncontrolled study, Winstead et al. found that
when given free access to diazepam on an as-needed basis for
anxiety control [66], psychiatric inpatients tended to use it
infrequently and conservatively. However, the major correlate
of high drug-seeking behaviour in this group was high trait
anxiety. In other words, when diazepam was administered
p.r.n. and regulated by the patient, those with higher levels
of anxiety consumed significantly more diazepam. Although
preliminary, this study suggests that on-demand access is
associated with increased BZ use among those with high
anxiety, and lesser BZ use in those with low anxiety.
In short, literature on conditioned drug effects may be
relevant to explaining possible differences in longer-term
anxiolysis between p.r.n. and regular BZ administration.
Specifically, the development of compensatory responses to
the sedative and anxiolytic effects of BZs may occur if drug
intake is repeatedly preceded by external and/or internal cues
(e.g., feared situations, bodily arousal), as may be operative
with p.r.n. BZ administration. This conditioning paradigm
would suggest that p.r.n. administration may, over time,
become relatively lacking in anxiolytic efficacy. In contrast,
regularly scheduled BZ administration may be relatively
more anxiolytic given the lack of contingency between
context (internal and external anxiety-provoking simuli) and
drug intake.
Perhaps the most well-established literature relevant to
the issue of manner of BZ use and consumption rates,
derives from investigations on the reinforcement potential of
BZs. These studies have utilized varying manners of BZ
administration. Here we compare across these studies to
determine whether sampling conditions mimicking
regularly-scheduled vs. p.r.n. administration differ in terms
of the observed reinforcement potential of BZs. The principal
outcome variable in these studies is drug preference, which is
presumed to serve as an index of a drug’s addictive potential
[59,60]. These investigations generally utilize a paradigm
whereby subjects are given a controlled number of sampling
trials of placebo and BZ. Following these initial sampling
trials, drug reinforcement potential is determined by
assessing participant’s preference for active medication or
placebo in a number of subsequent choice test trials.
Summary of Symptom Control and p.r.n. BZ Use
One might intuitively assume that since patient
satisfaction is greater with as-needed medication
administration (BZs and analgesics), this should result in
objectively more effective symptom control and improved
function. However, this assumption has not been
consistently supported in studies comparing the efficacy of
fixed and p.r.n. administration (for both BZs and analgesics).
Findings from the narcotics literature are variable, with some
suggesting equivalent pain control with as-needed
administration and others finding fixed-interval
administration to be superior for symptom control. Although
available data on p.r.n. BZ administration is limited,
findings are also variable and several studies are suggestive
Numerous investigations of this type have determined
that BZs are generally not reinforcing in nonanxious subjects
[64], as they are not significantly “preferred” or chosen over
placebo except among drug abusers [64,68,69]. In particular,
studies with regular, constrained and fixed administration
(i.e., conditions mirroring regularly scheduled BZ use in
66 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
vivo) have generally failed to support BZ reinforcement.
McCracken et al. [70], using a sample of patients with
generalized anxiety disorder diagnosed according to DSM-III
[71] criteria, found that diazepam was not significantly
preferred over placebo following fixed-interval
administration: diazepam was chosen on only 38% of the
choice occasions. DeWit et al. also studied BZ
reinforcement in clinically anxious [58], non-clinically
anxious, and nonanxious subjects sampling diazepam on a
fixed schedule. Similar to the findings of McCracken, et al.,
DeWit et al. [58,70] found that most subjects (anxious and
non-anxious alike) chose diazepam less often than placebo,
indicating no overall preference for diazepam following fixed
schedule sampling opportunities. These findings are
generally taken as evidence that BZs are only weak
reinforcers and consequently that the abuse potential is
minimal in those without a history of drug abuse (at least,
relative to other highly reinforcing drugs such as barbiturates
or amphetamines).
suggest the possibility that different BZ administration
schedules may be associated with differences in BZ
reinforcement potential. To the extent that the reinforcement
paradigm provides a useful indication of a given drug’s
addictive potential, these preliminary findings suggest that
rather than minimizing addictive potential, p.r.n. BZ
administration may be associated with increased addictive
potential of these drugs.
In conclusion, in subjects without histories of drug
abuse, BZs appear more likely to function as reinforcers
when subjects are given control over the time and total dose
administered. Arguably, these latter studies seem to better
simulate the actual p.r.n. drug-taking behaviour of anxious
patients (i.e., unconstrained drug-taking). These findings
cast doubt on the assumption that p.r.n. BZ use minimizes
drug taking relative to fixed interval use. Rather, it appears
that only under conditions of symptom-contingent drug
administration is increased choice of BZs established. In
controlled studies of drug preference, anxious patients
allowed to determine the timing of BZ administration,
choose BZs for administration at significantly higher rates
than those whose drug taking was constrained at fixed
intervals. However, no study has yet investigated whether
fixed-interval and p.r.n. BZ administration during initial
sampling trials produces different levels of preference for BZs
over placebo or higher doses of BZ administration during the
test phase. These questions appear important for future
research, and would be best investigated in a study where the
two BZ administration regimens are directly compared. If the
tentative conclusion that BZ reinforcement potential is
stronger under symptom-contingent p.r.n. sampling
conditions is supported, further research should be conducted
to determine the mechanisms underlying this effect. For
example, does greater anxiolysis at early sampling trials
(prior to development of anxiolytic tolerance) explain the
greater reinforcement potential of BZs when sampled under
p.r.n. conditions?
Although these findings suggest that reinforcement may
not be a significant factor in continued BZ use, these studies
have generally involved medication sampling and choice
sessions administered on a fixed schedule. That is drug
administration and choice were experimentally regulated and
constrained. Interestingly, several studies have provided
evidence of the reinforcement potential of BZ administration,
but only under as-needed administration conditions (i.e.,
drug administration conditions mirroring those involved in
p.r.n. use in vivo). Using a heterogenous group of subjects
with clinical anxiety disorders defined according to DSM-III-
R [72] criteria, Chutuape and deWit [59] found that higher
anxiety levels (i.e., trait anxiety scores and number of
anxiety disorder diagnoses) were related to increased
preference for diazepam over placebo in the subsequent choice
trials. As such, increased anxiety was related to increased
preference for diazepam over placebo with unconstrained
administration.
Roache et al. examined preference for diazepam in a
group of 4 patients with generalized anxiety disorder [73],
without histories of drug abuse. Subjects were given a 1-
week sampling period where they could ingest diazepam or
placebo on an as-needed basis for anxiety. Under these
conditions, 3 out of 4 subjects demonstrated an exclusive
preference for diazepam on subsequent preference testing. A
similar pattern of continued alprazolam administration was
reported by Roache et al. [60] using a group of anxiety
disorder patients diagnosed with either generalized anxiety
disorder (57% of sample) or panic disorder (43% of the
sample) according to DSM-III-R criteria [72]. Patients were
given unconstrained opportunities for initial sampling and
subsequent choice trials, of alprazolam and placebo. In
contrast to the study by Chutuape and DeWit [59] in which
both initial sampling and choice trials occurred in a
laboratory situation, patients in the Roache et al. [60] study
took home the medication capsules which they were
instructed to use “as needed” for the treatment of anxiety
over a two-week period. Daily fluctuations in anxiety within
subjects was positively correlated with increased alprazolam
use in a majority of cases. In both these studies [60,73],
patients developed low stable rates of BZ administration
during the medication choice test phase. These findings
It is difficult to argue logistically that as needed use of
BZs for the clinical management of anxiety results in
objectively greater drug taking than regular use two or three
times a day. However, no study has systematically examined
consumption rates between those patients instructed to
administer their BZ medications regularly vs. those
instructed to use BZ’s in a p.r.n. fashion. Clearly findings
from drug reinforcement studies do indicate that increased
preference for BZs is established with as-needed
administration and is not present with fixed administration.
Although this does not seem to result in out of control use
and dose escalation with p.r.n. use, it may very well have
implications for continued BZ use and ability to
discontinue.
As-needed Narcotic Administration and Consumption
Rates
Again, parallel findings are apparent in the literature on
as-needed narcotic administration. Here, it has not been
routinely found that increased patient satisfaction with on-
demand dosing translates into minimized levels of
medication use, relative to fixed administration. Tsang and
brush reported no group differences in dose between fixed
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 67
administration and patient-controlled analgesia morphine
[37]. Moreover, several studies have observed increased
medication usage with patient-controlled analgesia, despite
comparable pain control between on-demand and fixed
interval administration [26,31]. For example, Forst et al.
reported that a patient-controlled analgesia group required
twice as much medication relative to a fixed administration
post-operative pain therapy group [30]. In a recent review,
Lehmann concluded that although patients generally prefer
self-control over analgesic administration, opiod
consumption is generally higher [26], albeit still within safe
limits, relative to fixed-interval administration regimens.
These findings largely converge with available data on p.r.n.
BZ administration and cast doubt on the assumption that as-
needed administration necessarily produces more minimal
use relative to fixed-administration.
needed treatment of insomnia with a high potency, rapidly
eliminated BZ (triazolam) has been shown to promote
continued as-needed administration after only one or two
nights of initial medication use, relative to placebo [79].
Moreover, BZ withdrawal has been demonstrated to
trigger future drug administration, particularly in high dose
users [7,80,81]. Cappell et al. randomly assigned existing
BZ users to either a gradual BZ tapering group of diazepam
or placebo [80]. They reported significantly greater
“supplementation” (unauthorized use of BZ) in the group
abruptly switched to tapering on placebo compared to a
diazepam tapering group. These findings suggest that
pharmacological deprivation of BZ promotes future p.r.n.
administration to alleviate withdrawal symptoms among BZ
users. Although it is not known whether p.r.n. use to reduce
withdrawal results in poorer ability to discontinue BZs, it
may be reasonable to assume that such supplementation
would promote negative reinforcement. It may also be quite
practical to recommend non-p.r.n. use to patients when they
are attempting to discontinue, particularly with high potency
rapidly eliminated BZs. That is, withdrawal may
inadvertently be precipitated with intermittent, as-needed use
and this may promote a cycle of negative reinforcement with
continued use in order to alleviate aversive withdrawal
symptoms. Controlled studies of these issues are lacking
however.
(3) Is p.r.n. BZ Use Associated With Reduced Addictive
Potential And Enhanced Ability To Discontinue?
BZs are associated with physical dependence,
withdrawal, and discontinuation difficulties, even with
administration of therapeutic doses [74]. The reported
incidence of withdrawal symptoms upon discontinuation
among chronic BZ users has been estimated at between 40%
and 100% [75]. BZ withdrawal symptoms are characterized
by anticipatory anxiety, increased avoidance and intensified
physical symptoms of anxiety which can be as severe as (or
more severe than) the original symptoms for which BZs were
prescribed [67,76]. An increasing number of patients are
presenting to clinics for assistance with BZ discontinuation
[77,78] and studies of long-term BZ users commonly report
that patients have engaged in one or more unsuccessful
discontinuation efforts [18,7].
Bruce et al. [77] found that fear of body sensations (i.e.
“anxiety sensitivity”; [50]) was a significant and exclusive
predictor of relapse to alprazolam use in panic disorder
patients being tapered off this BZ. Those subjects who
continued to have strong fear of body sensations from
baseline to post-taper showed significantly poorer ability to
achieve and maintain BZ abstinence. It is interesting to
relate these findings to our study of panic disorder patients
reviewed earlier [10], which showed that p.r.n. BZ users
demonstrated significantly less change in fear of body
sensations (i.e., anxiety sensitivity) following CBT, relative
to regularly scheduled BZ users. Extrapolating, such
findings suggest that p.r.n. BZ users may be particularly at
risk for BZ discontinuation difficulties. This possibility is
deserving of future research.
As noted previously, many physicians in our recent
survey study endorsed a belief that dependence may be less
likely, and ultimate discontinuation enhanced, with p.r.n.
use of BZs for clinical anxiety management [17].
Accordingly, the naturalistically observed patient shift to
p.r.n. use over time [6] might represent an attempt by
patients to use BZs more minimally and ultimately
discontinue the drug. If the assumption of reduced addiction
potential and facilitated discontinuation with as-needed BZ
administration is accurate, this switch to p.r.n. use should
be desirable and encouraged. Again, controlled studies of
this issue are lacking however.
In short, although it may be intuitively hypothesized that
as-needed use of BZs may enhance discontinuation, there is a
paucity of data from controlled studies to evaluate this
assumption. Rather, existing findings seem to suggest that
as-needed use may detract from BZ discontinuation via
setting up a negative reinforcement cycle and potentially
undermining changes in fear of anxiety-related bodily
sensations needed for discontinuation success. In addition,
recall that BZs are only preferred and chosen for
administration over placebo under p.r.n. dosing conditions
[64]. These findings tentatively suggest that continued BZ
use may be more likely under as-needed administration
conditions. Anecdotally, anxious patients discontinuing BZs
commonly report a fear of discontinuation secondary to a
belief that the drug is responsible for anxiety control. This
belief may be arguably stronger in those with a history of
p.r.n. BZ use given the stronger contingency between
symptom control and medication taking experienced by
p.r.n. relative to regularly scheduled BZ users, at least in
“Negative reinforcement” has been cited as a factor
facilitating dependence on BZs. Negative reinforcement is a
term from the operant conditioning literature that refers to
removal of an aversive state increasing the response that
preceded it. The stronger the contingency between the
behaviour and the desired outcome, the more likely the
behaviour is to be learned and maintained. To extrapolate to
BZ use, future medication taking would be more likely when
a patient’s experience with medication taking results in
detectable removal of high anxiety, as it may with symptom
contingent, p.r.n. use. Although the dampening of anxiety
could also be achieved with regular BZ use, the contingency
between symptom relief and drug taking would presumably
be stronger with p.r.n. use. In support of this contention, as-
68 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
their initial experiences with BZ-taking. If these speculations
are empirically confirmed, this would suggest that clinically,
a switch to regularly scheduled BZ use may facilitate
discontinuation efforts, prior to dose tapering. Since patients
would no longer take BZs in response to acute experiences of
anxiety symptoms, a switch to a regularly scheduled BZ
administration regimen may be more likely to facilitate
tolerating anxious arousal which is needed for successful
discontinuation.
administration in treatment outcome research. A converse
argument regarding the impact of as-needed BZ use on
adjustment is also plausible. Given the empirically validated
tendency of anxious individuals toward external attributions
for positive events [94,95], it could be argued that symptom
relief produced by symptom-contingent BZ use may be
attributed to the medication rather than the individual’s own
efforts. This would arguably detract from a belief in personal
control or coping ability. This effect may be particularly the
case when medication use is repeatedly and strongly
associated with symptom control, as under as-needed as
opposed to fixed-interval dosing conditions.(4) Does p.r.n. BZ Use Have Implications For Cognitive
Factors Associated With Effective Long-Term Anxiety
Control?
Anecdotal support for this hypothesis was reported by
Clinthrone et al. [96] in a study investigating attitudes to
tranquilizers in 395 users. These investigators found that
users were ambivalent about taking tranquilizers. Although
87% of users agreed with the statement “tranquilizers work
very well to make a person more calm and relaxed”, a
significant proportion (i.e., 45%) also endorsed the item
“Using tranquilizers just prevents people from working out
their problems for themselves”. Moreover, 30% of users
reported a perception that “taking tranquilizers is a sign of
weakness”. Finally, 58% of users indicated that
“tranquilizers cause people to lose some control over what
they do”. Moreover, Dammen et al. found that those few BZ
users who did prefer a fixed administration schedule reported
feeling “less guilty” with this mode of medication
administration [22]. In short, although patients clearly derive
some sense of increased security and control through p.r.n.
BZ use, this may possibly be purchased at the price of
lowered self-perception of personal coping ability.
A final consideration of manner of BZ use concerns
potential implications for longer-term control of anxiety. It
has been commonly observed that while BZs appear to
confer good short-term benefit, they are associated with poor
long-term management of anxiety symptoms [48]. Moreover,
the addition of BZs to psychotherapies (e.g. Cognitive
Behavioural Therapy: CBT) associated with effective long-
term anxiety management results in poorly consolidated
treatment gains and increased anxiety relapse potential
[46,47,82,83].
Exposure to fear-provoking stimuli and situations has
been identified as a necessary, essential component of long-
term anxiety control (e.g. [84-86]). Moreover, the efficacy of
exposure has been postulated to be predicated on cognitive
change or the challenging of catastrophic beliefs and other
dysfunctional cognitive processes characteristic of anxiety
disorder patients [25,87,88]. In this next section, we
consider the impact of p.r.n. use of BZs on these cognitive
change processes associated with longer-term positive
adjustment. We consider the impact of manner of BZ use on
self-efficacy or perception of anxiety control, fear of body
sensations and catastrophic cognitions, selective attention to
threat, and learning and memory.
Catastrophic Beliefs
Reductions in catastrophic beliefs have been found to be
important and significant predictors of long-term anxiety
management [93,97,98] and even ability to discontinue BZs
[77]. Reductions in catastrophic beliefs have been
consistently demonstrated as a function of CBT but not BZs
[47,99,100]. For example, in a review of studies of
alprazolam efficacy in the treatment of panic disorder,
Uhlenhuth et al. [100] reported that treatment effects declined
progressively when moving from the physiological to the
more cognitive aspects of the disorder.
Self-efficacy
We have seen that as-needed use of BZs is associated
with an enhanced perception of control over anxiety
symptoms. The establishment of a sense of controllability
over somatic arousal results in less intense somatic
symptoms in panic disorder patients and non-clinical
research participants high in anxiety sensitivity or fear of
body sensations [89,90]. Moreover, among panic disorder
patients, self-efficacy has been found to be an important
predictor of panic and CBT outcome [91-93]. In short,
enhanced confidence in one’s ability to control anxiety
symptoms is associated with successful anxiety
management. As such the perceptions of increased control
reported by p.r.n. BZ users suggests that symptom-
contingent BZ use should be beneficial for anxiety control.
No controlled investigations have examined the impact of
manner of BZ use on catastrophic belief reduction. In the
Westra et al. naturalistic study [10], daily p.r.n. BZ use was
the best predictor of change in fear of body sensations
(anxiety sensitivity) over the course of CBT, even beyond
pre-CBT anxiety sensitivity levels. In this study, p.r.n. use
of BZs was significantly predictive of a lesser degree of
change in anxiety sensitivity from pre- to post-CBT.
Moreover, p.r.n. use, relative to other BZ parameters such as
chronicity, frequency, and dose, was the only significant
predictor of change in anxiety sensitivity. These short-term,
naturalistic findings clearly require replication and
assessment at follow-up intervals, as well as more controlled
experimental scrutiny. While tentative, they suggest that
p.r.n. use of BZs may inhibit changes in catastrophic beliefs
necessary for successful long-term anxiety management in
panic disorder. Theoretically, we argued that p.r.n. use of
The question of whether p.r.n. BZ use promotes self-
efficacy, and consequent positive adjustment, has not been
systematically investigated in treatment outcome research,
however. This issue is further complicated by the failure to
include measures of self-efficacy in clinical drug trials, as
well as by the failure to manipulate manner of BZ
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 69
BZs may fail to promote catastrophic belief changes by
reducing opportunities to gather negative belief challenging
evidence [48]. For example, predictions that panic results in
heart attack or death remain unchallenged if symptoms
interpreted as indicative of these catastrophic possibilities are
quickly controlled by p.r.n. BZ use.
rate). That is, BZs may actually inadvertently promote
attentional bias to bodily arousal symptoms as a prerequisite
for effective as-needed medication use. Clearly these
naturalistic findings are speculative and more controlled
research is needed to evaluate this hypothesis further.
However, most studies on Stroop interference in anxiety
have not controlled for medication use. Moreover, other
paradigms such as cardiac acuity or other measures of
attention to threat and threat hypervigilance are needed. In
using cardiac acuity as an outcome measure, Antony et al.
[111] found that panic disorder patients on anxiolytic
medication (90% on BZs) were actually more accurate in
estimating their own heart rates than unmedicated patients.
However, Antony et al. [111] did not examine cardiac acuity
in relation to BZ regimen (p.r.n. vs. regularly scheduled). If
replicated, the Stewart et al. [109] findings would suggest
that rather than reducing selective attention to bodily arousal
cues, p.r.n. BZ use may actually be associated with
facilitating this anxiety-maintaining factor.
BZs and Selective Attention to Threat
Selective attention to threat cues has been conceptualized
as an anxiety-maintaining factor in clinical anxiety [101]. A
preponderance of evidence indicates that anxious individuals
display an attentional bias for disorder-specific threat
information. Individuals with panic disorder have been found
to selectively attend to physical threat stimuli [102], social
phobia has been associated with preferential attention to
social threat cues [103], and generalized anxiety disorder
patients selectively process cues related to general threat
[104]. Moreover, while conflicting results have emerged,
some studies of successfully treated CBT patients have
demonstrated that this bias toward threatening information is
reduced, relative to less successful treatment responders
[105,106]. Moreover, this bias has not been found to be
reduced following treatment with diazepam [107,108]. These
findings suggest that CBT may attenuate selective attention
to threat, while BZs fail to do so.
Learning and Memory
A final consideration of the impact of manner of BZ use
concerns implications for learning and memory. Recall that
exposure and associated cognitive change are important for
long-term anxiety control. Moreover, the most frequently
cited rationale for p.r.n. BZ use by anxiety patients is
facilitation of exposure [10] and physicians also frequently
endorsed exposure facilitation as a reason underlying their
p.r.n. BZ prescriptions to anxiety disorder patients [17]. It is
of interest then to consider literature relevant to learning and
memory under conditions of BZ administration.
Westra and Stewart [48] postulated that BZs, particularly
p.r.n. use, may either attenuate or interfere with changes in
selective attention to threat. In examining subjects with
clinical anxiety presenting for CBT treatment, Stewart et al.
[109] compared the Stroop performance of anxiety patients
on BZs with that of unmedicated patients. In this commonly
used paradigm for assessing selective attention in anxiety,
subjects are required to name the ink-colour of stimulus
words, while ignoring their semantic meaning [110]. Delays
in colour-naming (or Stroop interference) occur when the
meaning of the word attracts and holds the subjects’
attention, despite their efforts to attend to ink colour. Stewart
et al. [109] covaried differences between BZ users and
medication non-users on age, anxiety severity, and panic
disorder diagnostic status, in all group comparisons. The
results indicated that BZ users showed significantly greater
Stroop interference than unmedicated controls. In fact,
unmedicated patients showed little to no Stroop interference
in this study. Moreover, this effect was specific to physical
threat: BZ medicated patients showed significantly greater
interference for physical threat stimuli (e.g. HEARTBEAT)
compared to unmedicated anxiety patients, but the groups
did not differ in the degree of social threat interference.
Interestingly, greater p.r.n. use of BZs for coping with
anxiety symptoms was positively correlated with greater
physical threat interference, but not social threat interference.
Many animal studies support the concept of state-
dependent learning with BZs. That is, behavioural evidence
of learning that has occurred under the influence of BZs does
not transfer to the non-BZ state [112] and is blocked
following drug withdrawal [113,114]. In using analogue
exposure exercises, BZs have been found to interfere with
developing tolerance for aversive stimuli [115] and to
produce slower rates of behavioural extinction [116]. Some
limited evidence also suggests that state-dependent learning
with BZs occurs in humans (see [25,117]). State-dependent
learning might be particularly relevant to the as-needed use
of BZs. As just one example, crucial learning experiences
(i.e. exposure to feared stimuli) may be more likely to occur
under the influence of BZs if these medications are taken on
an as-needed basis to facilitate exposure, relative to BZ use
independent of exposure activity. According to state-
dependent memory research, the material learned during
exposure exercises in p.r.n. users would then fail to
generalize to future episodes when the patient was not taking
BZ.
Stewart et al. [109] interpreted these findings as
consistent with the p.r.n. BZ enhancement hypothesis of
selective attention [48]. To elaborate, the need for early
identification of physiological threat cues is central to the
effective utilization of BZs, when they are used in an as-
needed manner. Paradoxically, the effect of using
physiological arousal as a cue for BZ administration would
then result in increased hypervigilance or attentiveness to
physical threat (e.g., increased attention to one’s own heart
BZs are also well known to interfere with human
memory abilities (for a more thorough review of cognitive
effects of BZs see [118]). BZs exert significant memory
impairments which are maximal at peak drug blood
concentrations and are not subject to tolerance with repeated
dosing [119]. More specifically, BZs induce anterograde
amnesia (information presented after drug administration is
poorly remembered; [119]). Other cognitive effects of BZs
include paradoxical facilitation of memory for information
70 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
learned prior to drug administration [120,121] and
impairment of ability to use effortful, relative to automatic,
cognitive functions [122] Finally, BZs appear to impair both
explicit (conscious recall) and implicit memory (memory
with no conscious awareness; [123,124]). Westra et al.
found that these BZ-induced memory impairements obtained
from laboratory-based studies may also be present in the
psychotherapeutic context [125]. We examined memory for
psychoeducational material presented in therapy among BZ
using and unmedicated patients with panic disorder. We
found that incidental recall of psychoeducational material was
significantly poorer among BZ users than among
unmedicated clinical anxiety patients undergoing CBT for
anxiety, even when controlling for education, anxiety, and
age.
suggests that anxiolysis may not be achieved under
conditions of as-needed administration but is demonstrated
under regularly scheduled BZ administration. Parallel
findings from the literature on narcotics also suggests that
while patients strongly prefer p.r.n. analgesic administration,
greater symptom control is not consistently apparent with
this mode of administration. In part, conditioned drug effects
may explain the differences in anxiolytic efficacy of these two
types of BZ administration regimens.
Considering hypotheses regarding the facilitation of more
minimal use and greater ease of discontinuation of BZs with
as-needed use, again available data fail to consistently
support these predictions. Rather, highly anxious subjects
administering BZs on an as-needed basis, demonstrate
higher rates of use than those low in anxiety. Studies of drug
preference indicate that enhanced reinforcement potential of
BZs under as-needed dosing conditions may account for
these observations of higher rates of use by anxious
individuals under p.r.n. dosing conditions. While this does
not translate into BZ dose escalation or loss of control over
drug use, it may very well hamper patients’ ability to
discontinue BZs. Findings that BZ withdrawal (which is
arguably more likely with p.r.n. administration) precipitates
continued use may further underscore arguments for regular
BZ use rather than administration at the discretion of users,
in order to minimize the drug’s negative reinforcement
potential.
Poor learning and generalization of learning may be
arguably optimized under conditions of exposure-contingent
or panic-contingent BZ use, relative to regularly scheduled
BZ use. Here, consolidation of learning that one can achieve
personal mastery or control over anxiety may be less likely if
BZs are used prior to exposure. Moreover, impairments in
memory, which are maximal at peak drug concentrations,
may render effective learning less likely when BZs are used
to facilitate mobility. As such, although p.r.n. BZ use may
possibly facilitate non-avoidance, learning from these
exposure experiences may be less likely given the cognitive
impairments induced by BZs. Clearly, these effects on
learning and memory are not absent with regularly scheduled
BZ use, but non-symptom contingent use may produce more
opportunities for effective learning (e.g., encountering feared
situations when BZ levels are not at their peak). Again, it
would be important to investigate these suggestions in
controlled empirical studies manipulating manner of BZ
administration.
Finally, symptom-contingent BZ administration may
inhibit or minimize cognitive changes necessary for effective
long-term anxiety control, to a greater extent than regularly
scheduled BZ use. Use of medication for symptom control
may promote reliance on (and attributions of success to)
external coping factors rather than internal self-control over
symptoms. This may particularly be the case with a history
of p.r.n. BZ use, where the contingency between controlling
symptoms and medication taking is presumably very strong.
In addition, some preliminary naturalistic data on p.r.n. BZ
administration suggests that this type of BZ use may inhibit
reductions in fear of body sensations and paradoxically
promote selective attention to physical threat cues associated
with maintenance of anxiety. Finally, data on state-
dependent learning and memory impairments with BZs may
imply poorer learning from psychotherapy and necessary
exposure to fear-provoking stimuli. These effects may
arguably be maximized among symptom-contingent BZ
users who are more likely to experience cognitive medication
effects during crucial learning activities.
SUMMARY AND CONCLUSIONS
The high frequency of BZ administration on a p.r.n.
basis, by patients and prescribers alike, in and of itself,
merits further consideration of this common parameter of BZ
use. Long-term users of BZs for anxiety control seem to
prefer p.r.n. BZ administration and increasingly switch to
p.r.n. use over time. Moreover, p.r.n. administration for
anxiety disorder management is apparently a common
practice among family physicians, and less so among
psychiatrists. Physicians’ recommendations for, and
patients’ use of, BZs on an as-needed as opposed to
regularly scheduled manner is curious in view of the paucity
of empirical data and prescription guidelines supporting one
or the other manner of BZ use.
Thus, increased subjective anxiety control in the short-
term may be purchased at the price of reduced medication
anxiolytic efficacy in the longer-term, reinforcement of
continued use and poorer long-term anxiety control. This
may be consistent with Pennebaker’s findings that
individuals notice short-term changes in symptom status to
a greater degree than longer-term changes [126]. As such,
individuals may be more likely to establish a belief in the
efficacy of medication, but fail to notice the waning long-
term benefits of medication. This effect may be more likely
when a strong contingency is established between
medication use and symptom control, as with as-needed BZ
use.
Although the rationales endorsed by patients and
prescribers for p.r.n. BZ use have intuitive appeal, empirical
data evaluating these assumptions does not consistently
support this practice. As-needed administration of BZs does
appear to be associated with heightened perceptions of
symptom control among anxiety disorder patients.
Unfortunately however, heightened perceptions of control are
not necessarily associated with objectively superior
anxiolytic effects of BZs, relative to regular administration
independent of symptoms. In fact, a review of the literature
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 71
These studies from disparate literatures are merely
suggestive of possible deleterious effects of BZ p.r.n.
administration. Controlled studies of this variable are
lacking and many opportunities exist for more systematic
investigation. At minimum, the findings of differential effects
between different manners of BZ administration suggest that
this variable should be systematically assessed in studies of
BZ effects. More importantly, controlled studies
manipulating and controlling manner of BZ use are required
in relation to BZ efficacy, discontinuation, and cognitive
factors associated with effective long-term anxiety control.
Moreover, it would be of interest to investigate any obtained
differences with manner of BZ administration across types of
BZs, duration of medication use, dose, and different patient
populations. It may be for example that high doses of
rapidly-acting BZs are more detrimental for p.r.n.
administration than lower doses of slower-release BZs. This
suggestion is consistent with studies demonstra-ting greater
reinforcement potential and greater liking of fast-release BZs
(e.g. [64]). Future studies along the lines suggested appear
particularly important given the commonality of p.r.n. BZ
use among anxiety disorder patients, inconsistent physician
prescription practices regarding manner of BZ
administration, and the high incidence of chronic BZ use for
anxiety disorder control in general. Delineation of the role of
this potentially important medication use parameter may
also have practical and clinical implications which may
assist clinicians in navigating BZ discontinuation efforts and
promoting more successful long-term anxiety management.
[8] Canadian Pharmacists Association.. Compendium of
Pharmaceuticals and Specialties, 34
th
Edition. Ottawa,
ON: Author. 1999.
[9] Gorman, J.M. Journal of Clinical Psychiatry, 1997, 58,
54-58.
[10] Westra, H.A., Stewart, S. H. Current Pharmaceutical
Design, 2002, 8, 77-97.
[11] Barlow, D.H. Journal of Clinical Psychiatry, 1997, 58,
37-38.
[12] Podhorna, J. Current Pharmaceutical Design, 2002, 8,
29-55.
[13] Otto, M. W., Hong, J. J., Safren, S. A. Current
Pharmaceutical Design, 2002, 8, 99-106.
[14] Goldberg, C. Psychiatry Quarterly, 1998, 69, 23-44.
[15] Gould, R.A., Otto, M.W., Pollack, M.H. Clinical
Psychology Review, 1996, 15, 819-844.
[16] Passnau, R.O. Psychiatry and Clinical Neuroscience,
1998, 52, S199-202.
[17] Westra, H.A. Stewart, S.H. Survey of benzodiazepine
prescription practices among family physicians and
psychiatrists in Nova Scotia. Paper presented at the
Annual Dalhousie University Psychiatry Research Day,
Halifax. 2000.
[18] Romach, M.K., Busto, U.E., Sobell, L.C., Sobell, M.B.,
Somer, G.R., Sellers, E.M. Psychopharmacology
Bulletin, 1991, 27, 391-395.
ACKNOWLEDGEMENTS
[19] Stewart, S.H. Westra, H.A. Motivations for as needed
benzodiazepine use in panic disorder patients. Paper
presented at the Annual Meeting of the Association for
the Advancement of Behaviour Therapy, New Orleans.
2000.
The authors wish to thank Dr. Patricia Conrod for her
helpful comments on an earlier version of this manuscript.
This project was funded through research grants from the
Queen Elizabeth II Health Sciences Centre and the Dalhousie
University, Department of Psychiatry Research Fund.
[20] Nisbett, R.E., Wilson, T.D. Psychological Review, 1977,
84, 231-259.
REFERENCES
[21] Sobell, L.C., Sobell, M.B. (Eds.). Behavioral
Assessment, 1990, 12(1).
[1] Boixet, M., Batlle, E., Bolibar, I. Addiction, 1996, 91,
549-556.
[22] Dammen, T., Haug, T., Gotestam, K.G. European Journal
of Psychiatry, 1994,8, 227-241.
[2] Uhlenhuth, E.H., Balter, M.B., Ban, T.A., Yang, K.
Journal of Clinical Psychopharmacology, 1998, 18,
27S-31S.
[23] Hecker, J.E., Fink, C.M., Fritzler, B.K. Journal of Anxiety
Disorders, 1993, 7, 373-384.
[3] Mant, A., Mattick, R.P., de Burgh, S., Donnelly, N., Hall,
W. Family Practice, 1995, 12, 37-43.
[24] Mellinger, G.D., Balter, M.B., Uhlenhuth, E.H. Journal
of the American Medical Association, 1984, 251, 375-
379.
[4] Balter, M.B., Ban, T.A.,Uhlenhuth, E.H. Human
Psychopharmacology, 1993, 8, 253-261.
[25] Barlow, D.H. Anxiety and its disorders: The nature and
treatment of anxiety and panic. New York, NY: Guilford.
1988.
[5] Uhlenhuth, E.H., Balter, M.B., Ban, T.A., Yang, K.
Journal of Affective Disorders, 1995, 35, 153-162.
[26] Lehmann, K.A. European Surgical Research, 1999, 31,
112-121.
[6] Romach, M., Busto, U., Somer, G., Kaplan, H.L., Sellers,
E. American Journal of Psychiatry, 1995, 152, 1161-
1167.
[27] Lindley, C. Support Care Cancer, 1994, 2, 319-326.
[7] Sellers, E.M., Somer, G.R, Sobell, L., Sobell, M. Clinical
Pharmacology and Therapeutics, 1990, PII-81, 168.
[28] Etches, R.C. Surgical Clinics of North America, 1999,
79, 297-312.
72 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
[29] Chumbley, G.M., Hall, G.M., Salmon, P. Anaesthesia,
1999, 54, 386-389.
[50] Reiss, S., Peterson, R.A., Gursky, D.M., McNally, R.J.
Behaviour Research and Therapy, 1986, 24, 1-8.
[30] Forst, J., Wolff, S., Thamn, P., Forst, R. Archives of
Orthopedic Trauma Surgery, 1999, 119, 267-270.
[51] File, S. E. Neuroscience and Biobehavioral Reviews,
1985, 9, 113-121.
[31] Knudsen, W.P., Boettcher, R., Vollmer, W.M., Griggs,
D.K. Hospital Pharmacology, 1993, 28, 117-122.
[52] Bateson, A.N. Current Pharmaceutical Design, 2002, 8,
5-27.
[32] Ripamonti, C. Bruera, E. Oncology (Huntingt), 1997, 11,
373-380.
[53] Goudie, A.J. Demellweek, C. Conditioning factors in
drug tolerance. In S.R. Goldberg I.P. Stolerman (Eds.),
Behavioural analysis of drug dependence (pp. 225-
285). New York: Academic Press. 1986.
[33] Chapman, C.R. Acta Anaesthesiology Belgium, 1992, 43,
41-52.
[54] Siegel, S. In F.C. Colpaert J.L. Slangen (Eds.), Drug
discrimination: Application in CNS pharmacology (pp.
287-302). Amsterdam: Elsevier Biomedical Press.1982.
[34] Gust, R., Pecher, S., Gust, A., Hoffmann, V., Bohrer, H.,
Martin, E. Critical Care Medicine, 1999, 27, 18-23.
[35] DeConno, F., Ripamonti, C., Gamba, A., Prada, A.,
Ventafridda, V. European Journal of Surgical
Oncology, 1989, 15, 242-246.
[55] Cannizzaro, G., Nigito, S., Provenzao, P. M., Vitikova, T.
Psychopharmacologia, 1972, 26, 172-184.
[56] Cook, L., Sepinwall, J. Behavioral analysis of the effects
and mechanisms of action of benzodiazepines. In E. Costa
P. Greengard (Eds.), Mechanisms of action of
benzodiazepines (pp. 1-28). New York: Raven Press.
1975.
[36] Berntzen, D. Gotestam, K.G. Journal of Consulting and
Clinical Psychology, 1987, 55, 213-217.
[37] Tsang, J. Brush, B. Anaesthiology in Intensive Care,
1999, 27, 464-470.
[57] Sepinwall, J., Cook L. In L. L. Iversen, S. D. Iversen, S. H.
Snyder (Eds.), Handbook of psychopharmacology, Vol.
13: Biology of mood and anxiety drugs (pp. 345-393).
London, UK: Plenum Press. 1978.
[38] Rachman, S. Behaviour Research and Therapy, 1984, 22,
59-70.
[39] Woody, S. Rachman, S. Clinical Psychology Review,
1994, 14, 743-753.
[58] deWit, H., Uhlenhuth, E.H., Hedeker, D., McCracken,
S.G., Johanson, C.E. Archives of General Psychiatry,
1986, 43, 533-539.
[40] Ballenger, J.C., Burrows, G.D., DuPont, R.L., Lesser,
I.M., Noyes, R. Pecknold, J.C., Rifkin, A., Swinson, R. P.
Archives of General Psychiatry, 1988, 45, 413-422.
[59] Chutuape, M.D. deWit, H. Psychopharmacology, 1995,
121, 91-103.
[41] Charney, D.S., Woods, S.W., Goodman, W.K., Rifkin, B.,
Kinch, M., Aiken, B., Quadrino, L.M., Heninger, G.R.
Journal of Clinical Psychiatry, 1986, 47, 580-586.
[60] Roache, J.D., Stanley, M.A., Creson, D.L., Shah, N.N.,
Meisch, R.A. Drug and Alcohol Dependence, 1997, 45,
143-155.
[42] Greenblatt, D.J., Shader, R.I., Abernethy, D.R. Journal of
Medicine, 1983, 309, 354-358.
[61] Dunner, D.L., Ishiki, D., Avery, D.H., Wilson, L.G.,
Hyde, T.S. Journal of Clinical Psychiatry, 1986, 47,
458-460.
[43] Jefferson, J.W. Journal of Clinical Psychiatry, 1995, 56,
18-24.
[62] Noyes, R., Jr., Garvey, M.J., Cook, B., Suelzer, M.
American Journal of Psychiatry, 1991, 148, 517-523.
[44] Marshall, R.D., Schneier, F.R., Fallon, B.A., Feerick, J.,
Liebowitz, M.R. Journal of Clinical Psychiatry, 1994,
55, 33-37.
[63] Cole, J.O. Chiarello, R.J. Journal of Psychiatric
Research, 1990, 24, 135-144.
[45] Rickels, K., Case, G., Downing, R.W., Winokur, A.
Journal of the American Medical Association, 1983,
250, 767-772.
[64] Griffiths, R.R. Weerts, E.M. Psychopharmacology, 1997,
134, 1-37.
[46] Echeburua, E., De Corral, P., Bajos, E.G., Borda, M.
Behavioural and Cognitive Psychotherapy, 1993, 21,
219-238.
[65] Woods, J.H, Katz, J.L., Winger, G. Pharmacological
Reviews, 1987, 39, 251-390.
[66] Winstead, D.K., Anderson, A., Eilers, M.K., Blackwell,
B., Zaremba, A.L. Archives of General Psychiatry, 1974,
30, 349-351.
[47] Marks, I. M., Swinson, R. P., Basoglu, M., Kuch, K.,
Noshirvani, H., O'Sullivan, G., Lelliott, P. T., Kirby, M.,
McNamee, G., Sengun, S., Wickwire, K. British Journal
of Psychiatry, 1993, 162, 776-787.
[67] Burrows, G.D, Norman, T.R., Judd, F.K., Marriott, P.F.
Journal of Psychiatric Research, 1990, 2, 65-72.
[48] Westra, H.A. Stewart, S.H. Clinical Psychology Review,
1998, 18, 307-340.
[68] Roach, J.D. Griffiths, R.R. Abuse liability of anxiolytics
and sedative/hypnotics: Methods assessing the
likelihood of abuse. In N.K. Mello M.W. Fischman (Eds),
Testing for abuse liability of drugs in humans (pp 123-
[49] Schweizer, E., Clary, C., Dever, A.I., Mandos, L.A.
Journal of Clinical Psychiatry, 1992, 53, 19-22.
Incidence and Implications Current Pharmaceutical Design, 2002, Vol. 8, No. 1 73
146). NIDA Research Monograph No. 92. Washington,
DC: U.S. government printing office. 1989.
[89] Sanderson, W.C., Rapee, R.M., Barlow, D.H. Archives of
General Psychiatry, 1989, 46, 157-162.
[69] Roache, J.D., Meisch, R.A., Henningfield, J.E., Jaffe, J.H.,
Klein, S., Sampson, A. Pharmacology, Biochemistry,
and Behaviour, 1995, 50, 171-179.
[90] Telch, M.J., Silverman, A., Schmidt, N.B. Journal of
Anxiety Disorders, 1996, 10, 21-36.
[91] Borden, J.W., Clum, G.A., Salmon, P.G. Cognitive
Therapy and Research, 1991, 15, 257-272.[70] McCracken, S.G., deWit, H., Uhlenhuth, E.H., Johanson,
C.E. Journal of Clinical Psychopharmacology, 1990,
10, 190-196.
[92] Williams, S.L., Kinney, P.J., Falbo, J. Journal of
Consulting and Clinical Psychology, 1989, 57, 436-
442.[71] American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders, 3
rd
Edition
(DSM-III). Washington, DC: Author. 1980.
[93] Williams, S.L. Falbo, J. Behaviour Research and
Therapy, 1996, 34, 253-263.
[72] American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders, 3
rd
Edition –
Revised (DSM-III-R). Washington, DC: Author. 1987.
[94] Cloitre, M., Heimberg, R.G., Liebowitz, M.R., Gitow, A.
Cognitive Therapy and Research, 1992, 16, 569-577.
[95] Hoffart, A. Martinsen, E.W. Cognitive Therapy and
Research, 1990, 14, 343-351.
[73] Roache, J.D., Stanley, M.A., Creson, D.L., Meisch, R.A.,
Shah, N.N. Experimental and Clinical
Psychopharmacology, 1996, 4, 308-314.
[96] Clinthorne, J.K., Cisin, I.H., Balter, M.B., Mellinger,
G.D., Uhlenhuth, E.H. Archives of General Psychiatry,
1986, 43, 527-532.
[74] Michelini, S., Cassano, G.B., Frare, F., Perugi, G.
Pharmacopsychiatry, 1996, 29, 127-134.
[97] Butler, G. Behaviour Research and Therapy, 1993, 31,
211-213.
[75] Rickels, K.E., Schweizer, W., Case, G., Greenblatt, D.
Archives of General Psychiatry, 1990, 47, 899-907.
[98] Hegel, M.T., Ravaris, C.L., Ahles, T.A. Behaviour
Therapy, 1994, 25, 183-195.
[76] Fyrer, A.J., Liebowitz, M.R., Gorman, J.M. American
Journal of Psychiatry, 1987, 144, 303-308.
[99] Mattick, R.P., Andrews, G., Hadzi-Pavlovic, D.,
Christensen, H. Journal of Nervous and Mental
Disease, 1990, 178, 567-576.
[77] Bruce, T.J., Spiegel, D.A., Gregg, S.F., Nuzzarello, A.
American Journal of Psychiatry, 1995, 152, 1156-1160.
[78] Spiegel, D.A., Bruce, T.J., Gregg, S.F., Nuzzarello, A.
American Journal of Psychiatry, 1994, 151, 876-881.
[100] Uhlenhuth, E.H., Matuzas, W., Warner, T.D., Thompson,
P.M. Psychopharmacology Bulletin, 1997, 33, 31-39.
[79] Kales, A., Manfredi, R.L., Vgontzas, A.N., Bixler, E.O.,
Vela-Bueno, A., Fee, E.C. Clinical Pharmacological
Therapy, 1991, 49, 468-476.
[101] McNally, R.J. Behaviour Research and Therapy, 1995,
33, 747-754.
[102] Ehlers, A., Margraf, J., Davies, S., Roth, W. T. Cognition
and Emotion, 1988, 2(3), 201-219.
[80] Cappell, H., Busto, U., Kay, G., Naranjo, C.A., Sellers,
E.M., Sanchez-Craig, M. Psychopharmacology, 1987,
91, 154-160.
[103] Hope, D.A., Rapee, R.M., Heimberg, R.G., Dombeck, M.J.
Cognitive Therapy and Research, 1990, 14, 177-189.
[81] Romach, M.K., Somer, G., Sobell, L., Sobell, M., Sellers,
E.M. Clinical Investigations in Medicine, 1990, 13,
B14.
[104] Mathews, A., Mogg, K., Kentish, J., Eysenck, M.
Behaviour Research and Therapy, 1995, 33, 293-303.
[82] Clum, G.A., Clum, G.A., Surls, R. Journal of Consulting
and Clinical Psychology, 1993, 61, 317-326.
[105] Mathews, A. MacLeod, C. Behaviour Research and
Therapy, 1985, 23(5), 563-569.
[83] Wardle, J., Hayward, P., Higgitt, A., Stabl, M., Blizard,
R., Gray, J. Behaviour Research and Therapy, 1994, 32,
203-215.
[106] Watts, F.N., McKenna, F.P., Sharrock, R., Tresize, L.
British Journal of Psychology, 1986, 77, 97-108.
[107] Golombok, S., Mathews, A., MacLeod, C., Lader, M.
Human Psychopharmacology Clinical and
Experimental, 1990, 5, 143-147.
[84] Hope, D.A., Heimberg, R.G., Bruch, M.A. Behaviour
Research and Therapy, 1990, 33, 637-650.
[85] Rachman, S., Wilson, G. The effects of psychological
therapy. Oxford: Permagon Press. 1980.
[108] Golombok, S., Stavrou, A., Bonn, J., Mogg, K.,
Critchlow, S., Rust, J. Cognitive Therapy and Research,
1991, 15, 459-467.
[86] Taylor, S. Behaviour Therapy and Experimental
Psychiatry, 1996, 27, 1-9.
[109] Stewart, S.H., Westra, H.A., Thompson, C.E., Conrad,
B.E. Cognitive Therapy and Research, 2000, 24, 67-85.
[87] Butler, G. Behaviour Research and Therapy, 1985, 23,
651-657.
[110] Stroop, J.R. Journal of Experimental Psychology, 1935,
18, 643-661.
[88] Mattick, R.P., Peters, L., Clarke, J.C. Behaviour Therapy,
1989, 20, 3-23.
74 Current Pharmaceutical Design, 2002, Vol. 8, No. 1 Westra
and Stewart
[111] Antony, M. M., Brown, T.A., Craske, M.G., Barlow, D.H.,
Mitchell, W.B., Meadows, E.A. Journal of Anxiety
Disorders, 1995, 9, 355-371.
[119] Curran, H.V. Psychopharmacology, 1991, 105, 1-8.
[120] McGaugh, J.L. Psychological Science, 1990, 1, 15-25.
[121] McGaugh, J.L. Affect, neuromodulatory systems and
memory storage. In S. A. Christianson (Ed.), The
handbook of emotion and memory research and theory
(pp.245-269). Hillsdale, NJ: Erlbaum. 1992.
[112] Wardle, J. British Journal of Psychiatry, 1990, 156,
163-168.
[113] Gray, J.A. Journal of Comparative and Physiological
Psychology, 1969, 69, 55-64.
[122] Weingartner, H.J., Hommer, D., Johnson, D., Andreason,
P., Eckardt, M. J. Alcohol Alcoholism, 1994, 2, 291-298.[114] McNaughton, N. Quarterly Journal of Experimental
Psychology, 1984, 36B, 319-330.
[123] Curran, H.V., Gorenstein, C. International Clinical
Psychopharmacology, 1993, 8, 37-42.[115] Gray, J.A. Interaction between drugs and behaviour
therapy. In H. J. Eysenck I. Martin (Eds.), Theoretical
foundations of behaviour therapy (pp. 433-447). New
York, NY: Plenum Press. 1987.
[124] Stewart, S.H., Rioux, G.F., Connolly, J.F., Dunphy, S.C.,
Teehan, M.D. Psychopharmacology, 1996, 128, 139-
149.
[116] Gorman, J.E., Dyak, J.D., Reid, L.D. Bulletin of the
Psychonomic Society, 1979, 14, 46-48.
[125] Westra, H.A., Stewart, S.H., Teehan, M.T., Johl, K.,
Dozois, D.J.A., Hill, T. Benzodiazepine use and memory
for psychotherapy information in clinical anxiety.
Paper presented at the World Congress of Cognitive and
Behavioural Therapy, Vancouver. 2001.
[117] Roy-Byrne, P.P. Swinson, R.P.L. In D. S. Crowley (Ed.)
Benzodiazepines in clinical practice: Risks and
benefits (pp. 202-211). Washington, DC: American
Psychiatric Press. 1991.
[126] Pennebaker, J.W. The psychology of physical symptoms.
New York: Springer-Verlag. 1982.[118] Buffett-Jerrott, S.E., Stewart, S.H. Current
Pharmaceutical Design, 2002, 8, 57-75.
... The study of the effects of substance administration regimens has proven useful in an older yet related area of research pertaining to the use of benzodiazepines in individuals with anxiety and related disorders (Busto & Sellers, 1986;Westra & Stewart, 2002a;Westra & Stewart, 2002b). In this literature there is evidence suggesting that a "pro re nata" ("PRN") regimen, or as needed use, in response to the occurrence of anxiety-related symptoms is more This preprint research paper has not been peer reviewed. ...
... Consistent with patterns revealed in benzodiazepine regimen research (Westra & Stewart, 2002a;2002b;Romach et al., 1991), and with H1, PRN regimens were much more common than an RS only regimen at the time of testing. In terms of changes in use regimen over time, many This preprint research paper has not been peer reviewed. ...
... Movement from PRN related regimens to an RS only regimen was rare: 9.7% of those who were originally PRN only users and 8.3% of those who were originally PRN+ users moved to be RS only users. As in previous findings with benzodiazepines in individuals with anxiety (Westra & Stewart, 2002a;2002b), large proportions of individuals in the sample who began as PRN remained as such (52.7%). However, unexpectedly, many PRN-only users switched to PRN+ use (37.6%). ...
... All rights reserved 1) Benzodiazepines are usually taken as needed, which may decrease drug levels leading to fewer reactions. 23 Furthermore, as discontinuation resolves photosensitivity reaction, testing may not be commonly done in clinical practice -leading to underreporting of photosensitivity. ...
... 3) Photoallergic reactions may resemble drug eruptions; thus, a patch test and oral photoprovocation test may be necessary to distinguish between a drug eruption and a photosensitivity reaction. 23 There are case reports of patients having skin eruptions likely due to benzodiazepines, however, two patients had negative patch tests and no follow-up photopatch or oral photoprovocation testing. 25,26 Lack of photosensitivity testing may lead to misclassification of photosensitivity as allergic skin reaction. ...
Article
Objective Benzodiazepines have been reported to cause photosensitivity reactions. We characterized the clinical presentation and diagnosis of benzodiazepine‐associated photosensitivity and adjudicated these cases for a causal association with benzodiazepines. Methods A literature search on PubMed’s “MeSH” search feature and CINAHL (1964 to 2019) was performed using search terms: benzodiazepine, photosensitivity, and photosensitivity disorders/chemically induced. We applied the Naranjo scale, a standardized causality assessment algorithm, to identified cases. Results We identified 8 published cases, with 50% of patients being female with a mean age of 46.3 years. Alprazolam, tetrazepam, clobazam, and clorazepate induced phototoxic reactions. Chlordiazepoxide induced one photoallergic reaction. Photosensitivity occurred between 1‐3 days (37.5%), 7‐14 days (25%), and >14 days (25%). Photosensitivity resolved after drug discontinuation within 2 weeks (62.5%). Benzodiazepine rechallenge confirmed photosensitivity in 75% of cases. Photopatch testing was negative in two patients; however, these patients had positive oral provocation testing. However, an oral photoprovocation test, an ideal diagnostic test, was not administered to several patients. Despite these challenges, the Naranjo scale identified 5 cases as definite benzodiazepine‐induced photosensitivity. Conclusion Five benzodiazepines induced photosensitivity reactions. Five patients showed a definite association with the Naranjo scale. Reporting to pharmacovigilance databases may help identify other benzodiazepines causing photosensitivity reactions.
... Long-term treatments with benzodiazepines are limited by the development of tolerance to the majority of their pharmacological actions and by the development of dependence (von Moltke and Greenblatt, 2003;Westra and Stewart, 2002). The development of tolerance to each pharmacological effect follows different temporal courses, suggesting the existence of multiple mechanisms. ...
Article
GABAA receptors mediate most of the fast inhibitory transmissions in the central nervous system. These receptors are pentameric complexes that exhibit high structural and pharmacological heterogeneity, as they can be constructed from 19 distinct subunits. GABAA receptors are the targets of numerous clinically relevant drugs used to treat various disorders such as anxiety, insomnia and epilepsy. These receptors are also the targets of many volatile anesthetics and drugs of abuse, such as alcohol. This review is focused on the effect of long-term treatment with GABA, and the positive allosteric modulators benzodiazepines, neurosteroids and ethanol on GABAA receptors. Prolonged exposure of GABAA receptors to these compounds triggers several adaptive mechanisms that lead to changes in the structure, function and localization of receptors. These changes include GABAA receptor subunit expression, intracellular trafficking and phosphorylation. These adaptations are relevant to different physiological, pathological and pharmacological conditions and, in most cases, are associated with the development of tolerance. Understanding the molecular mechanisms underlying these regulatory processes will be relevant for therapeutic benefits.
... [12] Most typically the benzodiazepines are seen to have value as 'rescue' agents, specifically alprazolam (including the extended-release formulation), [13] in spite of a lack of evidence-based data supporting such use, notwithstanding clinician and patient support and preference for such use. [14] ...
Article
Full-text available
div style="left: 70.8662px; top: 324.72px; font-size: 15.45px; font-family: serif; transform: scaleX(1.05793);" data-canvas-width="422.862">Panic disorder (PD) is a prevalent anxiety disorder with lifetime prevalence rates ranging from 1.1% to 3.7% in the general population and 3.0% to 8.3% in clinic settings. [1] The presence of agoraphobia in patients with PD is associated with substantial severity, comorbidity (e.g. major depression, other anxiety disorders, alcohol abuse) and functional impairment. [1]</div
... Moreover, as-needed use leads to daily benzodiazepine consumption. 21 A regular time-fixed dosing scheme that prevents anxiety symptoms is preferable to trying to reduce symptoms once they occur. ...
Article
Benzodiazepine receptor agonists (BZDs) should be appropriately used owing to the associated risks of delirium and falls. Since January 2018, the liaison team pharmacist at Iizuka Hospital has been applying digital labels with recommendations for the reduction of use and changes in the medication orders and prescriptions of BZDs on electronic medical records of patients in the surgical ward. This study aimed to verify the effectiveness of reducing the use of BZDs via the implementation of digital labels. Patients in the surgical ward were retrospectively assessed for changes in medication orders and prescription ratios of BZDs before and after the implementation of digital labels. The ratio of the number of digital labels implemented to the number of confirmations of medication orders and prescriptions of BZDs was 15.0% at the start of implementation; however, the ratio gradually and significantly decreased to 3.6%. The medication order ratio of BZDs was 52.2% before the implementation of digital labels; however, this ratio decreased to 2.7% and 5.6% immediately and 4 months after the implementation of digital labels, respectively. The present study showed that medication orders for BZDs were reduced after the implementation of digital labels and that the reduction effect was maintained for a certain period of time. Thus, the liaison team pharmacist-led approach can contribute to the proper use of BZDs.
Thesis
Full-text available
This thesis is presented as part of the PhD in Medical Sciences of the University of Liège (ULg). It provides an overview of the use of benzodiazepine hypnotics and anxiolytics, their likelihood of progression to high-dose dependence and their mortality risk. The general section summarizes the indications of benzodiazepines, their types and their characteristics. It reviews the diagnostic features of acute intoxication and provides a clinical classification of the problematic use of benzodiazepines. It also describes the withdrawal of these substances and summarizes the criticisms of the recommendation to move from benzodiazepines to selective serotonin reuptake inhibitors in the treatment of anxiety disorders. The empirical section includes epidemiological data and proposes a definition of "high-dose use". It raises the hypothesis that differences exist between benzodiazepines, especially between anxiolytic and hypnotic benzodiazepines. It presents the results of research done in relation to this thesis. It also outlines the preliminary results of mortality from all-cause associated with benzodiazepines and discusses the likely biases. The discussion debates the findings and actions to be taken on the basis of the results of this thesis. It summarizes the preventive measures already in place and sets out concrete recommendations. It suggests studies to be undertaken in the future.
Chapter
We developed this clonazepam guideline using drug prescribing information and reviewing the available literature on relevant neuropsychiatric ­disorders in populations without intellectual disabilities because of the dearth of available literature on the population with intellectual disabilities. This guideline includes indications; contraindications; assessments prior to and during treatment; dosing with particular focus on dosing modifications required by drug–drug interactions, personal characteristics, or genetic variants; and adverse drug reactions. The procedures contained in this guideline may not fully account for all of the possible risks of treatment in this population because of the limited studies available; thus, there will be a need to periodically update this guideline as new information becomes available. Nevertheless, we believe that this guideline provides a useful resource for clinicians who treat epilepsy, anxiety, or catatonia in adult individuals with intellectual disabilities. A clonazepam drug utilization review that summarizes this guideline is described.
Article
Full-text available
The reinforcing effects of diazepam were examined in 4 female patients with generalized anxiety disorder but without histories of drug abuse. Under double-blind conditions, patients were given medication capsules containing either 4 mg diazepam or placebo that they could use "as needed" for anxiety. After a 1-week sampling period for each medication, participants repeatedly chose which medication they preferred over the next 4 weeks. Reinforcing effects of diazepam were demonstrated in 3 participants who showed exclusive diazepam preference. The results suggest procedures whereby drug self-medication behavior may be studied in an outpatient environment. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Article
Full-text available
The present study examined the effects of naturalistic benzodiazepine (BZ) use on selective attention to threat cues in 50 patients diagnosed with anxiety disorders, according to DSM-IV (APA, 1994) criteria. Patients provided information on their BZ use histories, demographics, and severity of anxiety symptomatology, and completed a computerized Stroop task involving color naming of social threat, physical threat, and matched no-threat control words. Patients were selected to fill two age-, gender-, and diagnosis-matched groups based on self-reported BZ use histories: 25 current BZ users versus 25 medication nonusing controls. Planned comparisons were conducted to determine whether BZ use groups differed in degree of selective attention to either the physical and/or social threat stimuli, or overall. Even with BZ use group differences in anxiety severity covaried out, the BZ users demonstrated significantly greater selective attention to threat than the medication nonusers, particularly in the case of physical threat stimuli. These findings are consistent with Westra and Stewart's (1998) suggestion that BZ use may increase preferential attention to physical threat cues, since BZs are often taken on an as needed (prn) basis. This prn enhancement interpretation was further supported through the finding of a significant positive correlation between frequency of prn use of BZs and degree of physical threat-related interference on the Stroop among the BZ users group. Theoretical explanations and clinical implications of these findings are discussed.
Article
A between-group cross-over design was used in th assessment of the effects of two different schedules of administration in benzodiazepine treatment of generalized anxiety. Several rating scales were used to gather information on symptoms (DSM-III Axis I) as well as personality characteristics (DSM-III Axis II). Despite non-significant quantitative differences found on symptom measures between the different schedules of administration, most of the patients subjectively preferred an on-demand schedule of self administration, while a minority was indifferent to schedules. These data suggest a discrepancy between objectively defined effects and subjective preferences, the latter probably governed by individual coping strategies, drug discrimination learning as well as public attitudes and doctor's prescription practice.
Article
The discovery and introduction of valuable new antianxiety agents into clinical psychotherapeutics within the last 20 years was accompanied by a parallel advance in the development of behavioral pharmacology. Behavioral methodology helped to characterize these new agents. At the same time, the interesting and varied properties of these compounds stimulated the development of more comprehensive behavioral techniques so that possible subtle differences among the compounds could be evaluated. In this chapter, we shall review behavioral studies in animals that may shed light upon the therapeutic actions of benzodiazepines and other antianxiety drugs. Although many pharmacologic techniques have been used to study the central nervous system actions of antianxiety agents, e.g., antagonism of pentylene-tetrazol-induced convulsions or footschock-induced fighting, most of the material to be reviewed will involve operant conditioning techniques. The advantages of operant methods, including the use of animals as their own controls because of the availability of stable baselines in trained animals, have been described previously (Cook and Sepinwall, 1975a,b). Some material will be included, however, to indicate the Value of certain nonoperant behavioral procedures for studying anxiolytic compounds.
Article
• Using a choice procedure, these experiments tested whether diazepam is more highly preferred by anxious subjects than by normal control subjects. Subjects first sampled and then chose between two capsules containing diazepam (5 or 10 mg) and placebo, or amphetamine (5 mg) and placebo. The number of times each drug was chosen over placebo and the subjective effects of the drugs were measured. Anxious subjects did not differ from controls in their drug choices. Most subjects chose diazepam less often than placebo, especially at the higher dose, whereas they chose amphetamine more often than placebo. The subjective drug effects (including anxiety reduction after diazepam) were similar for anxious and nonanxious subjects, despite predrug differences in anxiety. The results suggest that individuals with high anxiety are not at greater risk for dependence on antianxiety drugs.
Book
Physical symptoms are fascinating phenomena to examine. We all experience them, use them as signals to guide our behavior, and usually assume that they accurately represent underlying physiological activity. At the same time, we implicitly know that bodily sensations are often vague, ambiguous, and subject to a variety of interpretations. It is not surprising, then, that there is often a disparity between what we think is going on in our bodies and what is objectively occurring. In short, phenomena such as physical symptoms are the stuff of psychology. My own research into physical symptoms started by accident several years ago. In a hastily devised experiment dealing with the effects of noise on behavior, I had to write a post-experimental questionnaire that would be long enough to allow the experimenter time to calibrate some equipment for a later portion of the study. I included some physical symptoms on the questionnaire as fillers. The experiment was a total failure, with the exception of the symptom reports. People's perceptions of symptoms were easily influenced by our manipulations, even though their actual physiological state had not changed. And so began the present inquiry. Despite the pervasiveness, importance, and sheer amount of time and money devoted to discussing and curing common physical symptoms and sensations, very little empirical work has been devoted to examining the psychological and perceptual factors related to sensory experience. Occa sional papers have tested a specific theory, such as cognitive dissonance, wherein physical symptoms served as an interesting dependent measure."
Article
The acceptability of three approaches to treating panic disorder was studied. One hundred eighteen family practice physicians read descriptions of imipramine therapy, cognitive-behavior therapy, and client-centered therapy of panic disorder and completed treatment acceptability measures for each description. The subjects also described their usual treatment and referral practices with panic disorder clients and rank-ordered the three forms of treatment based upon their beliefs about treatment effectiveness. Cognitive-behavior therapy was evaluated more positively than client-centered therapy on all measures of treatment acceptability. Subjects' ratings of the acceptability of drug therapy tended to fall between cognitive-behavior therapy and psychotherapy, but were not significantly different from either. Physicians reported that they tend to treat panic disorder clients pharmacologically and refer for psychological treatment relatively infrequently. Pharmacotherapy was ranked as the most effective intervention by most physicians, followed by cognitive-behavior therapy, and then by psychotherapy.
Article
Recent research on psychosocial models of panic disorder has demonstrated that panic patients become anxious about, and tend to focus on, the physical sensations of panic (e.g., palpitations, dizziness, and breathlessness). To assess whether patients' hypervigilance for panic sensations is associated with heightened awareness of internal sensations, subjects with panic disorder, social phobia, and no mental disorder were asked to count heartbeats at rest and following a period of exercise, while actual heartbeats were recorded using a polygraph. The groups did not differ in heartbeat awareness at rest. Following exercise, all groups became more aware of cardiac sensations, again with no between-group differences. Despite a lack of group differences, several variables were positively related to accuracy of heartbeat perception, including self-reported anxiety over relevant sensations and subjects' confidence in their estimations. In addition, actual heart rate was negatively related to accuracy of heartbeat tracking. The implications of these results are discussed.
Article
The aim of this work was to test the comparative efficacy of four therapeutic modalities (self-exposure, self-exposure + Alprazolam, Alprazolam, and self-exposure + placebo) and also to determine the combined effects of self-exposure with Alprazolam and self-exposure with placebo in the treatment of agoraphobia without current panic. The sample consisted of 31 patients selected according to DSM-III-R criteria. A multigroup experimental design with repeated measures of assessment (pre-treatment, post-treatment and 1, 3 and 6-month follow-up) was used. The results indicated that there was a similar therapeutic improvement (in about 75% of the cases) between pre- and post-treatment in all therapeutic modalities, except for the Alprazolam group, where improvement did not take place, was rather weak or tended to fade as time passed. This improvement increased at the follow-ups in the self-exposure + placebo group, remained stable in the self-exposure group, and was irregular or fairly unpredictable in the self-exposure + Alprazolam group. There was a positive combined action between self-exposure and placebo and a negative interaction between self-exposure and Alprazolam. The highest relapse rate appeared in the therapeutic modalities where the active drug was administered. The intratreatment evolution was faster in the self-exposure group than in the others, but it tended to remain stable in the second part of the therapy. It is therefore concluded that the efficacy of self-exposure therapy may be the same if reduced to half the number of sessions. Finally, several topics that may contribute to future research in this field are commented upon.
Article
Although clinical experience influences psychotropic drug treatment world-wide, it has been underutilized because it has not been systematically gathered and widely disseminated. An international survey was conducted to help remedy this situation. One of the major objectives was to develop a representative body of expert judgment and opinion on the clinical use of benzodiazepines in relation to other psychotherapeutic medications that might be used for the same purposes. A select international panel of psychiatric experts on the pharmacotherapy of the anxiety and depressive disorders (N = 73) was constituted on the basis of progressive peer nominations. The peer selection process began with primary nominators from 44 countries. Judgments and opinions about psychotherapeutic medications were elicited from the Expert Panel via a self-administered questionnaire. Completion rate: 90 per cent (66/73). Outcomes bear directly on current therapeutic and regulatory concerns. Topics addressed include: special indications for use, abuse liability, dependence potential, duration of treatment, high-risk treatments, and adverse effects. Although sometimes divided, agreement was generally high and indicative of a good benefit to risk ratio for the benzodiazepines. Expert judgments were more positive than would have been anticipated from media reports, public concern, and recent regulatory postures. The following inferences can be drawn from this body of expert judgment and opinion: (1) qualitative differences in abuse liability among the benzodiazepines are minimal; (2) physical dependence at therapeutic doses is not a major clinical problem; (3) when physical dependence occurs, it can be readily managed clinically by the treating physician; (4) the relative abuse liability of the benzodiazepines as a class is low. These expert evaluations do not support or justify the imposition of stronger or differential restrictions on the benzodiazepines. The data help define the reasonable limits within which clinical guidelines and regulatory mandates can be meaningfully promulgated.