Pharmacological Interventions

Psychological Interventions

Self-Change Interventions

Evaluating Treatment Options

• a recurrence of panic attacks which reduced improvement to less than 50 percent fewer attacks when compared to pre-treatment; or
• a worsening of symptoms requiring reinstatement of treatment or re-hospitalization.

Four classes of medications have been identified as potentially helpful for panic disorder with or without agoraphobia. All of these groups have been studied regarding their effectiveness for treating these disorders. The four groups are:

1. Low potency benzodiazepines such as Diazepam (Valium) and Chlordiazepoxide (Librium)
2. High-potency benzodiazepines such as Alprazolam (Xanax), Lorazepam (Ativan), Clonazepam (Klonopin), and Clorazepate (Tranxene)
3. Buspirone (Buspar) which is an anti-anxiety medication different from the benzodiazepines
4. Four classes of anti-depressant drugs
   • Tri-cyclic anti-depressants such as Imipramine (Tofranil), Amitryptaline (Elavil), and Clomipramine (Anafranil)
   • Mono-Amine-Oxidase inhibitors such as Phenelzine (Nardil) and Tranylcypromine (Parnate)
   • Selective Serotonin Re-uptake Inhibitors (SSRIs) such as Fluoxetine (Prozac), Fluvoxamine (Luvox), Sertraline (Zoloft), Citalopram (Celexa) and Paroxetine (Paxil)
   • Second-generation anti-depressants that are not SSRIs, such as Trazodone (Desyrel), and Bupropion (Wellbutrin)

The effectiveness and side effects that could result in dropouts, as well as the relapse rates after discontinuation will be discussed for each class of medication.

This class of drugs was very popular in the treatment of a variety of anxiety disorders. They act to reduce anxiety, agitation and fear by their actions on receptors in the lower parts of the brain - the amygdala and hyppocampus. They also have a sedative effect that is produced by their action on the cerebral cortex. In addition, they have a muscle relaxant effect that is produced indirectly by their anxiety-reduction effect and directly by their effect on receptors located in the spinal cord, cerebellum and brain stem. One of the principal effects of the low-potency benzodiazepines is on anticipatory anxiety that often accompanies panic disorder. Because of side effects discussed below and the tendency for chronic users to become dependent on these medications the use of low -potency benzodiazepines has decreased at least 25 percent since 1975. Nevertheless, this class of drugs was one of the first used to treat panic attacks.

When used to treat panic disorder with or without agoraphobia the dropout rate for individuals taking the low-potency benzodiazepines tends to be relatively low - approximately 21 percent of individuals started on the drug. This dropout level is not significantly different than the lowest dropout rates reported. It can, therefore, be safely concluded that there is nothing about this class of drugs that increases or decreases the dropout rate relative to other treatments. Side effects of medications are the primary reason for their discontinuance. While a number of side-effects have been found associated with the administration of these drugs, including impaired memory and concentration, irritability, insomnia, loss of appetite, muscle aches, sweating, and heart palpitations, the frequency and severity of these side effects is so low as to not result in a significant discontinuation of their use.

The effectiveness of the low-potency benzodiazepines, as defined by eliminating or reducing the frequency of panic attacks to less than 50 percent of the pre-treatment level, was 45 percent for a relatively small number of studies that systematically studied their effectiveness. As we shall see, when we compare these rates to those of other treatment this effectiveness rate is fairly low. In fact, when you compare the effectiveness rates of other drugs and psychological therapies to those of this class of drugs they are found to be significantly less effective. Most likely this low effectiveness rate has led to a decline in the frequency with which these medications are prescribed for the treatment of panic. This low rate of improvement may also account for the relative infrequency that the effectiveness of these drugs have been studied.

There has been almost no study that has examined the relapse rate when low-potency benzodiazepines have been used to treat panic disorder. In one study that was found, using a follow-up period of only 3 weeks, the relapse rate was 10 percent. While this rate is relatively low when compared to other treatments the period of time individuals were followed after the conclusion of treatment was relatively brief, so brief in fact that no real conclusions can be drawn about relapse rates.

One additional effect should be considered in evaluating the relapse rates of individuals who have been using this class of drugs for a year or longer. Recent studies have shown that individuals taking these drugs for an extended period of time develop a physical dependence on them. For these individuals withdrawal of the drug produces withdrawal symptoms, including psychological, physical, and perceptual problems. These problems are most apparent when the individual is abruptly terminated from the treatment. Taking an approach that slowly reduces the amount of drugs adminstered essentially eliminates the experience of withdrawal symptoms. Accordingly, withdrawal from these drugs should only be undertaken under the close supervision of a physician.

High-potency benzodiazepines are so-called because of the lower dosages required to produce the same anxiety-reduction effects as the low-potency benzodiazepines. Depending on the drug in question high-potency benzodiazepines can be from 5 to 20 times as potent as Valium. They are given, therefore, in much smaller doses, measured in milligrams. The duration of "clinically-significant effect" varies depending on the drug in question. Xanax is the shortest lasting drug, having a time of effectiveness of approximately 6 hours, independent of the dosage level. To reduce the probability of panic attacks throughout a 24-hour day requires that it be taken 3-4 times. Ativan clinical effect is slightly longer - about 8 hours, while Klonopin has a clinical effect of an entire day. The mechanisms of action are similar to those of the low-potency benzodiazepines, in that they reduce anxiety, reduce muscle tension, and increase the level of sedation.

As with the low-potency benzodiazepines the high-potency benzodiazepines produce very low dropout rates. A number of studies have been conducted on dropout rates when Xanax was administered. These rates were found to be among the lowest for any treatment - approximately 15 percent. This low rate is again largely attributable to how well taking these medications is tolerated because of the few or insignificant side effects produced by them. The dropout rate for Xanax is somewhat, but non-significantly lower than for the low-potency benzodiazepines. This may be attribuatable to a second reason that some people discontinue using a particular treatment, namely its ineffectiveness. Since, as we shall see, low potency benzodiazepines are less effective than high-potency benzodiazepines in the treatment of panic disorder, with or without agoraphobia, a greater number of individuals may stop taking the prescribed treatment because they are not experiencing significant improvement.

Using the criteria of effectiveness defined above high-potency benzodiazepines are one of the most effective treatment for panic disorder and agoraphobia. Treatment success rates were 68 percent when aggregated over a series of different studies. There appears to be a different effectiveness rate when Xanax is used to treat individuals diagnosed as panic disordered only compared to individuals diagnosed as panic disordered with agoraphobia. When combining a variety of studies in each group the effectiveness rate for individuals diagnosed as panic disordered only was 94 percent For individuals whose diagnosis included agoraphobia, however, the effectiveness rate dropped to 64 percent. These improvement rates likely account for the rapid increase in the prescription of this class of drugs for treating panic disorder. When Xanax is combined with either supportive or behaviorally-oriented psychotherapy no increase in improvement rates have been found. This conclusion must be interpreted cautiously, however, as it is based on very small samples of individuals.

Relapse rates are only meaningful if the individual being treated wants to stop taking the drug or if they have been recommended to stop taking the drug. Stopping the drug can be considered for a number of reasons, including the cost of taking the medication and a philosophy that one would prefer to be drug-free.

Discontinuation of the drug was often part of the treatment protocol in those studies that reported relapse rates. In an early study that combined results across a number of studies a relapse rate of 28 percent for high-potency benzodiazepines was found. This estimate may be somewhat low because a rather liberal definition of relapse was used in one of the largest studies conducted. More recent estimates of relapse place the figure at 55 to 70 percent of those discontinued from the drug. In one large study of panic disordered individuals followed from 3 to 6 years post-treatment, approximately 33 percent remained symptom free, and about 50 percent experienced only mild symptoms. Approximately 20 percent did poorly throughout the treatment and follow-up periods.

The tri-cyclic anti-depressants, developed in the 1950s, were among the first groups of drugs found to be effective in the treatment of depression. Somewhat later, in 1964, a landmark study Donald Klein found that individuals who experience panic attacks had a positive reaction when given Tofranil (Imipramine). Since then approximately 12 controlled studies have been conducted on the effectiveness of this class of drugs, though almost all of the studies have focused on either Tofranil or Anafranil. Tri-cyclics work by preventing the reuptake of serotonin into the presynaptic nerve channels. Serotonin is a neurotransmitter required for the smooth transmission of nerve impulses in the Central Nervous System. It must be present in sufficient quantities in the synapse for smooth transmission of these nerve impulses. When the amount of neurotransmitters decrease, as is the case when they are "taken-up" into the presynaptic nerve channels, the depressive symptoms of slowing, low energy and depressed mood are seen. Panic attacks are known to be frequently associated with depression, though it is unclear exactly how the tri-cyclics reduce the frequency of panic attacks.

Tri-cyclics that, as we shall see, are effective in the treatment of panic disorder and agoraphobia, also bring a number of side-effects with them. These side effects include over-stimulation which is experienced as increased anxiety, dry mouth, constipation, sexual impotence and weight gain, occur frequently enough and at a level high enough to lead many individuals who start on this treatment regimen to terminate it. In controlled studies evaluating the effectiveness of these medications these terminations are termed treatment dropouts. In a summary of these dropout rates Clum reported in a 1989 article that 28 percent of individuals terminated their participation in these studies. Over-stimulation, which occurs about 20 percent of the time, was given as the principal reason for terminating the drug. Weight gain, which occurs approximately 34 percent of the time was the next most important reason. The negative side effect of over-stimulation can be handled by beginning treatment with an anti-anxiety medication and then gradually building up the level of the tri-cyclic medication.

Over a series of studies Tofranil was found to lead to improvement, as defined above, in approximately 63 percent of the individuals who continued to take it. This rate of improvement is not significantly different than the most effective treatment approach. As with the high-potency benzodiazepines Tofranil is somewhat more effective in studies that target individuals with Panic Disorder (75 percent) than in those studies that target individuals with Agoraphobia (62 percent). The addition of cognitive-behavioral or supportive therapies was not found to improve effectiveness rates. As was previously noted, this latter conclusion must be viewed cautiously as it is based on a relatively small sample size.

Once started on a regimen of tri-cyclics, individuals who either experience few side-effects or can tolerate the side-effects they do experience can remain on the medication indefinitely as there are no long-term effects of taking these drugs. For those individuals who see themselves as drug-free sometime in the future, however, relapse rates are a consideration. Unfortunately, no studies were found which utilized a tricyclic alone, without the addition of some psychotherapeutic treatment , either supportive or behavioral. When combined with a supportive therapy intervention individuals taking Tofranil relapsed 40 percent of the time. When combined with a behavioral intervention, individuals relapsed 23 percent of the time. Studies have compared the relapse rates of individuals who terminate treatment after undergoing a cognitive-behavioral intervention, a tricyclic-only intervention or an intervention combining the two. Relapse rates were lowest in the cognitive-behavioral treatment. Adding a cognitive-behavioral intervention to the tricyclic did not reduce the relapse rate. Combined, these data suggest that relapse after termination of tricyclic treatment is a serious consideration.

Monoamine oxidase is an enzyme that breaks down normal neurotransmitters so that they cannot be used to transmit messages across nerve fibers. MAO inhibitors work by bonding with the enzyme and preventing it from breaking down the neurotransmitters norepinephrine and serotonin. As a consequence, large amounts of these transmitters accumulate in the nerve terminals and are released when the nerves are stimulated. MAO inhibitors were the first drugs to be used effectively to treat depression. Their effectiveness rate for treating depression is high and they often successfully reduce depression in individuals who are not effectively treated with the tricyclics. As we will see later the side effects associated with taking this group of drugs has worked against it being used more frequently. A newer class of MAO inhibitors does not have the long-lasting effect of the older drugs, increases cognitive functioning and produces fewer side effects.

Soon after their introduction in the 1950s the MAO inhibitors were found to have serious interactions with other substances, including adrenaline-like drugs such as those found in nasal sprays, cold medicines, etc., as well as foods that contained tyramine, found in wine, beer, cheese and some beans. These interactions included a sharp rise in blood pressure that were occasionally serious enough to be fatal. Like the tricyclics they also work slowly and may take several weeks or more to exert an anti-depressant effect. They also have similar side-effects as the tricyclics, including weight gain and interference with sexual function. There is very little data on dropout rates for individuals who begin treatment on these drugs, while aware of their possible interactions with other substances. When used to treat agoraphobia combined with supportive psychotherapy the dropout rate was 41 percent, which is at the high end.

MAO inhibitors have improvement rates as defined above of approximately 58 percent, which is somewhat, but not significantly lower than the improvement rates for the high-potency phenothiazines and tricyclics. Again, these estimates are based on limited data and should be interpreted cautiously. Improvement is not enhanced by adding supportive therapy to the intervention package. While there is little in the way of comparison data for individuals with panic disorder and individuals with agoraphobia there appears to be no differences in effectiveness rates for these two groups. Given the complications that can arise should dietary restrictions not be followed, and the relatively similar effectiveness rates for treating panic when compared to other anti-depressants, the MAO inhibitors have been studied and used clinically less frequently than other anti-panic medications.

Again these conclusions are based on minimal information as provided by research studies. For those studies that do exist the relapse rates for this class of drugs was high - 55 percent. This relapse rate is the highest for any pharmacologic agent examined. Further, the individuals on whom these rates were determined carried a primary diagnosis of panic disorder as opposed to agoraphobia.

There is a relationship between decreased amounts of serotonin - a neurotransmitter (substance that is required to transmit messages in the brain) - and depression. One way that serotonin gets depleted and cannot be used to aid neurotransmission of nerve impulses is when it is taken back (re-uptake) into the pre-synaptic nerve. The SSRIs block this re-uptake, increase the amount of serotonin available for nerve impulse transmission, and decrease depression. There are currently 5 different SSRIs that, while they all have anti-depressant and anti-anxiety effects, have been found to have different levels of potency, somewhat different side-effects and exert different effects depending on the type of anxiety disorder targeted. This brief review will focus on five of these SSRIs - Prozac, Paxil, Luvox, Zoloft and Celexa - all of which have been shown in at least one large-scale or several small-scale placebo-controlled studies to affect panic disorder and its correlates.

Dropouts occur with the SSRIs at a much lower rate than for the other anti-depressant medications, likely because this class of anti-depressants has a much lower side-effect profile. Chief among the side-effects of these medications are increased anxiety and restlessness, sweating, insomnia, nausea and sexual difficulties, including decreased libido and abnormal ejaculation. Celexa appears to be one SSRI with a particularly low rate of side effects, including no long-term weight gain, anxiety, agitation or nervousness. These side-effects occur in a relatively low percentage of individuals or with sufficiently low level of potency to result in low dropout rates. For example, individuals with panic disorder, the dropout rate for individuals receiving Paxil is approximately 10 percent while the dropout rate for individuals receiving Prozac has been found to vary. A recent study by Bakker, et al. (1999) in the Journal of clinical Psychiatry reported a 13 percent dropout rate for individuals taking Prozac, while a less recent study (Pecknold, et al., 1995) found a much higher dropout rate. Another reason that individuals stop taking these medications is that they can take from 6-8 weeks to begin having anti-panic effects, especially when started on low doses. This latter problem can be dealt with by more frequent examination of progress, especially after week three, when some studies begin to show an anti-panic effect. Another approach is to combine an SSRI with a high-potency Benzodiazepine such as Xanax in the first few weeks and then gradually reduce the role of Xanax.

It is clear from a variety of studies that the SSRIs are a first-line treatment modality for panic disorder. Effectiveness rates, defined as being panic-free at the end of treatment has been as low as 50 percent for low doses of Prozac, 75 percent for higher doses of Prozac, to 86 percent for Paxil. Improvements from these medications are also shown across a variety of outcome measures, including general anxiety, depression and global improvement. Improvement appears to be dose and time related with higher doses and treatment over longer time periods associated with greater improvement. In one study of the effectiveness of Prozac (Schneier, et al., 1990) an effectiveness rate of 76 percent was found after one year of treatment.

There are very few studies that directly compare the SSRIs to other types of medications. Several studies have been completed comparing Sertraline (Zoloft) to Clomipramine (Anafranil) with no significant differences found between the two. Anafranil has a much higher side-effect profile, however, and is typically less well tolerated. Further, a recent study comparing Clomipramine to Prozac (Bakker, et al., 1999), found that panic-free status was accomplished in 59 percent of individuals being treated with Clomipramine and 75 percent of individuals being treated with Prozac. Further, a recent review study that pooled data on drug effectiveness across all studies found that the SSRIs as a class had a greater treatment effect than the tri-cyclics or benzodiazepines for individuals with panic disorder with or without agoraphobia. These data seem to suggest that the SSRIs are currently the first choice for pharmacologically treating individuals with panic problems.

Psychological Interventions

Two types of psychological interventions have been studied in terms of their effectiveness for treating panic disorder and agoraphobia: cognitive-behavioral approaches and interpersonal approaches. Each of these has been provided in both individual (one-to-one) and group modalities. While there is supportive evidence for both of these approaches, the preponderance of research studies, and therefore the preponderance of supportive evidence, lies with the cognitive-behavioral approach. Accordingly, the summary of the effectiveness of psychological approaches to treatment will focus on the cognitive-behavioral approach.

Cognitve-Behavioral Treatments

The cognitive-behavioral (CB) approach utilizes specific treatment components that target each of the characteristics of panic disorder and agoraphobia. Panic attacks are known to occur much more frequently in individuals who have high baseline levels of anxiety, Accordingly, one aspect of the CB approach is to train the panic sufferer in relaxation techniques to be used to reduce the overall level of anxiety. Characteristically, panic sufferers also tend to respond to the initial symptoms that signal the onset of a full-blown panic attack by becoming fearful, with an increase in their level of arousal. As part of this response panic sufferers tend to hyperventilate, a response that increases the number of panic symptoms and increases the likelihood of their experiencing a full-blown attack. Accordingly, the CB approach often includes teaching breathing techniques, such as "deep-belly" or diaphragmatic breathing, to be used to counter the automatic reaction of hyperventilation. A third characteristic of panic sufferers is to interpret their panic symptoms catastrophically, anticipating that they are suffering from some serious ailment, will go crazy or do something harmful or embarrassing. These erroneous thoughts are countered using both corrective information and exercises that help the panic sufferer understand and challenge these panic-accelerating thoughts. The anticipatory anxiety that panic sufferers experience when they fear that a panic attack may occur is dealt with be teaching the above techniques and by learning that employing such techniques will provide the sufferer with a set of strategies to short-circuit the panic attack To counter the panic sufferers tendency to avoid situations in which panic attacks have been experienced (agoraphobia) graduated exposures are provided to both the actual external situations and to the internal experiences that are feared. This is accomplished in a variety of ways, but principally by having the panic sufferer practice the techniques described above and then introduce him gradually to the feared situations. This could be accomplished within the therapist's office, outside the office while accompanied by the therapist, and later, outside the office by the individual himself.

In a review of treatment studies using CB treatments published in the journal Behavior Therapy, Clum found that only 13 percent of individuals with panic disorders who received CB treatment dropped out of therapy, while only 18 percent of individuals with both panic disorder and agoraphobia dropped out of therapy. These dropout rates are equivalent to the best dropout rates for any of the pharmacologic interventions. It can therefore be concluded that there is nothing about the use of the CB treatment approach when compared to other treatments, that increases the likelihood that individuals will stop therapy . Since dropouts occur primarily because the treatment is producing significant side-effects or because it is not being effective, it is likely that the CB approach is not seen by individuals undergoing such treatment as problematic or ineffective. CB treatment can, however, produce discomfort. This discomfort occurs primarily when the panic sufferer is asked to put himself in anxiety-producing situations. Since exposing oneself to feared situations is an important part of the treatment process, this aspect of treatment may be responsible for the dropout rates that have been observed.

Behavior therapies in general have been found to be very effective. In a review study of these treatments by George A. Clum published in the journal Behavior therapy, 83 percent of panic disordered individuals and 61 percent of individuals with panic disorder and agoraphobia were found to be significantly improved by the completion of treatment. These findings have been very consistent over time with more recent treatments that combine information, relaxation, cognitive restructuring, and exposure to both feared external situations and feared internal events having the greatest success rates. Recent studies that have compared this approach to tri-cyclic anti-depresssant drugs have shown them to be equal to the drugs in effectiveness.

One of the most positive features of CB treatments for panic disorder and agoraphobia is that you learn skills for dealing with these problems. Because these skills are learned they continue to be available even after treatment is completed. When a former panic sufferer encounters stressful situations that increase his anxiety and makes him vulnerable to developing panic attacks anew, he can call on his learned skills to deal with the experiences he is having. The fact that such skills are learned may account for the relatively low relapse rate for individuals successfully treated for these problems. For individuals diagnosed with panic disorder only, relapse rates are the lowest for all treatments - approximately 8 percent. For individuals diagnosed as panicked disordered with agoraphobia the relapse rates increase to 18 percent, still a very low rate. Such rates compare very favorably with those found for individuals treated with pharmacologic treatments after these treatments have been discontinued. These rates are typically found when the treated individuals have been followed for a relatively brief period of time - from six months to one year. Because of the vulnerability of panic sufferers to stress and because stress tends to reoccur, panic tends to repeat throughout the life cycle. Relapse rates, for the very few studies that have examined them over longer time periods, tend to increase as the time from the end of treatment increases.

Self-Change Interventions

Self-change interventions, also referred to as self-help, bibliotherapy, web-based interventions, etc., have been used as interventions for individuals with anxiety for several decades. For the most part, these interventions tend to be prescriptive in format and provide recommendations based on principles developed from cognitive-behavioral therapies. Often, they involve imparting the same therapeutic information that a therapist would provide but without the interactive nature of the therapeutic process. These interventions, when applied to panic problems and avoidance behavior, include such prescriptions as:

1. providing the sufferer with information about their problem
2. providing information on the use of relaxation strategies with directions on how to implement such strategies on their own;
3. providing information on the importance of catastrophic cognitions and fear of having a panic symptoms on the maintenance of the disorder;
4. providing information on why avoidance becomes a prominent feature of the disorder and how to go about changing that avoidance; and
5. providing information on how to structure these interventions sequentially and by oneself.

We do not yet understand which individuals profit from such an approach and whether they differ from individuals who require individual or group therapy formats or pharmacologic interventions. One possibility is that the individuals who utilize such self-paced interventions are a select group, in that they identify themselves as people who can put effective change processes into action without professional help. While not much is known about who accesses such interventions a body of research literature has developed that evaluates the effectiveness of these interventions.

Early scientific reviews of the effectiveness of self-help treatments in general concluded that a relatively large number of individuals who started such programs terminated treatment prematurely. Very few research studies on the effectiveness of self-change interventions with panic have examined dropout rates. Of the few that have, the rate is much lower than the early reviews would lead us to believe. In a study published in Behavior Therapy by Gould and Clum (1995) the dropout rate was 17 percent. In a study by Lidren, et.al. (1994) even better results were found. These authors reported a zero percent dropout rate. A third research study by Hecker, et al. (1996) also reported a zero percent dropout rate. When you combine the results from the three studies a dropout rate just under six percent is obtained, considerably lower than those reported in pharmacologic studies or in studies of CB therapy using individual or group therapists. A fourth study that reports dropout rates for telephone therapy of panic disordered individuals with various levels of agoraphobia used with individuals in remote regions of Canada has also been conducted. In this study a nine percent dropout rate was reported. It is difficult to explain the low dropout rate for these self-change studies. One possible explanation is that there is a selection factor, and that individuals who elect to participate in these studies are highly motivated to change and thus dedicate themselves to completing the program. Another explanation is that the convenience of being able to complete an intervention program in the comfort of one's own home, at one's own pace, and for a relatively low cost, explains why panic sufferers appear to complete these treatments.

In a review study conducted by Drs. Gould and Clum (1993) self-change treatments were found highly effective. In 12 studies that directly compared self-change with therapist- administered treatment no differences were found, suggesting that for many self-administered interventions are as powerful as therapist-administered interventions, at least for individuals who elect to participate in such studies. When studies that targeted individuals with anxiety disorders were examined separately, the results were even more promising. Targeting anxiety resulted in one of the highest treatment effects of any problem treated via this method. Since the publication of this research, four studies that compared the effectiveness of self-change approaches to therapist-administered interventions for the treatment of panic disorder and agoraphobia have found self-administered interventions as powerful as therapist-administered interventions. One study did not find that a self-administered intervention was superior to no-treatment. Participants in this study were completely on their own, with no therapist contact. It may be necessary to provide some type of contact to continue to motivate panic sufferers to use the techniques they have been taught. No studies have compared self-change interventions with either placebo or with any pharmacologic treatments. Examining the effect sizes of self-administered interventions compared to those from drug-outcome studies indicates that the latter is likely to be more effective, especially as long as the individuals remain on the medication.

An important consideration in evaluating the effectiveness of self-administered interventions is whether the improvements that have been shown continue after the treatment is completed or whether individuals in this type of program relapse more frequently than with other approaches. In the studies conducted to date on the use of self-administered interventions for panic, follow-up results up to six months after the completion of treatment indicate: 1) no differences between therapist-administered and self-administered treatments at follow-up; and 2) little change in the number of panic sufferers who remain panic free at follow-up when compared to the number that are panic free at the conclusion of the intervention. These results, while still few in number, are promising. They suggest that self-administered treatments for panic can have long-term effects.