Thursday, 26 September 2013

Using adaptive treatments for smoking cessation may prove effective!

 This post was written by Andrew Jones for the Mental Elf blog, you can find the original here

Around 20% of adults in the UK smoke cigarettes regularly. Smoking was the primary cause of approximately 462,900 hospital admissions and 79,100 (18%) deaths in adults over the age of 35 last year. Therefore, smoking cessation represents a serious (and costly!) clinical challenge on an individual and global level.

Many smoking cessation aids are available, including over the counter treatments such as nicotine patches, gums and inhalers, known as Nicotine Replacement Therapy (NRT). Prescription drugs such as bupropion and varenicline are also widely used. These treatments have differing degrees of efficacy as well as potential side-effects (which we have discussed in a previous elf blog). Despite the treatment options available, quitting rates are low and there is little available guidance to inform health care providers as to which treatment should be provided for any given smoker.

A recent study published in the American Journal of Psychiatry set out to examine the optimal treatment strategies for individual smokers who did not respond well to initial treatment with NRT.

Methods and Results 

The authors performed a double-blind, parallel-arm adaptive treatment trial, consisting of two separate phases. Smokers who expressed a desire to quit were recruited using various media advertisements. To be eligible, participants had to smoke more than 10 cigarettes per day for 3 years and have an expired breath carbon monoxide (CO) reading of at least 10 ppm (parts per million). Following screening of 1,256 smokers, 606 were eligible and enrolled into the study.

The volunteers attended a research center weekly for two weeks prior to a target quit date, during each visit brief support was given and various measures including smoking diaries, expired CO and withdrawal symptoms were recorded. The first line treatment was open-label NRT which everybody received, chosen for its safety and tolerability.

There were two phases to the research, each with a different hypothesis.

Phase 1 

This phase examined whether individuals who did not respond adequately to first line NRT could be ‘rescued’ by switching to a different treatment strategy. Individuals were classed as ‘non responders’ if their expired CO did not reduce by more than 50% in the first week of NRT.

 These non responders were then randomly assigned to one of three treatment strategies:

1.  Switch from NRT to varenicline

2.  NRT augmented with bupropion

3. Continuation of NRT only

NRT was not augmented with varenicline due to increased risk of adverse effects. The primary measure of interest was continuous abstinence 8-11 weeks after the target quit date.

313 smokers were categorized as non responders and were randomly assigned to one of the three strategies. Individuals who remained on NRT alone following initial poor response had low abstinence success rates of 16% after 8-11 weeks, which reduced further to 5.8% after 6 months.

NRT augmented with bupropion significantly improved continuous abstinence rates compared to NRT only after 8–11 weeks  (28.3%; p=.04; odds ratio = 2.06, 95% CI=1.05 – 4.07). Significant improvements in both continuous and point abstinence were demonstrated up to 6 months later.

The switch from NRT to varenicline did not significantly improve continuous abstinence at 8-11 weeks. However, there was a significant improvement in 6 month point abstinence (16.5%; p=.03; odds ratio = 2.80, 95% CI 1.11 – 7.06).

Phase 2 

This phase set out to examine whether a change in treatment strategy could ‘rescue’ individuals who had lapsed one week after their quit date. Lapsers were identified as individuals who self reported taking even one puff of a cigarette, or who provided a CO reading over 10 ppm. 95 lapsers were randomly assigned to one of the three strategies used in phase 1 and given a second quit date exactly one week later. If individuals did not lapse they remained on NRT.

Individuals who lapsed after one week showed no significant improvements in continuous or point abstinence if NRT was augmented with bupropion, or if they were switched to varenicline, compared to those who continued with NRT only.


The results of the study support an adaptive smoking cessation strategy and suggest that individuals who do not respond to NRT as a first line treatment before their quit date may benefit from a switch to an alternative therapy. Specifically, augmenting NRT with Bupropion led to an abstinence rate 2-3 times higher than continuation of NRT alone at later follow-ups. There were also some improvements among smokers who switched to varenicline. However, using the same adaptive treatment strategies to rescue smokers who relapsed after one week of the quit date had no significant beneficial effects on abstinence.

The authors correctly point out that the results of phase 2 should be interpreted with caution due to relatively small group sizes. Furthermore, whilst the adaptive treatment led to a significant improvement in ‘non-responders’ the actual abstinence rates after 6 months were still relatively low (13.1% for NRT augmented with bupropion). Future research should investigate which smokers will respond best to adaptive treatment by examining predictive markers of smoking and treatment efficacy such as genes.

The principle finding from this research is that individuals who did not respond to NRT by reducing their smoking were identified before their target quit date, therefore no participants actually relapsed before receiving adaptive treatment. By doing this the authors suggest that:

 "we could minimize the deleterious effects of relapse on motivation and retention in treatment".

To conclude, it may be possible to ‘rescue’ smokers who do not respond well to NRT before their quit date, by offering alternative pharmacotherapies instead.


Rose, J.E. & Behm, F.M. (2013). Adapting smoking cessation treatment according to initial response to precessation nicotine patch. American Journal of Psychiatry; 170, 860-867. [PubMed abstract]

The target study is also discussed in a podcast (August 2013) from the American Journal of Psychiatry.

Statistics on Smoking: England, 2013 (PDF). Health and Social Care Information Centre, 15 Aug 2013.

Tuesday, 24 September 2013

Concurrent treatments may be effective in treating comorbid alcohol dependence and PTSD

This post was written by Abi Rose for the Mental Elf, you can find the original here.

In those with post-traumatic stress disorder (PTSD), alcohol abuse or dependence is the most common co-morbid disorder. Unfortunately, although individuals with PTSD and alcohol dependence (AD) often have complex needs, the treatments available usually only target one issue. This is often due to specialist treatment services not having access to treatments for comorbid disorders, and the belief that treatment for PTSD can increase drinking behaviour, while AD may interfere with PTSD treatment.
Following treatment, those with comorbid PTSD and AD are more likely to relapse, and relapse quickly, than patients with AD alone or AD with other comorbidities. These increased risks combined with the complex needs of this patient group make the development of effective concurrent treatments a priority.
Effective concurrent treatment combines two or more therapies, but it's vital that the interventions don't conflict
Effective concurrent treatments combine two or more therapies for two or more conditions
Foa and colleagues conducted a trial (recently published in JAMA) which primarily assessed the effectiveness of combining an evidence-based pharmacotherapy for AD (naltrexone) with an evidence-based psychosocial treatment for PTSD (prolonged exposure therapy).
Overall they found that:
  1. Naltrexone decreased percentage of drinking days
  2. Prolonged exposure therapy did not decrease PTSD symptoms compared with other groups during treatment
  3. Prolonged exposure therapy did not worsen percentage drinking behaviour


165 patients were recruited from two clinical sites in Pennsylvania and required a diagnosis of AD (defined by: DSM-IV, >12 drinks per week, >4 drinks in one day over past 30 days) and PTSD (DSM-IV, ≥15 score on the PTSD symptom severity scale). Patients did not have any other substance dependence (except nicotine), psychiatric condition, or physical condition which could interfere with the treatments provided.
Patients were randomised to 1 of 4 conditions:
  1. Prolonged exposure therapy & naltrexone (100 mg/d)
  2. Prolonged exposure therapy & placebo
  3. Naltrexone (100 mg/d)
  4. Placebo
Treatments were started after outpatient detoxification, and lasted 6 months. Prolonged exposure therapy was administered once a week for 12 weeks and then once a fortnight for 12 weeks, and involved: imagining, processing, and discussing traumatic memories; listening to these sessions at home; in vivo exposure to safe situations.
In addition, all participants received 18 supportive counselling sessions with a nurse (administered on the same time frame as the exposure therapy).
Outcome measures were PTSD symptom severity, percentage of days drinking and alcohol craving.
Outcome measures were taken at baseline, during treatment, and up to 52 weeks follow-up.


Of the 165 patients, 11-15 patients in each treatment group did not receive the intervention as randomised; 20-26 completed the 3 month follow-up, 19-30 completed the 6 month follow-up.
Alcohol dependence outcomes:
Naltrexone decreased the percentage of drinking days
Naltrexone decreased the percentage of drinking days
  • All patients reported a decrease in drinking during treatment, irrespective of which group they were in
  • Following treatment, those on naltrexone reported fewer drinking days than those on placebo
  • Craving was lower in those receiving naltrexone, supporting the suggestion that naltrexone’s therapeutic effect may be to reduce craving
  • There was no effect of prolonged exposure therapy on drinking outcomes
PTSD outcomes:
  • PTSD symptoms decreased during treatment, irrespective of treatment group
  • Following treatment, there were no differences in PTSD symptoms across the different groups
  • Further analysis found that 70% of patients in the Exposure + Naltrexone group had a low PTSD severity score (≤10) 6 months following treatment. This compared favourably with those in the Exposure + Placebo (55%), naltrexone (43.9%) and placebo (37.2%) groups


Patients receiving the prolonged exposure treatment only attended a third of the available sessions on average
Patients receiving the prolonged exposure treatment only attended a third of the available sessions on average
The finding that prolonged exposure therapy was no better than supportive counselling as a treatment for PTSD is inconsistent with other evidence. The authors suggest that supportive counselling (which all patients received) may have masked specific effects of prolonged exposure, and this is something that needs to be controlled for in future research. However, the lack of a significant finding may also be due to poor adherence rates; patients only attended an average of 6/18 exposure sessions.
Importantly, the PTSD treatment did not exacerbate AD symptoms and, at 6 month follow-up, there was a lower alcohol relapse rate in those from the Exposure + Naltrexone group.


Given the low adherence rates to the prolonged exposure treatment, this study is unable to comment on the long-term effectiveness of the PTSD therapy. The complex needs of this patient group make it possible that those with comorbid disorders have difficulties in engaging and sticking with treatment regimes. Therefore, future research needs to focus on ways to improve treatment adherence.
The now somewhat out of date NICE guidance (NICE, 2005) suggests treating alcohol dependence before PTSD treatment starts (although in severe cases collaborative care may be appropriate). However, the current study found that treatments did not exacerbate comorbid symptoms. It is important that research continues to investigate the effectiveness of concurrent treatments in complex patient groups, and for health services to look at ways to deliver such treatments packages.


Foa EB., et al. (2013) Concurrent naltrexone and prolonged exposure therapy for patients with comorbid alcohol dependence and PTSD: a randomized clinical trial. Journal of the American Medical Association 310(5): 488-495. [Abstract]
Jacobsen LK., et al (2001) Substance use disorders in patients with posttraumatic stress disorder: A review of the literatureAmerican Journal of Psychiatry 158(8): 1184-1190

Monday, 2 September 2013

Do people stop smoking if their doctor advises them to? Cochrane review says sometimes and it IS worth the effort

This post was originally written by Matt Field for the Mental Elf website, you can find the original here

As previously discussed on the Mental Elf, the rates of tobacco smoking in most developed countries are falling, helped in part by high levels of taxation and bans on smoking in public places.

However, over 20% of adults in the UK continue to smoke, and this means that smoking is the greatest single cause of premature illness and death.  Therefore, anything that can be done to persuade smokers to quit is likely to have a major impact on public health.

General Practitioners and other health professionals could play an important role if they simply advised their patients to stop smoking as a precursor to prescribing medication, or referring them to local Stop Smoking services. In this Cochrane review, Stead et al. (2013) looked at whether advice from doctors could help their patients to quit smoking.


This paper is an update of previous Cochrane reviews on this topic. The authors identified randomised controlled trials that examined the influence of smoking cessation advice from a medical practitioner on abstinence from smoking at least 6 months later.

Having identified relevant papers, the authors pooled the data and asked:

  • - Is advice from a doctor to give up smoking more effective than no advice, in terms of the number of patients who manage to quit?

  • - Are there any benefits of offering more intensive advice (as opposed to just giving brief advice), and of following patients up afterwards?


Brief advice and intensive advice both improve quit rates, but only everso slightly
Brief advice and intensive advice both improve quit rates, but only everso slightly
The total sample size was over 31,000, from 42 randomised trials that were published between 1972 and 2012.

The majority of the individual trials randomly assigned people to receive brief advice to give up smoking, or they got no advice at all (not even a wagging of the finger). There was enormous variation in what constituted ‘brief advice’, but this didn’t seem to affect the overall results too much.

There were 17 trials in which the intervention was classed as minimally intensive ‘brief advice’, and 11 trials in which the intervention was classed as more intensive.

  • Comparison of minimally intensive brief advice with no advice demonstrated that brief advice led to a small but significant increase in quit rates (Relative Risk (RR) = 1.66, 95% CI 1.42 to 1.94).

  • Comparisons of more intensive advice with no advice also demonstrated an increase in quit rates (RR = 1.84; 95% CI 1.60 to 2.13).

  • However, this indirect comparison between different studies revealed no statistically significant difference between brief advice and more intensive advice (note the overlapping confidence intervals).

There were fifteen additional trials that made a direct comparison between intensive advice and brief advice. This revealed significantly higher quit rates among people who received intensive advice compared to those who received only brief advice (RR = 1.37, 95% CI 1.20 to 1.56).

Other findings included a small incremental effect on quit rates if follow-ups were provided after the initial advice. However, provision of additional aids (such as self-help manuals) did not make much difference, and there were minimal differences between different types of intervention.


Giving smokers brief advice to help them quit is better than nothing, but only just!
Giving smokers brief advice to help them quit is better than nothing, but only just! However, it’s cheap and easy so definitely something that GPs should be doing.
Overall, patients who receive brief advice to quit smoking from their doctor are more likely to actually do so, compared to patients who receive no advice. The effect size is small, although it is slightly larger if the advice is more intensive, or if patients are followed up a bit later.

Can brief advice make a difference in the real world? The ‘unassisted quit rate’ (the percentage of smokers who will quit if given no help or advice whatsoever) is difficult to estimate, and it varied enormously between studies in this analysis. But a plausible estimate is 3%, and this gives us a ‘number needed to treat for an additional beneficial outcome’ (NNTB) ranging between 33 and 80. In other words, doctors would need to offer brief advice to between 33 and 80 of their patients in order to get one of them to stop smoking! While this may seem rather pitiful, we have to remember that most smokers do not manage to quit by themselves and anything that nudges smokers towards quitting, particularly something very quick and easy like brief advice from a doctor, should be put into practice whenever possible.  

The authors conclude their paper by stating that:

"Every smoker who does not receive advice represents a missed opportunity".

And the Mental Elf agrees with that!


Stead LF, Buitrago D, Preciado N, Sanchez G, Hartmann-Boyce J, Lancaster T. Physician advice for smoking cessation. Cochrane Database of Systematic Reviews 2013, Issue 5. Art. No.: CD000165. DOI: 10.1002/14651858.CD000165.pub4.

Cochrane summary of this review