GME Logo

User login

User menu

Issue 52, Aug 2016
Share This
Do all persons with MDD have to continue the antidepressant?

Guidi et al. The Sequential Integration of Pharmacotherapy and Psychotherapy in the Treatment of Major Depressive Disorder: A Meta-Analysis of the Sequential Model and a Critical Review of the Literature. Am J Psychiatry. 2016 Feb 1;173(2):128-37. PubMed PMID: 26481173.

Antidepressant monotherapy, psychotherapy only, and antidepressant plus psychotherapy. Are these the only three treatment approaches for continuation/maintenance treatment of major depressive disorder (MDD)?

This paper suggests a fourth option—to use medication in the acute phase and then start psychotherapy, with or without tapering off the antidepressant.


The “sequential model” of treatment discussed in this paper consists of using medication in the acute phase and, after the person has significantly improved, starting psychotherapy.

This paper reports on an updated meta-analysis of the several randomized controlled trials of this “sequential model” in persons with major depressive disorder that aimed to reduce the risk of relapse or recurrence.

Note that the sequential model does not refer to using psychotherapy as an augmentation treatment in persons with an inadequate response to an antidepressant. The sequential model refers only to using psychotherapy later in treatment and aiming to improve symptoms that did not improve with the antidepressant.


Multiple databases were searched. 

The meta-analysis included randomized controlled trials that assessed the efficacy of using psychotherapy after successful response to acute-phase treatment with medication, aiming to reduce relapse or recurrence in adults with major depressive disorder.


Thirteen “high-quality” studies were included that had a total of about 700 patients each in the sequential treatment arm and in the control treatment arm.

All these studies involved cognitive-behavioral therapy (CBT).

Several different comparisons were made:

1. Combining all the studies, the risk ratio for relapse or recurrence in persons in whom CBT was added later on (with or without continuation of the antidepressant) was 0.8. 

This is the ratio of the risk of relapse/recurrence in persons who received sequential treatment divided by the risk in persons who received control treatment. Since the ratio is less than one, the risk of relapse/recurrence was obviously lower in the sequential treatment group, right?

2. Next, they looked only at the persons who were randomly assigned to add CBT later but to also continue the antidepressant. They were compared to those in whom the antidepressant or “usual treatment” were continued but without addition of CBT. “Usual treatment” means treatment at the discretion of the clinicians and may include an antidepressant.

The risk of relapse/recurrence was lower in those in whom CBT was added after the acute phase; the risk ratio was 0.8.

3. Lastly, what about persons in whom the antidepressant was tapered off after acute phase response and only CBT was then given? These persons were compared to those in whom only antidepressant or “usual treatment” were given.

To me, this is the most interesting comparison. Persons who were taken off the antidepressant (after good acute phase response) and given CBT instead were significantly less likely to have a relapse/recurrence than those who continued on the antidepressant or had “usual treatment” but without addition of CBT; the risk ratio for this was 0.7.


The sequential use of medication followed by CBT is a viable strategy for preventing relapse/recurrence in persons with major depressive disorder.

Clinical Commentary

Given the high prevalence of MDD, the critical importance of preventing relapse/recurrence, and the reluctance of many patients to take medications longer-term, the sequential treatment approach is potentially of great clinical significance.

I hope that the field will pay a lot of attention to further discussing, researching, and using this approach.

Of course, the authors are not saying that the antidepressant can be tapered off in all patients after acute phase improvement. They are only saying the following:

1. Once the person with MDD has improved significantly with antidepressant treatment, it may be felt that since the antidepressant alone has done so well, why not just continue the antidepressant alone? However, this study shows that if now CBT is added along with the antidepressant, the risk of relapse/recurrence is reduced compared to only continuing the antidepressant.

2. In some patients, after significant acute phase improvement on antidepressant alone, it may be appropriate to taper off the antidepressant and provide CBT alone. In fact, in this meta-analysis, this strategy actually did better than continuing the antidepressant. So, we should accept this as one of the possible treatment options rather than insisting that all patients must continue on the antidepressant.

An important point from this paper is that addition of psychotherapy after the person has improved with acute phase antidepressant should be considered in many patients, even if the person’s depression has improved considerably. Usually, psychotherapy is added only in those who have not responded to the antidepressant.

Whether or not the antidepressant can be tapered off at that stage is a complex decision. It should be considered if the person is motivated to actively participate in CBT and wants to be tapered off the antidepressant for any reason. We should keep in mind that studies of any treatment (either CBT or antidepressant or a combination) are able to recruit only patients who are motivated to accept that treatment if randomized to it. If the person is unwilling or unable to do CBT, then switching to CBT is, of course, not a viable option.

Patients who are at high risk of relapse or in whom the consequences of any relapse would be more serious (e.g., higher suicidal risk) should probably be treated with a combination of medication and psychotherapy rather than tapering off either of the treatments.

Expand Content
Share this Section:
Topiramate for patients with schizophrenia?

Correll et al. Efficacy for Psychopathology and Body Weight and Safety of Topiramate-Antipsychotic Cotreatment in Patients With Schizophrenia Spectrum Disorders: Results From a Meta-Analysis of Randomized Controlled Trials. J Clin Psychiatry. 2016 Jun;77(6):e746-56. PubMed PMID: 27337425.

Weight gain is one of the most important problems we face in persons with schizophrenia. It impairs quality of life, increases medical morbidity, and is an important reason for non-adherence to antipsychotic treatment.

Topiramate is often used off-label to mitigate this weight gain. But, how well does it work? And will it help the symptoms of schizophrenia too? What about side effects?


This study aimed to combine many studies on this topic into a meta-analysis of the efficacy and tolerability of the combination of an antipsychotic and topiramate in persons with schizophrenia spectrum disorders.

In this study, the term “schizophrenia spectrum disorders” referred to schizophrenia, schizophreniform disorder, and schizoaffective disorder.


The PubMed/MEDLINE database was searched. Unfortunately, other databases were not searched so some studies may have been missed, especially unpublished studies.

The authors identified randomized controlled trials (RCTs) comparing the combination of an antipsychotic plus topiramate to antipsychotic plus placebo in persons with schizophrenia spectrum disorders.

Two evaluators extracted data from the papers, which is the right way to do this to reduce errors and possible bias from just one person doing it.


Eight RCTs that randomized a total of 439 patients were included. So, each RCT included a relatively small number of patients.

Open-label (un-blinded) studies with a control group were included, but were used to evaluate adverse effects only and not to assess efficacy.

The duration of treatment was a mean of 14 weeks, with a standard deviation of 5 weeks.

Patients were randomized to receive topiramate (100 to 400 mg/day) versus placebo (seven RCTs) OR topiramate versus antipsychotic alone (one RCT).

Topiramate was efficacious for total psychopathology. The Standardized Mean Difference, a measure of the magnitude of effect, was 0.57, which indicates a moderate degree of effect.

Topiramate was efficacious for positive symptoms (Standardized Mean Difference 0.56), negative symptoms (Standardized Mean Difference 0.62), and for general psychopathology (Standardized Mean Difference 0.69).

Body weight and Body Mass Index (BMI) were reduced in persons treated with topiramate. The measure of magnitude of effect used for reduction in body weight was Weighted Mean Difference (no pun intended), and it was a reduction of 3.1 Kg (6.8 lbs).

The investigators also tried to see if certain clinical characteristics could predict the benefit of topiramate for reducing total psychopathology and for weight reduction. However, they could not identify any effect of duration of treatment, dose of topiramate, gender, age, baseline Positive and Negative Syndrome Scale score, or baseline BMI.

What about when topiramate is combined with clozapine? When combining topiramate with clozapine was compared to combining topiramate with other antipsychotics, the clozapine-topiramate combination led to greater efficacy than other antipsychotic-topiramate combinations. However, when combined with clozapine, topiramate was not efficacious for weight loss.

The only adverse effect that statistically trended towards being more common on topiramate than on the comparator was paresthesia.

Only one study assessed cognitive effects using formal testing and, surprisingly, found no cognitive adverse effects of topiramate. The reason for this is not known but the authors speculated that one reason could be that the study did not have enough power to find this effect with statistical confidence. Alternatively, these patients with schizophrenia spectrum disorders may already have significant cognitive impairment that may have masked any additional impairment due to topiramate.


Two main findings emerged:

1. Clozapine: The combination of an antipsychotic and topiramate reduced total, positive, negative, and general psychopathology in patients with schizophrenia spectrum disorders treated with clozapine.

2. Other antipsychotics: The addition of topiramate helped reduce weight/BMI in persons with schizophrenia spectrum disorder treated with antipsychotics other than clozapine.

However, larger studies are needed to confirm and extend these findings.

Clinical Commentary

This meta-analysis suggests that clinicians should consider addition of topiramate to clozapine when clozapine alone is not sufficient. Also, it suggests considering topiramate as one option to mitigate weight gain due to antipsychotic treatment.

There are concerns about adverse effects though, including cognitive impairment and the risk of urinary stones (1.5% of adults taking topiramate). Yes, 1.5% is correct.

Clinicians should also be aware that there are a few case reports of worsening of psychosis as well as of new onset psychosis with topiramate treatment.

Expand Content
Share this Section:
Can rating scales be used to predict a repetition of self-harm?

Quinlivan et al. Which are the most useful scales for predicting repeat self-harm? A systematic review evaluating risk scales using measures of diagnostic accuracy. BMJ Open. 2016 Feb 12;6(2):e009297. PubMed PMID: 26873046; PubMed Central PMCID: PMC4762148.

Self-harm is a common and important clinical problem. Persons who engage in self-harm behavior are more likely to repeat such behavior.

An episode of self-harm increases the risk of suicide by between 30- and 100-fold in the subsequent year.

1 in 15 people who have self-harm behavior die by suicide within the next 9 years.


Risk assessment scales combine multiple risk factors into a single instrument.

This review aimed to determine the diagnostic utility of such scales that assess the risk of repetition of self-harm behavior in order to identify those that may be most useful in clinical practice.


 A systematic review of the literature was done. The search strategy was carefully designed and many databases were searched (Good).

The methodological quality of the studies was assessed (Good).

Three reviewers extracted the data independently (Good).

Cohort studies in adults in which the outcome was either repeat self-harm or attempted suicide were included.

Although suicide is the most important outcome following self-harm, only a small proportion of persons who engage in self-harm behavior subsequently attempt suicide. This makes it difficult to assess the predictive ability of these risk scales even in high-risk populations. Therefore, these investigators used either repeat self-harm or attempted suicide as the outcome because the incidence rate is higher and the prediction of repetition may more feasible than predicting only suicide attempts.


Eight papers were included.

The risk scales evaluated in these studies were: Barratt Impulsivity Scale (BIS), Drug Abuse Screening Test (DAST), Edinburgh Risk of Repetition scale (ERRS), Global Severity Index (GSI), Manchester Self-Harm Rule (MSHR), Modified SAD PERSONS Scale (MSPS), ReACT Self-Harm Rule (ReACT), Repeated Episodes of Self-Harm (RESH), SAD PERSONS Scale, Söderjukuset Self-harm Rule (SSHR), and the Suicide Assessment Scale.

Diagnostic parameters for the risk assessment scales were found to be as follows:

Sensitivity 6% to 97%. A very wide range!

Sensitivity was highest for the Manchester Self-Harm Rule (97%), ReACT Self-Harm rule (95%) and the Söderjukuest Self-harm Rule (89%).

Positive Predictive Value (PPV) 5% to 84%. Again, a very wide range!

The Positive predictive values were highest for the Repeated Episodes of Self-Harm score scale at its highest threshold (84%), Global Severity Index (73%), Drug Abuse Screening Test (71%), and the Barrett Impulsivity Scale (70%).  Note: the Repeated Episodes of Self-Harm score was evaluated on inpatients admitted to the hospital for self-harm behavior. Positive predictive value of any test increases as the risk in the population being evaluated increases.

Diagnostic Odds Ratio (odds of the risk scale being positive in those who repeated self-harm divided by the odds of the risk scale being positive in those who did not repeat self-harm) is an overall measure. It takes into account the need to identify as many as possible of those who will repeat self-harm with as few false positives as possible.

The Diagnostic Odds Ratio ranged from 1 to 16 (higher value being better). Also a very wide range.

The Diagnostic Odds Ratio was highest for the Repeated Episodes of Self-Harm scale at the highest threshold (odds ratio 16) and for the Manchester Self-Harm Rule (odds ratio 11).


It was difficult to determine from this review which scale(s), if any, are the most useful for self-harm risk assessment.

The authors concluded that no scales perform sufficiently well so as to be recommended for routine clinical use.

They suggested that risk assessment scales should only be used as an adjunct to clinical assessment.

Clinical Commentary

This study appropriately warns us against relying on risk assessment scales for attempting to predict repetition of self-harm.

If such scales are used it would be best to use a scale with high sensitivity to initially identify persons at high risk of self-harm. Later in the assessment process, a scale with high specificity could be used to reduce the number of false positives.

Most importantly, we definitely must not use risk scales as the main tool in risk assessment.

Please also see the other paper on predicting suicide after self-harm that is discussed this month in GME Research Review.

Expand Content
Share this Section:
After self-harm, how can we predict risk of suicide?

Chan et al. Predicting suicide following self-harm: systematic review of risk factors and risk scales. Br J Psychiatry. 2016 Jun 23.[Epub ahead of print] PubMed PMID: 27340111.

Self-harm behaviors are, unfortunately, common and are an important predictor of subsequent suicide.

An episode of self-harm increases the risk of suicide by between 30- and 100-fold in the subsequent year.

1 in 15 people who have self-harm behavior die by suicide within the next 9 years.

Therefore, it is of tremendous importance to be able to assess the risk of subsequent suicide in a person who has engaged in self-harm behavior.


The relationship between self-harm behavior and subsequent suicide is complex. It is, therefore, difficult to assess the risk of suicide in these persons.

This study is the first systematic review and meta-analysis of prospective studies of a) risk factors and b) risk assessment scales to predict suicide following self-harm.

Risk assessment scales combine multiple risk factors into a single instrument.


The literature was searched for prospective cohort studies of persons who had engaged in self-harm behavior that followed these persons to determine risk of dying by suicide.

For the review of risk assessment scales, the authors included studies in multiple settings including in persons receiving treatment from mental health specialists.


Twelve studies on risk factors and seven studies on risk scales were included.

Four risk factors emerged from the meta-analysis. Each of these had a significant effect and the magnitude of the effect for each risk factor was largely independent of other factors.

These four risk factors were ones that are already well-known to us:

1. Previous episodes of self-harm (hazard ratio = 2)

2. Suicidal intent (hazard ratio 3)

3. Physical health problems (hazard ratio = 2)

4. Male gender (hazard ratio 2)

Three risk assessment scales were used:

1. Beck Hopelessness Scale (BHS)

2. Suicide Intent Scale (SIS)

3. Scale for Suicide Ideation.

The data on risk assessment scales was weak. Neither their sensitivity nor their specificity was adequate. For example, the meta-analysis of data on the Suicide Intent Scale found relatively low sensitivity (0.73) and specificity (0.64). What this means is that the Suicide Intent Scale was “positive” in only 73% of persons who subsequently died by suicide and it was “negative” in only 64% of persons who did not.

The positive predictive values for these scales were low--ranging from only 1% to 17%. This means that there were too many false positives. Also, clinicians usually face the challenge of predicting short-term risk rather than long-term risk. The positive predictive values of these risk scales in the short-term are even lower than in the long-term.


The four risk factors are of limited practical use because these risk factors are relatively common in the patients we see (e.g., previous episodes of self-harm).

The authors concluded that “there is no robust evidence to support the use of one risk scale over another, and because all the scales reviewed had a low PPV with significant numbers of false positives these scales should not be used in clinical practice alone to assess the future risk of suicide.”

In fact, they point out that use of these risk assessment scales, or an over-reliance on the identification of risk factors in clinical practice, may provide false reassurance and is, therefore, potentially dangerous.

Clinical Commentary

The authors have done a great job of warning us against relying on crude risk factors or on currently available risk assessment scales. Rather, they recommend comprehensive psychosocial assessment of the risks and needs of the particular person being evaluated.

They note that, “People who self-harm often have complex and difficult life circumstances, and clearly need to be assessed – but we need to move away from assessment models that prioritize risks at the expense of needs.”

“…we may well be putting our own professional anxieties above the needs of service users.”

Please also see the other paper on predicting repetition of self-harm that is discussed this month in GME Research Review.

Expand Content
Share this Section:
Can SSRIs cause glaucoma?

Chen et al. Association of Selective Serotonin Reuptake Inhibitor Use and Acute Angle-Closure Glaucoma. J Clin Psychiatry. 2016 Jun;77(6):e692-6. PubMed PMID: 27135704.

Acute angle-closure glaucoma is a true medical emergency. If it is not treated immediately, it can lead to blindness.

There have been several reports over the years of a possible association between treatment with selective serotonin reuptake inhibitors (SSRIs) and increased risk of acute angle-closure glaucoma.


This study evaluated the potential relationship between recent SSRI use and the risk of acute angle-closure glaucoma.


A case-control study design was used. “Cases” were persons who had had acute angle-closure glaucoma and “Controls” were persons who had not. A case-control study design is the appropriate one for outcomes that are relatively rare.

Acute angle-closure glaucoma is more common in Asians, especially in ethnic Chinese persons. Data for about a ten-year period were obtained from the Taiwan National Health Insurance.

About 1500 cases and about 6000 controls were identified.

Cases and controls were matched to each other on gender, age, and the index year.

Persons who had received at least one prescription for an SSRI within seven days prior to being diagnosed with acute angle-closure glaucoma were considered to be “immediate SSRI users.”

Persons who had not received any prescription for an SSRI were considered to be “Nonusers.”

Statistical adjustments were made for possible confounding factors including use of other (non-SSRI) antidepressants, comorbid illnesses, etc.


A higher percentage of persons in the acute angle-closure group (0.7%) were immediate SSRI users compared to the non-acute angle-closure glaucoma group (0.1%).

Similarly, a higher percentage of persons in the acute angle-closure group were non-SSRI antidepressant users (4%) compared to the non-acute angle-closure glaucoma group (2%).

The adjusted odds ratio for acute angle-closure glaucoma was 5.8 in “immediate SSRI users” compared to “nonusers.”

The adjusted odds ratio was even higher in persons whose SSRI dose was higher.


The study suggests a strong relationship between being on an SSRI and having acute angle-closure glaucoma.

Clinical Commentary

This study adds to the literature on the relationship between SSRIs and glaucoma.

What is the potential mechanism by which SSRIs can cause angle-closure glaucoma? This is not clearly known but SSRIs can cause mydriasis (pupillary dilatation) and serotonin is also involved in regulating the dynamics of aqueous humor.

Expand Content
Share this Section:

Rajnish Mago, MD
Medical Editor, GME Research Review

GME Research Review is a monthly newsletter edited by Rajnish Mago, MD, who is author of "The Latest Antidepressants" and "Side Effects of Psychiatric Medications: Prevention, Assessment, and Management." Dr. Mago selects, summarizes, and provides a clinical commentary on the latest published research in psychiatry. 

We are always carefully evaluating which research papers to discuss in GME Research Review. Have come across a research paper published in the last 6 months that you thought is clinically relevant? Do you want me to analyze it for you and for the benefit of others? Please email Dr. Mago the citation at [email protected]

To contact GME, email us at [email protected]

GME does not provide medical advice. The website and articles are intended for informational purposes only. They are not a substitute for professional medical advice, diagnosis or treatment. Never ignore professional medical advice in seeking treatment because of something you have read on the GME Website. If you think you may have a medical emergency, immediately call your doctor or dial 911.

Log In

Join GME For Free

Become a GME subscriber and gain full access to our extensive library of 700+ psychiatric medical education videos, free CME webcasts, latest research updates, and more. To sample our content, watch the featured videos below.
Join Today!