Article Text

Download PDFPDF

Choosing the optimum strategy for rTMS
Free
  1. Toshi A Furukawa
  1. Departments of Health Promotion and Human Behavior, Kyoto University Graduate School of Medicine/School of Public Health, Kyoto, Japan
  1. Correspondence to Dr Toshi A Furukawa, Departments of Health Promotion and Human Behavior, Kyoto University Graduate School of Medicine/School of Public Health, Kyoto 606-8501, Japan; furukawa{at}kuhp.kyoto-u.ac.jp

Statistics from Altmetric.com

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Clinical case

Patient

A 45-year-old man.

Present illness

Mr A had been living a relatively eventless and successful life: after graduating from college, he found a post in the local government, got married with one of the colleagues at the office; the couple had two kids, and the patient was promoted to the section head at the age of 35 years, which is neither early nor late as a typical government employee. However, this section was in charge of the refurbishing of the old city centre, for which Mr A had to negotiate with all sorts of interested parties. The negotiation was endless while the deadline for the project was mandated by the mayor. His first major depressive episode set in, due to which he took a sickness leave for 6 months. After return to the office, he was transferred to a new section where he was exempted from important tasks. Since then he has had several major depressive episodes, with the most recent one lasting now almost 2 years. As he appeared less and less responsive to drug treatments, the psychiatrist in charge referred him to your clinic, which has recently acquired a repetitive transcranial magnetic stimulation (rTMS) device.

Present status

On presentation he was a man of medium build. He had been treated with venlafaxine up to 225 mg/day and escitalopram up to 20 mg/day, with augmentation attempts with aripiprazole and with lithium, all of which appeared somewhat effective in the beginning but every time the patient eventually returned to the same degree of depressive state. His current Hamilton Rating Scale for Depression Score was 24. No physical abnormality, including thyroid function, has been found or reported. He had already received good explanation of rTMS from his treating psychiatrist and is eager to try rTMS.1 You are aware that there exist several administration methods for rTMS but wonder if any is better than the others.

Formulate your clinical question

Patients: Patients with refractory major depression

Intervention: Various administration strategies of rTMS

Comparison: Against each other and against sham rTMS

Outcomes: Improvement in depressive symptoms

Literature search

When there are more than several treatment alternatives for the same medical condition, the network meta-analysis, if properly conducted, can provide the clinician with the most clinically useful information: which of the many available is the most, the second most, or the third most, and so on, effective? You go to the Mesh Database in Pubmed and identify two keywords corresponding with depression and rTMS ((‘Depressive Disorder’(Mesh)) AND ‘Transcranial Magnetic Stimulation’(Mesh)), combine them with free keywords (network meta-analysis). This PubMed search retrieves one reference, Brunoni AR et al. Repetitive transcranial magnetic stimulation for the acute treatment of major depressive episodes: a systematic review with network meta-analysis. JAMA Psychiatry 2017;74:143152.

Critical appraisal of a network meta-analysis

In order to appraise the quality of findings from a network meta-analysis, we will follow the checkpoints slightly modified from recent users’ guides published in Clinical Orthopaedics and Related Research,2 3 supplemented by more extensive checklists by the International Society For Pharmacoeconomics and Outcomes Research (ISPOR)4 and National Institute for Health and Care Excellence (NICE) Decision Support Unit.5

CRITICAL APPRAISAL OF NETWORK META-ANALYSIS

FIRST JUDGMENT: EVALUATE THE CREDIBILITY OF THE METHODS OF A SYSTEMATIC REVIEW

Did the review explicitly address a sensible clinical question?

Yes. The PICO of this review may be summarised as follows:

Patients: Patients with a primary diagnosis of an acute unipolar or bipolar depressive episode.

Intervention and Control: All well-characterised methods of administration of rTMS or sham rTMS.

Outcome: Response (defined as 50% or greater improvement) and acceptability (all-cause dropouts).

Was the search for relevant studies exhaustive?

Unclear. The reviewers searched major databases. The search strategy provided in supplement 2 shows that each contained ‘transcranial’ ‘magnetic’ with methodological filters: as the intervention in question is well defined, this search appears reasonably comprehensive.

On the other hand, they have not searched clinical trial registers nor did they use the Cochrane highly sensitive search strategy.

Were selection and assessments of studies reproducible?

Yes. The selection, the data extraction and the risk of bias assessment were done in duplicate. The inter-rater reproducibility was reported to be a κ of 0.84 for the risk of bias assessment.

Did the review present results that are ready for clinical application?

Yes. The ORs for response were used for the primary end points. This should be readily applicable to the expected event rate of the patient(s) to estimate the absolute benefit that they can expect.

Is the review up to date?

Yes. The last search was conducted in October 2016.

SECOND JUDGEMENT: RATE THE CONFIDENCE IN THE EFFECT ESTIMATES

Ideally these should be rated for each comparison, that is, for each network estimate. In the following we will concentrate on the bilateral rTMS against sham rTMS, which ranked the second best and was relatively well represented in the network (ie, a relatively large number of subjects were allocated to this intervention and were compared with several other treatments).

How serious is the risk of bias for each comparison?

No important bias. eFigure 9A in the supplement of the original publication shows the contribution of high risk, moderate risk or low risk of bias comparisons to the network estimates. The comparison of bilateral rTMS versus sham rTMS received contributions from low (approximately 15%) risk, moderate (approximately 80%) risk and high (approximately 5%) risk of bias comparisons.

In a sensitivity analysis excluding studies at high risk of bias, the OR for the comparison between bilateral rTMS and sham rTMS was essentially unchanged (3.86, 95% CI 2.24 to 6.65).

Are the results homogeneous across studies in direct comparisons?

No important bias. eTable 2 in the supplement of the original publication shows the OR for response for all the 11 direct comparisons between bilateral rTMS and sham rTMS. Although it does not provide I2 or τ for this comparison, the individual ORs from the included studies showed that they were basically at least not qualitatively heterogeneous (ie, all but one studies had ORs smaller than 1.0).

Are the results consistent between direct and indirect comparisons?

No important bias. eTable 6 provides the results of side-splitting tests. The side-splitting for the comparison between bilateral rTMS and sham rTMS was not statistically significant (P=0.547).

How precise are the results?

No important bias. The OR of bilateral rTMS over sham rTMS was 3.96 (95% CI 2.37 to 6.60). The lower limit of the 95% CI was therefore large enough to rule out possibility of statistically significant but clinically meaningless differences.

Do the results directly apply to my patient?

No important bias. Our patient suffered from refractory depression. The review included all types of depression. However, a sensitivity analysis focusing on resistant depression produced similar or greater ORs (5.21, 95% CI 3.27 to 8.30).

Is there concern about publication bias?

Important bias. Visual inspection of the comparison-adjusted funnel plot (plotting all active interventions against sham rTMS) suggested some asymmetry for response.

Based on these ratings, the overall certainty of evidence for the comparison between bilateral rTMS and sham rTMS was judged to be moderate.

What will you do with your patient?

Given the relatively large body of evidence supporting bilateral rTMS, although not meaningfully different from some other methods, you decide to use this approach for the patient.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.

Footnotes

  • Competing interests TAF has received lecture fees from Eli Lilly, Janssen, Meiji, Mitsubishi-Tanabe, MSD and Pfizer. He has received royalties from Igaku-Shoin and Nihon Bunka Kagaku-sha publishers. He has received research support from Mitsubishi-Tanabe and Mochida.

  • Provenance and peer review Commissioned; internally peer reviewed.