Abstract, title and references.

The goal of the study, as well as its structure, is clear. Title of the manuscript is clear and relevant to the study. References are relevant, recent, referenced correctly. Key studies are included in the reference list.

Introduction/background

The introduction section consists of four well-structured paragraphs. In the first paragraph cervicogenic headache (CH) is defined and its patterns are described. In the introduction section, Dunning and colleagues provided the reader with the information what their article is about. They clearly introduced the topic of their research, explained why current research is important and stated the purpose of the research question. They clearly stated the gap in the field of the research and the hypothesis of the research. Special terminology is defined; technical terms (manipulation, mobilization) are explained, acronym (CH) is written out in full. All the important references in the field of the topic are included.

Methods.

In the “methods” section, according to their aims, the authors applied a convenient purposeful method of sampling. They used acceptable inclusion and exclusion criteria, based on convenient diagnostic criteria, and presented (in the “results” section) a clear flow diagram of patient recruitment and retention. They clearly described standardization of the testing and treatment procedures and the experience of physiotherapists, taking part in the study.
To measure the effect of treatment, they used reliable and valid questionnaire (the Neck Disability Index [NDI]) and scales – (Numerical Rating Scale [NRS]) and Global Rating of Change Scale (GRC). Methods of data analyses are appropriate and clearly described.

Results.

In the “results” section, mostly, tables and figures are relevant to the study and data are presented in an appropriate way. However, I have two questions to the author concerning presenting data in table 1. 1) Page 7. Table 1. The third row. In the table1 duration of symptoms is presented in days. I can hardly imagine that patients with CH were keeping their everyday diary, marking each day of headache. I would prefer the authors to explain how they calculated these data. 2) Table 1. The sixth row. In table 1 disability (NDI 0–50) is presented in points. However, some patients (who are not drivers) keep section 8 of the NDI (driving) unfilled. In such cases, it is recommended to calculate disability in %, using the formula overall score/45 × 100%, where 45 is the maximal number of scores after filling nine sections of the NDI. I’d prefer the authors to clarify, where there any patients who did not fill section 8? If “yes”, why the authors prefer to present data on total disability score in points, instead of %?

The text of the results complements the data. The authors clearly showed statistically significant results between two groups – treatment with a neck and thoracic manipulations (1-st group) has a statistically larger effect than treatment with mobilization and exercises (2-nd group). However, how it became clear today, these results were incorrectly interpreted later in clinical guidelines for non-pharmacological treatment of cervicogenic headache: “spine manipulations are more effective than multimodal care that is why clinicians should not offer a multimodal program of care that includes a combination of exercise, spinal manipulation, and spinal mobilization” (P. Cote et al., 2019), This is an example of incorrect interpretation of statistically significant differences as clinically important results, because we can consider combination of mobilization and exercise not as “multimodal program”, but as insufficient treatment. The authors discuss this point later in their discussion section.

However, this is not a mistake of the authors, but an example of misinterpretation in the later systematic review.

Discussion and conclusion

The results discussed from multiple angles and placed into context without being overinterpreted. The conclusion does answer the aim of the study. The conclusion is supported by results and references. The weakness and limitations of the study are clearly discussed and are not fatal, they can stimulate further investigations on this topic.

Overall

To summarise: the study design is appropriate to answer its goal. The study approved our knowledge on the effectiveness of spinal manipulation in the treatment of cervicogenic headache and shows some benefits in applying manipulation techniques over the combination of mobilization techniques and exercises. However, in clinical practice, these techniques – manipulation and mobilization – are combined to achieve a better clinical effect. The article does not have major flaws as is consistent with itself.

Overall statement or summary of the article and its findings

The article “Upper cervical and upper thoracic manipulation versus mobilization and exercise in patients with cervicogenic headache: a multi-center randomized clinical trial” constitutes as the first high-quality clinical trial that not only confirms the effectiveness of spinal manipulation in treatment of cervicogenic headache but also shows some comparative advantages of spinal manipulation over combination of mobilization techniques with exercises. However, in clinical practice, manipulation and mobilization techniques are often applied together and combined with recommendations to perform special exercises.
In the comparison of two groups of patients (manipulation vs mobilization + exercises) statistically significant differences were observed. However, these findings should not be mistakenly extrapolated into clinical decision making as recommendations not to allow mobilization + exercises in the treatment of cervicogenic headache.

Overall strengths of the article and what impact it might have in your field

The high-quality design of this randomized clinical trial gives the possibility to consider spinal manipulation technique as an effective method for the treatment of cervicogenic headache.

Specific comments on the weaknesses of the article

Major points in the article which needs clarification, refinement, reanalysis, rewrites and/or additional information and suggestions for what could be done to improve the article.

1. I would prefer the authors to clarify how did they calculate the duration of cervicogenic headache in days.
2. I would prefer the authors to clarify if there were patients, who did not fill section 8 (driving) and if “yes”, why the authors did not show the overall disability score in %, as it was recommended by the developer of the NDI.