There are a number of limitations t

There are a number of limitations to the study. First, a study without a comparison group does not allow for inferences to be made regarding cause and effect. Therefore, as a result of a lack of control group it cannot be determined if the CPR predicted response to treatment or simply identified patients with a good prognosis. Future randomized clinical trials will be needed to validate the variables in the CPR before it can be suggested for widespread clinical application. In fact, it has been stated that single-arm CPRs are vulnerable to a regression effect, where the variables entered into the logistic regression may have resulted in over-fitting of the model, which can lead to spurious findings. However, in the development stage of a possible CPR, it is important and necessary to include all potential predictor variables. Nevertheless, as is the case with all statistical modeling, the results presented here will require validation, which can include performing the study on an independent sample of patients. Further, based on the prospective cohort design, it is also possible that the identified prognostic variables were simply predictors of the passage of time rather than contributed to the manual therapy program, although this is unlikely as the history evolution of the pain in our sample of TTH patient was stable and chronic (mean years with pain 7.3). Therefore, it is essential for all CPRs to be validated prior to widespread clinical implementation. Future clinical trials will be necessary to determine if the variables identified in the current study indeed predict response to treatment in a patient population with headache. In addition it is also possible that we did not capture every possible variable that could be a potential predictor during the examination, eg, other psychobiological variables. Second, we should recognize that we collected only data for short-term outcomes on these individuals and after a single session of physical therapy. We do not know whether those patients classified as having a successful outcome were still doing well at a longer-term follow-up, and if some patients classified as having a nonsuccessful outcome with 1 session can be classified as having a successful outcome with consecutive treatment sessions. Third, patients were recruited from 3 different hospitals involving 3 different experienced neurologists, which might affect the reliability of the diagnosis. However, Leone et al reported a substantial to almost perfect inter-rate reliability (kappa between 0.71 to 0.81) for the diagnosis of TTH according to the IHS diagnostic criteria by experienced neurologists. 70 In our study patients had to describe all pain characteristics of TTH, which can increase the reliability of the diagnosis. Finally, only women were included as TTH is much more prevalent in women than in men. We do not know if the predictors would be the same in men with TTH. Future studies are needed to elucidate these questions.