This study assessed the quality of investigations that reported cSMT associated with CAD by determining the frequency with which specific quality factors were described. Because it is recommended that case reports should be included in systematic reviews of relatively rare harms, the quality of these reports is of critical importance in knowledge synthesis. Overall, the quality of case reports examined in this study was low in that they infrequently contained more than 5 of the 11 relevant factors.
Certainly, previous papers have also identified limitations of case reports within the cSMT/CAD literature. A previous review provided limited data regarding subject demographics, time-to-onset of symptoms, and the profession of the cSMT provider 
. Based on this, Kawchuk et al. showed that two of the largest case-series involving cSMT/CAD to date did not report CAD location. Although this information was unreported, the information was available. As a result, a secondary analysis was performed that demonstrated that in cases where CAD was reported to follow cSMT, lesions in the vertebral arteries were not distributed randomly 
. This study demonstrates that by providing clinically relevant factors within case reports, further synthesis toward understanding the relation between cSMT and CAD is possible.
In studying the association between cSMT and CAD, it is important to understand the events preceding the application of cSMT and the onset of CAD. Unfortunately, the results from this study demonstrate a general deficiency in reporting events preceding cSMT or CAD other than that the patient presented to the emergency clinic following cSMT. Clearly, case reports discussing the development of a stroke should endevour to include an etiological work-up so that an alternative cause of the patient’s presentation might also be elucidated. For example, spinal manipulation is a therapeutic modality that is used to treat head and neck pain 
. The reasons for the presentation of head and neck pain may be minor trauma (i.e. motor vehicle accident). If a CAD then occurs following cSMT, it becomes difficult, if not impossible, to identify which event, if any, were associated with the injury. This paper clearly demonstrates that there is a critical need to report all events surrounding CAD, not just the event immediately preceding the injury.
Identifying the stroke type when reporting cSMT associated with CADs would be useful for those practitioners who frequently examine patients presenting to the emergency room (or clinic) with stroke-like symptoms following a therapeutic intervention. Understanding of the typical stroke presentation might help with rapid identification of injury location and assist in the determination of a management protocol. Our study demonstrates that while stroke type was reported often, the anatomical location of the stroke was not. Given that stroke symptoms are specific to the area affected by the lesion, further understanding of the management of patients with strokes thought to be associated with cSMT might occur with increased reporting of injury location rather than simply stating the type of stroke.
The lack of reporting regarding cSMT-specific factors was similar to the under-reporting of other factors. This is an important omission as there is an inherent variability of cSMT techniques used by manual therapists. While some investigators have examined the mechanical forces associated with instrument-assisted spinal manipulation 
, there are few studies that have examined the differences between the various types of cSMT 
. Clearly, understanding the distribution of cSMT and the type of cSMT provided would generate important data regarding the safety of various cSMT procedures.
Our study has demonstrated that there are deficiencies in reporting key factors associated with CAD and cSMT. While the temporality and location of the injury were reported consistently, additional efforts are needed to improve harms reporting so that clinicians are provided with accurate information about various therapies and their potential sequelae. One approach to improved reporting has been suggested by the EQUATOR (Enhancing the Quality and Transparency of health Research) network comprised of researchers working towards improved quality of the published literature. The network hosts an up-to-date library of reporting guidelines for health research on their website (www.equator-network.org
). Currently, at least two sets of guidelines have been identified which if used, may improve the quality of case report literature 
. In addition examples of standardized reporting tools used to evaluate the cause of adverse events exist such as the Consolidated Standard of Reporting Trials (CONSORT) Statement for reporting adverse events in clinical trials, 
and the Naranjo Causality Scale for reporting adverse reactions to pharmaceuticals. 
The use of these standardized reporting tools for monitoring the safety of treatments within clinical trials demonstrates that through effective use of clinical information, the causes of adverse events might be identified early so as to prevent further incidents. Currently, there is no standardized reporting tool for examining adverse events associated with cSMT. Therefore, in the interest of further understanding CAD in relation to cSMT, a standardized reporting tool should be developed. Implementation of consistent reporting of all data features for all CADs may provide clinicians and researchers with more, and better, information to 1) understand which patients are at risk of developing a post-cSMT CAD and 2) possibly decrease the overall incidence of post-cSMT CAD events.
In addition to a standardized tool for reporting cases where adverse events are associated with cSMT, it is important to standardize how the diagnosis of CAD is achieved. The diagnosis of CAD is exceedingly difficult and has to be performed with sufficient quality to ensure that the patient had a CAD rather than some other cause of arterial occlusion, stenosis, or hypoplasia. Failure to diagnosis a CAD accurately places a limitation in the interpretation of the data found in the case report. The diagnosis of CAD should be made from the visualization of a transmural hematoma or a pseudoaneurysm with long tapering stenosis and/or an intimal flap or double lumen 
. Further standardization of the diagnostic criteria for CADs is important for improving the quality of CAD case reports.
A potential limitation of this study was the lack of an existing tool to measure case report quality in this topic area. Given this void, and the recommendations of the Cochrane Collaboration to include case reports in systematic reviews designed to investigate infrequent harms, we developed a list of 11 factors to describe case report quality based on well-established criteria used to explore the relation between cause and effect 
. While this approach provides one way to measure the quality of case report material, it may not be the only relevant way to achieve this goal. As such, future measures of case report quality may arrive at a different conclusion; however, the observed deficiencies of the existing case report literature remain. By collecting and collating information from multiple reports, a better understanding about the association of cSMT, CADs and stroke will be possible.
Another potential limitation of this study is publication bias. Specifically, not every case of cSMT associated with CAD is published in the scientific literature. In fact, previous papers 
present data from medico-legal proceedings that were not published in the academic literature. This bias suggests that there is under-reporting of the cases of CAD associated with cSMT, and suggests that a more complete examination of data should include examination of those cases in medico-legal proceedings.
This paper examined the quality of literature describing an association between cSMT and CAD. Case reports represented the majority of this literature. Since these reports may contribute to further understanding CADs as they relate to manual therapy, it is important that they are of the highest quality. This study has demonstrated that the literature infrequently reports useful data toward understanding the association between cSMT, CADs and stroke. As a result, the value of these reports toward informing our understanding of the relation between cSMT and CAD is minimal. We suggest that through the systematic collection of data features presented in this paper, a clearer clinical picture of the association between cSMT and CAD would be possible. This study lays the groundwork for developing a universal reporting tool for adverse events related to cSMT.