J Manipulative Physiol Ther 2011 (Jun); 34 (5): 314–346
James W. Brantingham, DC, PhD, Tammy Kay Cassa, DC, Debra Bonnefin, DC, MAppSc, Muffit Jensen, DC, Gary Globe, DC, MBA, PhD, Marian Hicks, MSLS, Charmaine Korporaal, MTech Chiropractic
Director of Research and Associate Professor, Department of Research, Cleveland Chiropractic College, Los Angeles, CA.
OBJECTIVE: The purpose of this study was to conduct a systematic review on manual and manipulative therapy (MMT) for common shoulder pain and disorders.
METHODS: A search of the literature was conducted using the Cumulative Index of Nursing Allied Health Literature; PubMed; Manual, Alternative, and Natural Therapy Index System; Physiotherapy Evidence Database; and Index to Chiropractic Literature dating from January 1983 to July 7, 2010. Search limits included the English language and human studies along with MeSH terms such as manipulation, chiropractic, osteopathic, orthopedic, musculoskeletal, physical therapies, shoulder, etc. Inclusion criteria required a shoulder peripheral diagnosis and MMT with/without multimodal therapy. Exclusion criteria included pain referred from spinal sites without a peripheral shoulder diagnosis. Articles were assessed primarily using the Physiotherapy Evidence Database scale in conjunction with modified guidelines and systems. After synthesis and considered judgment scoring were complete, with subsequent participant review and agreement, evidence grades of A, B, C, and I were applied.
RESULTS: A total of 211 citations were retrieved, and 35 articles were deemed useful. There is fair evidence (B) for the treatment of a variety of common rotator cuff disorders, shoulder disorders, adhesive capsulitis, and soft tissue disorders using MMT to the shoulder, shoulder girdle, and/or the full kinetic chain (FKC) combined with or without exercise and/or multimodal therapy. There is limited (C) and insufficient (I) evidence for MMT treatment of minor neurogenic shoulder pain and shoulder osteoarthritis, respectively.
CONCLUSIONS: This study found a level of B or fair evidence for MMT of the shoulder, shoulder girdle, and/or the FKC combined with multimodal or exercise therapy for rotator cuff injuries/disorders, disease, or dysfunction. There is a fair or B level of evidence for MMT of the shoulder/shoulder girdle and FKC combined with a multimodal treatment approach for shoulder complaints, dysfunction, disorders, and/or pain.
From the FULL TEXT Article:
In 2008, McHardy et al  published the first extensive systematic review of chiropractic treatment of upper extremity conditions and disorders. McHardy et al required that research articles include “a peripheral diagnosis and chiropractic intervention.” Research articles were excluded “if (1) pain was referred from proximal or spinal sites, (2) the patient was referred for surgical intervention, (3) the condition was not amendable to chiropractic treatment, or (4) a red-flag condition or diagnosis was present (unless post-surgical rehabilitation occurred).” The authors also required that treatment had to be either “peripheral or spinal or a combination of both.”  They further wrote “there is a paucity of literature that describes the singular use of high-velocity, low-amplitude (HVLA) thrust manipulation of the extremities.” 
A number of extremity mobilizations, manipulations, techniques, or “moves” were included in a textbook by BJ Palmer as long ago as 1911; and teaching the use of adjustive extremity technique, including mobilization and adjunctive or multimodal therapies, such as exercise and/or what was later termed physical therapy, can be dated back at least 100 years. [2–5] As others have posited, there is an apparent disconnect between the services chiropractors actually provide, the public perception of the services provided by chiropractors, and what some within the profession believe should be provided.  Although many if not most chiropractors provide various physiotherapy modalities including exercise prescription, electrical modalities and ultrasound, and a range of soft tissue techniques as well as joint mobilization for extremity disorders, medicine and the public tend to focus solely on the traditional HVLA adjustment or manipulation applied to the spine. [1, 6] As many chiropractors actually use a broad multimodal approach to extremity care as outlined above, research should be directed to this broader, more inclusive definition of chiropractic care. [ 1, 7]
Building upon the work of McHardy et al and using similar methodology, structure, and format, this is an expansion and update of that seminal work. The present research review includes additional chiropractic studies subsequently published as well as other similar manual and/or manipulative therapy research. 
For the purposes of this updated and expanded literature review, the term chiropractic has been replaced by manipulative therapy to facilitate inclusion of all similar, related, peer-reviewed literature. For this review, the authors define manipulative therapy as inclusive of all “manual” or “adjustive” procedures and/or therapy that includes grades I to IV++ of mobilization techniques and procedures and grade V manipulation, or HVLA thrust manipulation, with and without adjunctive or multimodal therapy. [1, 8–11]
Since the publication of the McHardy et al review, Bronfort et al  (2010) have published a comprehensive summary of the scientific evidence regarding the effectiveness of manual therapy in the management of a broad spectrum of common musculoskeletal conditions seen by chiropractors including disorders of the spine and the lower and upper extremities, and nonmusculoskeletal complaints. Of interest, Bronfort et al  appraised literature regarding manual therapy for the shoulder. However, Bronfort et al restricted their selection of evidence to only the largest, highest-quality, and methodologically “best” randomized controlled and/or clinical trials (RCTs). They did not consider research that did not meet a stringent level for RCTs nor other types of studies.
[12, 13] Although the Bronfort review is of undoubted value to some, using such a limited number of studies does not fully align with evidence-based medicine or care (hereafter EBC) as conceived by Sackett et al  and others.
[12, 13] For example, the efficacy of a new drug therapy, initially tested in narrowly defined and stringent RCTs, may be later determined to be less effective in clinical practice because of the complexity of the heterogeneity of patient populations, comorbidities, as well as patient compliance. Furthermore, patient and practitioner preferences cannot be taken into account solely through RCTs; yet these variables are often found in different degrees in a variety of other studies. [12–21] There are flaws in most every research study and all research designs; one must be cognizant of these limitations and interpret the findings carefully, not discounting all findings outside of the most stringent of RCTs. Therefore, in the interest of painting as broad a view of the existing evidence, this review will accept a broader range of RCTs, as well as single-group pretest posttest designs (SGPPDs), case series, and case reports, with a consensus view that all are still needed in the context of a larger review as vital components in guiding the delivery of “best patient care” and in developing new lines and areas of research. [12–21]
As Johnson  suggests, Sackett et al  originally developed EBC to improve practice and best patient care, improved practice and best patient care never being intended to be derived solely from RCTs, but rather derived from “tracking down the best external evidence.” Regrettably, one large, apparently well-designed RCT can be misleading, skew and distort knowledge, and do much harm when used unscrupulously out of context. [23, 24] In this regard, Manchikanti et al  have suggested that “the hierarchy of evidence has done nothing more than glorify the results of imperfect experimental designs on unrepresentative populations in controlled research environments above all other sources of evidence that may be equally valid or far more applicable in given clinical circumstances.”
Haldeman et al  and others have suggested that up to 80% of the practice of medicine is still based on and supported by sources with lesser levels of evidence than only large, high- or very high quality, methodologically faultless RCTs. [14, 15, 22, 25–27] Where then, or from what other studies, can such types of evidence be found: information “to improve practice and best patient care … for each individual, taking into account singular, individual clinical characteristics, co-morbidities and personal values and preferences for each particular individual…”? [13–15, 22] As Shacklock  notes in a commentary about a systematic review of manual therapy for neural mobilization “patient-therapist interactions are critical in affecting patient compliance which inevitably produces physical effects in the tissues. So even though this systematic review is appropriately directed at the holy grail (high level evidence), the therapist should not be deterred from using their clinical acumen in dealing with subtle nuances that have not yet been measured. There is simply much more research to be done before we can base treatment on randomized controlled trials and I am not aware of any systematic review or meta-analyses system for evaluating large scale qualitative phenomena, yet.”
A variety of reviews and research looking at current interventions and how often such interventions are actually “evidenced based” has been forced to admit the lack of RCTs, particularly high-level RCTs, in the majority of cases, with consensus-based expert opinion required for best evidence guidelines. [20, 23, 29–31] However, so-called experts can be wrong. [20, 23, 32–34] It must also be acknowledged that, although it appears to be improving, there is as of yet no comprehensive consensus of internationally accepted and fully agreed upon gradated levels of EBC. [35-38]
How is it possible then to practice without RCTs, or to develop a linear understanding of literature gaps, or to develop research to fill those gaps and develop better designed trials and studies to improve best patient care without listing or reviewing all levels of evidence and RCTs? [26, 27, 39–43] In fact, most diagnoses have no RCTs undergirding them to guide practitioners.13 Indeed, one of the RCTs listed in this study was developed directly and indirectly from the McHardy et al review and was further dependent on information generated through the included case series and reports (studies now often and/or generally excluded). [1, 44] The answer is that all levels of evidence, as intended by Sackett et al  and others, must be considered. [ 13, 15–19, 28]
Therefore, building upon the McHardy et al  seminal effort and the recent work by Brantingham et al,  this review has adopted similar methodology using a parallel structure. This present review expands on and updates this work by reviewing all relevant professional sources, including chiropractic literature.  In addition, this review uses Bronfort et al12 and other systematic reviews yet, unlike either previous study, will examine the shoulder alone. While acknowledging the previous work of these groundbreaking 2008 and 2010 reviews, the conclusions in this article are solely those of the included authors. [1, 7, 12]
Various treatments included in this review of manipulative therapy suggest possible alternatives for (a) those who may not or should not have surgery, (b) those who may not or should not chronically use nonsteroidal anti-inflammatory drugs (NSAIDS), and (c) those for whom exercise alone has not been effective. [7, 12, 39, 45, 46] Research into the application of manual therapy techniques has erupted, including intensive investigation by nearly all professions that treat the shoulder with manual and manipulative therapy (MMT) techniques. These investigations with and without multimodal or rehabilitative care include exploration into the most common manipulative method used by chiropractors, HVLA manipulation or thrust technique. [ 12, 25, 47] Broadly revisiting MMT studies to review the quantity, quality, and types of research published is needed, with the goal of ranking, grading, and presenting common characteristics. The purpose of this study to provide an update and a fuller, broader, general, and more expansive review of past, current, new, and innovative multimodal MMT approaches being developed to treat common shoulder disorders, pain, and dysfunction.
This systematic review of MMT for shoulder pain and disorders, in keeping with the intent of EBC, has presented a broader and more complete review of evidence. This intent is to cautiously provide practitioners, particularly in the context of clinical expertise and patient preference, with a more comprehensive picture of the existing evidence supporting a variety of MMT therapies (with and without rehabilitation or multimodal treatment) that may be useful. It is our position that the best approach to patient care is not informed by restricting one solely to the most stringently controlled randomized trials. Evidenced-based care was never meant to exclude all other study designs along the research hierarchy.
Multimodal treatment appears at this time to be the most efficacious approach for shoulder conditions (Table 2-7). This review has shown that MMT, whether grade V HVLA thrust or grades III and IV mobilizations, should be considered for inclusion in the treatment of shoulder pain and disorders, applied appropriately for the benefit, effectiveness, and safety of the patient. Regarding MMT, evaluation of the GH, AC, SC, spinal, upper ribs, and FKC (such as the elbow) joint should be assessed for ROM, accessory glide, and end-range play, feel, or accessory motions. High-velocity, low-amplitude or mobilization grades I to IV (or up to IV++) should then be applied, after an adequate diagnosis has been made and contraindications have been ruled out, in the direction of the restriction when appropriate.
From the results of this review, the clinician should be guided to additionally evaluate the cervicothoracic spine and ribs when treating the shoulder. A number of trials treated the cervicothoracic spine only and reported good outcomes without including GH (or AC or SC) joint manipulation. The segmental fixation of the cervicothoracic spine may refer pain to the shoulder area (from the neck to the arm) or may be partially responsible for inhibition of the lower scapular stabilizers that cause altered biomechanics of the shoulder eventually ending in shoulder pain. [108, 110]
Rarely in clinical practice is there one diagnosis for a given shoulder condition. Often, myofascial soft tissue involvement will be accompanied with joint restrictions and neuromuscular movement dysfunction, which over time may cause tissue injury or failure resulting in a primary joint disorder. Travell and Simons  have revealed pain referral patterns into the shoulder area as a result of myofascial trigger points. One aspect this review did not address is the impact of myofascial adhesions and restrictions on shoulder function. This is a topic that needs more research, as there is some evidence that fascial disorders may have far-reaching effects on function and pain. [92, 112] Treatment that addresses all of these dysfunctions as well as joint estrictions/fixations may be more efficacious in improving function and decreasing pain. For example, the work of Kibler and McMullen  suggests that scapular dyskinesis (an alteration in the normal position or motion of the scapula during coupled scapulohumeral movements) is very often present in the most painful shoulder conditions; rotator cuff injuries have scapular dyskinesis present in 68% of cases and labral tears in 94%, and there is scapular dyskinesis in GH instability in 100% of cases. Scapular stabilization exercise or rehabilitation may often then be the foundation of a shoulder rehabilitation program (requiring scapular MMT and/or shoulder girdle MMT) for success. This review found that scapular stabilization was one of the most common exercises prescribed in the studies reviewed. Please see Table 8 for a description of the most common exercises prescribed in the studies reviewed.
One limitation is confusion surrounding and lack of standardization of the term shoulder girdle. Shoulder girdle has been defined variously by different authors at different times and in the past has been the combination of the GH, AC (including scapular glide), and SC joints and/or including the axial spine (cervical and thoracic spines). In this review, some authors described the shoulder girdle as the cervical and thoracic spines and upper ribs, whereas others used the (previously) more common definition given above. Some included it all. Others defined the shoulder as restricted to the GHJ. This confusion cannot be resolved in this article and may have led to different interpretations of findings in this review. The reader is directed to the particular article cited and Table 2-7 for clarification.  It is also not clear when manipulation is indicated for the spine and not the GHJ, or the GHJ, spine, scapula, and upper ribs; this is explicated in a minor way in the Table 3-7; and again, the reader is directed to the individual articles cited.
Another limitation is use of the WSR. The WSR is not yet demonstrated to be valid and reliable, and the number or WSR “score” that is given must be viewed with caution, should not be quoted as would a PEDro score, and is best seen as how this review deemed the importance and/or strength of the non-RCT study. Ultimately, the WSR score is this review's expert but consensus opinion. Finally, the literature base continues to grow. It is likely that some articles were published after submission and acceptance of this article and therefore were not able to be considered for this review. 
This study found a level of B or fair evidence for manual and manipulative therapy (MMT) of the shoulder, shoulder girdle, and/or full kinetic chain (FKC) combined with multimodal or exercise therapy for rotator cuff injuries disease or disorders (RCIDs). There is a fair or B level of evidence for MMT of the shoulder/shoulder girdle and FKC combined with a multimodal treatment approach for shoulder complaints, dysfunction, disorders and/or pain (SCDP). There is a fair (B) level of evidence for MMT with exercise that included proprioceptive retraining as helpful for frozen shoulder (FS) or adhesive capsulitis. There was a fair level of evidence (B) for MMT using soft tissue or myofascial treatments for soft tissue disorders (ST) of the shoulder. There is a limited level of evidence (C) for cervical lateral glide mobilization (CLGM) and/or HVLA manipulation with soft tissue release and exercise in the treatment of minor neurogenic shoulder pain (NSP). There is an insufficient level of evidence (I) for MMT with or without exercise or multimodal therapy in the treatment of OA of the shoulder. In particular, MMT must be combined, when safe, appropriate, and including no contraindications, with commonly indicated exercise or rehabilitative therapy, as it remains the standard care. For clinicians, however, this study is intended to guide them in the appropriate use of MMT, soft tissue technique, exercise, and/or multimodal therapy for the treatment of a variety of shoulder complaints in the context of the entire hierarchy of available evidence.
Funding Sources and Potential Conflicts of Interest
No funding sources or conflicts of interest were reported for this study.
This review allows a basic comparison of the diversity and commonalities of multiple manual therapy techniques used in the treatment shoulder disorders.
This review, dating from the mid 1980s to 2010, gives a broad overview of the type and quality of previous manual therapy randomized controlled trials and other studies for the treatment of shoulder pain, allowing practitioners to have an increased choice of therapy.
This review helps to elicit the best evidence along with lesser levels of evidence (which may still be useful in, or for, particular settings or patients) and to bring out the gaps in our understanding or literature.
Return to the SHOULDER Page