The search was conducted from March to September 2017. The following databases were analysed: Medical Literature Analysis and Retrieval System Online (MEDLINE - PubMed), Excerpta Medica dataBASE (EMBASE), Physiotherapy Evidence Database (PEDro) and Cochrane Central Register of Controlled Trials (CENTRAL), from their inception to September 2017. The references of retrieved articles were reviewed for additional studies. The following search terms or medical subject headings (MeSH) were used:

• •

  Chronic obstructive pulmonary disease, chronic obstructive airway disease, chronic obstructive lung disease.

• •

  Manual therapy, manipulation, musculoskeletal manipulation, joint mobilisation, chiropractic, chiropractic manipulation, osteopathic manipulation techniques and stretching.

The main investigator (M.Vig.) conducted the search of the bibliographic databases, screening all titles and/or abstracts for the inclusion criteria, and identified and removed duplicates. Titles and abstracts of studies were screened by the main investigator to exclude non-eligible studies. Full-text articles were retrieved where the abstract suggested a potentially eligible study. When details were missing from the abstract, full-texts were retrieved and checked for eligibility. Two independent reviewers (M.Vig., C.S.) assessed full-texts for eligibility. Discrepancies were resolved by discussion. The main investigator maintained records on all studies not meeting inclusion criteria, providing the rationale for their exclusion.

We included RCTs with full-text in English language, according to the following criteria:

Participants: adults with COPD, without any age or disease severity restriction. Studies with mixed population (e.g. combined asthma and COPD) were excluded.

Interventions: RCTs evaluating any form of MT as defined above5 were eligible. Studies were excluded if MT was not delivered through physical hand contact (e.g. by mechanical tools or devices) or consisted only of treatment with complementary and alternative medicine such as point application, acupressure, acupuncture, reflexology, Chinese herbal medicine, tai-chi, yoga or other traditional Chinese medicine techniques. In addition, studies based on MT interventions consisting uniquely of gentle massage, passive stretching, manual chest physiotherapy (such as chest percussion or vibration), and secretion clearance techniques (such as postural drainage or abdominal thrust) were also excluded. Studies assessing patients during an acute exacerbation were also excluded. No restriction on the professional who carried out the MT was placed.

Comparisons: all RCTs comparing MT to usual care, to sham techniques or light manual interventions (e.g. gentle massage), to PA, exercise alone, or to a combinations of these interventions, were included.

Outcome measures: RCTs assessing any of the following outcome measures: lung function [at least forced expiratory volume at 1st second (FEV1)], exercise capacity by six-minute walking test (6MWT), and HRQL, were included. Safety of the interventions (Adverse events) was also evaluated.

Data collection and items: using a dedicated electronic database with the use of Microsoft Excel software (2010 version, Microsoft, Redmond, WA, US), the main investigator extracted the data and the second reviewer (C.S.) checked and revised the database. All entered data were screened for accuracy. Authors of Papers were not asked for missing data. The information included in the database were: study design, setting, sample size, groups, characteristics of participants, details of interventions such as frequency, intensity, time and/or type (FITT), professionals involved, outcome measures and results. Adverse events were classified as ‘mild’, ‘moderate’ and ‘major’, according to the classification described by Carnes et al.15

Risk of bias within studies: all included RCTs were assessed using the Cochrane Risk of Bias tool.16,17 Overall risk of bias for individual studies (low, unclear, high) was assessed by two independent reviewers (M.Vig., C.S.) according to the Cochrane guidelines.17 Discrepancies had to be solved by consensus with a third reviewer (M.P.).

A flow chart of the selection of studies is shown in Fig. 1. From an initial search of databases, 515 potential studies were identified and another 40 studies were identified from the bibliographies of retrieved articles. After removal of duplicates, 343 titles and abstracts were screened for eligibility, of which 332 were excluded as not meeting the eligibility criteria. Out of the remaining 11, four studies were excluded after review of the full-text, one full-text article was not found. Hence, 6 studies fulfilled the criteria for eligibility and were included in the review. The level of agreement between the two main reviewers was always good and the judgement of the third reviewer was not required.

Figure 1.

Flow chart of selection of studies for the review.

(0.3MB).

The main characteristics of the included RCTs are shown in Table 1. All studies contained a similar hypothesis about the physiological effect of MT: to reduce respiratory muscle hyperactivity and increase the mobility of the thoracic structures involved in respiratory mechanics. All subjects in the included papers were outpatients with moderate to very severe COPD.1 As also shown in Table 1, the intervention schedules varied, consisting of 1–24 sessions lasting 5–45min. As control treatment, all studies18–23 administered sham manoeuvres or a technique that the authors deemed as non-therapeutic, and four RCTs18,22 compared MT also to various exercise programmes or PA.

Table 2 shows the outcome measures assessed. Most studies assessed 6MWT and reported adverse effects. Few study assessed lung function.

Results of the risk of bias analysis are reported in Table 3. All studies were classified as having a high risk of bias. In one study,18 patients were selected from volunteers recruited through advertisements and newspaper ads, while in another paper19 a pulmonologist selected patients for the study and referred them to the medical centre where they were then randomised and treated. Fig. 2 provides details of the heterogeneity of the characteristics of the interventions: some studies did not state the duration of the sessions,20,21 or did not describe the technique.23Fig. 3 shows the heterogeneity of study designs, inclusion criteria and outcome measures among the included studies. Study designs and sample sizes were described properly in all studies, COPD was diagnosed according to criteria other than GOLD in two papers19,23 or without any clear criteria in another.18

Table 3.

Risk of bias of included studies based on Cochrane Risk of Bias tool 17

Domain		References
		Engel et al. 18	Engel et al. 19	Wada et al. 20	Rocha et al. 21	Zanotti et al. 22	Noll et al. 23
Selection bias	Random sequence generation
	Allocation concealment
Performance bias	Blinding of participants and personnel
Detection bias	Blinding to outcome assessment
Attrition bias	Incomplete outcome data
Reporting bias	Selective reporting
Other bias	Other sources of biasa
Sum

Legend:

, low risk of bias;

, unclear risk of bias;

, high risk of bias.

Note: aOther sources of bias include no comparison of baseline data between controls and intervention group, and lack of description of the MT intervention.

Figure 2.

Details of the study protocol as described in the reviewed studies.

(0.18MB).

Figure 3.

Details of study design, inclusion criteria and outcome measures across the reviewed studies.

(0.17MB).

Due to the heterogeneity of the included studies, it was not appropriate to conduct a meta-analysis of the results.

A summary of the main results is shown in Table 4. Noll et al.23 investigated the immediate effects after a single manoeuvre, while all the other studies18–22 investigated the effect of a comprehensive programme. No significant effect on FEV1 was found. Analysing exercise capacity, a significant increase in 6MWT was found in three studies.18,19,22 In the studies18,19 assessing HRQL as an outcome measure, an improvement was reported only by Engel et al.18 who found a reduction in the dyspnoea domain alone. Concerning adverse events, Zanotti et al.22 reported no adverse events or side-effects from osteopathic manipulative therapy, while three studies18,19,23 reported mild adverse events after MT (muscle soreness) in 15% and 0.005% of MT sessions in the two studies by Engel et al.,18,19 respectively, and in two patients out of 35 in the study by Noll et al.23

This systematic review investigated the current evidence for effectiveness of MT interventions targeted at musculoskeletal system in COPD patients. The analysis of the six RCTs included18–23 show high heterogeneity of study design, patient sample, MT schedule, and outcome measures assessed. All studies showed a high risk of bias. Interventions were feasible, with only mild adverse events. However, the studies showed no effects on airway obstruction and inconclusive results in static volumes. Inconclusive results were found concerning exercise capacity and only one study showed benefit in HRQL. As a consequence, at present there is no evidence to support the inclusion of MT interventions in standard PRP for patients with COPD.

Our results confirm those by Heneghan et al.,5 which found that pulmonary function change minimally after MT and stated that evidence supporting MT is lacking. Wearing et al.24 instead found that MT, administered in addition to exercise, increased forced vital capacity and functional capacity, but that review24 was focused only on joint manipulation techniques, whereas our review and that by Heneghan et al.5 included a wider variety of techniques. However, that review5 included only MT techniques performed by physical therapists, while we did not limit the search in relation to the professional carrying out MT.

The MT techniques evaluated in the included studies18–23 were either patient- or therapist-dependent. The assumption that a reduction in muscular stiffness and an improvement in joint mobility could produce a COPD-associated decrease of rib cage rigidity was not supported in these studies.18–23 It is unclear whether these manoeuvres may really increase chest wall motion. It has been shown that thoracic high velocity low amplitude techniques improve pain and function in some syndromes such as neck pain,25 shoulder impingement26 and low back pain,27 but effects on chest wall mechanics have not been demonstrated in either healthy or patients with COPD.

The intervention schedules as well as the operator performing them varied among studies. Therefore, it is difficult to compare the interventions, and the contribution of each individual technique to the outcome measures is unknown.

The small sample size of the studies is a serious methodological limitation. It was not clear whether the diagnosis of COPD followed accepted guidelines1 or whether lung function assessments were performed according to standardised methods, which puts into question the comparability of the outcome measures of the studies. No study assessed the specific outcome measures for which MT techniques were aimed at, such as pain. Furthermore, disease-specific HRQL questionnaires were used only in two studies.18,19 Manual therapy was compared to different exercise modalities and schedules but no studies compared MT to a standardised multidisciplinary PRP including high-intensity exercise training.2

No study showed significant effects of MT on airway obstruction. In addition, as a potential adverse effect, in one study23 MT resulted in an immediate increase in static volumes confirming a previous report by the same authors28 but in contrast to another study showing a reduction in static volumes.22 Despite the fact that exercise capacity, as assessed by 6MWT, improved significantly in 3 out of 5 studies18,19,22 the related mechanisms such as a reduction in dyspnoea were assessed only in one study.18 The HRQL did not significantly improve in any study.

Missing data were not requested from the authors of papers. Gathering more specific data on the applied technique and on the diagnostic criteria of COPD used, could have reduced the level of bias. Furthermore, the inclusion of RCTs only, and the exclusion of non-English language articles may have induced a publication bias. Another limitation may be due to the fact that various professionals, with different educational paths and different experience, performed MT. Limiting the search to MT manoeuvres performed by specific professionals (e.g. physical therapists) can lead to different results.

The present evidence does not support the inclusion of MT in PRP in COPD patients. A standardisation of MT techniques, methods and schedules as well as of outcome measures is necessary to provide a common language, which may be shared across different countries.29 Also the professional carrying out the MT should be always clearly specified. An effort should be made to tailor these techniques to the individual patient,30 and to different COPD phenotypes.31 Future studies should also investigate the effects of MT on symptoms, such as pain, and disability related to musculoskeletal disorders in COPD.