Types of studies.

This research included single- and multi- center randomized controlled trial (RCT), quasi-RCT and controlled clinical trials. Interrupted time series (ITS) studies, controlled before and after (CBA) studies, observational studies, cohort studies, cross-sectional studies, case-control, case-series and case-report studies were also included because of the lack of controlled studies on the topic. Study reports must have been written in English. Abstracts were also excluded.

Types of participants.

This review included subjects with medical conditions classified also as CID, of either gender and any age. Moreover, studies with animal models were included in order to provide a wide variety of casuistry.

Types of interventions.

Included studies had to assess the effect of OMT compared with one or more of the following control groups: no treatment, placebo/sham, usual/routine care, or waiting list control. Interventions could be applied alone or in addition to conventional treatments (i.e. pharmacological co-interventions, counseling or advice prescription).

Types of outcome measures.

The primary outcome for this review was to quantify the effectiveness of OMT in patients with medical pathologies associated with CID compared with any other type of complementary medicine or usual medical care.

Secondary outcome was to analyze data on side effects caused by osteopathic treatment.

Search methods for identification of studies.

The identification of the studies was conducted by a comprehensive computerized search of MEDLINE (http://www.ncbi.nlm.nih.gov/pubmed), Scholar google (http://scholar.google.it), SCOPUS (http://www.elsevier.com/online-tools/scopus), clinicaltrial.gov, chiloras/MANTIS, OSTMED.DR (http://ostmed-dr.com/), Osteopathic Research Web (http://www.osteopathic-research.com/) and the Cochrane Library (http://www.thecochranelibrary.com). Other sources were considered as follows: grey literature, national trials registers, web searching, conference proceedings. Search terms included: osteopathic manipulative treatment(Free Terms), chronic inflammatory disease(MeSH Terms), bronchitis(MeSH Terms), pulmonary disease, chronic obstructive(MeSH Terms), asthma(MeSH Terms), pelvic inflammatory disease(MeSH Terms), prostatitis(MeSH Terms), otitis(MeSH Terms), vestibular neurotides(MeSH Terms), middle ear inflammation(MeSH Terms), neuropathy(MeSH Terms), polyradiculoneuropathy, chronic inflammatory demyelinating(MeSH Terms), myelitides(MeSH Terms), brain inflammation(MeSH Terms), inflammatory bowel disease(MeSH Terms), irritable bowel syndromes(Free Terms), gallbladder inflammation(Free Terms), gastritides(MeSH Terms), atherosclerosis(MeSH Terms), vasculitides(MeSH Terms), mediastinum inflammation(Free Terms), cardiomyopathy, arteritides, bone inflammation(Free Terms), arthritides(MeSH Terms), rheumatic disease(Free Terms). The research was conducted from journal inception to January 2014.

Duplicate records were identified in EndNOTE and eliminated.

Selection of study.

The two reviewers (LC and MM) with expertise in osteopathic medicine conducted independently study selection based on the explicit search strategy. Discrepancies were resolved by consensus with FC as an arbiter. There were discussions about two studies that in the end were excluded.

According to inclusion criteria, reviewers independently screened titles and abstracts and full-text were retrieved and assessed.

Data extraction and management.

Data extraction was performed independently by the two reviewers, in terms of number of patients, type of interventions, study results and all the other descriptive characteristics of the included trials. All disagreements were discussed and reported by consensus. If data was not reported in the study, the author was contacted. All the analyzed data was stored in a dedicated hard disk, accessible only by the two reviewers.

Assessment of risk of bias in included studies.

Each study was, independently, evaluated by the two reviewer authors. According to the Cochrane methods, the risk of bias was categorized in low, high and unclear across the following domains: Sequence Generation; Allocation Concealment; Blinding to Personnel; Blinding to Outcome Analysis and other bias [18]. Tools proposed by the Cochrane Effective Practice and Organization of Care Group were used to evaluate the risk of bias for CBA and ITS studies [19].

Measures of treatment effect.

For continuous data, mean differences with 95% confidence intervals (MD; 95% CI) were used. For dichotomous outcomes, results were presented as relative risk (RR) with 95% confidence intervals (CI).

Dealing with missing data.

With regard to missing data, the authors were contacted for more information also to know the reason for the absence. Where missing data were known the reasons were described.

Assessment of heterogeneity.

Studies were not pooled if there was significant heterogeneity. This may be methodological, statistical or clinical heterogeneity. Heterogeneity was assessed using the I² statistic, which assessed how much of the variation between studies is due to heterogeneity rather than to chance [20]. Values over 50% suggested substantial heterogeneity, and values over 75% suggested considerable heterogeneity, but its significance also depended upon the magnitude and direction of the effect, and the strength of the evidence, e.g. the p value from a statistical test. Funnel plots were used to help identify possible publication bias [20].

Data synthesis.

An intention-to-treat analysis i.e. including all those randomized to their original groups, whether or not they remained in the study, was conducted. Data was reported as mean, point estimate, percentage and range. Dispersion was presented as standard deviation (SD) and 95% confidence interval (CI). The relative risk (RR) was greater than 1 if more patients were successfully treated by the osteopathy group compared with the intervention group. An estimated pooled weighted average of RRs, using the Mantel-Haenszel fixed-effect method, with a 95% CI, was calculated. Where meta-analysis was not possible, results were presented using summary and descriptive statistics. The software used for statistical analysis was Review Manager v. 5.2.6.

Asthma.

Two studies used the variation of Peak Expiratory Flow (PEF) as primary outcome [21, 22].

Bockenhauer et al [21] enrolled 10 adult smokers and non-smokers women (mean age = 47) with chronic asthma, and assigned them randomly to OMT and sham therapy. Patients acted as their own control, thus underwent both OMT intervention and sham therapy in two sessions, one week apart. OMT consisted of four techniques administered in sequential order. Sham therapy consisted of gentle manual pressure applied on different bodily regions. PEF assessments were performed 15 min before and after each intervention by two examiners. Between and within group analysis demonstrated no statistically significant differences (Tables 1 and 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Overview of included studies of osteopathic manipulative treatment for chronic inflammatory diseases.

https://doi.org/10.1371/journal.pone.0121327.t001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Overview of main findings and side effects of included studies.

https://doi.org/10.1371/journal.pone.0121327.t002

Chronic obstructive pulmonary disease (COPD).

Three studies addressed the effectiveness of OMT on COPD [23, 24, 30].

Noll et al. [30] enrolled 35 elderly patients (mean age = 72) with primary diagnosis of COPD hypothesizing that a single multi-technique OMT session would produce an immediate effect on pulmonary function parameters compared to light-touch sham control treatment. A nonparametric ANCOVA showed statistically significant mean differences between groups in eight of the 21 pulmonary function parameters analyzed (Tables 1 and 2).

The day after the subjects received the treatment, a telephone survey was conducted with the aim at assessing the success of blinding, subjective perception of the intervention received and adverse effects (Tables 1 and 2).

Another piece of research from Noll and colleagues [23] was conducted looking at the immediate effect of four single-technique treatment sessions on a sample of 25 COPD patients (mean age = 68; SD = 8). This RCT, comparing four OMT groups to a minimal touch control, showed different results according to the type of technique used (Table 2). Overall, the use of any single osteopathic technique was associated with a moderate post treatment decrease of pulmonary function.

Zanotti et al. [24] included 20 patients with severe COPD aged 64 (SD = 5). The primary outcome measure of their study was the mean change of 6MWT in people treated with four OMT sessions plus pulmonary rehabilitation (PR) compared with people treated with PR only over a period of one month. Secondary outcomes were baseline changes in pulmonary parameters.

Results demonstrated that combining OMT with PR produced an additional gain in 6MWT of 49m (95% CI 17 to 81m). In addition, OMT plus PR led to a significant improvement of the residual volume compared to control group (Tables 1 and 2).

A meta-analysis was conducted on these three studies [23, 24, 30] with the aim to evaluated the efficacy of the type of osteopathic techniques on three pulmonary parameters: forced expiratory volume in the first second (FEV1), forced vital capacity (FVC) and residual volume (RV). Results showed no statistically significant differences between techniques on FEV1, FVC and RV parameters (Fig. 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Forest plot of comparisons, OMT for COPD parameters.

Outcomes: A, forced expiratory volume in the first second (FEV1); B, forced vital capacity (FVC) and C, residual volume (RV). CI, confidence interval; SD, standard deviation.

https://doi.org/10.1371/journal.pone.0121327.g002

Peripheral arterial disease (PAD).

The only study published on PAD was conducted by Lombardini et al [27]. Authors investigated whether 6-months osteopathic treatments, in combination with lifestyle modifications and pharmacological therapy, could improve endothelial function and functional performances in a sample of 30 adult male patients (mean age = 69; SD = 8) with PAD and intermittent claudication.

Within group analysis showed significant changes only in the OMT group when compared end-of-study and baseline endothelial and functional values (Tables 1 and 2). Moreover, OMT group significantly improved health scores (p < 0.05).

Irritable bowel syndrome (IBS).

Three studies were carried out enrolling patients with IBS [25, 26, 28] (Tables 1 and 2).

Hundscheid et al [26] conducted a RCT on 36 patients (mean age = 43) with IBS investigating the effect of five osteopathic sessions compared with standard care in changing symptoms and quality of life. The study used the standardized IBSQOL 2000 questionnaire and the Functional Bowel Disorder Severity Index as outcome tools. After six months, results showed a significant decrease of Functional Bowel Disorder Severity Index in the OMT group compared with the standard care group. Furthermore the IBSQOL score increased significantly in the OMT group compared to the standard care group.

Florance et al [28] conducted a RCT enrolling 30 subjects affected by IBS (mean age = 48, SD = 17). Results showed that osteopathic treatment improved quality of life and reduced the severity of IBS symptoms by 33% after 7 days and 25% after 28 days. Sham procedure reduced the severity of IBS by 16% at day 7 and 24% at day 28. Secondary outcomes concerning psychological aspects (depression and fatigue) did not show any significant difference at any time point between groups. Level of satisfaction was higher in the OMT group compared to control.

Attali et al [25] included 31 patients with refractory IBS aged 50 (SD = 2) and randomized them in a three sessions of visceral osteopathic manipulation and three sessions of sham therapy. Outcomes were baseline changes in constipation, diarrhea, abdominal distention, abdominal pain and depression. After 10 weeks, authors claimed that visceral osteopathic manipulation produced an amelioration of symptoms associated with IBS and a reduction of rectal hypersensitivity. No significant results were reported for colonic transit time.

Arthritis.

Considering arthritis, the only research carried out was performed by Hallas et al [29]. Authors created a unilateral arthritis rat model and assigned 18 animals to treated (OMT) and untreated groups. A healthy group (n = 8) was included as control. The primary outcome of the study was to determine whether non pharmacological treatment (OMT + exercise) could produce behavioral and biomechanical changes in treated subjects compared with healthy controls (untreated group).

Results showed significant increase in foot-based stride length at the end of the experiment in the treated group versus the untreated group which showed no changes in the same parameter over time (Tables 1 and 2). Furthermore, in the ankle-based stride length assessment, the OMT group increased the stride length over the untreated group but with a non significant difference. The OMT group also showed significantly greater improvement in vertical ankle lift over time compared to the untreated arthritic group (Table 1).

Side Effects.

Only seven studies reported data on side effects [21, 23–26, 28, 30]. In five studies [21, 24–26, 28] none of the participants showed side effects after osteopathic treatments (Table 2). In the study conducted by Noll et al. [23], 14 subjects over 25 reported mild side effects characterized by musculoskeletal soreness or pain. Post-hoc calculation of RR showed a reduction of side effects in the minimal touch control group compared to all other groups (data not showed). A further study conducted by Noll et al [30] reported two patients with symptoms of muscle soreness after the OMT session, while in the sham group the side effects were recorded in four subjects who reported “elevated blood pressure in the morning”,”mild heart palpitations”, “a little muscle soreness” and “back was a little sore”. Again, post-hoc RR computations demonstrated no significant reduction of side effects in the study group compared to controls (data not showed).

Interventions.

Considering osteopathic treatment and control procedures, high heterogeneity was revealed across studies in terms of: type of techniques used, length of the session, dose and duration of manipulations, practitioners’ background, settings for interventions and type of control group.

Bockenhauer et al [21] utilized pre-determined manual protocols based on four OMT techniques applied in the following sequential order: balanced ligamentous tension in the occipito-atloid and the cervicothoracic junctions; Still’s technique for “upward displacement” of the first rib; direct action release of “lower rib exhalation” restriction, and diaphragmatic release. The entire intervention last 10 to 15 minutes. The distance between treatments was at least 1 week. Sham procedures took place in the same room, with the subjects in the same position of the OMT session. Sham techniques in this study consisted in gentle manual pressure to the region of the thoracic outlet, occipito-atloid and cervicothoracic junctions and epigastric region. The upper extremities were circumducted at the shoulder through a partial range of passive motion. No details on settings and practitioners who carried out treatments were provided.

Guiney et al [22] used three pre-determined osteopathic manipulative techniques administered in random sequence: rib raising, muscle energy for ribs and myofascial release. An osteopathic physician performed all osteopathic sessions. Number of treatments, dose, length and period of treatment were not reported. An allopathic physician performed the sham therapy using soft touch on different bodily regions such as rib cage, paraspinal muscles and abdominal diaphragm areas mimicking the osteopathic techniques without applying any type of therapeutic process. No details on settings were provided.

Attali et al [25] performed a pre-determined standardized approach consisting of one general visceral technique in association with local techniques on hypersensitive areas of the body and one sacral technique. OMT group received 3 sessions, separated by two weeks interval, lasting 45 minutes each. As far as the placebo treatment is considered, the practitioner performed a superficial abdominal massage in the same areas of the OMT procedure, without any internal organ mobilization. Number of treatments and length of sessions overlapped the OMT criteria. All the manipulations were performed by a single senior osteopath. No details on settings were reported.

Hundscheid et al [26] used a patient need-based approach applying five osteopathic sessions one every 2–3 weeks. One osteopath was in charge of performing OMT. The standard care group received a diet rich in fiber. Moreover, extra fiber and laxatives could have been administered to the standard care group in case of constipation and loperamide was used if necessary in case of predominant diarrhea. In case of abdominal cramps, mebeverine was prescribed. No sham osteopathic treatment was applied. The osteopathic sessions were performed in a private osteopathic practice whilst standard care was administered in a public hospital. No details on length of each session was provided.

Florance et al [28] used a standardized osteopathic procedure consisting in three techniques applied in sequential order: one direct technique where the osteopath applied manual pressures on each segment of the spine for 90 seconds, followed by one indirect technique where the osteopath manipulated the spine segment in all directions and one final visceral technique to relieve imbalances between the motion of all the organs. Patients received two sessions of manipulative treatment at a 7-day interval. Each session lasted 60 min. One single osteopath administered OMT. The sham procedure consisted of a gentle massage in the same areas treated in the OMT sessions. No information on settings was provided.

Lombardini et al [27] planned eight sessions in six months with a session treatment time of approximately 30 minutes. Sessions were organized as follows: every two weeks during the first two months, every three weeks during the forth, fifth and sixth months. A wash-out period of one month was scheduled between month two and three. The OMT sessions were performed by an osteopath while physicians and nurses collected blood samples and assessed vascular parameters for the control group. The study was conducted in two adjacent rooms of a public hospital unit. No details on osteopathic approach were reported.

Zanotti et al [24] used a patient need-based approach programming one treatment per week for four weeks. Each session lasted 45 min. The examination procedures were conducted by an osteopathic practitioner with emphasis on neuromusculoskletal system. Information on soft manipulation procedure, control group and setting of the interventions were not provided.

Noll et al [23] applied five single standardized treatment protocols in random order. Techniques used were: minimal touch control, thoracic lymphatic pump with and without activation, rib raising and myofascial release. Between each treatment protocol session, a 4-week washout period was planned. The duration of each treatment protocol ranged from five to 10 minutes depending on the type of technique. The osteopathic treatments were administered in an outpatient office by two osteopathic physicians. The pulmonary function tests were conducted by certified respiratory therapists.

The other research programme conducted by Noll et al [30] administered seven standardized osteopathic manipulative techniques: soft tissue, rib raising, indirect myofascial release, sub-occipital decompression, thoracic inlet myofascial release, pectoral traction, thoracic lymphatic pump with activation. The treatment session duration was 20 minutes. Sham manipulations mimicked in terms of dose, length and duration the OMT session. An outpatient office setting was chosen for the study. No other details were reported.

Hallas et al [29] dealt with rats and utilized a sequence of pre-determined techniques consisting of passive range of motion of the right ankle and knee joint, modified muscle energy techniques and passive myofascial stretch. Techniques were applied 10 times a day for five days a week for six weeks by student physicians with at least one year of training in osteopathic manipulative medicine. The duration of each technique was 10 seconds. The control group performed sessions consisting of five minutes of moderate exercise on a mechanical wheel. Dose and total duration of treatment was consistent with the OMT group. All sessions were performed in the laboratory.

Allocation.

Out of 10 studies included, six were RCTs [22, 24–26, 28, 30] but only two employed an adequate randomization method: Attali et al [25] used random permutations generated by a computer software, while Zanotti et al [24] drawn a randomization based on computer-generated list of random numbers within the range of 1 to 20. The remaining four did not specify the randomization method used [22, 26, 28, 30] (Fig. 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Risk of bias for included studies.

+, low risk of bias;-, high risk of bias,?, unclear risk of bias.

https://doi.org/10.1371/journal.pone.0121327.g003

Two research used a 2:1 allocation ratio [22, 28]. Hundscheid et al [26] allocated patients using closed envelopes. Patients included by Noll et al [30] were randomly assigned to either the OMT or sham protocol group using stratification by the severity of airflow obstruction (Fig. 3).

The remaining four non-RCT studies were designed as observational [23], pilot cross-over [21], laboratory [29] and case control [27], exposed, therefore, to allocation biases (Fig. 3).

Blinding of participants and personnel.

The majority of the studies did not report data on blinding of participants and personnel producing an unclear risk of performing bias [21, 22, 25–27, 29, 30].

One study had a high risk of bias as unblinding of osteopathic physicians and patients [23].

Low risk of bias was revealed for Florance et al [28] as the osteopath was not aware of the clinical situation and was not involved in data analysis. Similarly Zanotti et al [24] claimed that patients remained blinded to the randomization and were not able to determine the study arm allocation (Fig. 3).

Blinding of outcome assessors.

Five studies reported an acceptable blinding of the outcome assessors as all measurements were performed by a clinician or specialized therapists, blinded to the treatment applied [21, 23–25, 30].

All the other five included studies did not report any information [22, 26–29] (Fig. 3).

Selective bias.

Study protocols were not available and although requested, the assessment was not possible to carry out (Fig. 3).

Other bias.

The quality of studies included was further assessed considering other reported information regarding: conflict of interest, reporting funding source, ethical approval, informed consent, confidentiality, declaration of interests, access to data, trial registration, data collection, data management and data monitoring committee. Taking into account conflict and declaration of interests only five studies specified appropriate information [23–25, 28, 30]. Regarding source of funding, only Noll et al [23, 30] and Hallas et al [29] reported complete details on grants received. All studies included apart from Hallas et al [29] declared ethical statement and informed consent approval. None of the research detailed any information regarding confidentiality, access to data, trial registration, data collection, data management and data monitoring committee (S1 Table).