Frozen Shoulder [Disease]

Methodology:

The authors of this review followed the PRISMA statement guidelines to conduct a systematic review and meta-analysis.

Eligibility criteria

• Types of studies

For our analysis, we used randomized controlled trials as well as retrospective and prospective studies that were written in English. Reviews, correspondence, remarks, opinions, conference papers, animal observations, case studies, and case series weren't part of our study.

• Types of Participants

We included patients who suffered from adhesive capsulitis, considering their ethnicity, gender, age , and diabetes.

• Types of Intervention

Physical therapist interventions such as physiotherapy, muscle relaxation, electroshock treatment, exercise, and education were among the publications that were taken into consideration for review. Articles contrasting a therapeutic approach with a surgical treatment or interventions that had no clear connection to physiotherapy, however, were not included. Additionally, articles that focused on shoulder disorders other than the one being studied, such as rotator cuff injuries or osteoarthritis, as well as economic evaluation studies, were also excluded.

• Types of outcome measures

The main measure of interest in this study was the level of shoulder pain, which was evaluated using various types of pain scales. The study also looked at secondary outcome measures including a range of motion (ROM), functional status, and quality of life, which were assessed using relevant scales.

• Information sources

A thorough search was done on several of databases, namely PubMed, Physiotherapy Evidence Database (PEDro), Web of Science, and Cochrane, to find pertinent papers. As the most recent evaluation was completed in 2013, the search was restricted to papers written in English and published after January 2013. On January 16, 2020, the last search was conducted.

• Search

To search for relevant articles, the expression "frozen shoulder" and "adhesive capsulitis" were used interchangeably with the "OR" operator, as these terms are commonly used to describe the condition. These terms were then combined with the "AND" operator with additional search terms including "physical therapy", "rehabilitation", "manual therapy", "mobilization", "exercise", "education", and "electrotherapy".

• Study selection process

The titles and abstracts of the articles generated from the search were independently screened by two authors. If an article met the inclusion criteria, the full paper was read. Any disagreements between the authors were resolved through discussion and consensus.

• Data collection process

The study's first author extracted data related to basic information such as authors, year of publication, gender, age, and duration of symptoms. The study design details including sample size, randomization methods, allocation, blinding, and reporting information were also extracted. Furthermore, intervention details such as type, session, frequency, intensity, dosage, and duration were noted, as well as primary and secondary outcomes. The extracted data was then checked by the second author, and any discrepancies were resolved through a discussion between the two authors.

• Risk of bias in individual studies

Two writers evaluated the research's methodological quality using a customized form of Mortenson and Eng's and Sackett's critical assessment criteria. Based on the research of Sackett, Butler, and Campbell, the degree of evidence that was provided from each original study and the recommendation's grade were also determined. Any disagreements between the two writers were settled through conversation and, if required, by consulting a third author (Jason et al., 2015).

Data Analysis (research papers):

The databases indicated that they produced a total of 1269 entries using different keyword combinations. 33 papers were judged to be pertinent and included in the assessment after 1236 articles were excluded. Thirty of the 33 publications were randomized controlled trials, while the other three were cross-sectional exploratory investigations, prospective cohort studies, and retrospective cohort studies. The research employed somewhere between 26 and 122 individuals. There were 51.97 patients on average per trial. Even though six studies omitted to provide the female-to-male ratio, on average 31.7 (61.1%) of these patients were female. Just six research had more male patients than that. Each group included a minimum of 12 patients and a maximum of 53, with just an average of thirty-four patients per group. The respondents in the trials had an average age of 53.64 4.79 years. Studies with individuals in phases 1, 2, and 3 of the disease had symptom durations ranging from 1 month to 11.57 months. The period of follow-up in the trials we analyzed ranged from two months to two years following the intervention. Just five research utilized a control group, and nine studies did not follow the patients after the intervention. There was no control group, and this was mostly due to ethical considerations. The 33 papers that were examined included 30 randomized controlled trials (RCTs), two cohort articles with both retrospective and prospective designs, one cross-sectional exploratory research, and one study with no design. Five studies were given level I evidence based on the inclusion of at least one comparison group, subject randomization, and substantial sample sizes of more than 30 individuals per group. Due to their modest sample sizes of between six and thirty patients per group but randomized design, 25 studies were given Level II evidence. Each study underwent a quality evaluation using Sackett's critical appraisal criteria, examining eight different factors (Page & Labbe, 2010). Seven studies avoided contamination and cointerventions, lowering the danger of results being contaminated by advising individuals not to receive any adjuvant treatments throughout the trial, whereas just two studies met all the requirements (Ibrahim, Donatelli, Hellman, & Echternach, 2014). However, the remaining investigations did not cover the prevention of co-interventions or contaminations. All 30 RCTs compared two or more therapies or comprised an experimental group as well as a control group. In one research, group assignment was not completely random, while the mechanism of randomization was not disclosed in one journal (Forsythe et al., 2021). Twenty of the 33 research employed a blinded study considered, whilst the remaining studies either didnot use one or did not specify it. In 32 trials, physical therapists oversaw the therapies; one research did not provide this information. While some patients left the study during the remedy or follow-up phase in 12 trials, all subjects were present for the intervention in 21 research. 23 studies examined the validity and reliability of outcome measures, focusing on motion range in seven, pain in five, and functional status in eleven. Twenty research addressed the validity of measures, with four focused on motion range, six on pain, and ten on functional status. 21 studies included an additional follow-up period. With a score of seven points or 75% is regarded as good quality, quality ratings were reported as percentages. The results of a prior study served as the foundation for this cut-off point (Shanahan et al., 2022). Just 14 of the randomized trials were determined to be of good quality, with their quality ratings reaching the threshold of six points or 75%. Three Level I studies were among them, and two of them had low-quality ratings: Study: 62.5% (Fares, Koa, & Abboud, 2023) and the other study (Kumar Digge et al., 2022).