Hypothesis / aims of study
Chronic constipation is a common, symptom-based condition estimated to affect one in seven adults: disproportionately affecting older adults, and other vulnerable groups [1]. Management of constipation in adults is varied and may include pharmacological, conservative, surgical interventions or a combination of approaches. Frontline treatment usually consists of laxatives, although this is not always successful at alleviating constipation, and guidelines to treat constipation in adults are currently lacking.
This has resulted in the re-emergence of alternative conservative (non-invasive) approaches such as abdominal massage. Some clinicians (and patients) have used abdominal massage to help relieve the symptoms of constipation (e.g. infrequent bowel movements, feeling bloated and uncomfortable). However, we do not know how effective it is, or what the optimal intervention regime (e.g. frequency, intensity, self-management) should be or which outcome is best to capture any potential benefit of this intervention. In this systematic review, we sought to systematically identify the outcome measures currently used in abdominal massage randomised controlled trials (RCTs) in adults with chronic constipation.
Study design, materials and methods
Study design:
Cochrane systematic review using established methodology [2].
Selection criteria:
We included RCTs and quasi-RCTs that compared the effectiveness of abdominal massage with no treatment, usual care or another conservative treatment. We included trials with adult participants irrespective of their aetiology as long as they had a clinical diagnosis of chronic constipation. Studies of participants taking medication known to cause constipation (e.g., pain medication for cancer) were excluded.
Searches:
We conducted systematic searches of the major electronic databases (i.e. Medline, AMED, British Nursing Index, Cochrane library databases) and clinical trial registers (i.e. WHO ICTRP, clinicaltrials.gov). No date or language restrictions were applied (date of last searches: November 2020).
Selection of studies:
Records were imported into Covidence and deduplicated. Titles and abstracts were screened independently by two reviewers. Relevant full text publications were retrieved and independently screened and an online consensus meeting was held to agree the final selection.
Data extraction and management:
We extracted the following data from each study: authors, year, methods, study design, key information about participants, comparators and interventions. We specifically extracted the following data: outcomes measured specifying the tool used, whether the tool was specified as a primary or secondary measurement, whether the tool had been measured at baseline and whether it was validated. One reviewer extracted the data, and it was cross-checked by a second reviewer.
Assessment of risk of bias in included studies:
We assessed the methodological quality of included studies using established criteria published in the Cochrane Handbook [2].
Data synthesis:
We calculated the frequency of each outcome and outcome measurement tool reported within each study. Evidence tables were constructed, and findings were summarised using a narrative synthesis.
Results
Our search identified 1993 records of which 14 studies (18 randomised paired comparisons) with 981 participants met the selection criteria. Half of the studies employed a parallel group RCT design. The remaining studies were multi-arm RCTs (n=5), cross-over RCT (n=1), and a quasi-RCT (n=1). Participant populations (age range: 24 -94 years) were heterogeneous and included healthy college students (diagnosed with constipation), individuals with complex needs living in residential care, older adults (aged > 60 years), and adults living with a variety of neurological conditions (Multiple Sclerosis, Stroke, and Parkinson's).
Of the 18 randomised paired comparisons:
• 11 evaluated the effectiveness of abdominal massage compared to no treatment, usual care or placebo control.
• 5 evaluated the effectiveness of abdominal massage compared to another type of massage (i.e.) acupressure (n=1), connective tissue massage (n=2), manual lymph drainage (n=1) and tensegrity massage (n=1)
• 2 evaluated the effectiveness of abdominal massage compared with another type of conservative intervention (i.e.) laxatives (n=1) and electrical stimulation (n=1).
Abdominal massage interventions were typically delivered daily in sessions which lasted 15 - 20 minutes. The duration of the intervention period ranged from 2 – 8 weeks (mean duration: 4.8 weeks). Abdominal massage was delivered by a range of providers including physiotherapists, nurses, experienced therapists, patients, carers or a mix of providers.
Our review identified 95 outcomes reported across 14 included studies, ranging from 3 – 16 outcomes in one study. The included studies used a large number of heterogeneous validated and non-validated outcome measures often capturing the same outcome using more than one tool.
The most frequently reported outcomes were:
• the severity of constipation (n=8 studies) using five different tools including the Constipation Scoring System, the Constipation Severity Instrument, the Constipation Severity Scale, the Neurogenic Bowel Dysfunction Score and several non-validated patient questionnaires.
• the type of stool (n=8 studies) using three different tools including the Bristol Stool Scale, bowel diaries and a variety of non-validated patient questionnaires.
Adverse events were the least frequently reported outcome (n=1 study).
The majority of outcomes were measured at baseline and immediately post-intervention (n=9 studies). Five studies measured outcomes at one further follow-up timepoint which ranged from a 2 week follow-up post-intervention to 24 week follow-up.
The methodological quality of the included studies was generally judged as high quality (low risk of bias) as shown in Figure 1. The exception to this was performance bias, as it was not possible to blind participants or staff to the intervention being delivered (Figure 1).
Interpretation of results
The outcomes in clinical studies evaluating the effectiveness of abdominal massage are highly varied, often reporting overlapping outcomes of different aspects of constipation severity and symptoms within the same trial using different validated and non-validated tools. Although many of these instruments are similar, the scoring system within each scale is different. Moreover, trialists present the data in a variety of ways, limiting our ability to compare and pool the data from different studies to examine overall efficacy and leading to research waste. When this heterogeneity in outcomes is considered alongside the diverse populations included in the studies, it dilutes the value of conducting often high-quality trials.