Chase 2015

Methods	Single‐Centre RCT (Australia)
Participants	Children aged 8‐18 years with slow‐transit constipation (STC) for more than 2 years Inclusion criteria: “All children had >2 years chronic constipation not responding to standard medical therapies (diet, laxatives, behaviour modification). STC was diagnosed by radionuclear transit studies performed within the previous year. STC was defined by retention of radioisotope in the ascending/ transverse colon but not rectosigmoid at 48 hours.” Exclusion criteria: “Children with Hirschsprung Disease, celiac disease, hypothyroidism or allergies that may impact on bowel function were excluded Children were excluded after bowel surgery (except for appendix stoma for antegrade colonic enemas), any contraindication to electrical current (eg. cardiac pacemaker), or previous electrical stimulation Children with neurological disorders or children/families with conditions that did not allow them to complete the questionnaires or bowel diaries were also excluded.” Forty‐six children in total were randomised, and 42 completed the study (21 from each group) Two children from each group did not complete the study
Interventions	Transcutaneous electrical stimulation (TES) delivered by trained physiotherapists using two channels of alternating current, one at a fixed frequency (4 kHz), while the other varied from 4080 to 4150Hz, producing a varying beat frequency of 80‐150 Hz. using 4 electrodes on the participants' belly and back, covering spinal outflow of T9 to L2 Two electrodes were placed paraspinally (T9‐L2), with the paired electrode positioned diagonally opposite on the anterior abdominal wall below the costal margin Stimulation was applied for 20 minutes per session by the physiotherapist, three times per week for four weeks Participants were allocated to receive either real or sham stimulation for four weeks and then were followed up for two months Radioisotope transit studies were performed pre and two months post intervention. Follow‐up period for the study was two months
Outcomes	Complete spontaneous bowel movement (CSBM), specifically, the number of participants with greater than 3 CSBM per week, soiling, colonic transit rate (measured by the geometric centre position of the radio‐isotope in the bowel) and quality of life (PedsQL)
Notes	This unpublished manuscript reports the same study as five other reports that were identified as eligible studies in this review (Chase 2009a; Clarke 2009; Clarke 2009a; Leong 2011; Southwell 2010) The other articles reported either interim analysis results or selected outcomes of the study, and they are classified as secondary references of this study The study was jointly funded by the National Health and Medical Research Council, Australia (Project Grants 384434, 546432, Senior Research Fellowship 436916‐ BRS) and Murdoch Childrens Research Institute Theme Investment Grants
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Randomisation, page 7 ”Instructions for active or sham stimulation were randomised (in blocks of 6) and sealed into numbered envelopes before recruitment. For each child recruited, the next envelope was mailed to the treating physiotherapist.” The method employed to generate random sequence was not stated
Allocation concealment (selection bias)	Unclear risk	Randomisation, page 7 ”Instructions for active or sham stimulation were randomised (in blocks of 6) and sealed into numbered envelopes before recruitment. For each child recruited, the next envelope was mailed to the treating physiotherapist” There was insufficient description on the generation of random sequence generation to enable an assessment of whether randomisation was performed independently from allocation. It was also unclear whether the envelops used were opaque
Blinding of participants and personnel (performance bias) All outcomes	High risk	Blinding “Patients and all trial staff (except for treating physiotherapist) were blinded to treatment. Because machines were connected to 240 V power, the treating physiotherapists could not be blinded Physiotherapists were given a written script and trained to present stimulations in identical fashion with both machines and to offer no other advice or intervention, thus ensuring patients received identical treatment with each machine" Active and sham stimulation “For sham stimulation, one machine was factory altered to deliver no current, but with output lights and dials indicating changing intensity levels to match active machines" The authors have stated that treating personnel (physiotherapist) could not be blinded, so the study was judged to have high risk of bias in terms of blinding of personnel The authors made a detailed explanation on the setting up of sham stimulation and measures to standardise the administration of active and sham IFTs to ensure that the patients and/or carers were unaware of the allocation that they received. However, it was unclear how effective the measures were, as the presence or absence of electrical currents might be felt by the patient, hence undermining the effectiveness of the efforts to mask the the patients”
Blinding of outcome assessment (detection bias) All outcomes	High risk	Blinding ”... physiotherapists mailed the treatment record and daily diary to the trial office. Blinded data entry was performed during the trial (CC, DR) and analysis was performed on conclusion of the trial by staff (KI, HA, YIY) not involved with the patients. The code was broken for final analysis” Despite blinding in data entry and analysis, It appeared that the physiotherapists were the assessors of at least some of the outcomes, including the subjective outcomes of symptoms as would be recorded in the diary. As the physiotherapists were clearly stated to be unblinded, the study was accorded high risk in the domain of blinding of outcome assessment. it was unclear whether the assessors of the other outcomes, such as colonic transit, were blinded to the allocation of the participants
Incomplete outcome data (attrition bias) All outcomes	High risk	Four out of 46 participants (two each group) did not complete the trials. The reasons given were: two participants lost their bowel diaries, and two “violated the trial protocol”. It was unclear whether the four participants underwent assessments of other outcomes such as colonic transit study Despite the relatively small number of participants with missing data for the major outcome of CSBM, it appeared that different subsets of the total participant group were included in different outcomes (as presented in different reports of the same study) For example, only 21 participants (14 in the TES group and 7 in the control group) were included in the outcome "number of participants with improved colonic transit, and as low as 16 participants (8 in each group) were included in the outcome (number of participants with improved quality of life) Considering the overall participation rate across all the outcomes, we decided to accord the study high risk under the domain of incomplete outcome data
Selective reporting (reporting bias)	Low risk	All major outcomes specified in the methods, including CSBM, soiling, colonic transit rate, soiling, abdominal pain and quality of life (PedsQL) were reported in the results in the primary and/or secondary references
Other bias	High risk	In one of the secondary references that reported the outcome of colonic transit (Clarke et al 2009), there were 61 radio imaging studies performed on 26 children, meaning some children undertook multiple studies The studies were taken as the unit of reporting rather than the participants The authors do not appear to have adjusted their data to account for this clustering effect, either by undertaking some form of generalised estimating equation or taking the mean reading of all radio imaging studies for each participant