Ma 2020.

Study characteristics
Methods	Allocation: randomised Blindness: single‐blind, block random coding table Duration: 12 weeks Funding: 'Chinese association for the study of pain, the minimally invasive interventional group,' research organisation Setting: China; multi‐centre, unclear about the settings Design: parallel
Participants	Diagnosis: Chinese people with moderate to severe cancer pain (mean pain intensity ≥ 5/10 cm on the VAS or breakthrough pain ≥ 3 times a day). n = 233 Age: 18–80 years Sex: 152 men; 81 women Cancer stage: not reported History: not reported Inclusion criteria: people with cancer pain aged 18–80 years; after standard treatment and opioid rotation according to the guidelines, oral morphine equivalent daily dose still > 200 mg with unsatisfactory analgesia or people with intolerable adverse effects caused by systemic opioids; mean pain intensity ≥ 5/10 cm on the VAS or breakthrough pain ≥ 3 times a day; people suitable for IDDS and with the ability to comply with the medical protocol and visit; and people with the indication of IDDS Exclusion criteria: with severe infection, respiratory dysfunction, serious liver dysfunction or renal dysfunction; history of allergy to narcotics or malignant hyperthermia; spinal deformation who are unable to receive the IT delivery system; intracranial metastasis, consciousness disorders, central nervous system infection or coagulation disorders; pregnant or lactating; plan to become pregnant within 1 month after the study; participated in another medical trial within 3 months before this study; family members of the study investigators; and people considered ineligible for the study as evaluated by the investigator
Interventions	Hydromorphone IT, mean starting daily infusion dose (mean reported as 0.276 (SD 0.53) mg; n = 121 (121 completed study, 13 completed follow‐up) Morphine IT, mean starting daily infusion dose (mean reported as 1.551 (SD 4.20) mg; n = 112 (112 completed study, 9 completed follow‐up) The starting infusion rates for IT opioids was based on the baseline oral morphine equivalent daily dose (pre‐IDDS consumption) of each participant, 1:300 for IT: oral equivalent dose, and the opioid equivalence of hydromorphone to morphine was set at 0.15:1 according to the 2016 National Comprehensive Cancer Network guideline (1.5:10, for parenteral dose). For breakthrough cancer pain, the bolus dose of each participant was set at 1/10 of the daily continuous infusion dose individually. If the participant's previous 24‐hour bolus press number was ≥ 4 times, daily continuous infusion dose increased according to the experience of the study physician; each time, it was increased by 50% of the previous 24‐hour IT opioid daily dose, and then reset the bolus dose according to the daily continuous infusion dose. The participants were not allowed to intake any additional analgesics except the IT participant‐controlled analgesia bolus doses of hydromorphone or morphine for the management of breakthrough pain. If the participant was diagnosed with neuropathic pain, they followed the study protocol, and the participant would not have received any other treatment except IDDS with hydromorphone/morphine. And according to the existing evidence and clinical experience, IT opioids could also be used for the treatment of neuropathic pain, although it was not the guideline recommendation.
Outcomes	Pain intensity VAS Clinical success rate Dropouts Unable to use: Quality of life measured by the quality of life score (36‐item Short Form)a, Drug cost‐effectivenessb, The frequency, duration and degree of the flare painb, Intensity of anxiety experienced by patientb, Intensity of depression experienced by patientb
Notes	aNo available data. bNot predefined by protocol.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "A total of 233 participants across all centers were randomized chronologically in a 1:1 ratio according to the block random coding table and given the corresponding medication" (page 2503). Comments: participants were randomly assigned using random coding table.
Allocation concealment (selection bias)	Low risk	There was no explicit description on allocation concealment, but we considered concealment was likely to have been used since the randomisation was done via a random table.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "All patients and investigators conducting follow‐up work were blinded to the treatment drug assignment throughout the study." Comments: although the report stated the trial was a single‐blind study, participants and investigators were blinded, so we judged it at low risk.
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Although the outcome assessment was not reported, as most of the outcomes were participant‐reported, we rated this item at low risk of bias.
Incomplete outcome data (attrition bias) All outcomes	High risk	Although the report stated that 'all patients completed assessments of [primary and secondary outcomes],' the CONSORT flow diagram showed a total of 211/233 (90%) participants dropped out in the follow‐up period. Therefore, we rated this at high risk.
Selective reporting (reporting bias)	High risk	The study protocol was available, but data on quality of life were not available.
Other bias	Unclear risk	Not obvious.