Table 1.

Literature review table—included studies

Study	Design	Sample	Intervention(s)	Findings	Findings short*	Quality of evidence	Quality level
Hypnotherapy
Katz 1987 [22]	RCT, evaluating hypnotherapy versus attention control for pain, anxiety, and distress associated with BMAs	Children with acute lymphoblastic leukemia (6–11 years) undergoing repeated BMA who experience significant anxiety, fear, and/or pain during BMA (n = 36)	Hypnotherapy: hypnotic induction, active imagery, individually tailored, deep muscle relaxation, and suggestions. Ending with a post-hypnotic suggestion Attention control: non-directed play sessions designed to control for the amount of time and attention	- Improvement was reported in self-reported pain (0–100 VAS) and distress over baseline with both interventions, with no differences between them. - No significant main effects were found in PBRS scores. - Girls exhibited more distress behavior than boys on three of four dependent measures used. - Results are discussed in terms of potential individual differences in responding to stress and intervention that warrant further research	Hypnosis vs attention control = no differences for pain and distress Post-treatment vs baseline + pain and distress (for hypnosis and control)	RCT, sufficient sample size, randomization process not entirely described, blinding of independent observers, nurses and observers, good inter-rater reliability. No selective reporting, adequate analysis, study completed as planned, no missing data	Moderate
Smith 1996 [30]	RCT, cross-over, repeated measures single group study evaluating hypnosis versus distraction for pain, anxiety, and distress associated with venipuncture or infusaport access	Children (3–8 years) with hematology and oncology diagnoses undergoing repeated venipuncture or infusaport access (n = 27)	Hypnosis: favorite place hypnotic induction. Both parents and children were taught the exercises. Attention control/distraction: activating the pop-up toy, noting interesting aspects of the toy	- Only children with high hypnotizability had reduced child self-reported pain (1–5 Global Rating Scale) and anxiety (1–5 Likert scale), parent-rated pain (1–5 Likert scale), and observer anxiety and distress from hypnosis intervention (OSBD-R scale) -Children with low hypnotizability in the distraction condition had significantly lower observer-rated anxiety only -Practical: parents and children were both trained in hypnosis exercises. Parents were very positive and exercises were easy to learn and practise.	Hypnosis vs control ++ for self-reported pain ++ for parented reported pain ++ for distress All only for children with high hypnotizability	RCT, cross-over design. Observers, trainers, and parents were told that both interventions were equally effective, observers were blind to high and low hypnotizability level of children, both self-reported measurements and observer measures. Adherence to the exercises at home was monitored and no significant differences in compliance were observed between the groups. Sufficient sample size, no selective reporting, adequate analysis, study generally completed as planned, some missing data due to death of participants	High
Liossi 2003 [23]	RCT, evaluating direct hypnosis and indirect hypnosis versus attention control and standard care for pain and distress associated with LPs	Children and adolescents (6–16 years) with leukemia or non-Hodgkin lymphoma undergoing repeated LPs (n = 80)	Indirect hypnosis: Using metaphors, imagination using various senses, develop cues to experience immediate relaxation, and ways to adapt to discomfort. Ending with a post-hypnotic suggestion. Directed by therapist and then self. Direct hypnosis: “Analgesic” suggestions. Directed by therapist and then self. Attention control: including elements such as development of rapport, non-medical play, and no medical verbal interactions Equivalent time was spend with the therapist as in hypnotherapy. Standard care: no contact with the therapist, medical care for pain with LP provided by the hospital staff.	-Direct and indirect hypnosis groups were equally effective and reported less pain and anxiety (both 0–5 Wong-Baker Faces scale) as compared with attention control or standard care groups. -Higher levels of child hypnotizability associated with increased treatment benefit. -Treatment benefit lessened with self -hypnosis as compared with therapist-directed	Hypnosis vs attention control or standard care + for pain and distress (indirect and direct hypnotherapy) Indirect vs direct hypnosis = for pain and distress	RCT, sufficient sample size, independent observers, doctors, and behavioral observers were blinded, blinding was measured, observers could only guess which children were in the direct hypnosis group (intervention 1), they could not distinguish between the other intervention groups and control group, no selective reporting, appropriate analysis, study completed as planned, no missing data.	High
Liossi 2006 [24]	RCT evaluating hypnosis versus attention control or standard care for pain and distress associated with LPs	Children and adolescents (6–16 years old) with leukemia or non-Hodgkin lymphoma undergoing repeated LPs (n = 45)	Hypnotherapy: Standard care + “Analgesic” suggestions, ending with a post-hypnotic suggestion. Directed by therapist and then self. Attention control: Standard care + including elements such as development of rapport, non-medical play, and no medical verbal interactions Equivalent time was spend with the therapist as in hypnotherapy Standard care: EMLA/analgesic cream. Medical care for pain with LP provided by the hospital staff	- Group receiving hypnosis, in addition to local anesthetic (EMLA), reported less pain and anxiety (both 0–5 Wong-Baker Faces scale), and less observed behavioral distress as compared with other groups. -Treatment superiority was maintained when switched to self -hypnosis following therapist-directed hypnosis. -Higher levels of child hypnotizability associated with increased treatment benefit	Hypnosis vs attention control or standard care ++ for self-reported pain and distress	RCT, sufficient sample size. Independent observers, doctors and behavioral observers were blinded. Blinding was measured, observers could not guess in which groups the children were allocated. Inter-rater reliability was tested and found to be good. No selective reporting, appropriate analysis, study completed as planned, no missing data	High
Liossi 2009 [25]	RCT evaluating self-hypnosis versus attention control or standard care for pain and distress associated with venipuncture	Children and adolescents (7–16 years) with cancer undergoing venipuncture (n = 45)	Self-hypnosis: Standard care + “Analgesic” suggestions, ending with a post-hypnotic suggestion. Following that, children were taught self-hypnosis. Attention control: standard care + including elements such as development of rapport and no medical verbal interactions. Equivalent time was spend with the therapist as in hypnotherapy Standard care: EMLA/analgesic cream. Medical care for pain with LP provided by the hospital staff	-Self-hypnosis + local anesthetic (EMLA) reported less anticipatory and experienced anxiety, pain (self-report 100 mm VAS) and observed behavioral distress as compared with other groups. -Parents experienced less anxiety in the hypnosis group	Self-hypnosis vs attention control or standard care ++ for self-reported pain and distress + anxiety parents	RCT, sufficient sample size. Independent observers, doctors, and behavioral observers were blinded, blindness was measured, observers could not guess in which groups the children were allocated. Inter-rater reliability was tested and found to be good No selective reporting, appropriate analysis, study completed as planned, no missing data.	High
Mind-body (including imagery, meditation, breathing techniques)
Pourmovahed 2013 [29]	RCT evaluating regular breathing versus standard care for pain associated with intrathecal injections	Children and adolescents (6–15 years) with leukemia undergoing a first intrathecal injection (n = 100)	Hey-Hu breathing technique: the child first takes a deep breath and exhales while whispering ‘hey’, then inhales deeply again and exhales whispering “hu” Standard care: current standard medical practice	- Children in the “Hey-Hu” breathing group reported significantly less pain (0–5 Wong-Baker Faces scale) than the control group, particularly among children aged above 10 years. -There was no significant difference between the two sexes. -Nurses could help children learn the method of ‘Hey-Hu’ breathing and implement it in hospitalized children who undergo painful procedures.	Hey-Hu breathing vs standard care + for pain ++ for pain in children > 10 years	RCT, sufficient sample size, sampling using random allocation software, some blinding (semi-blind, the performer of the procedure was aware of the aim of the study), no selective reporting, appropriate analysis, study completed as planned, some missing data.	Moderate
Massage
Phipps 2005 [28]	RCT, unbalanced pilot, evaluating professional massage and parent massage versus standard care for pain and distress experienced undergoing hematopoietic stem cell transplantation (HSCT)	Children (all ages) scheduled to undergo HSCT (n = 50)	Professional massage: therapeutic massage delivered by licensed massage therapists three times per week for the 4-week period from admission for HSCT through 3 weeks post-transplantation. Parent massage: parents learned basic massage techniques to use with the child. The routines taught to the parents were essentially the same as those provided in the therapist massage arm. Parents were asked to begin giving their child massage at least three times per week. - Standard care: usual care	-No significant differences were observed between the two massage interventions on distress and pain scores (self-report 100-mm VAS). -No significant differences between either massage group and standard care for pain and distress, although there were descriptive trends suggestive of benefit, some of which approached significance. Larger differences emerged on the outcomes of days in hospital and days to engraftment, pointing to the potential cost-benefits of a massage intervention in this setting. -Regarding narcotic usage, there were no significant differences between groups, but descriptively there was a trend for those in the massage arms to use less medication.	Professional vs parent massage = no differences for pain and distress between massage groups Massage vs standard care: = for pain, distress and narcotic medication use (for professional and parental massage)	RCT, insufficient sample size (underpowered, though the sample was representative of the population of patients who underwent transplantation), allocation to treatment arms was not equal but was designed so that participants were twice as likely to enter either intervention arm than the control arm, lack of blinding, no selective reporting, appropriate analysis, not described if study completed as planned, some missing data reported	Low
Mehling 2012 [26]	RCT, non-blinded pilot, feasibility study, evaluating a combined massage-acupressure intervention versus standard care, for decreasing treatment-related symptoms such as nausea, vomiting and pain associated with hematopoietic cell transplant	Children (5–18 years) undergoing hematopoietic cell transplant at an academic medical center (n = 23)	Combined massage-acupressure intervention: practitioner-provided, combined Swedish and acupressure massage three times a week throughout hospitalization. Parents were trained to provide additional acupressure as needed. - Standard care: usual care	-There was no statistically significant difference or change in pain (BASES subscale self-report) between the two groups -Intervention group versus control showed fewer days of mucositis, lower overall symptom burden, feeling less tired and run-down, having fewer moderate/severe symptoms of pain, nausea, and fatigue	Massage vs control = for pain	RCT, insufficient sample size (small feasibility study, aim to report standardized effect sizes that allow for sample-size calculations for future studies), no blinding, no selective reporting, appropriate analysis, study completed as planned, no missing data	Low
Celebioglu 2015 [21]	Controlled pretest/posttest quasi-experimental study, investigating the effect of massage therapy versus standard care, on pain and anxiety arising from intrathecal therapy or BMA	Children (4–15 years) with primary diagnosis of cancer (n = 25)	Massage therapy: one massage session from a licensed massage therapist. Massage techniques were a combination of effleurage and petrissage to the shoulders, neck, face, arms, lower back and waist. Standard care: standard treatment offered to patients undergoing IT or BMA.	-No difference between groups for pain (0–10 VAS self-report or by mother) or anxiety -It was determined that pain and anxiety levels in the massage group decreased significantly post-treatment versus baseline	Massage vs control = for pain and anxiety Post-treatment vs baseline: + for pain and anxiety (massage group)	Pretest/posttest quasi-experimental study with the control group, small sample size, non-probability convenience sampling, children were divided between the groups according to admission date, no blinding, no selective reporting, inappropriate analysis, study completed as planned, no missing data	Low
Healing touch
Wong 2013 [31]	RCT, evaluating healing touch versus attention control on feasibility in pediatric oncology	Children (3–18 years) diagnosed with childhood malignancy, receiving chemotherapy and/or radiation therapy (n = 9)	Healing touch: by certificated practitioner, standardized techniques. Attention control: reading or age-appropriate play activity for the same time as the intervention group	-There were statistically significant differences in pain scores (children and parents on 1–10 Wong-Baker Faces scale) and distress scores (parents) between the healing touch group and the control group. - Among the healing touch group, all scores (pain, distress, and fatigue) decreased significantly after the intervention. Scores among the control group did not show a statistically significant decrease. - The study demonstrates the feasibility of using energy therapy in the pediatric oncology patient population.	Healing touch vs control ++ for self-reported pain + for pain reported by parents = for pain reported by staff = for self-reported distress + for distress reported by parents = for distress reported by staff	RCT, insufficient sample size (recruitment rate 60%), the participants in the intervention group received approximately 6.5 times more treatments than the control group, which may bias results. High heterogeneity of groups (age, diagnose, and treatment protocols), no blinding. No selective reporting. Inappropriate analysis, study not entirely completed as planned (2 drop-outs, because of prolonged hospitalizations and complicated treatments and 1 participant died while in the study because of disease progression), some missing data	Low
Music therapy
Nguyen 2010 [27]	RCT, evaluating music vs control for pain and distress associated with LPs	Children (7–12 years) with leukemia undergoing LPs (n = 40)	Music group (earphones with music): Children choose their own music to be played into earphones from an iPod, 10 min before the LP procedure started. Control group (earphones without music): same procedure as music group, only without music	- As compared with the control group, children in the music group had significant reduction in self-reported pain (0–10 Numeric Rating Scale during and after procedure) and anxiety (before and after the procedure) -Significant reductions in heart rate and respiratory (during and after procedure) in music group; blood pressure and oxygen saturation did not differ between groups - The findings from the interviews confirmed the quantity results through descriptions of a positive experience by the children, including less pain and fear.	Music vs control ++ for self-reported pain during and after the lumbar puncture. ++ for heart- and respiratory rates during and after the lumbar puncture. = for blood pressure and O2 saturation	RCT, sufficient sample size, lack of blinding (all the children were given identical pre-procedural information, randomization was carried out using opaque envelopes, the researcher and the physician did not know to which group the patient belonged), no selective reporting, correct analysis, study completed as planned, no missing data	High

RCT, randomized controlled trial; LP, lumbar puncture; BMA, bone marrow aspiration; IV, intravenous; CBT, cognitive-behavior therapy; GA, general anesthesia; IM, intramuscular injection

*: + or − → P < 0.05

++ → P < 0.001

= → no significant difference