Table 1 Description of studies meeting eligibility for the systematic review
From: Electronic behavioral interventions for headache: a systematic review
Tool | Type of behavioral intervention | Design | First author, year | # of participants (N), setting, duration of treatment | Headache type/criteria | Outcomes measured | Results (HA freq, HA intensity, disability, adherence) | Other results | Drop out rate |
|---|---|---|---|---|---|---|---|---|---|
CD ROM | CBT | RCT (waitlist control) | Connelly, 2005 [44] | N = 50 (ages 7–12) Pediatric Neurology Clinic 4 weeks | Migraine, tension type or chronic daily HA min 4/month with symptom free period, assessed by neurologist or NP | HA duration; HA days; HA intensity; HA severity; Medication; Self-efficacy; QoL; Disability; Acceptability | • There was a significant reduction in HA frequency from baseline to post-tx in both the tx (Headstrong) + control groups (Univariate ANOVA, p < 0.001) with a trend suggesting a greater reduction in the tx group (60.16 % reduction in tx group vs. 45.43 % reduction in controls, p = 0.091). | • Significant “group by phase” interaction effect on the HA duration variable (p = 0.014) suggesting that there was different changes in HA duration from baseline to 1-month post-tx as a function of group assignment. | 6 % Overall (4 % tx group vs 2 % control group) |
• A clinically significant change in HI from baseline to 1 month post-tx was observed in 60 % of the tx group + only 8 % of the control, therefore using the adjunctive Headstrong program resulted in more children achieving clinically significant outcomes (Chi-square, p = 0.005) | |||||||||
• Significant “group by phase” interaction effect on the HA intensity variable (p = 0.004) suggesting that there was different changes in HA intensity from baseline to 1-month post-tx as a function of group assignment. | |||||||||
HA intensity decreased from baseline to 1 month post-tx in the tx group, while it remained fairly constant in the control group. | |||||||||
• No power to assess secondary outcomes including disability. | |||||||||
CBT + self-management | RCT | Rapoff, 2014 [45] | N = 35 (ages 7–12) Pediatric clinics & children’s hospitals 4 weeks | migraine with or without aura min 1/week | HA duration; HA frequency; HA days; HA severity; QoL; Disability | • NS change in HA frequency between tx + control groups | 50 % Overall-no allocated intervention (55 % tx group vs. 43.3 % control group) | ||
• There was a statistically significant difference in pain severity (10-point VAS) post-intervention, with tx group reporting lower pain severity than control group (5.06 vs. 6.25, p = 0.03, ES = 0.7). | |||||||||
18.6 % Overall were lost to follow-up (17.5 % tx group vs. 20 % control group) | |||||||||
• At 3 months post-intervention, parents reported lower migraine-related disability (PedMIDAS) in the tx group compared to control group (1.36 vs. 5.18, p = 0.04). | |||||||||
Internet | CBT | Parallel group unblinded RCT | Day, 2014 [46] | N = 36 (ages 19+) Physician referral, brochures + public service announcements 7 weeks | Migraine, Tension-type, cluster or other primary HA min 3 days/month | HA duration; HA index; HA frequency; HA intensity; HA severity; Medication; Self-efficacy; Disability; Acceptability; Alliance; Feasibility; Engagement; Other | • There was a statistically significant baseline to post-test decrease in HA frequency, HA peak intensity + HA average intensity in the total completer sample, however there were no significant differences in these variables between the tx + control groups. | • ITT analysis: Greater improvement in self-efficacy (p = 0.02) + pain acceptance (p = 0.02) in tx group compared to controls. | 11.3 % Overall prior to randomization |
2.1 % Overall after randomization (.53 % in tx group vs. 1.6 % in control group) | |||||||||
• Completer analysis: Improved pain interference (p < 0.01) + pain catastrophizing (p = 0.03) in tx group compared to controls. | |||||||||
• For the ITT analysis, there was a significant decrease in HA frequency overall, but again no difference in tx groups. | • MBCT was found to be feasible, tolerable + acceptable to patients. | ||||||||
CBT | Prospective parallel group design | Bromberg, 2012 [27] | N = 213 (ages 18–65) Website postings, electronic newsletter announcements, neurology clinics, + social networking/community sites 4 weeks | migraine with or without aura min 2/month | Pain catastrophizing; Self-efficacy; Disability; LoC; CPC; DAS; Other | • Reduction of HA frequency + severity could not be tested due to technical problems resulting in loss of data. | • Decrease in depression (DASS) in tx group compared to controls from baseline to 3-months post-intervention (p = 0.0009) + baseline to 6-months post-intervention (p = 0.0079). | 11.3 % Overall prior to randomization | |
2.1 % Overall after randomization (.53 % in tx group vs. 1.6 % in control group) | |||||||||
• Both tx + control subjects reported similar reductions in disability on MIDAS (12.8 % decrease + 13.0 % decrease respectively) immediately post-intervention. | |||||||||
• Decrease in stress (DASS) in tx group compared to controls from baseline to post-intervention (p = 0.0324) + from baseline to 3 month follow-up (p = 0.0045). | |||||||||
• Follow-up assessment completion in tx vs. control groups respectively were 80 % vs. 89 % at 1-month, 70 % vs. 82 % at 3-months, + 55 % vs. 82 % at 6-months. | • Reduction in pain catastrophizing (PCS) in tx group compared to controls from baseline to post intervention (p = 0.0030), 3-month follow up (p = 0.0099), + 6-month follow up (p = 0.0006). | ||||||||
• CPCI-42:Increase in relaxation (baseline to post-intervention, 3-month assessment + 6 month assessment), task persistence (baseline to post-intervention + 3-month assessment), exercising (baseline to post-intervention) + use of social support (baseline to post-intervention) in tx group compared to controls. | |||||||||
• Increases in self-efficacy in tx group compared to controls (baseline to post-intervention, 3-month assessment + 6 month assessment) | |||||||||
CBT (family-based) | RCT | Law, 2015 [43] | N = 83 (ages 11–17) Pediatric clinic 8 weeks | Recurrent HA (>3 months) | HA days; HA intensity; Activity limitation; DAS; Acceptability; Feasibility; Engagement; Other | • There was a statistically significant reduction in HA frequency from baseline to post-tx + baseline to 3-month follow up in both tx conditions, however there was NS difference in HA frequency between tx + control groups. | • There was a significant reduction in activity limitations, emotional functioning + parent response to pain behavior from baseline to post tx in both groups, but no significant difference between tx + control groups | 28.9 % Overall (29.5 % in tx group vs 28.2 % in control group | |
• There was a statistically significant reduction in HA pain intensity from baseline to post-tx + baseline to 3-month follow up in both tx conditions, however NS in HA frequency between tx + control groups. | |||||||||
CBT + PMR | RCT | Trautmann, 2010 [47] | N = 65 (ages 10–18) Newspaper ads, websites 6 weeks | Migraine, tension-type HA, or combined HA min 2 HA attacks/month | HA duration; HA frequency; HA intensity; Pain catastrophizing; QoL; DAS; Acceptability; Alliance; Other | • There was a significant reduction in HA frequency + duration post-tx in all groups, but NS between groups. | • Pain catastrophizing was significantly reduced post-assessment in all groups, but no difference was found between groups. | 7.7 % Overall (16.6 % in CBT group vs. 0 % in AR group vs. 5.3 % in EDU group) | |
• No significant difference in HA intensity was found in any group at post-assessment | • Responder rates (reduction in HA frequency of 50 % or more from baseline) were significantly higher in CBT (63 %) + AR (32 %) groups, compared to the EDU/ control group (19 %).This resulted in NNTs of 2.0 for CBT + 5.2 for AR. | ||||||||
• There was no significant difference in depression, psychopathological symptoms, + health-related quality of life in any group post-assessment. | |||||||||
CBT + Relaxation | RCT | Sorbi, 2015 [48] | N = 368 (18–65) HA centers, website, + flyers 8 weeks* | Migraine with 2–6 attacks in the month prior to randomization | HA index; HA intensity; Medication; Self-efficacy; QoL; Disability; LoC; Other | • NS in HA frequency or intensity in either group or between groups. | • HA duration decreased significantly more in telephone arm (p < 0.05) | 32 % Overall (29 % tx vs. 35 % control) | |
• NS in HI between groups | |||||||||
• Self-reported inventories (HADS depression subscale, HDI, PSS) showed significant improvements in both groups but not between groups. | |||||||||
Multimodal including CBT | RCT | Trautmann, 2008 [49] | N = 18 (ages 10–18) Participation was online, recruitment strategy not specified 6 weeks | Migraine +/or tension-type HA min 2 HA attacks/ month | HA duration; HA frequency; HA intensity; Pain catastrophizing; Acceptability; Alliance | • No significant difference found between HA frequency or intensity between groups post-tx. | • NS found between the two groups post tx in any of the outcome variables (HA frequency, intensity, duration, or pain catastrophizing). | 11.1 % Overall (5.6 % in tx vs. 5.6 % in control) | |
• Frequency of HA decreased significantly from pre-tx to post-tx in CBT group but not in control (EDU) group. | • Pain catastophizing was significantly decreased from baseline to post tx in CBT group but not in control. | ||||||||
• NS difference between the groups in satisfaction or “patient-therapist-alliance/assistance” | |||||||||
Multimodal including CBT | RCT | Hedborg, 2012 [50] | N = 76 (ages 22–65) Newspaper ads 24 weeks (MBT) + 36 weeks (hand massage) | Migraine at least 2 times monthly | Medication | • Decrease in total migraine drug intake at the end of the MBT program in the MBT group (13.0 vs. 10.1 drug doses/subject/56 days) compared to controls, no significant difference in total migraine medication drug intake in the control group (8.3 vs. 8.9 drug doses/subject/56 days) | 8.4 % Overall (7.4 % in MBT+ hand massage vs. 14.3 % in MBT vs. 3.6 % in control group) | ||
• Drug efficacy increased during MBT from 0.30 to 0.52 (p < 0.001), but this was mainly explained by the increase proportion of mild HAs | |||||||||
Multimodal including CBT | RCT | Hedborg, 2011 [51] | N = 83 (ages 22–65) Newspaper ads 24 weeks (MBT) + 36 weeks (hand massage) | Migraine at least 2 times monthly | HA frequency; QoL; DAS; Acceptability | • 40 % of patients receiving MBT alone + 42 % of patients receiving MBT+ hand massage had 50 % + reduction in migraine frequency when compared to control group. | 8.4 % Overall (7.4 % in MBT+ hand massage vs. 14.3 % in MBT vs. 3.6 % in control group) | ||
• Hand massage NS on migraine frequency compared to MBT alone. | |||||||||
• NS in depression (MADRS-S) scores from baseline to post-tx or across groups. | |||||||||
• Improvement in “perceived work performance” in hand massage + MBT group from baseline to all follow-up points. | |||||||||
Multimodal including CBT + applied relaxation | RCT | Andersson, 2003 [38] | N = 44 (ages 18–59) Newspaper ads, project website 6 weeks | Migraine, Tension-type HA, or cluster HA (Dx = self-report) | HA duration; HA index; HA days; HA intensity; Disability; CPC; DAS | • NS in HA frequency or intensity (either group or between groups). | • HA duration decreased more in telephone arm (p < 0.05) | 32 % Overall (29 % tx vs. 35 % control) | |
• NS in HI between groups | |||||||||
• Self-reported inventories (HADS depression subscale, HDI, PSS) showed significant improvements in both groups but not between groups. | |||||||||
Multimodal including PMR + biofeedback | RCT (delayed tx control which later crossed over) | Devineni, 2005 [52] | N = 139 (age not specified) Internet-based promotion channels e.g. online classified ads, websites 4 weeks | Migraine with or without aura, tension-type HA, or mixed. | HA duration; HA index; HA frequency; HA severity; Medication index; Disability; DAS; Cost | • Only a non-significant trend was found for # of HA days per week between groups post-tx. | • There was a trend towards a between group difference in medication index post tx (p = 0.12) | 38.1 % Overall (58.8 % in immediate tx group vs. 70.4 % in delayed tx group) | |
• % of tx completers with clinically significant improvement (50 % decrease in HI) was 38.5 % vs. 6.4 % (waitlist) | |||||||||
• There was a significant decrease in peak intensity between the tx and control groups post-tx | |||||||||
• Estimated time expenditure for the therapist =1.3 h/participant (range = 0.2–8.8 h), resulting in a cost-effectiveness estimate of 0.32. | |||||||||
• Greater compliance was associated with greater improvement in primary HA outcomes. | |||||||||
Multimodal including PMR | Randomized to intervention vs waitlist | Strom, 2000 [53] | N = 102 (ages 19–62) Newspaper articles + Internet magazines 6 weeks | Recurrent HA (>6 months, at least 1 HA per week) | HA duration; HA index; HA days; HA intensity; HA severity; Medication index; Disability; DAS; Cost | • Decrease in HA days + HA peak intensity post-tx in the tx group compared to the control group. | • Improvement in HI (average reduction in HI was 31 % for tx group vs. 3 % for control group, p = 0.028). | 56 % Overall | |
• Cost-effectiveness: Estimated sum of therapist time = 40 hrs/participant. Cost-efficiency estimate: 0.78 | |||||||||
• NS in Headache Disability Inventory (HDI) or Beck Depression Inventory (BDI). | |||||||||
Multimodal including relaxation | RCT | Kleiboer, 2014 [54] | N = 368 (ages 18–65) HA specialist referral, website, flyers, newspaper/ magazine ads 8 weeks-11.4 weeks, 8 lessons, to be done in 7–10 days | Migraine with or without aura + 2–6 attacks/30 days prior to randomization | HA frequency; HA days; HA severity; Self-efficacy; QoL; Disability; LoC | • A 20–25 % decrease in migraine frequency was found for both the tx + control groups, NS between groups. | • BT (tx group) had significantly more improvement that the control group in migraine-related self-efficacy (p < 0.001, ES = 0.86), + developed more internal control (p,0.001, ES = 0.57) but less external control (p < 0.001, ES = 0.78). | 27.4 % Overall (39.0 % in tx group vs. 14.5 % in control group) | |
• A significant but small decrease in average attack peak intensity was seen in the ITT BT (tx) group from baseline to post-tx, but NS between groups. | |||||||||
• Compliance was explored in a random sample of 60 participants, which showed that participants reported conducting at least one relaxation exercise on 45.5 % of days of being in training. | |||||||||
Other | Descriptive study | Sorbi, 2010 [55] | N = 10 (ages 31–68) Individuals with recently expressed interest in self-management training 10 weeks | Migraine with 1–6 attacks/month | Acceptability | • All lessons were rated positively regarding clarity, instructiveness, importance + easy execution by new participants | 40 % Overall (New Participants) + 0 % overall (Expert Patients) | ||
• Expert patients provided positive ratings for the web application, digital support, + web-adaptation of the protocol. | |||||||||
Self-Management Program | Descriptive study (Interviews + concept mapping to develop Web-prototype + study feasibility) | Donovan, 2013 [56] | N = 12 (ages 12–17, adolescents), 9 (ages 30–55, caregivers) 12 (adults, clinicians).** Newspaper ads + community message board (adolescents + caregivers). Emails/invitation at conference (clinicians) 60 min (interview) + 30 min (acceptance testing via telephone) | Migraine | Acceptability | • Disagreement over content areas for the website-clinicians but not adolescents felt diet + exercise were important to include. | N/A | ||
• During the prototype evaluation, most adolescents indicated that the website would be useful (especially the “Ask an Expert” feature) when they felt a migraine coming on or had a migraine. | |||||||||
• Caregivers reported being “somewhat” to ‘extremely likely” to use the range of features offered on the website. | |||||||||
PDA | Multimodal including relaxation | Descriptive Study + Case Control study | Kleiboer, 2009 [57] | N = 44 (ages 25–63)*** HA websites, newspaper ad, referral by HA specialists 3 weeks | Migraine | HA frequency; QoL; LoC; Acceptability | • There were no significant improvements in HA frequency in the ODA + BT group (tx) compared to BT alone (control). | • ODA was considered feasible, well-accepted + perceived to support self-care. | 29.5 % Overall |
• There were no significant improvements in internal control or migraine-specific QoL in the ODA + BT group compared to BT alone. | |||||||||
Other | Descriptive pilot study (To establish feasibility) | Sorbi, 2007 [39] | N = 5 (ages 24–52) Unknown 8.5 days on average (range 4–12 days) | Migraine without aura | Acceptability | • In the second run, adherence was 85 %. | • ODA had good acceptability evidenced by positive participant responses | 0 % Overall | |
• Loss of data due to technical problems amounted to 6.8 % of potential diary entries + lost internet connection contributed to loss of 5.6 % of lost diary entries. | |||||||||
Other | Multimodal including biofeedback | Prospective, single-arm, open-label pilot study | Shiri, 2013 [58] | N = 10 (ages 10–17.5) Pediatric neurology clinic 6 months- 60 months | Chronic migraine or Chronic Tension-type HA | QoL; Activity limitation; Other | • Patients reported a decline in HA severity (VAS 4.28 pre-test vs. 3.11 post-test, p 0.015) + signficant improvements in daily function + quality of life. | • Improvement pre-tx to post-tx in quality of life (Pedsi QL) + daily function (measured by 2 questions on VAS scale) | 10 % Overall |
• Overall the participants reported they were satisfied with the tx. | |||||||||
Multimodal including Biofeedback | RCT | Scharff, 2002 [59] | N = 36 (ages 7–17) Referred from children’s hospital 6 weeks | Migraine with or without aura +/-co-existing tension-type HA | HA index; HA days; HA severity; DAS; Acceptability | • Change in # of HAs recorded + highest intensity rating over time, but there were no significant between-group differences. Likely due to small n + low power of the study. | • 53.8 % (7) of children in the handwarming biofeedback group, 10 % (1) in the handcooling biofeedback group, and 0 %(0) in the waitlist control group had a 50 % or more decrease in HI at the post tx. The significantly higher proportion of participants achieving 50 % reduction in HI in handwarming group vs. handcooling group was maintained at 3 month + 6 month follow-up. | 9.4 % Overall (0 % in handwarming group vs. 9.1 % in handcooling group vs. 8.3 % in WLC group) | |
• Adherence: Data from home practice records of 29 participants in handwarming or handcooling group indicated the average # of practice sessions was 5.3 times per week. | |||||||||
• NS in CDI or STAIC scores. | |||||||||
• There was a temperature change between the handwarming + handcooling groups, with the handwarming group more likely to report that their temperatures increased. | |||||||||
Other-Sound therapy | RCT (double blind, placebo-controlled study with a parallel group add-on design) | Trinka, 2002 [60] | N = 32 (ages 16–60) Outpatient HA clinic 12 weeks | Migraine with or without aura assessed by neurologists | DAS; Acceptability; Other | • Raw values of the “headache” subtest of the GBB improved in both groups but NS between groups. | No Adherence Data | ||
• NS in FPI-R, STAI or SDS | |||||||||
PMR+ Biofeedback | Prospective non randomized | Arena, 2004 [61] | N = 4 (ages 52–64) Medical Center 8 weeks | Migraine or combined migraine-tension HA | HA index; HA days; HA severity; Medication index | • 1 subj had 50 % or greater reduction in HI, 2 had some clinical improvement, 1 subject demonstrated no tx response | 0 % Overall | ||
Biofeedback | Prospective non randomized | Folen, 2001 [62] | N = unknown**** U.S Army/Navy hospitals Not specified | Migraine, Chronic daily HAs | Acceptability | • When evaluating the viability of the system in 2 separate rooms of the medical center, patient satisfaction was high (8/10) + patients produced physiologic changes in desired direction. | No adherence data | ||
• Total cost of the system about $9000 |