Scandinavian Journal of Gastroenterology. 2014; 49: 387–408
REVIEW
Utilization and efficacy of internet-based eHealth technology in gastroenterology: a systematic review
SIMON R. KNOWLES1,2,3,4 & ANTONINA MIKOCKA-WALUS5 1
Faculty of Life and Social Sciences, Swinburne University of Technology, Melbourne, Australia, 2Department of Mental Health, St Vincent’s Hospital, Melbourne, Australia, 3Department of Psychiatry, The University of Melbourne, Melbourne, Australia, 4Department of Gastroenterology, Royal Melbourne Hospital, Melbourne, Australia, and 5 Department of Health Sciences, University of York, York, YO10 5DD, UK
Abstract Objective. While there have been several reviews exploring the outcomes of various eHealth studies, none have been gastroenterology-specific. This paper aims to evaluate the research conducted within gastroenterology which utilizes internetbased eHealth technology to promote physical and psychological well-being. Material and methods. A systematic literature review of internet-based eHealth interventions involving gastroenterological cohorts was conducted. Searched databases included: EbSCOhost Medline, CINAHL, and PsycINFO. Inclusion criteria were studies reporting on eHealth interventions (both to manage mental health problems and somatic symptoms) in gastroenterology, with no time restrictions. Exclusion criteria were non-experimental studies, or studies using only email as primary eHealth method, and studies in language other than English. Results. A total of 17 papers were identified; seven studies evaluated the efficacy of a psychologically oriented intervention (additional two provided follow-up analyses exploring the original published data) and eight studies evaluated disease management programs for patients with either irritable bowel syndrome, inflammatory bowel disease (IBD) or celiac disease. Overall, psychological eHealth interventions were associated with significant reductions in bowel symptoms and improvement in quality of life (QoL) that tended to continue up to 12 months follow up. The eHealth disease management was shown to generally improve QoL, adherence, knowledge about the disease, and reduce healthcare costs in IBD, although the studies were associated with various methodological problems, and thus, this observation should be confirmed in well-designed interventional studies. Conclusions. Based on the evidence to date, eHealth internet-based technology is a promising tool that can be utilized to both promote and enhance gastrointestinal disease management and mental health.
Key Words: eHealth, gastroenterology, internet intervention
Introduction Gastrointestinal (GI) conditions, both functional and organic, are highly prevalent in general community: one-third of individuals experience functional GI disorders [1], up to 505 per 100,000 are affected by ulcerative colitis (UC), and 322 per 100,00 are diagnosed with Crohn’s disease (CD) [2]. Owing to their widespread occurrence and challenges associated with treatment, GI conditions are a significant burden on the healthcare system. Further, this burden is likely to rise
with the evidence that: (1) incidences of organic and functional conditions are on the increase [2,3], and (2) these chronic gastroenterological illnesses require increasingly more complex integrated care models of care [4–6]. Therefore, new cost–effective approaches to augment traditional medical healthcare are needed. The eHealth telemedicine (such as internet-delivered information, support, and monitoring) is one such potential approach. The growth of studies evaluating eHealth technology has resulted in four Cochrane systematic
Correspondence: Antonina Mikocka-Walus, Senior Lecturer, Department of Health Sciences, University of York, York, YO10 5DD, UK. Tel: +0044 1904 321521. E-mail: [email protected]
(Received 20 September 2013; revised 6 November 2013; accepted 8 November 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.865259
388
S. R. Knowles & A. Mikocka-Walus
reviews of the area [7–10]. While identifying the potential benefits, the reviews also clearly identify that the evidence on the effectiveness of eHealth in the general medical population is limited. Key limitations of the research conducted include lacking cost/benefit analyses and utilizing heterogeneous outcome measures which cannot be combined to present a logical picture. As identified in Table I, there are many benefits in relation to utilizing eHealth internet technology, including accessibility to large clinical samples, accessibility 24 h a day, and accessibility to rural and remote clinical cohorts. However, its disadvantages include expenses associated with development and implementation, limitations with individualization, and increased attrition rates compared to face-to-face interventions. Within medicine, the mental health domain has conducted the most extensive research into the application of eHealth technology. Research has shown that when comparing face-to-face versus online psychotherapies, online psychotherapies can be equally informative and efficacious [11] and provide significant cost savings [12]. To date, only a few researchers have explored the potential utilization of eHealth internet technology in gastroenterology and no systematic review of this area has been undertaken. This paper systematically reviews the available studies with an aim to: (1) explore the potential of eHealth internet technology to augment traditional medical care in gastroenterology for the benefit of patients and the community and (2) to provide recommendations in relation to future internet-based eHealth gastroenterology interventions. Methods Types of studies Only experimental studies meeting the below inclusion criteria were included in the review.
Sources A systematic and comprehensive literature search of EbSCOhost Medline, CINAHL, and PsycINFO were conducted in September 2013 using the search strategy listed below. Search strategy The search strategy was developed in collaboration with a research librarian. The following keywords were used. EbSCOhost Medline: (MH “Electronic Mail” MH “Internet+” TI website* or TI etherap* or TI e-therap* or TI online or TI internet or TI web or TI ehealth or TI e-health) AND (MH “Gastrointestinal Diseases+” TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 490 journal articles. EbSCOhost CINAHL Plus with Full Text: (MH “Electronic Mail” MH “Internet+” TI website* or TI etherap* or TI e-therap* or TI online or TI internet or TI web or TI ehealth or TI e-health) AND (MH “Gastrointestinal Diseases+” TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 780 journal articles. EbSCOhost PsycINFO: (online therapy or DE “Online Therapy” or internet or DE “Internet”) AND (Gastrointestinal Disorders OR DE “Gastrointestinal Disorders” OR DE “Colon Disorders” OR DE “Gastrointestinal Ulcers” OR DE “Vomiting” or TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 23 journal articles. Reference lists of the identified papers were also searched for identifying relevant studies.
Inclusion criteria
Data collection and analysis
Studies reporting on eHealth interventions (both to manage mental health problems and somatic symptoms) in gastroenterology with no time restriction were included. The eHealth interventions were defined as those studies that undertake an experimental design using the internet as their primary method of delivery (not email, or mobile apps or telehealth, e.g., skype).
To reduce potential bias, two reviewers (SK and AMW) independently inspected the articles identified by the search after initially agreeing on data extraction and interpretation. Any disagreements regarding the inclusion of the studies were resolved by discussion. The systematic review was undertaken based on the recommended PRISMA statement (http://www. prisma-statement.org) guidelines (Suppl checklist).
Exclusion criteria Data extraction Non-experimental studies, studies using only email as primary eHealth method, and studies in language other than English were excluded from the review.
Extracted data included the following: author, year of publication, country of origin, setting, design (type of
eHealth interventions in gastroenterology
389
Table I. Benefits and challenges of eHealth interventions in gastroenterology. Benefits: Access to large clinical populations; accessible 24 h, 365 days a year. Patients can assess the eHealth service at a time and place convenient to them; it can be integrated with traditional clinical services to provide feedback, monitoring or additional augmented services; it can be relatively cheap to maintain and can utilize multiple audio-visual materials to promote learning and engagement. It enables access to patients in rural and remote areas and can promote patient engagement and empowerment and decision making in relation to their condition. Cheaper than traditional services which are also often limited, it may offer improved care and thus better QoL in patients and consequently less burden on healthcare system Challenges: Expensive to setup, especially if it involves online diagnosis, and condition-focused; expensive to make individualized high rates of attrition compared to face-to-face therapies. It is highly dependent on participant engagement and motivation, language and computer skills and thus not useful for people with cognitive impairment, low education, coming from various ethnic backgrounds. It is unable to validate physical diagnoses and symptoms; it is based on self-report Quality assurance – difficult for the patient to differentiate quality eHealth information and services. It is inappropriate for individuals with severe mental illness or an undiagnosed condition and is not available for people having limited access to computers or low literacy and thus may be disadvantageous for older people and people on limited income. There is a lack of personal relationship between a patient and a clinician which is a significant contributor to better patient outcome
trial, blinding, or control group), patient characteristics and sample size, treatment (type of intervention, number of sessions, duration, and content), outcome measures, monitoring of adherence and dropout rates, and results of main outcome measures. Of the 1292 abstracts identified, 17 papers met the inclusion criteria. Of the 17 identified, two studies were found to provide post-analyses of original published papers focusing on either attrition [13] or costeffectiveness [14] and thus the review included 15 separate studies presented in 17 papers (see the PRISMA flowchart attached) (Figure 1). Data analysis A meta-analysis was not undertaken due to methodological differences between the studies, particularly with respect to types of design (not all studies were randomized controlled trials [RCTs]), interventions, populations studied, and outcomes measured. Effect sizes are used where available to compare treatment effectiveness. Critical appraisal was conducted based on gastroenterology-specific recommendations adapted from Knowles et al. [15] (see Table II). Results Two trends in eHealth studies in gastroenterology were identified: (1) eHealth studies targeting GI symptoms and mental health and (2) eHealth studies in disease management. The general observation we made was that the former have been of general higher quality and better established than the latter, which have mostly been preliminary in nature and characterized by small sample sizes and various methodological flaws (see Table III). eHealth studies targeting GI symptoms and mental health A total of seven GI-focused studies (nine papers) were found to evaluate eHealth mental health
interventions; six involved irritable bowel syndrome (IBS) cohorts [16–21], and one involved an inflammatory bowel disease (IBD) cohort (see Table IV) [22]. Studies included 657 participants; six studies included middle-aged samples, while one study involved females adolescent and parent dyads. Five studies were RCTs (two pre-post only, one pre-post 3-month follow up, one pre-post 6-month follow up with an active control group [stress management], and one pre-post 12-month follow up) [18–21,23], two involved non-randomised with wait-list/control groups (one with post 3-month follow up and two with a 6-month follow up) [16,22]. In relation to the style of intervention, one involved skills-based training [22], five involved cognitivebehavior therapy (CBT) [18–21,23], and one involved expressive writing [16]. In their intervention, McCormick et al. [22] included adolescent IBD females and one of their parents to undertake a blended online/face-to-face intervention. Participants completed a group-based face-to-face 6-h training course and then weekly online chat sessions. In relation to the five CBT interventions involving IBS cohorts [18–21,23], pre-post analyses indicated that there were improvements in IBS symptoms and quality of life (QoL). One study reported reduction in pain [19], while three reported reduction in visceral sensitivity [19,21,23]. At follow up, all five studies reported improvements in QoL, three studies reported that positive changes were maintained in IBS symptoms [18,20,21]. Visceral sensitivity was also found to remain lower at post-follow up for three studies [18,20,21], with a further study indicating improvement being maintained [23]. In a cost–effectiveness evaluation using 12-month follow-up data from a 10-week CBT intervention with therapist email contact [19], Andersson et al. [14] concluded that 36% of the individuals who reported clinically reduced IBS symptoms (27 out of 76) also saved the community $16,806 per person. In a second cost–effectiveness study, Ljótsson et al. [23]
S. R. Knowles & A. Mikocka-Walus
Identification
390
Records identified through database searching (n = 1290)
Additional records identified through other sources (n = 2)
Included
Eligibility
Screening
Records after duplicates removed (n = 67)
Records screened (n = 1225)
Records excluded (n = 1140) based on the screening of title and abstract
Full-text articles assessed for eligibility (n = 85)
Full-text articles excluded, with reasons (n = 68)
Studies included in qualitative synthesis (n =17)
57 were not interventions 10 did not consider a GI population 1 used emails as the only eHealth intervention.
Studies included in quantitative synthesis (meta-analysis) (n = NA) Figure 1. PRISMA 2009 flow diagram. From [42].
conducted the only cost–benefit analysis showing that the treatment had an 87% chance of producing societal gains as well as health gains, saving around $39,821 in associated medical costs. Based on four consecutive 30-min expressivewriting intervention involving IBS patients, Halpert et al. [16] reported that there was an improvement in IBS symptoms post-intervention and that this continued at 3 months follow up. Using the blended face-to-face and online format with an IBD cohort, McCormick et al. [22] found that there were no significant differences between the treatment and wait-list groups at post-treatment with the exception of parent-reported approach subscale on the pain coping questionnaire (PCQ). Within-group analyses indicated that the adolescent treatment group reported significant reductions in somatic symptoms. Reports by parents indicated reduced children’s use of approach coping strategies and distraction techniques. In a follow-up study of McCormick et al. [22], ReedKnight et al. [24] evaluated the factors that influenced adolescent and parent participation and attrition. No significant differences were found in relation to medical or demographic factors and online participation
rates. Reduced adolescent participation was associated with increased adolescent- and parent-perceived abdominal pain, somatic symptoms, parent-reported catastrophizing and adolescent depressive symptoms. Parent participation was also found to have a significant positive association with adolescent participation. Owing to the small number of heterogeneous studies, a meta-analysis of the studies was not possible. However, review of the effect sizes from Ljótsson et al. IBS-based studies [19–21,23] indicated medium-tolarge effect sizes (e.g. pre-post IBS symptoms ranged from 0.38 to 1.21, QoL ranged from 0.51 to 91 and visceral sensitivity ranged from 0.33 to 0.79). Hunt et al. [18] found similar medium-to-large effect sizes (IBS symptoms = 1.21, QoL = 1.08), while McCormick et al. [22] found small-to-moderate effects sizes. When compared to Hunt et al., the smaller effect size found by McCormick et al. [22] may be due to several factors including differences in rates of change or possible change (e.g. pain symptoms may be less amenable to change vs. anxiety), associated with functional versus organic GI conditions (i.e. IBS vs. IBD), and intervention type and format (e.g. online vs. mixed-method). Mixed evidence was found in
eHealth interventions in gastroenterology
391
Table II. Ideal intervention eHealth gastroenterology requirements. Protocol characteristics: RCT design; makes sure that randomization is adequately conducted to avoid significant group differences; prior to intervention, power analyses are conducted (including correction for potential dropouts of at least 20%); comprehensive and clear details of intervention content are made fully available to allow other researchers to review and undertake accurate replication. When targeting psychological distress, the interventions should be GI-specific (i.e. if a patient has IBS and panic disorder, the intervention should be targeted to panic disorder using IBSoriented content). It includes measures that assess cost/benefit analysis and healthcare utilization; it clearly identifies intervention type (e.g. CBT) and format (e.g. modulized with email correspondence). Where possible, assessors and statisticians are blinded; in psychological trials, control for possible moderating/mediating factors is identified in the research literature, for example, stress, personality, social support, and self-efficacy. Base psychological interventions are made on validated theoretical approaches. Clear inclusion and exclusion criteria are identified. In psychological trials, it identifies (and where possible control for) if participants have, or have not, received, or currently receiving psychotherapy (including type and duration), have used psychotropic medication, and/or are on specialized diets. It includes long-term posttherapy efficacy review time points (i.e. 1- and 2-year post-intervention). Adequately selected control group, with active control groups are particularly encouraged Participant characteristics: Larger sample sizes are needed. It differentiates between non-active versus active condition status and condition type (e.g. IBS subtypes). It differentiates between disease subtypes (e.g. CD and UC) In psychological studies, it conducts psychological assessment to attain formal diagnosis of psychological disorders (e.g. generalized anxiety versus generalized anxiety with major depressive episode); pre-post intervention Data collection and analyses characteristics: It utilizes valid measures which can be accurately compared to other intervention studies. Where possible, it utilizes validated questionnaires and biomarkers and includes power analyses. Clear protocol for managing missing data is reported. It evaluates participant engagement in therapy (e.g. time accessing intervention materials and engaging in homework) and reports effect sizes and attrition rates. It evaluates results based on more statistically rigorous multilevel modeling techniques. This would allow for exploration of main effect changes as well as subgroups changes, while being able to control for confounding factors (e.g. stress, disease status, medication type and changes). It also evaluates differences between completers and non-completers Adapted from Knowles et al. [15].
regard to the improvement in mental health, with one study showing an improvement in depression [19] and two studies showing no significant change in anxiety and/or depression scores [20,23]. Regarding retention of participants in the study prepost intervention rates ranged from 46% to 99% and follow-up rates ranging from 36% (3-month follow up) to 85% (6-month follow up). Attrition rates were similar for those in the control groups. eHealth studies into disease management Overall, eight studies on eHealth in disease management have been identified (see Table V). Of these, three focused on patients with both CD and UC, three on UC only, one on CD only and one on celiac disease. Studies included 714 participants; six studies included middle-aged samples, while one study was focused on children and one on older participants. Four studies were RCTs (two with a 12-month follow up, one with a 9-month follow up, and one with a 3-month follow up), three studies were before-andafter trials (two with a 6-month follow up and one with a 5-month follow up), and one study was an openlabel trial with historical controls. Three studies were focused on monitoring disease activity, two on treatment adherence and side-effects, one on disease activity and adherence, one on overall management, and one on diet adherence. Studies by Elkjear et al. and Pendersen et al. [25–27] came from the same group as did the studies by Cross et al. [28,29], and
thus, of the seven included studies, there were only four different interventions (one from Elkjear et al., one from Cross et al., one from Hommel et al. and one from Krier et al.). In controlled studies [13,26–28,30], good experience was noted with eHealth applications and, in general, no group difference in QoL, medication adherence, hospitalization rates, surgery frequency or use of corticosteroids was noted. However, one open-label trial with historical controls observed significantly more adverse events in controls than in the intervention group (p = 0.002) [27]. In two RCTs [13,26], improved disease knowledge and compliance with treatment/diet were noted (p < 0.05). In withinparticipant trials [25,29,31], high acceptance and satisfaction with eHealth have been reported as have improved QoL, knowledge, and adherence and reduced costs, although the latter studies did not include control group comparisons. Regarding retention of participants in the study, at follow up, it ranged from 50% to 84% in intervention groups as compared to 74% to 92% in controls, thus indicating higher dropouts in participants receiving eHealth in comparison to standard care. Discussion The aim of this systematic review was to explore the potential of eHealth internet technology to augment traditional medical care in gastroenterology for the benefit of patients and the community and provide
392
S. R. Knowles & A. Mikocka-Walus
Table III. Study strength/weaknesses of the reviewed studies on eHealth in gastroenterology. eHealth studies on mental health Study
Strengths
Weaknesses
Hunt et al. [18]
First IBS online study, involving patient–therapist communication. Included wait-list control (WLC) group and a follow-up for those in the intervention group. Explored differences in participant characteristics of completers and non-completers.
McCormick et al. [22]
It combined face-to-face and online intervention; rewards were provided after each assignment completion/engagement in chat sessions; it utilized parent-adolescent dyads. Follow-up study [24] of the McCormick et al. [22] dataset explored participation and attrition rates.
Ljótsson et al. [19]
It conducted power analysis indicating that 50 participants would be valid and sufficient; however, authors aimed to recruit 80 patients. Assessed for diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patienttherapist communication, and patients were required to complete homework to continue to next module. A follow-up study [14] based on the Ljótsson et al. [19] dataset explored the cost–effectiveness of the intervention. It was conducted based on CONSORT statement for non-pharmacological trials, conducted power analyses, assessed for diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patient-therapist communication, and patients were required to complete homework to continue to next module. It used consecutive sample from a GI service, assessed healthcare costs and diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patient-therapist communication, and patients were required to complete homework to continue to next module. It explored 12-month follow up and reviewed completers versus non-completers.
No assessment (or possible controlling) of depression, stress, diet, illness perceptions was made, and excluded those with other GI conditions, but did not assess/ account for other physical or mental health conditions. It did not assess or control medication use. No poweranalysis was reported and no identification was made as to years with IBS; it had self-selected cohort. No assessment (or possible controlling) of depression, stress, diet, illness perceptions was made; there is no report of excluding other comorbid physical or mental health conditions. It had small sample size, nonrandomized allocation to conditions, therapy was conducted by combination of clinical psychologist or graduate clinical psychology students, and the study was limited to adolescent female sample. No power-analysis was reported. No assessment (or possible controlling) of medication use, stress, illness perceptions, and mindfulness was made. It had a self-selected cohort.
Ljótsson et al. [20]
Ljótsson et al. [23]
Ljótsson et al. [21] (12-month follow up of Ljótsson et al. [19]) Halpert et al. [16]
A useful pilot study, which explored 3-month follow up and reviewed completers versus non-completers.
eHealth studies into disease management Elkjaer et al. [26] Two international centers, used a preliminary study [25] to develop a well thought through intervention; two control groups were included. Elkjaer et al. [25] Pedersen et al. [27]
A useful companion study to Elkjaer et al. [26] Inclusion of a control group, and long follow up
Cross and Finkelstein [29] Cross et al. [28]
A useful pilot study informing a trial [28] RCT design and long follow up
No assessment (or possible controlling) of medication use, stress, illness perceptions, and mindfulness was made. It had a self-selected cohort.
It used VSI as a measure of coping and did not assess depression or anxiety, control for stress, and mindfulness.
There was no assessment (or possible controlling) of medication use, depression, anxiety, or stress, illness perceptions, and mindfulness. It had a self-selected cohort. There was no assessment (or possible controlling) of medication use, depression, anxiety or stress, no power analysis, and conducted an additional post-hoc exploratory subscale analyses to support hypotheses. It had a self-selected cohort. Power calculation was not presented and no comment to power of analyses was included; there was a small Irish sample which may explain some insignificant differences in this group. There were no control group and exclusion criteria. There were historical controls and no group comparisons on main outcome measures were conducted. It had a within-participant design and no control group Imbalance in baseline characteristics between groups; a significant proportion of intervention group participants discontinued intervention; sample size: the power calculation was for 84 participants but only
eHealth interventions in gastroenterology
393
Table III. (Continued). eHealth studies on mental health Study
Hommel et al. [31]
Krier et al. [30]
Sainsbury et al. [13]
Strengths
Weaknesses
A useful pilot study, first telehealth study on no adherence to treatment in IBD, prospective design, cost data An RCT design, long follow up, and excellent attrition rate An RCT design, adequate sample size with good calculation, and comprehensive methodology reporting
recommendations in relation to future internet-based eHealth gastroenterology interventions. Two trends in eHealth studies in gastroenterology were identified: (1) eHealth studies targeting GI symptoms and mental health and (2) eHealth studies in disease management. In relation to eHealth interventions targeting GI symptoms and mental health, seven studies were identified (six involving IBS cohorts and one involving an IBD cohort). Five of the seven eHealth interventions utilized an RCT design and included either 3-, 6-, or 12-month follow-up assessments, while the remaining studies utilized control groups with 3- and 6-month follow ups, respectively. Despite the heterogeneity of the IBS-focused studies (e.g. differences in intervention type, assessment methods, and time-points), the findings indicate that significant benefits were reported by participants with medium-to-large effect sizes in QoL and IBS symptoms. Although found to be beneficial, McCormick et al. [22] reported small-to-moderate effect sizesacrossmultiple adolescent–parent dyad measures (e.g. adolescent somatic symptoms, parent irrational cognitions, and protective responses to pain/IBD symptoms). These findings are consistent with face-toface/group-based interventions targeting GI symptoms and mental health (see reviews: Zijdenbos et al. [32], Timmer et al. [33], and Knowles et al. [15]), in that it can be argued that individuals who experience mental distress along with their physical health condition are likely to benefit from a focused psychological intervention. It is, however, too early to identify if targeted eHealth psychological interventions may be more beneficial for different illness cohorts (i.e. functional vs. organic GI conditions). In relation to eHealth interventions targeting disease management, eight studies were identified (seven involving IBD cohorts and one involving a celiac disease cohort). Four of the eight eHealth interventions utilized an RCT design and included either 3-, 9-, or 12-month follow-up assessments, while the remaining studies utilized either a before-and-after
41 completed intervention and thus study was underpowered It had no control group, a small sample, and did only basic cost analysis. No power calculation provided and thus unknown whether the study is underpowered, and had a small sample size Nearly 50% of the intervention group participants did not complete the intervention; short follow up
trials (with 5- or 6-month follow up) or an open-label trial compared with historical controls. The interventions targeting disease management to date suggest that eHealth may be a preferred way for patients to help monitor their illness and also enhance compliance with the recommended treatment plan. While still too early to argue its definitive benefits, the current positive findings in the eHealth studies reflect the published literature and indicate that interventions that enhance patient self-efficacy and that are developed to facilitate patient engagement are likely to be useful in managing chronic GI conditions [34]. Recommendations Based on the reviewed papers, we suggest that gastroenterology is in a strong position to utilize eHealth internet technology to provide augmented services for promoting patient general well-being. While participants generally identify a high acceptance of eHealth interventions, the retention rates may actually be poorer than traditional face-to-face interventions. Given this observation, while eHealth interventions seem to provide benefit, further research is needed to better identify when this (and what) form of treatment is more appropriate, what interpersonal/social factors mediate engagement (e.g. self-efficacy and parental involvement), and what participant characteristics help to determine this. For example, is an eHealth intervention efficacious when the patient cohort experiences significant distress or in an active phase of their GI condition? Given the diversity of treatment interventions and their format, elucidating these important factors and possible mediating processes will be important. Further, more research is needed to evaluate acceptability for e-health interventions and if, and how, differences in patient needs can be addressed in eHealth interventions. For example, traditional faceto-face and group-based interventions, while usually
IBS
IBD
Hunt et al. [18]
McCormick et al. [22]
Authors
Condition RCT; pre-post assessment, post 3-month assessment (intervention only)
NRTC; pre-post and follow up (6 months). Participant medical records (n = 215) screened for eligibility. Inclusion/exclusion criteria were not identified, although later reported in Reed-Knight et al. [24]
N = 31 female adolescent-parent dyads (n = 20 treatment, n = 11 wait-list). Regarding adolescents: mean age = 14.46 years; mean years diagnosed = 3.15
Design and assessments
N = 54 (44 females; n = 28 intervention, n = 26 wait-list; mean age = 38 years) were invited to participate via IBS relevant websites
Participant details
Participants completed a 6-h training skills course on managing pain, coping with IBD, communicating with others, and for parents only setting limits skills. Following this day, participants completed 6 weekly web-based activities (separate activities for adolescents and parents). Separate
5 modules (1 per week): Module 1: Psychoeducation and relaxation training; Module 2: cognitiveoriented stress management (SM); Module 3: challenging catastrophic thinking; Module 4: introduction to exposure therapy; Module 5: exposurebased experiments. Homework emailed to primary investigator who then provided individualized responses within 48 h. Inclusion criteria: self-report diagnosis of IBS by medical professional with no other GI condition
Intervention
Table IV. Summary of eHealth mental health interventions for gastroenterological conditions.
Pre-post: GSRS-IBS = 1.21; IBS-QoL = 1.08; ASI-GI = 0.63; ASI non-GI = 0.70 CPSQ GI = 1.83; SPSQ-non-GI = 1.17
Effect sizes
Main findings
Significant improvement across all measures in the intervention group, with QoL and VAS at post3-month follow up. Nonsignificant trend suggesting that GI symptoms and catastrophizing increased at post-3-month follow up; based on Jacobson and Truax [40] method for identifying clinical significance; intervention was found to be superior to control-group (GSRS unchanged intervention 62% vs. 83% control; IBS-QoL unchanged intervention 54%; 83% control) Attrition: Of the initial 28 participants in the intervention group, 13 (46%) completed the intervention and postintervention questionnaire, 10 (36%) completed post3-month assessment. For the control group, of the 26, 18 (69%) completed the control phase, 8 (31%) completed the intervention, and 2 (8%) completed the post-3-month follow up Adolescent primary Pre-post: No significant differences outcome measures: parent-reported approach were found between the Abdominal Pain Index subscale on the PCQ = 0.28 treatment and wait-list (API), Child Within-group pre-post groups at post-treatment Somatization Inventory analyses: with the exception of (CSI), PCQ, Pain Treatment group – parent-reported approach Catastrophizing Scale API = 0.05 (adolescent); subscale on the PCQ. (PCS) API = 0.03 (parent); Within-group analyses Parent primary outcome CSI = 0.41 (adolescent); indicated for the treatment measures: API = 0.38 (parent); group adolescents reported Abdominal Pain Index within-group pre-follow up: reductions in somatic (API), Child Treatment group – symptoms. ParentsSomatization Inventory API = 0.24 (adolescent); reported reductions were
GI Symptom Rating Scale-IBS (GSRS-IBS), Anxiety Sensitivity Index (ASI), Visceral Anxiety Sensitivity (VAS), IBS-QoL, and the Consequences of Physical Sensations Questionnaire (CPSQ)
Outcome measures
394 S. R. Knowles & A. Mikocka-Walus
Ljótsson et al. IBS [19]
Authors
Condition
Table IV. (Continued). Design and assessments
RCT; patients were screened via-telephone to rule out other GI conditions, food intolerances and completed the Miniinternational Neuropsychiatric Interview (MINI). Responses from interview were assessed by gastroenterologist to confirm eligibility for inclusion. The study excluded patients with severe depression, substance dependence, psychosis, mania, and anorexia
Participant details
N = 85 (72 females; n = 42 intervention, n = 43 wait-list; mean age = 34.6 years; mean years diagnosed = 6.3) self-referred IBSpatients from Sweden were recruited
Primary outcome measures: weekly GI symptom diary, weekly GSRS-IBS. Secondary outcome measures: Montgomery Asberg Depression Rating Scale (MADRS-S), IBS-QoL, VSI, Sheehan Disability Scales (SDS), and Treatment Credibility Scale(TCS)
(CSI), PCQ, Pain Catastrophizing Scale (PCS), Adult response to Children’s Symptoms: Protect Scale (ARCS)
adolescent/parent online 30-min chat sessions occurred once a week with a reminder call before the chat session. Participants received $20 for completing each weekly homework assignment and $25 at the end of the treatment program with a minimum completion of 5 out of 6 homework and chat sessions
10-week 5-module CBT-protocol based on a previously published group intervention ([41]). Module 1: rationale for treatment and mindfulness; Modules 2, 3, and 4 identified psychological model of IBS and the associated cognitive and behavioral factors that exacerbate or attenuate IBS symptoms and the focus on them; Module 5: focus on exposurebased activities. Completion of homework was the
Outcome measures
Intervention
Pre-post: GSRS-IBS = 1.21; IBS-QoL = 0.93; VSI = 0.64; MADRS-S = 0.43; SDS = 0.47
API = 0.08 (parent); CSI = 0.06 (adolescent); API = 0.35 (parent); within-group pre-post analyses: Control group – API = 0.07 (adolescent); API = 0.12 (parent); CSI = 0.12 (adolescent); API =
REVIEW
Utilization and efficacy of internet-based eHealth technology in gastroenterology: a systematic review
SIMON R. KNOWLES1,2,3,4 & ANTONINA MIKOCKA-WALUS5 1
Faculty of Life and Social Sciences, Swinburne University of Technology, Melbourne, Australia, 2Department of Mental Health, St Vincent’s Hospital, Melbourne, Australia, 3Department of Psychiatry, The University of Melbourne, Melbourne, Australia, 4Department of Gastroenterology, Royal Melbourne Hospital, Melbourne, Australia, and 5 Department of Health Sciences, University of York, York, YO10 5DD, UK
Abstract Objective. While there have been several reviews exploring the outcomes of various eHealth studies, none have been gastroenterology-specific. This paper aims to evaluate the research conducted within gastroenterology which utilizes internetbased eHealth technology to promote physical and psychological well-being. Material and methods. A systematic literature review of internet-based eHealth interventions involving gastroenterological cohorts was conducted. Searched databases included: EbSCOhost Medline, CINAHL, and PsycINFO. Inclusion criteria were studies reporting on eHealth interventions (both to manage mental health problems and somatic symptoms) in gastroenterology, with no time restrictions. Exclusion criteria were non-experimental studies, or studies using only email as primary eHealth method, and studies in language other than English. Results. A total of 17 papers were identified; seven studies evaluated the efficacy of a psychologically oriented intervention (additional two provided follow-up analyses exploring the original published data) and eight studies evaluated disease management programs for patients with either irritable bowel syndrome, inflammatory bowel disease (IBD) or celiac disease. Overall, psychological eHealth interventions were associated with significant reductions in bowel symptoms and improvement in quality of life (QoL) that tended to continue up to 12 months follow up. The eHealth disease management was shown to generally improve QoL, adherence, knowledge about the disease, and reduce healthcare costs in IBD, although the studies were associated with various methodological problems, and thus, this observation should be confirmed in well-designed interventional studies. Conclusions. Based on the evidence to date, eHealth internet-based technology is a promising tool that can be utilized to both promote and enhance gastrointestinal disease management and mental health.
Key Words: eHealth, gastroenterology, internet intervention
Introduction Gastrointestinal (GI) conditions, both functional and organic, are highly prevalent in general community: one-third of individuals experience functional GI disorders [1], up to 505 per 100,000 are affected by ulcerative colitis (UC), and 322 per 100,00 are diagnosed with Crohn’s disease (CD) [2]. Owing to their widespread occurrence and challenges associated with treatment, GI conditions are a significant burden on the healthcare system. Further, this burden is likely to rise
with the evidence that: (1) incidences of organic and functional conditions are on the increase [2,3], and (2) these chronic gastroenterological illnesses require increasingly more complex integrated care models of care [4–6]. Therefore, new cost–effective approaches to augment traditional medical healthcare are needed. The eHealth telemedicine (such as internet-delivered information, support, and monitoring) is one such potential approach. The growth of studies evaluating eHealth technology has resulted in four Cochrane systematic
Correspondence: Antonina Mikocka-Walus, Senior Lecturer, Department of Health Sciences, University of York, York, YO10 5DD, UK. Tel: +0044 1904 321521. E-mail: [email protected]
(Received 20 September 2013; revised 6 November 2013; accepted 8 November 2013) ISSN 0036-5521 print/ISSN 1502-7708 online 2014 Informa Healthcare DOI: 10.3109/00365521.2013.865259
388
S. R. Knowles & A. Mikocka-Walus
reviews of the area [7–10]. While identifying the potential benefits, the reviews also clearly identify that the evidence on the effectiveness of eHealth in the general medical population is limited. Key limitations of the research conducted include lacking cost/benefit analyses and utilizing heterogeneous outcome measures which cannot be combined to present a logical picture. As identified in Table I, there are many benefits in relation to utilizing eHealth internet technology, including accessibility to large clinical samples, accessibility 24 h a day, and accessibility to rural and remote clinical cohorts. However, its disadvantages include expenses associated with development and implementation, limitations with individualization, and increased attrition rates compared to face-to-face interventions. Within medicine, the mental health domain has conducted the most extensive research into the application of eHealth technology. Research has shown that when comparing face-to-face versus online psychotherapies, online psychotherapies can be equally informative and efficacious [11] and provide significant cost savings [12]. To date, only a few researchers have explored the potential utilization of eHealth internet technology in gastroenterology and no systematic review of this area has been undertaken. This paper systematically reviews the available studies with an aim to: (1) explore the potential of eHealth internet technology to augment traditional medical care in gastroenterology for the benefit of patients and the community and (2) to provide recommendations in relation to future internet-based eHealth gastroenterology interventions. Methods Types of studies Only experimental studies meeting the below inclusion criteria were included in the review.
Sources A systematic and comprehensive literature search of EbSCOhost Medline, CINAHL, and PsycINFO were conducted in September 2013 using the search strategy listed below. Search strategy The search strategy was developed in collaboration with a research librarian. The following keywords were used. EbSCOhost Medline: (MH “Electronic Mail” MH “Internet+” TI website* or TI etherap* or TI e-therap* or TI online or TI internet or TI web or TI ehealth or TI e-health) AND (MH “Gastrointestinal Diseases+” TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 490 journal articles. EbSCOhost CINAHL Plus with Full Text: (MH “Electronic Mail” MH “Internet+” TI website* or TI etherap* or TI e-therap* or TI online or TI internet or TI web or TI ehealth or TI e-health) AND (MH “Gastrointestinal Diseases+” TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 780 journal articles. EbSCOhost PsycINFO: (online therapy or DE “Online Therapy” or internet or DE “Internet”) AND (Gastrointestinal Disorders OR DE “Gastrointestinal Disorders” OR DE “Colon Disorders” OR DE “Gastrointestinal Ulcers” OR DE “Vomiting” or TI gastrointestinal or TI colitis or TI irritable bowel or TI inflammatory bowel or TI IBS or TI IBD or TI FGID). Papers were limited to those written in English and the search resulted in 23 journal articles. Reference lists of the identified papers were also searched for identifying relevant studies.
Inclusion criteria
Data collection and analysis
Studies reporting on eHealth interventions (both to manage mental health problems and somatic symptoms) in gastroenterology with no time restriction were included. The eHealth interventions were defined as those studies that undertake an experimental design using the internet as their primary method of delivery (not email, or mobile apps or telehealth, e.g., skype).
To reduce potential bias, two reviewers (SK and AMW) independently inspected the articles identified by the search after initially agreeing on data extraction and interpretation. Any disagreements regarding the inclusion of the studies were resolved by discussion. The systematic review was undertaken based on the recommended PRISMA statement (http://www. prisma-statement.org) guidelines (Suppl checklist).
Exclusion criteria Data extraction Non-experimental studies, studies using only email as primary eHealth method, and studies in language other than English were excluded from the review.
Extracted data included the following: author, year of publication, country of origin, setting, design (type of
eHealth interventions in gastroenterology
389
Table I. Benefits and challenges of eHealth interventions in gastroenterology. Benefits: Access to large clinical populations; accessible 24 h, 365 days a year. Patients can assess the eHealth service at a time and place convenient to them; it can be integrated with traditional clinical services to provide feedback, monitoring or additional augmented services; it can be relatively cheap to maintain and can utilize multiple audio-visual materials to promote learning and engagement. It enables access to patients in rural and remote areas and can promote patient engagement and empowerment and decision making in relation to their condition. Cheaper than traditional services which are also often limited, it may offer improved care and thus better QoL in patients and consequently less burden on healthcare system Challenges: Expensive to setup, especially if it involves online diagnosis, and condition-focused; expensive to make individualized high rates of attrition compared to face-to-face therapies. It is highly dependent on participant engagement and motivation, language and computer skills and thus not useful for people with cognitive impairment, low education, coming from various ethnic backgrounds. It is unable to validate physical diagnoses and symptoms; it is based on self-report Quality assurance – difficult for the patient to differentiate quality eHealth information and services. It is inappropriate for individuals with severe mental illness or an undiagnosed condition and is not available for people having limited access to computers or low literacy and thus may be disadvantageous for older people and people on limited income. There is a lack of personal relationship between a patient and a clinician which is a significant contributor to better patient outcome
trial, blinding, or control group), patient characteristics and sample size, treatment (type of intervention, number of sessions, duration, and content), outcome measures, monitoring of adherence and dropout rates, and results of main outcome measures. Of the 1292 abstracts identified, 17 papers met the inclusion criteria. Of the 17 identified, two studies were found to provide post-analyses of original published papers focusing on either attrition [13] or costeffectiveness [14] and thus the review included 15 separate studies presented in 17 papers (see the PRISMA flowchart attached) (Figure 1). Data analysis A meta-analysis was not undertaken due to methodological differences between the studies, particularly with respect to types of design (not all studies were randomized controlled trials [RCTs]), interventions, populations studied, and outcomes measured. Effect sizes are used where available to compare treatment effectiveness. Critical appraisal was conducted based on gastroenterology-specific recommendations adapted from Knowles et al. [15] (see Table II). Results Two trends in eHealth studies in gastroenterology were identified: (1) eHealth studies targeting GI symptoms and mental health and (2) eHealth studies in disease management. The general observation we made was that the former have been of general higher quality and better established than the latter, which have mostly been preliminary in nature and characterized by small sample sizes and various methodological flaws (see Table III). eHealth studies targeting GI symptoms and mental health A total of seven GI-focused studies (nine papers) were found to evaluate eHealth mental health
interventions; six involved irritable bowel syndrome (IBS) cohorts [16–21], and one involved an inflammatory bowel disease (IBD) cohort (see Table IV) [22]. Studies included 657 participants; six studies included middle-aged samples, while one study involved females adolescent and parent dyads. Five studies were RCTs (two pre-post only, one pre-post 3-month follow up, one pre-post 6-month follow up with an active control group [stress management], and one pre-post 12-month follow up) [18–21,23], two involved non-randomised with wait-list/control groups (one with post 3-month follow up and two with a 6-month follow up) [16,22]. In relation to the style of intervention, one involved skills-based training [22], five involved cognitivebehavior therapy (CBT) [18–21,23], and one involved expressive writing [16]. In their intervention, McCormick et al. [22] included adolescent IBD females and one of their parents to undertake a blended online/face-to-face intervention. Participants completed a group-based face-to-face 6-h training course and then weekly online chat sessions. In relation to the five CBT interventions involving IBS cohorts [18–21,23], pre-post analyses indicated that there were improvements in IBS symptoms and quality of life (QoL). One study reported reduction in pain [19], while three reported reduction in visceral sensitivity [19,21,23]. At follow up, all five studies reported improvements in QoL, three studies reported that positive changes were maintained in IBS symptoms [18,20,21]. Visceral sensitivity was also found to remain lower at post-follow up for three studies [18,20,21], with a further study indicating improvement being maintained [23]. In a cost–effectiveness evaluation using 12-month follow-up data from a 10-week CBT intervention with therapist email contact [19], Andersson et al. [14] concluded that 36% of the individuals who reported clinically reduced IBS symptoms (27 out of 76) also saved the community $16,806 per person. In a second cost–effectiveness study, Ljótsson et al. [23]
S. R. Knowles & A. Mikocka-Walus
Identification
390
Records identified through database searching (n = 1290)
Additional records identified through other sources (n = 2)
Included
Eligibility
Screening
Records after duplicates removed (n = 67)
Records screened (n = 1225)
Records excluded (n = 1140) based on the screening of title and abstract
Full-text articles assessed for eligibility (n = 85)
Full-text articles excluded, with reasons (n = 68)
Studies included in qualitative synthesis (n =17)
57 were not interventions 10 did not consider a GI population 1 used emails as the only eHealth intervention.
Studies included in quantitative synthesis (meta-analysis) (n = NA) Figure 1. PRISMA 2009 flow diagram. From [42].
conducted the only cost–benefit analysis showing that the treatment had an 87% chance of producing societal gains as well as health gains, saving around $39,821 in associated medical costs. Based on four consecutive 30-min expressivewriting intervention involving IBS patients, Halpert et al. [16] reported that there was an improvement in IBS symptoms post-intervention and that this continued at 3 months follow up. Using the blended face-to-face and online format with an IBD cohort, McCormick et al. [22] found that there were no significant differences between the treatment and wait-list groups at post-treatment with the exception of parent-reported approach subscale on the pain coping questionnaire (PCQ). Within-group analyses indicated that the adolescent treatment group reported significant reductions in somatic symptoms. Reports by parents indicated reduced children’s use of approach coping strategies and distraction techniques. In a follow-up study of McCormick et al. [22], ReedKnight et al. [24] evaluated the factors that influenced adolescent and parent participation and attrition. No significant differences were found in relation to medical or demographic factors and online participation
rates. Reduced adolescent participation was associated with increased adolescent- and parent-perceived abdominal pain, somatic symptoms, parent-reported catastrophizing and adolescent depressive symptoms. Parent participation was also found to have a significant positive association with adolescent participation. Owing to the small number of heterogeneous studies, a meta-analysis of the studies was not possible. However, review of the effect sizes from Ljótsson et al. IBS-based studies [19–21,23] indicated medium-tolarge effect sizes (e.g. pre-post IBS symptoms ranged from 0.38 to 1.21, QoL ranged from 0.51 to 91 and visceral sensitivity ranged from 0.33 to 0.79). Hunt et al. [18] found similar medium-to-large effect sizes (IBS symptoms = 1.21, QoL = 1.08), while McCormick et al. [22] found small-to-moderate effects sizes. When compared to Hunt et al., the smaller effect size found by McCormick et al. [22] may be due to several factors including differences in rates of change or possible change (e.g. pain symptoms may be less amenable to change vs. anxiety), associated with functional versus organic GI conditions (i.e. IBS vs. IBD), and intervention type and format (e.g. online vs. mixed-method). Mixed evidence was found in
eHealth interventions in gastroenterology
391
Table II. Ideal intervention eHealth gastroenterology requirements. Protocol characteristics: RCT design; makes sure that randomization is adequately conducted to avoid significant group differences; prior to intervention, power analyses are conducted (including correction for potential dropouts of at least 20%); comprehensive and clear details of intervention content are made fully available to allow other researchers to review and undertake accurate replication. When targeting psychological distress, the interventions should be GI-specific (i.e. if a patient has IBS and panic disorder, the intervention should be targeted to panic disorder using IBSoriented content). It includes measures that assess cost/benefit analysis and healthcare utilization; it clearly identifies intervention type (e.g. CBT) and format (e.g. modulized with email correspondence). Where possible, assessors and statisticians are blinded; in psychological trials, control for possible moderating/mediating factors is identified in the research literature, for example, stress, personality, social support, and self-efficacy. Base psychological interventions are made on validated theoretical approaches. Clear inclusion and exclusion criteria are identified. In psychological trials, it identifies (and where possible control for) if participants have, or have not, received, or currently receiving psychotherapy (including type and duration), have used psychotropic medication, and/or are on specialized diets. It includes long-term posttherapy efficacy review time points (i.e. 1- and 2-year post-intervention). Adequately selected control group, with active control groups are particularly encouraged Participant characteristics: Larger sample sizes are needed. It differentiates between non-active versus active condition status and condition type (e.g. IBS subtypes). It differentiates between disease subtypes (e.g. CD and UC) In psychological studies, it conducts psychological assessment to attain formal diagnosis of psychological disorders (e.g. generalized anxiety versus generalized anxiety with major depressive episode); pre-post intervention Data collection and analyses characteristics: It utilizes valid measures which can be accurately compared to other intervention studies. Where possible, it utilizes validated questionnaires and biomarkers and includes power analyses. Clear protocol for managing missing data is reported. It evaluates participant engagement in therapy (e.g. time accessing intervention materials and engaging in homework) and reports effect sizes and attrition rates. It evaluates results based on more statistically rigorous multilevel modeling techniques. This would allow for exploration of main effect changes as well as subgroups changes, while being able to control for confounding factors (e.g. stress, disease status, medication type and changes). It also evaluates differences between completers and non-completers Adapted from Knowles et al. [15].
regard to the improvement in mental health, with one study showing an improvement in depression [19] and two studies showing no significant change in anxiety and/or depression scores [20,23]. Regarding retention of participants in the study prepost intervention rates ranged from 46% to 99% and follow-up rates ranging from 36% (3-month follow up) to 85% (6-month follow up). Attrition rates were similar for those in the control groups. eHealth studies into disease management Overall, eight studies on eHealth in disease management have been identified (see Table V). Of these, three focused on patients with both CD and UC, three on UC only, one on CD only and one on celiac disease. Studies included 714 participants; six studies included middle-aged samples, while one study was focused on children and one on older participants. Four studies were RCTs (two with a 12-month follow up, one with a 9-month follow up, and one with a 3-month follow up), three studies were before-andafter trials (two with a 6-month follow up and one with a 5-month follow up), and one study was an openlabel trial with historical controls. Three studies were focused on monitoring disease activity, two on treatment adherence and side-effects, one on disease activity and adherence, one on overall management, and one on diet adherence. Studies by Elkjear et al. and Pendersen et al. [25–27] came from the same group as did the studies by Cross et al. [28,29], and
thus, of the seven included studies, there were only four different interventions (one from Elkjear et al., one from Cross et al., one from Hommel et al. and one from Krier et al.). In controlled studies [13,26–28,30], good experience was noted with eHealth applications and, in general, no group difference in QoL, medication adherence, hospitalization rates, surgery frequency or use of corticosteroids was noted. However, one open-label trial with historical controls observed significantly more adverse events in controls than in the intervention group (p = 0.002) [27]. In two RCTs [13,26], improved disease knowledge and compliance with treatment/diet were noted (p < 0.05). In withinparticipant trials [25,29,31], high acceptance and satisfaction with eHealth have been reported as have improved QoL, knowledge, and adherence and reduced costs, although the latter studies did not include control group comparisons. Regarding retention of participants in the study, at follow up, it ranged from 50% to 84% in intervention groups as compared to 74% to 92% in controls, thus indicating higher dropouts in participants receiving eHealth in comparison to standard care. Discussion The aim of this systematic review was to explore the potential of eHealth internet technology to augment traditional medical care in gastroenterology for the benefit of patients and the community and provide
392
S. R. Knowles & A. Mikocka-Walus
Table III. Study strength/weaknesses of the reviewed studies on eHealth in gastroenterology. eHealth studies on mental health Study
Strengths
Weaknesses
Hunt et al. [18]
First IBS online study, involving patient–therapist communication. Included wait-list control (WLC) group and a follow-up for those in the intervention group. Explored differences in participant characteristics of completers and non-completers.
McCormick et al. [22]
It combined face-to-face and online intervention; rewards were provided after each assignment completion/engagement in chat sessions; it utilized parent-adolescent dyads. Follow-up study [24] of the McCormick et al. [22] dataset explored participation and attrition rates.
Ljótsson et al. [19]
It conducted power analysis indicating that 50 participants would be valid and sufficient; however, authors aimed to recruit 80 patients. Assessed for diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patienttherapist communication, and patients were required to complete homework to continue to next module. A follow-up study [14] based on the Ljótsson et al. [19] dataset explored the cost–effectiveness of the intervention. It was conducted based on CONSORT statement for non-pharmacological trials, conducted power analyses, assessed for diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patient-therapist communication, and patients were required to complete homework to continue to next module. It used consecutive sample from a GI service, assessed healthcare costs and diet, and excluded those with other GI conditions and other physical or mental health conditions. It involved patient-therapist communication, and patients were required to complete homework to continue to next module. It explored 12-month follow up and reviewed completers versus non-completers.
No assessment (or possible controlling) of depression, stress, diet, illness perceptions was made, and excluded those with other GI conditions, but did not assess/ account for other physical or mental health conditions. It did not assess or control medication use. No poweranalysis was reported and no identification was made as to years with IBS; it had self-selected cohort. No assessment (or possible controlling) of depression, stress, diet, illness perceptions was made; there is no report of excluding other comorbid physical or mental health conditions. It had small sample size, nonrandomized allocation to conditions, therapy was conducted by combination of clinical psychologist or graduate clinical psychology students, and the study was limited to adolescent female sample. No power-analysis was reported. No assessment (or possible controlling) of medication use, stress, illness perceptions, and mindfulness was made. It had a self-selected cohort.
Ljótsson et al. [20]
Ljótsson et al. [23]
Ljótsson et al. [21] (12-month follow up of Ljótsson et al. [19]) Halpert et al. [16]
A useful pilot study, which explored 3-month follow up and reviewed completers versus non-completers.
eHealth studies into disease management Elkjaer et al. [26] Two international centers, used a preliminary study [25] to develop a well thought through intervention; two control groups were included. Elkjaer et al. [25] Pedersen et al. [27]
A useful companion study to Elkjaer et al. [26] Inclusion of a control group, and long follow up
Cross and Finkelstein [29] Cross et al. [28]
A useful pilot study informing a trial [28] RCT design and long follow up
No assessment (or possible controlling) of medication use, stress, illness perceptions, and mindfulness was made. It had a self-selected cohort.
It used VSI as a measure of coping and did not assess depression or anxiety, control for stress, and mindfulness.
There was no assessment (or possible controlling) of medication use, depression, anxiety, or stress, illness perceptions, and mindfulness. It had a self-selected cohort. There was no assessment (or possible controlling) of medication use, depression, anxiety or stress, no power analysis, and conducted an additional post-hoc exploratory subscale analyses to support hypotheses. It had a self-selected cohort. Power calculation was not presented and no comment to power of analyses was included; there was a small Irish sample which may explain some insignificant differences in this group. There were no control group and exclusion criteria. There were historical controls and no group comparisons on main outcome measures were conducted. It had a within-participant design and no control group Imbalance in baseline characteristics between groups; a significant proportion of intervention group participants discontinued intervention; sample size: the power calculation was for 84 participants but only
eHealth interventions in gastroenterology
393
Table III. (Continued). eHealth studies on mental health Study
Hommel et al. [31]
Krier et al. [30]
Sainsbury et al. [13]
Strengths
Weaknesses
A useful pilot study, first telehealth study on no adherence to treatment in IBD, prospective design, cost data An RCT design, long follow up, and excellent attrition rate An RCT design, adequate sample size with good calculation, and comprehensive methodology reporting
recommendations in relation to future internet-based eHealth gastroenterology interventions. Two trends in eHealth studies in gastroenterology were identified: (1) eHealth studies targeting GI symptoms and mental health and (2) eHealth studies in disease management. In relation to eHealth interventions targeting GI symptoms and mental health, seven studies were identified (six involving IBS cohorts and one involving an IBD cohort). Five of the seven eHealth interventions utilized an RCT design and included either 3-, 6-, or 12-month follow-up assessments, while the remaining studies utilized control groups with 3- and 6-month follow ups, respectively. Despite the heterogeneity of the IBS-focused studies (e.g. differences in intervention type, assessment methods, and time-points), the findings indicate that significant benefits were reported by participants with medium-to-large effect sizes in QoL and IBS symptoms. Although found to be beneficial, McCormick et al. [22] reported small-to-moderate effect sizesacrossmultiple adolescent–parent dyad measures (e.g. adolescent somatic symptoms, parent irrational cognitions, and protective responses to pain/IBD symptoms). These findings are consistent with face-toface/group-based interventions targeting GI symptoms and mental health (see reviews: Zijdenbos et al. [32], Timmer et al. [33], and Knowles et al. [15]), in that it can be argued that individuals who experience mental distress along with their physical health condition are likely to benefit from a focused psychological intervention. It is, however, too early to identify if targeted eHealth psychological interventions may be more beneficial for different illness cohorts (i.e. functional vs. organic GI conditions). In relation to eHealth interventions targeting disease management, eight studies were identified (seven involving IBD cohorts and one involving a celiac disease cohort). Four of the eight eHealth interventions utilized an RCT design and included either 3-, 9-, or 12-month follow-up assessments, while the remaining studies utilized either a before-and-after
41 completed intervention and thus study was underpowered It had no control group, a small sample, and did only basic cost analysis. No power calculation provided and thus unknown whether the study is underpowered, and had a small sample size Nearly 50% of the intervention group participants did not complete the intervention; short follow up
trials (with 5- or 6-month follow up) or an open-label trial compared with historical controls. The interventions targeting disease management to date suggest that eHealth may be a preferred way for patients to help monitor their illness and also enhance compliance with the recommended treatment plan. While still too early to argue its definitive benefits, the current positive findings in the eHealth studies reflect the published literature and indicate that interventions that enhance patient self-efficacy and that are developed to facilitate patient engagement are likely to be useful in managing chronic GI conditions [34]. Recommendations Based on the reviewed papers, we suggest that gastroenterology is in a strong position to utilize eHealth internet technology to provide augmented services for promoting patient general well-being. While participants generally identify a high acceptance of eHealth interventions, the retention rates may actually be poorer than traditional face-to-face interventions. Given this observation, while eHealth interventions seem to provide benefit, further research is needed to better identify when this (and what) form of treatment is more appropriate, what interpersonal/social factors mediate engagement (e.g. self-efficacy and parental involvement), and what participant characteristics help to determine this. For example, is an eHealth intervention efficacious when the patient cohort experiences significant distress or in an active phase of their GI condition? Given the diversity of treatment interventions and their format, elucidating these important factors and possible mediating processes will be important. Further, more research is needed to evaluate acceptability for e-health interventions and if, and how, differences in patient needs can be addressed in eHealth interventions. For example, traditional faceto-face and group-based interventions, while usually
IBS
IBD
Hunt et al. [18]
McCormick et al. [22]
Authors
Condition RCT; pre-post assessment, post 3-month assessment (intervention only)
NRTC; pre-post and follow up (6 months). Participant medical records (n = 215) screened for eligibility. Inclusion/exclusion criteria were not identified, although later reported in Reed-Knight et al. [24]
N = 31 female adolescent-parent dyads (n = 20 treatment, n = 11 wait-list). Regarding adolescents: mean age = 14.46 years; mean years diagnosed = 3.15
Design and assessments
N = 54 (44 females; n = 28 intervention, n = 26 wait-list; mean age = 38 years) were invited to participate via IBS relevant websites
Participant details
Participants completed a 6-h training skills course on managing pain, coping with IBD, communicating with others, and for parents only setting limits skills. Following this day, participants completed 6 weekly web-based activities (separate activities for adolescents and parents). Separate
5 modules (1 per week): Module 1: Psychoeducation and relaxation training; Module 2: cognitiveoriented stress management (SM); Module 3: challenging catastrophic thinking; Module 4: introduction to exposure therapy; Module 5: exposurebased experiments. Homework emailed to primary investigator who then provided individualized responses within 48 h. Inclusion criteria: self-report diagnosis of IBS by medical professional with no other GI condition
Intervention
Table IV. Summary of eHealth mental health interventions for gastroenterological conditions.
Pre-post: GSRS-IBS = 1.21; IBS-QoL = 1.08; ASI-GI = 0.63; ASI non-GI = 0.70 CPSQ GI = 1.83; SPSQ-non-GI = 1.17
Effect sizes
Main findings
Significant improvement across all measures in the intervention group, with QoL and VAS at post3-month follow up. Nonsignificant trend suggesting that GI symptoms and catastrophizing increased at post-3-month follow up; based on Jacobson and Truax [40] method for identifying clinical significance; intervention was found to be superior to control-group (GSRS unchanged intervention 62% vs. 83% control; IBS-QoL unchanged intervention 54%; 83% control) Attrition: Of the initial 28 participants in the intervention group, 13 (46%) completed the intervention and postintervention questionnaire, 10 (36%) completed post3-month assessment. For the control group, of the 26, 18 (69%) completed the control phase, 8 (31%) completed the intervention, and 2 (8%) completed the post-3-month follow up Adolescent primary Pre-post: No significant differences outcome measures: parent-reported approach were found between the Abdominal Pain Index subscale on the PCQ = 0.28 treatment and wait-list (API), Child Within-group pre-post groups at post-treatment Somatization Inventory analyses: with the exception of (CSI), PCQ, Pain Treatment group – parent-reported approach Catastrophizing Scale API = 0.05 (adolescent); subscale on the PCQ. (PCS) API = 0.03 (parent); Within-group analyses Parent primary outcome CSI = 0.41 (adolescent); indicated for the treatment measures: API = 0.38 (parent); group adolescents reported Abdominal Pain Index within-group pre-follow up: reductions in somatic (API), Child Treatment group – symptoms. ParentsSomatization Inventory API = 0.24 (adolescent); reported reductions were
GI Symptom Rating Scale-IBS (GSRS-IBS), Anxiety Sensitivity Index (ASI), Visceral Anxiety Sensitivity (VAS), IBS-QoL, and the Consequences of Physical Sensations Questionnaire (CPSQ)
Outcome measures
394 S. R. Knowles & A. Mikocka-Walus
Ljótsson et al. IBS [19]
Authors
Condition
Table IV. (Continued). Design and assessments
RCT; patients were screened via-telephone to rule out other GI conditions, food intolerances and completed the Miniinternational Neuropsychiatric Interview (MINI). Responses from interview were assessed by gastroenterologist to confirm eligibility for inclusion. The study excluded patients with severe depression, substance dependence, psychosis, mania, and anorexia
Participant details
N = 85 (72 females; n = 42 intervention, n = 43 wait-list; mean age = 34.6 years; mean years diagnosed = 6.3) self-referred IBSpatients from Sweden were recruited
Primary outcome measures: weekly GI symptom diary, weekly GSRS-IBS. Secondary outcome measures: Montgomery Asberg Depression Rating Scale (MADRS-S), IBS-QoL, VSI, Sheehan Disability Scales (SDS), and Treatment Credibility Scale(TCS)
(CSI), PCQ, Pain Catastrophizing Scale (PCS), Adult response to Children’s Symptoms: Protect Scale (ARCS)
adolescent/parent online 30-min chat sessions occurred once a week with a reminder call before the chat session. Participants received $20 for completing each weekly homework assignment and $25 at the end of the treatment program with a minimum completion of 5 out of 6 homework and chat sessions
10-week 5-module CBT-protocol based on a previously published group intervention ([41]). Module 1: rationale for treatment and mindfulness; Modules 2, 3, and 4 identified psychological model of IBS and the associated cognitive and behavioral factors that exacerbate or attenuate IBS symptoms and the focus on them; Module 5: focus on exposurebased activities. Completion of homework was the
Outcome measures
Intervention
Pre-post: GSRS-IBS = 1.21; IBS-QoL = 0.93; VSI = 0.64; MADRS-S = 0.43; SDS = 0.47
API = 0.08 (parent); CSI = 0.06 (adolescent); API = 0.35 (parent); within-group pre-post analyses: Control group – API = 0.07 (adolescent); API = 0.12 (parent); CSI = 0.12 (adolescent); API =
