An estimated 6 million children are admitted to hospitals annually in the United States, often under life-threatening circumstances [1]. In addition to the physical impact of pediatric injuries and illnesses, consequences of pediatric medical conditions can often include negative psychological sequelae such as posttraumatic stress symptoms (PTSS), anxiety, depression, and impaired health-related quality of life (HRQoL) for both patients and their caregivers [2-5]. Web-based or electronic health (eHealth) interventions have the potential to mitigate these effects and to extend the reach of in-person interventions [6,7].

Psychological outcomes for acute and chronic conditions are similar [8]. For both acute (eg, burns and traumatic brain injury [TBI]) and chronic medical conditions (eg, atopic dermatitis, chronic pain, cancer, and diabetes), youth and caregivers are at risk for developing significant psychological symptoms and impairment in HRQoL, which have been associated with negative health outcomes. Approximately 30% of children with an injury or illness develop significant PTSS [8]. PTSS post injury is linked to depression, poor health outcomes, and impaired HRQoL [9]. PTSS have been associated with impaired HRQoL and subsequent health problems in youth [10]. In addition, increased anxiety before surgery has been associated with poorer health outcomes and worse postoperative recovery [11]. Similarly, children with chronic conditions, such as cancer, may experience distressing emotional reactions and lower quality of life (QoL), sometimes for years after the completion of treatment [12-14]. Thus, developing effective interventions to address these symptoms and challenges is essential for promoting full recovery (ie, physical and emotional) in youth with medical conditions.

A number of interventions currently exist to promote emotional recovery and adaptation in youth experiencing medical conditions, including education-based interventions, behavioral therapy, cognitive behavioral therapy (CBT), problem-solving therapy (PST), family therapy, multisystemic therapy, and systemic treatment [15-17]. However, limited access to psychologists, high costs of therapy, and difficulty accessing resources because of location (eg, rural areas) are all factors that can play a role in preventing many children from obtaining these treatments [17,18]. eHealth interventions may provide an avenue to distribute evidence-based strategies and treatments to children and families that may not otherwise have access to emotional health resources. Efforts are underway to translate current evidence-based in-person interventions to eHealth platforms.

Multiple reviews have been conducted to examine the use of eHealth interventions to improve health outcomes in children, suggesting promising results for this technology [19-22]. In examining emotional health outcomes, a recent review by Canter et al [23] summarized current evidence for technology-delivered interventions in improving family-centered outcomes (eg, communication, problem solving, and caregiver-child relationship) for children with chronic medical conditions.

Results suggested that although eHealth interventions are generally effective at reducing family conflict, the findings for other family-centered outcomes varied [23]. This review expands upon the findings of the study by Canter et al [23], examining specific individual emotional health outcomes of children and caregivers. Other reviews have examined the impact of eHealth interventions on physical and dietary changes, behavior change, and various health conditions such as asthma but have generally focused on a single medical condition and often only assessed QoL outcomes [19-22]. To our knowledge, no other reviews to date have examined the use of eHealth interventions on psychological outcomes across both acute and chronic pediatric medical conditions.

eHealth interventions can be cost-effective and easily accessible, but their initial development and maintenance can be costly. Without evaluation, it is unknown whether evidence-based, effective interventions are able to maintain their effect when they are translated or adapted for eHealth platforms. Many eHealth interventions are in the early stages of development and have shown promising early results on their efficacy. We conducted this systematic review with the goals of (1) summarizing the state of the field for eHealth interventions that are designed to prevent or treat negative psychological sequelae (ie, anxiety, depression, PTSS, and HRQoL) resulting from pediatric medical conditions and (2) providing recommendations toward future development of eHealth interventions.

The following databases were searched for all studies conducted between January 1, 1998, and March 1, 2019: MEDLINE (PubMed) and PsycINFO. The search terms used are as follows: “medical trauma,” “medical event,” “medical condition,” “medical procedures,” “illness,” “injury,” “web-based,” “mobile,” “e-Health,” “internet,” “telehealth,” mHealth,” “text,” “application,” “posttraumatic stress symptoms,” “PTSD,” “PTSS,” “posttraumatic stress disorder,” “quality of life,” “anxiety,” “depression”, “psychological sequelae,” ” “intervention,” and “prevention.” Reference lists were reviewed for all studies meeting eligibility criteria. If the title in a reference list suggested that it could be relevant for this review, the full manuscript was reviewed.

Eligible studies (1) employed empirical methods examining the efficacy of an eHealth intervention to prevent or treat negative psychological sequelae in youth (aged 0-18 years), their caregivers, or both youth and caregivers as either a primary or secondary outcome, (2) included youth or caregivers of youth with acute or chronic medical conditions, and (3) were available in English language. Studies were excluded if they failed to meet the inclusion criteria listed above. For the purposes of this review, psychological sequelae are defined as PTSS, anxiety, depression, and QoL. Studies that did not include these outcomes as a primary or secondary study outcome for either the child or caregiver were excluded.

The initial search of PubMed and PsycINFO yielded 1606 results. A total of 2 authors independently reviewed titles and abstracts of relevant studies focusing on the efficacy of eHealth psychological interventions for preventing or treating psychological sequelae (ie, PTSS, anxiety, depression, and HRQoL) in youth with medical conditions. After removing 105 duplicate studies and excluding 1473 studies based on their title and abstract, the full text of the remaining 39 articles were reviewed for eligibility. A total of 11 additional studies were selected for review from reference sections of the articles in the original search. In total, 22 studies met full inclusion criteria and were included in the review. See Figure 1 for a Preferred Reporting Items for Systematic Reviews and Meta-Analyses diagram depicting the study design of article selection. Each article that met inclusion criteria was reviewed and coded for theoretical framework, intervention delivery method, intervention outcomes, and intervention barriers.

The Downs and Black checklist was used to assess study quality [24]. The checklist comprises 5 sections (27 items total) measuring reporting, external validity, internal validity, selection bias, and power. Studies were scored and placed into 1 of the 4 categories: poor (>14 points), fair (15-19 points), good (20-25 points), and excellent (26-27 points).

The majority of studies scored within the fair (50%, 11/22) and good (41%, 9/22) categories [25-45]. Only 2 studies [46,47] were rated as poor (9%, 2/22). No studies were rated as excellent. For the 2 studies rated as poor, a score of 0 was given for many items in the internal validity subscale (eg, either did not address or failed to blind participants and researchers, did not address participants lost to follow-up, and did not adjust for confounding in the analyses). Approximately 64% (14/22) of the included studies described characteristics of the participants that were lost to follow-up; 68% (15/22) of the studies accounted for this loss in analyzing study outcomes.

Of the 22 studies included in this review, 8 assessed for depression [27,33-36,40,41,43], 5 for anxiety [27,33,40,41,44], 4 for PTSS [27,31,39,45], and 10 for HRQoL [25,26,29,30,32, 37,38,44,46,47]. A total of 5 studies [27,33,40,44,45] examined more than 1 outcome. All included studies were randomized controlled trials (RCTs) per inclusion criteria. Some of the studies (n=12) [25-27,30,31,34,37-39,44-47] examined interventions that were solely delivered via Web, whereas 10 studies [29,32,33,35,36,38,40-43] used a combination of eHealth program components and sessions with a therapist or coach. See Multimedia Appendix 1 for additional study details.

Sample sizes for the studies ranged from 37 to 164. Child participants were aged 2 to 18 years. A total of 7 interventions were designed for children alone [25,26,31,32,34,37,46], 3 for caregivers alone [36,39,45], and 11 for children and caregivers to use together [27,29,30,33,35,38,40-42,44,47]. Moreover, 10 studies focused on chronic conditions such as asthma [30,37,47], chronic headache [33], juvenile idiopathic arthritis (JIA) [25,38], inflammatory bowel disease [26], complex congenital heart disease [32], chronic respiratory condition [34], and type-1 diabetes [46]. In addition, 6 studies focused on pediatric patients suffering from psychological sequelae following a TBI [35,36,40-43]. A total of 3 studies focused on the aftermath of injuries (not focused on TBI) [27,39,45], and 1 study included children with potentially traumatic medical events (eg, injury, new diagnosis, and exacerbation of chronic condition) [31]. Furthermore, 2 studies had additional relevant criteria: one study required that participants perceived their injury as potentially traumatic [31], and another study enrolled participants who medical staff perceived as socially isolated or disadvantaged [34].

Research varied in whether outcomes were assessed for children only (n=7 studies) [25,26,31,32,34,37,46], caregivers only (n=3 studies) [36,39,45], or both children and caregivers (n=12 studies) [27,29,30,33,35,38,40-44,47]. Of the included interventions, 13 [27,30,31,33,34,37,39-44,47] identified significant intervention effects on at least one targeted psychological outcome.

This review identified 22 studies of eHealth interventions designed to prevent or alleviate negative psychological sequelae (ie, anxiety, depression, PTSS, and HRQoL) in youth with a medical condition and their caregivers. This is this first review to examine multiple psychological sequelae components across acute and chronic medical conditions. By better understanding existing interventions across populations, we can improve our ability to design evidence-based, tailored interventions to improve targeted outcomes.

Overall results from this review indicate that research on eHealth interventions is in early stages and that results are mixed but promising. Results from this systematic review highlight some evidence to suggest that using eHealth interventions may help improve child and caregiver psychological outcomes (including anxiety and PTSS) and functional outcomes (ie, HRQoL) [27,30,31,33,37,40-44,47] but that more research is necessary to examine essential intervention targets, variable needs of different medical populations, and barriers to intervention implementation. In addition, there is room for improvement on designing research studies to thoroughly evaluate the selected outcomes and to explicitly report study methods, as only 41% of studies included were coded within the good range of study quality. As many of the studies included in this review were early stage research of new interventions, limitations to study quality are reasonable, as it makes sense to first examine intervention potential before investing in larger, complex research. However, to improve the scientific rigor of the methodology of this research as it continues to grow, future researchers may want to consider carefully examining and reporting on intervention compliance, participant dropout, and the effect of cofounders on outcomes. In addition, future studies may want to aim for larger sample sizes to improve power, the inclusion of larger proportions of the targeted population in the study, and the implementation of double blinding.

As there were very few studies that examined the effectiveness of eHealth studies on child and caregiver emotional health and QoL and as there was substantial variability in intervention components and modality, populations in which interventions are delivered, intervention modality, and outcomes assessed, the following summaries of key findings should be interpreted as preliminary. Results from this systematic review suggest that the intervention theory that guided intervention content might have influenced the outcomes: of the studies included in this systematic review, those that primarily utilized CBT (80%) had the highest proportion of achieving at least one intended outcome [31,33,39]. This was followed by problem solving (71%) [34,40-43], education alone (60%) [30,37,47], psychoeducation (50%) [27-44], and PCIT (0%) [36]. These findings are consistent with current research supporting CBT [50-52].

In examining more nuanced study results, findings suggest that the type of intervention needed may be dependent upon the situation and goal of the intervention (ie, prevention vs treatment of symptoms). For example, information provision or educational interventions appear to be useful for decreasing anxiety and improving QoL, such as preoperative education, or when interventions are delivered within the hours after experiencing trauma (eg, injury) to prepare families for the emotional and physical challenges that lie ahead [27,44]. However, education-based interventions may not be helpful in achieving targeted outcomes for chronic illnesses (eg, JIA, inflammatory bowel disease, and congenital heart disease) [26,32,38] or if the education is provided days after a trauma [39,45]. Thus, intervention modality may need to be selected based on the challenges presented and targeted outcomes.

Results also suggest that intervention modality was fairly comparable regardless of whether the intervention was fully Web-based or not (64% [14/22] indicated significant positive outcomes), compared with Web-based plus in-person contact (44%). This is promising in that fully Web-based interventions may be less costly and have wider reach.

There are multiple factors to consider that have the potential to interfere with Web-based intervention engagement and effectiveness. Our results indicate that demographic factors such as SES, level of caregiver education, and social advantage may impact intervention efficacy [36,40]. How SES affects technology-based intervention uptake remains mixed. Although low SES is a potential barrier to intervention efficacy, some studies reported that families of lower SES might equally benefit from eHealth interventions compared with families of higher SES, when provided with computers and internet access [35,36,40,42,43]. More research is needed to better understand how intervention can be tailored to be the most efficacious across family SES.

Another potential barrier to consider is knowledge of technology. Although not included in this review because primary results are presented in another publication [40], a study by Carey et al [53] assessed past experience with technology in participants completing the intervention and found that participants who used technology less frequently showed less improvement in anxiety and depression symptoms, compared with those who used technology more often. The wide range of technology experience in target pediatric populations and their parents can been explained by the digital divide, defined as the gap in the frequency of information technology use and what it is used for, which is thought to be moderated by demographic factors (eg, income, education level, gender, and race) [53]. Jackson et al [54] surveyed 515 children and found that children of parents with a full-time job used cell phones more frequently than children of parents with other employment situations. In addition, children of parents with higher education reported more use of computers and internet [54]. Female participants used cell phones more often than males, whereas males most intensively used video games [54]. Furthermore, African American males were the least frequent users of computers and internet, whereas African American females were the heaviest users of internet out of all groups [54]. For these reasons, care should be taken when designing interventions to provide the target sample with a feasible intervention platform (eg, adding more or less directions, using pictures, and content delivery method). Future studies should take into account demographical differences when formulating intervention content and deciding which delivery method to use (ie, mobile phone app, internet, or video game).

Barriers to intervention compliance should also be considered. Although not included in the final review, Worthen-Chaudhari et al [55] reported on barriers to study compliance and found that participants who dropped out of the study had discontinued medical care, faced problems with internet access, busy schedules, and experienced co-occurring illness during the study. Although these factors were not assessed after completion of the intervention, it is important to consider these as potential barriers to intervention usage that could affect study results.

Future research should directly examine potential intervention barriers such as SES, education level, location, knowledge of technology, and severity of medical condition in youth with various medical conditions. In addition, although existing literature does not indicate how the timing of the intervention affects its efficacy, this may be an important factor to explore in future research. Studies should also aim to clearly identify their treatment outcomes and create interventions designed specifically to improve those outcomes. Finally, additional studies should examine the impact of improvements in parent outcomes on child outcomes.

There are several notable limitations to this study that should be considered in interpreting and generalizing study results. Many studies reviewed had small sample sizes, with limited power as a result. In addition, research studies included were inconsistent in the measures used and the outcomes assessed, and many studies were not RCTs. Moreover, this study was unable to review research that was not published in English. Finally, the majority of the studies published identified an effect for at least one outcome; it is unknown whether these studies represent the majority of studies conducted in this area or whether there are a number of unpublished nonsignificant findings to take into account.

eHealth interventions have the capacity to broaden our reach to improve emotional health in families with children undergoing medical treatment. Although results are mixed, the results of this study suggest that eHealth interventions may be useful for improving psychological sequelae in pediatric populations with medical conditions such as TBI and other potentially traumatic injuries. More research is needed to identify the most important intervention components and how to ensure that these components are maintained in the translation to eHealth modalities.