FormalPara Key Summary Points

This meta-analysis delves into the complexities and considerations that must be considered when integrating information and communication technologies (ICT) into clinical practice.

Mobile health applications and telemonitoring show significant improvement in patients’ quality of life (Dermatology Life Quality Index, DLQI) and self-management (Patient Oriented Eczema Measure, POEM).

Our results promote a more patient-centric and technologically integrated approach to care, with an emphasis on improving patient outcomes and managing atopic dermatitis effectively and efficiently while enhancing patient satisfaction and engagement.

There are substantial benefits of these technologies in improving quality of life, fostering self-management, and potentially optimizing healthcare resources.

Introduction

Atopic dermatitis (AD) is recognized as the most prevalent chronic inflammatory skin disorder, commanding the highest rank among all skin conditions as per disability-adjusted life-years metrics [1]. Recent epidemiological data reveal that AD afflicts approximately 25% of children and 5.4−5.6% of the adult population in the USA [2]. The clinical manifestations of AD extend beyond cutaneous symptoms. AD significantly impairs quality of life (QoL), and implies social and economic challenges for both patients and their families [3, 4].

The existing multifaceted treatment paradigm, though aligned with current guidelines, has met with limited success in managing AD. This is evidenced by a control rate of 75% after 16 weeks of treatment with dupilumab [5], and new molecules such as Janus kinase (JAK) inhibitors, which block multiple proinflammatory cytokines involved in AD, have emerged as robust therapeutic options. With these advancements, the effectiveness of success rates has increased to 52% in long-term studies (52 weeks) of topical JAK inhibitors [6] and up to 62% efficacy in the oral form (12 weeks) [7], however the safety profile is still of concern in the literature [8]. Patients typically experience nine flare-ups annually, with a total of 136 days per year spent in AD flares [9]. These figures highlight the importance of achieving effective control, a critical factor for enhancing QoL, as elucidated in the International Study of Life with Atopic Eczema by patients with moderate to severe AD and their caregivers [10, 11].

A key aspect of AD management involves symptom alleviation, exacerbation prevention, and risk minimization through patient education, self-monitoring, and routine symptom assessment using patient-reported outcomes (PROs) [12, 13]. The emergence of information and communication technology (ICT) offers novel avenues for improving chronic skin disease management, including AD. ICT, encompassing electronic health (eHealth) technology such as telemonitoring, and mobile health (mHealth) intelligent devices (applications, smartphones, emails, chatbots, SMS, wearable devices, etc.), has the potential to facilitate self-management and treatment adherence and enhance communication between patients and healthcare providers [14, 15].

This systematic review aims to evaluate the effectiveness of mHealth (mobile health) apps, telemonitoring, and smart devices for managing dermatological conditions, with a focus on AD but inclusive of related conditions. By synthesizing existing evidence and providing a rigorous analysis, this research aims to contribute valuable insights to the ongoing discourse in this field. Such insights have the potential to inform both future scientific inquiry and practical clinical applications, offering a data-driven foundation for interventions aimed at enhancing the lives of those affected by AD and related conditions.

Methods

Search Strategy

We searched PubMed, Web of Science, Scopus, and Embase until May 2023 for peer-reviewed clinical trials evaluating the effectiveness of teledermatology, mobile apps, and electronic devices for managing atopic dermatitis. No language or date restrictions were applied during the searching process. The review adhered to the PRISMA 2020 guidelines [16]. The study’s protocol was registered in PROSPERO (PROSPERO 2023: CRD42023421127).

Eligibility Criteria

Inclusion and Exclusion Criteria

Inclusion: Reports on teledermatology, mobile apps, and electronic devices for AD for all age groups and settings and providing an assessment of the performance of the electronic tools on disease control were included. Studies were considered if they were in the English language and comprised randomized controlled trials (RCTs), nonrandomized controlled trials, and observational studies.

Exclusion: Studies were excluded if they were not AD-related, not related to mobile apps, not telemedicine, not electronic devices, or were reviews, nonhuman studies, guidelines.

PICO Framework:

P: Pediatric and adult patients with dermatological conditions including but not limited to atopic dermatitis.

I: mHealth Apps, Telemonitoring, and Smart Devices.

C: Standard care practices.

O: Enhanced Self-Management Outcomes for Dermatological Conditions (Improved Control; Patient-Reported Outcomes; Itch Intensity Reduction; Decreased Flare Frequency; Reduced Topical Medication Use; Enhanced Skin Barrier Function).

Data Extraction and Synthesis

The selection process for the included articles involved three independent reviewers (M.O., K.R., and A.O.) who proceeded in two phases. In the first phase, they assessed the titles and abstracts of the articles on the basis of predetermined eligibility criteria. In the second phase, they reviewed the full-text articles using the same criteria as in phase 1 and cross-checked all information obtained. Any discrepancies were resolved through discussion between the reviewers. If any essential data were missing or unclear, the corresponding authors of the study were contacted to provide clarification.

Data Synthesis

The meta-analysis was conducted for controlled trials in Review Manager 5.4.3 (Cochrane, USA). The data were analyzed and are presented as mean difference with 95% confidence intervals. In addition, the effect size was computed as standardized mean difference with 95% CI and are reported as Cohen’s d. Additionally, the χ2, I2, and Z values, in addition to the P value, are also presented as forest plots for the meta-analytical outcomes.

Risk of Bias Assessment

The risk of bias was evaluated at the study level based on characteristics such as study design, sample size, participants, exposure, and outcome measures. The assessment also included a risk of bias evaluation using the ROB2 by Cochrane.

Ethical Approval

This article is based on previously conducted studies and does not contain any new studies with human participants or animals performed by any of the authors.

Results

Of the 811 studies identified, 252 duplicates were removed before screening. In total, 559 studies were screened using their titles and abstracts. Of these, 527 were excluded owing to the lack of relevance and meeting the exclusion criteria. Thirty-two studies were sought for retrieval and were assessed for full-text eligibility. Finally, 12 studies were included in the systematic review, and 8 were included in the meta-analysis. The PRISMA flowchart is depicted in Fig. 1.

Fig. 1
figure 1

PRISMA flowchart depicting the study selection process

Full size image

Trial Characteristics

Of 811 studies, we included 12 trials, comprising 2424 pediatric/adult participants followed for 1–12 months. These individuals were intervened with mHealth applications/telemonitoring or standard care practices. The pooled median age for adults was 33.6 years with an interquartile range of 11.43 years, consolidating data from general adults, mothers, parents or carers, and other adult groups. For children, the overall median age was 3.75 years with an interquartile range (IQR) of 2.06 years. Additionally, Santer and colleagues’ (2014) age distributions indicate that, among carers, 32% are aged between 31–35 and 36–40 years each, while children’s age percentages peaked at 26% for 3-year-olds. The 12 studies included in the systematic review cover various interventions and methodologies aimed at understanding and treating various forms of eczema and other dermatological conditions [17,18,19,20,21,22,23,24,25,26,27,28]. The characteristics are listed in Supplementary file (https://doi.org/10.17632/hyr49bdmxd.1).

Technological Functions, Study Objectives, and Outcome Measures

In the realm of digital healthcare for dermatological conditions, particularly eczema, several studies have utilized various technologies, durations, functions, and engagement methods to attain specific objectives and measure certain outcomes. The study-by-study information is presented in Supplementary file (https://doi.org/10.17632/hyr49bdmxd.1).

Patient Characteristics

The patient characteristics are listed in the Supplementary file (https://doi.org/10.17632/hyr49bdmxd.1).

Meta-analytical Findings

For the POEM assessment, the meta-analysis of 1038 participants demonstrated an MD of −1.57, with a 95% confidence interval (CI) of [−2.24, −0.91]. Heterogeneity was observed (χ2 = 9.35, degrees of freedom [df] = 5; I2 = 47%), and the overall effect was found to be significant (Z = 4.61, P < 0.00001). Cohen’s d value was −0.29, with a 95% CI of [−0.42, −0.17]. Heterogeneity for Cohen’s d was calculated as χ2 = 7.65, df = 5 (P = 0.18), I2 = 35%, and the test for overall effect resulted in Z = 4.68 (P < 0.00001) (Fig. 2).

Fig. 2
figure 2

Forest plots for POEM outcomes in interventional and standard care groups. POEM The Patient Oriented Eczema Measure

Full size image

Regarding the DLQI analysis, the meta-analysis included 495 participants, yielding an MD of −0.59 (95% CI: [−0.95, −0.23]). Heterogeneity was high (χ2 = 15.51, df = 4; I2 = 74%), and the overall effect was significant (Z = 3.24, P = 0.001). Cohen’s d value was −0.15, with a 95% CI of [−0.33, 0.03]. Heterogeneity was measured as χ2 = 24.12, df = 4 (P < 0.0001), I2 = 83%, and the test for overall effect revealed Z = 1.61 (P = 0.11) (Fig. 3).

Fig. 3
figure 3

Forest plots for DLQI outcomes in interventional and standard care groups. DLQI Dermatology Life Quality Index

Full size image

The SCORAD analysis demonstrated a negligible MD of −0.12, with a 95% CI of [−2.08, 1.84] (P = 0.91). Cohen’s d was calculated as −0.03, with a 95% CI of [−0.30, 0.25]. Heterogeneity was minimal, with χ2 = 1.47, df = 2 (P = 0.48), I2 = 0%, and the test for overall effect resulted in Z = 0.12 (P = 0.91) for MD and Z = 0.19 (P = 0.85) for Cohen’s d (Fig. 4). These findings indicate varying degrees of effects across the different measures and highlight the complex relationships within the data. The datasheet utilized for the meta-analysis is attached in the Supplementary file (https://doi.org/10.17632/hyr49bdmxd.1).

Fig. 4
figure 4

Forest plots for SCORAD outcomes in interventional and standard care groups. SCORAD SCORing Atopic Dermatitis

Full size image

Risk of Bias Synthesis

The studies by Weigandt (2023), Armstrong (2015), Santer (2014), and Os-Medendorp (2012) demonstrate a low risk of bias across all evaluated domains. This category accounts for 33.3% (4 out of 12) of the total studies, reflecting a strong methodological foundation. Studies by Ando (2022), Rijsbergen (2020), Kornmehl (2017), Annette Mollerup (2016), Bergmo (2008), Joergensen (2019), and Miriam Santer (2022) exhibited some concerns in one or more domains. These studies constitute 58.3% (7 out of 12) of the total and imply potential issues that might slightly affect the confidence in the results. Only the study by Gudmundsdottir (2022) was rated with a high risk of bias, reflecting significant concerns in some areas that might substantially affect the interpretation of the findings. This represents 8.3% (1 out of 12) of the included studies. The summary and traffic light plots are depicted in Fig. 5.

Fig. 5
figure 5

Summary plot and traffic plots for risk of bias assessments of the included studies [22,23,24,25,26,27,28,29,30,31,32,33]

Full size image

Discussion

In this systematic review and meta-analysis, the effectiveness of telemonitoring and mobile applications was examined specifically for patients with AD. The data indicated a promising improvement in quality of life and self-management. The meta-analytical findings of this study encompass three distinct analyses that provide insight into the management of AD. In the POEM assessment involving 1038 participants, there was a significant overall improvement with an MD of −1.57 (P < 0.00001), with a moderate heterogeneity (I2 = 47%); Cohen’s d value of −0.29 further emphasizes the observed benefits. In the DLQI analysis of 495 participants, the study yielded an MD of −0.59 (P = 0.001) with high heterogeneity (I2 = 74%). Although a significant overall effect was found, the high variability necessitates careful interpretation. The SCORAD analysis, on the other hand, showed a negligible effect with an MD of −0.12 (P = 0.91) and minimal heterogeneity (I2 = 0%), underscoring the complex relationships within the data. Collectively, these findings reveal varying degrees of effects across different measures. While significant improvements are evident in POEM and DLQI outcomes, pointing to effective interventions in some aspects of AD management, the negligible effect in SCORAD analysis calls for further exploration. The distinct levels of heterogeneity across the analyses also indicate the necessity for nuanced interpretation considering individual study contexts and patient populations.

The resurgence of interest in teledermatology has been further accelerated by the coronavirus disease 2019 (COVID-19) pandemic, with forecasts suggesting continued growth due to evident patient and provider satisfaction [29]. In particular, telemonitoring emerged as an invaluable tool in enhancing access to specialized dermatology care, even for those in remote locations, a factor that simultaneously contributes to healthcare cost reduction [30]. As demonstrated by Giavina-Bianchi et al., telemedicine’s capacity to manage 72% of mild AD cases without in-person visits further underlines its potential to conserve resources, improve accessibility to specialized care, and significantly decrease wait times, a consideration especially pertinent in the context of public healthcare [31]. In a cohort study of 195 patients, Sieniawska and colleagues conducted online consultations and ascertained that nearly 50% of patients were discontented with its use, whereas one-third of them did not mind the use of telemedicine [32]. Overall, the authors reported lower life satisfaction, impaired QoL pertaining to mental health and overall health rating with AD during the COVID-19 pandemic [32]. The study emphasizes the usability of telemedicine applications during the pandemic and emphasizes the need for such interventions to improve anxiety and depression among patient groups [32].

This systematic review and meta-analysis highlights a broader acceptance and appreciation for digital technologies used to support healthcare, namely eHealth (electronic health) services, among both medical professionals and consumers [33]. With the integration of features such as medication reminders, symptom and trigger education, symptom tracking, patient education, and PROs, mobile applications have not only witnessed significant user engagement but also proven to be effective in enhancing symptom management and adherence to treatment among AD patients [19, 34]. These technological innovations have thus yielded tangible benefits, evident in improved clinical outcomes, as reflected in assessments such as POEM and DLQI.

Strengths

The strengths of our study include a comprehensive review utilizing a systematic approach that ensures an extensive overview of the existing literature related to the use of telemonitoring and mobile applications in AD management. By including studies with diverse methodologies, populations, and interventions, the study offers rich and multifaceted insights. The robust statistical analysis, coupled with meta-analytical techniques, provides more precise estimates of the effects and their statistical significance. Furthermore, the application of the ROB2 tool for assessing the risk of bias adds critical appraisal, allowing for an evaluation of the reliability and validity of the included studies.

Limitations

The study is not without limitations. The high heterogeneity observed in some analyses may impede definitive conclusions, reflecting inconsistency across studies. The potential for publication bias if negative or nonsignificant findings were underrepresented could also skew understanding. Limited generalizability may arise if the studies focus primarily on specific groups or locations. The quality of included studies, as some showed high or some concerns about the risk of bias, may impact overall confidence in the evidence. The lack of long-term outcome data could limit insights into the extended efficacy of telemonitoring and mobile applications in managing AD, and potential overlap in study populations might lead to duplication and distortion of effects. In sum, while this study’s comprehensive and methodologically rigorous approach offers valuable insights, these identified limitations must be considered in interpreting the findings and their implications for practice and policy.

Recommendations for Clinical Care

Healthcare providers should consider integrating telemonitoring and mobile applications into routine dermatology care for patients with AD, as this technology has been shown to improve the quality of life and self-management without negatively impacting severity. Telemedicine should be emphasized, particularly in remote areas, to ensure access to specialized care. Both providers and patients should be encouraged to engage with eHealth services, including mobile applications, to enhance symptom management and treatment adherence [35]. Integrated apps should include features such as medication reminders, symptom tracking, and personalized patient education. Continuous evaluation and collaboration with information technology (IT) specialists are essential to ensure up-to-date, user-friendly technology. Healthcare providers need adequate training in these technologies, and data should be used to create dynamic, real-time care plans. Attention must be paid to privacy, confidentiality, and adherence to relevant laws and regulations, with care tailored to individual preferences and technological literacy. Overall, these recommendations aim to leverage the benefits of telemonitoring and mobile applications to enhance patient outcomes and efficiencies within the healthcare system, contributing to more patient-centered care.

Conclusions

This systematic review and meta-analysis offers a robust evaluation of the integration of telemonitoring and mobile applications in the care of patients with AD, synthesizing evidence across diverse studies and methodologies. The findings highlight the substantial benefits of these technologies in improving quality of life, fostering self-management, and potentially optimizing healthcare resources. The success and enthusiasm shown toward telemonitoring and digital health interventions, especially against the backdrop of the COVID-19 pandemic, herald a transformative moment in dermatology care, extending the reach of specialized services to remote regions and amplifying patient engagement. Moreover, the nuanced insights derived from assessments such as POEM and DLQI demonstrate a complex interplay of factors contributing to better clinical outcomes. However, there are some concerns of bias in the studies analyzed and the variability in the results. Navigating these complexities requires a delicate balance of technology, medical expertise, patient preferences, legal considerations, and financial considerations. The strength of this review lies in its rigorous analysis, while its recommendations offer actionable, nuanced strategies tailored to diverse healthcare settings. The lessons drawn are applicable well beyond the confines of AD, serving as a beacon for the broader healthcare landscape, navigating toward a future where technology and human touch are harmoniously intertwined.