A cross-sectional research design was used to explore the use of two online self-management programs among patients with AD or FA. The measure of usage was the total number of log-ins in the study period. Data of usage was obtained during the patient’s use of the program and was embedded into the program design. Patients had access to the program for a three month period. The number of log-ins was measured for all participants of both programs, and the number of hits on the system’s core features was measured only among the participants who provided informed consent. To explore the patients’ characteristics a questionnaire-based online survey was conducted on a convenience sample of patients attending the online self-management programs. All patients who provided informed consent were included, and the questionnaires were incorporated at the start of the program.

Study population consisted of adult patients with FA or with AD who received an account for one of the online programs. To investigate patient characteristics, all patients who provided informed consent since the start of the programs were included. The gender of each patient and the health care provider who enrolled them in the program were registered, and from the start of the program the number of visits to the site was counted. To examine the usage of the program, patients were recruited from the participants of both online self-management programs between October 2012 and November 2013, because since October 2012 it has been possible to measure the number of hits on the system’s core features.

Eligible patients were at least 18 years old, Dutch speaking, and had a clinical diagnosis of AD or FA. Patients were referred to the online programs by GPs, specialists, dieticians or nurses. After referral, they received an account from which they were given access to the training. The account was valid for a period of three months in which patients could complete the program, but this period could be extended at the request of the participant. Informed consent was asked at the start of the online program through a specific letter and was incorporated as a link in the webpage of the program. Their reply was registered and they received a copy via email. A flowchart of the study can be found in Figure 1.

The Medical Ethics Review Committee of UMC Utrecht confirmed that the Medical Research Involving Human Subjects Act did not apply to this study.

Statistical analyses were performed using SPSS Statistics 20.0 (IBM Corporation, Somers, NY, USA). Standard descriptive statistics were used to summarize sample characteristics and to describe numbers of log-ins and hits per module and per functionality.

The total score on the DLQI, FAQLQ-AF and the sum scores on the eight different domains of the heiQ are at interval/ratio level measurements, and correlations with the number of log-ins were calculated using Pearson’s product-moment correlation. The correlation between the extent + severity part of the ISDL, and usage was also calculated using Pearson’s product moment.

For the analysis of factors associated with number of log-ins into the online program, multiple linear regression was used. Variables related to the number of log-ins with a significance level of ≤ 0.1 were included in the model. Categorical variables were converted into dummy variables to perform the regression analysis. Prior to each regression analysis, data were checked for linearity and normality by performing a residual analysis, and checked for multicollinearity.

Since the start of both programs, a total of 513 patients received an account for the online program by their physician, nurse or dietician: 299 patients for AD and 214 for FA (Figure 1). Of the AD patients, 62% were women; in the FA program 74% were women. Most patients were referred by the university hospital which developed and started the program (Table 1). The mean number of log-ins for the AD program was 1.4 (range 0-15). For the FA program the mean number of log-ins was 1.3 (range 0-16); 58% of patients of both programs logged in on at least one occasion.

Of the 299 AD patients referred to the online program, 173 logged in on at least one occasion and 95 gave informed consent for the use of their data. Eighty-one patients filled in questionnaires, but 6 were excluded because they were under 18 years of age (n=4) or age was unknown (n=2). Reasons for not using the program or no informed consent were not given, because patients were asked to give their informed consent (yes or no) online when first visiting the program.

Data of 75 patients was available for analysis (Table 2). Of them 67 % were women and the mean age was 34 years (SD 15). Mean number of log-ins was 3.1 (range 1-11; SD 2.6). Patients had mild to moderate AD; mean ISDL 18.8 (range 10-34) and a mean VAS score of itching of 6.0 (range 1-10). AD had a moderate effect on patients’ lives (mean score DLQI 9.6; range 0-27).

A low, but significant correlation was shown between the number of log-ins and age (Pearson correlation coefficient r=.38, P=.001), and between number of log-ins and quality of life (r=.32, P ˂.01). Correlations between number of log-ins and domains of self-management, such as emotional well-being and skills and techniques were .2 (P=.09) and -.19 (P=.1), respectively. No correlation was shown between number of log-ins and sex or with other domains of self-management. Linear regression with number of log-ins as dependent variable and entering age, quality of life, and two domains of self-management, namely emotional well-being and skills and techniques, led to an R ²of .23. Age and quality of life contributed the most to the model, with betas of .35 (P=.002) and .26 (P=.05), respectively. Two of the self-management domains, emotional well-being, and skills and techniques did not significantly contribute to the model.

Of the 214 FA patients referred to the online program, 124 logged in at least once and 61 gave informed consent for the use of their data. Forty-nine patients filled in questionnaires, but 4 were excluded because they were under 18 years of age. Reasons for not using the program or no informed consent were not given.

Data of 45 patients were available for analysis (Table 3). Of them 80 % were women and mean age was 35 years (SD 13). Mean number of log-ins was 3.0 (range 1-11; SD 2.3). The mean number of food allergies was 3.5 (range 1-8), with highest percentages of allergies for tree nuts, peanut, and fruit/vegetables. Of them, 73% had been prescribed an adrenalin pen.

A low, but significant negative correlation was shown between the number of log-ins and the domain of self-management “social integration and support” (r=-.36, P=.02). No correlation was shown between number of log-ins and sex, age, possession of adrenaline pen, aspects of quality of life, and other domains of self-management. Linear regression with number of log-ins as dependent variable and entering “social integration and support” led to an R ²of .13.

The modules “What is AD” and “Treatment of AD” were the most visited modules of the AD program with 34% and 32% of all hits respectively. The module “How is it diagnosed” in the FA program was most visited with 24% of all hits (Table 4).

Most hits (excluding hits on introduction of a module) were on the informational parts of the modules (55-58%), followed by exercises (30-32%). Ten percent or less of all hits were on videos and patient narratives (Table 5).

The online self-management programs “Living with eczema” and “Living with food allergy” were mostly used by patients of our university center. Of all referred patients, 58% logged in for at least one occasion. Patients who participated in the online program had an average of three log-ins and mostly visited the modules with information about the disease (AD or FA) and its treatment. Most hits were on the informational sections of the modules and on exercises. It seemed that higher age and lower quality of life influenced use of the AD program, while the lower scores on “social integration and support” influenced use of the FA program. However, explained variance of use was low.

The feasibility evaluations of both online programs, carried out previously, showed that usefulness and usability of the programs were well-appreciated. According to the TAM model [15], this could predict use of the program. However, the underlying study showed that about 40% of patients referred to the online program never logged in. Because no data are available about patients who received an account but chose not to use the program, we do not know if there were differences between users and non-users. It is known that high attrition rates in eHealth interventions are not uncommon [8,22,23]. We did not investigate reasons for the non-usage attrition, but some patients mentioned during face-to-face consultations at the outpatient department that they had already received enough information. So it could be that the “right” users, who could benefit the most from the program, were not enrolled in all cases, which might lead to increased non-use [22,24]. Moreover, the online program is not fully integrated in the usual clinical face-to-face care, but was offered as an addition to usual care, which could also lead to non-use [23]. During the program, there was little room for human interaction. Patients could only receive feedback from a nurse on one specific exercise in the AD program. In the FA program, patients received automatic feedback on most exercises and they received personal feedback from a dietician on only one exercise. It was earlier reported that involvement of interactive technologies with human interaction and support can reduce attrition rates [8,24,25].

Besides the factors related to the online program itself, patient characteristics may also influence the use of the program. We concluded that higher age and lower quality of life were associated with use of the “Living with AD” program, but the explained variance was 23%. The mean age of users of the AD program was 34 years. It is possible that relatively older patients more often took the opportunity to visit the program than young adults. Usage of eHealth interventions has also been studied in patients with other chronic diseases. For example, users of a health weight assistant [26] and a Web-based intervention for heart disease self-management [27] were also more often of older age. Besides age, it was pointed out that the need for information or the need for care influenced use of the eHealth intervention [24,27], which was confirmed by our finding that decreased quality of life increased use of the online AD program. The FA patients in this study had a moderately impaired quality of life (mean score 3.8, SD 1.1). However, quality of life was not correlated to the use of the online program as in the AD patients. One explanation might be that for both patient groups, disease specific quality of life questionnaires were used, which were not comparable due to different aspects of quality of life being measured. On the other hand, lower scores on “social integration and support”, a domain of self-management, influenced the number of log-ins by FA patients. It is known that the social consequences of having a food allergy, such as feeling isolated or embarrassed, also influence quality of life [6].

Because of the low explained variance in the use of both programs, other factors than demographics, disease severity, quality of life, and self-management may be of influence. We expect that eHealth literacy is such a factor. eHealth literacy is defined as “the ability to seek, find, understand, and appraise health information from electronic sources and apply the knowledge gained to addressing or solving a health problem” [28]. eHealth literacy is not a static skill but develops over time and can be influenced by the health status of a person, his educational background, motivation to look for information, and the technologies that are used [28]. The eHealth literacy of participants of this study are not known, but it is of interest to further explore these skills and investigate their relation to the use of these online programs.

There is no doubt about the importance of patient education and self-management support for patients with AD [29] or FA [30]. However, little is known about the most preferred method for these patients. In the studied online programs, the information parts were most frequently visited, while we expected that videos and patient narratives would be more appealing. It was shown in one earlier study that online videos providing patient education were effective and attractive to patients with AD [31]. Another Web-based program for AD patients also found lower use of videos than expected, despite careful development of that program and adaptation after initial feedback [32]. Parents of food allergic children reported that they preferred a variety of formats for patient education, because of the differences in learning styles: paper-based, Web-based, and video-based information. They also preferred that these materials were recommended by reliable organizations [33,34]. We expected that the high use of the informational parts of these programs might also be related to the low number of log-ins. The different modules of the AD and FA program always start with the information pages, and participants have to take further steps for the online exercises and videos. A change in the order in which the different functionalities are offered might give the users more freedom of choice in using the online program.

A limitation of the study was that disease severity, quality of life, and self-management were only measured at the start of the program; as a result, effects of the program on these clinical outcomes are unknown and it is not possible to analyze which patients will benefit most from the program. Further research with a longitudinal design will give additional insight into the effects of the programs. Moreover, most patients were referred by a university center. In this center specialized nurses or dieticians already support patients with information and education The actual need for information in this university center may be lower in than in less specialized or general hospitals. Implementation of the program at regional hospitals or in community care would probably increase the use of the programs and extend it to a more diverse patient group. Knowledge about use and factors influencing use of specific self-management programs can contribute to optimal usage of these programs, which in turn will increase the intended effects on clinical outcomes. However, generalizability of the findings of this study is limited due to the small sample size and specific adult patient group.

Physicians, dieticians, and nurses, mostly from a university center, regularly referred their patients to the online programs “Living with food allergy” or “Living with eczema” for online self-management training, in addition to the usual face-to-face care. Nearly 60% of all referred patients logged in. All modules seemed to be relevant, but there is room for improvement in use of the program. Age, quality of life, and lower social integration and support were related to use of the program, but only a part of variance in use could be explained. Further research is needed into predictors of use related to the program and characteristics of users, as well as further research into the effects of the program on clinical outcomes.