This is an observational prospective study. All the patients in the study were randomly assigned to first complete the questionnaire by paper and pencil and then by tablet or in the opposite order (Figure 1). After both questionnaires were completed, physicians asked the patients to indicate their preferred version and comment on the accuracy of the paper in terms of it being easy to read and interpret. These additional data were collected together with demographics in a dedicated database.

Sample size was calculated basing on the literature regarding the estimates for sample size requirements for reliability studies using an intraclass correlation coefficient (ICC) [13]. In particular, assuming a possible 25% dropout rate [14], 185 patients were sufficient to detect an expected reliability of 0.8 against an acceptable reliability of 0.7, with an 85% power and a significance level of .05.

There was no need for ethical approval for this study. All the patients signed an informed consent form.

All patients aged 18 to 90 years who attended the Azienda Socio-Sanitaria Territoriale Grande Ospedale Metropolitano Niguarda between July and September 2013 and met the American College of Rheumatology criteria for RA [15] were considered for enrollment.

The questionnaire consisted of 3 independent self-assessment visual rating scales, commonly used in different adult patients including those with rheumatic diseases [16].

1. Visual Analog Scale (VAS) pain [17]: RA patients reported the degree of their pain in a scale ranging from 0 (no pain) to 100 mm (worst imaginable pain).

2. Global Health (GH) score: patients reported the level of impact of the RA disease on their global health with a value ranging from 0 (no effect) to 100 mm (maximum effect).

3. Patient Global Assessment of Disease Activity (PGA) [18]: patients answered the question “In the last week, how active would you define your rheumatic condition?” on a 0 to 100 mm scale ranging from “not active at all” to “extremely active” as anchors. 

Patients received both a device with an app version and a paper version. In the first case, they had to touch the visual scale on the screen of the electronic device at the same point corresponding with the status of their response; once the selected questionnaire was completed and saved, the relative score was automatically calculated. For the paper version, patients were asked to place a perpendicular line in each scale at the point which best matched with the status of their response. The score was determined by measuring with a ruler the distance (mm) between the “0” dash and the patient's mark.

Electronic questionnaires are part of a dedicated Web-based tool that helps physicians managing patients affected by RA. Every piece of information is sent to a cloud system, and the security of the communication is ensured by authentication routines.

Therefore, the platform is accessible both through a link on the Web and through a dedicated app for the tablet. The Web form, accessible through a link, represents the core of the app and includes “entry,” “search,” and “edit” data functions. All data input from the tablet or Web are stored in a dedicated database in the main repository of the system.

At first, the physicians can create detailed and accurate patient profiles by filling in demographic and clinical assessments. Once the profile is created, patients receive a username and password to access the self-examination module and fill in the PROs.

The tool allows real-time data gathering, with data displayed in various ways including pictograms, balloon chart, and chromatic scales. All information is accessible both through the Web site and tablet app, which allow full access to the patients’ data as well as direct access to the report printing functionalities.

The normality of continuous variables was checked by the examination of histograms and confirmed by the Kolmogorov-Smirnov test. All nonnormally distributed data were ranked before further analysis. Demographic characteristics were summarized as count and percentage, mean and standard deviation, and median with range. Any difference between the orders of the questionnaire administration was assessed by using a chi-square test or Fisher's exact test for categorical data and independent sample t test for continuous data.

To evaluate possible order and format effects, any difference between paper and electronic results of each evaluated item was analyzed by an analysis of variance test for repeated measures with format as repeats and order of administration as factor.

For all the items, the level of agreement between the responses of electronic and paper formats was estimated with the ICC, expressed with 95% confidence interval. Fleiss recommendations [19] were followed to identify ICC cut-offs.

A Bland-Altman plot [20] was used to graphically confirm the results and visualize the concordance degree. This chart, for each subject, plots the difference between the 2 measurements (the y axis) as a function of the mean of the same values (the x axis). All statistical tests were 2-sided and the significance level (alpha error) was set at .05.

In the modern clinical practice, fast and reliable collection of clinical data is an important need. The widespread use of mobile technologies throughout the world has also involved the medical field, where the use of information technology products and services allows access to health care, both containing costs and improving the quality of data and clinical outcomes.

As reported by the World Health Organization, mHealth or mobile health is a “medical and public health practice supported by mobile devices, such as mobile phones, patient monitoring devices, personal digital assistants, and other wireless devices” [21].

This study proved that a new approach for telemonitoring, where desktop applications are fully integrated with external mobile devices, could play an important role in the patient follow-up. Response rate was 100% for both methods, Internet-based electronic questionnaires comply with the traditional paper formats, and the app is considered easy to use (electronic version is defined as the preferred choice by most patients).

In particular, for all the investigated items, no significant differences (by considering also any order effects) between the 2 approaches were found, and these findings are in agreement with previous studies reporting none or very few differences between computerized and paper-and-pencil assessments [22,23].

Furthermore, it is important to stress that ePRO can provide a valuable source of information. Data can be collected at any time of the day according to the wishes of the patient. In this way, the database can be fed continuously over time, providing the physicians useful tools for better defining strategies to deal with a long-lasting (chronic) disease. These findings have shed some light on the significant value to the patient, but they mostly highlight the benefit that the overall health care system can enjoy from this new digital app.

These first results are applicable to RA patients, but further studies are recommended because the comparison between 2 approaches in other populations and settings has not yet been studied.

Furthermore, performed comparisons should be weighted by considering the low amount of collected data, in this case 3 independent single items. It would be interesting to gather data in order to study more complex measures of outcomes.

Sufficient privacy and security should be guaranteed: data transferred to a health care provider may be subject to hacking. A disaster recovery system should be ready at all times, and technical support for corrupted or erased data must be set up.

Finally, it is important to point out that rheumatology specialist care should never be substituted: telemonitoring systems are to be considered as an important challenge in the modern times, but they are complementary tools and not alternatives in the routine medical practice.