In rheumatology, self-administered patient questionnaires provide important information for clinical decision-making, clinical studies and scientific evaluations.1^–3 Paper-based assessments are easily handled by patients, but their processing is time-consuming before they can be incorporated for clinical use, decision-making or scientific purposes.4 5 The use of mobile electronic devices such as Tablet PCs simplifies data acquisition and accelerates information transfer between patients and physicians by eliminating intermediate data collection and processing steps.6^–8 Thus, increased quality of care might be achieved.

Before introducing Tablet PCs into clinical routine, careful comparison of data obtained by paper–pencil and computerised versions of the assessments is crucial, because the equivalence of data obtained by the two application methods can not be taken for granted.4 5 7 However, only a few studies have evaluated mobile Tablet PCs in clinical settings, and data on patients attending rheumatology clinics are unavailable.8^–10

We developed the documentation software DocuMed.rh for physicians, used in routine rheumatology outpatient care. It also allows the direct entry of data of self-administered questionnaires into the system using a Tablet PC. We studied patients’ mobile data entry of a set of self-administered questionnaires using the Tablet PC. The capability of patients with rheumatoid arthritis (RA), systemic lupus erythematosus (SLE) or spondyloarthritis (SpA) to handle the Tablet PC and patients’ preferences in comparison with paper-based assessments were assigned.

METHODS

DocuMed.rh, established for the documentation of clinical information of patients with rheumatic diseases in Germany since 1995, runs as client server applications with platform-independent clients and regular web browsers. Common web-based forms enable data entry.

DocuMed.rh offers the application of computerised self-administered questionnaires: Hannover Functional Questionnaire (FFbH) with a calculated Health Assessment Questionnaire (HAQ) index, Bath Ankylosing Spondylitis Disease Activity Index (BASDAI), Medical Outcome Studies Short Form 36 (SF-36), and other instruments. Integrated algorithms calculate questionnaires’ scores.

On regular scheduled visits 153 outpatients completed sets of questionnaires. All patients answered both paper–pencil and electronic questionnaires, and the order in which they did this was randomised. Patients were familiar with the paper–pencil forms, except for the SF-36.

Data capture was conducted in a separate quiet room.11 The study co-ordinator logged into DocuMed.rh, selected the forms and gave standardised short instructions. Patients answered their questions and saved the data themselves.

Experiences with the Tablet PC and history of computer/internet use were documented according to an assessment we used previously.

Signed patients’ consents were obtained. Ethical approval had been received from the local ethic committee.

Arithmetic mean, standard deviation and Pearson’s correlation coefficient were used for statistical description. Differences of locations were tested non-parametrically (Kruskal–Wallis tests, Wilcoxon rank-sum test and Wilcoxon signed-rank test). Internal consistency reliability was computed as Cronbach’s α. All statistical tests were performed two-tailed, p-values less than 0.05 were considered significant. Statistical computations used SAS 9.1.3 (SAS Institute Inc., Cary, North Carolina, USA 2004).

RESULTS

From 172 consecutive outpatients that were screened, 19 patients refused to participate due to various computer-unrelated reasons. A total of 153 patients (61 RA, 60 SLE, 32 SpA) were included in the final analyses.

Sociodemographic data

The patients were predominantly women (69.3%) and mean age was 45.7 (14.4) years. Age varied significantly between disease groups (RA 52.9 (13.6) years, SLE 38.9 (12.3) years, SpA 44.4 (13.3) years; p<0.01). The mean disease duration was 9.2 (7.7) years (median 7.0) without significant differences. Functional disability measured by the FFbH did not differ between the disease groups. No significant difference in the education level was notable between the RA and SpA groups; however, fewer patients with SLE had an academic degree.

Computer and internet experience

Of the patients, 69.9% reported regular personal computer use for 9.8 (5.4) years (mean (SD)), and 68.6% stated regular internet use since 5.0 (2.6) years; 3.9% were professional computer users, 47.7% common users, 17.0% layperson and 25.5% beginners.

Comparison of data acquisition modes

Mean differences of the scores obtained by data entry on the Tablet PC and those obtained by paper–pencil questionnaires are shown in table 1. No significant differences were found between the application modes. Pearson’s correlation coefficients ranged from 0.87 to 0.98 for the total questionnaire scores and for all (sub)scales of the SF-36, except for the vitality scale of the SF-36. No factor explaining the difference in the vitality scale could be determined in detailed statistical analyses; therefore, difference was regarded at random. Scores correlated similarly in all age, gender, education and functional disability groups, and independently from self-reported previous computer knowledge. Figure 1 shows the correlation of the FFbH as an example for the other scores. Cronbach’s α for the SF-36 were nearly identical for the paper–pencil and the computerised questionnaires (range 0.69–0.94).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1 Scattergram of Hannover Functional Questionnaire (FFbH) Scores (paper–pencil versus Tablet PC).

Fully completed paper-based questionnaires were obtained from 97.4% of the patients for FFbH/HAQ, 100% for BASDAI and 86.3% for SF-36. Corresponding figures for the computerised version were 86.9%, 96.9% and 81.0% respectively. Numbers of missing items ranged from 0 to 11 and were significantly higher in the computerised versions. No correlations of missing items to patients’ rheumatic disease, gender, age, order of the application mode or detailed ratings of patients’ experiences were found.

Experiences using the Tablet PC: patients’ preferences

The diagnosis did not influence the capability to fill in the computerised questionnaires. Seven patients (4.6%, all female) felt uncomfortable with the Tablet PC due to their disabilities (five RA and two SpA). Technical support was necessary in 26.1%, and independent from the disease. However, patients who needed support were significantly older (50.9 (14.9) vs 43.8 (13.8) years).

On six-step Likert scales (1 (very good) to 6 (bad)) the handling of the Tablet PC was rated to 1.39 (0.8) (mean (SD)), handling of the pen to 1.55 (0.93) (mean (SD)) and the comprehensibility of the program to 1.33 (0.66) (mean (SD)).

Overall the computerised questionnaires were rated equivalent to or even better than the paper–pencil versions (see fig 2); readability of the paper-based questionnaires was rated significantly better compared with the computerised versions. Although 62.1% of the patients preferred data entry on the Tablet PC future use was preferred especially in younger patients (p<0.01) and those with better computer knowledge. Neither feeling handicapped by the rheumatic disease nor functional disability correlated with patients’ future preference.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2 Patients’ ratings and preferences of the two application modes.

DISCUSSION

Treatment benefits for patients with rheumatic disease in routine patient care are well reflected by self-reported outcome questionnaires.3 In our study we were able to show that data acquisition of FFbH/HAQ, BASDAI and SF-36 using DocuMed.rh software on a Tablet PC is efficient and capable in patients with RA, SLE and SpA. In our cohort—representative for German patients with inflammatory rheumatic diseases—data of the two different application modes were nearly completely exchangeable. Collecting self-reported data via electronic data capture is feasible to patients—even to those who are very ill or inexperienced in the use of computers—and useful in clinical settings as assessments get more straightforward and less time-consuming.4 12 In concordance with other studies electronic data acquisition via Tablet PC performs similarly to traditional paper–pencil versions independently of age, gender, profession/education, and computer knowledge, although patients who were younger and more computer skilled preferred electronic media in the future more often than others.1 4 5 7 10 13 Our application permits regular disease activity, efficacy and safety assessments, thereby contributing intensely to improved patient–doctor interactions, improved patient care and patient empowerment.2 11 12

Comparable approaches of electronic data capture have been made using other electronic devices as Personal Digital Assistants, touchscreen scenarios or others in rheumatology outpatients.1 4 11 13^–15 However, a major advantage of electronic data capture via Tablet PC is the use of the same web-based software application applied on a common client or desktop. No additional software developments or user training is needed. In addition it allows rapid adoption to other clinical scenarios and disciplines, other patient assessments as well as foreign languages.7

Data acquisition on a Tablet PC was not influenced by the disease or the functional disabilities. As was already reported by Cook et al our patients rated the Tablet PC easier to use despite their given functional disabilities.10 Patients’ preference for the electronic versions of the questionnaires in the future mirrors findings from other studies.4 5 7 13 Readability was the only disadvantage compared with the paper–pencil forms, but this can be modified technically.6

Limitations

Our data represent cross-sectional data only from outpatients of a university clinic, larger studies in different clinical settings are needed. The software used allowed some questions to be unanswered—an important feature of personal choice.15 Unintentional non-response unfortunately was not avoided by having the skipped items represented a second time, eg, by error prompts.6 14 This feature will be integrated into the software.

CONCLUSIONS

Our data show that remote data entry via Tablet PC is good and capable of incorporating frequent disease activity, efficacy and safety assessments directly into routine clinical practice. Patient assessments obtained via remote data entry can directly be used for clinical decision-making at the patient visit, enhance quality of care and patient–doctor interactions, and contribute to patients’ self-empowerment in rheumatology.