Background Optimal monitoring of disease course and individual fluctuation in patients with rheumatic diseases requires both frequent and real-time assessment of relevant outcomes. Traditionally, patient-reported outcomes (PROs) have been collected using questionnaires completed with pen and paper (P&P). However, as a measurement tool in more frequent data collection, the P&P method is impractical and P&P registrations are not always filled in at the intended time (1). In contrast, data collected with text messages (SMS) on mobile telephones are time stamped and thus real-time assessments. This precludes retrospective reporting of data with regard to real time. Mobile phone ownership is rapidly increasing worldwide, making it a well-suited tool for frequent data collection and real-time assessments of PROs.
Objectives To compare daily registrations of PROs assessed with SMS messages and P&P forms with regard to scores, variation of scores and feasibility of methods.
Methods Pain, fatigue, stiffness and ability to carry out daily activities were assessed daily on a numeric rating scale (NRS, 0 to 10, 0 = best). Every other day the scales were reported on SMS and every other day on P&P forms for a total of 28 consecutive days. When receiving the SMS message, the patients scored each of the four variables and separated the scores with a period. A correct SMS response could for instance be: “220.127.116.11”. Data from the SMS responses were electronically transferred to a secure data server. Completed P&P forms were mailed using postage prepaid envelopes, and data from the forms were entered manually into the statistical software package. Variations, expressed as the pooled standard deviation or as the average range between the maximum and minimum scores for the two formats were compared with paired sample t-tests or Wilcoxon Signed Rank Test.
Results Twenty-one patients with various rheumatic diseases, mean (SD) age 49.7 (12.2) years, 70 % female and 90 % with disease duration > 6 years, were included. The mean scores of the four variables did not differ significantly between the SMS and the P&P formats, and the variations were comparable (p>0.05)(table). Agreements between the two formats, illustrated by Bland-Altman plots for all four variables, were acceptable. Overall, the SMS data sets were more complete than the P&P data sets with 10.4% versus 28.6% missing. When asked for preference of format, 85 % preferred the SMS format while 15 % were indifferent to type of format. None of the patients preferred the P&P format.
Conclusions Collection of PROs provides more complete data sets on mobile phones than on paper questionnaires. The two formats are comparable with regard to scores and variation of scores. Our results encourage the use of text messages in frequent data collection aiming at real-time assessments.
Lane SJ, Heddle NM, Arnold E, Walker I. A review of randomized controlled trials comparing then effectiveness of hand held computers with paper methods of data collection. BMC Med Inform Decis Mak. 2006;6:23
Disclosure of Interest None Declared