Developing and Validating a Tablet Version of an Illness Explanatory Model Interview for a Public Health Survey in Pune, India

Abstract

BackgroundMobile electronic devices are replacing paper-based instruments and questionnaires for epidemiological and public health research. The elimination of a data-entry step after an interview is a notable advantage over paper, saving investigator time, decreasing the time lags in managing and analyzing data, and potentially improving the data quality by removing the error-prone data-entry step. Research has not yet provided adequate evidence, however, to substantiate the claim of fewer errors for computerized interviews.MethodologyWe developed an Android-based illness explanatory interview for influenza vaccine acceptance and tested the instrument in a field study in Pune, India, for feasibility and acceptability. Error rates for tablet and paper were compared with reference to the voice recording of the interview as gold standard to assess discrepancies. We also examined the preference of interviewers for the classical paper-based or the electronic version of the interview and compared the costs of research with both data collection devices.ResultsIn 95 interviews with household respondents, total error rates with paper and tablet devices were nearly the same (2.01% and 1.99% respectively). Most interviewers indicated no preference for a particular device; but those with a preference opted for tablets. The initial investment in tablet-based interviews was higher compared to paper, while the recurring costs per interview were lower with the use of tablets.ConclusionAn Android-based tablet version of a complex interview was developed and successfully validated. Advantages were not compromised by increased errors, and field research assistants with a preference preferred the Android device. Use of tablets may be more costly than paper for small samples and less costly for large studies.

Implementing electronic data capture at a well-established health and demographic surveillance site in rural northern Malawi

“…Our quantitative results on data quality are consistent with findings from other evaluations [3,6,7,10,11,15–19]. On paper forms, the level of missing data increased with the complexity of skip patterns, while almost the opposite pattern was seen in EDC, showing the usefulness of programmed automatic skip patterns in guiding interviewers to the correct questions.…”

“…With a study size above a certain threshold (as start-up costs can be high), costs tend to be lower with EDC [6,7,10,13]. We were able to reduce costs in some areas, for example by buying devices in the UK, as they were expensive to buy locally.…”

“…Data quality : Systems of double-entry, checks, and verification ensure a high level of data quality in a paper-based system, and EDC must meet the same standards: most evaluations found that EDC was less prone to errors than paper [3,6,10,11,15–18]. …”

Evaluation of Electronic and Paper-Pen Data Capturing Tools for Data Quality in a Public Health Survey in a Health and Demographic Surveillance Site, Ethiopia: Randomized Controlled Crossover Health Care Information Technology Evaluation (Preprint)

“…In addition to that, a study with a similar design in India revealed with 2% PPDC and 1.99% EDC error rates [29]. However, this magnitude is lower compared to other similar studies (7% and 1%, with no omission in EDC) [9], 3.6% PPDC and 5.2% in EDC [33].…”

“…Similar experience observed in a couple of recent studies [7,29]. The exact copy of the paper questionnaire converted to its electronic replica form using Excel and XLSForm.…”