Comparison of two data collection processes in clinical studies: electronic and paper case report forms
1 Département de la recherche clinique et du développement, AP-HP, Groupe hospitalier Cochin Hôtel-Dieu, URC Économie de la Santé Ile de France, F-75004 Paris, France
2 AP-HP, Hôpital Robert Debré, Unité d’Épidémiologie clinique, Groupe Hospitalier Robert Debré, 48, Bld Sérurier, F-75019 Paris, France
3 Université Paris Diderot, PRES Sorbonne Paris Cité, F-75019 Paris, France
4 Service de Santé Publique, AP-HP, Groupe hospitalier Albert Chenevier- Henri Mondor, F-94010 Créteil, France
BMC Medical Research Methodology 2014, 14:7 doi:10.1186/1471-2288-14-7Published: 17 January 2014
Electronic Case Report Forms (eCRFs) are increasingly chosen by investigators and sponsors of clinical research instead of the traditional pen-and-paper data collection (pCRFs). Previous studies suggested that eCRFs avoided mistakes, shortened the duration of clinical studies and reduced data collection costs.
Our objectives were to describe and contrast both objective and subjective efficiency of pCRF and eCRF use in clinical studies. A total of 27 studies (11 eCRF, 16 pCRF) sponsored by the Paris hospital consortium, conducted and completed between 2001 and 2011 were included. Questionnaires were emailed to investigators of those studies, as well as clinical research associates and data managers working in Paris hospitals, soliciting their level of satisfaction and preferences for eCRFs and pCRFs. Mean costs and timeframes were compared using bootstrap methods, linear and logistic regression.
The total cost per patient was 374€ ±351 with eCRFs vs. 1,135€ ±1,234 with pCRFs. Time between the opening of the first center and the database lock was 31.7 months Q1 = 24.6; Q3 = 42.8 using eCRFs, vs. 39.8 months Q1 = 31.7; Q3 = 52.2 with pCRFs (p = 0.11). Electronic CRFs were globally preferred by all (31/72 vs. 15/72 for paper) for easier monitoring and improved data quality.
This study found that eCRFs and pCRFs are used in studies with different patient numbers, center numbers and risk. The first ones are more advantageous in large, low–risk studies and gain support from a majority of stakeholders.