Table 2.

Comparison of eCohorts to traditional cohorts.

Characteristic	eCohort	Traditional cohort
Research question	Broad, multipurpose, interdisciplinary questions; questions may be rooted in citizen science and attached to methodological elements (eg, use of technology in epidemiological studies), change dynamically, and may be answered prospectively	Broad, multipurpose questions; questions change dynamically and are mostly answered with existing data
Sampling	Usually nonrandom sampling with self-selected volunteers	Random samples or clinic populations defined by event, exposure, or disease
Recruitment	Primarily online advertisement (eg, webpages, newsletters, forums, social media), but can be complemented by offline approaches (eg, flyers, posters) Recruitment usually online, through dedicated study webpages, possible at any place, any time Electronic consent procedures	Primarily offline advertisement (eg, flyers, posters, newspaper advertisements), but increasingly complemented by online approaches Recruitment usually within clinical (eg, by health care providers) or community setting and appointment-based Consent procedures usually face to face and paper-based
Baseline data collection	Primarily online and usually directly reported by participants (eg, web-based surveys). Sometimes complemented by offline data collection (eg, mailed surveys) and nonself-reported data (eg, medical record data)	Primarily offline (eg, paper-based questionnaires, data retrieval from existing records, personal interviews), and may be combined with medical examinations and biosampling
Follow up	Primarily online and usually directly reported by participants (eg, web-based surveys, personalized email, or SMS text message reminders) Rarely linked to medical care. Use of internet (eg, study website, social media, newsletters) for outreach and participant contact/engagement Data quality, reliability, and internal validity may be a concern Data quality tradeoffs due to self-reporting; need for simpler questions, better data management, and user-friendliness	Primarily offline (eg, paper-based questionnaires, data retrieval from existing records, personal interviews, medical examinations, and biosampling, mailed reminders) Usually linked to medical care; personal relationship (or at least personal interactions) between participant and study coordinators Strong focus on data quality, reliability, and internal validity
Analysis	Usually built on self-reported data Easier data access, preliminary analyses possible in shorter time frames Analyses tend to have a stronger participant (patient) focus	Built upon multiple data streams, and a combination of clinical and self-reported data Longer process, preliminary analyses more difficult in short time frames Analyses tend to have a stronger clinical/biomedical focus
Dissemination	In addition to publications, through a variety of online channels (eg, websites, social media) More frequent dissemination of findings Dissemination may be a part of an overall strategy to keep participants engaged Opportunities for reproducibility and open science	Primarily focused on scientific publications Subject to larger time gaps Dissemination of findings in form of clinical tools (eg, risk scores)