Skip to main content

Table 2 Social media data characteristics

From: Utility of social media and crowd-intelligence data for pharmacovigilance: a scoping review

Social media data characteristics (n = 70)

Count (%)

Number of social media sources included by study authors

1

38 (54.3%)

2

10 (14.3%)

3

8 (11.4%)

4

3 (4.3%)

> 5

7 (10.0%)

Not reported

4 (5.7%)

Type of social media sites

Patient-specific

35 (50.0%)

General population

27 (38.6%)

Both patient-specific and general population

7 (10.0%)

Not reported

1 (1.4%)

Region of origin of the social media posts

USA

5 (7.1%)

Spain

2 (2.9%)

Germany

1 (1.4%)

50+ countries

1 (1.4%)

USA, Canada, UK, Australia, New Zealand

1 (1.4%)

France

1 (1.4%)

UK, North America, Australasia

1 (1.4%)

Not reported

58 (82.9%)

Language of the social media posts

English

60 (85.7%)

Spanish

2 (2.9%)

French

2 (2.9%)

German

1 (1.4%)

Serbian

1 (1.4%)

Multilingual

2 (2.9%)

Not reported

2 (2.9%)

Method employed to collect social media data

Web crawling/ spidering software

19 (27.1%)

API of the host site

14 (20.0%)

Keyword search

7 (10.0%)

Multiple methods

4 (5.7%)

Other methods

13 (18.6%)

Not Reported

13 (18.6%)

Duration (years) of social media data

Median (Q1, Q3)

1.13 (0.5, 7.13)

Not reported

34 (48.6%)

Number of social media posts retrieved

Median (Q1, Q3)

42,594 (4608, 711,562)

Not reported

5 (7.1%)