Skip to main content
. 2022 Dec 13;30(3):588–603. doi: 10.1093/jamia/ocac240

Table 1.

Characteristics of the included studies

Characteristics Studies (n) %
Geographical location of study
 United States 34 87
 Europe (Spain, Finland) 2 5
 Asian (Taiwan, China) 2 5
 Australia 1 3
Year of publication
 Before 2001 3 8
 2005–2010 8 21
 2011–2015 11 28
 2016–2021 17 44
Application fielda
 Etiology 0 0
 Diagnosis 26 67
 Prognosis 11 28
 Therapy 3 8
Clinical domain
 Oncology 4 10
 Cardiovascular 6 15
 Infectious diseases 4 10
 Pulmonary diseases 7 18
 Radiology 3 8
 Adherence to clinical guidelineb 4 10
 Maintenance of a problem list 2 5
 Other (eg, epilepsy, depression, hospital admission) 9 23
Type of patient care
 Inpatient 23 59
 Outpatient 7 18
 Both 6 15
 Unknown 3 8
Sample size validationc
 <500 20 51
 500–10 000 13 33
 >10 000 5 13
Free-text used for text mining
 Reports (eg, radiology reports, pathology reports) 16 41
 Orders (eg, clinical orders) 2 5
 Discharge summaries 2 5
 Clinical notes/free-text narratives (eg, free-text documents) 14 36
 Other (eg, messages, scheduling data) 5 13
TM-CDS tool
 Software already developedd 16 41
 Developed for the studye 23 59
 Source available softwaref 15 38

CDS: clinical decision dupport; USMSTF: United States Multi-Society Task Force; TM: text mining.

a

One study was included in 2 categories being, diagnosis and therapy.

b

This includes studies that tried to improve adherence to clinical guidelines.

c

This includes patients and free-text, Supplementary Appendix S2 shows which are patients and which are free-text.

d

This includes tools that were developed before the study, meaning already existing tools that were used in this study.

e

This includes tools that were developed in the study by the authors.

f

Tools whereof the source was available for others, containing tools that were developed for the study or tools whereof the software was already developed.