Table 2 Purposes of AI in emergency departments

1 Underlying problem they aim to solve

  1a

To improve patient outcomes (including mortality, morbidity, and satisfaction)

15, 16, 20, 22–24, 26, 33–36, 38, 40, 43, 44, 46

  1b

To reduce or cope with overcrowded EDs

17, 19, 22, 25–32, 37, 41, 45

2 The means through which the studies aimed to solve the overarching problem

  2a

To accurately predict future outcomes

11, 15, 16, 18, 19, 22–31, 34, 36, 37, 39, 41–45, 47

  2b

To accurately identify outcomes

28, 30, 33, 35, 40, 42, 46, 47

  2c

To reduce the need for a physician

17, 35

3 The focus area of AI use

  3a

To improve ED triage in general or through the prediction or identification of serious or critical (adverse) outcomes

17, 18, 20–29, 32, 37, 41, 45

  3b

To assist clinicians with the prediction or identification of serious or critical (adverse) outcomes

15, 16, 30, 33, 39, 42, 46, 47

  3c

To assist in predicting or identifying non-critical (adverse) outcomes

11, 19, 31, 34–36, 38, 40, 43, 44