Artificial intelligence promotes shared decision-making through recommending tests to febrile pediatric outpatients

doi:10.5847/wjem.j.1920-8642.2023.033

Abstract

Abstract:

BACKGROUND: To promote the shared decision-making (SDM) between patients and doctors in pediatric outpatient departments, this study was designed to validate artificial intelligence (AI) -initiated medical tests for children with fever.

METHODS: We designed an AI model, named Xiaoyi, to suggest necessary tests for a febrile child before visiting a pediatric outpatient clinic. We calculated the sensitivity, specificity, and F1 score to evaluate the efficacy of Xiaoyi's recommendations. The patients were divided into the rejection and acceptance groups. Then we analyzed the rejected examination items in order to obtain the corresponding reasons.

RESULTS: We recruited a total of 11,867 children with fever who had used Xiaoyi in outpatient clinics. The recommended examinations given by Xiaoyi for 10,636 (89.6%) patients were qualified. The average F1 score reached 0.94. A total of 58.4% of the patients accepted Xiaoyi’s suggestions (acceptance group), and 41.6% refused (rejection group). Imaging examinations were rejected by most patients (46.7%). The tests being time-consuming were rejected by 2,133 patients (43.2%), including rejecting pathogen studies in 1,347 patients (68.5%) and image studies in 732 patients (31.8%). The difficulty of sampling was the main reason for rejecting routine tests (41.9%).

CONCLUSION: Our model has high accuracy and acceptability in recommending medical tests to febrile pediatric patients, and is worth promoting in facilitating SDM.

Key words: Artificial intelligence, Pediatric outpatient, Medical examinations, Shared decision- making

Wei-hua Li, Bin Dong, Han-song Wang, Jia-jun Yuan, Han Qian, Ling-ling Zheng, Xu-lin Lin, Zhao Wang, Shi-jian Liu, Bo-tao Ning, DanTian , Lie-bin Zhao. Artificial intelligence promotes shared decision-making through recommending tests to febrile pediatric outpatients[J]. World Journal of Emergency Medicine, 2023, 14(2): 106-111.

Figures/Tables 7

Figure 1.

Figure 2.

Table 1.

Table 2.

Table 3.

Table 4.

Table 5.

References 19

[1]	Brody DS. The patient’s role in clinical decision-making. Ann Intern Med. 1980; 93(5): 718-22. doi: 10.7326/0003-4819-93-5-718 pmid: 7212484
[2]	Fiks AG, Mayne SL, Karavite DJ, Suh A, O’Hara R, Localio AR, et al. Parent-reported outcomes of a shared decision-making portal in asthma: a practice-based RCT. Pediatrics. 2015; 135(4): e965-73. doi: 10.1542/peds.2014-3167
[3]	Allen KA. Parental decision-making for medically complex infants and children: an integrated literature review. Int J Nurs Stud. 2014; 51(9): 1289-304. doi: 10.1016/j.ijnurstu.2014.02.003 pmid: 24636443
[4]	Aarthun A, Akerjordet K. Parent participation in decision-making in health-care services for children: an integrative review. J Nurs Manag. 2014; 22(2): 177-91. doi: 10.1111/j.1365-2834.2012.01457.x
[5]	Committee on Hospital Care. American Academy of Pediatrics. Family-centered care and the pediatrician’s role. Pediatrics. 2003; 112(<W>3 Pt 1):691-6. doi: 10.1542/peds.112.3.691
[6]	Lipstein EA, Brinkman WB, Britto MT. What is known about parents’ treatment decisions? A narrative review of pediatric decision making. Med Decis Making. 2012; 32(2): 246-58. doi: 10.1177/0272989X11421528 pmid: 21969136
[7]	Fiks AG, Jimenez ME. The promise of shared decision-making in paediatrics. Acta Paediatr. 2010; 99(10): 1464-6. doi: 10.1111/j.1651-2227.2010.01978.x
[8]	Wyatt KD, Prutsky Lopez G, Domecq Garces JP, Erwin P, Brinkman WB, Montori VM, et al. Study protocol: a systematic review of pediatric shared decision making. Syst Rev. 2013; 2: 48. doi: 10.1186/2046-4053-2-48 pmid: 23816266
[9]	Hallström I, Elander G. Decision-making during hospitalization: parents’ and children’s involvement. J Clin Nurs. 2004; 13(3): 367-75. pmid: 15009339
[10]	Armstrong MJ, Shulman LM, Vandigo J, Mullins CD. Patient engagement and shared decision-making: what do they look like in neurology practice? Neurol Clin Pract. 2016; 6(2): 190-7. doi: 10.1212/CPJ.0000000000000240 pmid: 27104070
[11]	Dobler CC, Sanchez M, Gionfriddo MR, Alvarez-Villalobos NA, Singh Ospina N, Spencer-Bonilla G, et al. Impact of decision aids used during clinical encounters on clinician outcomes and consultation length: a systematic review. BMJ Qual Saf. 2019; 28(6): 499-510. doi: 10.1136/bmjqs-2018-008022 pmid: 30301874
[12]	Vandvik IH, Førde R. Ethical issues in parental decision-making. An interview study of mothers of children with hypoplastic left heart syndrome. Acta Paediatr. 2000; 89(9): 1129-33. pmid: 11071097
[13]	Bertille N, Fournier-Charrière E, Pons G, Chalumeau M. Managing fever in children: a national survey of parents’ knowledge and practices in France. PLoS One. 2013; 8(12): e83469. doi: 10.1371/journal.pone.0083469
[14]	Bao YH, Fan GR, Zou DD, Wang T, Xue D. Patient experience with outpatient encounters at public hospitals in Shanghai: examining different aspects of physician services and implications of overcrowding. PLoS One. 2017; 12(2): e0171684. doi: 10.1371/journal.pone.0171684
[15]	Stacey D, Légaré F, Lewis K, Barry MJ, Bennett CL, Eden KB, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2017; 4(4): CD001431.
[16]	Liang HY, Tsui BY, Ni H, Valentim CCS, Baxter SL, Liu GJ, et al. Evaluation and accurate diagnoses of pediatric diseases using artificial intelligence. Nat Med. 2019; 25(3): 433-8. doi: 10.1038/s41591-018-0335-9 pmid: 30742121
[17]	National Collaborating Centre for Women's and Children's Health (UK). Feverish illness in children: assessment and initial management in children younger than 5 years. London: Royal College of Obstetricians and Gynaecologists (UK); 2013.
[18]	Sepúlveda J, Velastın SA. F1 score assesment of gaussian mixture background subtraction algorithms using the MuHAVi dataset. 6th International Conference on Imaging for Crime Prevention and Detection ( ICDP-15), London, 2015.
[19]	McKenna K, Collier J, Hewitt M, Blake H. Parental involvement in paediatric cancer treatment decisions. Eur J Cancer Care (Engl). 2010; 19(5): 621-30. doi: 10.1111/j.1365-2354.2009.01116.x

Characteristics	n (%)
Gender
Male	6,317 (53)
Female	5,549 (47)
Age, years
0-1	2,325 (20)
1-3	3,100 (26)
3-6	3,964 (33)
6-12	2,320 (20)
12-18	158 (1)
Fever peak, °C
37.5-38.5	6,527 (55)
38.6-39.5	5,017 (42)
39.6-40.5	323 (3)
Caretakers
Parents	5,149 (43)
Grandparents	6,718 (57)
Diagnosis
Respiratory disease	7,677 (65)
Gastrointestinal disease	2,066 (17)
Neurological disease	1,116 (9)
Systematic disease	471 (4)
Others	537 (5)

Diseases	Total	Qualified recommendation	Unqualified recommendation		Sensitivity	Specificity	F1 score
Diseases	Total	Qualified recommendation	Insufficient	Unnecessary	Sensitivity	Specificity	F1 score
Respiratory disease	7,677	7,001	651	25	0.91	0.99	0.95
Gastrointestinal disease	2,066	1,754	295	17	0.86	0.99	0.92
Neurological disease	1,116	988	66	62	0.94	0.94	0.94
Systematic disease	471	412	45	14	0.90	0.97	0.93
Others	537	481	33	23	0.94	0.95	0.94
Total	11,867	10,636	1,090	141	0.91	0.98	0.94

Characteristics	Acceptance group (n=6,935)	Rejection group (n=4,932)	Univariate analysis		Multivariate analysis
Characteristics	Acceptance group (n=6,935)	Rejection group (n=4,932)	P-value	Odds ratio (95%CI)	P-value	Odds ratio (95%CI)
Gender
Male	3,692 (58.4)	2,625 (41.6)	0.982	1.00 (0.93-1.08)	0.647	1.03 (0.90-1.19)
Female	3,242 (58.4)	2,307 (41.6)	0.982	1.00 (0.93-1.08)	0.647	1.03 (0.90-1.19)
Age, years
0-1	1,372 (59.0)	953 (41.0)	0.383	0.99 (0.98-1.01)	0.507	1.00 (1.00-1.03)
1-3	1,808 (58.3)	1,292 (41.7)
3-6	2,317 (58.5)	1,647 (41.5)
6-12	1,357 (58.5)	963 (41.5)
>12	81 (51.3)	77 (48.7)
Fever peak, °C
37.5-38.5	6,040 (92.5)	487 (7.5)	< 0.01	0.05 (0.04-0.05)	< 0.01	0.05 (0.05-0.06)
38.6-39.5	822 (16.3)	4,195 (83.7)
39.6-40.5	73 (22.6)	250 (77.4)
Caretakers
Parents	1,281 (24.9)	3,868 (75.1)	< 0.01	0.06 (0.06-0.07)	< 0.01	0.07 (0.06-0.09)
Grandparents	5,654 (84.2)	1,064 (15.8)	< 0.01	0.06 (0.06-0.07)	< 0.01	0.07 (0.06-0.09)
Diagnosis
Respiratory disease	5,155 (67.1)	2,522 (32.9)	< 0.01	/	< 0.01	/
Gastrointestinal disease	983 (47.6)	1,083 (52.4)		0.44 (0.40-0.49)		0.42 (0.35-0.50)
Neurological disease	246 (22.0)	870 (78.0)		0.14 (0.12-0.16)		0.13 (0.10-0.17)
Systematic disease	121 (25.5)	350 (74.5)		0.17 (0.14-0.21)		0.18 (0.12-0.26)
Other disease	430 (80.1)	107 (19.9)		1.97 (1.58-2.44)		2.30 (1.58-3.36)

Inspection items	Respiratory diseases	Gastrointestinal diseases	Neurological disease	Systematic disease	Other diseases	Total
Pathogen examinations^a	983 (39.0)	560 (51.7)	217 (24.9)	170 (48.6)	37 (34.6)	1,967 (39.9)
Imaging examinations^b	1,332 (52.8)	329 (30.4)	507 (58.3)	99 (28.3)	37 (34.6)	2,304 (46.7)
Routine examinations^c	207 (8.2)	194 (17.9)	146 (16.8)	81 (23.1)	33 (30.8)	661 (13.4)
Total	2,522 (51.1)	1,083 (22.0)	870 (17.6)	350 (7.1)	107 (2.2)	4,932 (100.0)

Reason for rejection	Pathogen examination	Imaging examination	Routine examination	Total
Ask to consult a doctor before making a decision	416 (21.5)	717 (31.1)	67 (10.1)	1,200 (24.3)
Consider the tests time-consuming	1,347 (68.5)	732 (31.8)	54 (8.2)	2,133 (43.2)
Consider the tests unnecessary	107 (5.4)	636 (27.7)	175 (26.5)	918 (18.6)
Difficulty obtaining samples	26 (1.3)	22 (1.0)	277 (41.9)	325 (6.6)
Other reasons (such as machine failure)	71 (3.6)	197 (8.6)	88 (13.3)	356 (7.2)
Total	1,967	2,304	661	4,932