Physicians’ intentions to use digital tools – a comparative survey, before and after the COVID-19 pandemic, in Southern Sweden

Figures & data

Figure 1. Theory of planned behavior [7]. Theory of planned behavior schematic representation. The figure is adapted from Ajzen [7]. The items in the questionnaire was developed from the theoretical frame of theory of planned behavior [5]. Four items measure attitudes, two measure subjective norms, three measure perceived behavioral control and two measure intentions for each of the digital tools; digital consultations by text, video consultations, chronic disease monitoring and artificial intelligence (AI).

Table 1. Characteristics of the physicians that responded to the survey in 2019 and 2022.

	2019 (n = 198)	2022 (n = 93)	p Value
Age
Mean (SD)	45 (11)	44 (10)	p = .75*
Median (IQR)	44 (35–53)	43 (36–51)
Missing n	8	6
Years of experience
Mean (SD)	12 (10)	10 (8)	p = .25*
Median (IQR)	10 (4–19)	9 (3–15)
Missing n	3	0
Gender
Men	46%	32%	p = .06**
Women	53%	67%
Undeclared	1%	1%
Missing n	0	1
Workplace
Public %	78%	80%	p = .70**
Private %	22%	20%
Missing n	5	2
Education in e-health
Yes		15%
Missing n		0
Staying updated
Mean (SD)	4 (1)	4 (1)	p = .53***
Median (IQR)	4 (3–6)	4 (3–5)
Missing n	17	12

Notes: 2019 data has previously been published in the article ‘Swedish primary care physicians’ intentions to use telemedicine: A survey using a new questionnaire–physician attitudes and intentions to use telemedicine (PAIT)’, Pikkemaat et al. [5] and are here presented for comparison with the 2022 results. Acronyms: SD stands for standard deviation; IQR stands for interquartile range. n stands for number, * stands for statistical test Mann–Whitney’s test, ** Chi-2 test and *** Student’s T-test.

Table 2. Physicians’ self-reported use of digital tools in 2019 and 2022.

	2019 (n = 198)	2022 (n = 93)	p Value*
Physical visits
Median (IQR)	7 (6–7)	7 (6–7)	p = .75
Missing n	1	1
Telephone consultations
Median (IQR)	6 (5–7)	7 (6–7)	p = .24
Missing n	2	1
Video consultations
Median (IQR)	1 (1–1)	1 (1–2)	p = .03
Missing n	4	3
National e-service portal www.1177.se
Median (IQR)	4 (3–5)	4 (3–5)	p = .07
Missing n	3	2
Letter
Median (IQR)	5 (4–7)	6 (4–7)	p = .23
Missing n	1	1
E-mail
Median (IQR)	1 (1–2)	1 (1–1)	p = .10
Missing n	4	3
Chat
Median (IQR))	1 (1–1)	1 (1–1)	p = .66
Missing n	4	3
Chronic disease monitoring
Median (IQR)	1 (1–2)	1 (1–3)	p = .34
Missing n	6	3
Artificial intelligence
Median (IQR)	1 (1–2)	1 (1–2)	p = .48
Missing n	8	4

Notes: Responses were provided as Likert-items with values from ‘1’ to ‘7’ where ‘1’ represented ‘not at all’ and ‘7’ ‘to a great extent’. In case of multiple answers to the same question, this was reported as a missing value. 2019 data has previously been published in the article ‘Swedish primary care physicians’ intentions to use telemedicine: A survey using a new questionnaire–physician attitudes and intentions to use telemedicine (PAIT)’, Pikkemaat et al. [5] and are here presented for comparison with the 2022 results. Acronyms: SD stands for standard deviation, IQR stands for interquartile range, n stands for number and * stands for statistical test Mann–Whitney’s test. Statistically significant changes (in bold when applicable) were defined as a p value equal to or less than .05.

Table 3. The behavioral intentions and predictors to use digital consultations for physicians in 2019 and 2022.

Predictor or intention	2019 (n = 198)	2022 (n = 93)	p Value*
Attitudes text digital consultations
Mean (SD)	4.2 (1.6)	4.4 (1.5)	p = .39
Median (IQR)	4.3 (3.3–5.5)	4.5 (3.5–5.8)
Missing n	19	1
‘I do not know’	92%	100%
No response	8%	0%
Attitudes video digital consultations
Mean (SD)	3.9 (1.6)	3.8 (1.6)	p = .64
Median (IQR)	4.0 (3.0–5.0)	4.0 (2.5–5.0)
Missing n	29	33
‘I do not know’	78%	96%
No response	22%	4%
Subjective norms digital consultations
Mean (SD)	4.1 (1.8)	3.9 (1.7)	p = .30
Median (IQR)	4.0 (2.5–5.5)	4.0 (3.0–5.5)
Missing n	7	2
‘I do not know’	100%	100%
No response	0%	0%
PBC digital consultations
Mean (SD)	4.8 (1.2)	5.0 (1.2)	p = .16
Median (IQR)	5.0 (4.0–5.7)	5.0 (4.3–6.0)
Missing n	9	0
‘I do not know’	100%	0%
No response	0%	0%
Intentions digital consultations
Mean (SD)	5.0 (1.8)	5.0 (1.8)	p = .91
Median (IQR)	5.5 (4.0–6.5)	5.5 (3.5–7.0)
Missing n	5	0
‘I do not know’	100%	0%
No response	0%	0%

Notes: Responses were provided as Likert-items with values from ‘1’ to ‘7’ where ‘1’ represented ‘not at all’, ‘harmful’, ‘bad’, ‘unpleasant’ or ‘useless’ and ‘7’ ‘to a great extent’, ‘prosperous’, ‘good’, ‘pleasant’ or ‘useful’. The separate Likert-items were summed and then divided by the number of items to a mean Likert scale consisting of two to four Likert-items for behavioral intentions and predictors to use digital tools. The respondents were, for specific questions, able to answer, ‘I do not know’. In case of multiple answers to the same question, this was reported as a missing value. Mann–Whitney’s test used because of non-normally distributed data. 2019 data has previously been published in the article ‘Swedish primary care physicians’ intentions to use telemedicine: A survey using a new questionnaire–physician attitudes and intentions to use telemedicine (PAIT)’, Pikkemaat et al. [5] and are here presented for comparison with the 2022 results. Acronyms: SD stands for standard deviation, IQR stands for interquartile range. n stands for number, PBC stands for perceived behavioral control, * stands for statistical test Mann–Whitney’s test. Missing is the total number of ‘I do not know’ and ‘no response’, which were not included in the analysis. Below missing is the percentage of ‘I do not know’ and ‘no response’.

Table 4. The partially adjusted and the adjusted regression analysis data for intentions to use digital consultations.

Partially adjusted regression analysis for intentions to use digital consultations
	2019 (n = 153)			2022 (n = 58)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes text	0.12	.20*	−0.06–0.30*	0.39	0.08*	−0.05–0.83*
Attitudes video	0.38	<.01*	0.20–0.57*	0.34	0.08*	−0.04–0.72*
Subjective norms	0.16	.02*	0.06–0.27*	0.36	0.01*	0.09–0.62*
PBC	0.55	<.01*	0.33–0.78*	0.57	0.02*	0.83–1.07*
Adjusted R square	0.53			0.57
Adjusted regression analysis for intentions to use digital consultations
	2019 (n = 143)			2022 (n = 54)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes text	0.12	.27*	−0.09–0.33*	0.36	.19*	−0.19–0.91*
Attitudes video	0.42	<.01*	0.21–0.63*	0.23	.26*	−0.18–0.64*
Subjective norms	0.19	.02*	0.07–0.31*	0.36	.02*	0.06–0.65*
PBC	0.52	<.01*	0.26–0.78*	0.75	<.01*	0.27–1.23*
Gender	0.16	.41*	−0.22–0.53	−0.04	.91	−0.79–0.70
Age	−0.02	.35*	−0.05–0.02	−0.05	.09	−0.12–0.01
Years of experience	0.01	.67*	−0.03–0.05	0.06	.12	−0.02–0.14
Workplace (public/private)	−0.19	.42*	−0.65–0.28	−0.13	.79	−1.10–0.85
Adjusted R square	0.54			0.56

Notes: 2019 data has previously been published in the article ‘Swedish primary care physicians’ intentions to use telemedicine: A survey using a new questionnaire–physician attitudes and intentions to use telemedicine (PAIT)’, Pikkemaat et al. [5] and are here presented to be able to compare to the results from 2022. There were two iterations of regression analysis, one partially adjusted iteration and an adjusted iteration. Acronyms: U-Beta stands for unstandardized beta value, p stands for p value, CI stands for 95% confidence interval, * stands for that robust standard error was used due to heteroskedasticity, PBC stands for perceived behavioral control. Statistically significant predictors (in bold when applicable) were defined as a p value equal to or less than .05.

Table

Variable	Survey-item
Attitudes	Using a digital contact (as video consultation) in patient care is:
	1	2	3	4	5	6	7	I don’t know
	Harmful						Beneficial
	Monitoring chronical disease with digital tools is: good (1)/bad (7)
	Using AI in patient care is (for me): pleasant (1)/unpleasant (7)
Subjective norms	I feel under social pressure to use digital contact ways in patient care: strongly disagree (1)/strongly agree (7)
	It is expected of me to use digital tools for monitoring chronic disease to a larger extent. Strongly disagree (1)/strongly agree (7)
Perceived behavioral control (controllability)	The decision to use more digital tools for monitoring of chronic disease is beyond my control. Strongly disagree (1)/strongly agree (7)
Perceived behavioral control (self-efficacy)	Using AI is for me: difficult (1)/easy (7)
Behavioral intention	I plan to use digital contact ways to a larger extent in the future, if it is clinically adequate for the patient. Strongly disagree (1)/strongly agree (7)
	I want to use more AI in the future, if it is clinically adequate for the patient. Strongly disagree (1)/strongly agree (7)

Table

Predictor or intention	2019 (n = 198)	2022 (n = 93)	p Value*
Attitudes chronic disease monitoring
Mean (SD)	5.1 (1.4)	4.9 (1.2)
Median (IQR)	5.3 (4.0–6.0)	5.0 (4.0–6.0)	p = .23
Missing n	39	23
‘I do not know’	90%	96%
No response	10%	4%
Subjective norms chronic disease monitoring
Mean (SD)	3.0 (1.8)	2.3 (1.7)
Median (IQR)	3.0 (1.5–4.5)	1.5 (1.0–3.6)	p < .01
Missing n	22	16
‘I do not know’	95%	97%
No response	5%	3%
PBC chronic disease monitoring
Mean (SD)	5.1 (1.2)	5.4 (1.3)
Median (IQR)	5.3 (4.6–6.0)	5.5 (4.3–6.5)	p = .13
Missing n	27	18
‘I do not know’	95%	100%
No response	5%	0%
Intentions chronic disease monitoring
Mean (SD)	5.1 (1.7)	5.1 (1.5)
Median (IQR)	5.3(4.5–7.0)	5.3 (4.0–6.1)	p = .60
Missing n	17	13
‘I do not know’	100%	100%
No response	0%	0%
Attitudes artificial intelligence
Mean (SD)	4.5 (1.5)	3.3 (1.5)
Median (IQR)	4.5 (3.9–5.8)	3.3 (2.0–4.3)	p < .01
Missing n	71	35
‘I do not know’	91%	100%
No response	9%	0%
Subjective norms artificial intelligence
Mean (SD)	2.7 (1.8)	2.2 (1.7)
Median (IQR)	2.0 (1.0–4.0)	1.0 (1.0–4.0)	p = .02
Missing n	45	19
‘I do not know’	90%	97%
No response	10%	3%
PBC artificial intelligence
Mean (SD)	4.9 (1.4)	4.3 (1.3)
Median (IQR)	5.0 (4.0–6.0)	4.0 (3.4–5.9)	p < .01
Missing n	57	29
‘I do not know’	94%	99%
No response	6%	1%
Intentions artificial intelligence
Mean (SD)	4.8 (1.8)	4.2 (2.0)
Median (IQR)	5.0 (4.0–6.5)	4.0 (2.0–5.4)	p = .03
Missing n	37	27
‘I do not know’	92%	100%
No response	8%	0%

Table

Partially adjusted regression analysis for intentions to use chronic disease monitoring tools
	2019 (n = 131)			2022 (n = 57)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes	0.54	<.01	0.37–0.71	0.71	<.01	0.41–1.00
Subjective norms	0.10	.10	−0.02–0.22	0.12	.18	−0.06–0.30
PBC	0.40	<.01	0.18–0.62	0.09	.52	−0.20–0.38
Adjusted R square	0.48			0.46
Adjusted regression analysis for intentions to use chronic disease monitoring tools
	2019 (n = 122)			2022 (n = 51)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes	0.47	<.01	0.28–0.66	0.65	<.01	0.30–1.00
Subjective norms	0.09	.16	−0.04–0.22	0.12	.23	−0.08–0.32
PBC	0.47	<.01	0.22–0.72	0.20	.27	−0.16–0.59
Gender	−0.01	.98	−0.44–0.43	−0.09	.77	−0.73–0.54
Age	−0.02	.40	−0.05–0.02	0.00	.99	−0.05–0.05
Years of experience	0.01	.70	−0.03–0.05	0.00	.94	−0.06–0.06
Workplace (public/private)	−0.18	.50	−0.69–0.34	0.73	.20	−0.40–1.85
Adjusted R square	0.46			0.40
Partially adjusted regression analysis for intentions to use AI tools
	2019 (n = 100)			2022 (n = 47)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes	0.44	<.01	0.24–0.63	0.72	<.01	0.40–1.03
Subjective norms	0.02	.75	−0.10–0.14	0.09	.37	−0.11–0.29
PBC	0.51	<.01	0.29–0.72	0.48	<.01	0.14–0.82
Adjusted R square	0.56			0.64
Adjusted regression analysis for intentions to use AI tools
	2019 (n = 94)			2022 (n = 43)
Predictors	U-Beta	p	CI (95%)	U-Beta	p	CI (95%)
Attitudes	0.39	<.01	0.17–0.61	0.77	<.01	0.44–1.11
Subjective norms	0.00	.96	−0.13–0.13	0.18	.13	−0.05–0.40
PBC	0.54	<.01	0.30–0.77	0.44	.02	0.08–0.81
Gender	0.01	.97	−0.43–0.45	0.49	.21	−0.30–1.26
Age	−0.01	.54	−0.05–0.03	0.00	.93	−0.06–0.05
Years of experience	0.01	.55	−0.03–0.06	−0.02	.54	−0.01–0.05
Workplace (public/private)	0.22	.40	−0.30–0.75	0.21	.75	−1.09–1.50
Adjusted R square	0.51			0.64

Ajzen I. The theory of planned behavior. Org Behav Hum Decis Process. 1991;50(2):179–211. doi: 10.1016/0749-5978(91)90020-T.

 Web of Science ®Google Scholar

Pikkemaat M, Thulesius H, Milos Nymberg V. Swedish primary care physicians’ intentions to use telemedicine: a survey using a new questionnaire–physician attitudes and intentions to use telemedicine (PAIT). Int J Gen Med. 2021;14:3445–3455. doi: 10.2147/IJGM.S319497.

 PubMed Web of Science ®Google Scholar