Knowledge and attitudes among high school students toward medical misinformation and their willingness to pursue medical-related majors

Abstract

With the rise of the Internet and social media, misinformation has become more prevalent and easier to access. This study aimed to assess the knowledge and attitudes among high school students toward medical misinformation and their willingness to pursue medical-related majors. This cross-sectional study enrolled high school students in China between December 2022 and January 2023. A total of 2002 valid questionnaires were collected. The knowledge, attitude, and willingness scores were 7.81 ± 3.02 (possible range: 0–12), 34.73 ± 4.56 (possible range: 9–45), and 30.98 ± 7.50 (possible range: 9–45), respectively. The multivariable analysis showed that attitude score (β = 0.803, 95% CI: 0.730–0.875, P < .001), age beyond 19 (β = 2.558, 95% CI: 0.596–4.520, P = .011), nontechnical high school (β = −1.549–3.358, all types P < .01), Grade III (β = 1.598, 95% CI: 0.752–2.444, P < .001), and having immediate family member engaged in medicine (β = 1.814, 95% CI: 0.736–2.892, P = .001) were associated with a higher willingness score. Additionally, structural equation modeling indicated that knowledge was associated with attitude (β = 0.727, P < .001), and attitude was associated with willingness (β = 0.806, P < .001). High school students demonstrated limited knowledge and low confidence in dealing with medical misinformation, along with only moderate willingness to pursue medical-related majors, despite generally recognizing the importance of combating misinformation. This study underscores the need to enhance high school students’ ability to distinguish medical misinformation and increase their willingness to pursue medical-related majors.

1. Introduction

Medical misinformation, also known as false or misleading medical information, encompasses any content that deviates from facts or scientific evidence.^[1] The COVID-19 pandemic has magnified the urgency of addressing misinformation in healthcare. During the pandemic, various media outlets have resorted to sensationalized reporting to attract attention and maximize viewership.^[1,2] Such increased focus on capturing audiences’ attention has inadvertently facilitated the spread of inaccurate medical information, hindering public health efforts and exacerbating the challenges faced by healthcare providers.

With the rise of the Internet and social media, misinformation has become more prevalent and easier to access. High school students, as an impressionable and information-seeking demographic, are particularly susceptible to the influence of inaccurate medical information. Still, it is a critical population for improving health literacy, as these individuals are at an age when they form attitudes and behaviors that will impact their future health-related willingness.^[3] Unfortunately, due to their underdeveloped ability to critically evaluate and discern the reliability and validity of information, exposure to misinformation poses a serious concern for the healthcare literacy of future adults and their willingness to enter medical professions. Still, exposure to misinformation must be evaluated to design effective interventions to counter it.

According to the World Health Organization (WHO), there is a significant global shortage of healthcare workers, with an estimated deficit of nearly 4.3 million.^[3] The shortage of healthcare professionals in China has also become a critical issue that demands immediate attention.^[4,5] One key factor contributing to the shortage of healthcare workers is declining interest among university students in pursuing medical majors. Notably, a study analyzing the effects of COVID-19 on career and specialty choices among Chinese medical students found that greater exposure to negative pandemic information and greater satisfaction with their major were associated with an increased willingness to become a doctor.^[6] Similarly, a study investigated factors affecting psychological health and career choice among medical students in eastern and western China after the COVID-19 pandemic. It showed that COVID-19 led medical students with psychological problems to a negative attitude toward future employment.^[7] The dissemination of medical misinformation may negatively impact adolescents’ health literacy and their willingness to apply for medical-related majors and careers.^[2,8,9]

Nevertheless, although studies are available on adolescent health literacy,^[10,11] misinformation exposure,^[12,13] and medical career intentions^[14,15] in China, the knowledge and attitudes of high school students regarding medical misinformation and its association with their willingness to pursue medical-related fields remain largely unknown. Therefore, this study aimed to use the knowledge, attitude, and willingness (KAW) structured survey method^[16] to assess the knowledge and attitudes of high school students towards the propagation of medical misinformation and their willingness to pursue medical-related majors, exploring the factors associated with the 3 KAW dimensions.

2. Materials and methods

2.1. Study design and participants

The present cross-sectional study enrolled high school students in China between December 2022 and January 2023. The inclusion criteria were 1) high school students and 2) the ability to complete the questionnaire through the following mediums: paper-based questionnaire, cell phone, or computer. The exclusion criteria were 1) incomplete questionnaire responses or 2) individuals who indicated that they did not plan on pursuing higher education in the future. The study was approved by the Ethics Committee of the Affiliated Hospital of Zunyi Medical University (No. KLL-2022-826), and informed consent forms have been obtained from all students and their parents or guardians.

2.2. Questionnaire

The questionnaire was designed with reference to the literature and the Guidance on Accelerating the Innovative Development of Medical Education.^[17,18] The questionnaire was reviewed and modified by 2 experts: Professor Xiong Zheng-ai (Department of Obstetrics and Gynecology, The Second Affiliated Hospital of Chongqing Medical University, Professor/Doctoral Supervisor, 31 years of experience, specializes in obstetrics and gynecology, research direction: gynecologic oncology, clinical medical education) and Deputy Chief Physician Chen Wenjuan (Department of Gynecologic Oncology, The Affiliated Cancer Hospital of Chongqing University, Deputy Chief Physician, 14 years of experience, specializes in gynecologic oncology, research direction: gynecologic oncology, clinical medical education). The first draft of the questionnaire was pilot-tested on a small scale (n = 40) with a Cronbach α of 0.722 (0.673, 0.773, and 0.799 for knowledge, attitude, and practice, respectively), indicating good internal reliability. A confirmatory factor analysis (Supplemental Digital Content Figure S1, Supplemental Digital Content, https://links.lww.com/MD/R776) was performed. The confirmatory factor analysis was evaluated using the root mean square error of approximation, goodness-of-fit index, incremental fit index, Tucker-Lewis index, and comparative fit index, all of which indicated appropriate model fit (Table S1, Supplemental Digital Content, https://links.lww.com/MD/R777). The Kaiser-Meyer-Olkin test revealed a value of 0.919 (P < .001), suggesting sampling adequacy.

The final version of the questionnaire was in Chinese and consisted of 4 dimensions: demographic characteristics, knowledge, attitudes, and willingness (Supplement Material). The knowledge dimension comprised 12 items, while the attitudes and willingness dimensions included 9 and 10 items, respectively. Each correct answer for the knowledge items received a score of 1 point, while incorrect answers received 0 points, resulting in a possible score range of 0 to 12. The attitude and willingness dimensions were assessed using a 5-point Likert scale, where responses ranged from the most positive option (5 points) to the least positive option (1 point). Item W10 of the willingness dimension, which investigated the “Application ranking for medical-related majors,” was presented by descriptive analysis and did not receive any scores. For W10, the participants could select and rate any item that applied to them. Therefore, the possible score ranges for attitudes and willingness dimensions were 9 to 45.

The data were collected using an online questionnaire hosted on Sojump (http://www.sojump.com). The online questionnaire was distributed through various social media platforms, including WeChat, Internet forums, and web links. This approach aimed to enhance the participation rate and engagement of the high school students in the study. The school grade was the actual grade the students were in.

2.3. Statistical analysis

The sample size was calculated using the formula.

$${\displaystyle \begin{array}{l}{\displaystyle n={\left(\frac{Z_{1-\alpha /2}}{\delta }\right)}^2\times p\times (1-p)}\end{array}}$$

where α was set to 0.05, Z_1-α/2 was 1.96, and p was assumed to be 0.5, resulting in a sample size of 384. Assuming an 80% effective questionnaire recovery rate of 80%, the sample size was 480.

STATA 17.0 (Stata Corporation, College Station) was used for statistical analysis. The continuous variables were expressed as means ± standard deviations, and the categorical variables were expressed as n (%). The continuous variables that conformed to a normal distribution were tested by the t test or ANOVA. Pearson’s correlation was used to analyze the relationships among knowledge, attitude, and willingness. To identify factors associated with the 3 continuous KAW scores, multivariable linear regression models were employed. Univariable liner regression analyses were first performed. Then, all variables with P < .05 were included in the multivariable analyses. Multicollinearity was evaluated using variance inflation factors (VIFs). A VIF <5 indicated the absence of multicollinearity.

A structural equation modeling (SEM) analysis was performed to examine the relationships among KAW dimensions. Based on the Knowledge, Attitude, Practices paradigm and extensions of the Theory of Planned Behavior, we conceptualized willingness to pursue medical-related majors as a proximal intention-like construct primarily shaped by students’ attitudinal evaluations.^[19,20] In this framework, knowledge is treated as a more distal cognitive factor that exerts its influence on willingness mainly through its effect on attitudes rather than directly. This ordering is also consistent with contemporary health literacy models, which view knowledge as one component of broader health literacy that enables, but does not, by itself, determine motivated decision-making and behavioral intentions. Accordingly, the SEM specified paths from knowledge to attitude and from attitude to willingness, and included a direct path from knowledge to willingness to empirically test whether any additional direct effect remained after accounting for attitudes. Given the cross-sectional design, these directional paths reflect theory-driven assumptions and should be interpreted as indicating pattern and strength of associations rather than definitive causal effects. Hence, the SEM was based on the hypotheses that 1) knowledge affected attitude, 2) attitude affected willingness, and 3) knowledge affected willingness. Given that this theory-driven path model is just-identified (i.e., saturated, with zero degrees of freedom), conventional goodness-of-fit indices (e.g., CFI, TLI, RMSEA) are not applicable as the model fit is perfect by definition. Two-sided P < .05 was considered statistically significant.

3. Results

A total of 2002 valid questionnaires were collected. The mean score of knowledge, attitudes, and willingness were 7.81 ± 3.02 (possible range: 0–12; n = 691, 34.52% had scores > 10), 34.73 ± 4.56 (possible range: 9–45; n = 686, 34.27% had scores > 37), and 30.98 ± 7.50 (possible range: 9–45; n = 687, 34.32% had scores > 35), respectively. The knowledge scores varied among individuals of different genders (P < .001), ages (P = .030), residences (P < .001), fathers’ educational backgrounds (P < .001), mothers’ educational backgrounds (P = .017), high school types (P < .001), grades (P = .007), academic rankings (P = .021), and immediate family members engaged in medicine (P = .031). Regarding attitude scores, there were differences among participants based on gender (P = .029), residence (P < .001), fathers’ education background (P = .012), mothers’ education background (P = .001), High School type (P < .001), Grade (P = .003), and academic rankings (P = .001). Additionally, there were differences in willingness scores among participants of different ages (P < .001), mothers’ educational background (P = .014), grades (P < .001), academic rankings (P = .001), and having an immediate family member engaged in medicine (P < .001) (Table 1).

Table 1.

Demographic characteristics and KAP scores.

	N(%)/Mean ± SD	Knowledge score		Attitude score		Willingness score
		Mean ± SD	P	Mean ± SD	P	Mean ± SD	P
Total	2002	7.81 ± 3.02		34.73 ± 4.56		30.98 ± 7.50
Gender			<.001		.029		.68
Male	922 (46.05)	7.51 ± 3.10		34.49 ± 4.86		30.90 ± 7.70
Female	1080 (53.95)	8.12 ± 2.92		34.94 ± 4.29		31.04 ± 7.32
Age (y)	16.55 ± 1.08		.03		.121		<.001
≤15	315 (15.73)	8.22 ± 2.93		34.93 ± 4.30		30.39 ± 7.58
16–18	1627 (81.27)	7.74 ± 3.04		34.66 ± 4.62		30.93 ± 7.48
≥19	60 (3.00)	7.52 ± 2.92		35.78 ± 4.35		35.18 ± 6.12
Residence			<.001		<.001		.066
Rural	1013 (50.60)	7.51 ± 3.10		34.37 ± 4.53		30.67 ± 7.26
Urban	989 (49.40)	8.12 ± 2.92		35.11 ± 4.56		31.29 ± 7.72
Father’s education			<.001		<.012		.091
Middle school and below	1166 (58.24)	7.63 ± 3.08		34.55 ± 4.50		30.64 ± 7.41
High school and technical high school	585 (29.22)	8.04 ± 2.88		34.94 ± 4.63		31.33 ± 7.68
Junior college and bachelor’s degree	227 (11.34)	8.29 ± 2.93		35.35 ± 4.36		31.80 ± 7.46
Master’s degree and above	24 (1.20)	6.54 ± 3.66		32.96 ± 6.68		30.54 ± 6.84
Mother’s education			.017		.001		.014
Middle school and below	1354 (67.63)	7.69 ± 3.07		34.49 ± 4.55		30.70 ± 7.38
High school and technical high school	433 (21.63)	7.94 ± 2.91		35.00 ± 4.61		31.18 ± 7.75
Junior college and bachelor’s degree	202 (10.09)	8.37 ± 2.84		35.81 ± 4.32		32.48 ± 7.39
Master’s degree and above	13 (0.65)	7.31 ± 3.68		34.46 ± 5.68		29.92 ± 9.60
Household per capita income, CNY			.114		.611		.983
<2000	250 (12.49)	7.48 ± 3.29		34.96 ± 4.63		31.12 ± 7.66
2000–5000	883 (44.11)	7.76 ± 2.98		34.60 ± 4.37		30.92 ± 7.41
5000–10,000	607 (30.32)	7.90 ± 2.95		34.75 ± 4.75		30.96 ± 7.43
10,000–20,000	186 (9.29)	8.23 ± 2.98		35.10 ± 4.47		31.20 ± 7.83
>20,000	76 (3.80)	7.74 ± 3.20		34.54 ± 5.24		30.75 ± 7.86
High school type			<.001		<.001		.121
Technical high school	275 (13.74)	5.72 ± 3.44		32.24 ± 4.61		30.53 ± 7.49
Non-key general high school	723 (36.11)	8.04 ± 2.83		34.93 ± 4.47		30.87 ± 7.37
District-level key high school	569 (28.42)	8.06 ± 2.80		35.21 ± 4.45		31.63 ± 7.66
Municipal key high school	329 (16.43)	8.33 ± 2.83		35.35 ± 4.20		30.73 ± 7.52
Provincial key high school	106 (5.29)	8.70 ± 2.65		35.42 ± 4.72		30.08 ± 7.30
Grade			.007		.003		<.001
I	666 (33.27)	8.09 ± 2.92		34.84 ± 4.40		30.26 ± 7.49
II	701 (35.01)	7.57 ± 3.18		34.29 ± 4.55		30.47 ± 7.18
III	635 (31.72)	7.78 ± 2.94		35.12 ± 4.71		32.29 ± 7.68
Grade rank			.021		.001		.035
Top 1%–20%	381 (19.03)	8.20 ± 2.94		35.41 ± 4.58		31.59 ± 7.73
Top 21%–40%	495 (24.73)	7.90 ± 3.05		34.99 ± 4.64		30.98 ± 7.57
Top 41%–60%	533 (26.62)	7.62 ± 3.09		34.40 ± 4.58		31.20 ± 7.45
Top 61%–80%	319 (15.93)	7.77 ± 2.96		34.64 ± 4.38		30.91 ± 7.37
Top 81%–100%	274 (13.69)	7.52 ± 3.00		34.11 ± 4.46		29.76 ± 7.17
Immediate family member engaged in medicine.			.031		.141		<.001
Yes	159 (7.94)	8.31 ± 2.51		35.25 ± 4.69		33.11 ± 7.71
No	1843 (92.06)	7.77 ± 3.06		34.69 ± 4.55		30.79 ± 7.45

CNY = Chinese Yuan, KAP = Knowledge, Attitude, Practices.

Students demonstrated strong theoretical understanding, with high correctness rates on items concerning reliable sources (K2, 87.96%), health literacy initiatives (K1, 81.72%), and the dangers/distortion of medical misinformation (K6, 79.97%). However, their ability to practically identify specific false medical incidents was notably lower, with correctness rates dropping significantly for items assessing real-world examples like the “Mr. Liu incident” (K10, 31.02%), “infants’ proper bowel movements” (K9, 46.40%), and “anal suture in pregnant women” (K8, 48.30%). It highlights a gap between conceptual knowledge and practical application (Table S2, Supplemental Digital Content, https://links.lww.com/MD/R777).

Students overwhelmingly recognized the importance of identifying medical misinformation (A1, ~77% agree/strongly agree) and acknowledged the commendable efforts of medical professionals (A7, ~86% agree/strongly agree) and the damage misinformation causes to institutional credibility (A3, ~80% agree/strongly agree). Despite this recognition, confidence in their own ability to discern fake medical news was much lower (A2, 38.7% confident). Approximately 46.55% expressed a strong aspiration for medical careers (A9) (Table S3, Supplemental Digital Content, https://links.lww.com/MD/R777).

A high proportion (~82%) actively sought medical knowledge (W1). Willingness to pursue clinical/technical medical careers (W7) or recommend medical majors (W9) was moderate (~50% and ~46%, respectively), as was the desire for overseas medical education (W8, ~49%) (Table S4, Supplemental Digital Content, https://links.lww.com/MD/R777).

A correlation analysis was performed to assess the relationship between knowledge, attitudes, and willingness. It was shown that knowledge and attitude were positively correlated (R = 0.482, P < .001), and knowledge and willingness were also positively correlated (R = 0.194, P < .001). Additionally, there was a positive correlation between attitudes and willingness (R = 0.470, P < .001) (Supplemental Digital Content Table S5, Supplemental Digital Content, https://links.lww.com/MD/R777).

No multicollinearity was detected among the variables (Supplemental Digital Content Table S6, Supplemental Digital Content, https://links.lww.com/MD/R777). The multivariable liner regression analysis showed that female gender (vs male, β = 0.619, 95% CI: 0.361–0.877, P < .001), father’s education of master’s degree or above (vs middle school and below, β = −1.973, 95% CI: −3.324–−0.622, P = .004), Non-key General High School (vs technical high school, β = 2.173, 95% CI: 1.763–2.582, P < .001), District-level Key High School (vs technical high school, β = 2.099, 95% CI: 1.670–2.527, P < .001), Municipal Key High School (vs technical high school, β = 2.347, 95% CI: 1.873–2.822, P < .001), Provincial Key High School (vs technical high school, β = 2.970, 95% CI: 2.299–3.641, P < .001), grade III (vs grade I, β = −0.384, 95% CI: −0.759–−0.008, P = .045), grade rank of Top 41%–60% (vs 1%–20%, β = −0.393, 95% CI: −0.777–−0.010, P = .045), and grade rank of Top 81%–100% (vs 1%–20%, β = −0.584, 95% CI: −1.040–0.128, P = .012) were associated with sufficient knowledge (Table 2).

Table 2.

Univariable and multivariable linear regression analysis of knowledge.

	Univariable		Multivariable
	β (95% CI)	P-value	β (95% CI)	P-value
Gender
Male	Ref.		Ref.
Female	0.701 (0.436, 0.965)	<.001	0.619 (0.361, 0.877)	<.001
Age (y)
≤15	Ref.		Ref.
16–18	–0.473 (−0.837, −0.108)	.011	−0.176 (−0.599, 0.247)	.414
≥19	−0.699 (−1.534, 0.135)	.1	−0.238 (−1.105, 0.629)	.59
Residence
Rural	Ref.		Ref.
Urban	0.61 (0.346, 0.874)	<.001	0.246 (−0.036, 0.528)	.087
Father’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.416 (0.116, 0.715)	.007	0.201 (−0.118, 0.52)	.217
Junior college and bachelor’s degree	0.664 (0.235, 1.092)	.002	0.043 (−0.482, 0.569)	.872
Master’s degree and above	−1.085 (−2.304, 0.133)	.081	−1.973 (−3.324, −0.622)	.004
Mother’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.246 (−0.08, 0.573)	.139	−0.059 (−0.411, 0.292)	.74
Junior college and bachelor’s degree	0.68 (0.233, 1.127)	.003	0.116 (−0.432, 0.665)	.677
Master’s degree and above	−0.384 (−2.033, 1.266)	.648	−0.446 (−2.198, 1.305)	.617
Household per capita income, CNY
<2000	Ref.		Ref.
2000–5000	0.286 (−0.138, 0.711)	.186	0.196 (−0.213, 0.604)	.348
5000–10,000	0.424 (−0.022, 0.869)	.062	0.247 (−0.196, 0.689)	.275
10,000–20,000	0.755 (0.181, 1.329)	.01	0.438 (−0.138, 1.014)	.136
>20,000	0.261 (−0.515, 1.037)	.51	0.345 (−0.447, 1.138)	.393
High school type
Technical high school	Ref.		Ref.
Non-key General High School	2.323 (1.919, 2.726)	<.001	2.173 (1.763, 2.582)	<.001
District-level key high school	2.342 (1.923, 2.76)	<.001	2.099 (1.67, 2.527)	<.001
Municipal key high school	2.608 (2.143, 3.074)	<.001	2.347 (1.873, 2.822)	<.001
Provincial key high school	2.978 (2.327, 3.629)	<.001	2.97 (2.299, 3.641)	<.001
Grade
I	Ref.		Ref.
II	−0.515 (−0.835, −0.195)	.002	−0.356 (−0.721, 0.01)	.056
III	−0.307 (−0.636, 0.021)	.066	−0.384 (−0.759, −0.008)	.045
Grade rank
Top 1%–20%	Ref.		Ref.
Top 21%–40%	−0.308 (−0.711, 0.096)	.135	−0.208 (−0.594, 0.179)	.292
Top 41%–60%	−0.582 (−0.979, −0.185)	.004	−0.393 (−0.777, −0.01)	.045
Top 61%–80%	−0.437 (−0.886, 0.013)	.057	−0.352 (−0.786, 0.081)	.111
Top 81%–100%	−0.683 (−1.152, −0.214)	.004	−0.584 (−1.04, −0.128)	.012
Immediate family member engaged in medicine
Yes	0.54 (0.051, 1.03)	0.031	0.385 (−0.092, 0.862)	0.114
No	Ref.		Ref.

CNY = Chinese Yuan.

The knowledge scores (β = 0.682, 95% CI: 0.621–0.743, P < .001), father’s education level of master’s degree or above (vs middle school or below, β = −1.882, 95% CI: −3.698–0.066, P = .042), mother’s junior college and bachelor’s degree (vs middle school or below, β = 1.039, 95% CI: 0.291–1.788, P = .007), Non-key General High School (vs technical high school, β = 0.974, 95% CI: 0.396–1.553, P = .001), District-level Key High School (vs technical high school, β = 1.141, 95% CI: 0.536–1.745, P < .001), Municipal Key High School (vs technical high school, β = 1.171, 95% CI: 0.501–1.842, P = .001), Provincial Key High School (vs technical high school, β = 0.946, 95% CI: 0.004–1.889, P = .049), and grade rank of Top 81%–100% (vs 1%–20%, β = −0.767, 95% CI: −1.397–0.138, P = .017) were associated with positive attitude (Table 3).

Table 3.

Univariable and multivariable logistic regression analysis of attitude.

	Univariable		Multivariable
	β (95% CI)	P-value	β (95% CI)	P-value
Knowledge	0.727 (0.669, 0.785)	<.001	0.682 (0.621, 0.743)	<.001
Gender
Male	Ref.		Ref.
Female	0.445 (0.045, 0.846)	.029	−0.114 (−0.469, 0.242)	.531
Age (y)
≤15	Ref.
16–18	−0.273 (−0.824, 0.277)	.331
≥19	0.853 (−0.406, 2.113)	.184
Residence
Rural	Ref.		Ref.
Urban	0.746 (0.347, 1.145)	<.001	0.063 (−0.323, 0.448)	.749
Father’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.39 (−0.062, 0.843)	.091	−0.194 (−0.633, 0.245)	.387
Junior college and bachelor’s degree	0.8 (0.152, 1.448)	.016	−0.516 (−1.235, 0.203)	.159
Master’s degree and above	−1.59 (−3.431, 0.252)	.091	−1.882 (−3.698, −0.066)	.042
Mother’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.516 (0.023, 1.008)	.04	0.444 (−0.037, 0.924)	.07
Junior college and bachelor’s degree	1.323 (0.65, 1.996)	<.001	1.039 (0.291, 1.788)	.007
Master’s degree and above	−0.027 (−2.512, 2.458)	.983	0.878 (−1.534, 3.291)	.475
Household per capita income, CNY
< 2000	Ref.
2000–5000	−0.355 (−0.996, 0.287)	.278
5000–10,000	−0.21 (−0.882, 0.463)	.541
10,000–20,000	0.146 (−0.721, 1.013)	.741
>20,000	−0.417 (−1.589, 0.756)	.486
High school type
Technical high school	Ref.		Ref.
Non-key general high school	2.689 (2.071, 3.308)	<.001	0.974 (0.396, 1.553)	.001
District-level key high School	2.967 (2.326, 3.609)	<.001	1.141 (0.536, 1.745)	<.001
Municipal key high school	3.11 (2.396, 3.823)	<.001	1.171 (0.501, 1.842)	.001
Provincial key high school	3.175 (2.177, 4.174)	<.001	0.946 (0.004, 1.889)	.049
Grade
I	Ref.		Ref.
II	−0.55 (−1.033, −0.067)	.026	−0.164 (−0.601, 0.272)	.46
III	0.279 (−0.216, 0.774)	.27	0.381 (−0.057, 0.819)	.088
Grade rank
Top 1%–20%	Ref.		Ref.
Top 21%–40%	−0.424 (−1.031, 0.184)	.172	−0.161 (−0.694, 0.372)	.554
Top 41%–60%	−1.014 (−1.612, −0.415)	.001	−0.502 (−1.031, 0.027)	.063
Top 61%–80%	−0.773 (−1.45, −0.096)	.025	−0.4 (−0.997, 0.197)	.189
Top 81%–100%	−1.3 (−2.006, −0.594)	0	−0.767 (−1.397, −0.138)	.017
Immediate family member engaged in medicine
Yes	−0.556 (−1.295, 0.184)	.141
No	Ref.

CNY = Chinese Yuan.

The attitude score (β = 0.803, 95% CI: 0.730–0.875, P < .001), age beyond 19 (vs <15, β = 2.558, 95% CI: 0.596–4.520, P = .011), Non-key General High School (vs technical high school, β = −1.859, 95% CI: −2.811–0.907, P < .001), District-level Key High School (vs technical high school, β = −1.549, 95% CI: −2.539–0.559, P = .002), Municipal Key High School (vs technical high school, β = −2.435, 95% CI: −3.530–1.340, P < .001), Provincial Key High School (vs technical high school, β = −3.358, 95% CI: −4.905–1.811, P < .001), Grade III (vs grade I, β = 1.598, 95% CI: 0.752–2.444, P < .001), and having immediate family member engaged in medicine (vs no, β = 1.814, 95% CI: 0.736–2.892, P = .001) were associated with strong willingness (Table 4).

Table 4.

Univariable and multivariable logistic regression analysis of willingness.

	Univariable		Multivariable
	β (95% CI)	P-value	β (95% CI)	P-value
Knowledge	0.481 (0.374, 0.587)	<.001	−0.039 (−0.15, 0.072)	.492
Attitude	0.772 (0.709, 0.836)	<.001	0.803 (0.73, 0.875)	<.001
Gender
Male	Ref.
Female	0.139 (−0.521, 0.798)	.68
Age (y)
≤15	Ref.
16–18	0.539 (−0.362, 1.439)	.241	0.011 (−0.946, 0.967)	.983
≥19	4.79 (2.729, 6.851)	<.001	2.558 (0.596, 4.52)	.011
Residence
Rural	Ref.
Urban	0.615 (−0.042, 1.272)	.066
Father’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.69 (−0.054, 1.434)	.069	0.417 (−0.293, 1.127)	.25
Junior college and bachelor’s degree	1.159 (0.093, 2.224)	.033	0.387 (−0.781, 1.555)	.516
Master’s degree and above	−0.102 (−3.13, 2.927)	.948	1.893 (−1.084, 4.871)	.212
Mother’s education
Middle school and below	Ref.		Ref.
High school and technical high school	0.484 (−0.326, 1.294)	.241	−0.027 (−0.81, 0.757)	.947
Junior college and bachelor’s degree	1.78 (0.673, 2.886)	.002	0.529 (−0.696, 1.754)	.397
Master’s degree and above	−0.773 (−4.861, 3.316)	.711	−1.183 (−5.145, 2.779)	.558
Household per capita income, CNY
<2000	Ref.
2000–5000	−0.207 (−1.261, 0.847)	.701
5000–10,000	−0.168 (−1.274, 0.937)	.765
10,000–20,000	0.08 (−1.344, 1.505)	.912
>20,000	−0.374 (−2.301, 1.553)	.704
High school type
Technical high school	Ref.		Ref.
Non-key general high school	0.347 (−0.694, 1.387)	.513	−1.859 (−2.811, −0.907)	<.001
District-level key high school	1.1 (0.021, 2.179)	.046	−1.549 (−2.539, −0.559)	.002
Municipal key High School	0.205 (−0.995, 1.405)	.737	−2.435 (−3.53, −1.34)	<.001
Provincial key high school	−0.452 (−2.131, 1.227)	.598	−3.358 (−4.905, −1.811)	<.001
Grade
I	Ref.		Ref.
II	0.214 (−0.576, 1.004)	.595	0.638 (−0.183, 1.459)	.127
III	2.028 (1.218, 2.838)	<.001	1.598 (0.752, 2.444)	<.001
Grade rank
Top 1%–20%	Ref.		Ref.
Top 21%–40%	−0.61 (−1.61, 0.39)	.232	−0.326 (−1.201, 0.548)	.464
Top 41%–60%	−0.387 (−1.372, 0.597)	.441	0.213 (−0.656, 1.082)	.631
Top 61%–80%	−0.679 (−1.793, 0.435)	.232	−0.055 (−1.034, 0.924)	.912
Top 81%–100%	−1.832 (−2.995, −0.67)	.002	−0.718 (−1.752, 0.315)	.173
Immediate family member engaged in medicine
Yes	2.323 (1.112, 3.534)	<.001	1.814 (0.736, 2.892)	.001
No	Ref.

CNY = Chinese Yuan.

Moreover, we further explored participants’ willingness to pursue admission to medical-related majors. Basic Medicine is the most popular major, chosen by 573 students, accounting for 28.62% of the total. Following closely behind are Clinical Medicine, with 385 students (19.23%), and Stomatology, with 247 students (12.34%). These majors are highly sought after among students. In contrast, Public Health and Preventive Medicine is the least popular, with only 46 students (2.30%) opting for it as their first choice. Additionally, Nursing Science (2.50%) and Medical Technology (3.20%) are relatively less popular among high school students. Nursing Science (76.77%), Public Health and Preventive Medicine (67.58%), and Traditional Chinese Pharmacology (66.18%) feature prominently in the “Not applied for” category, indicating a significantly diminished level of interest among the students (Table S7, Supplemental Digital Content, https://links.lww.com/MD/R777).

The SEM (Fig. 1) showed that knowledge was associated with attitude (β = 0.727, P < .001) and attitude was associated with willingness (β = 0.806, P < .001), but the knowledge was not associated with willingness directly (P = .058) (Table S8, Supplemental Digital Content, https://links.lww.com/MD/R777).

Figure 1.

Structural equation modeling analysis.

4. Discussion

Overall, high school students showed limited knowledge and modest applied skills for identifying medical misinformation, coupled with only moderate willingness to pursue medical-related majors, even though they generally endorsed the importance of addressing misinformation and valued medical professionals. This study underscores the need to enhance high school students’ ability to distinguish medical misinformation and increase their willingness to pursue medical-related majors.

The high school students demonstrated sufficient knowledge regarding the definition, characteristics, and properties of medical misinformation. Quite the opposite: the correctness rate was lower for questions that asked them to discern whether the information was real or fake. This finding aligns with recent research indicating that 52% of teenagers struggle to differentiate between true and fake online health messages.^[21] Similarly, a study conducted in China found that although college students heavily relied on online health information for managing their health, they lacked sufficient knowledge or skills to identify misinformation and fake information.^[22] These findings collectively suggest that, while students acknowledge the potential harm associated with false information, their limited knowledge may hinder their ability to distinguish reliable from unreliable sources.

Several factors may contribute to the insufficient knowledge and negative attitudes among high school students regarding medical misinformation. The study found that students residing in urban areas demonstrated higher levels of knowledge about Medical Misinformation than their counterparts in rural areas. This result is similar to the finding that students with urban residency were 5.6 times more knowledgeable about COVID-19 than rural students.^[23] It can be attributed to factors such as greater access to information and resources in urban environments, including a wider range of educational opportunities, exposure to diverse media outlets, and higher connectivity to digital platforms.^[24] Meanwhile, although our study revealed that parents’ education level had some impact on high school students’ knowledge of medical misinformation, no significant difference was observed in the multivariable analysis. It is partly because adolescents nowadays acquire medical knowledge mainly through the Internet, rather than relying solely on their family members for information.^[25–27]

A notable contrast exists between students’ recognition of the importance of identifying medical misinformation (77% agree/strongly agree) and their confidence in doing so (only 38.7% are confident). It indicates a gap between perceived importance and self-assessed ability. The discrepancy raises questions about the factors contributing to students’ lack of confidence in discerning medical misinformation. It could be due to various reasons, such as the complexity of medical information, the prevalence of misleading content, or a lack of knowledge and critical-thinking skills necessary to evaluate information accurately.^[28,29] Exploring these factors and addressing them through educational initiatives and critical-thinking training could help bridge this gap and empower students to become more effective consumers of medical information.

The present study demonstrated that only 14.34% of high school students fully confirmed their willingness to pursue medical majors, consistent with previous findings. A previous study of high school students in 4 Chinese provinces showed that only 5.6% of senior high school students clearly intended to study medicine.^[30] It was also reported that medical careers are becoming less attractive to academically talented students, with only 20.60% of students considering it a desirable career option.^[31] While the percentage may vary by region, sample size, and other factors, it is indisputable that the medical profession is becoming less appealing to high school students. Meanwhile, immediate family members in medical professions positively influence perceptions of healthcare careers and motivate teenagers to pursue them, consistent with previous findings.^[32,33] Such exposure may also provide them with practical knowledge and valuable insights into the profession, further enhancing their interest in this field.^[34]

The popularity of different medical majors among students indicates the current social recognition of medical professions and employment prospects. For instance, the popularity of Basic Medicine and Clinical Medicine in this study could be due to their perceived status and employment opportunities. In contrast, the low popularity of Nursing Science, Public Health, and Preventive Medicine may indicate that these fields are perceived as less prestigious or offer lower employment prospects.^[35] Universities should promote these less popular majors and provide students with more information about their potential benefits and career prospects. Comprehensive guidance and counseling should also be provided to help students make informed decisions about their future careers. Nevertheless, these preferences may change in the future due to factors such as changes in the job market, advances in technology, and shifts in social attitudes.^[36,37]

In the SEM, knowledge significantly predicted attitudes, and attitudes in turn showed a strong association with willingness to pursue medical-related majors, whereas the direct path from knowledge to willingness was not statistically significant. Taken together, these results indicate that knowledge about medical misinformation is associated with a willingness to pursue medical-related majors, mainly indirectly through its influence on students’ attitudes rather than via a strong direct pathway. This indirect-only pattern aligns with extended applications of the Theory of Planned Behavior and health literacy models, in which informational factors, such as knowledge, typically shape behavioral intentions through attitudinal and related motivational constructs rather than exerting large, independent, direct effects.^[19] In the present context, willingness to pursue medical-related majors can be viewed as an intention-like construct, whereas knowledge about medical misinformation serves as a distal cognitive resource that helps shape students’ attitudinal evaluations of medical misinformation and medical careers. The findings are also consistent with integrated health literacy models that define health literacy as encompassing people’s knowledge, motivation, and competencies to access, understand, appraise, and apply health information to make informed health-related decisions. Within these frameworks, knowledge is necessary but insufficient on its own; motivational components, including attitudes toward information and health-related choices, are key proximal determinants of behavioral intentions.^[20,38] The SEM therefore aligns with the view that high school students who possess greater knowledge about medical misinformation may develop more favorable and confident attitudes, which, in turn, are associated with a stronger willingness to pursue medical-related majors, rather than knowledge exerting a direct effect on willingness independent of attitudes. At the same time, the cross-sectional design precludes firm causal inference, so these directional pathways should be interpreted as theory-informed patterns of association that warrant confirmation in longitudinal research. Accordingly, our conclusions focus on the importance of enhancing both students’ knowledge and their attitudes toward medical misinformation, rather than implying that knowledge alone is sufficient to directly increase willingness to pursue medical-related majors.

From a practical perspective, integrating structured media and misinformation literacy modules into high school curricula could help students move from general awareness to more confident, applied evaluation of health information. Such modules might include analyzing real online rumors, guided practice in checking sources, and discussing how misinformation spreads on social media. Given the lower knowledge scores among students from rural areas and technical high schools, targeted programs for these groups (such as workshops delivered through local schools or youth centers) may be particularly important for reducing inequities in misinformation literacy. In addition, collaboration between the education and health sectors to co-develop teaching materials (e.g., case-based exercises, digital media literacy toolkits) could provide teachers with concrete resources to implement these skills in the classroom.

This study has some limitations. The study was conducted in specific regions of China, which may limit the generalizability of the results to other areas within the country. Additionally, these limitations may limit the applicability of the findings to other countries with distinct educational systems and societal characteristics. There is a possibility of a center effect if 2 or more students were from the same center, but the previous study suggested that there is no need to consider the center effect in Knowledge, Attitude, Practices studies.^[39] Secondly, our questionnaire did not evaluate the sources of health information-seeking, which could have provided insight into ways to raise awareness of medical misinformation among high school students. Thirdly, it was a cross-sectional study, and causality could not be examined. Still, a SEM analysis was performed to obtain causality estimates, but interpreting these results must be done cautiously, as causality is statistically inferred from predefined hypotheses rather than directly observed.^[40,41] Fourthly, the questionnaire was disseminated through social media platforms, favoring students who were interested in social media, had better internet access, higher digital literacy, or greater academic motivation. In addition, because the questionnaire was disseminated through social media and did not ask about the participant’s school (for confidentiality reasons), there is no way to determine the geographical coverage or the number and characteristics of the schools involved. Finally, it must be stressed that differences in educational systems, media environments, and healthcare perceptions limit external validity and generalizability.

In conclusion, these findings suggest that interventions should not only increase students’ factual knowledge but also strengthen their confidence and practical skills to evaluate medical information, thereby supporting more informed consideration of medical-related careers. The findings suggest that further education and intervention programs are necessary to enhance high school students’ ability to distinguish medical misinformation and promote their interest in pursuing medical-related careers.

Author contributions

Data curation: Yingge Mou, Guoyuan Li, Xiaosong Li, Yuanzhi Wu, Bing Yang, Hong Zhao.

Investigation: Yingge Mou, Guoyuan Li, Yuanzhi Wu, Bing Yang, Hong Zhao.

Project administration: Yingge Mou, Zhiliang Wang, Guoyuan Li, Yuanzhi Wu.

Validation: Yingge Mou, Zhiliang Wang, Xiaosong Li, Hong Zhao.

Conceptualization: Zhiliang Wang, Ben Wang.

Formal analysis: Zhiliang Wang, Tenghua Yu, Ben Wang.

Visualization: Zhiliang Wang, Guoyuan Li, Xiaosong Li, Ben Wang, Yuanzhi Wu, Bing Yang, Hong Zhao.

Methodology: Xiaosong Li, Ben Wang, Bing Yang.

Software: Xiaosong Li, Tenghua Yu.

Supervision: Ben Wang, Bing Yang.

Resources: Hong Zhao.

Writing – review & editing: Zhiliang Wang, Guoyuan Li, Xiaosong Li, Ben Wang, Yuanzhi Wu.

Writing – original draft: Tenghua Yu.