CoChat
Sign Up
Cited by 29Digital Mental Health InterventionsHealthcare professionals’ stress and burnoutStress and Burnout Research

Development of an artificial intelligence-based tailored mobile intervention for nurse burnout: single-arm trial

Open access
Cho, Aram, Cha, Chiyoung, Baek, Gumhee·2024·Journal of Medical Internet Research
DOI 10.2196/54029PMC PMC11226930

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Background: Nurse burnout leads to an increase in turnover, which is a serious problem in the health care system. Although there is ample evidence of nurse burnout, interventions developed in previous studies were general and did not consider specific burnout dimensions and individual characteristics. Objective: The objectives of this study were to develop and optimize the first tailored mobile intervention for nurse burnout, which recommends programs based on artificial intelligence (AI) algorithms, and to test its usability, effectiveness, and satisfaction. Methods: In this study, an AI-based mobile intervention, Nurse Healing Space, was developed to provide tailored programs for nurse burnout. The 4-week program included mindfulness meditation, laughter therapy, storytelling, reflective writing, and acceptance and commitment therapy. The AI algorithm recommended one of these programs to participants by calculating similarity through a pretest consisting of participants’ demographics, research variables, and burnout dimension scores measured with the Copenhagen Burnout Inventory. After completing a 4-week program, burnout, job stress, stress response using the Stress Response Inventory Modified Form, the usability of the app, coping strategy by the coping strategy indicator, and program satisfaction (1: very dissatisfied; 5: very satisfied) were measured. The AI recognized the recommended program as effective if the user’s burnout score reduced after the 2-week program and updated the algorithm accordingly. After a pilot test (n=10), AI optimization was performed (n=300). A paired 2-tailed t test, ANOVA, and the Spearman correlation were used to test the effect of the intervention and algorithm optimization. Results: Nurse Healing Space was implemented as a mobile app equipped with a system that recommended 1 program out of 4 based on similarity between users through AI. The AI algorithm worked well in matching the program recommended to participants who were most similar using valid data. Users were satisfied with the convenience and visual quality but were dissatisfied with the absence of notifications and inability to customize the program. The overall usability score of the app was 3.4 out of 5 points. Nurses’ burnout scores decreased significantly after the completion of the first 2-week program (t=7.012; P<.001) and reduced further after the second 2-week program (t=2.811; P=.01). After completing the Nurse Healing Space program, job stress (t=6.765; P<.001) and stress responses (t=5.864; P<.001) decreased significantly. During the second 2-week program, the burnout level reduced in the order of participation (r=–0.138; P=.04). User satisfaction increased for both the first (F=3.493; P=.03) and second programs (F=3.911; P=.02). Conclusions: This program effectively reduced burnout, job stress, and stress responses. Nurse managers were able to prevent nurses from resigning and maintain the quality of medical services using this AI-based program to provide tailored interventions for nurse burnout. Thus, this app could improve qualitative health care, increase employee satisfaction, reduce costs, and ultimately improve the efficiency of the health care system.

AI snapshot

Written from the paper’s full text.

TL;DR

This single-arm trial developed and optimized Nurse Healing Space, an AI-based mobile intervention that recommends tailored burnout-reduction programs to nurses based on user similarity. Among 300 completers, burnout, job stress, and stress responses decreased significantly, while coping strategy did not change significantly and usability was rated 3.4 out of 5.

Key findings

  • Burnout decreased significantly from 62.36 (SD 18.00) at pretest to 55.22 (SD 18.79) at posttest 1 (t=7.012; P<.001) and to 52.70 (SD 20.15) at posttest 2 (t=8.692; P<.001); the posttest 1 to posttest 2 reduction was also significant (t=2.811; P=.01).
  • Job stress decreased from 88.39 (SD 12.65) to 81.85 (SD 17.34) (t=6.765; P<.001), and stress response decreased from 42.00 (SD 17.95) to 35.56 (SD 18.39) (t=5.864; P<.001); coping strategy did not change significantly (70.25 to 70.97; t=–1.367; P=.17).
  • Of 600 total data points, 45.7% (274/600) were selected as valid data; laughter therapy accounted for the largest proportion of valid data (32.8%), followed by mindfulness meditation (28.1%), ACT (20.4%), and storytelling/reflective writing (18.6%).
  • Overall app usability was 3.4 out of 5; satisfaction increased across participant groups for program 1 (F=3.493; P=.03) and program 2 (F=3.911; P=.02), with storytelling/reflective writing receiving the highest satisfaction score (4.69).

Methods

  • Single-arm trial: the app was developed December 2021–July 2022, pilot-tested in August 2022 (n=10), then optimized with 300 nurses recruited from two web-based nursing communities between September 2022 and February 2023.
  • Nurse Healing Space delivered two sequential 2-week programs over 4 weeks; each program had 6 sessions of 10–15 minutes, and the AI recommended one of four programs (mindfulness meditation, laughter therapy, storytelling/reflective writing, ACT) based on similarity scores from demographics, job stress, stress response, coping strategy, and Copenhagen Burnout Inventory dimensions.
  • Outcomes included burnout (CBI), job stress, stress response, coping strategy, app usability, and satisfaction; analyses used paired 2-tailed t tests, ANOVA with Scheffe post hoc tests, and Spearman correlations.

Limitations

  • The paper states that the 300-user sample is not truly representative of the entire population, so results are not generalizable, and the absence of a control group restricts causal conclusions about intervention effects.
  • The paper states that algorithm optimization is an ongoing process, even though satisfaction increased as more users participated.
  • Evident from the text: the pilot test had only 10 completers although the authors cite a recommended minimum of 12, and recruitment occurred through web-based nursing communities, which may introduce self-selection bias.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.2196/54029

Cites

68 of 68 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 36 cited sources checked · 0 supported · 0 flagged

  • Findingown data

    Nurse Healing Space was implemented as a mobile app equipped with a system that recommended 1 of 4 programs based on similarity between users through AI.

    “Nurse Healing Space was implemented as a mobile app equipped with a system that recommended 1 program out of 4 based on similarity between users through AI.”
    Quote in the paperSubstantial claim
  • Findingown data

    Of the total data, 45.7% (274/600) were selected as valid data.

    “Of the total data, 45.7% (274/600) were selected as valid data.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: A reduction of at least 5 points in burnout dimension scores after program completion indicates that the program effectively lowered burnout.
  • Findingown data

    Reduction in work-related burnout from posttest 1 to posttest 2 was statistically significant (t=2.306; P=.02).

    “The reduction in work-related burnout from posttest 1 to posttest 2 was also statistically significant (t=2.306; P=.02).”
    Quote in the paperSubstantial claim
  • Findingown data

    Nurses' burnout scores decreased significantly after completion of the first 2-week program (t=7.012; P<.001).

    “Nurses’ burnout scores decreased significantly after the completion of the first 2-week program (t=7.012; P<.001)”
    Quote in the paperSubstantial claim
  • Findingown data

    Reduction in personal burnout from posttest 1 to posttest 2 was statistically significant (t=2.597; P=.01).

    “In addition, the reduction in personal burnout from posttest 1 to posttest 2 was statistically significant (t=2.597; P=.01).”
    Quote in the paperSubstantial claim
  • May combine two claims. 34
  • Classified as probably not a substantial claim. 25
  • The quote only partly states this claim. 1

  • May combine two claims.

    Findingown data

    User satisfaction increased for both the first (F=3.493; P=.03) and second programs (F=3.911; P=.02).

    “User satisfaction increased for both the first (F=3.493; P=.03) and second programs (F=3.911; P=.02).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The program completed by the participant with the lowest similarity was laughter therapy, which was recommended for the 300th participant.

    “It was confirmed that the program completed by the participant with the lowest similarity was laughter therapy, which was recommended for the 300th participant.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Laughter therapy accounted for the largest proportion of valid data (32.8%), followed by mindfulness meditation (28.1%), ACT (20.4%), and storytelling and reflective writing (18.6%).

    “In terms of programs, laughter therapy accounted for the largest proportion of valid data, with 32.8%, followed by 28.1% for mindfulness meditation, 20.4% for ACT, and 18.6% for storytelling and reflective writing.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: A reduction of at least 5 points in burnout dimension scores after program completion indicates that the program effectively lowered burnout.

  • May combine two claims.

    Findingown data

    Work-related burnout decreased statistically from 22.22 (SD 6.39) at pretest to 19.83 (SD 6.49) at posttest 1 (t=6.816; P<.001) and 19.06 (SD 6.68) at posttest 2 (t=8.286; P<.001).

    “Work-related burnout decreased statistically from 22.22 (SD 6.39) at pretest to 19.83 (SD 6.49) at posttest 1 (t=6.816; P<.001) and 19.06 (SD 6.68) at posttest 2 (t=8.286; P<.001).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Of participants, 97.7% (293/300) were female, 87% (261/300) were staff nurses, and 62.3% (187/300) worked in hospitals with more than 500 beds.

    “Of these, 97.7% (293/300) were female, 87% (261/300) were staff nurses, and 62.3% (187/300) worked in hospitals with more than 500 beds (Table 3).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The mean age of participants was 32.8 (SD 5.75) years.

    “The mean age of the participants was 32.8 (SD 5.75) years.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Users scored ease of use, good visual design, and high quality of content graphics more than 4 out of 5.

    “Users scored ease of use, good visual design, and high quality of content graphics more than 4 out of 5.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Users were satisfied with the convenience and visual quality of the app but were dissatisfied with the absence of notifications and inability to customize the program.

    “Users were satisfied with the convenience and visual quality but were dissatisfied with the absence of notifications and inability to customize the program.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Burnout decreased from 62.36 (SD 18.0) at pretest to 55.22 (SD 18.79) at posttest 1 (t=7.012; P<.001) and 52.70 (SD 20.15) at posttest 2 (t=8.692; P<.001).

    “The decrease in burnout was statistically significant, from 62.36 (SD 18.0) at pretest to 55.22 (SD 18.79) at posttest 1 (t=7.012; P<.001) and 52.70 (SD 20.15) at posttest 2 (t=8.692; P<.001).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    Users felt most uncomfortable with the absence of reminders and notification functions and the inability to interact with the app.

    “However, users felt most uncomfortable with the absence of reminders and notification functions and the inability to interact with the app.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Personal burnout decreased statistically from 20.40 (SD 6.39) at pretest to 18.15 (SD 6.38) at posttest 1 (t=6.134; P<.001) and 17.30 (SD 7.33) at posttest 2 (t=7.535; P<.001).

    “Personal burnout decreased statistically from 20.40 (SD 6.39) at pretest to 18.15 (SD 6.38) at posttest 1 (t=6.134; P<.001) and 17.30 (SD 7.33) at posttest 2 (t=7.535; P<.001).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The highest burnout subtype for the most similar participant was work-related burnout, and the participant's data were classified as valid data because of the effect of laughter therapy.

    “The highest burnout subtype for the most similar participant was work-related burnout, and the participant’s data were classified as valid data because of the effect of laughter therapy.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The calculated similarity with the participant was 1.17.

    “The calculated similarity with the participant was 1.17.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Client-related burnout decreased statistically from 19.72 (SD 6.63) at pretest to 17.22 (SD 6.91) at posttest 1 (t=6.439; P<.001) and 16.32 (SD 7.15) at posttest 2 (t=8.329; P<.001).

    “Client-related burnout decreased statistically from 19.72 (SD 6.63) at pretest to 17.22 (SD 6.91) at posttest 1 (t=6.439; P<.001) and 16.32 (SD 7.15) at posttest 2 (t=8.329; P<.001).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    Personal burnout was the highest subdimension with 57.4% of valid data, work-related burnout was next with 25.9%, followed by client burnout with 40% including duplicates.

    “Personal burnout was the highest subdimension with 57.4% of valid data and work-related burnout was the next highest subdimension with 25.9% of valid data, followed by client burnout with 40% of valid data including duplicates.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: A reduction of at least 5 points in burnout dimension scores after program completion indicates that the program effectively lowered burnout.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The similarity scores between the 300th participant and all accumulated valid data ranged between 1.17 and 8.14.

    “The similarity scores between the 300th participant and all the accumulated valid data ranged between 1.17 and 8.14.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The similarity system worked in the same way for the 299th participant, and the program recommended to the most similar participant among the valid data was the same.

    “we checked whether the similarity system worked in the same way for the 299th participant, and it was confirmed that the program recommended to the most similar participant among the valid data was the same.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Findingown data

    The 300th participant had the highest work-related burnout and client-related burnout among the burnout dimensions, and laughter therapy was recommended.

    “The 300th participant had the highest work-related burnout and client-related burnout among the burnout dimensions, and laughter therapy was recommended.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Findingown data

    Coping strategy did not change significantly from pretest to posttest 2 after the intervention (t=-1.367, P=.17).

    “Although participants experienced reduced job stress and stress responses after the intervention, no statistically significant change was reported in the coping strategy level.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    When programs were compared, the satisfaction score for storytelling and reflective writing (4.69) was the highest, followed by mindfulness meditation (4.61), laughter therapy (4.58), and ACT (4.52).

    “When programs were compared, the satisfaction score for storytelling and reflective writing (4.69 points) was the highest, followed by mindfulness meditation (4.61 points), laughter therapy (4.58 points), and ACT (4.52 points) (Table 5).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The correlation of coping strategy differences between pretest and posttest 2 and the order of participation was 0.140 (P=.08).

    “The correlation of coping strategy differences between pretest and posttest 2 and the order of participation was 0.140 (P=.08).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For program 1, the most frequently recommended program was laughter therapy (46.3%), followed by storytelling and reflective writing (20%), ACT (18.3%), and mindfulness meditation (15.3%).

    “For program 1, the most frequently recommended program was laughter therapy (46.3%), followed by storytelling and reflective writing (20%), ACT (18.3%), and mindfulness meditation (15.3%).”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    The correlation of stress response differences between pretest and posttest 2 and the order of participation was 0.094 (P=.13).

    “The correlation of stress response differences between pretest and posttest 2 and the order of participation was 0.094 (P=.13).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For program 2, the satisfaction score was 4.49 for the initial 100 participants, 4.66 for the next 100, and 4.70 for the last 100, showing a gradual increase (F=3.911; P=.02).

    “For program 2, the satisfaction score was 4.49 points for the initial 100 participants, it was 4.66 points for the next 100 participants, and the score was 4.70 for the last 100 participants; thus, the satisfaction for program 2 also gradually increased (F=3.911; P=.02).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    For program 2, the most frequently recommended program was mindfulness meditation (41.3%), followed by laughter therapy (23.7%), storytelling and reflective writing (21.3%), and ACT (13.7%).

    “For program 2, the most frequently recommended program was mindfulness meditation (41.3%), followed by laughter therapy (23.7%), storytelling and reflective writing (21.3%), and ACT (13.7%).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Laughter therapy reduced personal burnout by 10.4%, the highest reduction.

    “Laughter therapy reduced personal burnout by 10.4%, which was the highest reduction.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingnone given

    The users were highly satisfied with the app's convenience and high-quality visuals.

    “The users were highly satisfied with the app’s convenience and high-quality visuals.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The Nurse Healing Space program effectively reduced burnout, job stress, and stress responses.

    “This program effectively reduced burnout, job stress, and stress responses.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Changes in burnout, job stress, and stress responses from pretest to posttest are attributable to the Nurse Healing Space intervention rather than to other factors.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    A total of 325 potential participants showed interest in the study, of whom 300 completed the study.

    “A total of 325 potential participants showed interest in the study, of whom, 300 completed the study.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    Storytelling and reflective writing as an intervention for oncology nurses increased self-awareness and self-compassion and decreased insomnia and loneliness.

    “In a recent study that used storytelling and reflective writing as an intervention for oncology nurses, self-awareness and self-compassion reportedly increased, and insomnia and loneliness decreased [46].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [46] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Burnout patterns differ depending on coping strategies, whether working night shifts, working at a tertiary hospital, age, and department.

    “Burnout patterns differ depending on coping strategies, whether working night shifts, working at a tertiary hospital, age, and department [23-25].”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [23] source

      Not checked in the free run.

    • cites [24] source

      Not checked in the free run.

    • cites [25] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    To ensure the algorithm operates as intended, valid data were extracted from 10 pilot tests and 299 optimizations, and the similarity score of the 300th participant was manually calculated using Excel and compared with the suggested program.

    “To ensure that the algorithm operates as intended, we extracted valid data from 10 pilot tests and 299 optimizations. Subsequently, we manually calculated the similarity score of the 300th participant using Excel and checked the suggested program and compared it with what was suggested to the user.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: A reduction of at least 5 points in burnout dimension scores after program completion indicates that the program effectively lowered burnout.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductnone given

    The study was reviewed and approved by the institutional review board (ewha-202208-0005-01) at the principal investigator's institution.

    “This study was reviewed and approved by the institutional review board (ewha-202208-0005-01) at the principal investigator’s institution.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The collected data were encrypted and stored on a computer accessible only to the researchers.

    “The collected data were encrypted and stored on a computer accessible only to the researchers.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Burnout occurs simultaneously in 3 dimensions: personal, work-related, and client-related.

    “Burnout occurs simultaneously in 3 dimensions [1], that is, personal, work-related, and client-related dimensions [2].”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [1] source

      Not checked in the free run.

    • cites [2] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Laughter therapy is an effective way to improve life satisfaction and reduce burnout in nurses.

    “it has also recently been reported as an effective way to improve life satisfaction and reduce burnout in nurses [38].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [38] source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductnone given

    Informed consent was obtained from all participants.

    “Informed consent was obtained from all participants.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Premisecitation

    A previous study demonstrated that video-based exercise programs were effective and had a low dropout rate.

    “Moreover, a previous study demonstrated that video-based exercise programs were effective [70] and had a low dropout rate [71].”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites [70] source

      Not checked in the free run.

    • cites [71] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Interventions that provide tailored programs suited to individual characteristics are scarce in the literature.

    “However, interventions that provide tailored programs suited to individual characteristics are scarce in the literature.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The developed program proved to be an effective intervention by increasing satisfaction and reducing burnout, job stress, and stress response.

    “The program that we developed proved to be an effective intervention by increasing satisfaction and reducing burnout, job stress, and stress response.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: The pre-post changes in burnout, job stress, and stress response are attributable to the intervention rather than to other factors.

  • May combine two claims.

    Premisecitation

    A tailored program is an effective intervention that provides motivation to continue the program, reduces stress by increasing participant engagement, and increases participant satisfaction.

    “A tailored program is an effective intervention that provides motivation to continue the program, reduces stress by increasing participant engagement, and increases participant satisfaction [67].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [67] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    Traditional and conventional care is less effective while personalized care is considered the best.

    “Therefore, traditional and conventional care is less effective while personalized care is considered the best [68].”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [68] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    In recent years, ACT has been increasingly adopted to prevent and reduce professional staff burnout in medical and other workplace settings.

    “In recent years, ACT has been increasingly adopted to prevent and reduce professional staff burnout in medical and other workplace settings [50].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [50] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The most researched program for nurse burnout is mindfulness meditation.

    “This is interesting considering that the most researched program for nurse burnout is mindfulness meditation [29].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [29] source

      Not checked in the free run.

  • May combine two claims.

    Premisecitation

    In that study, AI algorithms were used only to guide programs and programs were conducted in group settings, thus overlooking diversity.

    “AI algorithms were used only to guide programs and programs were conducted in group settings, thus overlooking the aspect of diversity.”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [65] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S4 (based).

    Premisenone given

    Previous studies provided general burnout programs even though dimensions of nurse burnout have long been studied.

    “Previous studies provided general burnout programs even though dimensions of nurse burnout have long been studied.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationcitation

    The resulting sample is not truly representative of the entire population and hence the results are not generalizable.

    “The resulting sample is not truly representative of the entire population and hence the results are not generalizable [72].”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [72] source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Interpretationnone given

    This program is the first customized burnout reduction program for nurses.

    “This program is the first customized burnout reduction program for nurses, and nurse managers can make use of AI-based systems to offer specialized programs for nurse burnout.”
    Quote in the paperNot a substantial claim
48 of 48

  • The paper relies on this without support.

    Changes in burnout, job stress, and stress responses from pretest to posttest are attributable to the Nurse Healing Space intervention rather than to other factors.

    “This program effectively reduced burnout, job stress, and stress responses.”

  • The paper relies on this without support.

    A program that was effective in reducing burnout for a previous user who is most similar to a new user will also be effective for the new user.

    “One of 4 programs that were effective for a previous user who is most similar to the new user, is recommended to the new user.”

  • The paper relies on this without support.

    Previous burnout interventions were general and did not consider specific burnout dimensions and individual characteristics.

    “Although there is ample evidence of nurse burnout, interventions developed in previous studies were general and did not consider specific burnout dimensions and individual characteristics.”

  • The paper relies on this without support.

    The web-based nursing communities used for recruitment are representative of the population of nurses.

    “We used 2 large web-based nursing communities with 94,000, and 170,000 members for data collection.”

  • The paper relies on this without support.

    A reduction of at least 5 points in burnout dimension scores after program completion indicates that the program effectively lowered burnout.

    “Having valid data implies that the program effectively lowers burnout and data are stored in the AI algorithm.”

  • The paper relies on this without support.

    The order of participation is a valid proxy for the degree of algorithm optimization.

    “The satisfaction scores increased over time from participant 1 in October 2022 to participant 300 in January 2023 by optimization (Figure 4).”
    “As the algorithm was optimized with an accumulation of valid data, satisfaction increased significantly.”

  • The paper relies on this without support.

    The pre-post changes in burnout, job stress, and stress response are attributable to the intervention rather than to other factors.

    “The program that we developed proved to be an effective intervention by increasing satisfaction and reducing burnout, job stress, and stress response.”
    “Moreover, the absence of a control group restricts our capacity to determine the causal links underlying the effects of the intervention.”

  • The paper relies on this without support.

    The satisfaction scores reflect the app's usability.

    “Our mobile app’s usability was acceptable, as was discovered through postchecking and satisfaction measurement , which showed that the algorithm functions normally.”

  • The paper relies on this without support.

    Recommending the most similar valid data to new users is a valid measure of the algorithm's accuracy.

    “Furthermore, when we examined whether the algorithm was being applied correctly, we verified that the system recommends the most similar valid data to new users, proving its accuracy.”
9 of 9

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The AI algorithm was trained on data from 300 Korean volunteer nurses who were mainly female and in their 30s.

    “Our AI algorithm was trained on data from 300 Korean volunteer nurses who were mainly female and in their 30's (Cho et al. 2024).”

    From <scp>AI</scp>‐Assisted Tailored Intervention for Nurse Burnout: A Three‐Group Randomized Controlled Trial (2025) · cites this paper as Cho et al. 2024

    Methods > Data Collection and Setting says 300 nurses completed the study and their data were used to optimize the AI algorithm. Results > Intervention Effects reports that 97.7% were female and their mean age was 32.8 years. Our copy does not establish that the participants were Korean or that most, rather than the average participant, were in their 30s. (LLM, conf=0.84)

    Experimental — accuracy not yet measured

  • The 18 included studies were rated as having a moderate risk of bias.

    “Overall, the 18 studies were rated as having a moderate risk of bias (29–48).”

    From Artificial intelligence in nursing: an integrative review of clinical and operational impacts (2025) · cites this paper as [36]

  • Cho et al. (36) evaluated the Nurse Healing Space app, an AI-based mobile intervention designed to reduce burnout; the app significantly decreased burnout scores, job stress, and stress responses among nurses while increasing user satisfaction through personalized recommendations.

    “Cho et al. (36) evaluated the Nurse Healing Space app, an AI-based mobile intervention designed to reduce burnout. The app significantly decreased burnout scores, job stress, and stress responses among nurses while increasing user satisfaction through personalized recommendations.”

    From Artificial intelligence in nursing: an integrative review of clinical and operational impacts (2025) · cites this paper as [36]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Signs of two significant correlations are reversed in the results text

    Major

    The results text reports positive correlations for second-program burnout change and job-stress change, whereas Table 6 reports negative correlations for both. The sign changes the reported direction of these relationships.

  • User C's reductions are described as greater than five points but equal five

    Minor

    The text says two of user C's burnout subdimensions fell by more than 5 points. Table 2 records a reduction of exactly 5 points in each of those two subdimensions.

  • Signs of two other correlations are reversed in the results text

    Minor

    The text gives positive correlations for stress-response and coping-strategy changes, but Table 6 gives negative values for both.

  • Work-related burnout is ranked above a larger percentage

    Minor

    Work-related burnout is called the next highest subdimension at 25.9%, followed by client burnout at 40%. The stated percentages put client burnout above work-related burnout.

4 of 4

Cites

Flagged first, then by citations

Research works printed but not linked: 7

  1. SchaferJBKonstanJJohnRRecommender systems in e-commerce1999Proceedings of the 1st ACM Conference on Electronic CommerceNovember 3-5, 1999Denver, CO158166
  2. LeeKJBaekJHwangboYJKimYCBaeSWBaekTHAMPER (Aim-Measure-Predict-Evaluate-Recommend): The paradigm of digital me23rd International Conference on Electronic Commerce202223rd International Conference on Electronic Commerce 2022June 22-24, 2022Daegu, South KoreaSouth KoreaDBpia182184
  3. Bittman, B, 2003, Adv Mind Body Med, 19, 4
  4. Hayes, S, 2005, Get Out of Your Mind and Into Your Life: The New Acceptance and Commitment Therapy
  5. HamMPath analysis of emotional labor and burnout of nurses [thesis]Gyeongsang National University20112024-06-17https://dcollection.gnu.ac.kr/jsp/common/DcLoOrgPer.jsp?sItemId=000000014045
  6. Choi, S, 2006, J Korean Neuropsychiatr Assoc, 45, 541
  7. Shin, H, 2002, Korean J Couns Psychother, 14, 919

Cited by

Sorted by citations

Related

Authors
Cho, A., Cha, C., & Baek, G.
Affiliations
Ewha Womans University (KR)
Corresponding author
Cha, Chiyoung
ORCID
Venue
Journal of Medical Internet Research
Publisher
JMIR Publications Inc.
Published
2024-06-21
Volume
26
Pages
e54029
ISSN
1438-8871 (ISSN-L)
DOI
10.2196/54029
PMC
PMC11226930
OpenAlex
W4396714278
Semantic Scholar
008372366afd08733661b1ddc439a9b2bf2bce7f
Cited by
29
References
68
Citations by year
2026: 16 · 2025: 12 · 2024: 1
FWCI
11.50
Type
journal article
Language
en
Open access
Licence
cc-by
Locations
  • Journal of Medical Internet Research (publishedVersion, cc-by)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
Topics
  • Social Sciences › Psychology › Applied Psychology › Digital Mental Health Interventions
  • Health Sciences › Health Professions › General Health Professions › Healthcare professionals’ stress and burnout
  • Social Sciences › Psychology › Social Psychology › Stress and Burnout Research
Keywords
nurse burnout, artificial intelligence, mindfulness meditation, acceptance and commitment therapy, occupational stress, stress response, coping strategies, app usability, user satisfaction, burnout dimensions, nurse turnover, digital mental health interventions, algorithm optimization, storytelling, laughter therapy, reflective writing, Copenhagen Burnout Inventory
MeSH terms
Adaptation, Psychological, Adult, Artificial Intelligence, Burnout, Professional, Female, Humans, Male, Middle Aged, Nurses, Mobile Applications

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-10-09.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-10-09.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review