tetano
Editor, Senior Moderator
J Med Internet Res
. 2025 Jan 28:27:e63634.
doi: 10.2196/63634. Health and Experiences During the COVID-19 Pandemic Among Children and Young People: Analysis of Free-Text Responses From the Children and Young People With Long COVID Study
Natalia K Rojas[SUP] 1 [/SUP], Sam Martin[SUP] 1 [/SUP], Mario Cortina-Borja[SUP] 2 [/SUP], Roz Shafran[SUP] 2 [/SUP], Lana Fox-Smith[SUP] 2 [/SUP], Terence Stephenson[SUP] 2 [/SUP], Brian C F Ching[SUP] 3 [/SUP], Anaïs d'Oelsnitz[SUP] 2 [/SUP], Tom Norris[SUP] 1 [/SUP], Yue Xu[SUP] 4 [/SUP], Kelsey McOwat[SUP] 5 [/SUP], Emma Dalrymple[SUP] 2 [/SUP], Isobel Heyman[SUP] 2 6 [/SUP], Tamsin Ford[SUP] 6 [/SUP], Trudie Chalder[SUP] 7 [/SUP], Ruth Simmons[SUP] 5 [/SUP]; CLoCk Consortium[SUP] 8 [/SUP]; Snehal M Pinto Pereira[SUP] 1 [/SUP]
Affiliations
Background: The literature is equivocal as to whether the predicted negative mental health impact of the COVID-19 pandemic came to fruition. Some quantitative studies report increased emotional problems and depression; others report improved mental health and well-being. Qualitative explorations reveal heterogeneity, with themes ranging from feelings of loss to growth and development.
Objective: This study aims to analyze free-text responses from children and young people participating in the Children and Young People With Long COVID study to get a clearer understanding of how young people were feeling during the pandemic.
Methods: A total of 8224 free-text responses from children and young people were analyzed using InfraNodus, an artificial intelligence-powered text network analysis tool, to determine the most prevalent topics. A random subsample of 411 (5%) of the 8224 responses underwent a manual sentiment analysis; this was reweighted to represent the general population of children and young people in England.
Results: Experiences fell into 6 main overlapping topical clusters: school, examination stress, mental health, emotional impact of the pandemic, social and family support, and physical health (including COVID-19 symptoms). Sentiment analysis showed that statements were largely negative (314/411, 76.4%), with a small proportion being positive (57/411, 13.9%). Those reporting negative sentiment were mostly female (227/314, 72.3%), while those reporting positive sentiment were mostly older (170/314, 54.1%). There were significant observed associations between sentiment and COVID-19 status as well as sex (P=.001 and P<.001, respectively) such that the majority of the responses, regardless of COVID-19 status or sex, were negative; for example, 84.1% (227/270) of the responses from female individuals and 61.7% (87/141) of those from male individuals were negative. There were no observed associations between sentiment and all other examined demographics. The results were broadly similar when reweighted to the general population of children and young people in England: 78.52% (negative), 13.23% (positive), and 8.24% (neutral).
Conclusions: We used InfraNodus to analyze free-text responses from a large sample of children and young people. The majority of responses (314/411, 76.4%) were negative, and many of the children and young people reported experiencing distress across a range of domains related to school, social situations, and mental health. Our findings add to the literature, highlighting the importance of specific considerations for children and young people when responding to national emergencies.
Keywords: AI; COVID-19; InfraNodus; artificial intelligence; children and young people; discourse analysis; experiences; free-text responses; long COVID; sentiment analysis; text mining.
. 2025 Jan 28:27:e63634.
doi: 10.2196/63634. Health and Experiences During the COVID-19 Pandemic Among Children and Young People: Analysis of Free-Text Responses From the Children and Young People With Long COVID Study
Natalia K Rojas[SUP] 1 [/SUP], Sam Martin[SUP] 1 [/SUP], Mario Cortina-Borja[SUP] 2 [/SUP], Roz Shafran[SUP] 2 [/SUP], Lana Fox-Smith[SUP] 2 [/SUP], Terence Stephenson[SUP] 2 [/SUP], Brian C F Ching[SUP] 3 [/SUP], Anaïs d'Oelsnitz[SUP] 2 [/SUP], Tom Norris[SUP] 1 [/SUP], Yue Xu[SUP] 4 [/SUP], Kelsey McOwat[SUP] 5 [/SUP], Emma Dalrymple[SUP] 2 [/SUP], Isobel Heyman[SUP] 2 6 [/SUP], Tamsin Ford[SUP] 6 [/SUP], Trudie Chalder[SUP] 7 [/SUP], Ruth Simmons[SUP] 5 [/SUP]; CLoCk Consortium[SUP] 8 [/SUP]; Snehal M Pinto Pereira[SUP] 1 [/SUP]
Affiliations
- PMID: 39874576
- DOI: 10.2196/63634
Background: The literature is equivocal as to whether the predicted negative mental health impact of the COVID-19 pandemic came to fruition. Some quantitative studies report increased emotional problems and depression; others report improved mental health and well-being. Qualitative explorations reveal heterogeneity, with themes ranging from feelings of loss to growth and development.
Objective: This study aims to analyze free-text responses from children and young people participating in the Children and Young People With Long COVID study to get a clearer understanding of how young people were feeling during the pandemic.
Methods: A total of 8224 free-text responses from children and young people were analyzed using InfraNodus, an artificial intelligence-powered text network analysis tool, to determine the most prevalent topics. A random subsample of 411 (5%) of the 8224 responses underwent a manual sentiment analysis; this was reweighted to represent the general population of children and young people in England.
Results: Experiences fell into 6 main overlapping topical clusters: school, examination stress, mental health, emotional impact of the pandemic, social and family support, and physical health (including COVID-19 symptoms). Sentiment analysis showed that statements were largely negative (314/411, 76.4%), with a small proportion being positive (57/411, 13.9%). Those reporting negative sentiment were mostly female (227/314, 72.3%), while those reporting positive sentiment were mostly older (170/314, 54.1%). There were significant observed associations between sentiment and COVID-19 status as well as sex (P=.001 and P<.001, respectively) such that the majority of the responses, regardless of COVID-19 status or sex, were negative; for example, 84.1% (227/270) of the responses from female individuals and 61.7% (87/141) of those from male individuals were negative. There were no observed associations between sentiment and all other examined demographics. The results were broadly similar when reweighted to the general population of children and young people in England: 78.52% (negative), 13.23% (positive), and 8.24% (neutral).
Conclusions: We used InfraNodus to analyze free-text responses from a large sample of children and young people. The majority of responses (314/411, 76.4%) were negative, and many of the children and young people reported experiencing distress across a range of domains related to school, social situations, and mental health. Our findings add to the literature, highlighting the importance of specific considerations for children and young people when responding to national emergencies.
Keywords: AI; COVID-19; InfraNodus; artificial intelligence; children and young people; discourse analysis; experiences; free-text responses; long COVID; sentiment analysis; text mining.