A systematic review of artificial intelligence chatbots for promoting physical activity, healthy diet, and weight loss

May combine two claims.

Over half (56%) of studies used a constrained chatbot, while the remaining studies used unconstrained chatbots.

“Over half (56%) of the studies used a constrained chatbot (i.e., rule-based), while the remaining studies used unconstrained chatbots that resemble human-to-human communication.”

• Also stated in: § Summary of intervention and chatbot characteristics · plain

  “In 5 studies, the chatbots were constrained (i.e., users could only select pre-programmed responses in the chat) [31, 33–36], and in 4, the chatbots were unconstrained (i.e., users could freely type or speak to the chatbot) [20, 21, 32, 37].”

May combine two claims.

Of the 9 included studies, 4 were randomized controlled trials and 5 were quasi-experimental studies.

“Of the 9 studies, 4 were randomized controlled trials and 5 were quasi-experimental studies.”

• Also stated in: § Summary of study designs and sample characteristics · plain

  “Five studies employed quasi-experimental designs [20, 21, 35–37], and 4 were RCTs [31–34].”

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

Inter-rater agreement for title and abstract screening was 97.4% (Cohen's Kappa = .725).

“The agreement in the abstract and title screening between the two reviewers was 97.4% (Cohen’s Kappa = .725).”

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

Inter-rater agreement for full-text screening was 91.9% (Cohen's Kappa = .734).

“The agreement for full text screening was 91.9% (Cohen’s Kappa = .734).”

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

Eighty-nine percent of studies specified a chatbot name and 33% specified chatbot gender, which was woman.

“Eighty-nine and thirty-three percent of the studies specified a name and gender (i.e., woman) of the chatbot, respectively.”

• Also stated in: § Summary of intervention and chatbot characteristics · plain

  “Regarding chatbot characteristics, identity features, such as specific names (n = 8) [20, 21, 31–33, 35–37] and chatbot gender (n = 3) [20, 31, 33], were specified.”

• Also stated in: § Summary of intervention and chatbot characteristics · plain

  “Notably, the chatbot gender was woman in the 3 studies that reported it [20, 31, 33].”

May combine two claims.

The database search retrieved 1692 citations, and 9 studies met the inclusion criteria.

“The database search retrieved 1692 citations, and 9 studies met the inclusion criteria.”

• Also stated in: § Study selection · plain

  “The search yielded 2360 references in total, from which 668 duplicates were removed. A total of 1692 abstracts were then screened, among which 1630 were judged ineligible, leaving 62 papers to be read in full text. In total, 9 papers met the eligibility criteria and were included.”

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

One study used 4 different platforms to deliver the intervention, and 2 studies used a chatbot app (Ally) available on both Android and iOS systems.

“One study used 4 different platforms to deliver the intervention [21], and 2 studies used a chatbot app (i.e., Ally app) that was available on both Android and iOS systems [31, 33].”

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

A variety of commercially available technical platforms were used to host the chatbot and deliver the interventions, including Slack (n=2), KakaoTalk (n=1), Facebook messenger (n=3), Telegram messenger (n=1), WhatsApp (n=1), and SMS (n=2).

“A variety of commercially available technical platforms were used to host the chatbot and deliver the interventions, including Slack (n = 2), KakaoTalk (n = 1), Facebook messenger (n = 3), Telegram messenger (n = 1), WhatsApp (n = 1), and short messaging services (SMS) (n = 2).”

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

One chatbot communicated only on a weekly basis.

“One chatbot communicated only on a weekly basis [20], and 1 communicated daily, weekly, on weekends or weekdays or at a scheduled date and time [35].”

• cites [20] Maher CA, Davis CR, Curtis RG, Short CE, Murphy KJ. A physical activity and diet program delivered by artificially intelligent virtual health coach: proof-of-concept study. JMIR mHealth uHealth. 2020;8(7):e17558. sourceNot checked yet

  Not checked in the free run.

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

Two studies were conducted in the United States and the remaining 7 were conducted in Switzerland, Australia, South Korea, and Italy, with 1 not reported.

“Two studies [21, 37] were conducted in the United States and the remaining 7 were conducted in Switzerland [31, 33, 36], Australia [20], South Korea [32], and Italy [34] (1 not reported [35]).”

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

One study did not specify when and how often messages were delivered.

“One study did not specify when and how often the messages were delivered [21].”

• cites [21] Stephens TN, Joerin A, Rauws M, Werk LN. Feasibility of pediatric obesity and prediabetes treatment support through Tess, the AI behavioral coaching chatbot. Transl Behav Med. 2019;9(3):440–7. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

One study solely targeted an adolescent population, most studies targeted an adult population, and one study did not report the target population's age.

“One study [21] solely targeted an adolescent population, whereas most studies targeted an adult population [20, 31–35, 37]. One study [36] did not report the target population’s age.”

May combine two claims.

In total, 891 participants were represented in the 9 studies, with sample sizes ranging from 19 to 274 participants.

“In total, 891 participants were represented in the 9 studies, with sample sizes ranging from 19 to 274 participants.”

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

All 9 chatbots delivered messages in text format.

“All 9 chatbots delivered messages in text format.”

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

In addition to text, 3 chatbots used graphs, 2 used images, 1 used voice, and 1 used a combination of graphs, images, and videos.

“In addition to text, 3 chatbots used graphs [31, 33, 37], 2 used images [32, 35], 1 used voice [21], and 1 used a combination of graphs, images, and videos [36].”

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

The 9 included papers were recently published: 3 in 2020, 4 in 2019, and 2 in 2018.

“The 9 included papers had been recently published (3 were published in 2020 [20, 31, 32], 4 in 2019 [21, 33–35], and 2 in 2018 [36, 37]).”

May combine two claims.

The mean age of samples ranged from 15.2 to 56.2 years (SD range 2.0 to 13.7), and females/women represented 42.1 to 87.9% of the sample.

“The mean age of the samples ranged from 15.2 to 56.2 years (SD range = 2.0 to 13.7), and females/women represented 42.1 to 87.9% of the sample.”

May combine two claims.

The theories used in the 5 studies included the Health Action Process Approach (n=2), Habit Formation Model (n=1), Technology Acceptance Model (n=1), AttrakDiff Model (n=1), Cognitive Behavioral Therapy (n=1), Emotionally Focused Therapy (n=1), Behavioral Activation (n=1), Motivational Interviewing (n=1), and Structured Reflection Model (n=1).

“The theories used in the 5 studies included the Health Action Process Approach (n = 2), the Habit Formation Model (n = 1), the Technology Acceptance Model (n = 1), the AttrakDiff Model (n = 1), Cognitive Behavioral Therapy (n = 1), Emotionally Focused Therapy (n = 1), Behavioral Activation (n = 1), Motivational Interviewing (n = 1), and the Structured Reflection Model (n = 1).”

May combine two claims.

Most chatbots were equipped with relational capacity (n=8) and persuasive capacity (n=9).

“Most chatbots were equipped with relational capacity (n = 8; i.e., conversation strategy to establish, maintain, or enhance social relationships with users) and persuasive capacity (n = 9; i.e., conversation strategy to change user’s behaviors and behavioral determinants), meaning that the conversations were designed to induce behavioral changes while engaging with users socially.”

May combine two claims.

The 9 studies varied in intervention and program length, lasting from 1 week to 3 months.

“The 9 studies varied in intervention and program length, lasting from 1 week to 3 months.”

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

Self-reported dietary intake was measured in 4 included studies.

“Self-reported dietary intake was measured in 4 studies [20, 34–36].”

May combine two claims.

One study used multi-intervention components, and participants had access to an AI chatbot along with a study website with educational materials.

“One study used multi-intervention components, and participants had access to an AI chatbot along with a study website with educational materials [20].”

• cites [20] Maher CA, Davis CR, Curtis RG, Short CE, Murphy KJ. A physical activity and diet program delivered by artificially intelligent virtual health coach: proof-of-concept study. JMIR mHealth uHealth. 2020;8(7):e17558. sourceNot checked yet

  Not checked in the free run.

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

The one study reporting participants' race/ethnicity included 43% Hispanic, 39% White, 9% Black, and 9% Asian participants.

“The study [21] that reported participants’ race/ethnicity information included 43% Hispanic, 39% White, 9% Black, and 9% Asian participants.”

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

Only 3 chatbots were available on-demand so that users could initiate conversation at any preferred time.

“Only 3 chatbots [20, 21, 32] were available on-demand so that users could initiate conversation at any preferred time.”

May combine two claims.

In 5 studies, the chatbots were constrained, and in 4, the chatbots were unconstrained.

“In 5 studies, the chatbots were constrained (i.e., users could only select pre-programmed responses in the chat) [31, 33–36], and in 4, the chatbots were unconstrained (i.e., users could freely type or speak to the chatbot) [20, 21, 32, 37].”

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

One chatbot communicated daily, weekly, on weekends or weekdays, or at a scheduled date and time.

“One chatbot communicated only on a weekly basis [20], and 1 communicated daily, weekly, on weekends or weekdays or at a scheduled date and time [35].”

• cites [35] Fadhil A, Wang Y, Reiterer H. Assistive conversational agent for health coaching: a validation study. Methods Inf Med. 2019;58(1):009–23. sourceNot checked yet

  Not checked in the free run.

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

One study was guided by 3 theories and another by 4 theories.

“One was guided by 3 theories [35] and another by 4 theories [21].”

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

Identity features such as specific names (n=8) and chatbot gender (n=3) were specified.

“Regarding chatbot characteristics, identity features, such as specific names (n = 8) [20, 21, 31–33, 35–37] and chatbot gender (n = 3) [20, 31, 33], were specified.”

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

The chatbot gender was woman in the 3 studies that reported it.

“Notably, the chatbot gender was woman in the 3 studies that reported it [20, 31, 33].”

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

Among the 4 studies reporting education/income, the majority included undergraduate students or people with graduate degrees.

“Among the 4 studies [31, 34, 35, 37] that reported the information, the majority included undergraduate students or people with graduate degrees.”

May combine two claims.

Among the 7 included studies that measured physical activity, 2 used objective measures, 4 used self-reported measures, and 1 used both.

“Among 7 studies that measured PA [20, 21, 31–33, 35, 37], 2 used objective measures [31, 33], 4 used self-reported measures [20, 21, 32, 35], and 1 used both [37].”

May combine two claims.

Most included studies focused on changes in physical activity, and the majority reported significant improvements in physical activity-related behaviors.

“Most studies focused on changes in PA, and majority [20, 31–33, 37] reported significant improvements in PA-related behaviors.”

May combine two claims.

In study [35], participants with high intervention adherence showed a marginal increase in healthy diet intention, whereas those with low adherence showed decreased healthy diet intention.

“In this study, participants with high intervention adherence showed a marginal increase, whereas, those with low adherence showed decreased healthy diet intention”

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

Five included studies reported engagement using various types of measurements.

“Five studies reported engagement [20, 21, 31, 33, 37] using various types of measurements”

May combine two claims.

Although only 1 included study reported weight-related outcomes, it reported significant weight change after the intervention.

“Although only 1 study [20] reported weight-related outcomes, it reported significant weight change after the intervention.”

May combine two claims.

Among the 5 quasi-experimental studies, 2 reported only physical activity change outcomes, 1 reported only diet change outcomes, and 2 reported both outcomes.

“Among the 5 quasi-experimental studies, 2 [21, 37] reported only PA change outcomes, 1 [36] reported only diet change outcomes, and 2 [20, 35] reported both outcomes.”

May combine two claims.

Study [21] only reported descriptive statistics and showed that participants experienced positive progress towards physical activity goals 81% of the time.

“[21]. only reported descriptive statistics and showed that participants experienced positive progress towards PA goals 81% of the time.”

May combine two claims.

Non-significant findings and a lack of statistical reporting were more prevalent in the quasi-experimental studies, but the direction of intervention effects was similar to those reported in the RCTs.

“To summarize, non-significant findings and a lack of statistical reporting were more prevalent in the quasi-experimental studies, but the direction of intervention effects were similar to those reported in the RCTs.”

May combine two claims.

Three studies reported other measures, and 1 found that pragmatic, hedonic, appealing, and social ratings of the chatbot were significantly higher than the middle score.

“Three studies reported other measures [20, 35, 37], and 1 found that pragmatic, hedonic, appealing, and social ratings of the chatbot were significantly higher than the middle score [35].”

Classified as probably not a substantial claim.

Despite the rapid increase in publications about chatbot designs and interventions, standard measures for evaluating chatbot interventions and theory-guided chatbots are still lacking.

“Despite the rapid increase in publications about chatbot designs and interventions, standard measures for evaluating chatbot interventions and theory-guided chatbots are still lacking.”

May combine two claims.

Over half of the included studies used a constrained AI chatbot, while the remaining used unconstrained AI chatbots.

“With regard to user input, over half of the studies [31, 33–36] used a constrained AI chatbot, while the remaining [20, 21, 32, 37] used unconstrained AI chatbots.”

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

To the best of the authors' knowledge, this is the first review to systematically examine the characteristics and potential efficacy of AI chatbot interventions in lifestyle modifications.

“First, to the best of our knowledge, this is the first review to systematically examine the characteristics and potential efficacy of AI chatbot interventions in lifestyle modifications, thereby providing crucial insights for identifying gaps and future directions for research and clinical practice.”

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

The systematic review was conducted in accordance with PRISMA guidelines.

“The systematic review was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines.”

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

Six electronic databases were searched in consultation with a medical librarian using pre-planned systematic search strategies.

“In consultation with a medical librarian (MF), pre-planned systematic search strategies were used for six electronic databases (PubMed, EMBASE, ACM Digital Library, Web of Science Core Collection, PsycINFO, and IEEE).”

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

Two researchers independently screened titles/abstracts and full texts, and discrepancies were resolved by a third researcher.

“In the first phase of screening, two researchers (YO and JZ) independently assessed all study titles and abstracts against the eligibility criteria in Table 1.”

• Also stated in: § Study selection · plain

  “Discrepancies at each stage were resolved through discussion with a third researcher (YF).”

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

AI chatbots can participate in a broad range of conversations, from constrained (rule-based) to unconstrained (human-to-human-like).

“Powered by natural language processing and cloud computing infrastructures, AI chatbots can participate in a broad range, from constrained (i.e., rule-based) to unconstrained conversations (i.e., human-to-human-like communication) [1].”

• cites [1] Laranjo L, Dunn AG, Tong HL, Kocaballi AB, Chen J, Bashir R, et al. Conversational agents in healthcare: a systematic review. J Am Med Inform Assoc. 2018;25(9):1248–58. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Technology-based interventions do not require traditional in-site visits and help reduce participants' time and financial costs.

“Moreover, these interventions do not require traditional in-site visits, and thus, help reduce participants’ time and financial costs [16].”

• cites [16] Manzoni GM, Pagnini F, Corti S, Molinari E, Castelnuovo G. Internet-based behavioral interventions for obesity: an updated systematic review. Clin Pract Epidemiol Ment Health. 2011;7:19. sourceNot checked yet

  Not checked in the free run.

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

The systematic review protocol was registered in PROSPERO (CRD42020216761).

“The protocol of this systematic review was registered at the International Prospective Register of Systematic Reviews (PROSPERO) (ID: CRD42020216761).”

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

The NIH tools were used to assess confidence in the review's evaluations and conclusions, and poor-quality studies were not excluded.

“These tools were used to assess confidence in the evaluations and conclusions of this systematic review. We did not use these tools to exclude the findings of poor quality studies.”

May combine two claims.

AI chatbots have great potential to increase the accessibility and efficacy of personalized lifestyle modification programs.

“These limitations can be overcome by utilizing AI chatbots, which have great potential to increase the accessibility and efficacy of personalized lifestyle modification programs [20, 21].”

• cites [20] Maher CA, Davis CR, Curtis RG, Short CE, Murphy KJ. A physical activity and diet program delivered by artificially intelligent virtual health coach: proof-of-concept study. JMIR mHealth uHealth. 2020;8(7):e17558. sourceNot checked yet

  Not checked in the free run.

• cites [21] Stephens TN, Joerin A, Rauws M, Werk LN. Feasibility of pediatric obesity and prediabetes treatment support through Tess, the AI behavioral coaching chatbot. Transl Behav Med. 2019;9(3):440–7. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Digital technology-based programs have resulted in increased physical activity and reduced body weight, at least for a short duration.

“Several systematic reviews and meta-analyses [15–19] have shown that these digital technology-based programs resulted in increased PA and reduced body weight, at least for a short duration.”

• cites [15] Mateo GF, Granado-Font E, Ferré-Grau C, Montaña-Carreras X. Mobile phone apps to promote weight loss and increase physical activity: a systematic review and meta-analysis. J Med Internet Res. 2015;17(11):e253. sourceNot checked yet

  Not checked in the free run.

• cites [16] Manzoni GM, Pagnini F, Corti S, Molinari E, Castelnuovo G. Internet-based behavioral interventions for obesity: an updated systematic review. Clin Pract Epidemiol Ment Health. 2011;7:19. sourceNot checked yet

  Not checked in the free run.

• cites [17] Beleigoli AM, Andrade AQ, Cançado AG, Paulo MN, Maria De Fátima HD, Ribeiro AL. Web-based digital health interventions for weight loss and lifestyle habit changes in overweight and obese adults: systematic review and meta-analysis. J Med Internet Res. 2019;21(1):e298. sourceNot checked yet

  Not checked in the free run.

• cites [18] Laranjo L, Arguel A, Neves AL, Gallagher AM, Kaplan R, Mortimer N, et al. The influence of social networking sites on health behavior change: a systematic review and meta-analysis. J Am Med Inform Assoc. 2015;22(1):243–56. sourceNot checked yet

  Not checked in the free run.

• cites [19] Laranjo L, Ding D, Heleno B, Kocaballi B, Quiroz JC, Tong HL, et al. Do smartphone applications and activity trackers increase physical activity in adults? Systematic review, meta-analysis and metaregression. Br J Sports Med. 2021;55(8):422–32. sourceNot checked yet

  Not checked in the free run.

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

Two researchers independently validated the extracted data.

“Two researchers (YF and JZ) independently validated the extracted data.”

May combine two claims.

Technology-based programs can provide instrumental help in finding healthier alternatives or facilitating supportive social groups.

“While digital technologies may not address environmental factors that constrain an individual’s health environment, technology-based programs can provide instrumental help in finding healthier alternatives or facilitating the creation of supportive social groups [13, 14].”

• cites [13] Zhang J, Brackbill D, Yang S, Becker J, Herbert N, Centola D. Support or competition? How online social networks increase physical activity: a randomized controlled trial. Prev Med Rep. 2016;4:453–8. sourceNot checked yet

  Not checked in the free run.

• cites [14] Zhang J, Brackbill D, Yang S, Centola D. Efficacy and causal mechanism of an online social media intervention to increase physical activity: results of a randomized controlled trial. Prev Med Rep. 2015;2:651–7. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Due to heterogeneity in outcomes, measures, and designs, results were qualitatively synthesized and no meta-analysis or publication bias assessment was conducted.

“Due to the heterogeneity in the types of study outcomes, outcome measures, and clinical trial designs, we qualitatively evaluated and synthesized the results of the studies. We did not conduct a meta-analysis and did not assess publication bias.”

Classified as probably not a substantial claim.

Two reviewers independently evaluated risk of bias using the two NIH quality assessment tools.

“Two reviewers (YO and JZ) independently evaluated the risk of bias of included studies using the two National Institutes of Health (NIH) quality assessment tools [30].”

• cites [30] NIH National Heart, Lung, and Blood Institute. Study quality assessment tools. Available from: https://www.nhlbi.nih.gov/health-topics/study-quality-assessment-tools.Not checked yet

  Not checked in the free run.

Classified as probably not a substantial claim.

The AI Chatbot Behavior Change Model provides a comprehensive framework for analyzing and evaluating chatbot designs and technologies.

“Data extraction forms were developed based on the AI Chatbot Behavior Change Model [22], which provides a comprehensive framework for analyzing and evaluating chatbot designs and technologies.”

• cites [22] Zhang J, Oh YJ, Lange P, Yu Z, Fukuoka Y. Artificial intelligence Chatbot behavior change model for designing artificial intelligence Chatbots to promote physical activity and a healthy diet. J Med Internet Res. 2020;22(9):e22845. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Sample sizes of the 4 RCTs ranged from 106 to 274, and a priori power analyses were reported in 3 of them, showing sufficient power for analyzing the specified outcomes.

“Sample sizes of the 4 RCT studies ranged from 106 to 274 and a priori power analyses were reported in 3 [31, 32, 34], which showed that the sample sizes had sufficient power for analyzing the specified outcomes.”

May combine two claims.

Results from systematic reviews and meta-analyses suggest chatbots have high potential for healthcare and psychiatric use, such as promoting antipsychotic medication adherence and reducing stress, anxiety, and/or depression symptoms.

“Results from a few systematic reviews and meta-analyses suggest that chatbots have a high potential for healthcare and psychiatric use, such as promoting antipsychotic medication adherence as well as reducing stress, anxiety, and/or depression symptoms [1, 25, 26].”

• cites [1] Laranjo L, Dunn AG, Tong HL, Kocaballi AB, Chen J, Bashir R, et al. Conversational agents in healthcare: a systematic review. J Am Med Inform Assoc. 2018;25(9):1248–58. sourceNot checked yet

  Not checked in the free run.

• cites [25] Gentner T, Neitzel T, Schulze J, Buettner R. A Systematic literature review of medical chatbot research from a behavior change perspective. In 2020 IEEE 44th Annual Computers, Software, and Applications Conference (COMPSAC). IEEE; 2020. p. 735-40. sourceNot checked yet

  Not checked in the free run.

• cites [26] Vaidyam AN, Wisniewski H, Halamka JD, Kashavan MS, Torous JB. Chatbots and conversational agents in mental health: a review of the psychiatric landscape. Can J Psychiatry. 2019;64(7):456–64. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

Overall, reporting of participants' sociodemographic information was inconsistent and insufficient across the studies.

“Overall, reporting of participants’ sociodemographic information was inconsistent and insufficient across the studies.”

May combine two claims.

Of the 5 quasi-experimental studies, 1 was rated as fair and 4 as poor due to flaws with regard to several critical items.

“Of the 5 quasi-experimental studies [20, 21, 35–37], 1 study was rated as fair [20] and 4 studies were rated as poor [21, 35–37] due to flaws with regard to several critical.”

May combine two claims.

The poor-quality quasi-experimental studies reported neither a power analysis nor follow-up rates after baseline.

“These studies reported neither a power analysis to ensure that the sample size was sufficiently large, nor follow-up rates after baseline.”

Classified as probably not a substantial claim.

Inconsistent reporting of user engagement, acceptability, and satisfaction measures may be problematic because it could contribute challenges to the interpretation and comparison of results across different chatbot systems.

“Inconsistent reporting of user engagement, acceptability, and satisfaction measures may be problematic because it could contribute challenges to the interpretation and comparison of the results across different chatbot systems [1].”

• cites [1] Laranjo L, Dunn AG, Tong HL, Kocaballi AB, Chen J, Bashir R, et al. Conversational agents in healthcare: a systematic review. J Am Med Inform Assoc. 2018;25(9):1248–58. sourceNot checked yet

  Not checked in the free run.

May combine two claims.

The poorly rated RCT did not report overall dropout rates or differential dropout between treatment groups, did not report a power analysis, and did not prespecify outcomes for hypothesis testing.

“The poorly rated study did not report overall dropout rates or the differential dropout rates between treatment groups, did not report that the sample size was sufficiently large to be able to detect differences between groups (i.e., no power analysis), and did not prespecify outcomes for hypothesis testing.”

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

The lack of theoretical foundations of existing chatbot systems has also been noted in previous literature.

“The lack of theoretical foundations of existing chatbot systems has also been noted in previous literature [42].”

• cites [42] Fadhil A. Can a chatbot determine my diet?: Addressing challenges of chatbot application for meal recommendation. arXiv preprint arXiv:180209100. 2018.Not checked yet

  Not checked in the free run.

May combine two claims.

In the poor-quality quasi-experimental studies, statistical methods did not examine pre-to-post changes in outcome measures and lacked reporting of statistical significance.

“Additionally, the statistical methods did not examine pre-to-post changes in outcome measures and lacked reporting of statistical significance.”

May combine two claims.

Sample sizes for the 5 quasi-experimental studies were small, ranging from 19 to 36 participants, suggesting they may lack statistical power to detect potential intervention effects.

“In contrast, sample sizes for the 5 quasi-experimental studies were small, ranging from 19 to 36 participants, suggesting that these studies may lack statistical power to detect potential intervention effects.”

May combine two claims.

Of the 4 RCT studies, 3 were rated as fair and 1 as poor due to its lack of reporting of several critical items.

“Of the 4 RCT studies [31–34], 3 were rated as fair [31, 32, 34] and 1 was rated as poor [33] due to its lack of reporting of several critical.”

May combine two claims.

Chatbots may improve physical activity, but the review was not able to make definitive conclusions on the potential efficacy of chatbot interventions on promoting physical activity, healthy eating, or weight loss.

“In summation, chatbots can improve PA, but the study not able to make definitive conclusions on the potential efficacy of chatbot interventions on promoting PA, healthy eating, or weight loss.”

• Also stated in: § Conclusions · hedged

  “Chatbots may improve PA, but this study was not able to make definitive conclusions of the potential efficacy of chatbot interventions on PA, diet, and weight management/loss.”

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

Constrained AI chatbots are rule-based, well-structured, and easy to build, control, and implement, thus ensuring quality and consistency in the structure and delivery of content.

“Constrained AI chatbots are rule-based, well-structured, and easy to build, control, and implement, thus ensuring the quality and consistency in the structure and delivery of content [42].”

• cites [42] Fadhil A. Can a chatbot determine my diet?: Addressing challenges of chatbot application for meal recommendation. arXiv preprint arXiv:180209100. 2018.Not checked yet

  Not checked in the free run.

May combine two claims.

Due to heterogeneity of outcome measures and the limited number of RCT designs, the review was not able to conduct a meta-analysis and make firm conclusions of the potential efficacy of chatbot interventions.

“Lastly, due to the heterogeneity of outcome measures and the limited number of RCT designs in this systematic review, we were not able to conduct a meta-analysis and make firm conclusions of the potential efficacy of chatbot interventions.”

May combine two claims.

Constrained AI chatbots are not able to adapt to participants' inquiries and address emergent questions, and are thus not suitable for sustaining more natural and complex interactions with participants.

“However, they are not able to adapt to participants’ inquiries and address emergent questions, and are, thus, not suitable for sustaining more natural and complex interactions with participants [42].”

• cites [42] Fadhil A. Can a chatbot determine my diet?: Addressing challenges of chatbot application for meal recommendation. arXiv preprint arXiv:180209100. 2018.Not checked yet

  Not checked in the free run.

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

The review limited its search to peer-reviewed studies published as full-text in English only.

“Second, we limited our search to peer-reviewed studies published as full-text in English only.”

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

The review did not search gray literature.

“First, we did not search gray literature in this systematic review.”

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

The review developed comprehensive search strategies with an MLS for six electronic databases to increase the sensitivity and comprehensiveness of the search.

“Second, we developed comprehensive search strategies with an MLS for six electronic databases to increase the sensitivity and comprehensiveness of our search.”

May combine two claims.

Unconstrained AI chatbots are known to simulate naturalistic human-to-human communication and may strengthen interventions in general, particularly in the long-term, due to their flexibility and adaptability in conversations.

“In contrast, unconstrained AI chatbots are known to simulate naturalistic human-to-human communication and may strengthen interventions in general, particularly in the long-term, due to their flexibility and adaptability in conversations [1, 38, 42].”

• cites [1] Laranjo L, Dunn AG, Tong HL, Kocaballi AB, Chen J, Bashir R, et al. Conversational agents in healthcare: a systematic review. J Am Med Inform Assoc. 2018;25(9):1248–58. sourceNot checked yet

  Not checked in the free run.

• cites [38] Milne-Ives M, de Cock C, Lim E, Shehadeh MH, de Pennington N, Mole G, et al. The effectiveness of artificial intelligence conversational agents in health care: systematic review. J Med Internet Res. 2020;22(10):e20346. sourceNot checked yet

  Not checked in the free run.

• cites [42] Fadhil A. Can a chatbot determine my diet?: Addressing challenges of chatbot application for meal recommendation. arXiv preprint arXiv:180209100. 2018.Not checked yet

  Not checked in the free run.