Table 2.

Results from Final Sample of Articles (n = 27)

Authors/Purpose	Design & Sample	Intervention(Comparison/Control)	Main Outcomes	Main Findings
Branson et al.33 Examine effects of TM appointment reminders to improve mental health treatment attendance	Quasi-experiment, post-test only; Measurement: at 3-mos 48 patients (female, n = 24) in outpatient child mental clinic for low-income ethnic minority youth in New York City; (40% African American, 46% Latino, 14% multi-racial/other)Mean age: 15.1 yrs (± 1.5)	IG (n =24): Received TM reminders for time/date of next appointment CG (n = 24): Historical control group from same clinic & time period; received no reminder TM	Treatment attendance Self-reported satisfaction with intervention Presence of psychiatric disorder Reminder outcomes	Attendance rates higher for IG than CG; after controlling for demographic & mental health factors, this difference remained significant (p = .02) IG received TM reminders before 88% (226/257) of sessions Most (82%–100%) participants reported satisfaction with TM reminders
Brown et al.16 Evaluate TM for delivering a health promotion intervention to adolescent, minority mothers	Qualitative interviews; Measurement: 1 interview/mo × 6 mos 5 females attending a supplemental nutrition program in a Midwestern U.S. state (3 African American & 2 Latino) Mean age: 18.2 yrs (± 0.84)	IG: Received essentials for postpartum care information by TM weekly × first 6 mos postpartum; educational content in the form of text &/or pictures	Intervention evaluation Intervention impact	4 themes identified: social support; gaining information to overcome barriers; parenting validation; fit & benefits of using mobile phone for intervention Positive impact: all mothers provided breast milk to their children; 100% adherence to childhood immunization; all infants met well-baby care guidelines by 6 mos
Carroll et al.21 Assess feasibility & acceptability of a mobile phone glucose monitoring system for adolescents with diabetes & their parents	Single-subject experiment Measurements: at 3 & 6 mos 39 patients (female, n =19) seen in an Indiana adolescent diabetes clinic Age range: 13–19 yrs	IG: Received a Glucophone smartphone × 6 mos to enable testing/reporting blood glucose levels & interaction (via TM & voice call) with a nurse practitioner	System usability & satisfaction	TM helped participants remember to check their blood sugar
Chandra et al.17 Evaluate acceptability & feasibility of TM for promoting positive mental health & as a helpline among adolescent females	Single-subject experiment; Measurements: at 1 mo & 1 mo after study conclusion 40 females living in urban India Mean age: 16.8 yrs (± 1.68)	IG: Received 1 TM/day × 1 mo regarding positive mental health or helpline information; participants could call/text back with question or concerns	TM intervention perceptions	62% liked receiving the TM; 50% said the TM made them feel happy 8% faced family objections about TM 62% preferred helpline TM over moodlifting TM
Cornelius et al.41 Examine a HIV prevention intervention delivered via mobile phones to adolescents	Single-subject experiment; Measurements: at baseline, 7 & 19 wks 40 African Americans (female n = 21) recruited from community organizations & schools in a Southeastern U.S. state Mean age: 15.4 yrs (± 1.7)	IG: Attended weekly in-person meetings × 7 wks; then received daily multimedia TM to serve as “boosters” × 3 mos	TM evaluation HIV-related knowledge & attitudes	HIV knowledge increased after inperson meetings; no change from completion of meetings to conclusion of TM No change in attitudes toward condoms after meetings or TM Increased confidence in avoiding HIV after receiving TM No change in HIV risk behaviors over time Participants reliably responded to TM 97% said number of TM was “just right”
Dewar et al.10 Evaluate impact of a school-based multi-component program (NEAT Girls) on adolescent girls’ PA & sedentary behaviors	RCT; Measurements: at baseline & 12 mos 357 girls attending 12 secondary schools in low-income communities in Australia (ethnicity not reported) Mean age 13.2 yrs (±.50)	IG (n = 178; 6 schools): Received enhanced PA sessions, interactive seminars, student handbooks, nutrition workshops, pedometers, parent newsletters; TM to encourage PA, healthy eating, & decreased sedentary behavior. CG (n = 179; 6 schools): Wait-list CG	PA Sedentary behaviors Social-cognitive mediators	No group by time interactions for PA or social cognitive mediators Greater reductions in recreational computer use (p = .02) & sedentary activity (p = .04) in IG than CG
Fabbrocini et al.22 Evaluate adherence to therapy in acne patients using mobile phones & TM	RCT; Measurement: at baseline & after 12 wks 160 patients (female, n = 87) enrolled from outpatient acne service (ethnicity not reported) Mean age 19.5 yrs (IG) 18.5 yrs (CG)	IG (n = 80): Received 2 TM addressing acne × 2/day × 12 wks CG (n = 80): Did not receive TM	Adherence QOL Satisfaction with TM	Greater increases in adherence to treatment (p <0.0001) & improvement in QOL (p <0.0001) in IG than CG 95% of participants were “very much” or “quite” satisfied with TM
Haug et al.26 Test efficacy of an individually tailored TM intervention for smoking cessation in youth	RCT; Measurements: at baseline & at 6 mos 755 students (female, n = 392) attending schools in Switzerland who were smokers at baseline Mean age 18.2 yrs (± 2.3)	IG (n = 372; 90 classes): Received online assessment of individual smoking behavior, weekly TM assessment of smoking-related target behaviors, 2 weekly tailored TM, & integrated quit day preparation & relapse-prevention program CG (n = 383; 88 classes): Did not receive the intervention	Smoking behavior change	No significant difference in 7 day abstinence rates, stage of change, or quit attempts Decreased mean number cigarettes smoked/day greater in IG than CG (p = .002)
Haug et al.27 Test appropriateness & effectiveness of an individually tailored TM intervention to reduce problem drinking in vocational school students	Single-subject experiment; Measurements: at baseline & 12 wks 477 students (female, n = 111) attending 7 schools in Switzerland; 72% reported ≥1 instances of RSOD in past 30 days Mean age 18 yrs (±2.4)	IG: Received 1–2 tailored TM/wk; TM tailored for age, gender, number of standard drinks per wk, & RSOD	RSOD behavior change	Decreased percentage had RSOD within the last mo from baseline for at least 1 RSOD occasion (p <.001) & > 2 RSOD occasions (p = .01) Decreased number of drinks in a typical wk (p = .002), percentage with 1+ alcohol-related problems in the last 3 mos (p = .009), & maximum number of drinks on a single occasion (p = .08)
Herbert et al.34 Investigate adolescents’ use of a diabetes TM program & determine whether certain groups more likely to respond to TM	Single-subject experiment; Measurements: at baseline & 6 wks 23 adolescents (female, n = 11) with diabetes from a Mid-Atlantic U.S. state (78% White) Mean age 15.13 yrs (± 1.14)	IG: Received 2 TM/day for majority of intervention; TM included information/tips & a request to respond to a specific question;TM topics included blood glucose monitoring, nutrition, PA, & sleep/mood	TM evaluation Glucose monitoring	Participants responded to 78% of TM; most to nutrition TM, least to blood glucose TM Correlation between females & overall TM response rate & number personal TM sent/day (p < .05) Trend for participants with lower blood glucose to respond to more TM (p =.08).
Hingle et al.23 Evaluate a skin cancer prevention TM intervention among adolescents	Single-subject experiment; Measurements: at baseline &12 wks 113 adolescents (female, n =60) from 3 Arizona middle schools who had completed a sun safety education program 2 wks prior to enrollment (ethnicity not reported) Age range: 11–14 yrs	IG: Received 3 TM/wk × 13 wks; TM addressed skin cancer risk, sun protection benefits, & beliefs inconsistent with public health recommendation	Sun safety behavior, knowledge & attitudes	Increased self-reported use of sunscreen (p = .001), hats (p = .02), & sunglasses (P = .02) Greater consideration of sun avoidance during peak hours (p = .02) Increase overall skin cancer knowledge (P = .03)
Huang, Dillon et al.35 Compare a tailored versus generic weight management intervention among adolescent survivors of childhood leukemia	RCT; Measurements: at baseline & 4 mos 38 overweight survivors (female, n = 23) off therapy for 2+ yrs recruited from a clinical trial (89.5% Hispanic, 2% black) Mean age 13 yrs (range 10–16)	IG: (n=18): Received Web-&-TM information (tailored TM & queries) & weekly (mo 1) to biweekly (mos 2–4) counseling-based intervention CG (n=17): Received printed materials; biweekly phone call	Weight/BMI PA Dietary intake Depression	IG demonstrated greater, but not statistically significant, change in weight across study period compared to CG (p = .06) & no difference in changes in BMI, PA, or daily calories consumed IG reported reduced negative mood over time compared to CG (p = 0.01)
Huang, Terrones et al.36 Evaluate improved generic, Internet & mobile phone-delivered intervention on disease management, self-efficacy, & communication	RCT; Measurements: at baseline, 2 mos, & 8 mos 81 patients (female, n = 44) with inflammatory bowel disease, cystic fibrosis, diabetes seen at tertiary care pediatric center (49% Hispanic, 9% black, 1% Native American) Mean age 17 yrs (range 12–20)	IG: (n=38): Received access to a website for disease management, communication skills, & lifestyle tips + tailored TM CG: (n=37): Received monthly email messages on general health issues	Disease self-management Health-related self-efficacy	Group × time interaction for disease self-management (p = .02) & self-efficacy (p = .02); IG reported increased self-management & self-efficacy, but CG remained constant
Juzang et al.18 Evaluate a TM HIV prevention program among young adults	Quasi-experiment; Measurements: at baseline, 3, & 6 mos 60 young black men in Philadelphia Median age 17 yrs (CG) 19 yrs (IG)	IG (n=20): Received HIV prevention TM CG (n=19): Received nutrition TM Both groups: TM designed to increase positive outcome expectancies, norms, self-efficacy & intentions for condom use; received TM ×3/wk × 12 wks	Sexual health knowledge, awareness, & risk-prevention behavior	Greater condom norms & sexual health awareness in CG than IG at all time points (no p values given) No changes in condom use intention
Lana et al.28 Assess impact of a Web-based intervention supplemented with TM to reduce TCBR	RCT; Measurements: at baseline & 9 mos 2001 adolescents (female, n = 1081) attending Spanish & Mexican schools Age range: 12–16 yrs	IG 1(n=177): Received access to website with information about main cancer risk behaviors IG 2 (n= 244): Same as IG1 + weekly TM to encourage adherence CG (n=316): Not specified	Weight BMI (Kg/m2) Diet behavior: TCBR (smoking, unhealthy diet, alcohol use, obesity, sedentary lifestyle, sun exposure)	TCBR scores reduced in all groups with significant drop in IG-1 & IG-2, but not CG (no p values reported) IG-2 intervention increased the probability of improving post-test TCBR score & giving up at least 2 risky behaviors
Lau et al.29 Evaluate Internet & TM intervention for promoting PA among adolescents	Quasi-experiment; Measurements: at baseline & 8 wks 78 Chinese school children (female, n = 51) in Hong Kong Mean age IG 12.29 yrs (± 0.87) CG 13.26 yrs (± 1.14)	IG (n=38): Received Internet- & stages of change-based PA program × 2/wk & daily TM on weekdays 5 TM types: motivational, informational, behavioral skills, reinforcement of PA benefits, & solutions for PA barriers CG (n=40): No intervention	PA level over last 7 days SMR	No time × condition interactions for PA or SMR; significant increases in PA (p = .05) & SMR (p = .01) in IG but not in CG Positive correlation between number of TM read & SMR (p < .01)
Love-Osborne et al.30 Evaluate feasibility of adding a health educator to school-based health center teams to deliver preventive services for overweight adolescents (TM used to reinforce goals between visits)	RCT; Measurements: at baseline & 9 mos 165 adolescents (female, n = 86) with BMI ≥85% recruited from 2 centers (88.5% Hispanic) Mean age 15.7 yrs (± 1.5)	IG (n=77): Received MI with goal-setting plus 2 TM/wk (1 individualized goal-related & 1 reminder to turn in log) CG (n=72): No intervention	Self-monitoring: of weight & lifestyle behaviors BMI (both groups) Cardiovascular fitness (IG)	Greater proportion of CG decreased or maintained a stable BMI than IG (p =.025); Greater proportion of CG decreased BMI z-score by .1 or more than CG (p =.02) Sports participation higher in CG than IG (p = .02)
MacDonell et al.37 Assess feasibility of using ecological momentary assessment via TM on personal cell phones to measure medication adherence	Single subject experiment; Measurement: at 14-days 16 African Americans (female, n = 7) with asthma enrolled from a large, urban hospital & university student health center Mean age 19.75 yrs (± 1.77)	IG: Received TM daily to prompt a response about asthma medications/symptoms; sent event-based TM when they experienced asthma symptoms or took asthma rescue/controller medications	Asthma control Medication adherence Asthma symptoms Rescue medication use	Adherence to controller medication from 0 – 14 days during the TM trial (M =8.69 ± 5.39)Asthma-related symptoms or limitations reported 28.1% of trial days but use of rescue medications 18.8% of days Responded to 78.5% of all time-based TM with a relevant response
Mulvaney et al.25 Improve diabetes adherence using individually tailored TM	Quasi-experiment; Measurement: at baseline & 3 mos 28 adolescents (female, n = 15) with diabetes enrolled through a diabetes clinic (6% African American, 1% Hispanic & Pacific Islander) Mean age IG 15.9 yrs (± 2.9) CG 15.8 yrs (± 2.7)	IG (n=23) Received TM tailored to participant’s reported top 3 barriers to adherence; 8–12 unrepeated TM/wk CG (n not stated): Historical controls from same clinic; matched with IG on age, gender, & HgbA1c values	Glycemic control System usability & satisfaction	Interaction between group × time (p < .01) for HgbA1c; values in IG were unchanged, but increased in CG. High system usability & satisfaction
Nguyen et al.19 Evaluate effectiveness of additional therapeutic contact as an adjunct to an extended weight-loss maintenance intervention	Randomized trial (no control group); Measurement: at baseline, 12, & 24 mos 161 overweight & obese adolescents (gender not reported) seen in New Zealand community health centers (ethnicity not reported) Age range 13–16-yrs	IG 1 (n=78): Received Loozit program including 7 × 75-min weekly group sessions; maintenance of 5×60-min quarterly adolescent booster group sessions IG 2 (n=79): Received Loozit + additional contact every 2 wks (overall 14 telephone coaching sessions & 32 TM &/or email messages)	Baseline to 24-mos changes in BMI z-scores & waist: height ratio Psychosocialwell-being Dietary intake PA	No statistically significant group effects or group-by-time interactions for primary outcomes & very few for secondary outcomes From baseline to 24 months, reductions in BMI & triglycerides in IG2 (p < .05).
Rhee et al.38 Develop & evaluate a comprehensive mobile phone-based asthma self-management aid for adolescents (mASMAA)	Descriptive study with focus groups component; Measurement: at 2 wks Adolescents with asthma & their parents (16 dyads) (female, n = 6) recruited from emergency department & primary care clinics in a university medical center (40% black, 7% Asian) Mean age: 15 yrs (± 1.5)	Intervention (2 wks); Received TM at a time chosen by each adolescent based on preference & medication schedule; adolescents encouraged to initiate asthma-related TM at least ×2/day	Asthma symptoms Asthma control level Activity level Frequency of rescue & other medication use & asthma control	60% of adolescents experienced uncontrolled asthma for 2 + days during study Each adolescent on average submitted 19 self-initiated TM (range 3–38) regarding symptoms (69%), activity-(48%), & medication (10%) or 2 or more categories (29%) Intervention increased awareness of symptoms & triggers, improved asthma self-management, & medication adherence
Seid et al.20 Evaluate an intervention that integrates MI, problem solving skills training, & TM for adolescents with asthma	RCT; Measurements: at baseline, 1 & 3 mos 29 adolescents (gender not described) with moderate & severe asthma) enrolled from an Ohio hospital (76.9% African American) Mean age 15.76 yrs (± 1.67)	IG (n=12): Received 2 brief in person sessions 1 wk apart (asthma education, MI, problem skills training) & 1 mo of tailored TM CG (n=14): Received asthma education but no tailored TM	Manipulation Mechanism of effect Efficacy (asthma symptoms, HRQOL)	All participants found intervention appealing & acceptable At 1 & 3 mos, motivation, intentions, asthma symptoms, & barriers had clinically meaningful Cohen’s d medium to large effect sizes (.5–.96)
Shi et al.31 Test TM smoking behavior intervention to increase self-reported smoking abstinence & reduce daily cigarette consumption among adolescents	RCT; (cluster randomization of 6 schools); Measurements: at baseline & 12 wks 179 Chinese adolescent smokers (female, n = 8) attending 6 vocational high schools Age range 16–19 yrs	IG (n=76): Received daily tailored TM, interactive communication, & adjuvant online support CG (n=46) Received a self-help pamphlet	Smoking cognitions, attitudes, & behaviors	Attitude toward disadvantages of smoking & mean reduction of cigarettes/day higher in IG (p < .01) Intervention effectively inhibited cigarette (p < .01) & nicotine dependence (p = .04) psychologically Difference in number of TM sent to investigator not significant No difference between groups in self-reported 7- & 30-day tobacco abstinence
Shrier et al.39 Evaluate the MOMENT intervention for marijuana use cessation among adolescents	Single-subject experiment; Measurements: at baseline, 2 & 17 wks 27 patients (female, n = 19) in 2 adolescent clinics in a Northeastern U.S. city (44% black, 37% Hispanic) Median age 19 yrs (range, 15–24)	IG (n=16) Received 2 brief motivational enhancement therapy sessions; 2 wks of mobile reports with TM supporting self-efficacy & coping strategies	Marijuana use & desire	TM motivated non-use of marijuana; were interesting, motivating, & helpful Average use events/day declined over the study Desire to use during & after a triggering context decreased from baseline to 3-mo follow-up (p < .0001 & p = .03, respectively) Non-significant change in motivational scale scores
Skov-Ettrup et al.24 Compare 2 versions of an Internet- & TM-based smoking cessation intervention	RCT; Measurement: at baseline & 12 mos 2,030 newly registered users of xhale.dk (female, n = 1204) (ethnicity not reported) Mean age IG 19.4 yrs (± 3.1) CG 19.5 yrs (± 3.2)	IG 1: Untailored intervention (n=371): Received TM about smoking cessation sent once daily × 5 wks; weekly TM × next 3 wks IG 2 Tailored intervention (n=383): Received weekly TM 4 wks before quit date & daily TM 1–3 days before quit date; then 2 tailored TM/day × 4 wks; then 4–5 TM/wk × 4 wks	Smoking cessation perceptions & behavior	79.8% chose to receive supporting TM No significant difference between IG-1 & IG-2 in changes in self-efficacy & beliefs about smoking from baseline to 12 mos follow-up
Ting et al.40 Investigate the effects of TM reminders on adherence to clinic visits & use of HCQ among adolescents with lupus	RCT; Measurement: at baseline & 14 mos 70 patients (female, n = 65) in a lupus registry with childhood-onset systemic lupus erythematosus with unlimited access to TM (36% black)Mean age 18.6 yrs (± 2.5)	All participants received visit adherence intervention: TM reminder sent 7, 3, & 1 day (s) prior to appointment IG 1: Received printed information about HCQ benefits & side effects IG 2: Received printed information + a standardized daily TM reminder regarding HCQ intake	Clinical visit & medication adherence	19% of patients were nonadherent to clinic visits at baseline; among them, there was improved visit adherence during the TM intervention (p = 0.01)After IG1 concluded, adherence rates declined (p = 0.02), but rates remained higher compared to baseline (p = 0.005 Medication adherence poor in more than two-thirds of cohort based upon HCQ blood levels, self-reports, & pharmacy refill data
Whittaker et al.32 Test mobile delivery of a depression prevention intervention for adolescents	RCT; Measurement: at baseline & 12 mos 855 students (female, n = 584) (White European, Asian, Maori, Pacific Islander ethnicities) in New Zealand schools Mean age 14 yrs (range 13–17)	IG: Received 2 TM (mixed formats)/day × 9 wks based on cognitive-behavioral therapy, followed by monthly TM & access to a mobile website CG: Received non-depression focused message (e.g., environment sustainability, cybersafety)	Depression incidence Program perceptions	Perceptions of being more positive & ridding of negative thoughts higher in IG vs CG (p < .001) 82.4% of participants reported finding the intervention to be useful

Common abbreviations: BMI = body mass index; CG = control/comparison group; HRQOL = health-related quality of life; HCQ = hydroxychloroquine; IG = intervention group; MI = motivational interviewing; PA = physical activity; QOL = quality of life; RCT = randomized controlled trial; RSOD = risky single-occasion drinking; SMR = Stage of motivational readiness; TCBR = total behavioral cancer risk; TM = text messages(ing); mo(s) = months; wk(s) = week(s); yr(s) = year(s)