Table 2.

Selected patient-centered digital health interventions for primary care (direct engagement).

First author, year	Study design	Description of technology	Sample size	Selected outcomes
Grant, 2015 [97]	RCTa	Informatics surveillance and reminder system connected to EHRb lab test orders that generates mailed letters requesting patient completion of labs for hyperlipidemia, diabetes, and HTNc monitoring.	4038 patients	aHRd 1.26 (95% CI 0.99-1.62) for decreased time to LDLe goal. aHR 1.15 (95% CI 1.01-1.32) for earlier LDL lab assessment.
Hess, 2014 [110]	Observational cohort	PHRf delivering active notifications regarding gaps in preventive chronic disease monitoring until patient logs on to the PHR or closes the prevention gap.	584 patients	58% of all prevention gaps were closed over 12 months. 61% of notified patients accessed the PHR or closed the triggering care gap after the 1st message and 73% after the 2nd message.
Hojat, 2020 [112]	Controlled trial	EHR bulk-ordered HCVg antibody testing plus automatic PHR messages requesting patients to go to the lab.	1024 patients	14% increase in completed HCV tests (P<.001; ORh 1.7, 95% CI 1.2-2.1). Only 3.5% of patients responded to PHR messages, and repeat messaging had no effect on completion.
Langford, 2019 [138]	Observational cohort	SMS text message contact to help underserved patients with diabetes find their optimal basal insulin dose.	113 patients	84% of patients reached optimal insulin dose. Age, copay status, and initial fasting blood glucose were significantly associated with 100% SMS response (P≤.03).
Mehta, 2018 [163]	RCT	Patient portal message containing either opt-in or opt-out for FITi colorectal cancer screening test.	127 patients	28% higher FIT completion rate for patients receiving opt-out messages.
Quanbeck 2018 [185]	Observational cohort	Patient discussion board, interactive modules for health tracking, and self-management and coping with cravings for addiction management. Clinician web portal for patient-generated data.	268 patients	44% reduction in risky drinking days (P=.04), and 34% reduction in illicit drug use days (P=.01), over 12 months. 53%-60% of patients accessed the intervention during the final week of the implementation period.
Smallwood, 2017 [211]	RCT	Patient portal decision support tool for fracture risk and prevention. Includes educational information, risk calculation, and a treatment decision values elicitation exercise.	50 patients	Improved decision quality (P<.001) and conflict (P<.001) scores after the intervention. 25.7% (P=.046) increase in treatment decisions 3 months after the intervention.
Turvey, 2016 [234]	RCT	Patient portal link to a downloadable and printable CCDj for sharing with non-VAk providers for continuity of care.	52 patients	73% increase in the proportion of patients sharing the CCD with non-VA providers with training on accessing the CCD (P<.001). No improvement in medication reconciliations, but significant reduction in duplicate laboratory tests ordered by non-VA providers (P=.02).
Woo, 2016 [241]	Pilot	Daily customized spinal cord injury/disorder disease management questions delivered to patients via a data messaging device. Provider web portal with patient responses and risk level ratings.	33 patients	Average total response rate of 56%, ranging from 10% to 93%. Nearly 20% decrease in the DUSOIl score over 6 months.
Yakovchenko, 2019 [245]	RCT	Customized SMS reminder messages about HCV treatment appointments, labs, adherence, and motivation.	71 patients	Lower distress about failing treatment (P=.05) and better medication adherence (P=.06). 96% of texters vs 94% of nontexters achieved SVRm.

^aRCT: randomized controlled trial.

^bEHR: electronic health record.

^cHTN: hypertension.

^daHR: adjusted hazard ratio.

^eLDL: low-density lipoprotein.

^fPHR: personalized health record.

^gHCV: hepatitis C virus.

^hOR: odds ratio.

ⁱFIT: fecal immunochemical test.

^jCCD: continuity of care document.

^kVA: Veterans Affairs.

^lDUSOI: Duke Severity of Illness Checklist.

^mSVR: sustained virologic response.