Table 3.

Overview of system designs used by clinical decision support systems for telemonitoring.

Study (intervention), patient characteristics	Clinical decision support system							Outcomes of intervention
	Software/hardware	Source of input	Input	Task	Output (content)	Output users (ie, decision-makers)
Fried et al [32] (TRIMa), men aged ≥65 years with polypharmacy (diabetes mellitus, hypertension, and others)	Web application/ computer	EHRsb; patient telephone assessment	Patient cognition, social support, potential overtreatment, renal dosing, patient reports of adverse drug events	Evaluate medical appropriateness; generate algorithms for medication management	Patient-specific medication management (complete medication reconciliation, recommendations for discontinuation or dosage changes, recommendation of simplified regimen for patient with poor drug compliance)	Clinicians	Higher patient involvement rate; more active patient communication; more facilitative clinician communication; more medication-related communication; more correction of medication discrepancies; no reduction of medication number
Peleg et al [37] (MobiGuide), atrial fibrillation patients	App/smart phone	Personal health records (EHRs, sensors, patient-reported symptoms, data)	Patient status, symptoms, ECGc measurement, blood pressure, international normalized ratio (taking warfarin), weight and exercise	Analyze data from ECG; map computer-interpretable guidelines to health record data	Patient recommendation, notification for patients (ECG, medications), notification for provider	Patients, clinicians	Higher compliance to ECG measurements and blood pressure measurements; more clinicians change their diagnosis and treatment
Peleg et al [37] (MobiGuide), women aged 30-40 years with gestational diabetes mellitus (with or without hypertension)	App/smart phone	Personal health records (EHRs, sensors, patient-reported symptoms)	Patient status, symptoms, blood pressure, blood glucose level, ketonuria, exercise	Analyze data; map computer-interpretable guidelines to health record data	Patient recommendation, notification for patients, notification for provider	Patients, clinicians	Higher compliance to blood glucose levels, ketonuria, blood pressure; more clinicians start insulin therapy earlier; small number of patients received insulin therapy earlier
Schiff et al [33], patients aged >18 years who newly received target medication with >7 doses	Interactive voice response/ telephone	EHRs, patient interviews	Patient status; number of primary care visits; target drug; history of diabetes mellitus, hypertension, depression, and insomnia; target medication adherence; drug-specific symptoms	Detect a positive symptom of adverse drug events	Adverse drug event alert	Pharmacists	More adverse drug event documentation in medical record; slightly more medication discontinuation with reasons

^aTRIM: Tool to Reduce Inappropriate Medications.

^bEHR: electronic health record.

^cECG: electrocardiogram.