With the rapid advancement of technology in healthcare, there has been an increase in the growth of software technologies created for the purpose of improving healthcare delivery. The U.S. Food and Drug Administration (FDA) refers to these as “device software functions,” a category of software that also includes mobile medical applications (MMAs), which may be deployed on various platforms (e.g., mobile platforms, other general-purpose computing platforms, or in the operation or control of a hardware device), and are designed to enable consumers to better manage their health and well-being, assist healthcare providers to improve and facilitate patient care, diagnose a condition, or trigger a command necessitating patient action. Some examples of the previously mentioned MMA functionalities include the Radiation Emergency Medical Management (REMM) app and the National Institute of Health's LactMed (FDA, 2021b).

Additionally, the Center for Devices and Radiological Health, which functions within the FDA, takes a customized, risk-based approach with a priority on the subset of software functions that qualify under the regulatory definition of "device" and ensure those that have a potentially greater risk must require FDA review (FDA, 2021f). Furthermore, software functions that the FDA specifies as device software functions requiring regulatory oversight include:

The FDA will not require manufacturers to submit premarket review applications or registration their software with the FDA for software functions that qualify under the regulatory definition of a "device" when such software functions pose a minimal risk to patients and consumers. Software functions that belong to this FDA discretionary approach include functions that are as follows (FDA, 2021f):

More recently, a novel therapeutic class referred to as prescription digital therapeutics (PDTs) have entered into the digital healthcare space. This therapeutic class is different from other traditional health and wellness apps in that it possesses the following unique characteristics (Digital Therapeutics Alliance, 2021):

In order to provide regulatory oversight for software-based medical devices that is both streamlined and efficient, the FDA launched the Pre-Cert Pilot Program test phase in 2019. "In the Pre-Cert program, the FDA is proposing that software products from precertified companies would continue to meet the same safety and effectiveness standard that the agency expects for products that have followed the traditional path to market." A proposed aim is to focus on the software developer or digital health technology developer, rather than mainly on the product. Additionally, "the FDA's Total Product Lifecyle (TPLC) approach enables the evaluation and monitoring of a software product from its premarket development to postmarket performance, along with continued demonstration of the organization's excellence." Proposed key components of the FDA's TPLC methodology include the following (FDA, 2021d):

In September 2017, the following 9 companies out of over 100 candidates were chosen by the FDA to participate in the development of the Software Pre-Cert Pilot Program: Apple, Fitbit, Johnson & Johnson, Pear Therapeutics, Phosphorous, Roche, Samsung, Tidepool, and Verily (FDA, 2021d).

Other professional organizations such as the American Medical Association, American Psychiatric Association, and the Academy of Managed Care Pharmacy are also beginning to develop a framework and provide guidance to healthcare practitioners as they begin to integrate mobile health technologies, mobile apps, and digital therapeutics as a component in the delivery of patient care.

BlueStar Rx

WellDoc (Columbia, MD) developed the BlueStar Rx System which is indicated for use by healthcare providers and their patients who are 18 years of age and older to aid in their self-management of type 1 or type 2 diabetes. The BlueStar Rx System is an FDA-cleared software app that is complimentary to the patient's current therapies (e.g., pharmacologic, diet, exercise, and counseling). Patients can use the mobile app or the web version of BlueStar. The software app includes an always-on, fully-automated software coach that sends a report to the patient's healthcare provider team via facsimile, email, or electronic medical record and a patient portal is managed by an administrator who can manage, review, report, survey and communicate with the patient. In addition to reporting blood glucose results and supporting medication adherence, the BlueStar Rx System delivers coaching messages based on current time blood glucose results and trends. A prescription is required by a licensed healthcare professional for the BlueStar Rx system which also includes an insulin dose calculator that enables patients to use their prescribed regimen to determine insulin dosage based on a given amount of carbohydrates and/or blood glucose values. WellDoc states on their website that "The BlueStar Rx System is not intended to replace the care provided by a licensed healthcare professional, including prescriptions, diagnosis, or treatment" (Digital Therapeutics Alliance, 2021b; WellDoc, 2021).

Quinn and colleagues (2011) conducted the Mobile Diabetes Intervention Study, a cluster-randomized clinical trial to assess whether the addition of mobile application coaching and patient/provider web portals to community primary care compared to standard diabetes management would decrease glycosylated hemoglobin levels in patients with type 2 diabetes. This study randomly assigned 26 primary care practices consisting of 163 participants to one of three stepped treatment groups or a control group (usual care). The primary outcome was a change in glycated hemoglobin levels over a 1-year treatment duration and secondary outcomes included changes in patient-reported diabetes symptoms, diabetes distress, depression, and other clinical (blood pressure) and laboratory (lipid) values. Maximal treatment included a mobile- and web-based self-management patient coaching system and provider decision support. Automated, real-time educational and behavioral messaging were sent to patients via mobile phone in response to individually analyzed blood glucose results, diabetes medications, and lifestyle behaviors. Quarterly summary reports regarding patient’s glycemic control, diabetes medication management, lifestyle behaviors, and evidence-based treatment options were sent out to providers. Results included 1.9% mean declines in glycated hemoglobin in the maximal treatment group and 0.7% in the usual care group, a difference of 1.2% (p<0.001) over 12 months. Significant differences were not noticeable between groups for patient-reported diabetes distress, depression, diabetes symptoms, or blood pressure and lipid levels (all p>0.05). The investigators concluded that the combination of behavioral mobile coaching with blood glucose data, lifestyle behaviors, and patient self-management data individually evaluated and presented with evidence-based guidelines to providers significantly decreased glycosylated hemoglobin levels over 1 year.

Agarwal and colleagues (2019) evaluated BlueStar mobile app, an FDA-approved mobile prescription therapy, to determine if app usage results in improved hemoglobin A

(HbA

) for diverse participants in real-life clinical contexts. The study involved of a multicenter pragmatic randomized controlled trial consisting of 110 participants randomized to the immediate treatment group (ITG) receiving the intervention for 6 months, and 113 participants randomized to the wait-list control (WLC) group receiving usual care for the first 3 months and then receiving the intervention for 3 months. The primary outcome was glucose control measured by HbA

levels at 3 months and secondary outcomes determined intervention impact on patient self-management, experience of care, and self-reported health utilization using validated scales (i.e., the Problem areas in Diabetes, the Summary of Diabetes Self-Care Activities, and the EuroQo1-5D). The BlueStar mobile app captured the intervention usage data. The results did not show evidence of intervention impact on HbA

levels at 3 months (mean difference [ITG-WLC] -0.42, 95% Confidence Interval [CI] -1.05 to 0.21; p=0.19). Additionally, no intervention effect on secondary outcomes measuring diabetes self-efficacy, quality of life, and healthcare utilization behaviors were observed. Significant variation in app usage by site was noted such that participants from one site logged in to the app a median of 36 days over 14 weeks (interquartile range [IQR] 10.5-124), whereas participants at another site showed a notable decrease in app usage (median 9; IQR 6-51). The investigators concluded that there was no difference between intervention and control arms for the primary outcome of glycemic control measured by HbA

levels and the low usage of the app among participants warrants further study of patient and site-specific factors that increase app usage.

Canvas Dx

Cognoa (Palo Alto, CA) developed Canvas Dx which is an FDA-cleared software medical device that is indicated for use by healthcare providers as an aid in the diagnosis of Autism Spectrum Disorder (ASD) for patients ages 18 months through 72 months who are at risk for developmental delay based on concerns of a parent, caregiver, or healthcare provider. Canvas Dx utilizes a clinically validated artificial intelligence (AI) technology that integrates three separate user-friendly inputs. The inputs include a parent/caregiver questionnaire regarding the child's behavior and development collected via a parent/caregiver facing app, a questionnaire completed by a video analyst who reviews parent/caregiver recorded videos of the child, and a healthcare provider questionnaire completed by a physician during child and parent/caregiver interaction via a healthcare provider portal. A device output is then generated after an algorithm evaluates all of these inputs which will be used by the physician in addition to their clinical judgement. The device is by prescription only and Cognoa states on the www.canvasdx.com website "The device is not intended for use as a stand-alone diagnostic device but as an adjunct to the diagnostic process" (Canvas Dx, 2021; Cognoa, 2021).

Abbas and colleagues (2017) applied machine learning (ML) to gold standard clinical data captured across thousands of children at-risk for autism spectrum disorder to develop a low-cost, quick, and easy to use autism screening tool. Two algorithms to identify autism, included one based on short, structured parent-reported questionnaires and short, semi-structured home videos of children identifying key behaviors which are then combined in an algorithm to yield a single assessment of higher accuracy. The performance of these algorithms and their combination was assessed in a multicenter clinical study comprised of 162 children. While significant accuracy improvement compared to standard screening tools in measurements of AUC, sensitivity, and specificity was demonstrated, the authors discuss a myriad of confounding factors in the clinical analysis and also note the results are statistically limited. Additional clinical studies are warranted to firmly support the findings of this study that a mobile, machine learning process can be a reliable method for detection of autism outside of clinical settings.

Abbas and associates (2020) evaluated a multi-modular, machine learning-based assessment of autism via a mobile app in a blinded, multi-site clinical study comprised of 375 children who were 18 to 72 months of age. The machine learning-based assessment of autism consisted of three complimentary modules for a unified outcome of diagnostic-grade reliability. The complimentary modules (i.e., Cognoa assessment modules) included a 4-minute, parent-report questionnaire presented via a mobile app, a list of key behaviors identified from 2-minute, semi-structured home videos of children, and a 2-minute questionnaire presented to the clinician at the time of clinical assessment. The results demonstrated that the machine learning-based assessment outperformed baseline autism screening assessments (i.e., the Child Behavior Checklist [CBCL], the Modified Checklist for Autism in Toddlers, Revised [M-CHAT-R], and the Social Responsiveness Scale – Second Edition [SRS]) administered to children by 0.35 (90% Confidence Interval [CI]: 0.58 to 0.81) in specificity when operating at 90% sensitivity. Additionally, in children less than 48 months of age, the investigators’ machine learning-based assessment outperformed the most accurate baseline screening assessment by 0.18 (90% CI: 0.08 to 0.29 at 90%) in AUC and 0.30 (90% CI: 0.11 to 0.50) in specificity when operating at 90% sensitivity. The investigators discuss the limitations of the study, including that the children preselected have a high risk of autism, and that there is a need to validate this new machine learning-based assessment in the primary care clinic setting.

d-Nav Insulin Management Program

Hygieia offers the d-Nav Insulin Management Program, a digital therapeutic which is indicated for adult patients with type 2 diabetes who manage their condition with insulin injections. The d-Nav Insulin Management Program combines an FDA-cleared software mobile app enabled by AI technology, and virtual clinical support to make autonomous (algorithm-based) adjustments to insulin doses based on the patient’s glucose levels. Hygieia embeds the d-Nav application onto a blood glucose monitor and the technology is activated when there is a prescription. The d-Nav program relies on use of a handled device such as a smartphone or a device provided by Hygieia. Patients use the d-Nav technology before every insulin injection by entering their most recent glucose reading, and then receive a personalized dose recommendation on their handheld device. Data is sent to the cloud where it is monitored by d-Nav Care Specialists who are available for support.

Bergenstal and colleagues (2012) hypothesized that the Diabetes Insulin Guidance System (DIGS™) (Hygieia, Inc., Ann Arbor, MI) software, which automatically advises patients on adjustment of insulin dosage, would provide safe and effective weekly insulin dosage adjustments. The authors conducted a 16-week feasibility study (designed as a prospective, open-label, uncontrolled, single-arm, single-center study with intention-to-treat) in which they enrolled patients (n=46) with type 1 (n=20) and type 2 diabetes (n=26 ) treated with a variety of insulin regimens and having suboptimal glycemic control. Thirty-eight patients completed the study. The 12-week intervention period followed a 4-week baseline run-in period. During the intervention, DIGS processed patients' glucose readings and provided insulin dosage adjustments on a weekly basis. If approved by the study team, the adjusted insulin dosage was communicated to the patients. Insulin formulations were not changed during the study. The primary outcome was the fraction of DIGS dosage adjustments approved by the study team, and the secondary outcome was improved glycemic control. The authors found that during a cumulative period of 8.9 patient-years, the DIGS software recommended 1,734 insulin dosage adjustments, of which 1,731 (99.83%) were approved. During the run-in period the weekly average glucose was stable at 174.2±36.7 mg/dL (9.7±2.0 mmol/L). During the following 12 weeks, DIGS dosage adjustments resulted in progressive improvement in average glucose to 163.3±35.1 mg/dL (9.1±1.9 mmol/L) (p<0.03). Mean glycosylated hemoglobin decreased from 8.4±0.8% to 7.9±0.9% (p<0.05). Concomitantly, the frequency of hypoglycemia decreased by 25.2%. The authors concluded that the DIGS software provided patients with safe and effective weekly insulin dosage adjustments. Widespread implementation of DIGS may improve the outcome and reduce the cost of implementing effective insulin therapy. The authors acknowledge that their study was limited by lack of a control group. Thus, they could not have unequivocally excluded the possibility that improved glycemia resulted from participation in the study. Furthermore, the study duration was relatively short; therefore, it is possible that different HbA1c levels would have been recorded during a longer follow-up. Because weekly mean glucose levels were stable during the run-in period and improved particularly toward the middle of the active phase, it is the authors’ belief that HbA1c would have further improved. Randomized controlled trials are warranted, as well as additional studies to focus on improved glycemic balance among the minority of patients who experience frequent hypoglycemia.

Bashan and Hodish (2012) hypothesize that frequent insulin dosage adjustments based on glucose readings alone are sufficient for a safe and effective therapy. The authors conducted a three-month open-label prospective pilot study recruiting 14 subjects with suboptimal controlled insulin-treated Type-2 and Type-1 diabetes (n=11 with Type-2 diabetes, n=3 with Type-1; n=12 completed the study follow-up). Subjects were treated with basal-bolus insulin therapy that was titrated weekly for 12 weeks. Dosage adjustments were made by the study Endocrinologist by reviewing subjects' glucose readings, exclusively based on log sheets and contingent upon the approval of the on-site study team. To corroborate that the glucose readings were sufficient for making dosage adjustments, the authors used software to process only glucose readings and recommend insulin dosage adjustments. The recommendations made by the software were retrospectively compared to the ones made by the study Endocrinologist. The authors found that all 568 recommendations were approved by the study team and in 99.3% of the cases the recommendations were clinically similar to the ones made by the software. No hazardous disagreements were found. The mean A1C improved from 9.8% (± 2.0) to 7.9% (± 1.3) (p=0.001) in 12 weeks and the weekly mean glucose progressively improved from 220.3 mg/dl (± 51.9) to 151.5 mg/dl (± 19.2) (p<0.0001). The frequency of minor hypoglycemia was 22.7 per patient-year in subjects with Type-2 diabetes and 42.7 in the subjects with Type-1 diabetes. No severe hypoglycemic events occurred. The authors concluded that glucose readings are sufficient to adjust insulin therapy in a safe and effective manner, when adjustments are made on a weekly basis. Thus, dedicated software may help adjust insulin dosage between clinic visits. The authors acknowledge that the main weakness of this pilot study was its relatively small subject number and short follow-up (3.25 patient year); however, since the investigated parameter was the process of insulin dosage adjustments, the actual “N” value was considerable (n= 568).

Bergenstal et al (2019) conducted a multicenter, randomized controlled trial to determine whether the combination of the d-Nav device and health-care professional support is superior to health-care professional support alone. The investigators recruited patients (n=181) from three diabetes centers in the USA that were aged 21-70 years, diagnosed with type 2 diabetes with a glycated hemoglobin (HbA1c) concentration of 7.5 percent or higher and 11 percent or lower, and had been using the same insulin regimen for the previous 3 months. Exclusion criteria included body-mass index (BMI) of 45 kg/m2 or higher; severe cardiac, hepatic, or renal impairment; and more than two severe hypoglycemic events in the past year. Eligible participants were randomly assigned (1:1), with randomization blocked within each site, to either d-Nav and health-care professional support (intervention group, n=93) or health-care professional support alone (control group, n=88). Both groups were contacted seven times (three face-to-face and four phone visits) during 6 months of follow-up. The primary objective was to compare average change in HbA1c from baseline to 6 months. Safety was assessed by the frequency of hypoglycemic events. The primary objective and safety were assessed in the intention-to-treat population. The investigators used Student's t test to assess the primary outcome for statistical significance. At baseline, mean HbA1c was 8.7 percent (SD 0.8; 72 mmol/mol [SD 8.8]) in the intervention group and 8.5 percent (SD 0.8; 69 mmol/mol [SD 8.8]) in the control group. The mean decrease in HbA1c from baseline to 6 months was 1.0 percent (SD 1.0; 11 mmol/mol [SD 11]) in the intervention group, and 0.3 percent (SD 0.9; 3.3 mmol/mol [9.9]) in the control group (p<0.0001). The investigators found that the frequency of hypoglycemic events per month was similar between the groups (0.29 events per month [SD 0.48] in the intervention group vs 0.29 [SD 1.12] in the control group; p=0.96). The investigators concluded that the combination of automated insulin titration guidance with support from health-care professionals offers superior glycemic control compared with support from health-care professionals alone. Such a solution facilitated safe and effective insulin titration in a large group of patients with type 2 diabetes, and now needs to be evaluated across large health-care systems to confirm these findings and study cost-effectiveness.

Garg (2019) reviewed the Bergenstal et al (2019) study and reported doubts about its reproducibility in clinical practice. The author provided correspondence to The Lancet which states that “The control group seems to have received negligible clinical care from their health-care providers. The insulin dose in the control group was almost the same at the end of 6 months as at the baseline. I do not know any diabetes centre where providers will not make insulin dose adjustments in a patient with poor glycaemic control over a period of 6 months, despite seven patient contacts. Therefore, at best, it seems that the d-Nav Insulin Guidance System helps providers with clinical inertia overcome their clinical inertia. The premise that d-Nav will help in saving time for health-care providers is not proven by the results of this study”.

Drowzle Pro

Resonea (Scottsdale, AZ) developed the Drowzle Pro which is an FDA-cleared mobile application digital home sleep test. Drowzle Pro functions via a digital platform utilizing a smartphone for in-home screening of obstructive sleep apnea (OSA) in adults. The mobile software is used to gather symptom data for sleep apnea risk, including severity of daytime sleepiness and personal chronic disease risk factors. Drowzle Pro also records sleep breathing patterns and transmits the sound files to secure servers in the cloud. By being a stand-alone software medical device, The mobile application then analyzes and interprets the sleep breathing results, along with profile data provided by the patient, to measure and monitor sleep-related health risks over time. The mobile application’s deployment via the patient's own phone, enables it to be an option when polysomnography in the lab or conventional home sleep testing may not be feasible. The results assist the healthcare professional in determining the need for further diagnosis and evaluation. Furthermore, Drowzle Pro is not intended as a substitute for full polysomnography when additional parameters such as sleep stages, limb movements, or electroencephalogram (EEG) activity is required. Drowzle Pro is only available by prescription for adults 21 years of age and older (FDA, 2022a; Resonea, 2022).

EndeavorRx

Akili Interactive Labs, Inc. (Boston, MA) developed EndeavorRx which is an FDA-authorized digital therapeutic indicated to improve attention function as measured by computer-based testing in children ages 8 to12 years old with primarily inattentive or combined-type ADHD, who have a demonstrated attention issue. This digital treatment is delivered through an action video game experience and is designed to challenge a child’s attention span during gameplay with the necessary focus and flexibility to perform multiple tasks at the same moment. EndeavorRx should be considered for use as part of a therapeutic program that may consist of clinician-directed therapy, medication, and/or educational programs, which target symptoms of the disorder. Specifically, EndeavorRx is a prescription only medical device where one prescription will provide 3 months of access to this treatment. The duration of EndeavorRx daily treatments last approximately 25 minutes and should be completed by the patient without interruption (Akili Interactive Labs, 2021; Digital Therapeutics Alliance, 2021d).

In the Software Treatment for Actively Reducing Severity of ADHD (STARS-ADHD) study, Kollins and colleagues (2020) evaluated an investigational digital therapeutic, AKL-T01, for improved attentional performance in pediatric patients with attention-deficit hyperactivity disorder (ADHD). AKL-T01 (Akili Interactive Labs, Boston, MA) targets attention and cognitive control delivered through a video game-like interface through at-home play for 25 minutes per day, 5 days per week for 4 weeks. The STARS-ADHD study consisted of a randomized, double blind, parallel group, controlled trial of 348 pediatric patients aged 8 to 12 years with confirmed ADHD and Test of Variables of Attention (TOVA) Attention Performance Index (API) scores of -1.8 and below performed by 20 research institutions in the USA. Study participants were randomly assigned 1:1 to AKL-T01 or a digital control intervention which was in the form of a challenging and engaging word game. The study’s primary outcome was a mean change in TOVA API from pre-intervention to post-intervention. Additionally, participant safety, tolerability, and compliance were also evaluated. Study participants who received AKL-T01 (n=180 [52%]; mean [SD] age, 9.7 [1.3] years) or control (n=168 [48%]; mean [SD] age, 9.6 [1.3] years), the non-parametric estimate of the population median change from baseline TOVA API was 0.88 (95% Confidence Interval [CI] 0.24-1.49; p=0.0060). The mean SD change from baseline on the TOVA API was 0.93 (3.15) in the AKL-T01 group and 0.03 (3.16) in the control group. No serious adverse events or discontinuation occurred. Participant compliance was a mean of 83 (83%) of 100 expected sessions played (SD, 29.2 sessions). The investigators concluded based on the evidence, AKL-T01 might be used to improve objectively measured inattention in pediatric patients with ADHD with minimal adverse events.

Freespira

Freespira, Inc. (Kirkland, WA) developed Freespira which is an FDA-cleared digital therapeutic that utilizes a proprietary sensor, physiologic feedback display, and coaching to instruct patients over 28-days to normalize the respiratory irregularities underlying a key physiological mediator of anxiety attacks and post-traumatic stress disorder (PTSD) symptoms (carbon dioxide hypersensitivity). Freespira is an adjunctive digital treatment for symptoms of panic disorder (PD) and PTSD used under the supervision of a healthcare professional, in combination with other pharmacological and/or non-pharmacological interventions. Specifically, Freespira consists of a small, portable case with a commercial-grade portable sensor that is capable of measuring real-time carbon dioxide (CO

) and respiratory rate with wireless connectivity to a tablet computer that comes with a pre-installed app to guide treatment. Training is provided from a clinically supervising coach via telehealth in the form of guidance and support on appropriate use and best practices over the 28 day duration. The functionality of Freespira is based on breath sample delivery via a nasal canula connected to the Freespira sensor and by teaching patients to breath in synch and at different rates with rising and falling audio tones. Additionally, visual graphs of respiratory rate and exhaled CO

serve as a prompt to adjust breathing volume in order to achieve normal CO

targets. The coach is able to see the patient’s uploaded physiologic data from the app and provide patient-tailored and specific coaching to further augment engagement, adherence, and symptom reductions over time. Freespira is used for 17 minutes twice daily for 28 days at home and although a prescription is not required from a physician, this digital therapeutic must be authorized by a licensed healthcare provider (Digital Therapeutics Alliance, 2021e).

Tolin and colleagues (2017) evaluated Freespira (Palo Alto Health Sciences, Inc., Danville, CA) in a multicenter, single arm trial consisting of 69 adult participants with panic disorder (PD). Study participants received 4 weeks capnometry guided respiratory intervention (CGRI) using Freespira, which provided feedback of end-tidal CO

(PETCO

) and respiration rate (RR) transmitted by a sensor device. The intervention was delivered via home use after initial training by a clinician and provided remote monitoring of participant adherence and progress by the clinician. Outcomes assessment occurred post-treatment at 2- and 12-month follow-up. CGRI was associated with a response rate of 83% and remission rate of 54%. Additionally, large decreases in panic severity were noted as well as similar decreases in functional impairment and in global illness severity. The investigators noted that gains were largely sustained at follow-up and PETCO

moved from the slightly hypocapnic range to the normocapnic range. This study served as a benchmarking analyses against a prior published controlled trial and confirmed prior clinical results and further supported the viability of CGRI in the treatment of PD.

Halo AF Detection System

LIVMOR, Inc. (Frisco, TX) developed the Halo AF Detection System which is an FDA-cleared digital technology that is delivered on a Samsung wearable smartwatch device and provides continuous monitoring of pulse rhythms for the detection of atrial fibrillation (AF), on demand during the day and automatically overnight. A prescription is required from a physician for patients to use the Halo AF Detection System (LIVMOR, 2020).

Currently, there is a lack of published peer-reviewed evidence available.

Insulia

Voluntis (Cambridge, MA) developed the Insulia app which is an FDA-cleared software medical device that is indicated for use by healthcare professionals (HCPs) and their type 2 adult diabetes patients who are receiving treatment with a long-acting insulin analog. Insulia facilitates insulin titration for patients using any brand of basal insulin including Lantus, Levemr, Toujeo, Tresiba, and Basaglar. This app is complimentary to basal insulin therapy and may be used on a compatible smartphone or computer. The Insulia app’s functionality includes the secure capture, storage, and transmission of the patient’s diabetes related data via a web portal. Additionally, the visual reports and graphs supported by this app enables the HCP to review, analyze, and evaluate patient data to better manage the patient’s diabetes. The app also comes with an accompanying coaching feature to ensure continual patient support. A prescription is required from a qualified healthcare provider for the patient to use the Insulia app (Digital Therapeutics Alliance, 2021f; Voluntis, 2021).

In the TeleDiab-2 study, Franc and colleagues (2019) evaluated the efficacy and safety of two telemonitoring systems to optimize basal insulin (BI) in participants with inadequately controlled type 2 diabetes. The study was a 13-month randomized controlled trial consisting of 191 individuals (mean age 58.7 years, mean hemoglobin A

[HbA

] 8.9%). Study participants were randomized into three groups including group 1 (standard care, n=63), group 2 (interactive voice response system, n=64), and group 3 (Diabeo-BI app software, n=64). At 4 months follow-up, HbA

reduction was significantly higher in the telemonitoring groups (group 2: -1.44% and group 3: -1.48% vs group 1: -0.92%; p< 0.002). Furthermore, target fasting blood glucose was achieved by twice as many patients in the telemonitoring groups as in the control group, and insulin doses were also titrated to greater levels. The absence of severe hypoglycemia was noted in the telemonitoring groups. Mild hypoglycemia frequency was similar in all groups. The investigators concluded both telemonitoring systems improved glycemic control to a similar extent without an increase in hypoglycemic episodes.

leva Pelvic Health System

Renovia Inc. (Boston, MA) developed the leva Pelvic Health System which is an FDA-cleared medical device and consists of an intravaginal wand with motion sensors and app-based software program. This medical device is indicated for urinary incontinence in women with the aim of strengthening pelvic floor muscles and rehabilitating and training weak pelvic floor muscles in order to manage stress, mixed and mild to moderate urgency urinary incontinence, including overactive bladder. Under the guidance of the leva app, the patient performs 2 and a half minute exercise sessions twice a day for 8 to 12 weeks or until patient satisfaction with results. The patient performs the exercise while standing with the leva wand placed intravaginally for the exercise duration followed by immediate removal after use. Exercise data is transmitted from the wand to the software program on the patient’s smartphone and the healthcare provider receives a monthly summary and individual patient reports. This securely captured and transmitted data highlights therapy adherence, symptoms, perceived improvement, and material remarks from leva’s care management team which can then be used for short- and long-term follow up care. The leva Pelvic Health System requires a prescription from a qualified healthcare provider for patients to use this medical device as first-line therapy either alone or in combination with other therapies (Digital Therapeutics Alliance, 2021g; Renovia, 2021).

Rosenblatt and colleagues (2019) evaluated the effectiveness and patient satisfaction of the leva Pelvic Digital Health System (leva), a pelvic floor muscle training (PFMT) with an accelerometer-based system for the treatment of female urinary incontinence (UI). This prospective, single-center, open label study consisted of 23 female participants who were premenopausal with mild to moderate stress or mixed UI for 6 weeks duration with supervision. The study results were as follows: the Urogenital Distress Inventory (UDI) score decreased from 36.7 ± 4.7 to 1.45 ± 0.8 at 6 weeks (p<0.0001), the Patient’s Global Impression of Severity score decreased from 1.5 ± 0.1 to 0.2 ± 0.1 (p<0.0001) at study endpoint, the pelvic floor muscle (PFM) contraction duration increased from 13 ± 2.6 at baseline to 187 ± 9.6 seconds (p<0.0001), repeated contractions in 15 seconds increased from 5.9 ± 0.4 at enrollment to 9.6 ± 0.5 at 6 weeks (p< 0.0001), and maximum pelvic floor angle (a measure of lift) increased from 65.1 ± 2.0˚ to 81.1 ± 1.8˚ (p<0.0001). Additionally, increasing PFM contraction duration and maximum pelvic floor angle correlated with decreasing UDI-6 scores, r = -0.87, p=0.01; r = -0.97, p=0.0003, respectively. Device-related adverse events were absent.

Weinstein and colleagues (2022) evaluated whether the use of an intravaginal motion-based digital therapeutic device for pelvic floor muscle training (PFMT) was superior to PFMT alone in women with stress-predominant urinary incontinence (SUI). This study was a multicenter, randomized-controlled trial consisting of 61 female participants with SUI or SUI-predominant mixed urinary incontinence. The intervention group (n=29) was treated with PFMT using the device and the control group (n=32) received treatment with PFMT alone. Primary outcomes were measured at 8 weeks and included change in Urinary Distress Inventory, short-version and improvement in the Patient Global Impression of Improvement. In addition, participants completed Pelvic Organ Prolapse and Colorectal-anal Distress Inventories, Pelvic-Floor-Impact Questionnaire and a 3-day bladder diary. Study results were as follows: no statistical difference was noted in Urinary Distress Inventory, short-version scores between the intervention (-13.7 ± 18.7) and the control group (-8.7 ± 21.8; p=0.85), or in Patient Global Impression of Improvement (interventions 51.7% and control group 40.6%; p=0.47). Furthermore, Pelvic Organ Prolapse and Colorectal-anal Distress Inventories and Pelvic-Floor-Impact Questionnaire scores improved significantly more in the intervention group than the control group (all p<0.05) and median number of SUI episodes decreased from baseline to 8 weeks by -1.7 per day [(-3)-0] in the intervention group and -0.7[(-1)-0] in the control group, (p=0.047). Notably, this study was prematurely stopped due to device technical considerations.

MindMotion GO

MindMaze (Lausanne, Switzerland) developed MindMotion GO which is an FDA-cleared medical device software used in combination with the Microsoft Kinect v2 and Leap Motion controller that supports the physical rehabilitation of adults in the acute inpatient settings, outpatient clinics, and at home. The software employs game-based digital therapies which includes rehabilitation exercises for the upper extremity, trunk, and lower extremity; audio-visual feedback and graphic movement representations for individuals; and individual performance metrics for the healthcare professional. Prior to use of MindMotion GO, individual assessment, exercise guidance, and approval by the healthcare professional is required (FDA, 2022c; MindMaze, 2022).

My Dose Coach

Sanofi (Cambridge, MA) developed My Dose Coach which is an FDA-cleared basal titration app for adult patients with type 2 diabetes who have been prescribed a once-daily long-acting basal insulin. This app is intended to function as aid to the patient by providing dose suggestions based upon the healthcare provider's independent professional judgement. Prior to My Dose Coach use, the healthcare provider sets up the dose instructions for the specific patient and initiates the app using specific patient instructions. My Dose Coach utilizes dose plan instructions given by the patient's healthcare provider to give dose suggestions of once-daily long-acting basal insulin (i.e., basal insulin titration) that are based on the patient's fasting blood glucose as well as hypoglycemia occurrence. It is available by prescription only and is not intended to replace the care or advice of a healthcare provider (FDA, 2022d).

myVisionTrack (Home Vision Monitor [HVM])

Visual Art and Science, LLC (Richardson, TX) developed myVisionTrack which is an FDA-cleared mobile app designed as a vision function test. It is intended for the detection and characterization of central 3 degrees metamorphopsia (visual distortion) in individuals with maculopathy, including age-related macular degeneration and diabetic retinopathy, and aids in monitoring progression of disease factors causing metamorphopsia. The myVisionTrack app allows patients to regularly perform a simple self-test at home who have this capability. It is not intended to diagnose and a prescription is required for use (FDA, 2022e).

Nerivio

Theranica Bio-Electronics Ltd. (Montclair, NJ) developed Nerivio which is a wireless wearable neuromodulation device that is operated by a smartphone software application. Nerivio device is FDA-cleared via the De Novo Pathway and is indicated for the acute treatment of migraine with or without aura in patients 12 years of age or older. Nerivio can serve as a replacement for current migraine therapy or work in combination with existing therapy. The functionality of the Nerivio device is based on it being applied to the patient’s upper arm at the onset of migraine with self-administered treatment that is adjusted at an intensity that is not painful for a duration of 45 minutes. Notably, patients with congestive heart failure, severe cardiac disease, cerebrovascular disease, or uncontrolled epilepsy are not candidates for Nerivio treatment. Additionally, patients with active implantable medical devices, such as a pacemaker or hearing aid implant, should not use Nerivio. A prescription from a qualified healthcare provider is required for patients to use Nerivio (Digital Therapeutics Alliance, 2021h).

Grosberg and colleagues (2021) evaluated the efficacy and safety of remote electrical neuromodulation (REN) in patients with chronic migraine. This was an open-label, single-arm study consisting of 91 participants with chronic migraine and whose headaches were treated with the REN device (Nerivio, Theranica Bio-Electronics Ltd, Israel) for 4 weeks. In addition, participants used an electronic diary to record their symptoms at treatment initiation, 2 hours after treatment, and 24 hours after treatment. The primary outcome was the percentage of participants who achieved pain relief at 2 hours post-treatment. Secondary outcomes included pain freedom and improvement of associated symptoms and functional disability. Study results were as follows: pain relief and pain disappearance at 2 hours were achieved by 59.3% (54/91) and 20.9% (19/91) of participants, respectively, and sustained pain relief at 24 hours was observed in 64.4% (29/45) of those who achieved pain relief at 2 hours. REN had a favorable effect on nausea, photophobia, and phonophobia and improved functional ability. A device-related adverse event was observed.

Hershey and colleagues (2021) conducted a post-hoc analysis from a clinical study consisting of 35 adolescent participants which compared the efficacy of remote electrical neuromodulation (REN) to that of standard-care medications (triptans or over-the-counter medications) for the acute treatment of migraine. Specifically, efficacy was compared between a run-in phase in which attacks were treated with standard-care medications, and an intervention phase in which attacks were treated with REN. Efficacy was compared using the McNemar’s test at four endpoints (two hours post-treatment); single-treatment pain freedom and pain relief, and consistency of pain freedom and pain relief (defined as response in at least 50% of the available first four treatments). Post-hoc analysis results were noted as follows: at two hours post-treatment, pain freedom was achieved by 37.1% of participants with REN, vs. 8.6% of participants with medications (p=0.004), pain relief was achieved by 71.4% with REN, vs. 57.1% with medications (p=0.225), consistency of pain freedom was achieved by 40% with REN, vs.8.6% with medications (p<0.001), and consistency of pain relief was achieved by 80.0% with REN, vs. 57.2% with medications (p=0.033). The investigators concluded that REN may have a higher efficacy than certain standard-care medications for the acute treatment of migraine in adolescents.

NightWare

NightWare, Inc. (Hopkins, MN) developed NightWare which is an FDA-cleared medical device with a Breakthrough Device designation and is indicated for the reduction of sleep disturbance associated with nightmares in adult patients 22 years of age or older who suffer from nightmare disorder or have nightmares from post-traumatic stress disorder (PTSD). The functionality of this medical device is based on artificial intelligence (AI) and smart technology on the Apple Watch. NightWare is driven by the Apple Watch heart rate monitor sensor and other biometric sensors to continually evaluate the patient’s level of sleep disturbance (i.e., stress index) during sleep by tracking heart rate and body movements to determine nightmare occurrence. Once a nightmare is detected, the NightWare system quickly sends vibrations to interrupt nightmares without waking the patient. Through AI algorithms, both intensity and frequency of vibrations are based on an individual’s specific needs at that moment. As the NightWare system captures more data, it adapts to the patient’s sleep patterns. A prescription from a physician is required for NightWare to be used by patients (NightWare, 2021).

Currently, there is a lack of published peer-reviewed evidence available.

Parallel

Mahana Therapeutics, Inc.'s (San Francisco, CA) Parallel (formerly known as ReguI8) is an FDA-cleared digital therapeutic mobile application designed to deliver cognitive behavioral therapy for patients 22 years of age and older who have been diagnosed with irritable bowel syndrome (IBS). The mobile application uses the patient's mobile phone or tablet to deliver therapy on demand as a complement to the provider's care. It is available by prescription only as a 3-month treatment for patients with IBS and is intended to reduce the severity of symptoms when used as an adjunct with other IBS treatments (FDA, 2022a; Mahana Therapeutics, 2020).

Regulora

metaMe Health Inc. (Chicago, IL) developed Regulora which is an FDA-cleared prescription-only digital therapeutic software indicated for use in the treatment of abdominal pain due to irritable bowel syndrome (IBS). It is considered as a software as a medical device housed on and is accessed through the user's mobile device which can be performed at home. Regulora is intended to provide behavioral therapy through gut-directed hypnotherapy for patients 22 years of age and older who have been diagnosed with irritable bowel syndrome. It is indicated as a 3-month treatment for patients with abdominal pain due to IBS and is intended to be used in combination with other IBS treatments (FDA, 2022f).

RelieVRx

AppliedVR, Inc. (Van Nuys, CA) developed RelieVRx (formerly EaseVRx) is an FDA-authorized prescription-use immersive virtual reality system designed to provide adjunctive treatment based on cognitive behavioral therapy skills and other evidence-based principles for patients 18 years of age or older with a diagnosis of chronic lower back pain. The device is designed for in-home use for the reduction of pain and pain interference associated with chronic lower back pain. Therapy is provided via a virtual reality display using a software program containing the behavioral therapy content. The patient's pain centers are engaged through mindful escapes, pain education, diaphragmatic breathing, and relaxation/interoception. The virtual reality treatment is self-administered over 8 weeks with an average daily session of 7 minutes duration (Applied VR, 2022; FDA, 2022g).

reSET

Pear Therapeutics, Inc. (Boston, MA) developed reSET which is an FDA-cleared software application that provides cognitive behavioral therapy for substance abuse disorder as an adjunct to a contingency management system for patients 18 years of age and older who are enrolled in outpatient treatment under the supervision of a healthcare provider. Specifically, reSET delivers therapy established on the community reinforcement approach (CRA), an intensive form of validated neurobehavioral therapy for substance abuse disorder in addition to contingency management and reinforcement of concept mastery to augment learning. reSET consists of 62 interactive modules (32 core modules and 30 supplemental modules). The core modules involve CRA concepts, skill building to reinforce behavior change and prevent relapse. The supplemental modules focus on specific topics (e.g., relationship skills, living with hepatitis C). Modules may typically take 10 to 20 minutes to complete. The reSET app is supported on a mobile operating system (e.g., smartphone or tablet). A prescription is required from a licensed healthcare provider for a patient to use reSET which provides a 12 week duration of therapy (Digital Therapeutic Alliance, 2021i; Pear Therapeutics, 2021a).

Luderer and colleagues (2022) performed an exploratory analysis of data from a study to determine how patient engagement with a digital therapeutic for substance abuse disorder (SUD) in the clinic setting was associated with abstinence outcomes. The investigators evaluated engagement for 206 participants enrolled in a treatment program for SUDs related to cocaine, alcohol, cannabis, or other stimulants with randomization to receive treatment as usual (TAU) or reduced TAU plus the digital Therapeutic Education System (TES) for 12 weeks. Participant eligibility for contingency management incentives for module completion (Community Reinforcement Approach topic areas were covered) and negative urine drug screens were noted. The association of module completion with end-of-treatment abstinence was analyzed. Participants completed a mean of 38.8 (range 0-72) TES modules over 12 weeks of treatment. Study completers (n = 157) completed a mean of 45.5 (range 9-72) TES modules, whereas study noncompleters (n = 49) completed a mean of 17.4 (range 0-45) TES modules. A strong positive correlation between TES engagement (i.e., total number of modules completed) and the probability of abstinence during weeks 9-12 of treatment among 157 study completers (OR = 1.11; 95% Confidence Interval [CI] 1.08-1.14) was observed. Each module completed increased the odds of abstinence during weeks 9-12 by approximately 11% for study completers and 9% for the full sample. Additionally, a similar, but weaker, association between engagement and abstinence among 49 patients who did not complete the study (OR = 1.02; 95% CI 0.98-1.07) was observed. The investigators concluded that a greater engagement with a digital therapeutic for patients with SUD (i.e., number of modules completed over time) showed strong association with probability of abstinence in the last four weeks of treatment among those who completed the recommended 12-week treatment.

Maricich and colleagues (2022) performed a secondary analysis of patients with substance use disorders related to alcohol, cannabis, cocaine, or other stimulants (n = 399, patients with opioid use disorder [OUD] excluded) from a previously-published randomized controlled trial. Patients received 12-weeks of outpatient treatment-as-usual (TAU; n = 193) or TAU with reduced counseling plus a digital therapeutic (DT) (n = 206) providing computerized cognitive behavioral therapy and contingency management. Primary outcomes were abstinence in weeks 9-12 and retention in treatment. The 399 patients in this analysis (206 in the DT group and 193 in the TAU group) reported substance use disorders related to alcohol, cannabis, cocaine, or other stimulants (e.g., methamphetamines). Demographic and baseline characteristics such as age, sex, race, education, and reported primary substance use disorder were balanced between treatment groups. Abstinence was significantly higher in the DT group compared to TAU (40.3 vs. 17.6%; p < 0.001) as was retention in therapy (76.2 vs. 63.2%, p = 0.004). Intergroup adverse event rates were not significantly different (p = 0.68). The investigators concluded that the use of a DT safely increased abstinence (reduced substance use) and retention in treatment among patients with substance use disorders related to alcohol, cannabis, cocaine, or other stimulants (including methamphetamines).

reSET-O

Pear Therapeutics, Inc. (Boston, MA) developed reSET-O which is an FDA-cleared software application that provides cognitive behavioral therapy for opioid use disorder as an adjunct to outpatient treatment that includes transmucosal buprenorphine and contingency management outpatient treatment for patients 18 years of age or older who are under the supervision of a healthcare provider. Specifically, reSET-O delivers behavioral therapy based on the community reinforcement approach (CRA), a type of cognitive therapy targeting opioid use disorder. In addition, reSET-O combines CRA with reinforcement of concept mastery which should be initiated concurrently with contingency management and buprenorphine treatment to aid patient retention with opioid use disorder in outpatient treatment. reSET-O is supported on a mobile operating system (e.g., smartphone or tablet) and is a 12-week software application that requires a prescription from a licensed healthcare provider for patient use (Digital Therapeutics Alliance, 2021j; Pear Therapeutics, 2021a).

Christensen and colleagues (2014) examined the benefit of adding an internet-delivered behavior therapy to a buprenorphine medication program and voucher-based motivational incentives. This was a block-randomized, unblinded, parallel, 12-week treatment study consisting of 170 opioid-dependent adult participants (mean age = 34.3 years; 54.1% male; 95.3% white). Study participants received either an internet-based community reinforcement approach intervention plus contingency management (CRA+) and buprenorphine or contingency management alone (CM-alone) plus buprenorphine. The primary endpoints, measured over the course of treatment, were longest continuous abstinence, total abstinence, and days retained in treatment. Study results were as follows: in comparison to CM-alone participants, CRA+ participants displayed, on average, 9.7 total days more of abstinence (95% confidence interval [CI=2.3, 17.2]), and had a reduced hazard of dropping out of treatment (hazard ratio=0.47; 95% CI [0.26, 0.85]). Previous treatment for opioid dependence significantly mediated the additional improvement of CRA+ for longest continuous days of abstinence. The investigators concluded that an internet-based CRA+ treatment has efficacy and adds clinical benefits to a contingency management/medication based program for opioid dependence.

Maricich and colleagues (2021) conducted a study to evaluate real-world digital therapeutic (PDT) use and associated clinical outcomes among patients with opioid use disorder (OUD). Specifically, this study involved a real-world evaluation of patients who filled either a 12- or 24-week (refill) prescription for the reSET-O® PDT, a PDT consisting of 67 interactive lessons that unlock in sequence during use with an opportunity to earn rewards for progress and/or negative urine screens. The investigators collected engagement/retention data (ongoing engagement in weeks 9-12, or 21-24) via the PDT and performed analysis with descriptive data. Evaluation of substance use was from a composite of patient self-reports and urine drug screens (UDS). Missing UDS data were assumed to be positive. A regression analyses of hospital encounters for 12- vs. 24-week prescriptions controlling for covariates was conducted. In a cohort of 3,817 individuals with OUD who completed a 12-week PDT prescription, a cohort of 643 was prescribed a second 12-week 'refill' prescription, for a total treatment time of 24 weeks. Mean age of the 24-week cohort was 39 years, 56.7% female. At 24 weeks of total treatment: abstinence in the last 4 weeks of treatment was 86% in an analysis in which patients with no data were assumed to be positive for illicit opioids. Over 91% of patients were retained in treatment. An analysis of matched insurance claims showed that those treated for 24 weeks had a 27% decrease in unique hospital encounters compared to those who got the first 12-week prescription only. In summary, 93% of this cohort completed 8 or more core lesson modules in the second prescription period, 85% completed at least half of core modules, and 64% completed all 32 core modules. Patients used the PDT outside of clinic hours about 40% of the time. 94.4% of patients had 80% or greater negative reports of opioid use across the second 12 weeks of treatment. A 27% decrease in unique hospital encounters was observed in patients who completed a second prescription vs. patients who completed only one prescription. The data demonstrated that a second prescription (24 weeks) of a PDT for OUD is associated with improved outcomes, high levels of retention, and fewer hospital encounters compared to a single prescription for a PTD. Patients showed durable and high levels of engagement with the PDT, reduced substance use, and improved treatment retention through 24 weeks of treatment.

A Veteran's Administration and Department of Defense Clinical Practice Guideline on substance abuse disorders (VA/DoD, 2021) stated that "there are currently FDA-cleared apps in clinical use for the treatment of SUD (e.g., ReSET and ReSET-O), but literature leading to clearance did not meet the inclusion criteria for this CPG’s systematic evidence review."

Somryst

Pear Therapeutics, Inc. (Boston, MA) developed Somryst which is an FDA-cleared software app that provides digital cognitive behavioral therapy for insomnia (CBT-I) for chronic insomnia in patients who are 22 years of age and older. With the aim of improving a patient’s insomnia symptoms, Somryst is accessible on a mobile device (e.g., smartphone or tablet) and consists of 6 treatment cores focused on CBT-I concepts (e.g., sleep restriction and consolidation, stimulus control and cognitive restructuring). Patients should complete one core every 7 days and complete their daily sleep diary and follow the sleep restriction window recommendation provided by this software app. Somryst uses sleep restriction and consolidation and, therefore, is not to be used in individuals with any disorder worsened by sleep restriction (e.g., bipolar disorder, schizophrenia, other psychotic spectrum disorders), untreated obstructive sleep apnea, parasomnias, epilepsy, high risk of falls, pregnancy, and unstable or degenerative illness considered to be exacerbated by sleep restriction delivered as a part of CBT-I. Somryst is a 9-week therapy duration that is complimentary to current therapy. Additionally, a prescription from a licensed healthcare provider is required for a patient to use this software app (Digital Therapeutics Alliance, 2021k; Pear Therapeutics, 2021b).

Ritterband and colleagues (2022) conducted a retrospective investigation to evaluate outcome and patient engagement data of Sleep Health Using the Internet (SHUTi), a digital therapeutic delivering Cognitive Behavioral Therapy for insomnia (CBT-I) in a large real-world dataset of adults with insomnia. This real-world analysis is based on a dataset of consecutive users of SHUTi, the precursor program to the first FDA-authorized prescription digital therapeutic (PDT) Somryst (equivalent clinical content and enhanced features). SHUTi is a fully automated, interactive digital CBT-I intervention accessible via an internet-connected browser on mobile devices and computers. It delivers six sequential treatment modules (called Cores) based on key elements of CBT-I, which include an overview of insomnia, sleep restriction, stimulus control, cognitive restructuring, sleep hygiene, and relapse prevention. This real-world dataset analysis included 7216 adults who purchased access to SHUTi between December 2015 and February 2019. The Insomnia Severity Index (ISI) was given at the start of each of six treatment Cores of the intervention. Users recorded sleep diaries between Cores to track changes in sleep over time and obtain tailored sleep recommendations. Program usage was determined from the number of Cores completed and sleep diaries recorded. Users demonstrated a reduction in mean ISI scores and a corresponding increase in effect size at the start of each subsequent Core (compared to Core 1) (range: d = 0.3-1.9). Effect sizes at the last Core relative to the first were moderate-to-large for diary-derived sleep onset latency and wake after sleep onset. A reduction in number of medicated nights was also noted, with those with severe insomnia displaying the largest reduction from last-to-first week of treatment (d = 0.3). At the last Core, 61% met criteria for meaningful treatment response (reduction of >7 points on ISI) and 40% met criteria for remission (ISI<8). Engagement was comparable to SHUTi research trials. The investigators concluded that real-world data suggest that digital therapeutics can result in relatively high levels of engagement and clinically meaningful sleep improvements.

Christensen and colleagues (2016) evaluated whether an online self-help insomnia program could reduce depression symptoms. This was a randomized controlled study consisted of 1149 participants (aged 18-64 years) with insomnia and depression symptoms, but who did not meet criteria for major depressive disorder. Study participants were randomly assigned (1:1) to receive SHUTi (a 6 week, modular online insomnia program based on cognitive behavioral therapy for insomnia) or HealthWatch (an interactive, attention-matched, internet-based placebo control program). The primary endpoint was depression symptoms at 6 months, as measured with the Patient Health Questionnaire (PHQ-9). Results were based on 581 (51%) participants completing the study program assessments at 6 weeks and 504 (44%) participants completing 6 months follow up. SHUTi recipients had significantly lower depression symptoms on the PHQ-9 at 6 weeks and 6 months compared with HealthWatch (F[degrees of freedom 2,640.1] = 37.2, p<0.0001). Major depressive disorder was diagnosed in 22 (4%) participants at 6 months (n=9 in the SHUTi group and n=13 in the HealthWatch group), with no superior effect of SHUTi vs. HealthWatch (Fisher’s exact test=0.52; p=0.32). No adverse events were noted. The investigators concluded that online cognitive behavior therapy for insomnia treatment is a pragmatic and effective method to reduce depression symptoms and may have the capability to reduce depression at the population level.

Ritterband and colleagues (2017) evaluated a web-based, automated cognitive behavior therapy for insomnia (CBT-I) to improve insomnia in 9 weeks (short-term) and 1 year (long-term). This was a randomized clinical study consisting of 303 participants with chronic insomnia. Participant randomization occurred 1:1 where participants either received the internet CBT-I (Sleep Healthy Using the Internet [SHUTi]) or the online patient education program. SHUTi was a 6-week fully automated, interactive, and tailored web-based program incorporating the primary tenets of face-to-face CBT-I, whereas the online patient education program consisted of nontailored and fixed online information about insomnia. The primary sleep outcomes consisted of self-reporting online ratings of insomnia severity (Insomnia Severity Index) and online sleep diary-based values for sleep-onset latency and wake after sleep onset, collected prospectively for 10 days at each assessment period. The secondary sleep outcomes were comprised by sleep efficiency, number of awakenings, sleep quality, and total sleep time. The results of the three primary sleep outcomes revealed that the overall group x time interaction was significant for all variables, favoring SHUTi recipients (Insomnia Severity Index [F3, 1063 = 20.65, p<0.001), sleep-onset latency [F3, 1042 = 12. 68, p<0.001]). With-in group effect sizes exhibited improvements from baseline to post-assessment for the SHUTi recipients (range, Cohen d=0.79 [95% confidence interval [CI], 0.55-1.04] to d=1.90 [95% CI, 1.62-2.18]). Treatment effects were sustained at the 1 -year follow-up (SHUTi Insomnia Severity Index d=2.32 [95% CI, 2.01-2.63], sleep-onset latency d=1.41 [95% CI, 1.15-1.68], and wake after sleep onset d=0.95 [95% CI, 0.70-1.21]), with 56.6% (69 of 122) reaching remission status and 69.7% (85 of 122) deemed treatment responders at 1 year based on Insomnia Severity Index data. Secondary sleep outcomes, with the exception of total sleep time, showed significant overall group x time interactions, favoring the SHUTi group. The investigators concluded that internet-delivered CBT-I may have a pivotal role in the communication of effective behavioral treatments.

Natural Cycles App

The Natural Cycles (NC) is an FDA-cleared birth control app, which is powered by a proprietary algorithm that determines a woman’s fertility status based on her basal body temperature (BBT). Users measure their BBT with the NC thermometer, enter it into the app, and the proprietary algorithm will use temperature, period, and cycle data to determine the user’s fertility status. Users can use their daily fertility status to plan or prevent pregnancy.

Pearson et al (2021a) noted that digital fertility awareness-based methods of birth control are an attractive alternative to hormonal or invasive birth control for modern women. They are also popular among women who may be planning a pregnancy over the coming years and wish to learn about their individual menstrual cycle. In a prospective, real-world cohort study, these researchers examined the effectiveness of the NC app at preventing pregnancy for a cohort of women from the U.S. and described the key demographics of current users of the app in such a cohort. This trial included users who purchased an annual subscription to prevent pregnancy. Demographics were assessed via answers to in-app questionnaires. Birth control effectiveness estimates for the entire cohort were calculated using 1-year pearl index (PI) and 13-cycle cumulative pregnancy probability (Kaplan-Meier life table analysis). The study included 5,879 women who contributed an average of 10.5 months of data for a total of 5,125 woman-years of exposure. The average user was 30 years old with a body mass index (BMI) of 24 and reported being in a stable relationship. With typical use, the app had a 13-cycle cumulative pregnancy probability of 7.2 % and a 1-year typical use PI of 6.2. When the app was used under perfect use, the PI was 2.0. The authors concluded that the data presented in this study provided insights into the cohort of women using the NC app in the U.S. and gave country-specific effectiveness estimates. The contraceptive effectiveness of the app was in line with previously published figures from NC (PI of 7 for typical use and 2 for perfect use). Digital fertility awareness-based methods (FABMs) are a reality for a growing number of women of reproductive age and it is important that country-specific scientific evidence describing the effectiveness of such methods is encouraged.

Pearson et al (2021b) stated that digital FABM offers an alternative choice for women who do not wish to use hormonal or invasive methods for birth control. In a prospective, observational real-world study, these investigators examined the key demographics of current users of the NC app and evaluated the contraceptive outcomes of women preventing pregnancy in a U.K. cohort of women. The typical-use effectiveness of the method was calculated using both 13-cycle cumulative probability of pregnancy (life table analysis) and PI for the entire study cohort. Perfect-use PI was calculated using data from cycles where sexual intercourse during the fertile window was marked as protected and no unprotected sex was recorded on fertile days. A total of 12,247 women were included in the study and contributed an average of 9.9 months of data for a total of 10,066 woman-years of exposure. The mean age of the cohort was 30, mean BMI was 23.4, the majority were in a stable relationship (83.2 %) and had a university degree or higher (83 %). The 1-year typical use, PI was 6.1 (95 % CI: 5.6 to 6.6) and with perfect-use was 2.0 (95 % CI: 1.3 to 2.8); 13-cycle pregnancy probability was 7.1 %. The authors concluded that this was the 1st study that described the use of a digital contraceptive by women in the U.K. It provided the demographics of users and how they correlated with the apps effectiveness at preventing pregnancy.

A National Institute for Health and Clinical Excellence’s Medtech Innovation Briefing on “Natural Cycles for monitoring fertility” (NICE, 2021) noted that the innovative aspects are that NC is the 1st fertility-awareness app that comes with a basal thermometer and has been CE-marked as a medical device. The intended place in therapy is as a fertility-awareness contraception method. It would be used as a strategy to monitor ovulation, predict fertility, and may be used alongside abstinence or a barrier contraceptive method. The main points from the evidence were from 3 studies (2 retrospective analyses and 1 prospective observational study). These included 70,113 people using the app at home with typical follow-up of 6 to 9 months. They showed that NC can be used as a fertility-awareness contraception method. No evidence was identified on using the app to help plan a pregnancy.

Manhart and Duane (2022) noted that the NC app uses daily BBT to define the fertile window via a proprietary algorithm and is clinically established effective in preventing pregnancy. These researchers compared the app-defined fertile window of NC to that of CycleProGo, an app that uses BBT and cervical mucus to define the fertile window; and compared the app-defined fertile windows to the estimated physiologic fertile window. Daily BBT were entered into NC from 20 randomly selected regularly cycling women with at least 12 complete cycles from the CycleProGo database. The proportion of cycles with equivalent (± 1 cycle day) fertile-window starts and fertile-window ends was determined. The app-defined fertile windows were then compared to the estimated physiologic fertile window using Peak mucus to estimate ovulation. A total of 57 % of cycles (136/238) had equivalent fertile-window starts and 36 % (72/181) had equivalent fertile-window end days. The mean overall fertile-window length from the NC app was 12.8 days compared to 15.1 days for CycleProGo (p < 0.001). The NC algorithm declared 12 % to 30 % of cycles with a fertile-window start and 13 % to 38 % of cycles with a fertile-window end within the estimated physiologic fertile window. The CycleProGo algorithm declared 4 % to 14 % of cycles with a fertile-window start and no cycles with a fertile-window end within the estimated physiologic fertile window. The authors concluded that the NC app designated a higher proportion of cycles days as infertile within the estimated physiologic fertile window than CycleProGo.

Scope of Policy

This Clinical Policy Bulletin addresses prescription digital therapeutics.

Medical Necessity

Aetna considers FDA approved or cleared mobile apps for contraception based on fertility awareness (e.g., Natural Cycles) to be medically necessary per federal preventive care mandates, when prescribed by a treating provider.

Natural Cycles is currently the only FDA-cleared fertility app. One annual subscription to Natural Cycles is covered per benefit period; no additional supplies or services are covered.

Experimental and Investigational

The following prescription digital therapeutics (PDTs) are considered experimental and investigational because there is insufficient evidence in the published peer-reviewed literature of the the effectiveness of these PDTs: