Quality of Evidence in Remote Monitoring of Patients With Liver Cirrhosis: A Systematic Review Authors: Britt van Ruijven, Marten A Lantinga, Joost PH Drenth, Marieke J Pierik, Tom JG Gevers, Govert Veldhuijzen Published: 2026 ABSTRACT AbstractBackground and AimsRemote monitoring (RM) enables out-of-hospital surveillance and may benefit patients with liver cirrhosis. However, implementation in routine liver cirrhosis care is limited, most likely due to the lack of conclusive data on clinical effectiveness. This systematic review aims to analyze the quality of evidence regarding the effect of RM tools on disease outcomes for liver cirrhosis.MethodsThis Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA)-guided systematic review included studies of RM interventions per the World Health Organization’s definition in patients with liver cirrhosis reporting at least 1 liver disease-related outcome. Retrospective or cross-sectional studies were excluded. Risk of bias was assessed using the Cochrane Risk Of Bias Tool for Randomized Trials and the Newcastle-Ottawa Scale for other study types. RM development quality was evaluated against the V3+ framework.ResultsOne randomized controlled trial, 4 prospective cohorts with controls, and 7 prospective studies without controls were identified (n = 12; 636 patients; 17,680 studies screened). All studies with controls reported lower hospital readmissions in the RM groups. Interventions included smartphone applications, with or without additional measurement instruments, and automated calls. RM development quality varied substantially, with multiple RM interventions not reporting on verification (n = 6), usability (n = 4), or analytical validation (n = 5). The identified evidence informed the synthesis of a proposal framework to harmonize future RM research.ConclusionAlthough studies show clinical benefits of RM tools in liver cirrhosis, evidence is limited by the (inherent) heterogeneity in interventions, design, and outcomes, and quality of the RM development process. This review emphasizes the need for consensus on RM monitoring frameworks and high-quality data to inform better clinical decision-making.Keywords: Telehealth, Telemedicine, e-Health, End-Stage Liver Disease Introduction Healthcare delivery, including the management of liver cirrhosis, is increasingly incorporating telehealth, particularly through remote monitoring (RM).1, 2, 3, 4, 5 RM facilitates the surveillance of a patient’s condition outside the hospital, using technologies such as connected sensors and digital platforms.6 In patients with liver cirrhosis, RM platforms have demonstrated promising findings regarding disease-related benefits and cost-effectiveness.5 RM can support patient empowerment and could help in the earlier recognition of subtle changes preceding decompensation events in patients with liver cirrhosis. Several prospective studies and reviews have emphasized the potential of RM tools for various liver disease-related events.5,7 Despite encouraging early results, RM remains inconsistently integrated into routine liver cirrhosis care. Substantial heterogeneity in study designs, RM modalities, and reported outcomes constrains the synthesis of conclusive findings and has recurrently been acknowledged as a barrier for real-world translation and successful implementation.8, 9, 10, 11, 12, 13 Available systematic reviews on this topic focused on patients with decompensated liver cirrhosis while excluding studies without a control group or short follow-up, potentially missing out on promising technologies, or included populations with different needs, such as those infected with hepatitis C virus (HCV) without cirrhosis or post-transplant patients.9, 10, 11, 12, 13 Furthermore, current evidence has not yet been evaluated against structured quality validation frameworks for the development of digital tools. This systematic review aims to provide a comprehensive methodological analysis of the quality of evidence and existing knowledge gaps regarding the effect of RM tools on disease outcomes in compensated and decompensated liver cirrhosis. In addition, it assesses current evidence against a structured quality framework for validation of digital tools and proposes a framework informed by the identified evidence base to harmonize future research and improve the impact of RM interventions on liver cirrhosis patient care. Methods Search Strategy For this systematic review, the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline for reporting systematic reviews was used (Supplementary Table 1).14,15 The search strategy was developed in collaboration with a librarian with expertise in health sciences literature to optimize sensitivity. This review was registered in PROSPERO, an international systematic review registry produced by the Centre for Reviews and Dissemination (CRD), with the number CRD42024620147. The following databases were searched, with the final search on January 14, 2026: Medline via Ovid, Embase via Ovid, Web of Science, Cochrane Library, the PROSPERO database of systematic reviews, ClinicalTrials.gov, CIHNAHL via EBSCO, and Epistemonikos Health Evidence. Clinicaltrialsregister.eu and the International Clinical Trials Registry Platform Search Portal were screened to identify studies in preparation. Snowballing was performed on titles in the bibliographies of eligible studies. We did not use any date restrictions. Search strategies for each database and deduplication strategies are outlined in Supplementary Tables 2 and 3. Study Selection and Eligibility Criteria B.R. and T.G. independently screened titles and abstracts via Covidence systematic review software; both reviewers were blinded to each other’s decisions.16 Full texts were reviewed if the reviewers could not determine a study’s eligibility through title and abstract screening. Any disagreements were resolved through consensus meetings with a third reviewer (G.V.) when necessary. Studies evaluating patients with liver cirrhosis were selected. Studies assessing patients living with hepatitis B virus and HCV, metabolic dysfunction–associated steatotic liver disease (MASLD), alcohol-related liver disease (ALD), and autoimmune hepatitis patients were included if separate liver cirrhosis data were available. We anticipated that many studies encompassed pilot phases; therefore, studies were also eligible for inclusion even if they did not include a comparative group. The intervention must consist of technologies used for RM, for example, electronic portals, wearables, digital questionnaires, or message functions. We used the World Health Organization's definition of RM: services that enable health-care providers to monitor an individual’s condition remotely, using technologies such as implanted devices and sensors with wireless or wired connections.6 Publications required evaluation of at least 1 clinical outcome, reported as either primary or secondary outcomes: liver cirrhosis–related complications (hepatic encephalopathy, [refractory] ascites, gastrointestinal bleeding, acute kidney injury-hepatorenal syndrome), hospital admissions, mortality, transjugular intrahepatic portosystemic shunt–free survival, transplant-free survival, or treatment adherence. Study Designs We included the following study designs: randomized controlled trials (RCTs), non-RCTs, prospective cohort studies, routine health-care databases, uncontrolled and controlled before-and-after studies (including interrupted time series), case-control studies, observational studies, epidemiological studies, experimental or quasi-experimental studies, and feasibility studies. Studies with alternative study designs were evaluated for eligibility in a consensus meeting. We assessed individual studies used in reviews and meta-analyses for eligibility. We excluded retrospective and cross-sectional study designs, as they are less suitable for assessing the causal relationship between interventions and clinical outcomes, the focus of this review.17 Additional exclusion criteria comprised no availability in the English language, assessment of first-line RM without involvement of gastroenterologists/hepatologist, exclusive assessment of telephone or video consultations between patient and health-care provider, between patients, or between health-care providers, and insufficient data. Data Extraction and Study Quality Assessment Methods Data were collected in standardized data extraction forms in Excel. For homogeneity purposes, proof-of-concept studies and (prospective) pilot studies without a control group were recorded as feasibility studies. Follow-up periods were converted to days, with 1 week representing 7 days and 1 month representing 30 days. Nonalcoholic fatty liver disease or nonalcoholic fatty liver disease terms for underlying etiologies were collectively reported as metabolic dysfunction–associated steatohepatitis and MASLD etiologies. We separately defined liver cirrhosis etiologies ALD, MASLD, metabolic and alcohol-related liver disease, as lifestyle-related causes and grouped other etiologies. Quality of RCTs was assessed with the Cochrane Risk Of Bias Tool for Randomized Trials (RoB 2, Supplementary Figure 1).18 All other study types received quality assessment using the Newcastle-Ottawa Scale (NOS) Quality assessment scores of observational cohort studies, a 0–9 scale (Supplementary Material 1, and Supplementary Table 4).19 We defined “poor,” “fair,” and “good” quality as total NOS scores of 0–3, 4–6, and 7 or higher, respectively. Remote Monitoring Development Quality Assessment Methods RM interventions in studies were evaluated against the recently updated and widely adopted Digital Medicine Society’s V3+ framework.20,21 The Digital Medicine Society’s V3+ framework provides one of the most widely adopted structures for the development process of sensors, also applicable to many RM systems.20,21 The framework distinguishes 4 phases in the development process of digital health technologies: (1) verification, evaluating the performance or output of data of the RM tool against prespecified criteria; (2) usability validation, reporting whether the technology can be used with ease, efficiency, and user-satisfaction to achieve specified goals; (3) analytical validation, assessing whether the platform accurately measures, detects, or predicts physiological or behavioral metrics; and (4) clinical validation, which determines whether use of the technology leads to meaningful clinical benefits in the target population and situation. We assessed if included RM studies answered 4 questions, each representing a phase of the V3+ framework (Supplementary Table 6).20, 21, 22 If studies referred to previously conducted research in which one or more of the phases of the framework were reported, we reported that they complied with the respective phase of the framework. We reported the verification phase as performed in case a quality certification was available, or the RM tool was federal regulation compliant, also if this data was not open-sourced. Proposal for a Framework of Development and Evaluation Trajectories of Remote Monitoring Studies in Liver Cirrhosis The included studies were categorized and synthesized to provide an overview of prevailing methodologies and to identify gaps in the current evidence base. We especially examined the key study characteristics, including study designs, recruitment sites, outcomes, singlecenter vs multicenter designs, the availability of a validation cohort, and follow-up duration. Based on these observations, we aimed to move toward a more structured interpretation of the existing evidence by deriving an author-based consensus framework. This framework intends to help structure the development and evaluation of RM interventions in liver cirrhosis. Results Literature Search Our search strategy identified 17,680 individual studies. We included a total of 12 studies (n = 8 full papers, n = 4 abstracts, Figure 1). Snowballing through citation tracking did not yield any additional inclusions. Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 flow diagram. Presenting the identification of 17,680 screened records and 12 included studies. Study Designs The study designs included 1 RCT, 4 prospective cohorts with control groups, and 7 prospective cohorts without controls (Table 1). The majority of patients with liver cirrhosis were hospitalized at the time of inclusion (9/12). The primary study setting was a single academic liver transplant center (10/12). All control groups consisted of internal cohorts, of which 2 were historic populations. Follow-up durations ranged from 28 to 90 days. The RCT was a follow-up study on an open-label feasibility study with the Patient Buddy App and scored a “low risk of bias” in the Cochrane Risk of Bias Assessment Tool of RCTs (Table 1; Supplementary Figure 1).23 Cohort studies were rated with “good” (n = 2) and “fair” (n = 9) overall quality, respectively, through the NOS (Supplementary Table 4). Table 1Study DesignsDesignYearAuthorInpatient or outpatient recruitmentRM interventionPrimary outcomeSettingValidation cohortFollow-up durationRandomized controlled trial1. 2025Shaw et al23InpatientSmartphone application (Patient Buddy App and EncephalApp)Avoidable readmissionsThree transplant centers, of which 2 academic centersInternal30 dProspective cohort with control group2. 2023Kazankov et al7InpatientSmartphone application (CirrhoCare) with monitoring devices:(1)Wristwatch (Withings Move)(2)Blood pressure (BP) cuff(3)Weighing scales (Withings Body+) with bioimpedanceFeasibility of daily remote managementSingle academic transplant centerInternal, matched controls84 d 3. 2024Penrice et al24InpatientTablet with Bluetooth-enabled vital signs uploads from blood pressure monitor, pulse oximeter, and weighing scaleQualitative aspects of design,implementation, and patient satisfactionSingle academic transplant centerInternal, matched retrospective controls90 d 4. 2017Khungar et al25AbstractInpatientTablet with wireless blood pressure monitor, pulse oximeter, and weighing scaleNot reportedSingle academic transplant centerInternal90 d 5. 2024Ballesteros et al26AbstractInpatientSmartphone application (DOCCLA’s digital platform)Not reportedSingle academic transplant centerInternal, retrospective controls42 d intervention90 d follow-upProspective cohort without control group6. 2022Lin et al27OutpatientWrist personal activity trackers (PAT) and smartphone application (EL-FIT)Incidental hospital admission and mortalitySingle academic transplant center.NoPredefined follow-up not reported. Median follow-up 223 d. 7. 2015Thomson et al28InpatientProactive automated scheduled callsTime to first hospital admissionSingle academic transplant centerNo90 d 8. 2025Ngu et al29Inpatient and outpatientSmartphone application (LivR well) with Bluetooth weighing scales, and a cloud-based database for referral, screening and data collectionAcceptability and feasibilitySingle academic liver transplant centerNo28 d intervention84 d follow-up 9. 2020Bloom et al30Inpatient and outpatientSmartphone application with Bluetooth weighing scale connected to EMR% enrolled d during which weight data were successfully transmitted to the EMR and the % weight alerts that prompted a response by the providerSingle academic transplant centerNoPredefined follow-up: 28 d.Actual follow-up: 979 patient-d. 10. 2021Qian et al31AbstractInpatientSpecific devices not reported.Not reportedSingle academic transplant centerNo90 d 11. 2018Verma et al32AbstractInpatientTablet application with web-based provider portal to retrieve the resultsNot reportedOverall purpose: assess feasibilityNot reportedNo90 d 12. 2017Ganapathy et al33InpatientSmartphone application (Patient Buddy App)Total 30-d admissions, HE-related readmissions, dropoutsSingle academic transplant centerNo30 dEL-FIT, Exercise and Liver FITness; EMR, electronical medical record. Study Designs (1)Wristwatch (Withings Move) Wristwatch (Withings Move) (2)Blood pressure (BP) cuff Blood pressure (BP) cuff (3)Weighing scales (Withings Body+) with bioimpedance Weighing scales (Withings Body+) with bioimpedance EL-FIT, Exercise and Liver FITness; EMR, electronical medical record. Population Characteristics Sample size in the RM group ranged from 19 to 116 patients (Table 2). Baseline disease severity was mostly reflected in the mean and median Model for End-stage Liver Disease or Model for End-stage Liver Disease-sodium scores, with a range of 13–19.5. The mean or median age ranged from 51 to 67.2 years. A total of 636 patients participated in RM, of whom the majority were male (62.9%). ALD was the predominant etiology of liver cirrhosis (36%, 207/583 reported), 164 (28%) were known with metabolic dysfunction–associated steatohepatitis or MASLD, and 212 (36%) had other etiologies. Table 2Population CharacteristicsAuthorInpatient or outpatient recruitmentN (control group vs intervention group)Disease status at baseline (control group vs intervention group)% liver cirrhosis etiologies (control group vs intervention group)% males (control group vs intervention group)Age (y) (control group vs intervention group)Shaw et al23Inpatient116 vs 116 dyadscMELD-Na: 18.9 ± 8.4 vs 16.9 ± 7.9aAlcoholMASH/MASLDOther31.0 vs 31.026.7 vs 28.442.2 vs 40.565.5 vs 68.959.1 ± 12.38 vs 58.4 ± 8.88aKazankov et al7Inpatient20 vs 20MELD-Na: 18 ± 4.6 vs 16.1 ± 4.2aAlcoholMASH/MASLDOther85 vs 8010 vs 105 vs 1045 vs 7056 ± 14 vs 59 ± 10aPenrice et al24Inpatient74 vs 41MELD-Na: 17.3 (13.0–18.42) vs 15.1 (9.8–18.3)bAlcoholMASH/MASLDOther52.3 vs 46.330.1 vs 26.817.6 vs 26.854.0 vs 58.562.4 ± 12.5 vs 60.9 ± 9.0aKhungar et al25AbstractInpatient143 vs 19Median MELD-Na: 19% Child-Pugh B: 57.9% Child-Pugh C: 31.6Not reported for control groupAlcoholMASH/MASLDHepatitis CNot reportedNot reported53Not reported for control group% males intervention group: 47.4Not reported for control groupMedian age intervention group: 58Ballesteros et al26AbstractInpatient39 vs 43Decompensated patients with ascites.MELD-Na or Child-Pugh score not reportedNot reported.41 vs 65Not reported.Lin et al27Outpatient116: 71 PAT, 45 PAT + EL-FIT applicationMELD: 14.3 ± 5.9aMELD-Na: 15.2 ± 6.9aChild-Pugh score: 8 (5–10)bAlcoholMASH/MASLDOther3330375556 ± 11aInclusion criteria: age 40–70 yThomson et al28Inpatient79Mean MELD: 13 (6–40)Mean Child-Pugh score: 8 (6–13)AlcoholMASH/MASLD/cryptogenicOther1139495157 (23–85)bNgu et al29Inpatient and outpatient59Patients with ACLFMELD: 16 (12–21)b% Child-Pugh A: 22% Child-Pugh B/C: 78Inclusion criteria: MELD ≥ 10AlcoholMASH/MASLDOther745216651 (37–59)bBloom et al30Inpatient and outpatient25Patients actively receiving ascites management.MELD: 15.8 ± 5.9aAlcoholMASH/MASLDOther4436204857.6 ± 12.8aQian et al31AbstractInpatient4817 dropped out immediatelyMean MELD-Na: 13.5 (6–24)% Child-Pugh A: 33% Child-Pugh B/C: 67AlcoholMASH/MASLDOther345016100Mean age: 67.2 (47–78)Verma et al32AbstractInpatient30Mean MELD-Na: 19.5 (11–29)Inclusion criteria: ≥1 liver disease related complication.Exclusion criteria: patients discharged to nursing homes.AlcoholMASH/MASLDOther40105060.0Mean age: 55.7 (28–71)Ganapathy et al33Inpatient40MELD: 19.5 ± 5.2aAlcoholMASH/MASLDOther17.532.5506058 ± 10aEL-FIT, Exercise and Liver FITness; MASH, metabolic dysfunction–associated steatohepatitis; MELD-Na, Model for End-stage Liver Disease-sodium; PAT, physical activity tracker.aMean ± SD.bMedian (IQR).cDyad: a cirrhosis inpatient and an adult caregiver. Population Characteristics EL-FIT, Exercise and Liver FITness; MASH, metabolic dysfunction–associated steatohepatitis; MELD-Na, Model for End-stage Liver Disease-sodium; PAT, physical activity tracker. Mean ± SD. Median (IQR). Dyad: a cirrhosis inpatient and an adult caregiver. Intervention Characteristics of Remote Monitoring Seven studies employed smartphone applications as RM intervention, of which 4 incorporated additional wireless sensors/wearables, including weighing scales and physical activity trackers. Tablet applications were utilized in 3 studies. One study relied on automated, scheduled calls, while another did not specify the RM tool used, although heart rate, blood pressure, and weight were monitored (Table 3). Continuous monitoring was implemented in 1 study, whereas 5 studies applied daily monitoring protocols. Two reported weekly monitoring, and 1 publication described RM conducted 3 times per week in conjunction with various paramedical consultations. Body weight was the most frequently assessed parameter, being reported in 11 RM interventions. A message function was available in 4 RM tools. Most tools assessed multiple clinical or behavioral parameters, like orientation (7), temperature (6), blood pressure/heart rate (6), dietary intake (4), and physical activity (2). Table 3Intervention Characteristics of Remote MonitoringAuthorAimRM frequencyMessage functionWeightTemperatureBlood pressure and heart rateOrientationDietary intakeActivityOther:Shaw et al23Reducing 30-d readmissions.Daily monitoring: Weight and temperature.Weekly monitoring: Other parameters.✓✓✓Self-reportedx✓EncephalApp and judged by caregiver✓Sodium intakexSelf-reported:Bowel movementsGI-bleedingInfectionsMedication adherenceKazankov et al7Assess feasibility and patient engagement of monitoring advanced cirrhosis.Daily monitoring✓Text and voice messages✓x✓✓✓✓WristwatchxPenrice et al241.Design and implement a cirrhosis-specific RPM (CiRPM) program.2.Show utility through measurement of potential clinical benefit, patient satisfaction, and patient adherence.Daily monitoring: Parameters.Every 3 d: six alternating liver-complication yes-no questions.✓✓✓✓xxxSelf-reported:Bowel movementsChanges in attentionSwollen anklesAbdominal tensionShortness of breathMissed medicationKhungar et al25AbstractPilot a wireless mobile device monitoring system to detect early symptoms and signs, thereby preventing readmissions and keeping patients engaged and feeling cared for daily.RM frequency not reportedNot reported✓x✓✓xxSelf-reported:Hepatic encephalopathyFluid overloadGI bleedingInfectionsMedication adherenceAll not specifically reported.Ballesteros et al26AbstractDemonstrate the impact of real-world remote monitoring of patients with decompensated cirrhosis.RM frequency not reportedNot reported✓✓✓Not reportedNot reportedNot reportedLiver-related symptoms/complaints, not specifically reported.Lin et al27Assess the utility of daily step count in predicting hospital admission and mortality rates.Continuous monitoring and promotion of exercise trainingxxx✓Heart ratexx✓xThomson et al28Test the IVR system and determine if gathered information can predict complications.Weekly IVR calls.x✓Self-reportedxx✓Self-reportedxxSelf-reported:SwellingJaundiceWeaknessMedication changesOverall healthNgu et al29Investigate the acceptability and feasibility of a multimodal community intervention for ACLF.Three times per week, home nursing.Weekly medical review.“Regular” dietitian, and pharmacy consultations, with adjunct interventions by physiotherapy, social work, addiction medicine, or neuropsychiatry.Not reported✓xxxxxHome nursing to assess liver-related symptomsMedical review and urine (ETG) sample to assess alcohol use.Bloom et al30Evaluate the feasibility of a smartphone app in facilitating outpatient ascites management.Every weekday: Weight monitoringx✓xxxxxxQian et al31AbstractNot reportedRM frequency not reportedNot reported✓✓Self-reported✓✓Self-reported✓Self-reported salt intakexSelf-reported:Bowel movementsSwellingShortness of breathDizzinessFallsPainMedicationAlcohol useVerma et al32AbstractAssess the feasibility and utility of TeleMonitoring of Symptoms and Cognitive function/TeMSCog for ESLD inpatients.Daily questions about symptoms on a 0–10 scale, including pain, depression, and anxietyNot reported✓✓x✓EncephalAppxxLiver-related symptoms/complaints, not specifically reported.Self-reported:BreathlessnessPainDepressionAnxietyFatigueGanapathy et al33Define the feasibility of a smartphone app and its impact on 30-d readmissions.Daily monitoring: Weight and temperature.Weekly monitoring: Other parameters✓✓✓Self-reportedx✓EncephalApp and judged by caregiver✓Sodium intakexSelf-reported:Bowel movementsGI-bleedingInfectionsMedication adherenceJudged by caregiver:Fall riskTime up and go testACLF, acute on chronic liver failure; CiRPM, Cirrhosis-specific Remote Monitoring Program; ESLD, end-stage liver disease; ETG, ethyl glucuronide; GI, gastrointestinal; IVR, interactive voice response; TeMSCog, TeleMonitoring of Symptoms and Cognitive function. Intervention Characteristics of Remote Monitoring Bowel movements Bowel movements GI-bleeding GI-bleeding Infections Infections Medication adherence Medication adherence 1.Design and implement a cirrhosis-specific RPM (CiRPM) program. Design and implement a cirrhosis-specific RPM (CiRPM) program. 2.Show utility through measurement of potential clinical benefit, patient satisfaction, and patient adherence. Show utility through measurement of potential clinical benefit, patient satisfaction, and patient adherence. Bowel movements Bowel movements Changes in attention Changes in attention Swollen ankles Swollen ankles Abdominal tension Abdominal tension Shortness of breath Shortness of breath Missed medication Missed medication Hepatic encephalopathy Hepatic encephalopathy Fluid overload Fluid overload GI bleeding GI bleeding Infections Infections Medication adherence Medication adherence Swelling Swelling Jaundice Jaundice Weakness Weakness Medication changes Medication changes Overall health Overall health Bowel movements Bowel movements Swelling Swelling Shortness of breath Shortness of breath Dizziness Dizziness Falls Falls Pain Pain Medication Medication Alcohol use Alcohol use Breathlessness Breathlessness Pain Pain Depression Depression Anxiety Anxiety Fatigue Fatigue Bowel movements Bowel movements GI-bleeding GI-bleeding Infections Infections Medication adherence Medication adherence Judged by caregiver: Judged by caregiver: Fall risk Fall risk Time up and go test Time up and go test ACLF, acute on chronic liver failure; CiRPM, Cirrhosis-specific Remote Monitoring Program; ESLD, end-stage liver disease; ETG, ethyl glucuronide; GI, gastrointestinal; IVR, interactive voice response; TeMSCog, TeleMonitoring of Symptoms and Cognitive function. Study Aims and Reported Outcomes Two studies primarily evaluated the effect of RM on clinical outcomes.23,26 Seven studies primarily aimed to pilot or define the feasibility of the RM intervention, for example, evaluate whether patients and health-care providers are willing and able to use the RM tool (Table 3). Two studies assessed both feasibility and RM effectiveness,24,33 and 1 study did not report on the primary research aim.31 Primary outcomes ranged from avoidable readmissions to incidental hospital or total admissions, and time to hospital admission (4/12), to feasibility and related subdomains (5/12). In 3 abstracts, primary outcomes were not specifically reported (Table 1). All studies addressed hospital (re)admissions after inclusion (Table 4). Half of the studies recorded patient contacts, medication changes, or other interventions and mortality, while adherence and satisfaction were outlined in 7 and 8 publications, respectively (Table 4). They all documented positive results regarding adherence and satisfaction, but the reporting variety limited a summary synthesis of these outcomes in this review.7,23,24,28, 29, 30, 31, 32, 33 Only Ngu et al presented specific cost outcomes of the intervention and assessed quality of life.29 Table 4Reported OutcomesAuthorHospital (re)admissionsPatient contacts, medication changes, or interventionsMortalityLiver transplantation or workupRM adherenceRM evaluation by patientsOther reported outcomesShaw et al23✓Including avoidable readmissionsa✓x✓✓✓LVPsKazankov et al7✓✓✓✓✓✓Unplanned LVPsChanges MELD-Na and CLIF-C AD scoresPenrice et al24✓x✓x✓✓LVPsKhungar et al25Abstract✓Including avoidable readmissionsaxxxxxxBallesteros et al26Abstract✓a✓a✓axxxPlanned LVPsDecompensation rateLin et al27✓x✓xxxDaily step count association with hospital admission and mortality, adjusted for MELD-Na and EL-FIT useThomson et al28✓x✓✓✓xxNgu et al29✓x✓✓✓✓Changes MELD-Na and Child-Pugh scoresSarcopenia incidenceCLDQ and EuroQoL VAS scores (self-rated health questionnaires) CostsBloom et al30✓✓xx✓✓LVPsQian et al31Abstract✓xMentioned, but no specific data reportedxx✓xMentioned, but no specific data reportedxVerma et al32Abstract✓✓x✓xMentioned, but no specific data reported✓xGanapathy et al33✓✓x✓✓✓RM measurements:Changes in EncephalApp scores and up and go timesCLDQ, Chronic Liver Disease Questionnaire; CLIF-AD, Chronic Liver Failure Consortium Acute Decompensation; EL-FIT, Exercise and Liver FITness; LVP, large-volume paracentesis, MELD-Na, Model for End-stage Liver Disease-sodium.aP value <.05 in favor of intervention group. Reported Outcomes Changes in EncephalApp scores and up and go times Changes in EncephalApp scores and up and go times CLDQ, Chronic Liver Disease Questionnaire; CLIF-AD, Chronic Liver Failure Consortium Acute Decompensation; EL-FIT, Exercise and Liver FITness; LVP, large-volume paracentesis, MELD-Na, Model for End-stage Liver Disease-sodium. P value <.05 in favor of intervention group. Clinical Outcomes All 5 studies with control groups demonstrated clinical benefits of RM. Although not powered for clinical outcomes, 3 reported statistically significant benefits regarding (preventable) hospital readmissions compared to the control group, ranging from 10 vs 19.8% after 30 days to 12 vs 28% (Table 4, Supplementary Table 5).23,25,26 Ballesteros et al also presented a significant reduction in large-volume paracentesis and mortality.26 Kazankov et al and Penrice et al reported similar positive trends in readmission rates, large-volume paracenteses, and mortality.7,24 Thirty-day hospital readmission rates ranged from 15% to 42.5% in studies without a control group29,33 and from 3.9% to 68% in studies with follow-up until 90 days (Supplementary Table 5).27,28,30, 31, 32 Remote Monitoring Development Quality Assessment The Patient Buddy App, investigated by Shaw et al and Ganapathy et al, and the CirrhoCare platform of Kazankov et al presented reports on progress through all 4 phases of the V3+ quality framework (Figure 2, Supplementary Table 6, see Methods section for explanation of phases).7,23,33 Half of the studies used an RM tool, of which verification, quantifying the accuracy of a measurement, was traceable.7,23,26,27,30,33 Eight studies used an RM tool with reported usability, assessing platform ease of use and efficiency.7,23,24,28, 29, 30,33 Analytical validation, monitoring whether the platform accurately captures the intended outcomes, was reported in RM applications of 7 studies.7,23,24,27,30,32,33 Parts of clinical validation were performed in all studies, as they all reported on clinical outcomes of the RM platform in the target population. Number of studies reporting on remote monitoring development. Phases: verification (n = 6), usability validation (n = 8), analytical validation (n = 7), and clinical validation (n = 12). Proposal for a Framework of Development and EEvaluation Trajectories of Remote Monitoring Studies in Liver Cirrhosis The identified methodological strategies and gaps in the current evidence base synthesized a framework aimed to structure the development and evaluation of RM interventions in liver cirrhosis (Table 5). The framework comprises 4 sequential phases. First, early-phase feasibility studies should be conducted using harmonized methodologies to assess usability, adherence, and technical performance and to inform the selection of RM platforms for further evaluation. Second, prospective clinical evaluation should be undertaken in parallel with implementation efforts, enabling the generation of clinical evidence while facilitating integration into routine care. Third, confirmatory prospective trials should incorporate predefined core outcome sets, extended follow-up periods, multicenter designs, patient-reported outcome measures, patient-reported experience measures, and economic evaluations. Finally, external validation studies should assess generalizability across patient subgroups stratified by disease severity, comorbidity burden, and sociodemographic characteristics. Table 5Proposal for a Framework of Development and Evaluation Trajectories of RM Studies in Liver CirrhosisDesignRecruitment siteOutcomesSettingValidation cohortFollow-up durationMost promising RM tools optimizationFeasibility (pilot) studyInpatientPrimary:•Dropout rate and time to dropout (usability and clinical validation)•Compliance rate (usability validation)•Patient and provider satisfaction through PROMs and PREMs (usability validation)•Number of alerts (verification and analytical validation)•Number of (alert-based) contacts and interventions (analytical and usability validation)•Costs: healthcare consumption, RM use, QALY (clinical validation)Secondary (clinical validation):•Hospital readmission rate•Mortality rate•Decompensation events according to Baveno VII classificationSingle or multicenterNot applicable≥6 moOutpatientProspective cohort with control group, randomized or nonrandomizedInpatientPrimary (clinical validation):•Hospital readmission rate•Mortality rate•Decompensation events according to Baveno VII classificationSecondary:•Dropout rate and time to dropout (usability and clinical validation)•Compliance rate (usability validation)•Patient and provider satisfaction through PROMs and PREMs (usability validation)•Number of alerts (verification and analytical validation)•Number of (alert-based) contacts and interventions (analytical and usability validation)•Costs: healthcare consumption, RM use, QALY (clinical validation)Multicenter: Secondary center, and academic transplant centerProspective internal or external cohort matched for: age, sex, etiology, MELD-Na and Child-Pugh scores, treatment center, living situation, education, and digital health literacy≥3 moOutpatient≥6 mo for decompensated patients at baseline≥12 mo for compensated patients at baselineProspective cohort comparing different RM characteristics, for example:RM contentRM frequencyRM configurations, for example: additional vital parametersPREM, patient-reported experience measures; PROM, patient-reported outcome measures; QALY, quality-adjusted life year. Proposal for a Framework of Development and Evaluation Trajectories of RM Studies in Liver Cirrhosis •Dropout rate and time to dropout (usability and clinical validation) Dropout rate and time to dropout (usability and clinical validation) •Compliance rate (usability validation) Compliance rate (usability validation) •Patient and provider satisfaction through PROMs and PREMs (usability validation) Patient and provider satisfaction through PROMs and PREMs (usability validation) •Number of alerts (verification and analytical validation) Number of alerts (verification and analytical validation) •Number of (alert-based) contacts and interventions (analytical and usability validation) Number of (alert-based) contacts and interventions (analytical and usability validation) •Costs: healthcare consumption, RM use, QALY (clinical validation) Costs: healthcare consumption, RM use, QALY (clinical validation) •Hospital readmission rate Hospital readmission rate •Mortality rate Mortality rate •Decompensation events according to Baveno VII classification Decompensation events according to Baveno VII classification •Hospital readmission rate Hospital readmission rate •Mortality rate Mortality rate •Decompensation events according to Baveno VII classification Decompensation events according to Baveno VII classification •Dropout rate and time to dropout (usability and clinical validation) Dropout rate and time to dropout (usability and clinical validation) •Compliance rate (usability validation) Compliance rate (usability validation) •Patient and provider satisfaction through PROMs and PREMs (usability validation) Patient and provider satisfaction through PROMs and PREMs (usability validation) •Number of alerts (verification and analytical validation) Number of alerts (verification and analytical validation) •Number of (alert-based) contacts and interventions (analytical and usability validation) Number of (alert-based) contacts and interventions (analytical and usability validation) •Costs: healthcare consumption, RM use, QALY (clinical validation) Costs: healthcare consumption, RM use, QALY (clinical validation) RM content RM content RM frequency RM frequency RM configurations, for example: additional vital parameters RM configurations, for example: additional vital parameters PREM, patient-reported experience measures; PROM, patient-reported outcome measures; QALY, quality-adjusted life year. Discussion Main Findings This systematic review shows that RM in clinical care for patients with liver cirrhosis benefits the outpatient management of patients with liver cirrhosis. All studies with a control group, including 1 RCT, reported outcomes favoring RM, most commonly reflecting reductions in hospital readmissions or related healthcare utilization. The majority of included studies primarily assessed feasibility, consistently demonstrating high patient adherence and positive patient-reported experiences. Nonetheless, this review demonstrates the research gaps in RM interventions evaluated for disease outcomes in patients with liver cirrhosis. Most studies were small, single-center feasibility studies, with substantial variety in study design, intervention characteristics, and reported outcomes, limiting assessment of clinical endpoints. Evaluation using the V3+ framework revealed that reporting on verification, usability, and analytical validation was incomplete for several RM tools. Signals of Clinical Potential and Feasibility The included studies consistently demonstrated clinical effects in reducing hospital (re)admissions, as has been acknowledged in previous reviews, indicating a robust and coherent pattern across different study designs and settings.5,11 High adherence and patient satisfaction emphasize the acceptability of digital tools in this population, including patients with lifestyle-related etiologies, supporting RM applicability in real-world contexts.7,24 Importantly, the inclusion of patients with advanced cirrhosis is highly relevant, given the high readmission rates and potential health gains through RM.34 Heterogeneity as a Limiting Factor for Evidence Synthesis In this study, we observed heterogeneity across different domains, characteristic of an emerging research field still in an exploratory and developmental phase. This includes variation in study objectives (feasibility vs effectiveness), intervention characteristics, outcome measures, and follow-up duration, complicating comparisons across studies. These observations are supported by statements by Gananandan et al, highlighting the need for standardization of digital tools across healthcare settings with generalizable components.5 Evidence generation was further constrained by predominantly single-center designs, limited follow-up, and the frequent use of internal or retrospective control groups, limiting causal inference and increasing the risk of selection bias. Finally, the predominance of academic transplant populations and limited reporting on key feasibility determinants, like sociodemographic factors, raises concerns about the generalizability to broader hepatology settings, in compensated, outpatient groups, and low-resource contexts.9 Certainly, these restrictions reflect the reality that developing, evaluating, and implementing innovations necessarily involves an iterative and practice-oriented approach. Contextualization Within Existing Digital Health Literature The clinical benefits of RM have been more extensively studied in patients with cardiac failure, as presented by meta-analyses that suggest a more advanced stage of RM development and methodological rigor. Despite this relatively robust evidence base, the evidence diversity has resulted in only a weak recommendation for non-invasive RM in the most recent European guidelines for acute and chronic cardiac failure.35 Similarly, heterogeneity in study outcomes has been observed by Kumar et al in RM for patients with decompensated liver cirrhosis or inflammatory bowel disease (IBD).9 The extrapolability constraints identified in this review further align with findings in the IBD, chronic obstructive pulmonary disease (COPD), and acute care population, underscoring the pervasiveness of these limitations in the field of digital health across diseases.8,36,37 This indicates that limited extrapolability likely reflects differences in RM development and scientific evaluation practices rather than disease-specific effects, necessitating a structured assessment of RM development and research. Remote Monitoring Development Quality and Implications of the V3+ Framework Assessment Accordingly, evaluating available RM projects along a systematic quality framework like the V3+ framework clarifies why cross-study comparability remains limited. By distinguishing between different stages of development, the assessment highlights that clinical validation alone is insufficient without reports on verification, usability testing, and analytical validation, aligning with the findings on intervention heterogeneity. As a result, most studies evaluating RM tools in cirrhosis primarily focus on technical feasibility rather than long-term outcomes or cost-utility, explaining the limited ability to identify the most clinically valuable RM configurations. Assessment through this framework underlines why current effect estimates remain preliminary. Implications for Future Research This review proposes a framework for the development and research trajectory of RM studies in liver cirrhosis, derived from the evidence gaps identified, to facilitate harmonization research and generation of comparable evidence. We recommend future research to build on the existing evidence base while explicitly distinguishing feasibility studies from evaluations of disease-related outcomes. Initial studies need to demonstrate reliable RM performance and consistent benefits, which are essential prerequisites for resource-intensive research and implementation strategies. Feasibility studies should prioritize the performance of monitoring of ascites and hepatic encephalopathy through body weight and orientation assessments, as these domains are currently most frequently evaluated (Table 3) and are likely to yield the largest clinical impact in follow-up studies, given their predominance as causes of hospital (re)admissions.34,38 Following confirmation of feasibility and technical reliability, prospective evaluation studies should incorporate core outcomes, long-term follow-up, and multicenter designs. Notably, Muftah et al showed that telehealth in the HCV management achieves clinical outcomes comparable to standard care.10 Accordingly, quality-of-care indicators like patient-reported outcome measures, patient-reported experience measures, and healthcare and societal costs should complement clinical endpoints to capture the total value of the intervention, as has been previously proposed in the IBD population.8 Furthermore, increasing the inclusion of compensated patients may be advantageous for RM tools focusing on general well-being and prevention of a first decompensation event, given the adverse prognostic impact of decompensation.39,40 To avoid dilution of observable clinical effects, future studies should consider stratifications or separate evaluations of compensated and decompensated patient populations. Ongoing large RCTs may further clarify the clinical benefits of RM, many of which use RM tools that have been presented in this review.41, 42, 43, 44 More importantly, collaboration and consensus among RM developers and hepatology experts are required to address the identified research gaps and optimize the design and impact of future research. Strengths and Limitations This review has several limitations. Studies focusing on self-management, without evaluating disease outcomes, were excluded, which may have led to the omission of promising pilot RM interventions. However, restricting inclusion to RM tools that have been evaluated for clinical outcomes ensured the analysis reflected their impact on clinical care. Study heterogeneity precluded quantitative analysis, including evaluation for publication bias and comparison of RM tools. However, this constitutes a bias inherent to the emerging field, as all starting RM platforms inevitably differ in certain technical or content aspects. Additionally, we highlight that the framework of development and evaluation trajectories of RM studies in liver cirrhosis is a proposal that has not been established within a consensus meeting but could serve as a start for consensus and collaboration. Strengths include a comprehensive search strategy and a systematic overview of RM interventions across multiple modalities. Importantly, this review provides a pragmatic framework of methodological quality, highlights evidence gaps, and proposes directions to advance the research in the field. Conclusion RM for patients with liver cirrhosis promotes outpatient clinical care and improves hospital readmissions. However, evidence on RM for patients with liver cirrhosis remains fragmented and methodologically limited, with highly variable quality of development processes, precluding firm conclusions on clinical outcomes or comparisons of RM tools. Although feasibility studies show promise, they should naturally be distinguished from and followed by well-powered trials using standardized protocols, including secondary hepatology centers, core outcome sets, established follow-up durations, and matched (external) validation, as proposed in this review. We call for consensus and collaboration among RM and hepatology experts, which are essential to optimize RM development processes and future studies. Meeting these requirements will provide a substantial boost to the research field and enable the successful implementation of RM in patients with liver cirrhosis. Declaration of AI and AI-Assisted Technologies in the Writing Process During the preparation of this work, the authors used ChatGPT in order to assist with language refinement and optimize readability in accordance with American English standards. The authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Acknowledgments The authors thank Remy van Alebeek from the Maastricht University Library Systematic Literature Review Support Team for his support in optimizing the search strategy. Tom J.G. Gevers and Govert Veldhuijzen contributed equally to this study and are co-senior authors. Authors’ Contributions Govert Veldhuijzen, Tom J.G. Gevers, and Britt van Ruijven conceptualized and designed the study and interpreted the data. Britt van Ruijven developed the search strategy, performed the search, and drafted the manuscript. Tom J.G. Gevers and Britt van Ruijven screened the search results and performed the risk of bias assessment. Govert Veldhuijzen assisted in the consensus meetings to resolve disagreements. Govert Veldhuijzen, Joost P.H. Drenth, Marten A. Lantinga, Marieke J. Pierik, and Tom J.G. Gevers participated in the critical revision of the manuscript for important intellectual content. Footnotes Conflicts of Interest: The authors disclose no conflicts. Funding: G. Veldhuijzen and Tom J.G. Gevers have received funding from the Dutch Digestive Health Fund, Gelre Wetenschapsfonds, and Gilead Sciences Inc. Marten A. Lantinga has received speaker fees from GORE (2024) and Takeda (2025) and a nonrestricted research grant from Vaillant fonds (2023). Joost P. H. Drenth is a consultant for CAMURUS, all proceeds go to the Amsterdam UMC. Marieke J. Pierik has received nonrestricted research grants from Horizon 2020, ZONMW (Dutch national research fund), the Dutch Digestive Health Fund, Takeda, Johnson & Johnson, AbbVie, and Galapagos, as well as nonfinancial support from Immunodiagnostics. Additionally, Marieke J. Pierik has received speaker’s fees from BMS, Janssen Cilag, and Takeda, all outside the submitted work. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. Ethical Statement: Ethical approval was not required as this was a systematic review, together with included studies, performed in accordance with the Declaration of Helsinki, for which no patients were recruited. Data Transparency Statement: The data used and/or analyzed during the current study are available from the corresponding author on reasonable request. Reporting Guidelines: PRISMA guideline for reporting systematic reviews. Material associated with this article can be found, in the online version, at https://doi.org/10.1016/j.gastha.2026.100997. Supplementary Materials Supplementary Material Extended PDF References Associated Data Supplementary Materials Supplementary Material Extended PDF