Original investigation
Published: 2019-06-15

Is telemonitoring effective in older adults affected by heart failure? A meta-analysis focused on this population

Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy
Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy
Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy
Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy
Heart failure Telemonitoring Aged Meta-analysis

Abstract

Background and aims. Telemonitoring might improve outcomes in older adults affected by heart failure. However,
study results are contrasting, and no meta-analyses are available focused on this population. The objective
of this meta-analysis was to assess the effectiveness of telemonitoring in reducing all-cause mortality,
all-cause and heart failure-related hospitalizations and emergency department visits in older adults affected
by heart failure.
Methods. We performed a systematic search on Pubmed for randomized controlled trials published up to 31st
October 2018 studying the effectiveness of telemonitoring in improving outcomes, compared to usual care,
in older adults with HF or with a pre-specified sub-analysis on this population. Both fixed and random-effects
models were used to calculate the pooled RR (95% CI).
Results. 8 randomized controlled trials (1909 participants, mean age 76.8 years, 53% female) were included
in the meta-analysis. Telemonitoring did not reduce mortality (RR 0.87, 95% CI 0.69-1.09), emergency department
visits (RR 1.00, 95% CI 0.72-1.39), nor all-cause hospitalizations (RR 0.96, 95% CI 0.84-1.10), while was
evident a reduction in heart failure-related hospitalizations (RR 0.76, 95% CI 0.56-1.03). Heterogeneity across
studies was low (I2 39% for emergency department visits, and about 0% for all the other outcomes).
Conclusions. Telemonitoring might reduce the risk of heart failure-related hospitalizations in older adults affected
by heart failure. Further studies are needed to assess the role of telemonitoring in this population, taking
into account comorbidities, disability, frailty, in order to clearly identify which patients may most beneficiate of
this technology.

INTRODUCTION

Heart failure (HF) is a chronic progressive condition with prevalence that increases with age, reaching 10% in older adults 1. The natural history of this disease is characterized by a progressive course and frequent hospitalizations, that are more frequent in the advanced stages of the disease 2 3.

In the latest years remote telemonitoring was proposed as a technology for the management of HF patients, and many randomized controlled trials and meta-analyses 4 showed a reduction in HF-related hospitalizations and all-cause death in participants allocated to telemonitoring respect to usual care. Considering these evidences, the latest ESC guidelines on diagnosis and treatment of HF recommended utilization of communication technologies to increase access to health care 5 but also pointed out that recent large randomized controlled trials (RCTs) documented that telemonitoring did not improve outcomes of patients affected by HF 6 7. However, a subsequent meta-analysis documented that telemonitoring could be more effective in reducing all-cause mortality and HF-related hospitalizations in selected populations or using a more technologic and multi-parameter evaluating telemonitoring (defined by the authors “complex telemonitoring”) 8.

Few evidences are available on older adults, the population with highest prevalence of HF and HF-related hospitalization, and in which the presence of multimorbidity, reduction of functional autonomy, and frailty, may influence the effectiveness of telemonitoring. To the best of our knowledge, there are not large studies evaluating the effectiveness of telemonitoring in patients affected by HF and focused exclusively on older adults. Some meta-analyses showed data stratified by age, but stratification was based on mean or median age, and thus the “older” group could also include participants with relatively young age 8 9. The inclusion of people younger than 65 years old might have significantly changed the results, because of different clinical (comorbidities) and physical conditions of younger people respect to older adults. Furthermore, the most recent meta-analysis 8 included also observational studies, that might have introduced a bias in the results.

The objective of this study was to review the literature and to perform a meta-analysis on the effectiveness of remote telemonitoring in reducing deaths, hospitalizations and emergency department (ED) visits in older adults affected by HF.

MATERIALS AND METHODS

SEARCH STRATEGY AND STUDY SELECTION

This systematic review and metanalysis was performed according to the Cochrane Guidelines. We searched the Pubmed database for articles published up to 31st October 2018. The search keywords were “heart failure” and “telemonitoring”, “telemedicine”, or “telehealth”. Telemonitoring was defined as a transmission of objective clinical data, such as blood pressure, oxygen peripheral saturation and body weight. Telephone interviews were included only if there was a transmission of clinical objective data. Eligible outcomes were all-cause and cardiovascular mortality, all-cause and HF-related hospital admissions, and ED visits.

A systematic screening of titles and abstracts for the search keywords was performed by two blinded reviewers; the potential eligible ones were obtained in full-text for evaluation. The bibliography of these studies was also searched to retrieve studies. Once selected the articles, when needed, corresponding authors were contacted to obtain additional data for the meta-analysis.

We included only RCTs evaluating the effectiveness of telemonitoring performed on participants affected by HF with age ≥ 65 years, or with a pre-specified sub-analysis for participants aged 65 years or over and for HF. We excluded studies published in languages different than English or performing invasive monitoring or tele-rehabilitation only. Clinical trials that studied only health-related costs and/or quality of life were also excluded.

DATA EXTRACTION AND QUALITY ASSESSMENT

We abstracted the information about first author’s name, study population (setting, number of participants per group, gender, mean age), intervention and usual care characteristics, outcomes, and mean follow-up time.

A systematic assessment of bias in the included studies was performed using the Cochrane Criteria (low risk, high risk, unclear) 10: the items used for the assessment of each study were adequacy of sequence generation, allocation concealment, blinding, addressing of dropouts (incomplete outcome data), selective outcome reporting, and other potential sources of bias.

DATA SYNTHESIS

We calculated the Relative Risk (RR) and 95% confidence interval (CI) of each study and the pooled adjusted RR (95% CI) using both fixed and random-effect models. Heterogeneity across studies was evaluated using the I2 statistics. I2 > 50% was consider reflecting statistically significant heterogeneity. The risk of publication bias was evaluated using funnel plots and the trim and fill method.

Statistical analyses were performed using R version 3.5.2 (R Foundation for Statistical Computing, Vienna, Austria).

RESULTS

From 944 titles retrieved from the Pubmed search, 202 were identified for abstract evaluation; 80 of those were obtained for full-text examination. Ten RCTs met the inclusion criteria, and their bibliography was searched to retrieve studies, with no additional RCTs included. Two RCTs were excluded because reported only composite outcomes 11 or risk ratio 12. Eight RCTs had sufficient information available and were therefore included in the meta-analysis (Fig. 1).

Six out of 8 studies were focused exclusively on older adults; the other 2 had a pre-specified sub-analysis on older adults 7 13. For the latest two studies no information on age and sex was available. Mean age ranged from 72 14 to 80 years 15, and prevalence of female sex from 28% 14 to 65% 16. Mean follow-up time ranged from 3 17 to 21.5 months 13. Most of the studies included participants hospitalized for HF 14 17 18, or ambulatory HF participants 7 13. The characteristics of the telemonitoring intervention differed across studies: they ranged from weekly telephone interviews with reassessment of pharmacological therapy and/or scheduling of new clinical visits if indicated 18 to daily remote monitoring evaluated daily by a nurse or a physician with reassessment of pharmacological therapy and/or scheduling of new clinical visits if indicated 13-17 19. Furthermore, the parameters evaluated ranged from body weight only 7 16 17 to the evaluation of three or more parameters 13-15 18 (Tab. I).

A total of 1909 participants was included in the meta-analysis (telemonitoring: 975, usual care: 934); mean age of the study population was 76.8 years, 53% were female (excluding the two studies without information on age and sex). All-cause mortality was collected in all the 8 RCTs considered, while cardiovascular mortality was available for 2 RCTs only, and was not analysed.

Telemonitoring did not reduce all-cause mortality respect to usual care, as documented by both fixed and random-effects models, that were superimposable: RR: 0.87, 95% CI: 0.69-1.09 (Fig. 2). There was no heterogeneity across studies, as documented by the I2, that was 0%. The funnel plot documented a risk of publication bias for this outcome, with three potentially missing study on the right side identified using the trim and fill method (Fig. 3).

All-cause hospital admissions were available in 5/8 studies. Telemonitoring did not reduce the risk for this outcome, as documented by fixed and random-effects models (RR: 0.96, 95%CI: 0.84-1.10 for both models) (Fig. 4). The heterogeneity across studies was very low, with a I2 of 0.01%; there was a risk of publication bias, but with only one potentially missing study, on the right side of the funnel plot, as identified with the trim and fill method (Fig. 5).

HF-related hospital admissions were reported in 4/8 studies. Telemonitoring reduced the risk of HF-related hospital admission of 24%, reduction that, however, did not reach the statistical significance (RR: 0.76, 95% CI: 0.56-1.03) (fixed and random-effects model were superimposable) (Fig. 6). There was not heterogeneity across studies, as documented by the 0% of I2. The funnel plot documented a low risk of publication bias, as confirmed by the trim and fill correction, that identified only one potentially missing study (Fig. 7).

The number of ED visits for each group was available for 4/8 RCTs. The random-effects model documented that telemonitoring did not influence the number of ED visits: RR: 1.00, 95% CI: 0.72-1.39) (Fig. 8). Similar results were found using a fixed-effects model (RR: 1.03, 95% CI: 0.83-1.28) (Fig. 9). The heterogeneity between studies was low, with a I2 of 39.5%. The risk of publication bias was very low, as documented by the funnel plot and the trim and fill correction (Fig. 10).

Figure 11 shows the risk of bias of the studies included in the meta-analysis: the highest risk is related to the random sequence generation and the allocation concealment, that was unknown in three and four studies, respectively, and in data outcomes, that were incomplete in 3 studies or with a selected reporting in other 3 studies.

DISCUSSION

The results of our meta-analysis suggest that telemonitoring does not reduce the risk of negative outcomes in older adults affected by HF, except for HF-related hospital admissions.

Currently, the role of telemonitoring for the management of patients with HF is debated: the latest ESC guidelines for diagnosis and treatment of HF pointed out that telemonitoring might be ineffective in these patients 5. This consideration was related to the negative results of recently published large RCTs on this topic 6 7. However, in one of these RCTs, stating that in stable, optimally treated, ambulatory chronic HF patients telemonitoring did not reduce mortality, the authors suggested that some factors may influence the effectiveness of telemonitoring in this population, such as the history of a recent hospitalization for acute HF 6. In fact, in their subsequent RCT, the TIM-HF2 trial, the authors selected participants at higher risk of negative outcomes, such as in NYHA class II-III, or with at least one HF-related hospitalization in the previous year, and found a reduction in all-cause mortality and in proportion of days lost for cardiovascular-related hospitalization in telemonitored patients 20. Similarly, Aronow and Shamliyan, in their meta-analysis, documented that telemonitoring was not effective in improving the outcomes, except for HF-related hospitalization, while an effectiveness in reducing all-cause mortality and all-cause hospitalizations and a further reduction in HF-related hospitalization was evident for patients performing complex monitoring and for patients aged 70 years or over 8. Thus, the results of above-mentioned meta-analysis suggested that telemonitoring could to be more effective in older adults. However, the results of our meta-analysis are in contrast with these previous mentioned findings, showing that telemonitoring only slightly reduce the risk of HF-related hospitalizations in older adults. This discrepancy might be related to the different characteristics of the two meta-analyses, such as age inclusion criteria: we included exclusively older adults, while Aronow and Shamliyan included all the studies having a mean or median age of 70 years or over, thus not completely excluding younger participants; furthermore, the authors included also observational studies, that might have introduced a bias in the results, and did not include all the RCTs considered in our study. Compared to other meta-analyses, our inclusion criteria, especially the focus on studies not enrolling participants with age < 65 years, might have selected patients with more comorbidities, older, more frequently with a multi-drug therapy, with a higher prevalence of frailty and disability, all factors that can negatively influence the outcomes, independently from the diagnosis of HF 21-25. Unfortunately, most of the studies did not consider these factors, thus we could not perform a sub-analysis to assess the veracity of this speculation. However, our hypothesis is supported by the evidence of the effectiveness of telemonitoring exclusively in reducing HF-related hospitalization and not in all-cause hospitalizations, death or ED visits, all outcomes that could be influenced by the above-mentioned factors.

This is the first meta-analysis evaluating the effectiveness of telemonitoring in the management of HF exclusively focused on older adults. Nonetheless the large differences across studies in characteristics of telemonitoring and in inclusion criteria, the documented low heterogeneity across studies made reliable the reported results. This study has many limitations: there was a relative high publication bias, that may have influenced our results; furthermore, our analysis included a relatively small number of participants, due to the few available RTCs analysing this population and their relatively small sample size.

In conclusion, telemonitoring might be effective in reducing HF-related hospitalizations in older adults affected by HF, while the other outcomes might be primarily influenced by other factors, such as comorbidities or disability, that characterized this population. A careful selection of the patients that may benefit of telemonitoring, such as patients with moderate-severe HF, or without other diseases that may influence the primary outcomes (e.g. moderate-severe COPD, poli-vasculopathy), might improve the effectiveness of this technology in this population.

Further studies are needed to assess the role of telemonitoring in older adults, with a special attention to comorbidities, disability, frailty, in order to clearly identify which patients may most benefit of this technology.

Figures and tables

Figure 1.Flow-diagram of the study selection.

Figure 2.Forest plot of relative risk for all-cause mortality.

Figure 3.Risk of publication bias for all-cause mortality. Open circles represent the potential missing studies identified using the Trim and Fill method.

Figure 4.Forest plot of relative risk for all-cause hospitalization.

Figure 5.Risk of publication bias for all-cause hospitalization. Open circles represent the potential missing studies identified using the Trim and Fill method.

Figure 6.Forest plot of relative risk for heart failure-related hospitalization.

Figure 7.Risk of publication bias for heart failure-related hospitalization. Open circles represent the potential missing studies identified using the Trim and Fill method.

Figure 8.Forest plot of relative risk for emergency department visit (random-effects model).

Figure 9.Forest plot of relative risk for emergency department visit (fixed-effects model).

Figure 10.Risk of publication bias for emergency room visit. Open circles represent the potential missing studies identified using the Trim and Fill method.

Figure 11.Risk of bias of the included studies.

First author, year, nation TM/UC group size Mean age (SD) Sex (female, %) Primary outcomes FU time (months) Parameters considered TM intervention characteristics UC characteristics Setting
Antonicelli, 2008 Italy 18 28/29 78(7) 39 HA + D 12 BP, HR, BW, ECG, diuresis Weekly collection of parameters by telephone interviews; therapy reassessed and visit scheduled if needed Lifestyle and therapy adherence counseling; scheduled HF visits Acute ward for HF worsening
Chaudhry, 2010 US 7 330/306 - - HA, D 6 BW Daily answer to automatic questions through phone keypad. Daily check by site coordinator. Phone call in case of alert generation or missing data for two consecutive days Delivery of educational materials and, if needed, of a scale Cardiology practices
Koehler, 2012 Germany 13 155/158 - - D 21.5 ECG, BP, BW Data transmission through a smartphone, daily evaluated by a physician. Phone calls or visits if indicated. Structured phone calls monthly Visits every 3 months for the 1 year, then every 6 months Stable ambulatory HF patients from cardiology, internal medicine or general medicine
Pedone, 2015 Italy 15 47/43 80(7) 61 HA + D 6 BP, BW, POS Data transmission through a smartphone, daily evaluated by a geriatrician. Alert if measurement outside predefined range. Phone calls or visits if indicated Lifestyle and therapy counseling; 1 month-FU visit Geriatric acute care ward or outpatient clinic
Schwarz, 2008 US 17 51/51 78.1(7) 50 HFHA 3 BW Data transmission through telephone line, daily evaluated by a nurse. Phone calls if values outside prescribed parameters NA Acute care ward
Soran, 2008 US 16 160/155 76(7) 65 CD + HFHA 6 BW Daily transmission of BW via phone line, daily reviewed by a nurse. Phone call if alerts; office visits with physician if indicated Education to daily weigh; education materials. Visits at baseline and at 6 months Three centers of primary care
Villani, 2014 Italy 14 40/40 72(3) 28 CD, HFHA, ED 12 BP, HR, BW, ECG A cardiologist at discharge decides which parameter monitor and with which frequency. Remote transmission of data. At a pre-scheduled time (NA) the cardiologist analyzes the information and contact the patient if indicated Education on everyday management of the disease by a specialized nurse. Follow-up visits every 3 months Acute ward for HF worsening
Wade, 2011 US 19 164/152 78.1 48 HA, ED, D 6 BP, BW Wireless data collection and transmission every weekday. The information was checked for alerts. Care coordination assistance as needed Education, medication management, care coordination needs. Scheduled calls (initially 2-3 times/week) Aetna Medicare Advantage members
Table I.Characteristics of the studies included in the meta-analysis.

References

  1. Kannel WB. Current status of the epidemiology of heart failure. Curr Cardiol Rep. 1999; 1:11-9.
  2. Quinones PA, Seidl H, Holle R. New potential determinants of disability in aged persons with myocardial infarction: results from the KORINNA-study. BMC Geriatr. 2014; 14:34.
  3. Lelli D, Pedone C, Rossi FF. Clinical and echocardiographic characteristics of elderly hospitalized patients with high levels of NT-proBNP without clinical diagnosis of heart failure. Aging Clin Exp Res. 2014; 26:607-13.
  4. Inglis SC, Clark RA, Dierckx R. Structured telephone support or non-invasive telemonitoring for patients with heart failure. Cochrane Database System Rev. 2015. Publisher Full Text | DOI
  5. Ponikowski P, Voors AA, Anker SD. 2016 ESC Guidelines for the diagnosis and treatment of acute and chronic heart failure. Eur J Heart Fail. 2016; 18:891-75.
  6. Koehler F, Winkler S, Schieber M. Impact of remote telemedical management on mortality and hospitalizations in ambulatory patients with chronic heart failure: the Telemedical Interventional Monitoring in Heart Failure study. Circulation. 2011; 123:1873-80.
  7. Chaudhry SI, Mattera JA, Curtis JP. Telemonitoring in patients with heart failure. N Eng J Med. 2010; 363:2301-9.
  8. Aronow WS, Shamliyan TA. Comparative effectiveness of disease management with information communication technology for preventing hospitalization and readmission in adults with chronic congestive heart failure. J Am Med Directors Assoc. 2018; 19:472-9.
  9. Inglis SC, Conway A, Cleland JG. Is age a factor in the success or failure of remote monitoring in heart failure? Telemonitoring and structured telephone support in elderly heart failure patients. Eur J Cardiovasc Nurs. 2015; 14:248-55.
  10. Higgins JPT, Altman DG, Gøtzsche PC. The Cochrane collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011; 343:d5928.
  11. Madigan E, Schmotzer BJ, Struk CJ. Home health care with telemonitoring improves health status for older adults with heart failure. Home Health Care Serv Q. 2013; 32:57-74.
  12. Tompkins C, Orwat J. A randomized trial of telemonitoring heart failure patients. J Healthcare Manag. 2010; 55:312-22.
  13. Koehler F, Winkler S, Schieber M. Telemedicine in heart failure: pre-specified and exploratory subgroup analyses from the TIM-HF trial. Int J Cardiology. 2012; 161:143-50.
  14. Villani A, Malfatto G, Compare A. Clinical and psychological telemonitoring and telecare of high risk heart failure patients. J Telemed Telecare. 2014; 20:468-75.
  15. Pedone C, Rossi FF, Cecere A. Efficacy of a physician-led multiparametric telemonitoring system in very old adults with heart failure. J Am Geriatrics Soc. 2015; 63:1175-80.
  16. Soran OZ, Piña IL, Lamas GA. A randomized clinical trial of the clinical effects of enhanced heart failure monitoring using a computer-based telephonic monitoring system in older minorities and women. J Cardiac Failure. 2008; 14:711-7.
  17. Schwarz KA, Mion LC, Hudock D. Telemonitoring of heart failure patients and their caregivers: a pilot randomized controlled trial. Progress Cardiovasc Nurs. 2008; 23:18-26.
  18. Antonicelli R, Testarmata P, Spazzafumo L. Impact of telemonitoring at home on the management of elderly patients with congestive heart failure. J Telemed Telecare. 2008; 14:300-5.
  19. Wade MJ, Desai AS, Spettell CM. Telemonitoring with case management for seniors with heart failure. Am J Managed Care. 2011; 17:9.
  20. Koehler F, Koehler K, Deckwart O. Efficacy of telemedical interventional management in patients with heart failure (TIM-HF2): a randomised, controlled, parallel-group, unmasked trial. Lancet. 2018; 392:1047-57.
  21. Vermeiren S, Vella-Azzopardi R, Beckwée D. Frailty and the prediction of negative health outcomes: a meta-analysis. J Am Med Dir Assoc. 2016; 17:1163.e1-1163.e17.
  22. Leistner DM, Münch C, Steiner J. Effect of physical disability on mortality in elderly patients of ≥ 80 years of age undergoing percutaneous coronary intervention. Am J Cardiol. 2018; 122:537-41.
  23. Testa G, Cacciatore F, Bianco A. Chronic obstructive pulmonary disease and long-term mortality in elderly subjects with chronic heart failure. Aging Clin Exp Res. 2017; 29:1157-64.
  24. Herr M, Robine J-M, Pinot J. Polypharmacy and frailty: prevalence, relationship, and impact on mortality in a French sample of 2350 old people. Pharmacoepidemiol Drug Saf. 2015; 24:637-46.
  25. Martín-Sánchez FJ, Rodríguez-Adrada E, Vidán MT. Impact of geriatric assessment variables on 30-day mortality among older patients with acute heart failure. Emergencias. 2018; 30:149-55.

Affiliations

D. Lelli

Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy

R. Antonelli Incalzi

Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy

V. Adiletta

Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy

C. Pedone

Area di Geriatria, Università Campus Bio-Medico di Roma, Rome, Italy

Copyright

© Società Italiana di Gerontologia e Geriatria (SIGG) , 2019

How to Cite

[1]
Lelli, D., Antonelli Incalzi, R., Adiletta, V. and Pedone, C. 2019. Is telemonitoring effective in older adults affected by heart failure? A meta-analysis focused on this population. JOURNAL OF GERONTOLOGY AND GERIATRICS. 67, 2 (Jun. 2019), 87-95.
  • Abstract viewed - 32 times
  • PDF downloaded - 15 times