Telemonitoring in Patients with Heart Failure - Lessons from Recent Randomised Multicentre Trials

Login or register to view PDF.
Abstract

Home telemonitoring (TLM) has been proposed as an effective tool to reduce cardiac events in patients with chronic heart failure (HF). In contrast to older and more recent meta-analyses, large randomised multicentre trials have failed to demonstrate any positive effect of TLM on HF rehospitalisation and all-cause mortality. However, these negative results do not preclude the potential role of TLM as an effective system for managing patients with HF. In this article, possible explanations for these negative results are presented and discussed, and a new model of TLM, with investments in technology but above all in personnel and organisation, is suggested.

Disclosure
The author has no conflicts of interest to declare.
Correspondence
Andrea Mortara, Heart Failure Unit, Cardiology Department, Policlinico di Monza, Via Amati 111, 20900 Monza, Italy. E: andreamortara@libero.it
Received date
09 January 2012
Accepted date
18 February 2012
Citation
European Cardiology, 2012;8(2):84-7
DOI
http://dx.doi.org/10.15420/ecr.2012.8.2.84

Data from Italian, European and US registries clearly show that a large proportion of acute hospitalisations for heart failure (HF) could be avoided by improving the follow-up of patients at home after discharge.1–3 There are many factors that may contribute to decompensation, including reduced coronary perfusion, anaemia, hypertension, infections, arrhythmias and renal failure. The important question is whether these phases of acute decompensation can be prevented, and whether telemonitoring (TLM) can help.

As one might expect, data from the literature are conflicting. Theoretically, TLM can be helpful:

  • to monitor weight and fluid balance and adjust diuretic therapy;
  • to avoid over-diuresis and hypotension;
  • to control heart rate, blood pressure and arrhythmias;
  • to enable infection surveillance;
  • to better titrate therapy; and
  • to enhance educational intervention and improve self-management counselling.

However, from a practical point of view, three crucial factors are fundamental to make TLM work. First, the physiological variables meant to reflect an early phase of decompensation need to be transmitted efficiently. Second, information received by qualified personnel need be translated into specific recommendations to correct the perturbation. Third, patients who receive new recommendations must correctly implement the interventions. At the end of this process, confirmation is needed that the perturbation has been solved, or alternatively that other interventions are needed. This last part of the process is the core of TLM because it translates a simple monitoring action into effective management, with a possible decrease of events such as HF rehospitalisation or cardiac death. Telemedicine may support healthcare when distance separates the patient from the healthcare provider. However, the remote care of patients with HF is not an easy task because the disease is often complex, with many possible co-morbidities and multiple factors contributing to new phases of decompensation.

Evidence from Recent Multicentre Trials

Returning to the question of whether TLM can help, one should consider the data published so far. Many reviews and meta-analyses have been published in recent years. A recent review by Inglis et al.4 confirmed that both HF hospitalisation and all-cause mortality are positively affected by TLM, with a relative risk (RR) of 0.79 (95 % confidence interval [CI] 0.67–0.94) and 0.66 (95 % CI 0.54–0.81), respectively, and that TLM significantly reduces costs.5 Meta-analyses have been largely criticised because they are often small, single-centre studies and not randomised studies. When looking at the most recently published multicentre studies, data appear, at first glance, to be less convincing.

The largest multicentre study, published in the US at the end of 2010 by Chaudhry et al.,6 concluded that TLM is not helpful in reducing cardiac events after discharge from hospital in patients admitted for acute HF. Patients were enrolled if they had been hospitalised in the previous 30 days. Their mean age was 61 years, 42 % had New York Heart Association (NYHA) functional class I–II symptoms and 30 % had preserved left ventricular ejection fraction (LVEF) (>40 %). The TLM group was instructed to make daily toll-free calls to the system. During each call, patients were asked a series of questions and they entered responses using the telephone keypad. Information was reviewed every weekday by site co-ordinators. After a follow-up of six months, no significant differences were observed for all primary and secondary endpoints between the TLM group and the control group. Because patients were enrolled after an acute episode of decompensation, medical treatment was not optimised and the incidence of cardiac events was high – nearly 50 % at six months.

The second important study was published by Koehler et al. in 2011.7 This German multicentre study enrolled ambulatory patients who had stable disease. Their mean age was 67 years, the mean LVEF was 27 %, 50 % of patients had NYHA functional class III symptoms and more than 90 % were optimally treated with either angiotensin receptor blockers and β-blockers or angiotensin-converting enzyme inhibitors. The TLM system was very advanced, with direct bluetooth transmission of vital sign parameters to the core lab and a parallel system for handling emergency calls. Again, death and hospitalisation for HF were similar in the TLM and control groups. An analysis of the Kaplan–Meier curves shows that the percentage of patients who experienced cardiac events was only 10 %, versus 50 % in the US study by Chaudhry et al.,6 which enrolled patients admitted for acute HF.

Two other European randomised multicentre studies have been published in recent years. In 2005, data from the Trans-European network home care management system (TEN-HMS) study were published.8 The trial was conducted in the UK, the Netherlands and Germany and enrolled patients 48 hours to six weeks after hospitalisation for acute HF. Eighty-five patients were randomised to a control group, 170 to a nurse-led monitoring system group and 163 to a TLM group. Patients in the TLM group had to transmit clinical parameters daily to the co-ordinating centre and, at the end of the study, compliance to this intensive programme was found to be quite low. A significant effect of TLM was demonstrated for cardiac death but not for HF hospitalisation. Similar to the US study,6 cardiac events in the control group occurred in nearly 50 % of patients.

In 2009, Mortara et al. presented the results of the Home or hospital in heart failure (HHH) study.9 Patients with chronic HF were enrolled during a stable phase of the disease in three European countries: Italy, Poland and the UK. The incidence of cardiac events at six months was 10 %, as in the German study,7 and again no differences in the primary endpoints (cardiac death and HF hospitalisation) were observed between controls and the TLM group. Interestingly, when looking at the Italian arm of the study, which accounted for more than 50 % of the total study population, the TLM group exhibited a significantly lower number of cardiac events. This was explained by the extensive experience of Italian centres in the use of TLM and remote management of HF patients.

Why the Trials May Have Failed to Demonstrate Benefits from Telemonitoring

The main message of all these studies seems to be negative. Why did these trials fail to demonstrate any benefit from telemonitoring? Possible explanations are listed in Table 1. Of course, the obvious one is that TLM does not work and cannot have a positive effect on prognosis, whether after an acute event or in a chronic phase of the disease. However, this is not necessarily the case. Further explanations for these negative results are discussed below.

In these trials, the impression is that TLM has been used in the same way that a drug is tested in a randomised population (the drug is given and a positive or negative response is awaited during follow-up). None of them reported in any detail the interventions suggested by the core lab to address initial signs of decompensation, or the processes put in place to formulate recommendations for patients after reception of their clinical parameters. The telemedicine system should be supported by an expert in HF who knows the patients’ general disease characteristics and co-morbidities. However, it is likely that, when centres enrolled patients shortly after an acute event, they did not have time to gather detailed information about these patients for correct home monitoring after discharge. Moreover, even when the team had extensive expertise in HF and knowledge of their patients, was it sufficiently prepared and organised to manage rapidly and efficiently a large flow of TLM data? Enrolling centres often become involved in such multicentre studies because of their experience of HF, not because they have a structured organisation that can handle large volumes of clinical data and actively manage information derived from these data. The HHH study may be considered a good example of this. Italian centres contributed to testing and ameliorating the TLM system used later on in the trial and, in the preceding pilot study, they had improved their abilities in telemonitoring and telemanagement. This expertise translated into a significant reduction in clinical events that was not observed to the same extent in the other participating European centres (see Figure 1).

Thus it is time to design specific training courses in telemedicine and to establish new methods of compensation for all the personnel involved in complex clinical systems remotely caring for HF patients at home.

Another important factor that may have affected the results of the TLM trials is the short observational time in all the trials (6–12 months) and, as a consequence, the fact that the enrolling centres only had a short period of time to gain experience in telemedicine and become familiar with the new technologies.

The use of clinical indicators that were perhaps not sufficiently effective in predicting future destabilisation may also explain the results of the TLM trials. Classical indicators are heart rate, blood pressure, compliance with therapy and weight changes. However, although an increase in weight is recognised as an important sign of worsening congestion related to unfavourable prognosis, it has recently been shown that changes in weight per se explain less than 20 % of the destabilisation process.10 One single parameter may not be enough to capture initial signs of decompensation. Interesting findings were reported in the recent Program to access and review trending information and evaluate correlation to symptoms in patients with heart failure (PARTNERS HF) study,11 which included nearly 700 patients with an implantable cardioverter defibrillator (ICD) already in place. The authors developed a dynamic algorithm and score based on intrathoracic impedance fluctuation, low patient activity, reduced heart rate variability, high heart rate at night and number of ICD shocks, which conferred a 5.5-fold increased risk of acute HF in the subsequent month (RR 5.5, 95 % CI 3.4–8.8). Therefore it can be concluded that a multi-parameter approach may be more effective than home monitoring of single parameters.

It should be emphasised that, in the PARTNERS HF study, the devices enabled the automatic monitoring of parameters without direct intervention from the patients. One of the leading companies in the field (Boston Scientific, USA) as recently developed an integrated communication system that is placed in the patients’ home and directly collects parameters from a cardiac resynchronisation therapy (CRT) device or an ICD, plus weight and blood pressure from a scale and sphygmomanometer. All parameters are then automatically transmitted through a wireless connection to a web database that can be checked daily and reviewed by personnel involved in the TLM system. Also, new haemodynamic sensors for monitoring atrial or pulmonary arterial pressures have been tested in the last few years.12 All these new parameters, combined with the classical indicators, may make TLM more beneficial in future.

Multicentre trials may have failed to show benefits from TLM because telemonitoring mainly relied on patient-initiated communication, and this may have led to an underuse of the system, particularly if there was no adequate feedback provided. In the US study, Chaudhry et al.6 reported that 14 % of the patients assigned to the TLM arm never used the system and, at the end of the study, only 55 % of patients in the TLM arm were still sending data. By contrast, the randomised trial of telephone intervention in chronic heart failure (DIAL trial), in which TLM was not patient- but nurse-driven (with periodic phone calls to patients at home) showed reduced HF hospitalisations during the study itself and during follow-up.13

Finally, TLM may not have been effective because it was not integrated into an HF programme and because the alerting flags were not tailored to individual patients. For example, some fluctuations in weight (e.g., ±2 kg) may be acceptable for some patients but not for others, and only healthcare professionals who have detailed knowledge of a particular patient’s case can decide how much fluctuation should be tolerated. This individualisation of parameters for the home setting is normally very difficult in a multicentre study, particularly if randomisation is done after the patient has been discharged from hospital following an acute event. It is well known that patients recently admitted for acute HF with de novo disease (first manifestation of HF) need to be integrated into an HF programme soon after discharge so that response to therapy, aetiology and the presence of co-morbidities can be monitored. In these patients, TLM may be helpful in titrating therapy and checking for haemodynamic stabilities, but it is unlikely to have an effect on events such as HF rehospitalisation or death. However, patients followed up in an HF clinic who are deemed to be at high risk of deterioration (mainly with NYHA functional class III–IV) should be introduced to a TLM programme, particularly if there are geographical and socioeconomic barriers to their care. Unfortunately, no validated score has been produced in recent years to recognise HF patients at high risk of future hospitalisations. The only published experience is from the Outcomes of a prospective trial of intravenous milrinone for exacerbations of chronic heart failure (OPTIME CHF) study,14 which demonstrated a high risk of death or rehospitalisation in patients who had a history of HF hospitalisation, reduced systolic blood pressure, high blood urea nitrogen, reduced haemoglobin and a history of coronary intervention. However, this multivariate model has not been reproduced or translated into a specific score of possible clinical value.

A New Telemonitoring Model

Although a lack of benefit from TLM has been observed in the aforementioned multicentre studies, if all the factors discussed here are considered, one could argue that TLM still has the potential to be a useful tool for HF management in the future. Taking into account the experiences described in this article, a new TLM model needs to be developed (see Figure 2). In this new model, the central role is played by an HF clinic that invests in dedicated and trained personnel who can gain experience in the remote management of patients. After the clinic has set up a specific HF programme, patients with HF who are deemed to be at risk of a cardiac event during follow-up can be introduced to TLM, specific multiple indicators can be chosen and personalised algorithms can be developed. Ideally, TLM should not be patient-driven but, as is now possible with new devices, parameters should automatically be transmitted to a web database accessible by the HF clinic personnel, the GP and even the patient. Data should be frequently reviewed and regular feedback given to the patient.

It has been demonstrated that TLM improves communication whatever the level of technology used. The systems are practicable and compliance appears excellent. Multiple physiological indicators and personalised algorithms are probably required and, because patient-initiated communication is not the appropriate strategy, automatic transmission of disease indicators – mainly from an implanted cardiac device – should be put in place. Better definitions are needed to identify patients with HF who are at high risk of destabilisation by producing, from large clinical HF trials, validated scores that may be dynamic and periodically readjusted. Finally, it is mandatory not only to invest in technology but also in personnel and organisation, because it is effective management, and not TLM per se, that will determine the effects on morbidity and mortality.

References
  1. Fonarow GC, Yancy CW, Heywood JT, et al., for the ADHERE Scientific Advisory Committee, Study Group, and Investigators, Adherence to heart failure quality of care indicators in US hospitals, Arch Intern Med, 2005;165:1469–77.
    Crossref | PubMed
  2. Tavazzi L, Maggioni AP, Lucci D, et al., on behalf of the Italian survey on Acute Heart Failure Investigators, Nationwide survey on acute heart failure in cardiology ward services in Italy, Eur Heart J, 2006;27:1207–15.
    Crossref | PubMed
  3. Maggioni AP, Dahlstrom U, Filippatos G, et al., on behalf of the Heart Failure Association of ESC (HFA), EURObservational Research Programme: The Heart Failure Pilot Survey (ESC-HF Pilot), Eur J Heart Failure, 2010;12:1076–84.
    Crossref | PubMed
  4. Inglis SC, Clark RA, McAlister FA, et al., Which components of heart failure programmes are effective? A systematic review and meta-analysis of the outcomes of structured telephone support or telemonitoring as the primary component of chronic heart failure management in 8323 patients: Abridged Cochrane Review, Eur J Heart Fail, 2011;13:1028–40.
    Crossref | PubMed
  5. Klersy C, De Silvestri A, Gabutti G, et al., Economic impact of remote patient monitoring: an integrated economic model derived from a meta-analysis of randomized controlled trials in heart failure, Eur J Heart Fail, 2011;13:450–9.
    Crossref | PubMed
  6. Chaudhry SI, Mattera JA, Curtis JP, et al., Telemonitoring in patients with heart failure, N Engl J Med, 2010;363:2301–9.
    Crossref | PubMed
  7. Koehler F, Winkler S, Schieber M, et al., Telemedical Interventional Monitoring in Heart Failure Investigators, 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.
    Crossref | PubMed
  8. Cleland JG, Louis AA, Rigby AS, et al., TEN-HMS Investigators, Noninvasive home telemonitoring for patients with heart failure at high risk of recurrent admission and death: the Trans-European Network-Home-Care Management System (TEN-HMS) study, J Am Coll Cardiol, 2005;45:1654–64.
    Crossref | PubMed
  9. Mortara A, Pinna GD, Johnson P, et al., HHH Investigators, Home telemonitoring in heart failure patients: the HHH study (Home or Hospital in Heart Failure), Eur J Heart Fail, 2009;11:312–8.
    Crossref | PubMed
  10. Zhang J, Goode KM, Cuddihy PE, Cleland JG, TEN-HMS Investigators, Predicting hospitalization due to worsening heart failure using daily weight measurement: analysis of the Trans-European Network-Home-Care Management System (TEN-HMS) study, Eur J Heart Fail, 2009;11:420–7.
    Crossref | PubMed
  11. Whellan DJ, Ousdigian KT, Al-Khatib SM, et al., PARTNERS Study Investigators, Combined heart failure device diagnostics identify patients at higher risk of subsequent heart failure hospitalizations: results from PARTNERS HF (Program to Access and Review Trending Information and Evaluate Correlation to Symptoms in Patients With Heart Failure) study, J Am Coll Cardiol, 2010;55:1803–10.
    Crossref | PubMed
  12. Samara MA, Tang WH, Device monitoring strategies in acute heart failure syndromes, Heart Fail Rev, 2011;6:491–502.
    Crossref | PubMed
  13. Ferrante D, Varini S, Macchia A, et al., GESICA Investigators, Long-term results after a telephone intervention in chronic heart failure: DIAL (Randomized Trial of Phone Intervention in Chronic Heart Failure) follow-up, J Am Coll Cardiol, 2010;56:372–8.
    Crossref | PubMed
  14. Felker GM, Leimberger JD, Califf RM, et al., Risk stratification after hospitalization for decompensated heart failure, J Card Fail, 2004;10:460–6.
    Crossref | PubMed