Abstract
During an eight-year period, 358 patients with chronic heart failure (CHF) were enrolled in a six-month home-based telemanagement (HBT) programme. The efficacy of the programme was evaluated in two four-year periods, based on changes in clinical, functional, Quality of Life (QoL) status and rate of hospital readmission. The New York Heart Association (NYHA) class and the number of patients with comorbidities increased significantly in the second period, while the number of patients with beta-blockers decreased significantly (P < 0.01). Following the HBT programme, non-cardiovascular hospital readmission rate and all-cause readmission rate increased by 11% (P < 0.03) and 13% (P < 0.05), respectively. On re-evaluation after six months (238 patients) there was a general improvement in clinical, functional and QoL status and a significant increase in the mean daily dosage of beta-blockers prescribed. Our experience confirms that HBT for patients with CHF is associated with favourable effects on hospital readmission for cardiovascular reasons and on QoL. However, a more comprehensive multidisciplinary approach will probably be required to obtain favourable effects on total morbidity.
Introduction
Chronic heart failure (CHF) is a major public health problem in Western industrialized countries. 1 Despite new and more effective pharmacological and non-pharmacological treatments, the prognosis for patients with CHF remains poor. 2 Hospital treatment of heart failure accounts for nearly 2% of the total health-care costs in European countries. 3
Multidisciplinary management has become the standard in CHF, based on evidence from meta-analyses of randomized trials. 4 Several trials of structured telephone support or remote telemonitoring have been published and the efficacy of monitoring as a component of multidisciplinary CHF management has been confirmed. 5
In Italy, home-based telemanagement (HBT) for patients with CHF was introduced by the IRCCS Salvatore Maugeri Foundation in the year 2000. In a multicentre randomized study we demonstrated that HBT – which included multidisciplinary management and remote telemonitoring – could improve access to health care, reduce hospital readmissions, identify haemodynamic instability and reduce cardiac-related costs. 6 Based on these preliminary results, the model was more extensively tested in the Lombardy Region between 2003 and 2006. In 2006, HBT was formally adopted as a regional health service with a dedicated fee. At the end of 2009, the regional database included 802 patients recruited from 34 hospitals.
We have evaluated the effects of HBT in patients with CHF during an eight-year period.
Methods
Patients enrolled in the HBT programme at the Cardiovascular Rehabilitation Department of Fondazione Salvatore Maugeri between January 2000 and December 2007 were identified. All patients had a confirmed diagnosis of CHF, New York Heart Association (NYHA) class II-IV, echocardiograph evidence of left ventricular systolic dysfunction [left ventricular ejection fraction (LVEF) <40%] and at least one hospitalization for acute heart failure (AHF) in the previous six months. All patients gave their written informed consent.
The exclusion criteria were: non-cardiac debilitating illness (active malignancy, severe renal insufficiency, clinically evident cognitive impairment), myocardial infarction or revascularization within the previous 30 days, planned coronary revascularization or valve surgery, heart transplant, residing outside the Lombardy Region.
The study was approved by the appropriate ethics committee. Before enrolment, all patients were given advice on self-measurement of weight and blood pressure (BP), schedule for blood examinations, dietary restrictions, including sodium and fluid, and signs and symptoms of heart failure (HF) exacerbation.
Home-based telemanagement
The HBT programme provided multidisciplinary care with monitoring data and medical/nursing intervention via the telephone. 6 The model provided care for 24 hours/day every day. Briefly, a nurse-tutor (NT) followed-up the enrolled patients for six months managing the weekly contacts with them, mainly through scheduled appointments (Monday to Friday, 08:30–16:00). Occasional appointments could also be made in case of symptoms, signs of possible decompensation or any doubt about therapy.
Patients received a portable 1-lead ECG monitor (Card-Guard 2206, Israel). Each 1-lead ECG trace was transmitted by a fixed or mobile telephone call to a receiving centre, where a nurse or a doctor was available for consultation. At the end of each telephone call, the NT scheduled a new appointment if there were stable conditions, or planned treatment changes with either cardiologists or general practitioners (GPs), or asked for further investigations or consultations with cardiologists, or contacted the patient's GP and/or cardiologist in case of ECG changes or of signs or symptoms of haemodynamic instability. All conversations were recorded. An electronic record with all the patient's clinical data (BP, biochemical data) and ECG tracings was filled in by the nurse.
Once a week, the cardiologist and the NT met to reassess the clinical course of the enrolled patients. The cardiologist supervised all calls received by the NT in the previous week. The cardiologist also supervised implementation of therapy proposed by the NT and adjusted treatment for specific drugs (i.e. diuretics, ACE-inhibitors and beta-blockers). Only a cardiologist or a GP could decide to refer patients to the emergency or cardiology department. Details of the technical organization of the call centre have been reported previously. 6,7
Before the enrolment and at end of the HBT programme, all patients underwent a medical check-up including ECG, six-min walking test (6-min WT) 8 and Quality of Life (QoL) measured by Minnesota Living Questionnaire (MLHFQ). 9
Data collection
The information retrospectively obtained from patient records included: age, gender, comorbidities, HF aetiology, onset of symptoms, NYHA class, weight, heart rate, BP, LVEF, biochemical data, six-min WT, QoL evaluation and medications taken.
The outcome measurements included: all-cause cardiovascular mortality, HF exacerbation without hospitalization, cardiovascular and non-cardiovascular re-hospitalizations and AHF re-hospitalizations. The data and the cause of readmission were obtained from the GP and confirmed by hospital records. Episodes of clinical instability were confirmed by the GP.
Statistical analysis
We compared patients enrolled in two four-year periods (from January 2000 to December 2003, and from January 2004 to December 2007). The differences were analysed by chi-squared test for discrete variables, by the Student's t-test for normally distributed continuous variables and by the Mann-Whitney test for non-normally distributed continuous variables. Repeated-measures ANOVA and ANCOVA models were constructed to analyse the effect of time, year of enrolment and time-by-enrolment interaction for clinical and functional evaluations performed before and after the six-month HBT programme, with adjustment for baseline values when appropriate. Post-hoc tests were used to compare means when a significant F-ratio was found in ANOVA models of main effects.
Results
Of 938 patients registered in both the regional database and the electronic database of our telemedicine service, 831 fulfilled the inclusion criteria. Of these, 358 were recruited over an eight-year period: 210 from January 2000 to December 2003 and 148 from January 2004 to December 2007. The baseline characteristics of the patients are shown in Table 1.
Baseline characteristics of the patients
ns, P ≥ 0.05
Home-based telemanagement
The duration of the telemanagement was similar in the two periods: 167 days (SD 35) vs. 165 days (SD 52). In the second four-year period, the total numbers of both scheduled and unscheduled calls increased significantly: from 1423 to 3329 calls (P < 0.01) and from 82 to 510 calls (P < 0.0001), respectively. As a consequence, the number of scheduled calls/patient doubled (P < 0.01), as well as the number of calls for symptoms (P < 0.0001). The number of ECGs transmitted and the treatment changes prescribed per patient also increased significantly (P < 0.0001), see Table 2.
Home-based telemanagement: telephone calls and actions taken
ns, P ≥ 0.05
The time spent by the NT for telephone support and telemonitoring activities increased from 2.1 h/patient (SD 0.8) to 5.1 h/patient (SD 1.9) between the two periods (P < 0.0001).
Clinical data
With ageing, NYHA functional class and presence of comorbidities increased significantly as well as the number of patients with more recent onset of HF symptoms. There was a significant reduction in the percentage of patients treated with beta-blockers during the two study periods (P < 0.01).
The clinical events that occurred during HBT are summarised in Table 3. Rates of HF decompensation without hospitalization, cardiovascular and AHF hospital readmissions as well as all-cause mortality were similar between the two four-year periods. Conversely, non-cardiovascular hospital readmission and all-cause readmission rates increased over the time: 8% (P < 0.03) and 13% (P < 0.01), respectively. AHF was the most frequent reason for all-cause hospital admission in the whole population (59%) and in both four-year periods considered (56% and 56%, respectively). Chronic obstructive pulmonary disease exacerbation, bone and joint diseases and kidney/urinary tract disorders accounted for a large proportion (>70%) of non-cardiovascular hospital readmission in both periods.
Clinical events. Values shown are percentages
ns, P ≥ 0.05
Clinical, functional and QoL re-evaluation at six months were available for 238 patients (66%). In the whole sample (Table 4), a general improvement of clinical, functional data and QoL was observed with a significant increase in the mean daily dosage of beta-blockers prescribed. In both four-year periods, similar results were observed (Table 5). However, after adjustment for baseline values, the changes observed at the end of HBT programme were significantly different between the two periods only for NYHA class, MLHFQ score and haemoglobin concentration.
Clinical and functional data before and after HBT
ns, P ≥ 0.05
Clinical and functional data before and after HBT in the two four-year periods
*P < 0.003 between groups
†P < 0.001 within groups
‡P < 0.01 interaction
Discussion
To our knowledge, the present study is the first which describes the long-term experience of a single site and the evolution of a multidisciplinary HBT programme in patients with CHF in terms of efficacy on quantitative end-points. Comparison with previous studies reporting the effects of a structured telemanagement programme on clinical and functional outcome in patients with CHF is quite difficult due to the diversity of the studies. The management strategy of previous studies was based on telephone support or remote telemonitoring and, generally, the programmes lasted for one year. Our HBT programme included both telemonitoring and 24-hour telephone support managed by a NT, and it lasted for six months.
The all-cause and HF readmission rates observed in the present study are similar to those reported in the treatment arms of randomized trials recently reviewed by Inglis et al. 5 : telephone support (38% all-cause, 16% HF), telemonitoring (47% all-cause, 22% HF). Our data are also similar to the results reported in the telemonitoring arm of a randomized trial recently published by Chaudhry et al. 10 Finally, in our study, AHF readmission was confirmed to be the main reason for all-cause re-hospitalization with an incidence similar to that reported for telemonitoring and telephone support (51% and 44% respectively) 5 and in the Tele-HF study (56%). 10
The rate of clinical events observed in the present study was similar to that of the treatment arm of our previous randomized trial. 6 However, the all-cause readmission rates were higher than those previously reported (38% vs. 29%). This may be due to differences between the two populations studied and the shorter duration of the HBT in the present study. 5 Moreover, in the present study, the differences of all-cause readmissions between the two four-year periods were mainly related to non-cardiovascular reasons whose incidence almost doubled during the study period.
Considering the differences between the population of the present study and that of our previous randomized trial, 6 the favourable effect of HBT on AHF hospital readmissions would probably be evident in a more heterogeneous CHF patient population at higher cardiovascular risk. The comparison between the two four-year periods of the present study confirms this hypothesis. The patients enrolled between 2004–2007 were older, in more advanced NYHA functional class, with more comorbidities and other indices of unfavourable prognosis, but the efficacy of HBT to prevent AHF readmissions was unchanged.
The epidemiological changes documented in the sample over time were consistent with changes in the clinical profile of patients referred to our department. Both the inclusion criteria and the nurse/medical team remained unchanged, as well as the technology and the organization of the call centre. Few studies have evaluated the effects of HBT on QoL and functional capacity. Of the 25 randomized studies recently reviewed by Inglis et al. 5 changes in QoL and six-min WT were reported only in 11 and one, respectively. The positive effect on QoL (mostly evaluated by MLHFQ) was evident in eight out of 11 studies and the increase in six-min WT reported by Ramachandran et al. was about 18%. 11 In our study, QoL improved significantly after HBT and this result seemed to be maintained over time. In the whole sample, the increase observed in six-min WT was small (4%) and probably related to the baseline value which was higher (427 m) than that reported by Ramachandran et al. (202 m). 11 This difference could explain why the six-min WT did not improve in our study. Therefore, the clinical, functional and therapeutic results obtained demonstrate the flexibility of our HBT programme. They also suggest the need to improve the multidisciplinary approach, which should include staff other than cardiologists (internists and geriatrics) who are probably better suited to managing older patients with several comorbidities.
It is noteworthy that epidemiological changes, documented in the population over time, produced an adaptation of the HBT programme to the patients' needs. This was clear from the increased time spent by the NT for scheduled and unscheduled calls, and the therapy modifications for the patients.
The main limitation of the present study is the generalization of the results. The applicability of our findings to patients managed similarly in different hospital settings needs to be confirmed.
In conclusion, the present study confirms that our HBT programme in patients with CHF shows favourable effects on hospital readmission for cardiovascular reasons and on QoL. However, a more comprehensive multidisciplinary approach will probably be required to obtain favourable effects on total morbidity.
Footnotes
Acknowledgements
We thank the NTs Mrs Giovanna Martinelli and Mrs Doriana Baratti for their support, Mrs Silvia Brognoli for data analysis, and Dr Laura Comini and Dr Alessandro Bettini for their assistance with the manuscript. The work was a part of a project (ICS 030.8/RF00.91) financed by the Italian Ministry of Health (grant 502/92) and by the Lombardy Region (grant 15882).