Design
The HF-GERIATRICS study is a trial in elderly patients who are admitted with a primary diagnosis of HF in acute care units of eight geriatric services in Spain. Patients are assigned randomly to a DMP (intervention group) or to receive usual care (control group). Randomisation is stratified by hospital and is conducted with concealment of the randomisation list. The two trial groups are followed for 12 months.
The project has been approved by the clinical research ethics committee of all participating hospitals (Comité de Etica y Ensayos Clínicos (CEIC) de Galicia, CEIC del Complejo Hospitalario de Caceres, CEIC del Hospital Universitario La Paz, CEIC del Complejo Universitario de Albacete, CEIC del Hospital Universitario de Guadalajara, CEIC del Hospital Central de Asturias, CEIC del Hospital Universitario de Getafe, CEIC del Hospital Universitario de San Carlos, Spain).
Study participants
The study will include patients aged 75 years or over who are admitted with a primary diagnosis of HF according to the European Society of Cardiology and Framingham criteria [
14,
27]. Each patient must also meet at least one of the following comorbidity criteria: Charlson index ≥ 3 (regardless of age), dependence in ≥ 2 basic activities of daily living, prescription of ≥ 5 drugs, emergency hospitalisation in the last 3 months, active treatment for ≥ 3 diseases, limitations in daily life due to visual or hearing loss, cognitive impairment, Parkinson's disease, diabetes mellitus, chronic obstructive pulmonary disease (COPD), anaemia or constitutional syndrome whose causes have not been determined previously.
Exclusion criteria are terminal or rapidly fatal disease and life expectancy of less than 6 weeks, severe functional or cognitive impairment preventing patients from understanding their illness and lack of a caregiver who agrees to participate in the study, patients waiting for cardiac surgery, those who cannot be followed up (e.g., due to change of address), and institutionalised persons for whom a primary caregiver cannot be identified.
Identification of participants, informed consent and randomisation
In each hospital a case manager with a degree in health sciences and special training in HF will identify and register cases admitted for HF each day by reviewing the diagnoses of those aged ≥ 75 years admitted in the acute care unit of the geriatric service. The case manager will then assess each patient and carry out the following activities:
i- Verify the study inclusion and exclusion criteria.
ii- Inform patients (verbally and in writing) and their caregivers (when considered necessary and with the patient's consent) of the characteristics of the DMP and confirm their agreement to participate by asking them to sign the informed consent document.
iii- Assign the patient to the intervention or control group. Randomisation will be conducted independently in each centre. Consecutively numbered envelopes, containing the group to which each patient should be assigned, will be used for this purpose. The envelope will be opened only at the time of patient assignment, to assure concealment of the randomisation list.
iv- Coordinate patient assessment, record the information in the case report form, and transcribe the information to a computerized database.
Description of the intervention: Disease Management Programme
The DMP, which will be carried out by the case manager, has three main components: education of the patient and the main caregiver to improve disease knowledge and self-management, monitoring of the patient's clinical condition, and monitoring of treatment compliance.
a) DMP components to be conducted by the case manager during the hospital stay
i- Provision of information to patients and their caregivers or family members. An individual session will be held to teach the characteristics, causes and prognosis of HF. Emphasis will be given to recognition of symptoms of decompensation, precipitating factors, and treatment (drug knowledge and treatment compliance).
ii- Provision of written and graphic educational materials with the foregoing information.
iii- Health education on warning signs and symptoms and information on drugs for the most common comorbidities: diabetes, COPD and depression.
iv- Coordination of hospital discharge with ward nursing staff
v- Development of a follow-up plan. A date for an outpatient appointment will be set, and a telephone contact number will be provided to patients so they can ask questions or report any incidents.
b) Components of the DMP to be conducted by the case manager after hospital discharge
The case manager will devote 2 days/week to the clinical assessment of outpatients. The first contact with the patient will be by telephone, in the second week after hospital discharge: during this call, the case manager will verify the patient's knowledge of HF, reinforce the information provided during hospitalisation, and strengthen compliance with recommended lifestyles and pharmacological treatment. Control of minor symptoms of decompensation by use of diuretics will be emphasized. The first visit will take place at the hospital outpatient clinic 4 weeks after discharge. Future visits will be agreed according to the patient's clinical and treatment needs, but some minimum controls will be established at months 1, 2, 3, 6, 8 and 12 after discharge (table ). The case manager will be provided with a cell phone to facilitate coordination with patients, caregivers and other health professionals. Messages left with the case manager will be answered within 24 hours.
| Table 1Design and data collection in the HF-Geriatrics study |
The DMP will be supervised by the geriatricians on the research team, who will progressively delegate clinical functions to the case manager. A training programme for case managers will be conducted before the beginning of the trial. The DMP includes protocols for HF diagnosis and treatment will be implemented, based on the clinical guidelines of the European Society of Cardiology [
14]. It also includes protocols to optimise the diagnosis and treatment of COPD, diabetes mellitus and depression. All cases of COPD are to be confirmed by spirometry, and patients and their caregivers will be trained on the use of inhaled drugs, smoking cessation, use of vaccines, and identification and management of exacerbations; in addition, treatment will be established in accordance with the Global Initiative for Chronic Obstructive Lung Disease (GOLD) recommendations. Patients with diabetes mellitus will receive an educational intervention on the symptoms of hypoglycaemia and hyperglycaemia, foot care, and, for patients treated with insulin, on the proper technique for its administration and self-monitoring of capillary blood glucose. In cases of depression, a clinical assessment of the diagnosis will be made if the score on the 15-item Yessavage test is > 5, and appropriate treatment will be initiated.
At outpatients clinic visits and telephone contacts, case managers will identify warning signs and symptoms of clinical deterioration from records of blood pressure and heart rate, changes in body weight, night-time cough, degree of dyspnea, decreased physical activity, and biomarkers such as renal function, natriuretic peptides and haemoglobin saturation by pulse oximetry.
For patients in the control group, the same intermediate and outcome variables will be recorded as in the intervention group, but the previously described educational activities and monitoring by the case manager will not be carried out. Follow-up will begin with the index admission and ends 12 months after discharge or in case of the patient's death.
Study variables
In addition to the inclusion and exclusion criteria, four types of variables will be collected: psychosocial, biomedical, treatment prescribed, and intermediate and final outcome variables (table ). Episodes of hospital emergency care will also be recorded. The intermediate outcome variables are knowledge of the nature and management of HF and of comorbidity, and compliance with recommendations on lifestyle and drug treatment. The final outcome variables are health-related quality of life (assessed using the Minnesota Living with Heart Failure Questionnaire [MLWHFQ]), hospital readmissions, and total mortality. The outcome variables are measured at 6 and 12 months after the DMP begins. The information is collected through interview with the patient or primary caregiver and review of the clinical record.
Study size
In previous studies we observed that about 30% of HF patients are readmitted every 6 months [
11]. These studies were conducted in hospitals and with patient's inclusion criteria different from those in the HF-Geriatrics, and within patients with less use of effective drug treatments than at present. Accordingly, under the conservative assumption that the annual frequency of readmission will be 40%, and that the frequency of readmission is similar in all study hospitals, 700 patients will need to be recruited (350 in each group) to detect, with an alpha error of 5% and a statistical power of 80%, a 20% reduction in hospital admission. This analysis assumes, as can be expected, that no patients will be lost to follow-up. A total of 700 patients also allows to detect, with an alpha error of 5% and 80% power, differences of 10 points (clinically relevant) in the MLWHFQ, assuming that patients in the control group have a mean score of 50 points on this questionnaire [
11].
Statistical analysis
We will first verify the effectiveness of randomisation. The calculated sample size should be sufficient to assure an even balance of the predictors of the outcome variables between the intervention and control groups.
Second, we will estimate the time to readmission or death for each patient, and will obtain curves for time free of such events (Kaplan-Meier method) in the intervention and control groups. These curves will be compared using univariate methods. The effectiveness of DMP will be summarized with hazard ratios (and their 95% confidence intervals) for readmission and mortality obtained by Cox regression, using the hospital as stratification variable. Finally, if the DMP is shown to be effective, we will estimate the number of patients needed to treat (NNT) to prevent one outcome event.
For health-related quality of life, the MLWHFQ score at baseline, 6 months and 12 months will be calculated for each group in the trial. The effect of the DMP on quality of life will be estimated as the difference between groups in the change in MLWHFQ scores during the study. These differences will be accompanied by their respective 95% confidence intervals.
The analyses of the effect of DMP will be performed according to the intention-to-treat principle.