Patients and Intervention

The HELLEN-II trial (NCT01213875) [16] included 252 adult patients (≥18 years old) with left
ventricular systolic dysfunction (LVEF ≤45%) who were admitted because of acute decompensated
heart failure to two mixed university hospitals in the city of Porto Alegre, Brazil, with acute
decompensated HF. The 246 patients who completed the trial were selected for cost-effectiveness
analysis.After hospital discharge, patients were randomly allocated to receive specialist nurse-led
HVs and telephone calls (123 patients) or usual care (return to the outpatient clinic) (129 patients)
over 6 months. Visiting nurses provided guidance on disease awareness, self-care practices, early
recognition of the signs and symptoms of decompensation and decision making in the event of
decompensation, annual immunizations, and the importance of adherence to a low-sodium diet,
fluid restriction (symptomatic patients only), and medications [16].

Data Collection and Outcomes Assessment

Baseline data were collected during hospitalization for the index event with the aid of Sphinx® data
collection and analysis software. The primary outcome was a composite end point of afirst visit to
the emergency room (lasting o24 h), a first hospital readmission (424 h) due to HF or all-cause death.
Outcomes assessment was carried out by trained nurses, blinded to group allocation, at 6-month
follow-up. The costs of the index admission and HV intervention were measured.

Decision Model

An analytical decision model, type decision tree (Fig. 1) was constructed for simulation of the trial
intervention based in actual data from clinical trial. A microsimulation model was developed,
including the possibility of emergency room visits and hospital admission at the time horizon for
each strategy. Regarding cost of the HV intervention, the model assumed that no patients would be
lost to follow-up during the study period—a conservative assumption in that it takes the full cost of
the intervention to all patients into account. In view of the 24-week time horizon, no discount rate
was employed. Analysis was conducted both from the perspective of the Brazilian Public Health care
System (PHS) and from the perspective of private plans.

Base-Case Scenario

We used data from the HELEN-II trial: 123 HF patients in the study intervention group and 129 HF
patients in the usual care group. Participants were predominantly male, mostly with elementary
education, had a mean income of US$596, and lived with their families. Mean age was 62 years old
(±13), New York Heart Association (NYHA) functional class was predominantly II–III, and mean left
ventricular ejection fraction was 29.6% (±8.9). The most prevalent HF etiologies were ischemic or
hypertensive (Table 1). The baseline characteristics of the two groups are not statistically different.

Effectiveness of Intervention

At 6-month follow-up, the risk of a first HF-related visit to the emergency room, hospital
readmission, or all-cause death was 27% lower among patients in the nurse-led intervention group
as compared with those receiving usual care. Overall, individual clinical events occurred less
frequently in patients who received the nurse-led intervention, although none individually reached
statistical significance (Table 2). There were no losses to followup and all analyses were by intention
to treat.

Incremental Cost

The total cost of the study intervention was R$467 per patient. Taxi fares (R$235) accounted for the
largest portion of this cost. The total cost of nurse honoraria was R$55 per visit (Table 3). Within the
PHS framework, costs were R$5,934 for baseline hospitalization and R$37 for each emergency room
visit, versus R$14,177 for hospitalization and R$1,325 per emergency room visit in the private health
care sector. The costs of laboratory tests and imaging and days spent in intensive care were the main
cost drivers in both scenarios. In the private health care sector, room fees accounted for a larger
portion of emergency room visit costs. The full cost of all four HVs was allocated to all patients in the
nurse-led intervention group.

Cost-Effectiveness

Cost-effectiveness was measured by means of the incremental cost-effectiveness ratio (ICER),

considering the cost of each hospital admission and emergency room visit avoided, and expressed as
difference in mean cost between the Intervention (nurse-led management) group and the Control
(usual care) group (CostI – CostC) divided by the difference in mean effect between these groups
(EffI – EffC):ICER ¼ (CostI – CostC)⁄(EffI – EffC) [17]. The strategy more expensive and less effective
than its comparator is described as dominated and as dominant or costsaving when it reports more
health benefits and lower expected costs, with regard to its comparator.

Effectiveness Data

Rates of hospitalization (length of stay 424 hours), emergency room visits (length of stay less than 24
hours) due to HF or allcause death were estimated, using relative risks as the effect measure at the
24-week time horizon from the HELEN-II trial (NCT01213875) [16].
Cost

Costs were estimated within the Brazilian PHS perspective, using a service provider perspective. For
the private health care system, actual hospitals bills changed for the insurance companies or fee for
services were used as proxies to opportunity costs. As prices paid by the PHS do not reflect actual
costs, analyses were also conducted from a private health care system perspective. Reimbursement
table codes from PHS were applied. All cost data were calculated and expressed in 2016 Brazilian
currency (Reais, R$). The costs of intervention included nurse honoraria, private transportation by
taxi to all patient homes for all four HVs (meter fare within a 20-km range of the study hospital or up
to R$70 for visits to patients residing outside city limits), and the four intervention telephone calls.
The total duration of nursing intervention throughout the study period was defined as 8 hours (2
hours per HV: 1 hour for transport and 1 hour for patient care). Hourly wages were calculated on the
basis of the average base salary for nurses in the country under Brazilian labor law.