Real-life effectiveness of sacubitril/valsartan in older Belgians with heart failure, reduced ejection fraction and most severe symptoms

We assessed the real-world effectiveness of sacubitril/valsartan in patients with chronic heart failure (HF) and reduced ejection fraction (HFrEF) with an emphasis on those with older age (≥ 75 years) or with New York Heart Association (NYHA) class IV, for whom greater uncertainty existed regarding clinical outcomes. We conducted a retrospective cohort study based on patient-level linkage of electronic healthcare datasets. Data from all adults with HFrEF in Belgium receiving a prescription for sacubitril/valsartan between 01-November-2016 and 31-December-2018 were collected, with a follow-up of > 6 years. The total study population comprised 5446 patients, older than the PARADIGM-HF trial participants, and with higher NYHA class (all P < 0.0001). NYHA class improved following sacubitril/valsartan initiation (P < 0.0001 baseline vs. reassessment). Most concomitant medications were reduced. Remarkably, the risk of hospitalization for a cardiovascular reason and for HF was reduced by > 26% in the overall cohort, and in subgroups of patients ≥ 75 years, with NYHA class III/IV (all P < 0.0001) or with NYHA class IV (P < 0.05), vs. baseline. All-cause mortality did not increase in real-world patients with NYHA class III/IV. The results support the long-term beneficial effects of sacubitril/valsartan in older patients and in those experiencing the most severe symptoms.

NYHA Classification II, III or IV.It is important to mention the following limitations: i) providing the NYHA status was not mandatory for subsequent reimbursement requests and was therefore underreported, ii).reimbursement requests have only been reported until December 31, 2018.
-The National Register of Natural Persons (NRNP) database originates from the Federal Public Service.
The data have been collected between November 1, 2016 and March 16, 2023 and include NISS, Gender, birth month and year, Address (city), Date and Place of Death.
-Intermutualistic Agency (IMA) database: IMA is a platform where data from the health funds is brought together and analyzed.The collection of data is done through the public pharmacies, pricing services of the pharmacists and insurance institutions.However, no data on patient identification, medical records and diagnosis are available.Data from IMA includes sacubitril/valsartan prescribed and dispensed to the All the electronic data sources available at the time of data extraction were collected.
b. Databases from the national federation of independent pharmacists (APB) APB data were also made available, independently from healthdata.be.A decentralized database managed by the pricing services (TariferingsDiensten, TD) and APB has been developed, to perform longitudinal studies, a project named BelPharData (80% of all Belgian pharmacies divided into 14 geographically clustered TDs).APB also provides data related to the delivery of medication, as for IPhEB.The TDs collect the data from its members and then sends it to the various insurance institutions to which the patients are affiliated.The insurance institutions in turn send the data to the NIHDI-RIZIV-INAMI.Part of the information from TD is also sent to the BelpharData database.

Analyses of the healthdata.be datasets
a. Overall number of patients, determination of age, date of treatment initiation and censoring date in the Belgian Registry -Healthdata.begranted access to a first set of data in 2021, enabling to perform a first analysis.This first dataset (healthdata.bedataset 1) was used to determine the number of requests for sacubitril/valsartan and the evolution of NYHA functional classification during treatment.In 2023, healthdata.beprovided access to a second set of data (healthdata.bedataset 2), corresponding to the extended data of the patients already present in the sacubitril/valsartan Registry.The analysis of healthdata.bedataset 2 was performed to determine the characteristics of patients with chronic HFrEF in real-world, as well as hospitalization rates and mortality rates.
Healthdata.be dataset 1 included a total of 5606 patients.From these 5606 patients the following were excluded i) the patients without records in the IMA database (i.e., for whom there is no evidence that sacubitril/valsartan was dispensed), ii) patients without data besides IMA records (i.e., no information regarding any follow-up), iii) patients who started the study medication on December 31, 2018 (i.e, cannot contribute any information), iv) patients whose Death Date in the NRNP is prior to sacubitril/valsartan initiation.The total number of patients included in the analyses was therefore N = 5317.An identical method was applied for the analysis of the healthdata.bedataset 2 data, therefore including N = 5408 out of the 5779 total patients.For both analyses of healthdata.bedataset 1 and 2, the date of birth was based on NRNP for most patients, whereby the day of birth was assumed to be on the 15 th of the recorded birth month and year.When not available (N=132 patients in the 1 st analysis and N=117 in the 2 nd analysis), the birth year of patient was retrieved from the hospitalization records.In these cases, the date of birth was imputed as July 01 of that year.The first day of treatment or Day 0 was defined as the date of the first medication delivery.The age at start of treatment was calculated using the imputed dates of birth and start dates.
For analyses of healthdata.bedataset 1, the censoring date was December 31, 2018.For healthdata.be dataset 2, the censoring date for the analysis of hospitalization rates was December 31, 2019, and the censoring date for the analysis of mortality rates was March 16, 2023.The total follow-up time for the analysis of mortality rates was calculated as the number of days between Day 0 and the date of death (as retrieved from the NRNP database).Otherwise, the total follow-up time was the number of days between Day 0 and March 16, 2023.

b. Determination of Hospitalizations, Cardiovascular (CV) and HF Hospitalizations
Overlapping Hospital Records (i.e., multiple discharge date with the same hospital admission date) were collapsed.In addition, consecutive hospital records with start date one day after the previous stop date were also collapsed unless the record was an admission to the emergency ward.After collapsing hospital records as describe above, a hospitalization was deemed to be due to CV or HF if the Diagnosis at Admission (ICD-10 code) of at least one of the collapsed records were related to CV or HF, respectively.Hospitalization records with special codes (e.g., UUUUUU for emergency ward) or missing codes were deemed to be non-CV, non-HF.Hospital diagnoses were available for >85% of patients starting from 30-720 days prior to treatment and during treatment for all patients, the last day for hospitalization records being December 31,

2019.
The period at risk was calculated as the total follow-up time (from day 0, first day of treatment, to either December 31, 2019 or death, whichever comes first) from which the total number of in-hospital days was subtracted.Patients who first received sacubitril/valsartan in the hospital and died during the same hospitalization were not considered at risk for hospitalization and therefore not included in the analysis.

Table S4. Number of reimbursement requests in the Belgian registry
The absolute number of patients and percentage are provided.Abbreviation: New York Heart Association (NYHA).
In parallel to this information, between November 2016-December 2018, a total of 65.6% of patients exhibited high compliance, indicating that they had no treatment gap based on the supply of their last prescription, while 17.8% of patients stopped treatment (i.e., theoretical treatment gap of 3 months or more) as determined by the collected prescription information (by APB).The information in the pharmacy databases should be interpreted as a proxy for delivery, not for administration by a clinician or use by the patient.

Requests
individual patient with HFrEF, Prescription Date, Dose and Package Size.-Minimal Hospital Data (MHD) database: For financial reimbursement purposes, Belgian hospitals need to comply with compulsory registration of the Databases MDH.These data include the International Classification of Diseases, Clinical Modification, version 10 (ICD-10-CM) coding of admission diagnoses and the ICD-10-PC for procedures.MDH are managed by the Federal Public Service (FPS) for Health, Food Chain Safety and Environment.MDH include the following: Verified Admission Diagnosis, Type of Discharge, Indicator Age, Code Re-admission, Gender, Code of Cause of Death, Code of Verified Admission Diagnosis, Zip code, Type of Hospitalization, Month/Year of Hospitalization, Month/Year of Discharge, Number of Fully/Non-/Partially Invoiced Days in the Previous and Current Registration Year, Year of Birth, Cause of Death, Code Diagnosis, Code Procedure and Date, Type of Test, Results and Date, Reason for Emergency Admission and Treatment, Invoice Details, Code Main and Secondary Diagnosis.Some hospitals might experience a delay in the registration of MHD.-The IFSTAT Database is collected from the Institute for Pharmaco-Epidemiology in Belgium (IPhEB) via pharmacies (i.e., pricing services).This database contains data used by the invoicing offices to obtain reimbursement of the pharmaceutical supplies for the affiliated pharmacies and aims at showing that the beneficiaries of the compulsory health insurance in Belgium were delivered.Data are aggregated by district without enabling the potential identification of the prescriber or the patients.Data include Reimbursement Category of Medicine, CNK Code Medicine, Delivery Date, Year and Month of Invoicing.Reimbursement can be requested up to 2 years after the procedure.
while prior diagnoses (right) only include those occurring before treatment initiation (between 01-November-2016 and treatment start).Abbreviations: Heart Failure and reduced ejection fraction (HFrEF), chronic obstructive pulmonary disease (COPD).

Figure S1 .
Figure S1.Kaplan-Meier curve for death from any cause in the Belgian registry and comparison with the PARADIGM-HF trial

Figure S2 .
Figure S2.Kaplan-Meier curve for death from any cause in the Belgian registry, over a follow up period of > 6 years

Table S5 . Changes in NYHA classification in the Belgian registry after treatment initiation
Abbreviation: New York Heart Association (NYHA).