Protocol of a Pilot Study to Optimize Care for Heart Failure Patients in Primary Care

Objectives:

The objective of this pilot study is to assess the feasibility of a multifaceted intervention to optimize care for HF patients in Belgian general practice. It is not a primary objective to study the effectiveness of the intervention on the defined outcomes, since follow-up might be too short, however the evolution in quality will be evaluated. Specific research questions are:

• Do general practitioners use the NT-proBNP POC test and how does it influence practice?
• Do general practitioners use the assistance of the HF nurse and how does it influence practice?
• Does this intervention improve quality of care for HF patients in primary care?
• Does this intervention improve quality of life for HF patients in primary care?
• Is this intervention/data collection method feasible? How do participating GPs and investigators experience the interventions?

Data collection:

• Measurements at baseline and after 6 months of follow-up
• Baseline demographic data on patient age, gender, years since the diagnosis of HF, cardiovascular comorbidities (hypertension, ischemic heart disease, valvular heart disease, atrial fibrillation, cerebrovascular disease, peripheral arterial disease), non-cardiovascular comorbidities (diabetes, depression, chronic obstructive pulmonary disease, dementia), the last two blood pressure measurements, last 2 creatinine measurements and the last two values of the estimated glomerular filtration rate (eGFR), last date of GP consultation, involvement of cardiologist (yes/no), last date of cardiologist appointment, echocardiography (yes/no), last date of echocardiography report, ejection fraction (EF), diastolic dysfunction (yes/no), clinically relevant valvular dysfunction (yes/no), left ventricular hypertrophy (yes/no), left atrial enlargement (yes/no) and heart failure treatment with dosage (diuretics, ACE-I, ARB, B-blockers, MRA, digoxin) will be collected. Additionally, the number of chronic medications and the number of chronic comorbidities will be collected. Furthermore, the MLHF-Q will be collected to evaluate patients' quality of life and grip strength (only at baseline) as a predictor of functional decline, disability and mortality. At baseline, HF hospitalization during the last three years will be registered based on discharge letters in the EHR. All hospitalizations and mortality will be collected at practice level during the study course. Data on physician age, gender, years of experience and practice organization (group/single-handed/number of practitioners/EMHS) will be requested as well.

Patient level

Outcomes:

All data will be extracted from the EHR of the general practitioner for all patients who satisfy the eligibility criteria. Data will be extracted at baseline and after 6 months. First, at baseline and after 6 months, all patients with a registered diagnosis of HF will be identified in the EMHS. Demographic data can be extracted with an automated search; other data will need to be collected by manual chart review. Additionally, all identified HF patients will be visited by the HF nurse at baseline and after 6 months and will be asked to fill in the MLHF-Q. At this occasion grip strength will be measured in the dominant hand using a JAMAR® Plus digital hand-held dynamometer. Three attempts at maximal squeeze will be recorded.

Process evaluation:

GPs will be asked to fill in a study registration form about the indication and results of their NT-proBNP POC test use and how it influenced their decision-making. The HF nurse will keep a log of all actions. Consequence of these actions will be collected from the patients' medical record. GPs will be asked to take note of the HF nurse advice in the EHR and indicate if it influenced decision-making.

Experiences of GPs and investigators:

The feedback meeting at 6 months will be organized as a focus group in each GP practice to get more insight in the experiences of the participating physicians and investigators. The meetings will be audiotaped and typed out. Any harm as a result of the trial will be explicitly collected and reported.

Practice level

Impact of the intervention on hospitalizations and mortality cannot be studied in this pilot trial, however, in a later stage this would be the outcome in the larger cluster-randomized trial, therefore, the feasibility of collecting data on hospitalizations and mortality will be tested. These outcomes will be measured at practice level in all patients aged 40 years or older, enlisted in one of the participating GP practices. GPs will be asked to register date of death and cause of death in the EHR of all patients who die during the study course. A list of hospitalized patients of each participating GP will be collected in all the surrounding hospitals (ZOL Genk, AZ Vesalius, Jessa Hasselt) and handed over to the GPs. Cause of hospitalization will be deducted from the discharge letter in the EHR by the GP in a standardized manner (trained by principal investigator (PI)). The anonymized data will be handed over to the PI.

Data management:

Data collection and data entries will be done by the PI and an assisting data manager. All patient data will be coded and anonymized. All other authors will have full access to the coded data (including statistical reports and tables) in the study and will be able to take responsibility for the integrity of the data and the accuracy of the data analysis.

Data analysis:

Descriptive statistics will be provided regarding baseline variables of the HF patients and general practitioners/general practices. Outcomes measured before and after the study intervention will be compared with a chi-squared test. Experiences of participating GPs and investigators will be analyzed qualitatively using thematic analysis as method.

Ethics:

Before the study start the investigators will present the participating physicians with an informed consent that outlines the intervention and the purpose of the trial. Additionally, all identified HF patients will be visited by the HF nurse before the start of the study. On that occasion the HF nurse will inform all HF patients about the study outline and purpose and ask informed consent. Furthermore, patients eligible for a NT-proBNP POC test will be informed by their GPs about the study outline and purpose and asked informed consent. However, informed consent cannot be sought for all patients on practice level (hospitalizations/mortality). Therefore, information about the study intervention and purpose of the trial will be presented on a poster in each practice, visible for all patients, who can opt-out if they object to study participation.

Inclusion criteria

Heart failure patients are included if:

• They have their electronic medical records ("globaal medisch dossier") registered with one of the participating general practitioners.
• They have a registered diagnosis of HF (ICPC-2 code K77 or free text) in their electronic health record (EHR).

Patients can be included during the study if:

• They give their informed consent to participate in the study.
• Are eligible for a NT-proBNP POC test.

Exclusion criteria

• No exclusion criteria will be set