Background: Heart failure (HF) is in many cases the final way for coronary heart diseases. In older than 65 years is the most important cause of hospitalization. Many studies have analyzed the causes of decompensation and the possible causes of hospital re-admission but most evidences come from the hospital setting. Nevertheless, It is known that many of these patients are usually attended in primary care setting, being treated by their General Practitioner and are not referred to the hospital.

The determinants which condition whether stable HF patients will be hospitalized as a consequence of a decompensation remain still unclear.

Objective: To develop and validate a predictive model based on clinical variables easy to measure at primary care to predict short-term hospitalization or mortality in a cohort of patients attended in primary care setting as a consequence of a HF decompensation.

Methods: A cohort of patients with established HF attended in primary care setting (regular care or out of hours) as a consequence of an episode of HF decompensation will be follow up. To develop the predictive model we will select variables which previously have been found involved in the risk of hospitalization or mortality as a consequence of HF as well as other ones which could be relevant and have not been studied yet. Variables selected should be easily determined in primary care setting..

The predictive model obtained in the derivation cohort from Spain will be validated in an external (several European countries) validation cohort.

Relevance: Our study will help GP to identify HF patients at higher risk of hospitalization or death and manage them according to an accurate prognostic rule (escala pronostica) made in primary care setting.[DISCUTIR]

DESIGN:

Multicenter prospective cohort study in which patients diagnosed of Heart Failure and suffering from a clinical decompensation will be followed for a 30 days in order to predict short-term hospitalització or mortality.

SETTING:

1. Derivation cohort will consist of HF patients recruited in more than 20 general practices settled in the city of Barcelona.
2. Internal validation cohort: population recruited will come from the same general practices than derivation cohort.
3. External Validation population will come from the rest of European countries participants.

Since the health system could be different in the participant countries we will consider as a primary care all consultations attended by General Practitioners out of the hospital setting (which includes out of hours clinics managed by GP).

Previous versions of this protocol have been presented to two European networks: Euroohnet (European research network in out of hours) and EGPRN (European General Practice Research Network).

In order to be able to generalize the results to the European territory, all European countries would be invited to participate. Up to data, Barcelona (Spain), France, Slovenia, Ireland, Italia, Turkey and Portugal have accepted to participate.

RECRUITMENT

Participants will be consecutively recruited from September 2014 to September 2016, amongst those previously diagnosed from HF who need to be attended in primary care from a clinical decompensation Study protocol will be previously presented to all potential GP participants and a hard copy will be distributed to all of them as well as informed consents to be signed by the patients who accept to be included. In this protocol all the instructions regarding recruitment, inclusion criteria, measurements and follow up will be written. GP participants must agree with the study protocol, and to consent to participate and to provide necessary data from their patients.

Since it is an observational study no randomization will be needed. Recruitment will end when the sample size is raised.

STUDY POPULATION

Heart Failure patients included in the study will come from primary care setting. We consider primary care, the health assistance provided in outpatients by a General practitioner, or out of hours care in primary care setting.

Inclusion criteria:

Patients aged 45 or more with previous diagnoses of HF in its medical record and a decompensation of HF treated/attended in primary care settings at the moment of the inclusion.

Decompensation is defined as the presence of at least one of the following clinical conditions (12):

• The worsening functional class according to the criteria of the New York Heart Association (NYHA).
• A weight gain of more than one kilogram in a period of 24 hours or greater than two kilograms in a maximum of 72 hours.
• Increased ankle edema (self-reported by the patient, caregiver or physician)

Exclusion criteria:

• Patients with severe psychiatric illness or severe cognitive impairment, not able complete the study protocol.
• Refusal to participate in the study.
• Patients directly attended in the hospital in the previous 30 days.

Every patient will be followed up to 30 days after the decompensation episode. We are intended to consecutively including patients in the study during a period of two years, from December 2014 to December 2016.

DATA COLLECTION: Procedures, measurements and Instruments

Sources of information

Variables will be obtained by:

• Interview to the patients
• Clinical exploration
• Primary medical record
• Hospital discharge reports. We will create a specifically designed questionnaire for the study where will registry information about personal medical history, clinical examination at the moment of the visits. These data will be supplemented with information from primary care electronical medical records.

Once the patient has signed the informed consent we will proceed to obtain the variables and measurements for the study.

In the final database only an identification number should be available for each patient. This ID should be easily associated with the data of patient, but only by their GP. Identification of patients should not be allowed to the rest of researchers. Every GP participant will have a list with their patients included in the study.

The ID will be as follows:

Country: SP (Spain), FR (France), PO(Portugal), TU (Turkey)…….. Surname of the GP responsible for the patient (initial) Date of birth of the patient (day/month/year) (--/--/----) ID number labeled by the main researcher of each country.

The coordinator centre (.Institut Català de la Salut. IDIAP-Jordi Gol, Spain) will design an electronic on-line questionnaire (https://es.surveymonkey.com/) which will be hosted in the website of IDIAP-Jordi Gol: (http://www.idiapjgol.org/) and e-mailed to all of GP participants in the study.

Potential predictive variables:

• Age (years)
• Gender
• Personal medical history (information obtained from medical record or interview) as dichotomous variables:

Diabetes mellitus ( ICD 10: E10, E11, E12, E13, E14) (YES/NOT) History of cardiovascular diseases (CHD or stroke) (ICD 10: I20-125; I63-I66) (YES/NOT) Chronic Kidney disease (ICD 10 N18) (YES/NOT) Current smoking (ICD 10: F17.1) (YES/NOT) Chronic Obstructive Pulmonary Disease (ICD 10: J43, J44) (YES/NOT) Previous hospital admission as a consequence of HF in the last year (YES/NOT)

- Clinical exploration at the baseline visit (decompensation episode) NYHA (III-IV vs I-II) (YES/NOT) Crackles (YES/NOT) Heart rate: beats/min Respiratory frequency Body Mass Index (to register weight and height) : km/m2 Temperature (ºC) Bilateral ankle edema with fovea (YES/NOT) ECG : Normal, Atrial fibrillation, Branch block, Acute Ischemia, Other arrhythmias Systolic Blood pressure : mmHg Diastolic Blood pressure: mmHg Treated with furosemide (a)/torasemide (b) before the visit : mg/day Not 1-40 mg (a) or 5mg (b) 40-80mg(a) or 5-10mg (b) >= 80 mg(a) or >10mg(b)

Other possible causes of decompensation:

• Dietary sodium excess: supplementation with salt to the aliments : (YES/NOT)
• Compliance with drugs : (YES/NOT)
• Excessive liquid intake: taking >2, 5 l/day
• Suspected Respiratory infection: (YES/NOT)
• Suspected other infections (YES/NOT)
• Known anemia (YES/NOT)

Others (By interviewing):

• Time since the first symptoms: hours

• Paroxistic dispnea: (YES/NOT)

• Weight gaining in the last week (self perceived) : (YES/NOT) 10. Data of diagnostic of HF (ICD 10: I.50) in medical record. 11. Setting of assistance:

  1. Regular care (usual GP)
  2. Out of hours care
  3. At home

Inclusion Criteria: Patients aged 45 or more with previous diagnoses of HF in its medical record and a decompensation of HF treated/attended in primary care settings at the moment of the inclusion.

Decompensation is defined as the presence of at least one of the following clinical conditions (12):

• The worsening functional class according to the criteria of the New York Heart Association (NYHA).
• A weight gain of more than one kilogram in a period of 24 hours or greater than two kilograms in a maximum of 72 hours.
• Increased ankle edema (self-reported by the patient, caregiver or physician)

Exclusion Criteria:

• Patients with severe psychiatric illness or severe cognitive impairment, not able to complete the study protocol.
• Refusal to participate in the study.
• Patients directly attended in the hospital in the previous 30 days.

Every patient will be followed up to 30 days after the decompensation episode. We are intended to consecutively including patients in the study during a period of two years, from December 2014 to December 2016.

Heart Failure patients included in the study will come from primary care setting. We consider primary care, the health assistance provided in outpatients by a General practitioner, or out of hours care in primary care setting.