Don Thiwanka Dilshan Wijeratne, Queen’s University

Supervisor: Stephen Archer, Queen’s University

I’ m currently initiating a population level retrospective cohort study describing the epidemiology of pulmonary hypertension (PH) in Ontario using health administrative databases. I will be focusing on different sub-groups of PH.

Databases – This is a retrospective cohort study using administrative healthcare databases in Ontario, Canada available through the Institute of Clinical Evaluative Sciences ICES. We will identify patients with a diagnosis of PH through databases which include records for admission to hospital for acute care obtained from the Canadian Institute for Health Information Discharge Abstract Database (CIHI DAD) and the National Ambulatory  Care  Reporting  System  (CIHI  NACRS).  These  use  nomenclature  from  the  ICD-9/ ICD-10 (international classification of diseases, 9th/10th revision) to provide detailed diagnostic records. Computerized pharmacy records of the Ontario Drug Benefit Database which records prescription drugs dispensed to all Ontario residents aged 65 years or older will be used to capture medication dispensing.

Cohort definitions – The cohort will consists of Ontario residents with a diagnosis of pulmonary hypertension between January 1st 1993 and December 31st 2012. Cases will be captured both from ER visits (CIHI NACRS) and hospitalizations (CIHI DAD). The following specific aims were addressed in the study:

1. Prevalence of PH (Table 1) – The prevalence of PH in Ontario will be estimated for two dates: December 31, 2002 and December 31,
2. Subgroups of patients with PH [ie left heart disease (LHD), lung disease ILD), chronic thromboembolic pulmonary hypertension (CTEPH)] – Patients with PH will be classified into the four PH subgroups based on diagnoses in the 5 years before their index Prevalence of PH in patients with these co-morbidities will also be calculated.
3. PH Medications – The Ontario Drug Benefit Claims date (ODB) will be used to identify patient All prescriptions for calcium channel blockers, endothelin receptor antagonists, PDE5 inhibitors and prostanoids in 1993-2012 for patients who reached the age of 65 years will be identified using ODB records.
4. Incidence – A case will be considered incident if the patient had no diagnosis of PH in previous 10 Incidence rates will be calculated for 2003-2012. Crude Mortality Rates, Survival Rates, Standardized Mortality Rate (SMR) will be calculated for the incident cohort

This project resonates well with the CVN – Bayer Research Fellowship mandate improving the health of all Canadians through innovative and effective prevention and detection of vascular conditions. This study would facilitate a better understating of PH patients at a population level, establish a cohort of PH patients for future work and thereby impact health outcomes.