Queen’s Researchers develop algorithm to identify pulmonary hypertension from available Ontario healthcare data

KINGSTON, ON—Pulmonary hypertension is an often under-recognized chronic disease that involves a congestion of blood supply in the lungs and heart, and is associated with other life-threatening diseases like heart failure.

A team of researchers from the Department of Medicine at Queen’s University in Kingston, including Stephen Archer, Don Thiwanka Wijeratne, and Diane Lougheed along with other clinicians and epidemiologists have found a way to use available Ontario healthcare data to identify pulmonary hypertension in patients. They have developed an algorithm that can recognize people with this disease from this data, although patients will remain unidentified for privacy purposes. This work was published in Pulmonary Circulation in February.

The first of its kind in Canada, the study used secure deidentified data of patients in Kingston confirmed to have pulmonary hypertension and compared it to provincial healthcare data for patients with the same diagnosis. The researchers then developed a simple algorithm based on a pattern of emergency care visits or hospitalizations that captures a diagnosis of pulmonary hypertension. Patients in Ontario exhibiting this pattern of care can now be recognized in the provincial health database.

“To our knowledge, this is the first such study conducted in Canada that attempts to reliably identify such patients with this disease using healthcare data in Ontario. Previously, researchers in Canada and other parts of the world have developed similar algorithms to identify patients with other common diseases like diabetes, hypertension, asthma, and chronic obstructive pulmonary disease (COPD). As a result of these algorithms, there have been significant changes to practices on screening and treatment, and a better understanding of these diseases and what they do to patients,” says Dr. Wijeratne.

Pulmonary hypertension is associated with other common diseases like heart failure, COPD, sleep apnea, and blood clots within the lung. Over time, patients with such conditions develop pulmonary hypertension as a complication that goes undiagnosed until their symptoms get worse. A diagnosis of pulmonary hypertension brings a risk of death seven times higher than those in the general population of the same age.

Until now, only a small number of patients with pulmonary hypertension who attended specialized clinics could be studied, but now with the algorithm, a larger group of patients can be studied over time. This algorithm will provide valuable information on how patients with pulmonary hypertension get sick, what other illnesses they may have, why they get sick and what complications they develop, and how often and why they die. The effectiveness of treatments can then also be gauged over time. 

This research is one of the first steps in facilitating better care for pulmonary hypertension patients and will hopefully contribute to better quality of life and survival.

“This opens up significant opportunities for research, intervention and knowledge translation. Without such algorithms to identify pulmonary hypertension patients, they can only be studied at individual clinics that specialize in treating pulmonary hypertension patients. Now, we are able to track a large group of patients, look at their other comorbidities and risk factors, and the effects of treatment over time,” says Dr. Wijeratne.

Now that this algorithm exists in Ontario, it can be easily adapted for similar healthcare databases across Canada and the world.