Making an Accurate Diagnosis of Chronic Hypersensitivity Pneumonitis by K. Johannson & C. Ryserson, published in Canadian Journal of Respirology (2014) DOI: 10.1136/thoraxjnl-2016-208286 

Lay Summary

Why We Did This Study:

Chronic hypersensitivity pneumonitis (HP) is a type of interstitial lung disease (ILD) caused by long-term exposure to something in the environment. It can be difficult to tell chronic HP apart from other types of ILD. Making the correct diagnosis is important because chronic HP has its own treatment approach and expected disease course. However, previous tools designed to help diagnose chronic HP have been complex or have not worked well enough for chronic disease.

Study Goals:

The goal of this study was to develop a practical way to diagnose chronic hypersensitivity pneumonitis (HP) using information that is commonly available during patient care. The researchers wanted the model to identify chronic HP with high confidence and potentially reduce the need for a surgical lung biopsy in some patients.

How The Study Was Done:

To build the diagnosis model, patients with chronic hypersensitivity pneumonitis (HP), and other interstitial lung disease (ILDs) were identified. All patients had a CT scan of their lungs, and all the patients with chronic HP had their diagnosis confirmed with a lung biopsy. Researchers collected patient information, including age, smoking history and possible environmental exposures such as birds, down feathers, mould and farming. An expert radiologist reviewed the CT scans for specific findings and recorded whether scans appeared to show chronic HP, without knowing the patients’ clinical information or diagnosis. Factors linked to a diagnosis of chronic HP were then considered for the models.

Researchers then developed two models using different combination of patient information and CT scan findings. One model used specific findings from the CT scan, while the other used the radiologist’s overall impression of whether the scan showed chronic HP. The models were developed in one group of patients and then tested in a separate group to see how well they identified chronic HP.

Findings:

A total of 66 patients with chronic HP, and 124 patients with other ILD were included. They were split into two groups, a larger group to develop the models and a smaller group to test how well the models did. In the end two models were developed to differentiate HP from other ILDs.

The first model used four main pieces of information: age, patient exposure to birds or down feathers, and two CT findings called “ground-glass opacity”, and “mosaic perfusion”. Using this model, an HP score of 63 or higher was very good to identify people who actually had chronic HP. But using this model, some people with scores lower than 63 actually had chronic HP and the model did not identify them.

The second model also used age and bird or down-feather exposure and then used radiologist’s overall impression of whether the CT scan showed chronic HP instead of a specific CT scan finding. A score of 57 or higher also was very good at identifying people who had chronic HP, but missed chronic HP in people who scored lower than 57 about half the time. Both models continued to perform well when validated in the separate group of patients.

Interpretation:

These models suggest that information already commonly available during a patient's assessment may help doctors diagnose chronic hypersensitivity pneumonitis with high confidence in selected patients. For patients who meet the higher score cut-offs, the authors suggest that a surgical lung biopsy may be avoidable.

However, the models did not include several other tests that may also help diagnose chronic HP because these tests were not routinely performed in this group. The study also excluded patients whose interstitial lung disease could not be classified, so it is not known how well the models would work for these patients. Environmental exposures may also differ between parts of the world. For this reason, the authors state that the models need to be tested in patients from different geographic areas.

Summary:

This study developed two models to help identify chronic hypersensitivity pneumonitis (HP) using readily available clinical and imaging information. In patients with a high model score, chronic HP could be diagnosed with greater confidence, potentially reducing the need for a surgical lung biopsy. A low model score, however, could not rule out the disease. The models need to be tested in other patient populations and geographic regions to be more widely applicable.

Downloadable PP Coming SOON