Diagnosis of COPD before cardiac surgery

To the editor:

We recently performed a retrospective study of the utility of respiratory function tests prior to cardiac surgery at our hospital. The study aimed to identify whether an association exists between pre-operative respiratory function tests and post-operative outcomes. The study also aimed to determine how the definition of chronic obstructive pulmonary disease (COPD) by Global Initiative in Chronic Obstructive Lung Disease (GOLD) [1] or by lower limit of normal (LLN) [2] may alter peri-operative management.

Between January 2011 and January 2014, 1330 patients underwent cardiac surgery at Holy Spirit Northside Private Hospital (HSNPH), Brisbane, Australia, a private hospital facility operating as part of the St Vincent's Health Australia group. Of this number, 289 mainly symptomatic patients were referred for respiratory function tests which were performed at HSNPH within 3 days of admission. The patients then proceeded with the intended cardiac surgery.

A patient was classified as COPD (GOLD) if the ratio of forced expiratory volume (FEV₁) to forced vital capacity (FVC) was <0.70 [1]. A patient was classified as COPD (LLN) if the FEV₁/FVC ratio was <LLN. The LLN criteria identify the fifth percentile of the study population, defined as 1.645 relative standard deviations below the mean predicted value [2]. The Global Lung Function Initiative 2012 equations were used to calculate predicted values and z-scores for spirometry [3].

125 patients (43% of the cohort) were classified as having COPD by the GOLD criteria (hereafter referred to as GOLD) and only 45.6% of these (n=57; 19.7% of the cohort) were also classified as COPD by the LLN criteria (hereafter referred to as LLN). Table 1 shows the demographics for the GOLD and LLN groups compared to controls (non-COPD) group (n=164).

Of the 289 subjects in the study, eight (2.8%) died within 30 days of their surgery and six (2.1%) had pneumonia in the post-operative period. Seven of the eight deaths occurred in hospital and one death occurred after discharge to a rehabilitation facility.

Patients with COPD defined by either GOLD or LLN criteria had longer intubation time, longer intensive care unit (ICU) stay and longer length of hospital stay than patients in the control group. Interestingly, only patients with COPD defined by LLN criteria (n=57) had higher mortality and pneumonia than the control group.

Patients with COPD as defined by the LLN criteria (n=57) had a lower FEV₁ (p<0.001) and lung transfer factor for carbon monoxide (p<0.001) compared to patients with COPD defined by GOLD criteria (n=125).

Patients in the discordant group which comprised those with COPD defined by GOLD criteria but not by LLN criteria (COPD GOLD only) (n=68) did have a longer intubation time (p=0.003) and a longer length of stay in ICU (p=0.014) than patients in the control group (n=164). However, there was no difference in length of stay in hospital compared to the control group (p=0.084). When assessed by logistic regression, patients with COPD GOLD only (n=68) showed no difference in the risk of death and/or pneumonia when compared to those in the control group (n=164) (OR 2.45, 95% CI 0.34–17.79; p=0.37). In comparison, having COPD as defined by LLN criteria (n=57) was associated with a 9.53-fold (95% CI 1.86–48.68; p=0.007) increased risk of death and/or pneumonia following cardiac surgery compared to those in the control group.

Patients with COPD defined by either GOLD or LLN criteria were significantly older than those in the control group (mean 72.5, 71.9 and 67.5 years, respectively). Considering that FEV₁/FVC ratio declines as part of normal ageing processes [3–5], COPD defined by GOLD criteria alone has the potential to “over-diagnose” COPD in elderly patients [4–11]. Indeed, 125 patients (43%) in our study group could have been diagnosed as having COPD by GOLD criteria, compared with only 57 patients (20%) by LLN criteria (p<0.001). It is likely that the use of COPD (LLN) criteria would identify patients with genuine disease, whereas using COPD (GOLD) criteria would lead to the inclusion of patients who are healthy, elderly patients. A false diagnosis of COPD by GOLD criteria could lead to some elderly patients being denied potentially useful surgery for serious cardiac problems.

The use of LLN criteria rather than GOLD criteria to diagnose COPD prior to cardiac surgery is likely to reduce the number of elderly patients who would be labelled as having COPD. This could potentially avoid unnecessary drug treatment of COPD in these patients and thus ameliorate costs and reduce side-effects from β₂-agonists. In a population of patients with cardiac disease, it may be an advantage to avoid β₂-agonists, as they can cause tachy-arrhythmias and exacerbation of angina [12–14]. A future study examining both the utility and cost of drug treatment for COPD and the cost savings that could be made by using LLN criteria to diagnose COPD would be very interesting.

Based on our study, we recommend the use of LLN criteria rather than GOLD criteria to diagnose COPD prior to cardiac surgery. The use of LLN criteria may reduce the “over-diagnosis” of COPD in elderly patients and may reduce costs and side-effects from unnecessary treatment of “COPD” and better identify those at risk of poor outcomes following surgery.