Why Anemia is Associated with Increased Mortality in AECOPD, What Should We Do? [Letter]

Dear editor

We read with great interest and peruse the recently published article by Garcia-Pachon Eduardo and Isabel Padilla-Navas that investigates the impact of anemia on acute exacerbation of chronic obstructive pulmonary disease (AECOPD) patients’ long-term mortality.¹ The authors have ensured a minimum follow-up of 3 years of 125 patients, and conducted comprehensive statistical analysis including Kaplan–Meier analysis and Cox proportional hazard regression analysis. We thank the authors for their hard work and contributions to AECOPD clinical study. As is known to all, AECOPD management is really a big challenge for every pulmonologist.

However, the study results and discussion maybe could focus more on the main factor of the age of the included patients. We believe the findings of the authors in this paper, while the important factor of age should be also taken into consideration by our readers.

Firstly, as shown in Table 1, AECOPD patients with anemia were significantly older than patients without anemia (median, 78 vs 71 years). Although the comparable baseline characteristics in observational retrospective analysis is not necessary, the potential influence should to be noticed. Secondly, as shown in Table 2, deceased patients were also significantly older than surviving patients (median, 77 vs 69 years). Thirdly, as in Table 4, multivariate hazard ratios by Cox analysis were conducted to explore the difference variables, and anemia, age, dyspnea, exacerbations, and length of admission were found as independent factors of mortality. Previous studies reported that persons aged 70 years or older have high prevalence of anemia.² Meanwhile, the elderly have high risk of all-cause mortality, and especially old patients with anemia are associated with higher risk of mortality.³ Age, anemia, and mortality are reported to have strong correlation with each other.^2,3 Therefore, the calculated value of anemia (HR = 3.8) was probably affected by statistical correlation, because Cox analysis firstly requires independence of each variable and correlated variables might enhance the effect size of HR.

So it follows that age plays an unavoidable role in anemia-associated mortality of AECOPD. Further statistical analysis by using propensity score matching (PSM) may be useful to investigate this in the future.⁴ Furthermore, as pulmonologists what should we do for AECOPD patients with anemia in clinical practice, to treat or not to treat? By our experience, anemia needs to be regarded as a “result” of decompensated status of AECOPD and/or COPD. Different underlying causes such as infection, chronic kidney disease-related erythropoietin resistance, congestive heart failure, malnutrition, or iron deficiency should be confirmed at first. Then individualized therapy rather than transfusion was recommended to treat anemia if possible. And by our observation, partial COPD patients due to malnutrition or iron deficiency may substantially benefit from treatment, especially regarding aspects of the quality of life, as well as mortality.