TABLE 1

Characteristics of studies included in the meta-analysis of association between baseline C-reactive protein (CRP) level and mortality in chronic obstructive pulmonary disease (COPD) patients

First author [ref.]	Year	Study design; recruitment period; location	Patients n	Type of patients^#	Main aim	Outcome	CRP^¶	Prognostic value⁺	Cut-off mg·L⁻¹
Kuller [35]	1996	Retrospective, multicentric; 1973–1976; UK/USA	737	Smokers; stable	To investigate the relationship between CRP, α₁-acid glycoprotein and albumin and subsequent risk of myocardial infarction and coronary heart disease death in a nested case–control study among the Multiple Risk Factor Intervention Trial participants	Late mortality	Alone	Multivariate; 2.8 (1.4–5.4); age, tobacco dose, blood pressure, triglycerides, cholesterol levels	Not reported (fourth quartile)
Man [36]	2006	Prospective, multicentric; 5 years; USA/Canada	4803	COPD; stable	To determine whether serum CRP is associated with increased risk of all-cause and disease-specific causes of mortality, increased risk of fatal and nonfatal cardiovascular events, and an accelerated decline in lung function in COPD patients	Late mortality	Alone	Multivariate; 1.79 (1.25–2.56); age, sex, race, smoking status, pack-years, BMI, FEV₁	7.06
Dahl [37]	2007	Prospective, multicentric (epidemiological study); 1991–1994; Denmark	1302	COPD; stable	To determine whether increased serum CRP in individuals with airway obstruction predicts future hospitalisation and death from COPD	Late mortality	Alone	Multivariate; 1.4 (1.1–1.8); age, sex, tobacco consumption, FEV₁, cardiovascular disease	3
de Torres [38]	2008	Prospective, multicentric; 2000–2004; Spain/USA	218	COPD; stable	To determine if CRP levels are associated with survival in patients with moderate to very severe COPD in comparison with other well-known prognostic parameters	Late mortality	Alone	Multivariate; 1.00 (0.82–1.22); age, sex, pack-years, cardiovascular disease, corticosteroids	Not reported
Mehrotra [39]	2010	Prospective, community-based observational cohort (Pittsburgh, PA and Memphis, TN); 1997–1998; USA	268	COPD; stable	To identify significant covariates in addition to spirometry that predict mortality in elderly subjects with obstructive airway disease	Late mortality	Alone	Multivariate; 1.12 (0.90–1.30); age, sex, race, smoking status, cardiovascular disease	Not reported
Liu [40]	2011	Prospective, single-institution; 2005–2006; China	114	COPD; stable	To investigate the predictive value of combined serum CRP and BODE index score for mortality in COPD patients	Late mortality	Alone	Multivariate; 5.15 (1.65–16.60); BODE index	3
Zhang [41]	2011	Prospective, single-institution; 2001–2003; the Netherlands	405	COPD; stable	To quantify the effect of cardiovascular determinants on mortality in patients with a diagnosis of COPD	Late mortality	Alone	Multivariate; 1.78 (1.15–2.82); age, FEV₁, angina pectoris	3
Høiseth [42]	2012	Prospective, single-institution; 2005–2006; Norway	99	COPD; acute	To test the hypothesis that N-terminal pro-brain natriuretic peptide independently predicts long-term mortality following acute exacerbations of COPD	Late mortality	Alone	Multivariate; 2.4 (1.7–3.2); age, sex, comorbidity, BMI, S_aO₂, troponin level	50
Deng [43]	2014	Prospective, single-institution; 2009–2012; China	116	COPD; stable	To evaluate whether circulating CRP levels are a biomarker of systemic inflammation and a significant predictor of future COPD outcome	Late mortality	Alone	Univariate; 2.71 (1.05–6.99)	3
Moberg [44]	2014	Prospective, single-institution; 2005–2011; Denmark	423	COPD; stable	To investigate if leukocytes, CRP and vitamin D are independent predictors of mortality and hospitalisation after adjusting for disease severity with an integrative index (i-BODE index)	Late mortality	Alone	Multivariate; 1.50 (1.07–2.10); age, sex, i-BODE index	10
Cano [45]	2014	Prospective, multicentric; recruitment period 1 year; France	637	COPD; stable	To investigate predictors of long-term survival, including respiratory, nutritional and inflammatory dimensions, in a prospective cohort of home-treated patients with chronic respiratory failure	Late mortality	Alone	Multivariate; 1.51 (1.13–2.02); age, P_aO₂, P_aCO₂, BMI, FEV₁/FVC, 6MWT distance, transthyretin	5
Ford [46]	2015	Prospective, multicentric; 1988–1994; USA	1144	COPD; stable	To examine the association between elevated inflammatory marker count (white blood cell count, CRP and fibrinogen) on all-cause mortality in a national sample of US adults with obstructive lung function	Late mortality	Alone/combined (white blood cell count, fibrinogen)	Univariate for CRP alone; 1.26 (0.95–1.66)/multivariate for CRP combined; 2.08 (1.29–3.37); age, sex, race, education, smoking status, comorbidity, physical activity, alcohol use, BMI, FEV₁/FVC, blood pressure, cholesterol level, urinary albumin/creatinine ratio, cancer history	3
Kleber [47]	2015	Prospective, single-institution; 1997–2001; Germany	777	Smokers; stable	To characterise the diagnostic value of two independent risk factors for cardiovascular events (high-sensitivity CRP and lipoprotein-associated phospholipase A2), which provide information on inflammation and plaque stability in active smokers and never-smokers of the Ludwigshafen Risk and Cardiovascular Health (LURIC) study	Late mortality	Combined (lipoprotein-associated phospholipase A2)	Multivariate; 1.94 (1.10–3.45); age, sex, comorbidity BMI, cholesterol level, triglycerides	3.6
Blumenthal [48]	2015	Prospective (randomised controlled trial), multicentric; 2009–2014; USA	326	COPD; stable	To examine the prognostic value of select biobehavioural factors in patients with COPD in a secondary analysis of participants from the INSPIRE-II trial	Late mortality	Alone	Multivariate; 2.25 (1.02–4.96); age, Charlson score, COPD duration, GOLD, corticosteroids, coping skills training	Not reported
Loprinzi [49]	2016	Retrospective, single-institution; 2003–2006; USA	385	COPD; stable	To examine the association between objectively measured physical activity and all-cause mortality among a national sample of COPD patients, with stratification by inflammatory status	Late mortality	Alone	Multivariate; 1.33 (1.02–1.72); age, sex, comorbidity, race, BMI, physical activity, poverty/income ratio, cotinine	Not reported
Saltürk [50]	2015	Retrospective, single-institution; 2013–2014; Turkey	647	COPD; acute	To assess whether eosinophilic COPD exacerbations have better outcomes than noneosinophilic COPD exacerbations in the intensive care unit	Early mortality	Alone	Multivariate; 1.78 (1.01–3.14); age, sex, BMI, NLR, eosinophilia, invasive mechanical ventilation, noninvasive mechanical ventilation, APACHE II score, septic shock, resistant pathogen	500
Murphy [51]	2010	Retrospective, single-institution; 2004–2007; UK	60	COPD; acute	To determine if routine clinical assessment could reliably predict in-hospital death in patients admitted with acute exacerbation of COPD	Early mortality	Alone	Multivariate; 1.22 (0.95–1.14); Charlson score, pH, urea	Not reported
Zhao [52]	2014	Prospective, single-institution; 2010–2011; China	159	COPD; acute	To investigate the COPD assessment test, serum copeptin, procalcitonin and CRP levels as potential predictive factors for recurrence of acute exacerbation and all-cause mortality in 6 months in COPD inpatients	Early mortality	Alone	Univariate; 0.90 (0.82–1.01)	Not reported
Stolz [53]	2008	Prospective, single-institution; 2003–2005; Germany	167	COPD; acute	To investigate whether plasma pro-endothelin-1 and/or pro-adrenomedullin on admission to the hospital for acute exacerbation predict survival in patients with COPD	Early Survival	Alone	Multivariate; 1.02 (0.92–1.13); age, Charlson score, P_aO₂, P_aCO₂, pro-adrenomedullin, endothelin-1, BMI, FEV₁, leukocyte counts, procalcitonin, pulmonary arterial hypertension	Not reported
Duman [54]	2015	Retrospective, single-Institution; 2014 (1 year); Turkey	1704	COPD; acute	To evaluate mortality and outcomes of eosinophilic and noneosinophilic COPD exacerbations, and identify new biomarkers that predict survival	Early mortality	Alone	Multivariate; 1.32 (1.01–1.71); cardiovascular disease, corticosteroids, length of stay, readmission, NLR	19
Tofan [55]	2012	Prospective, single-institution; 1999–2010; Iran	60	COPD; acute	To assess the clinical utility of serum high-sensitivity CRP at admission in predicting outcome in hospitalised patients with acute exacerbation COPD	Early mortality	Alone	Univariate; 4.04 (1.32–12.33)	100
Zhang [56]	2014	Prospective, single-institution; 2007–2012; China	378	COPD; acute	To assess the association of high-sensitivity CRP with in-hospital outcomes in patients with COPD undergoing percutaneous coronary intervention	Early mortality	Alone	Multivariate; 1.78 (1.15–2.82); FEV₁, left ventricular ejection fraction, three-vessel disease, β-blocker use	3
Andreassen [57]	2014	Retrospective, multicentric; 2005; Norway/Sweden	731	COPD; acute	To find the proportion of patients with pneumonia among admissions due to acute exacerbations COPD and whether pneumonia has an impact on the length of stay, usage of noninvasive ventilation or in-hospital mortality	Early mortality	Alone	Multivariate; 0.71 (0.28–1.82); age, sex, GOLD	40
Guertler [58]	2011	Prospective, multicentric; 2006–2008; Switzerland	877	COPD; acute	To investigate the long-term prognostic performance of the Pneumonia Severity Index score and the association of clinical parameters and different blood biomarkers with long-term mortality rate in a large cohort of patients with community-acquired pneumonia	Early mortality	Alone	Multivariate; 0.3 (0.2–0.5); age, sex, comorbidity, temperature, chills, pro-adrenomedullin	Not reported (fourth quartile)
Moghbeli [59]	2005	Prospective, multicentric; 1997–1999; USA	1862	Smokers; acute	To investigate the role of inflammation, as measured by high-sensitivity CRP levels, in cardiovascular risk in smokers who have acute coronary syndrome	Early mortality	Alone	Multivariate; 2.60 (1.42–4.79); age, comorbidity, previous treatment, troponin level	15
Haja Mydin [60]	2013	Prospective, single-institution; 2009–2010; 23 months; UK	65	COPD; acute	To identify factors associated with inpatient mortality from hypercapnic respiratory failure with respiratory acidosis due to COPD	Early mortality	Alone	Univariate; 3.43 (0.38–30.55)	3

HR: hazard ratio; BMI: body mass index; FEV₁: forced expiratory volume in 1 s; BODE: BMI, airflow obstruction, dyspnoea and exercise capacity; S_aO₂: arterial oxygen saturation; P_aO₂: arterial oxygen tension; P_aCO₂: arterial carbon dioxide tension; FVC: forced vital capacity; 6MWT: 6-min walk test; GOLD: Global Initiative for Chronic Obstructive Lung Disease; NLR: neutrophil/lymphocyte ratio. ^#: COPD or smokers; stable or acute; ^¶: alone or combined; ⁺: univariate or multivariate; HR (95% CI); adjusted covariables.