Methods

We selected studies of participants who underwent PCI and reported any of the following adverse outcomes: mortality, MACE by any definition or combination of (adverse cardiovascular events and mortality), reinfarction, and bleeding in patients who were anemic and nonanemic. We also included studies that evaluated the risk of adverse outcomes for incremental increase or decrease in hemoglobin.

A search of MEDLINE and EMBASE was performed on OVID SP on April 8, 2014. The exact search strategy is presented in Supplementary Table 1 . There was no restriction on the search based on language and both abstracts and unpublished reports were included.

Three reviewers (CSK, AP, and AA) independently screened all titles and abstracts for studies potentially meeting the inclusion criteria. The full reports of these studies were retrieved, and data were extracted independently by 2 reviewers (CSK and DT). Data extracted included study design, participant characteristics, participant inclusion criteria, definition of anemia or incremental hemoglobin change, adverse outcomes, and follow-up. Additional data were collected on quality of studies that included ascertainment of anemia, ascertainment of outcomes, loss to follow-up, and the use of adjustments for confounding. Publication bias was assessed using funnel plots for analyses with >10 studies and no evidence of substantial heterogeneity.

We used RevMan, version 5.3 (Nordic Cochrane Centre, Copenhagen, Denmark) to perform random effects meta-analysis using inverse variance methods for pooling risk ratios. We assumed similarity between odds ratio and other relative measures such as relative risk, rate ratios, and hazard ratios because cardiovascular events were rare events. We chose to pool adjusted results when available and crude results when adjusted results were not available. For data sets with the multiple time points, we chose to pool the results of the longest time point (up to 4.5-year interval) in the individual primary analysis. The I ² statistic was used to assess statistical heterogeneity. We performed the primary analysis considering unadjusted and adjusted results for anemic and nonanemic patients for the outcomes mortality, MACE, reinfarction, and bleeding. Secondary analysis was performed considering both the incremental decrease in hemoglobin per unit and incremental decrease in hematocrit and the risk of adverse outcomes. For the main analysis of anemia and risk of mortality, we also performed sensitivity analysis considering only studies that adjusted for baseline hemoglobin, renal impairment, and the severity of anemia. Further analysis was performed considering the effect of elective or acute coronary syndrome (ACS) patients on outcomes.

Results

The process of study selection is shown in Figure 1 . A total of 44 studies ^S1-S44 ( Supplementary Material ) were included with a total of 230,795 participants. The number of participants ranged from 100 to 73,067, and the overall prevalence of anemia was 26,514 of 170,914 (15.5% [15.3 to 15.7]) and for individual studies, the prevalence ranged from 3% to 41%. The study design, study dates, country of origin, and indications for PCI are presented in Table 1 . There were 4 post hoc analyses of randomized controlled trials, 14 prospective cohort studies, 14 retrospective cohort studies, and 12 cohort studies of unclear design. There were 15 studies of patients with ST elevated myocardial infarction, 1 study of only patients with non–ST elevated myocardial infarction, and all patients included in the analysis underwent PCI.

Flow diagram of study selection.

Supplementary Table 2 lists the risk of bias table for the included studies. Almost all the studies specified that blood measurements were taken before PCI except one study with unclear timing of blood sampling. A variety of methods were used to ascertain adverse outcomes after PCI, and loss to follow-up was reported in 18 studies. Most studies reported at least 1 adjusted analysis except for 5 studies.

The study definitions of anemia, lengths of follow-up, and results are presented in Table 2 and Supplementary Table 3 . The most frequent definition of anemia was the World Health Organization criteria, and the mean follow-up of the studies ranged from in-hospital events up to 4.5 years. The risk of mortality and MACE (as defined by each study in Supplementary Table 4 ) with anemia compared with no anemia was risk ratio (RR) 2.39 (2.02 to 2.83; 32 studies, 134,042 participants; Figure 2 ) and RR 1.51 (1.34 to 1.71; 20 studies, 47,552 participants; Figure 3 ), respectively. There was a significant difference between the unadjusted and adjusted results for anemia and mortality (p = 0.02). The risk of reinfarction and bleeding with anemia compared with no anemia was RR 1.33 (1.07 to 1.65; 13 studies, 36,316 participants; Figure 4 ) and RR 1.97 (1.03 to 3.77; 11 studies, 34,388 participants; Figure 5 ), respectively. A summary of key results are shown in Figure 6 .

Table 2

Study results

Study ID	Definition of Anemia		Follow Up (mo)	Results
	Hb (mg/dL)	Hct (%)
Akgul 2013	<13(M)<12(F)	–	6	Odds of cardiovascular mortality with anemia vs no anemia: aOR 3.9 (1.52-10.2).
Ali 2004	<13(M)<12(F)	–	31	Odds of long term mortality with anemia: aOR: 1.8 (1.6-2.0). In-hospital major bleeding: aOR: 3.3 (2.3-4.6).
Ayhan 2011	<13(M)<12(F)	–	In-hospital	Odds of long term cardiovascular mortality with anemia vs no anemia: aOR 2.2 (1.2-4.0).
Bolinska 2011	<13(M)<12(F)	–	In-hospital	Odds of death or cardiovascular complication with incremental Hb: aOR 0.89 (0.754-1.051).
Catakoglu 2007	–	<39%(M)<36%(F)	12	Odds of nonfatal coronary events with anemia vs no anemia: aOR 2.507 (1.379-4.555).
Chi 2012	<13(M)<12(F)	–	17	Risk of long term mortality with anemia vs no anemia: RR 3.293 (1.431-7.578).
Cho 2011a	<13(M)<12(F)	–	6	See crude results in Supplementary Material .
Cho 2011b	<13(M)<12(F)	–	24	Risk of mortality with anemia vs no anemia: aHR 1.52 (0.90-2.56). MACE: aHR 1.78 (1.14-2.78).
Dada 2009	<13(M)<12(F)	–	9	Odds of in-hospital mortality with anemia vs no anemia: OR 2.87(1.58-6.23). 9 month mortality OR 1.46(1.07-1.99).
Dunbar 2012	<13(M)<12(F)	–	In-hospital	Odds of in-hospital mortality per 1 g/dl decrease in Hb: aOR 1.28 (1.01-1.63).
Feldman 2009	≤12	–	25	See crude results in Supplementary Material .
Greenberg 2010	–	<39%(M)<36%(F)	12	Odds of 1 month mortality with anemia vs no anemia: OR 3.5 (1.6-7.5).
Gurm 2004	–	<35%, 35-40%	36	Risk of 3 year mortality per % of continuous hematocrit: aHR 0.953 (0.930-0.977).
Hanna 2013	<10, 10-12	–	In-hospital	Odds of death per 1 g/dl decrease in Hb <15 g/dl: aOR 1.07 (1.02-1.11).
Hosseini 2014	<13(M)<12(F)	–	12	MACE with anemia mild aHR 1.249(0.652-2.39) moderate aHR 1.462(0.584-3.66) severe aHR 4.623(1.642-13.021).
Husemann 2007	<13(M)<12(F)	–	“Long term”	Risk of long term mortality per unit increase in Hb: HR 0.711 (0.616-0.819).
Jones 2010	<13(M)<12(F)	–	36	Odds of 3 year mortality with anemia vs no anemia: aOR 2.4 (1.1-3.7).
Kim 2012	<13(M)<12(F)	–	25	Risk with anemia: MACE aHR 1.479(1.025-2.134), death aHR 1.943(1.241-3.043), MI aHR 1.182(0.476-2.936).
Kitai 2013	<13(M)<12(F)	–	36	See results in Supplementary Material for mild/moderate-severe anemia.
Kruk 2010	–	<39%(M)<36%(F)	In-hospital	Anemia: death HR 2.03(1.19-3.46) HF HR 1.09(0.79-1.51) death/HF HR 1.34(0.98-1.82) bleed HR 1.67(1.12-2.45)
Kurek 2010	<13(M)<12(F)	–	19	Risk of 1 year outcomes with anemia: death aHR1.31 (1.14-1.48) MACE aHR1.14(1.03-1.25). See Supplementary Material .
Liu 2008	<13(M)<12(F)	–	Unclear	Risk of mortality with anemia: aRR 2.216 (1.019-4.428).
Liu 2009	–	–	Unclear	Risk of long term mortality and Hb: HR 0.952 (0.921-0.984).
Maluenda 2009	–	–	12	Risk with incremental hematocrit: death HR 0.92(0.91-0.93), MI HR 0.94(0.91-0.97), death/MI HR 0.92(0.91-0.93).
Manzano 2008	<13(M)<12(F)		19	Risk of major bleed with anemia: aHR 2.15 (1.08-4.30).
McKechnie 2004	<13(M)<12(F)		In-hospital	Anemia: death OR 2.29 (1.79-2.92) MACE OR 1.2 (1.05-1.34) ↓Hb: death (M) aOR 1.21(1.14-1.28), (F) aOR 1.05(0.97-1.14).
Nikolsky 2004a	–	<39%(M)<36%(F)	12	Risk of outcome with anemia: in-hospital mortality: aHR 3.26 (1.01-10.52), 1 year mortality aHR 2.38 (1.18-4.81).
Nikolsky 2004b	–	<39%(M)<36%(F)	12	Odds of 1 year mortality with anemia vs no anemia: aOR 1.88 (1.46-2.43).
Oduncu 2013	<13(M)<12(F)	–	48	Risk of MACE with anemia vs no anemia: aHR 3.12 (1.15-6.59).
Ozasa 2012	<13(M)<12(F)	–	Unclear	Risk of mortality with anemia vs no anemia aHR 1.96 (1.49-2.58) and cardiac mortality: aHR 1.36 (0.90-2.06).
Park 2012	<13(M)<12(F)	–	24	Odds of 2 year mortality with anemia vs no anemia: OR 3.78 (2.19-6.52).
Poludasu 2009	<13(M)<12(F)	–	38	See crude results in Supplementary Material .
Rathod 2014	<13(M)<12(F)	–	36	Risk of mortality with anemia vs no anemia: propensity score matched HR 1.14 (0.52-2.49).
Reinecke 2003	<13	–	23	Risk of death with incremental increase in Hb HR 0.68 (0.59-0.78) and anemia vs no anemia HR 4.09 (1.52-11.05).
Rodriguez 2013	<13(M)<12(F)		27	Risk of mortality with anemia vs no anemia: aHR 2.1 (1.1-4.1).
Schroder 2013	–	–	32	Risk of mortality with incremental increase in Hb: aHR 0.784 (0.62-0.991).
Sgura 2010	<13(M)<12(F)	–	30	See crude results in Supplementary Material .
Shishehbor 2009	–	<40%(M)<36%(F)	54	Risk of mortality with incremental increase in hematocrit aHR 0.93 (0.91-0.96).
Tsujita 2010	–	<39%(M)<36%(F)	12	Risk of with anemia vs no anemia: 1 year mortality aHR 1.98 (1.05-3.73), major bleeding: aHR 2.15 (1.43-3.24).
Uchida 2013	<13(M)<12(F)	–	Unclear	Risk of adverse events with anemia vs no anemia: aHR 2.58 (1.01-6.60).
Varma 2010	<13(M)<12(F)	–	30	See crude results in Supplementary Material .
Vis 2010	<13(M)<12(F)	–	12	Odds of 1 year mortality with incremental Hb: aOR 0.992 (0.992-1.423).
Voeltz 2007	<13(M)<12(F)	–	12	Risk of 1 year mortality with anemia vs no anemia: aHR 1.84 (1.22-2.78).
Vrsalovic 2012	<13(M)<12(F)	–	1	Odds of 30 day mortality with anemia vs no anemia: aOR 2.69 (1.24-5.86).

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

1. Angiographic Findings and Revascularization Success in Patients With Acute Myocardial Infarction and Previous Coronary Bypass Grafting
2. A Comparison of the Quality of Life of Patients With an Entirely Subcutaneous Implantable Defibrillator System Versus a Transvenous System (from the EFFORTLESS S-ICD Quality of Life Substudy)
3. Usefulness of the Hepatocyte Growth Factor as a Predictor of Mortality in Patients Hospitalized With Acute Heart Failure Regardless of Ejection Fraction
4. Frequency of Methamphetamine Use as a Major Contributor Toward the Severity of Cardiomyopathy in Adults ≤50 Years

Stay updated, free articles. Join our Telegram channel

Join