Introduction

As the burden of cardiovascular disease increases, cardiovascular diseases have become the world's leading cause of death, and the number of coronary angiography continues to rise [1]. Contrast-associated acute kidney injury (CA-AKI) is one of the major complications of coronary angiography (CAG), and the occurrence of CA-AKI is related to the toxicity of contrast agent, which leads to kidney function loss, with apoptosis and tubular necrosis [2, 3]. The high osmolality contrast agent and ionic contrast agent were related to the high risk of CA-AKI [4, 5]. At the same time, the risk of CA-AKI increases with each additional 100 ml of contrast agent [6]. In addition, renal ischemic injury caused by vasoactive substances such as endothelin, nitric oxide and prostaglandins may increase the risk of CA-AKI [7,8,9]. The occurrence of CA-AKI is significantly associated with prolonged hospitalization and with an increase in short and long-term mortality [10, 11].

Numerous studies have suggested that the incidence of CA-AKI ranged from 3 to 50% [12,13,14]. Chalikias et al. mentioned that the current incidence of CA-AKI was difficult to calculate due to the influence of many factors (CA-AKI definition, prevalence of comorbidity, including diabetes mellitus, pre-existing renal disease, heart failure, anemia) and was also related to clinical setting [15]. Furthermore, it has been confirmed in previous studies that age > 75 years, hypotension, intra-aortic balloon pump and hypoproteinemia are risk factors for CA-AKI [6, 16]. In this context, Hoste suggested that the current incidence of CA-AKI and the prognosis varied, depending on the diagnostic criteria and population characteristics [17]. Up to now, the incidence of CA-AKI around the world has not been systematically reviewed, and further research is needed to determine its stages of severity and associated mortality.

This meta-analysis was conducted to estimate the world incidence and mortality associated with CA-AKI after coronary angiography, and to describe variations according to countries, regions, comorbidities and economic aspects.

Materials and methods

Data sources and literature search

A literature search was performed in MEDLINE, Embase and the Cochrane Database with the searching terms “contrast-induced acute kidney injury”, “risk” and “acute kidney injury” (Supplemental Item 1). The search was limited to human studies written in English and published before 30th June, 2019 (PROSPERO register number: CRD42019121534).

Study selection

Studies were included if they fulfilled the following criteria: (1) studies reporting on the risk factors of CA-AKI and (2) retrospective and prospective observational studies. Studies were excluded if they were classified as (1) duplicate articles; (2) contained no precise definitions of CA-AKI; (3) randomized controlled trials, meta-analyses, case reports, editorials, animal studies. At least two authors independently assessed the citations retrieved in the electronic search and identified eligible studies.

Data extraction and quality assessment

At least two independent authors (ZB Lun and LW Liu) assessed each study by screening the title, abstract or full-text independently. Then these two reviewers extracted data on the study characteristics, the incidence of CA-AKI and associated potentially modifiable risk factors. Newcastle–Ottawa Scale was performed to evaluate the quality of the included studies. The authors resolved discrepancies in study selection, data extraction and assessment of study quality through discussion with an arbitrator (Y Liu).

Definitions and outcomes

According to the geographical scheme designed by the United Nations Statistics Division, countries are classified into different continents and world regions [18]. Countries’ economies were assessed according to three ranges of gross national income per capita derived from the World Bank’s classification of income of economies: lower middle, upper middle, and high income countries [19, 20]. Using the World Health Organization’s World Health Statistics data, countries are also classified according to the percentage of total national health expenditure to gross domestic product (GDP) [21].

We evaluated the incidence of CA-AKI by three definitions. The definition of CA-AKI was an increase in serum creatinine ≥ 0.5 mg/dl or ≥ 25% from baseline within 72 h after exposure to contrast. In addition, we also analyzed the incidence of two other definitions of CA-AKI, as reported in Supplementary material. The two definitions of CA-AKI were: an increase in serum creatinine ≥ 0.3 mg/dl or ≥ 50% from baseline within 72 h and the criteria of acute kidney injury network (AKI stage 1, ≥ 0.3 mg/dl absolute or 1.5 to 2.0-fold relative increase in serum creatinine; AKI stage 2, > 2- to threefold increase in serum creatinine; AKI stage 3, > threefold increase in serum creatinine or serum creatinine > 4.0 mg/dl with an acute increase of > 0.5 mg/dl) or RIFLE criteria (risk, injury, failure, loss of kidney function and end-stage kidney disease) [22] or Kidney-disease-improving-global outcomes[23]. The CA-AKI associated mortality was considered as CA-AKI associated all-cause mortality.

Statistics analysis

The pooled incidence rates of CA-AKI and mortality were assessed with random-effects or fixed-effects models, based on the heterogeneity of included studies. Heterogeneity was quantified by the Q statistic and I2 statistic, which describes the percentage of total variation across studies due to heterogeneity and not due to sampling error [24]. If I2 was > 50%, a random-effects model was used. Otherwise, the fixed-effects model was adopted [25]. Subgroup analyses and meta-regression were performed for the following subsets of studies: percutaneous coronary intervention related studies, CAG or percutaneous coronary intervention related studies, clinical comorbidities, CA-AKI definitions, countries, continents, latitude, countries income classification, countries total wealth. Funnel plot was used to assess publication bias [26]. Publication bias was considered significant when p < 0.05. All analyses were performed with STATA (version 13.0) and R software (version 3.6.1; R Core Team, Vienna, Austria).

Results

Study characteristics and quality assessment

A total of 18,868 potentially relevant citations were identified and screened; 17,732 articles were retrieved for detailed evaluation, 134 of which fulfilled eligibility criteria (Supplemental References), representing 1.2 million patients from 24 countries worldwide (Fig. 1). Characteristics of all 134 studies are displayed in Supplemental Table 1. All studies were published in English and publication spanned 17 years. In the Table 1, most studies (23.2%) originated from Turkey (23 studies), followed by China (20 studies), United States (15 studies), Japan (10 studies), Italy (9 studies), Korea (7 studies). The top-three world zones where studies were conducted were Asia (67 studies), North America (15 studies), and Europe (14 studies). Most studies (48.5%) originated from high-income countries (48 studies) followed by upper middle income countries (47 studies). Most studies also originated from countries that spent over 5% of GDP on total health expenditure (94 studies). There were 97 studies (98.0%) that originated from countries located north of the equator. All studies were considered high-quality studies. Among them, 78 studies got 9 points, 45 studies got 8 points, and the rest got 7 points (Supplement Table 2).

Fig. 1
figure 1

Literature search and selection

Full size image
Table 1 Pooled incidence of CA-AKI
Full size table

Data synthesis

Pooled incidence rate of CA-AKI

We pooled the incidence rate of CA-AKI by 99 studies, the pooled rate of CA-AKI was 12.8% (95% CI 11.7–13.9%). In addition, the pooled incidence rate of CA-AKI in patients with percutaneous coronary intervention was 13.3% (95% CI 11.9–14.6%). When we study by clinical setting, the incidence rate in patients with chronic kidney disease, diabetes mellitus and ST-elevation myocardial infarction were 14.4%, 12.0% and 12.8%, respectively (Table 1). No publication bias was found, as confirmed by Egger’s tests (p = 0.07), and the funnel plot was shown in Fig. 2. As Table 2 knows, the pooled incidence rate of CA-AKI will not change with the growth of the year (rate change: 0.23%, 95% CI − 0.050 to 0.510, p = 0.103). At the same time, we found that compared with the incidence from 2002 to 2013, the incidence of CA-AKI from 2014 to 2019 did not increase (rate change: 0.08%, 95% CI − 0.1.490 to 3.070, p = 0.497).

Fig. 2
figure 2

Pooled incidence rate and mortality of CA-AKI by world zones. CA-AKI: increase in serum creatinine ≥ 0.5 mg/dl or ≥ 25% from baseline within 72 h after exposure to contrast

Full size image
Table 2 Meta-regression analysis examining the association of CA-AKI incidence rate and its associated mortality rate
Full size table

The pooled incidence rate of other definitions of CA-AKI were 15.1% (95% CI 12.9–17.4%) and 16.0% (95% CI 12.6–19.5%), respectively (detailed in Supplementary Table 3).

Pooled CA-AKI Associated Mortality Rate

As shown in Table 3, the pooled CA-AKI associated mortality rate was 20.2% (95% CI 10.7–29.7%). As Table 2 knows, the pooled CA-AKI associated mortality rate will not change with the growth of the year (rate change: − 1.05%, 95% CI − 3.070 to 0.970, p = 0.308). Compared with the pooled CA-AKI associated mortality rate from 2002 to 2013, the mortality rate of CA-AKI from 2014 to 2019 did not increase (rate change: − 10.2%, 95% CI − 0.27.710 to 7.340, p = 0.497).

Table 3 Pooled CA-AKI associated mortality rate
Full size table

The pooled CA-AKI associated mortality of other two definitions were 20.6% (95% CI 9.1–32.1%) and 27.5% (95% CI 7.8–47.2%) (Supplementary Table 3).

Pooled incidence and mortality associated with CA-AKI and variability around the World

The incidence rate in Turkey and the United States was 14.7%, in China was 12.4%, in Japan was 14.5% and in Italy was 12.9%. The pooled incidence rate in Asia was higher than in Europe and North America (13.2% versus 12.7% and 10.4%). According to country income classification, the incidence of CA-AKI in high income countries was slightly lower than that of upper middle income countries (11.4% versus 14.1%). And the pooled incidence of CA-AKI in countries with spent over 10% GDP on total health expenditure was less than the countries with 5–10% GDP on total health expenditure (12.2% versus 13.1%). Mortality associated with CA-AKI in Turkey and the United States were 14.6% and 35.1%, China was 3.9%, Japan was 19.9%. Compared with upper middle income countries, the high income countries has higher mortality associated with CA-AKI (26.4% versus 10.0%). The pooled CA-AKI associated mortality in countries with spent over 10% GDP on total health expenditure was higher than the countries with 5–10% GDP on total health expenditure (28.7% versus 12.1%) (Tables 1, 3 and Fig. 3).

Fig. 3
figure 3

Funnel plot of standard error. CA-AKI: increase in serum creatinine ≥ 0.5 mg/dl or ≥ 25% from baseline within 72 h after exposure to contrast

Full size image

Discussion

Our meta-analysis is the first article with the aim to assess the global incidence of CA-AKI. Since 1946 (the earliest period of Medline retrieval), we have identified a total of 18,868 large cohort studies. The pooled incidence and mortality of CA-AKI after coronary angiography were 12.8% and 20.2%, respectively. Our results showed that the incidence was not significantly related to national economic conditions and the percentage of gross domestic product spent on total health expenditure. And, as time goes on, the incidence of CA-AKI has not changed.

We have pooled the incidence of CA-AKI through 99 studies and found that this is similar to the incidence reported in previous studies [27, 28]. And we also compared different definitions and found that the incidence of other definitions of CA-AKI was both higher than that defined an increase in serum creatinine ≥ 0.5 mg/dl or ≥ 25% from baseline within 72 h after exposure to contrast. Both Chen and Centola’s studies have confirmed that the incidence of CA-AKI is affected by the presence of different definitions [29, 30]. In addition, we also analyzed the incidence of patients with different complications. We found that the incidence of patients with ST-elevation myocardial infarction and chronic kidney disease were higher than the pooled incidence (12.8%), because previous studies have confirmed that ST-elevation myocardial infarction and chronic kidney disease are risk factors of CA-AKI [31, 32]. However, the incidence of CA-AKI in diabetic patients was lower than pooled incidence, which may be related to the small sample size.

We observed that the pooled incidence of CA-AKI in most countries was close to the incidence estimated by us. However, the incidence data reported in Germany, Brazil and Iran were significantly higher, which may be associated with the number of studies and patient characteristics. The patients included in the studies from Germany were elderly, and this was a risk factor for CA-AKI [33, 34]. The increased incidence in Brazil and Iran may be related to the small number of patients included. Because only two studies from Africa and South America were included, it is not possible to draw conclusions from these settings. The incidence rate was 13.2% in Asia, followed by Europe (12.7%) and North America (10.4%, 15 studies were from the United States). The incidence in high income countries was lower than upper middle income countries. Countries that spend more than 10% of GDP on health also have lower rates than countries that spend more than 5–10% of GDP. This may be related to the scale of national economic development and the performance  of the medical and health security system. Moreover, a previous study suggested that AKI had a high incidence in intensive care unit in high income countries, making it difficult to prevent. Conversely, AKI in middle income countries often appears in rural health centers and hospitals as well as in large urban hospitals, and it is possible that it can be prevented through public health initiatives [17]. This may be one of the reasons for the difference. However, the opposite was true for CA-AKI associated mortality, which was higher in high income countries and in countries with high health expenditures as a percentage of GDP. The reason may be that the high mortality rate was at least partly due to other concomitant chronic diseases, which are more frequent in high income countries. Previous studies have confirmed that, in high income countries, patients with AKI were more likely to have risk factors that affect prognosis such as old age, heart failure, diabetes, and hypoproteinemia [35,36,37,38,39].

The incidence of CA-AKI is unclear, which may leads clinicians to underestimate CA-AKI. Our meta-analysis may therefore be of great significance to the future development and formulation of CA-AKI-related public health policies for the scientific community, government, and medical staff. Through meta-regression analysis, we can observe that the incidence and mortality of CA-AKI have not significantly changed since 2002. In addition, we also found that the incidence of CA-AKI from 2014 to 2019 did not change compared with the incidence from 2002 to 2013. Compared with the guidelines on myocardial revascularization of the European Society of Cardiology in 2010, the 2014 guidelines on myocardial revascularization of the European Society of Cardiology recommended that patients who undergo coronary angiography should be assessed for risk of CA-AKI [40, 41]. Moreover, the 2014 guidelines canceled the recommendation of ß-blockers, angiotensin-converting enzyme inhibitors or statins and reduced the level of evidence for N-acetylcysteine for chronic kidney disease patients. This may mean that the relevant prevention and treatment strategies of CA-AKI have not achieved actual benefits.

Our meta-analysis had several limitations. Firstly, the studies included in our analysis were targeted to the identification of risk factors of CA-AKI. However, our meta-study included as many as 134 articles, and the sample size was large. Secondly, we had included few studies from Africa and South America, which may lead to the lack of representativeness of results. Studies from the southern hemisphere were also very rare, and more studies are needed in these settings in the future. Thirdly, the various definitions of CA-AKI may have added to the heterogeneity. However, we analyzed the incidence of common definitions of CA-AKI, which makes our results equally representative. Our research compares morbidity and mortality in various countries, but international comparisons may vary depending upon geentic, environmental and social elements.

Conclusion

According to our metaanalysis, CA-AKI is a common complication in various regions of the world, and incidence and mortality are still high and did not decrease with time. This means that health care managers and and clinicians should pay more attention to CA-AKI to reduce its incidence and improve its prognosis.