Contrast induced nephropathy in patients undergoing intravenous (IV) contrast enhanced computed tomography (CECT) and the relationship with risk factors: A meta-analysis

European Journal of Radiology 82 (2013) e387–e399

Contents lists available at SciVerse ScienceDirect

European Journal of Radiology journal homepage: www.elsevier.com/locate/ejrad

Review

Contrast induced nephropathy in patients undergoing intravenous (IV) contrast enhanced computed tomography (CECT) and the relationship with risk factors: A meta-analysis Shira I. Moos ∗ , David N.H. van Vemde, Jaap Stoker, Shandra Bipat Department of Radiology, Academic Medical Centre, University of Amsterdam, Amsterdam, The Netherlands

a r t i c l e

i n f o

Article history: Received 7 February 2013 Received in revised form 4 April 2013 Accepted 24 April 2013 Keywords: Computed tomography Contrast medium Nephropathy Prevention Risk factors

a b s t r a c t Purpose: To summarize the incidence of contrast-induced nephropathy (CIN) and associations between CIN incidence and risk factors in patients undergoing intravenous contrast-enhanced computed tomography (CECT) with low- or iso-osmolar iodinated contrast medium. Methods: This review is performed in accordance with the preferred reporting items in systematic reviews and meta-analysis (PRISMA) guidelines. We searched the MEDLINE, EMBASE and Cochrane databases from 2002 till November 2012. Two reviewers included papers and extracted data. The pooled data were analysed by either fixed or random-effects approach depending on heterogeneity defined as the I2 index. Results: 42 articles with 18,790 patients (mean age 61.5 years (range: 38–83 years)) were included. The mean baseline eGFR was 59.8 mL/min and ranged from 4 to 256 mL/min. Of all patients 45.0% had an estimated glomerular filtration rate (eGFR) < 60 mL/min, 55.2% had hypertension; 20.2% had diabetes mellitus (DM) and 6.5% had congestive heart failure (CHF). The overall pooled CIN incidence, defined as a SCr increase of ≥25% or ≥0.5 mg/dL, was 4.96% (95%CI: 3.79–6.47). Data analysis showed associations between CIN and the presence of renal insufficiency, DM, malignancy, age > 65 years and use of non-steroidal anti-inflammatory drugs (NSAID’s) with odds ratios of 1.73 (95%CI: 1.06–2.82), 1.87 (95%CI: 1.55–2.26), 1.79 (95%CI: 1.03–3.11), 1.95 (95%CI: 1.02–3.70) and 2.32 (95%CI: 1.04–5.19), respectively while hypertension, anaemia and CFH were not associated (p = 0.13, p = 0.38, p = 0.40). Conclusion: The mean incidence of CIN after intravenous iodinated CECT was low and associated with renal insufficiency, diabetes, presence of malignancy, old age and NSAID’s use. © 2013 Elsevier Ireland Ltd. All rights reserved.

1. Introduction Contrast-induced nephropathy (CIN) is a major adverse effect of intravascular administration of iodinated contrast medium [1]. In most studies it is defined as an absolute (≥0.5 mg/dL) or relative (≥25%) increase in serum creatinine (SCr) within 48–72 h after iodinated contrast medium administration in absence of another explanation for the rise in SCr [1]. CIN has been associated with an increase in morbidity, mortality and medical resource consumption [2]. In an effort to reduce CIN, guidelines have been developed. Most (inter)national guidelines indicate that patients at risk should be identified by screening for the presence of risk factors in combination with renal

∗ Corresponding author at: Department of Radiology, G1-215, Academic Medical Centre, University of Amsterdam, Meibergdreef 9,1105 AZ Amsterdam, The Netherlands. Tel.: +31 (0)020 5662630; fax: +31 (0)020 5669119. E-mail address: [email protected] (S.I. Moos). 0720-048X/$ – see front matter © 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.ejrad.2013.04.029

insufficiency [3–6]. Usually this concerns verification of the estimated glomerular filtration rate (eGFR) or SCr and risk factors such as diabetes, hypertension and old age. If patients are at risk, preventive measures should be taken [3–6]. These guidelines have led to discussion, mostly about the identification of patients at risk and the type of prevention measures that usually are recommended by these guidelines [7–9]. One of the problems is that most of the evidence used to develop these guidelines is based on studies evaluating patients undergoing intra arterial cardiac interventions with high volume of sometimes high osmolar iodinated contrast media [9,10]. This population differs significantly from the patient population undergoing iodinated contrast enhanced computed tomography (CECT), as does the occurrence of adverse events [7]. This could be related to a difference in association with risk factors and the development of CIN [10,11]. A recently published systematic review showed a pooled incidence of CIN of 6.4% in patients undergoing intravenous CECT and higher incidences of CIN were seen in patients with renal

e388

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

insufficiency or diabetes mellitus compared to patients without these risk factors [12]. No associations were seen between CIN incidence and hypertension or with the volume of administrated iodine. However in most guidelines other risk factors are also mentioned, such as age, gender, race, anaemia, congestive heart diseases, use of nephrotoxic medication, dehydration, and cardiovascular diseases and these were not taken into account in the aforementioned meta analysis [3–6,12]. We hypothesize that the CIN incidence is lower in a population representing the large majority of patients referred for CECT, mostly outpatients who are hemodynamically and respiratory stable. Further it is important to identify whether risk factors mentioned in guidelines are associated with CIN incidence in patients receiving intravenous iodinated contrast medium. Up till now the results of clinical trials about the association between CIN and different risk factors as mentioned in most prevention guidelines, differ in outcome and very seldom mention more than three risk factors. The purpose of our meta-analysis was to summarize the incidence of CIN in patients undergoing intravenous CECT and to study associations between CIN and several risk factors that are mentioned in most prevention guidelines. 2. Materials and methods 2.1. Search strategy and study selection We searched the MEDLINE, EMBASE and Cochrane databases from 2002 till 10th of November 2012 to identify all relevant studies on CIN. In the previously mentioned meta-analysis of Kooiman et al., no papers published before 2002 were found to be relevant [12]. The systematic review was conducted in accordance with the Preferred Reporting Items in Systematic Reviews and Meta-analysis (PRISMA) guidelines [13]. We used the following search terms: (CIN (Title, Abstract, Keyword) AND Nephropathy (Title, Abstract, Keyword) OR Contrast-induced nephropathy (Title, Abstract, Keyword) OR Contrast induced nephropathy (Title, Abstract, Keyword)). The search strategy is described in detail in Appendix 1. Firstly all studies not related to CIN were excluded. Secondly all comments, reviews and conference papers were excluded to select potential relevant papers. Two reviewers checked all potential relevant data to select relevant papers. Of all relevant papers full text were retrieved for further checking of inclusion and exclusion criteria. 2.2. Inclusion and exclusion criteria Inclusion and exclusion criteria were checked independently by two reviewers SM and SB and disagreements were resolved by consensus. Papers were included when: (1) written in English, German, Dutch, French, Italian and Spanish; (2) patients underwent intravenous CECT (if data on intra-arterial examinations were also given and data could not be selected for only intravenous administration of contrast medium, the study was not included); (3) patients underwent intravenous CECT with low- or iso-osmolar contrast medium (if data on high osmolar medium was also given, and could not be split for low osmolar medium, the study was not included); (4) CIN incidence and risk factors were presented and (5) follow-up period for determining CIN between 24 h and 1 week after intravenous CECT (if follow-up was done >1 week and the follow-up data between 24 h and 1 weeks could not be selected, the study was not included). The exclusion criteria were: (1) duplicate publication (most recent paper was included for analysis); (2) ICU patients included

and these data could not be separately identified; (3) less than 10 patients with intravenous CECT. Of all included articles, data on methodological assessment, baseline patient characteristics, preventive measures, characteristics of computed tomography (CT), CIN incidence determination were assessed by the same reviewers independently SM and SB. A third reviewer [bDV] checked all collected data and was the decisive factor in case of disagreement between the first two reviewers.

2.3. Methodological assessment Methodological assessment of the included studies was done according to the Delphi list for randomized controlled trials (RCT) [14], combined with the signalling aspects of the QUADAS-2 tool for diagnostic accuracy studies [15]. The following characteristics were assessed whether: (1) the study was a cohort or RCT; (2) the study was a single centre or multicentre study; (3) data were extracted prospectively or retrospectively; (4) a consecutive or random sample of patients was enrolled; (5) inclusion/exclusion criteria were specified; (6) the spectrum of patients was representative of the patients who will receive the test in real life practice; (7) the administration of contrast medium was described in sufficient detail to permit its replication; (8) the time period between contrast medium administration and follow-up was reasonable (performed within 2–4 days, 48–92 h); (9) the whole (or random sample) underwent follow-up for occurrence/determination of CIN. In case of a RCT, the following data were also assessed, whether: (10) the method of randomization was described; (11) groups were similar at baseline regarding the most important indicators; (12) relevant data presented with confidence intervals (CI) and (13) the RCT was double blinded.

2.4. Baseline patient characteristics The following data were assessed: (1) number of patients included and analysed; (2) age of patients (mean ± SD, median and/or range); (3) proportion of patients >60 years and/or >75 years; (4) male: female ratio; (5) baseline eGFR (mean ± SD, median and/or range); (6) proportion of patients with an eGFR < 60 mL/min and proportion of patients with an eGFR < 45 mL/min [3,5]; (7) method for calculation eGFR (Cockcroft-Gault formula or Modification of Diet in Renal Disease (MDRD-4 or MDRD-6)) (because both formula’s were used in the papers included in the metaanalysis the eGFR is expressed in mL/min throughout the article); (8) baseline SCr (mean ± SD, median and/or range); (9) proportion of patients with renal insufficiency: (10) proportion of patients with diabetes mellitus: (11) proportion of patients with hypertension; (12) proportion of patients using nephrotoxic medication; (13) proportion of patients with anaemia; (14) proportion of patients with congestive heart disease and (15) proportion of patients with cardiovascular diseases. In case data on other risks factors, such as liver disease or stroke were presented, the proportions of patients with these conditions were also extracted. If data on proportion of patients with renal insufficiency or diabetes mellitus were not complete, we contacted the corresponding authors of these papers for additional data.

2.5. Preventive measures We also recorded whether prevention measures, before or after the CT examination, were performed. If this was done, details on the program were assessed and also the proportion of patients who received prevention measures (e.g. hydration and discontinuation of nephrotoxic medication).

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

e389

2.6. Characteristics of CT protocols The following data on CT protocol characteristics were assessed: (1) whether different CT protocols (e.g. different volumes of contrast medium, different types of contrast media) were compared, (2) information on type of contrast medium (brand, concentration of iodine, amount in milliliter or gram, monomer or dimer and osmolarity) and (3) the type of CT examination performed (e.g. abdomen or chest, brain). 2.7. Overall CIN incidence and risk factors The following data on CIN incidence determination were extracted: (1) definition of CIN, defined as a relative increase SCr ≥ 25% or an absolute increase of SCr ≥ 0.5 mg/dL; (2) number of patients in whom CIN was determined and (3) time between CT examination and occurrence or determination of CIN (within/mean/median/SD/range). If data were not complete a request for additional data was sent to corresponding author. In addition, 2 × 2 tables for the following risk factors were obtained: renal insufficiency, diabetes mellitus, hypertension, congestive heart failure, use of nephrotoxic medication, age and anaemia. In case data on other risks factors, such as liver disease or stroke were presented we also extracted these data. 2.8. Statistical analysis Firstly, we calculated the CIN incidence as follows: number of patients with CIN divided by the total number of patients undergoing follow-up after intravenous CECT. The incidence was presented as percentage per study with corresponding 95% confidence interval (CI). The I2 index was used to quantify heterogeneity for the incidence [16]. In case of I2 ≥ 25%, we used a random-effects model and in case of I2 < 25%, we used a fixed-effects model to pool the CIN incidence. This analysis was performed on logit-transformed incidence, because these are assumed to follow a normal distribution across studies and, therefore, the mean logit CIN incidence with corresponding standard errors was calculated. After antilogit transformation, pooled estimate of CIN incidence was obtained (95%CI). Pooled estimates of CIN were determined for studies defining CIN as a relative increase of SCr ≥ 25%, for studies defining an absolute increase of SCr ≥ 0.5 mg/dL or studies defining CIN as a relative increase ≥25% or absolute increase ≥0.5 mg/dL. In addition, per criteria subgroup analyses were performed on the type of contrast agents (use of low-osmolar monomer or iso-osmolar dimer or using both contrast media). All above mentioned analyses were executed by using SAS software (SAS 9.3 procNlmixed; SAS Institute, Cary, NC). For the association between CIN and risk factors, the odds ratio (OR) was calculated based on the 2 × 2 tables. The data were pooled using fixed-effects (I2 < 25%), or the random-effects (I2 ≥ 25%) models if at least three data sets were available. The results were presented in Forest plots. Statistical significance for the association was set at p = 0.05. These data were analysed using Cochrane RevMan software (version 5.0.; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark).

After checking inclusion and exclusion criteria, 42 papers fulfilled all criteria and methodological quality and all relevant data were extracted. The remaining 47 were excluded due to missing data on patients undergoing intravenous CECT (n = 29); or it was not clear whether patients received low- or iso-osmolar contrast medium (n = 2); there were no data on CIN incidence and/or risk factors mentioned (n = 6); there was no adequate follow-up period (24 h to 1 week) for determining CIN (n = 4). In addition, three papers were found to be duplicates (n = 3) and three papers included ICU patients (n = 3). The results of the search strategy and selected studies are shown in Fig. 1 and Appendix 2.

3. Results

3.3. Methodological assessment

3.1. Search strategy and study selection

Of the 42 papers, there were seven RCTs; randomization was described in all but one paper. In three RCTs the baseline characteristics were not similar in both groups, the number of patients with diabetes were significantly different between groups. Two RCTs did not conceal treatment allocation during data analysis. All papers presented their relevant data with 95%CIs. Five RCTs compared

The initial strategy resulted in 4892 papers of which 1739 were CIN related. After excluding case-reports, comments, letters, reviews and conference papers, eventually 244 studies were found to be potential relevant. Eighty-nine papers were found relevant

Fig. 1. The results of the search strategy, study selection and inclusion.

after reading the abstract and full texts were retrieved for further selection. 3.2. Inclusion and exclusion criteria

e390

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

Table 1 Methodological and design criteria of all 42 included articles. Author, year of publication

Type of studya

Study designb

Data collectionc

Patient selectionc,d

Inclusion/ exclusion criteriad ,e

Spectrum of Description patientsf contrast administrationg

Time interval 2–4 days (48–96 h)h

Complete verificationi

Abujudeh (2009) Balemans (2012) Barrett (2006)

Cohort Cohort RCT comparing contrast agents Cohort Cohort Cohort RCT comparing contrast agents Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort RCT comparing NAC vs. NaCla Cohort Cohort Cohort Cohort Cohort RCT comparing contrast agents Cohort

Single centre Single centre Multicentre

Retrospective Prospective Prospective

Unclear Yes Yes

Yes Yes Yes

Yes No No

Yes Yes Yes

Yes (within 48 h) No (within 3–5 days) Yes (within 48–72 ± 6 h)

Yes Yes Yes

Single centre Single centre Single centre Single centre

Prospective Retrospective Retrospective Prospective

Yes Yes Yes Yes

Yes Yes Yes Yes

Yes Yes Yes No

Yes No Yes Yes

Yes (within 3 days) Yes (within 3 days) Yes (within 48–72 h) Yes (within 48–72 h)

Yes Yes Yes Yes

Single centre Single centre Single centre Single centre Single centre Single centre Single centre Single centre Multicentre Single centre

Retrospective Prospective Prospective Retrospective Retrospective Retrospective Prospective Retrospective Retrospective Prospective

Unclear Yes Yes Unclear Yes Yes Unclear No Unclear Yes

No Yes Yes Yes No Yes Yes Yes Yes Yes

Yes No No Yes Yes Yes Yes Yes Yes No

Yes Yes Yes Yes Yes Yes No Yes Yes Yes

Yes (within 48 h) No (within 3–5 days) Yes (within 72 h) Yes (within 48 h) No (within 1 week) Yes (within 72 h) Yes (within 48–72 h) No (within 1 week) No (24 ± 4 h) Yes (within 48–72 h)

Yes Yes Yes Yes No No Yes Yes Yes Yes

Single centre Single centre Single centre Single centre Single centre Multicentre

Retrospective Retrospective Prospective Retrospective Retrospective Prospective

Unclear Unclear Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes No

Yes Yes Yes Yes Yes Yes

Yes (within 48–96 h) Yes (within 3 days) Yes (within 2–4 days) No (within 5 days) No (within 1–5 days) Yes (within 48–72 h)

Yes Yes Yes Yes Yes Yes

Single centre

Retrospective

Unclear

Yes

Yes

Yes

Yes (within 48–72 h)

Yes

Cohort Cohort Cohort Cohort Cohort Cohort

Single centre Single centre Single centre Single centre Single centre Single centre

Prospective Retrospective Prospective Retrospective Prospective Retrospective

Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes Yes Yes Yes Yes Yes

Yes (within 48–72 h) Yes (At 72 h) No (within 24–72 h) No (within 1 week) Yes (within 2 days) Yes (within 72 h)

Yes Yes Yes No Yes Yes

Cohort Cohort Case–control RCT comparing contrast agents Cohort RCT comparing NAC vs. NaCla Cohort Cohort RCT comparing contrast agents Cohort Cohort Cohort

Single centre Single centre Single centre Single centre

Prospective Retrospective Retrospective Prospective

Yes Yes Yes Yes

Yes Yes Yes Yes

Yes No Yes Yes

Yes Yes No Yes

No (within 2–7 days) Yes (Within 48–72 h) Yes (within 48–72 h) No (within 1–3 days)

Yes Yes Yes Yes

Single centre Single centre

Retrospective Prospective

Unclear Yes

Yes Yes

Yes Yes

Yes Yes

Yes (within 48 h) Yes (within 2–4 days)

Yes Yes

Single centre Single centre Multicentre

Prospective Prospective Prospective

Unclear Yes Yes

Yes Yes Yes

No Yes Yes

No Yes Yes

Yes (within 24–72 h) Yes (within 72 h) Yes (within 48–72 h)

Yes Yes Yes

Single centre Single centre Single centre

Retrospective Prospective Prospective

Unclear Yes Unclear

Yes Yes Yes

Yes No No

No Yes No

Yes (within 48 h) Yes (within 48–96 h) No (within 48–96 h)

Yes Yes No

Becker (2005) Bruce (2009) Cheruvu (2007) Dillman (2012) Dittrich (2007) El-Hajjar (2008) Garcia-Ruiz (2004) Hipp (2007) Holmquist (2006) Hopyan (2008) Hsu (2012) Josephson (2005) Karlsberg (2011) Khalili (2006) Kim (2010) Kim (2011) Kingma (2010) Kooiman (2010) Krol (2007) Kuhn (2008) Kulvatunyou (2011) Langner (2008) Lencioni (2010) Lima (2010) Lodhia (2009) Lufft (2002) Matsushima (2011) Mitchell (2010) Murakami (2010) Najjar (2002) Nguyen (2008) Pahade (2011) Poletti (2007) Sancak (2002) Sandstede (2007) Thomsen (2008) Tremblay (2005) Weisbord (2008) Yoshikawa (2011) a

The study was a cohort or randomized controlled trial (RCT). The study was a single centre or multicentre study. Data were extracted prospectively or retrospectively. d A consecutive or random sample of patients was enrolled. Inclusion/exclusion criteria were specified. f The spectrum of patients was representative of the patients who will receive the test in daily practice. g The administration of contrast medium was described in sufficient detail to permit its replication. h The time period between contrast medium administration and follow-up was reasonable (performed within 2–4 days). i The whole (or random sample) underwent follow-up for occurrence/determination of CIN. b c

e

different types of contrast media and two RCTs compared different CIN prevention methods (e.g. hydration methods). The other 35 papers were cohort studies in which 15 collected their data in a prospective manner. Twelve papers only included patients with renal insufficiency (SCr > 1.5 mg/dL or eGFR < 60 mL/min) and three papers excluded all patients with renal insufficiency defined as SCr > 1.5 mg/dL. In Table 1, the results of the methodological criteria are presented in detail. 3.4. Baseline patient characteristics We included 42 papers containing 18,790 patients, with a mean age of 61.5 years (range 38–83 years). The mean eGFR at baseline

was 59.8 mL/min ranging from 4 to 256 mL/min, the mean SCr at baseline was 1.3 mg/dL ranging from 0.14 to 6.6 mg/dL. The male/female ratio was 10,064:6435 (1.6:1.0).

3.4.1. Renal insufficiency If renal insufficiency was defined as a SCr > 1.5 mg/mL, 13.5% of the patients (1458/10,776) had renal insufficiency at baseline. When renal insufficiency was defined as an eGFR < 60 mL/min 45.0% (4943/10,979) had renal insufficiency at baseline. The proportion of patients with an eGFR < 45 mL/min was 8.8% (648/7355) at baseline. Four papers reported the number of patients with renal insufficiency by a different definition; therefore these data were not

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

comparable and not further mentioned. Six papers did not report the number of patients with renal insufficiency. 3.4.2. Diabetes mellitus At baseline 20.2% (3128/15,517) of the patients had diabetes mellitus. Thirty-two papers mentioned the number of patients with diabetes mellitus; in the other seven papers data were not complete despite our effort to obtain additional information. 3.4.3. Other risk factors Twenty papers described the number of patients that used nephrotoxic medication. Four excluded all patients using nephrotoxic medication. The proportion of patients using nephrotoxic medication was 30.6% (1292/4223). The proportion of patients with hypertension (HT) was 55.2% (2371/4298) in the 14 papers that mentioned HT. The proportion of patients with congestive heart failure (CHF) was 6.5% (507/7845) in 21 papers that mentioned CHF. In 10 papers patients with CHF were excluded. Detailed information on all patient characteristics are shown in Table 2. 3.5. Preventive measures Preventive intravenous or oral hydration was described in 29 papers. Five papers excluded patients if they received any form of prevention. In the remaining eight papers it was unclear whether patients received prevention. The mean percentage of hydrated patients was 65.1% (range 9–100%). Intravenous hydration was the method of prevention in 27 studies. In the other two studies patients received oral hydration. In five studies patients received a combination of oral hydration and intravenous hydration. Nine studies described that patients received N-acetylcysteine orally or intravenously before the CT examination. Four used sodiumbicarbonate and the other 17 papers stated they hydrated with normal saline 0.9% or did not further specify. 3.6. Characteristics of CT protocols Sixteen studies compared different CT protocols. Six papers did this to determine the difference between contrast medium in CIN incidence and the other 10 studies used different media for different types of CT examinations. The type of contrast medium was either low-osmolar monomer or iso-osmolar dimer. The mean contrast volume was 120.6 mL (range 60–160). Three mentioned the grams of iodine administrated this ranged from 6.4 to 40 g of iodine. Thirty papers specified what type of CT examination was done (e.g. CT abdomen and CT cerebrum), the other nine did not specify. 3.7. Overall CIN incidence Thirty-nine papers defined CIN as a relative (≥25%) and/or absolute (≥0.5 mg/dL) increase of SCr from baseline value; one paper defined CIN as an increase of SCr > 20%, one paper defined CIN as a decrease of eGFR ≥ 25% and one paper reported CIN incidence but how it was defined was not mentioned in the paper. CIN was defined within 48–72 h after intravenous contrast administration in 30 papers. In the remaining 12 papers, CIN was also defined outside the range of 48–72 h. 3.7.1. Relative increase of SCr ≥ 25% If CIN was defined as relative increase of SCr ≥ 25%, the overall pooled incidence was 4.72% (95%CI: 3.51–6.33). The pooled incidence in studies using iso-osmolar dimer contrast medium was 5.02% (95%CI: 3.68–6.82). In studies using low-osmolar monomer contrast medium the pooled incidence was 4.72% (95%CI:

e391

3.01–7.35) and in studies using both iso-osmolar monomer or lowosmolar dimer 6.42% (95%CI: 4.50–9.07). No significant differences were seen between the pooled incidences of the different contrast media (p = 0.83, p = 0.30, p = 0.29). The I2 values for all three subgroups were 46%, 89% and 76% respectively. 3.7.2. Absolute increase of SCr ≥ 0.5 mg/dL If CIN was defined as absolute increase of SCr ≥ 0.5 mg/dL, the overall incidence was 2.77% (95%CI: 2.01–3.80), and the subgroup analyses showed pooled incidence in datasets using isoosmolar dimer contrast medium, low osmolar monomer medium or datasets using both of 3.71% (95%CI: 2.41–5.66), 2.26% (95%CI: 1.19–4.24) and 3.57% (95%CI: 2.32–5.44), respectively. No significant differences were seen between contrast media (p = 0.20, p = 0.90, p = 0.24). The I2 values for all three subgroups were 68%, 83% and 66% 3.7.3. Relative increase of SCr ≥ 25% or absolute increase of ≥0.5 mg/dL If CIN was defined as either a relative increase (SCr ≥ 25%) or absolute increase (≥0.5 mg/dL), the overall incidence was 4.96% (95%CI: 3.79–6.47), in studies using iso-osmolar dimer contrast medium 4.28% (95%CI: 2.61–6.95), and in studies using lowosmolar monomer contrast medium 5.37% (95%CI: 3.84–7.47) and when using both contrast media 7.47% (95%CI: 4.26–12.76). No significant differences were seen between contrast media (p = 0.45, p = 0.14, p = 0.32). The I2 values for all three subgroups were 51%, 85% and 80%. 3.7.4. Comparison of the three commonly used criteria In datasets with CIN defined as an absolute SCr increase (≥0.5 mg/dL), the lowest overall CIN incidence was seen, compared with datasets using a relative SCr increase (≥25%) as criterion (p = 0.02) or datasets using both criteria (relative increase of ≥25% or absolute increase of SCr of ≥0.5 mg/dL) (p = 0.01). 3.8. CIN incidence and associated risk factors For the associations with risk factors, not all data could be extracted. For major risk factors, such as diabetes and renal insufficiency we tried to complete the information by contacting the corresponding authors but did not receive any results. Therefore data analyses were performed if 2 × 2 data could be extracted from the papers. 3.8.1. Renal insufficiency In datasets evaluating association between SCr > 1.5 mg/dL and CIN the OR was 4.10 (95%CI: 2.26–7.42, p < 0.001). For eGFR < 60 mL/min and CIN the OR was 1.73 (95%CI: 1.06–2.82; p = 0.03). For eGFR < 45 mL/min, the OR was 2.41 (95%CI: 1.12–5.16, p = 0.02). The I2 values were 68%, 33% and 55%. The datasets are summarized in Fig. 2a–c. 3.8.2. Diabetes mellitus The pooled analysis of the relationship between diabetes mellitus and CIN showed a homogenous dataset (I2 = 2%) and a positive association, OR 1.87 (95%CI: 1.55–2.26) p < 0.001. The data are summarized in Fig. 3. 3.8.3. Old age In datasets evaluating association between old age (>65 years) and CIN, an OR of 1.95 (95%CI: 1.02–3.70), p = 0.04 was seen. I2 value was 64%, data are shown in Fig. 4.

e392

Table 2 Patient characteristics of all 42 included papers. N

Age (years) Mean ± SD; median (range)

Male:female

eGFR (mL/min) Mean ± SD; median (range)

eGFR calculation

SCr (mg/dL) Mean ± SD; median (range)

Abujudeh (2009)

164

90:74

747

539:405

88.2 ± 34.2 (8.9–256.9) 47.8 ± 9

MDRD 4

Balemans (2012)

56.3 ± 19.2 (14–98) 71.3 ± 10

1.03 ± 0.7; 0.3–6.6 1.37 ± 0.38

Barrett (2006) Iopamidol (isovue)

77

67.3 ± 13

54:23

44 ± 14

Barrett (2006) Iodixanol (visipaque)

76

67 ± 11.5

51:25

45.2 ± 11.3

Becker (2005) Bruce (2009) Iohexol (omnipaque) Bruce (2009) Iodixanol (visipaque) Cheruvu (2007) Iohexol (omnipaque) Cheruvu (2007) Iodixanol (visipaque) Dillman (2012) Iohexol (omnipaque) Dillman (2012) Iopamidol (isovue) Dittrich (2007)

100 5328

65 ± 13 53 ± 18.4

75:25 2454:1982

37 ± 12 85.6 ± 31.1

CockcroftGault formula CockcroftGault formula MDRD 4

462

63 ± 15.6

219:118

51.3 ± 19.6

194

58 (7–85)

85:109

0.3–1.4c

374

173:201

0.4–2.4c

190

63.4 (15–92) 56.1 ± 14.7

85:105

87.6 ± 26.2

199

56.7 ± 14.7

100:99

85.9 ± 24.6

162

66 ± 14; 69 (23–90)

104:58

87 ± 34

El-Hajjar (2008)

400

76 ± 8.9

314:86

Garcia-Ruiz (2004)

50

66.9 ± 8.8 (40–80)

41:9

Hipp (2007)

235

188:47

Holmquist (2006)

29

44 ± 20 (13–92) 83 (68–93)

7:22

Hopyan (2008)

198

65.4 ± 18.9

99:99

Hsu (2012)

209

79.6 ± 9.8

156:53

Josephson (2005)

1075

36 (12–49)

49.7 ± 28.8

MDRD 4

>60 years % (n)/>75 years % (n)

NA/40.8 (944a )

Number of patients with risk factorsg

17.7 (29)a 6.7 (11) 100 (747)a

DM: 30, nephrox: 35, hypertension: 63, other risk: 21 (CIN), non-caucasian: 38 DM: 177, nephrox:362, hypertension: 537, CHF: 116, anaemia: 309 DM: 15, nephrox: 9, hypertension: 56, CHF: 0, other risk: 66

1.6 ± 0.4

100 (77)a

1.6 ± 0.5

100 (76)a

DM: 21, nephrox: 4, hypertension: 56, CHF: 0, other risk: 76

2.1 ± 0.6 0.98 ± 0.32

100 (100) 19.2 (1023)a 6.3 (337) 69.3 (320)a 58.4 (270)

0.89 ± 0.24

0 (0)

DM: 16, CHF: 0, nephrox: 0 DM: 721, other risk: 217 (African American) DM: 160, other risk: 17 (African American) DM: 22, nephrox: 90, hypertension: 62, CHF: 3 DM: 91, nephrox: 197, hypertension: 193, CHF: 25 DM: 28, African American: 12

0.91 ± 0.42

0 (0)

DM: 16, African American: 12

7.4 (12)

DM: 34, nephrox: 111, hypertension: 138, CHF: 0, anaemia: 29

1.8 ± 0.2

72 (287)a 100 (400)

DM: 165, other risk 5 (single kidney) 101, 5 (contrast volume)

2.92 ± 1.39

100 (50)

DM: 14, nephrox: 31, other risk: 11 (heart disease), 5 (cerebrovascular), 14 (peripheral) NA

1.52 ± 0.42

MDRD 4

eGFR < 60 mL/min % (n)/SCr > 1.5 mg/dL

CockcroftGault formula CockcroftGault formula CockcroftGault formula MDRD 4

1.1 ± 0.3

CockcroftGault formula

1.37 (0.54–3.25)

CockcroftGault formula MDRD 4

0.99

34 (59)b

1.6 ± 0.6

75.6 (158)b

NA/23(37)

NA/10.6(25) 100(29)/NA

8.1 (19)a 7.2 (17) 100 (29)a , d

6.7 (72)

DM: 2, CHF: 9, other risk factors: 8 (COPD), 2 active malignancy, 2 angina pectoris, 3 other (chronic aortic dissection, AAA) DM: 22, other risk: 134 (ischaemic stroke), 64 (haemorrhagic stroke) DM: 57, nephrox: 19, hypertension: 132, CHF: 30, other risk factors: 158 (renal insufficiency), 22 (shock), 42 (coronary artery disease), 30 (CVA) NA

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

Author, year of publication Subgroups

Table 2 (Continued) Author, year of publication Subgroups

N

Age (years) Mean ± SD; median (range)

Male:female

Karlsberg (2011)

264

168:96

Khalili (2006) N-Acetylcysteine Khalili (2006) Control Kim (2010)

35

64.7 ± 10.93; 65 (39.0–88.0) 59.76 ± 1.99

35

eGFR (mL/min) Mean ± SD; median (range)

eGFR calculation

SCr (mg/dL) Mean ± SD; median (range)

>60 years % (n)/>75 years % (n)

eGFR < 60 mL/min % (n)/SCr > 1.5 mg/dL

Number of patients with risk factorsg

17:13

2.1 ± 1.2; 1.8 (0.6–11.6) 1.43 ± 0.5

NA

55.89 ± 12.92

19:11

1.31 ± 0.15

NA

520

64.4 ± 11.2

410:110

43.0

MDRD 4

1.75 ± 0.65

100 (520)a

Kim (2011) Kingma (2010) Kooiman (2010)

750 148 237

55.9 ± 18.9 65 (21–83) 57 ± 1.1

418:332 88:66 107:130

77 ± 19

MDRD 4

1.1 ± 0.5 0.98 ± 0.26

13.7 (103) 10.1 (15)a , e 22.8 (54)a

Krol (2007) Kuhn (2008) Iopamidol (Isovue) Kuhn (2008) Iodixanol (visipaque) Kulvatunyou (2011) Langner (2008)

224 125

68.2 ± 14.1 69.5 ± 10.05

138:86 54:71

47.6 ± 13.5

MDRD 4

1.46 ± 0.44

100 (125)a

DM: 158, nephrox: 30, hypertension: 344, CHF: 25, other risks: 119 (chronic liver disease), 15 (previous kidney transplant), 55 (cerebrovascular disease) DM: 79, hypertension: 147 NA DM: 27, nephrox: 82, anaemia: 154, CHF: 28, other risks: 5 (multiple myeloma) DM: 85 DM: 125, CHF: 0

123

68.3 ± 9.19

62:61

49.9 ±11.6

1.41 ± 0.38

100 (123)a

DM: 123, CHF: 0

543 100

38.8 ± 1.1 65.4 ± 13.9

380:163 52:48

1.07 ± 0.05 0.96 ± 0.35

20 (20)

Lencioni (2010)

493

68.1 ± 13.6 (21–97)

283:210

64.7 ± 33.3

1.31 ± 0.94

34.3 (169)a

Lima (2010)

575

67 ± 15

299:276

76 (61–89)

Lodhia (2009)

216

53.2

128:88

Lufft (2002)

33

56 ± 14

20:13

Matsushima (2011)

1184

38 (18–95)

822:362

Mitchell (2010)

633

46.9 ± 14.9

270:363

Murakami (2012) Najjar (2002)

938 144

70 ± 11.2 51.5 ± 2.0

504:434 78:66

48.5 ± 8.55

MDRD 4

1 ± 0.27 0.88 ± 0.05

Nguyen (2008) Iodixanol (visipaque)

61

63 ± 11.7 (29–84)

45:16

54.8 ± 16.58

1.77 ± 0.24

54.1(33)/NA

100 (61)b

Nguyen (2008) Iopromide (ultravist)

56

65.8 ± 13.4 (18–86)

38:18

52.98 ± 26.02

CockcroftGault formula CockcroftGault formula

1.75 ± 0.32

53.6(30)

100 (56)b

Pahade (2011)

56

65 (37–93)

32:24

1 (0.8–1.1)

0f

CockcroftGault formula MDRD 4

36.4 (12) 4.3 (51)

0.9 (0.2–4.8)

2.4 (15)a

0.97 ± 0.28

100 (938)a 2.1 (3)

Other risks: trauma scores DM: 41, nehrox: 12, hypertension: 68, other risk factors: 56 (coronary disease), 55 (hypercholesterolemia) DM: 198, nephrox: 119, CHF: 144

DM: 97, hypertension: 339, CHF: 34, other risks: 126 (coronary art disease), 40 (black), 575 (stroke patients) DM: 66, nephrox: 128, CHF: 0, other risk factors: 114 (ascites)

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

CockcroftGault formula MDRD 4

NA

DM: 3, nephrox: 19, hypertension: 33, CHF: 0 DM: 95, CHF: 0 DM: 45, anaemia: 23, CHF: 22, other risk factors: 28 (vascular disease) CHF: 0 DM: 29, CHF: 3, other risk factors: 34 (severe cirrhosis) DM: 23, hypertension: 31, nephrox: 0, other risk factors: 22 (dyslipidemia), 23 (diabetes and hypertension) DM: 10, hypertension: 19, nephrox: 0, other risk factors: 11 (dyslipidemia), 10 (diabetes and hypertension) DM: 17 e393

e394

Table 2 (Continued) Age (years) Mean ± SD; median (range)

Male:female

Poletti (2007) N-acetylcysteine Poletti (2007) NaCl Sancak (2002)

44

69.5 ± 18.7

26:18

1.65 ± 0.4

DM: 9, nephrox: 12, CHF: 0

43

72.7 ± 17.2

29:14

1.67 ± 0.41

DM: 6, nephrox: 15, CHF: 0

20

11:9

95.5 ± 15.1

Sandstede (2007) Thomsen (2008) Iodixanol (visipaque)

99 72

42 ± 10 (27–71) 64.5 ± 13.4 65.4 ± 12.1

68:31 46:26

44.4 ± 11.8 43 ± 13.3

Thomsen (2008) Iomeprol (Iomeron)

76

67.1 ± 14.1

58:18

41.5 ± 13.1

Tremblay (2005) Weisbord (2008)

56 421

48.2 ± 22.3 69 ± 10

50:6 403:18

Yoshikawa (2011)

180

67 ± 8

87:93

a

c d e f g

Renal insufficiency defined as eGFR < 60 mL/min. Renal insufficiency defined as SCr > 1.5 mg/dL or eGFR < 60 mL/min. Only ranges were given without median value. Renal insufficiency defined as eGFR < 50 mL/min. Assuming all patients identified as being at risk had an eGFR < 60 mL/min. Renal insufficiency defined as SCr > 1.8 mg/dL. Nephrox indicates number of patients using nephrotoxic medication.

eGFR calculation

MDRD 6 CockcroftGault formula CockcroftGault formula

SCr (mg/dL) Mean ± SD; median (range)

>60 years % (n)/>75 years % (n)

eGFR < 60 mL/min % (n)/SCr > 1.5 mg/dL

0.62 ± 0.13

0 (0)

DM: 0, nephrox: 0, hypertension: 0

1.7 ± 0.7

100 (72)b

DM: 20, nephrox: 0 DM: 21, nephrox: 7, CHF: 0

1.7 ± 0.6

100 (76)b

DM: 9, nephrox: 10, CHF: 0

1.51 ± 0.23 100 (421)a

MDRD 4

70 ± 15

MDRD 4

0.77 ± 0.19

0

DM: 10 DM: 172, CHF: 68, other risk factors: 45 (cerebrovascular disease), 54 (peripheral vascular disease) DM: 71, hypertension: 153, other risk factors: 77 (hyperlipidemia), 70 (smoking)

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

N

b

eGFR (mL/min) Mean ± SD; median (range)

Number of patients with risk factorsg

Author, year of publication Subgroups

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

e395

Fig. 2. (a) A significant association (p < 0.00001) between baseline serum creatinine > 1.5 mg/dL and CIN incidence with a pooled odds ratio of 4.10 (95%CI: 2.26–7.42). (b) A significant association (p = 0.03) between eGFR < 60 mL/min or stage ≥ 3 and CIN incidence with a pooled odds ratio of 1.73 (95%CI: 1.06–2.82). (c) A significant association (p = 0.02) between eGFR < 45 mL/min or stage ≥ 3b and CIN incidence with a pooled odds ratio of 2.41 (95%CI: 1.12–5.16). Crit 1: CIN defined as increase in serum creatinine ≥ 25%. Crit 2: CIN defined as increase in serum creatinine ≥ 0.5 mg/dL. Crit 3: CIN defined as increase in serum creatinine of ≥25% or ≥0.5 mg/dL.

Fig. 3. A significant association (p < 0.00001) between diabetes mellitus and CIN incidence with a pooled odds ratio of 1.87 (95%CI: 1.55–2.26). Crit 1: CIN defined as increase in serum creatinine ≥ 25%. Crit 2: CIN defined as increase in serum creatinine ≥ 0.5 mg/dL. Crit 3: CIN defined as increase in serum creatinine of ≥25% or ≥0.5 mg/dL. DM: diabetes mellitus.

3.8.4. Other risk factors The presence of gender (female), hypertension, congestive heart failure and no hydration with NAC showed no significant association with CIN incidence, the odds ratios were 0.82 (95%CI: 0.65, 1.04), p = 0.10, and 1.33 (95%CI: 0.91–1.95) p = 0.13 and 1.43 (95%CI:

0.63–3.23), p = 0.40 and 1.32 (95%CI: 0.51–3.41), p = 0.56 respectively. The I2 values were 10%, 10%, 52% and 67% respectively. Amongst the use of NSAID’s, use of other nephrotoxic drugs, presence of anaemia and presence of malignancy, only the association with NSAID’s and malignancy were significant

e396

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

Fig. 4. A significant association (p = 0.04) between old age (>65 years) and CIN incidence with a pooled odds ratio of 1.95 (95%CI: 1.02–3.70). Crit 1: CIN defined as increase in serum creatinine ≥ 25%. Crit 2: CIN defined as increase in serum creatinine ≥ 0.5 mg/dL. Crit 3: CIN defined as increase in serum creatinine of ≥25% or ≥0.5 mg/dL. Old age defined as >65 years.

(p = 0.04 and p = 0.04). The I2 values were: 0%, 49%, 65% and 0%. All data on these risk factors are presented in Table 3. Data on all other risk factors such as race, vascular disease (coronary, cerebrovascular or peripheral), diuretics, angiotensin converting enzyme inhibitors, angiotensin receptor blocker, calcium antagonists, repeated contrast medium administration or contrast medium volume > 250 mL, hypotension, proteinuria, single kidney or kidney transplant, liver disease, chemotherapy are shown in Appendix 3. No pooled analysis was performed for these factors due to the low number of datasets.

4. Discussion Our results show that the incidence of CIN after intravenous administration of iodinated contrast medium in general is low. The incidence is influenced by the definition used: when CIN is defined as an absolute increase of SCr (≥0.5 mg/dL) the incidence seems to be lower than when defined as a relative increase of SCr (≥25%). The CIN incidence does not seem to differ between the different subgroups of contrast media.

In the systematic review published by Kooiman et al. an incidence of 8.8% (95%CI: 6.2–12.3) was seen in patients with renal insufficiency versus 5.2% (95%CI: 3.3–8.3) in patients without renal insufficiency (p < 0.001) and an incidence of 9.3% (95%CI: 5.6–15.2) in patients with diabetes mellitus versus 3.7% (95%CI: 2.2–6.4) in patients without diabetes mellitus (p < 0.001) [12]. Further, in that meta-analysis several studies evaluating other examinations than CECT were also included as well as intensive care patients (ICU). Critically ill patients show substantial daily fluctuation of SCr with or without administration of iodinated contrast medium, this might influence the incidence of the diagnosis of CIN and might have led to overestimation of CIN incidence [9,17]. Our results show that out of all risk factors mentioned in present guidelines only a few were found to be significantly associated with CIN after intravenous iodinated contrast medium administration for CECT. Namely: renal insufficiency, diabetes and age above 65 years and use of NSAID’s. In addition significant association between CIN incidence and malignancy was found, this is not mentioned in present guidelines. Other risk factors such as hypertension, congestive heart failure, contrast volume, hydration status and anaemia do not appear

Table 3 Pooled data on other risk factors. Risk factor

Heterogeinity

Risk factor present

Risk factor absent

CIN

No. of patients

CIN

No. of patients

Female (n = 12)

I2 = 10%

130

2852

202

3876

Hypertension (n = 5)

I2 = 23%

59

1357

64

1589

Congestive heart failure (n = 6)

I2 = 52%

22

376

150

3013

Without NAC hydration (n = 5)

I2 = 67%

52

414

35

309

NSAID’s (n = 3)

I2 = 0%

13

144

41

1523

Nephrotoxic drugs (n = 3)

I2 = 49%

29

441

37

784

Anaemia (n = 3)

I2 = 65%

34

543

76

1232

Malignancy (n = 4)

I2 = 0%

28

693

54

1289

Odds ratio

0.82 [0.65, 1.04] Test for overall effect: Z = 1.66 (p = 0.10) 1.33 [0.91, 1.95] Test for overall effect: Z = 1.50 (p = 0.13) 1.43 [0.63, 3.23] Test for overall effect: Z = 0.85 (p = 0.40) 1.32 [0.51, 3.41] Test for overall effect: Z = 0.58 (p = 0.56) 2.32 [1.04, 5.19] Test for overall effect: Z = 2.06 (p = 0.04) 1.55 [0.71, 3.37] Test for overall effect: Z = 1.11 (p = 0.27) 1.49 [0.61, 3.62] Test for overall effect: Z = 0.87 (p = 0.38) 1.79 [1.03, 3.11] Test for overall effect: Z = 2.06 (p = 0.04)

Crit 1: CIN defined as increase in serum creatinine ≥ 25%. Crit 2: CIN defined as increase in serum creatinine ≥ 0.5 mg/dL. Crit 3: CIN defined as increase in serum creatinine of ≥25% or ≥0.5 mg/dL.

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

to have a significant association with CIN incidence. To our knowledge this is the most complete meta-analysis up till now including association analyses of most frequent mentioned risk factors in CIN prevention guidelines and specific for patients undergoing CECT. Our results confirm the finding by Kooiman et al. that there seems to be no association between volume of contrast medium and CIN [12]. Our meta-analysis also has some limitations. In 20 papers included in our meta-analysis data were collected in a retrospective manner which could have led to an interpretation bias since other factors contributing to CIN might not be documented. In these retrospective studies patients were only included when the renal function was determined several consecutive days before and after intravenous iodinated contrast medium exposure. This could provide biased information because there might have been other reasons than screening for risk factors for CIN to determine renal function. Also patients with decreased renal function could have been excluded from receiving intravenous iodinated contrast medium by their treating physician. Another factor that could have contributed to an interpretation bias is that almost none of the papers elaborated on other causes of renal insufficiency (e.g. concomitant medication, dehydration) or the natural background fluctuation of serum creatinine. As in any meta-analysis, not all data could be extracted due to missing information. We contacted the corresponding authors of papers that were included to complete the data for renal insufficiency and diabetes mellitus but unfortunately did not succeed. We therefore performed stratified analyses to study associations between risk factors and CIN. Ideally we would like to perform a multivariate regression analysis with CIN incidence as dependent variable and all risk factors as independent variables. However due to the missing data on risk factors this analyses would not represent real relation between CIN and the risk factors in this population. We cannot assume in case of missing data this meant that risk factors were not present. Another well-known limitation is the pooling of heterogeneous studies with regard to selection of patients which make it difficult to interpret results. For instance studies without patients with renal insufficiency or diabetes mellitus at one hand and on the other hand studies containing high risk patients only [18–20]. One might argue that this does not represent a general patient population, however the proportions of the distribution show a normal distribution and we therefore assume that the mean represents the general population. In addition, we performed random effects approaches to summarize our findings, in case the I2 test was >25%. The influence of contrast volume on CIN is difficult to establish and might not be found to be significant due to the fact that the variation of contrast volume (mL) administration was low. However because the concentration of iodine differs between different types of contrast media ideally these data would be presented in grams of iodine but our data set did not permit this. In patient populations where were intra-arterial contrast agent was used CIN has been related to grams of iodine administered [21]. Some studies have shown a significant difference in the incidence of CIN between these different contrast media, especially when risk factors were present [22,23]. However this was not always confirmed in other randomized controlled trials [18,19]. Our pooled data suggest that there is no difference between low-osmolar monomer and iso-osmolar dimer contrast medium and we therefore consider the effect to be small. The effect in some analyses might have been underestimated due to the small number of data e.g. nephrotoxic drugs and anaemia. Also the different definitions of CIN (absolute or relative) could influence the CIN incidence. Thomsen and colleagues showed that

e397

these two definitions of CIN are not interchangeable [23]. Our results confirm this, which could have introduced an interpretation bias since not all papers reported the incidence according to both definitions. Another problem was the methodological assessment of observational studies. There are no proper methodological assessment tools for observational trials. One tool developed for observational trials was designed to optimize reporting instead of methodology [24]. Since good reporting does not necessarily mean good methodology we decided to use the QUADAS-2 tool for these studies studies [15].

5. Conclusion The overall incidence of CIN seems low. The main risk factors that are probably related to CIN in patients undergoing intravenous CECT are pre-existent renal insufficiency, old age, use of NSAID’s and diabetes. Also the presence of malignancy was associated with CIN this could be a result of the co morbidity and the nephrotoxic chemotherapy administered to these patients in combination with intravenous iodinated contrast medium. The other risk factors did not show a significant association with CIN. More certainty could be obtained by evaluating the risk of CIN in large scale studies with well-defined risk factors taking into account other factors such as amount of iodine per contrast dose, eGFR (contrast mediumdose/eGFR ratio).

Funding None.

Conflict of interest None.

Appendix 1. Search strategy in details PUBMED #1 #2 #3 #4 #5 #6 EMBASE #1 #2 #3 #4 #5 #6 Cochrane #1 #2 #3 #4 #5 #6 #7

Contrast induced nephropathy (tw) Contrast-induced nephropathy (tw) CIN (tw) Nephropathy (tw) #3 AND #4 #1 OR #2 OR #5 filters: publication date from 2002/01/01 to 2013/12/31 Contrast induced nephropathy (ti, ab, sh, hw, tn, ot, dm, mf, dv, kw) Contrast-induced nephropathy (ti, ab, sh, hw, tn, ot, dm, mf, dv, kw) CIN (ti, ab, sh, hw, tn, ot, dm, mf, dv, kw) Nephropathy (ti, ab, sh, hw, tn, ot, dm, mf, dv, kw) #3 AND #4 #1 OR #2 OR #5 filters: publication date from 2002/01/01 to 2013/12/31 Contrast-induced nephropathy (ti, ab, kw) Contrast induced nephropathy (ti, ab, kw) CIN Nephropathy #3 AND #4 (#1 OR #2 OR #5) #6 from 2002 to 2012

5090 935 7843 421,247 604 2843

1847 1849 8315 59,387 793 1847

186 229 509 3294 103 231 202

e398

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

Appendix 2. Included papers Included for data analyses (n = 42) 1. Abujudeh H, Gee M, Kaewlai R. In emergency situations, should serum creatinine be checked in all patients before performing second contrast CT examinations within 24 hours? Journal of the American College of Radiology 2009;6:268–73. 2. Balemans C, Reichert L, Van Schelven B, et al. Epidemiology of contrast material-induced nephropathy in the era of hydration. Radiology 2012;263:706–13. 3. Barrett B, Katzberg R, Thomsen H, et al. Contrast-induced nephropathy in patients with chronic kidney disease undergoing computed tomography: a double-blind comparison of iodixanol and iopamidol. Investigative Radiology 2006;41(11):815–21. 4. Becker C, Reiser M. Use of iso-osmolar nonionic dimeric contrast media in multidetector row computed tomography angiography for patients with renal impairment. Investigative Radiology 2005;40(10):672–5. 5. Bruce R, Djamali A, Shinki K, et al. Background fluctuation of kidney function versus contrast-induced nephrotoxicity. American Journal of Roentgenology 2009;192(3):711–8. 6. Cheruvu B, Henning K, Mulligan J, et al. Iodixanol: risk of subsequent contrast nephropathy in cancer patients with underlying renal insufficiency undergoing diagnostic computed tomography examinations. Journal of Computer Assisted Tomography 2007;31(4):493–8. 7. Dillman J, Al-Hawary M, Ellis J, et al. Comparative investigation of i.v. iohexol and iopamidol: effect on renal function in low-risk outpatients undergoing CT. American Journal of Roentgenology 2012;198(2):392–7. 8. Dittrich R, Akdeniz S, Kloska S, et al. Low rate of contrast-induced nephropathy after CT perfusion and CT angiography in acute stroke patients. Journal of Neurology 2007;254(11):1491–7. 9. El-Hajjar M, Bashir I, Khan M, et al. Incidence of contrast-induced nephropathy in patients with chronic renal insufficiency undergoing multidetector computed tomographic angiography treated with preventive measures. American Journal of Cardiology 2008;102:353–6 [0002-9149 (print)]. 10. Garcia-Ruiz C, Martinez-Vea A, Sempere T, et al. Low risk of contrast nephropathy in high-risk patients undergoing spiral computed tomography angiography with the contrast medium iopromide and prophylactic oral hydration. Clinical Nephrology 2004;61(3):170–6. 11. Hipp A, Desai S, Lopez C, et al. The incidence of contrast-induced nephropathy in trauma patients. European Journal of Emergency Medicine: Official Journal of the European Society for Emergency Medicine 2008;15(3):134–9. 12. Holmquist F, Nyman U. Eighty-peak kilovoltage 16-channel multidetector computed tomography and reduced contrast-medium doses tailored to body weight to diagnose pulmonary embolism in azotaemic patients. European Radiology 2006;16(5):1165–76. 13. Hopyan J, Gladstone D, Mallia G, et al. Renal safety of CT angiography and perfusion imaging in the emergency evaluation of acute stroke. American Journal of Neuroradiology 2008;29(10):1826–30. 14. Hsu T, Huang M, Yu S, et al. N-acetylcysteine for the prevention of contrast-induced nephropathy in the emergency department. Internal Medicine 2012;51(19):2709–14. 15 Josephson S, Dillon W, Smith W. Incidence of contrast nephropathy from cerebral CT angiography and CT perfusion imaging. Neurology 2005;64(10):1805–6. 16. Karlsberg R, Dohad S, Sheng R. Contrast medium-induced acute kidney injury: comparison of intravenous and intraarterial administration of iodinated contrast medium. Journal of Vascular and Interventional Radiology 2011;22(8):1159–65. 17. Khalili H, Dashti-Khavidaki S, Tabifar H, et al. Evaluating the efficacy of single daily dose of 1200 mg of N-acetyl-cysteine in preventing contrast agent-associated nephrotoxicity. The Internet Journal of Internal Medicine 2006;6(1). 18. Kim K, Kim K, Hwang S, et al. Risk stratification nomogram for nephropathy after abdominal contrast-enhanced computed tomography. The American Journal of Emergency Medicine 2011;29(4):412–7. 19. Kim S, Cha R, Lee J, et al. Incidence and outcomes of contrast-induced nephropathy after computed tomography in patients with CKD: a quality improvement report. American Journal of Kidney Diseases 2010;55(6):1018–25 [1523-6838 (electronic)]. 20. Kingma H, Algra P, Tegelaers F, et al. Effect van een protocol op incidentie van contrastmiddel- geïnduceerde nefropathie bij CT-scans van poliklinische patiënten. wetenschappelijk platform 2010;4(7/8):1–4. 21. Kooiman J, Klok F a, Mos I, et al. Incidence and predictors of contrast-induced nephropathy following CT-angiography for clinically suspected acute pulmonary embolism. Journal of Thrombosis and Haemostasis 2010;8(2):409–11.

22. Krol A, Dzialowski I, Roy J, et al. Incidence of radiocontrast nephropathy in patients undergoing acute stroke computed tomography angiography. Stroke: A Journal of Cerebral Circulation 2007;38(8):2364–6. 23. Kuhn M, Chen N, Sahani D, et al. The PREDICT study: a randomized double-blind comparison of contrast-induced nephropathy after low- or isoosmolar contrast agent exposure. American Journal of Roentgenology 2008;191(1):151–7. 24. Kulvatunyou N, Rhee P, Carter S, et al. Defining incidence and outcome of contrast-induced nephropathy among trauma: is it overhyped? The American Surgeon 2011;77(6):686–9. 25. Langner S, Stumpe S, Kirsch M, et al. No increased risk for contrast-induced nephropathy after multiple CT perfusion studies of the brain with a nonionic, dimeric, iso-osmolal contrast medium. American Journal of Neuroradiology 2008;29(8):1525–9. 26. Lencioni R, Fattori R, Morana G, et al. Contrast-induced nephropathy in patients undergoing computed tomography (CONNECT) – a clinical problem in daily practice? A multicenter observational study. Acta Radiologica (Stockholm, Sweden: 1987) 2010;51(7):741–50. 27. Lima F, Lev M, Levy R, et al. Functional contrast-enhanced CT for evaluation of acute ischemic stroke does not increase the risk of contrast-induced nephropathy. American Journal of Neuroradiology 2010;31(5):817–21. 28. Lodhia N. Risk of contrast-induced nephropathy in hospitalized patients with cirrhosis. World Journal of Gastroenterology 2009;15(12):1459. 29. Lufft V, Hoogestraat-Lufft L, Fels L, et al. Contrast media nephropathy: intravenous CT angiography versus intraarterial digital subtraction angiography in renal artery stenosis: a prospective randomized trial. American Journal of Kidney Diseases: The Official Journal of the National Kidney Foundation 2002;40(2):236–42. 30. Matsushima K, Peng M, Schaefer E, et al. Posttraumatic contrast-induced acute kidney injury: minimal consequences or significant threat? The Journal of Trauma Injury, Infection and Critical Care 2011;70(2):415–9 [discussion 419–20]. 31. Mitchell A, Jones A, Tumlin J, et al. Incidence of contrast-induced nephropathy after contrast-enhanced computed tomography in the outpatient setting. Clinical Journal of the Amercan Society of Nephrology 2010;5:4–9 [1555-905X (electronic)]. 32. Murakami R, Hayashi H, Sugizaki K, et al. Contrast-induced nephropathy in patients with renal insufficiency undergoing contrast-enhanced MDCT. European Radiology. 2012;22(10): 2147–52. 33. Najjar M, Hamad A, Salameh M, et al. The risk of radiocontrast nephropathy in patients with cirrhosis. Renal Failure 2002;24(1):11–8. 34. Nguyen U, Suranyi S, Ravenel P, et al. Iso-osmolality versus low-osmolality iodinated contrast medium at intravenous contrast enhanced CT: effect on kidney function. Radiology 2008;248(1):97–105. 35. Pahade J, LeBedis C, Raptopoulos V, et al. Incidence of contrast-induced nephropathy in patients with multiple myeloma undergoing contrast-enhanced CT. American Journal of Roentgenology 2011;196(5):1094–101. 36 Poletti P, Saudan P, Platon A, et al. I.v. N-acetylcysteine and emergency CT: use of serum creatinine and cystatin C as markers of radiocontrast nephrotoxicity. American Journal of Roentgenology 2007;189:687–92 [1546-3141 (electronic)]. 37. Sancak A, Derici U, Arinsoy T, et al. Effects of contrast media on endothelin and netric oxide system after computed tomography. Gazi Medical Journal 2002;13:81–5. 38. Sandstede J, Roth A, Machann W, et al. Evaluation of the nephrotoxicity of iodixanol in patients with predisposing factors to contrast medium induced nephropathy referred for contrast enhanced computed tomography. European Journal of Radiology 2007;63(1):120–3. 39. Thomsen H, Morcos S, Erley C, et al. The ACTIVE trial: comparison of the effects on renal function of iomeprol-400 and iodixanol-320 in patients with chronic kidney disease undergoing abdominal computed tomography. Investigative Radiology 2008;43(3):170–8. 40. Tremblay L, Tien H, Hamilton P, et al. Risk and benefit of intravenous contrast in trauma patients with an elevated serum creatinine. The Journal of Trauma: Injury, Infection, and Critical Care 2005;(November):1162–7. 41. Weisbord S, Mor M, Resnick A, et al. Incidence and outcomes of contrast-induced AKI following computed tomography. Clinical Journal of the American Society of Nephrology 2008;3(5):1274–81. 42. Yoshikawa D, Isobe S, Sato K, et al. Importance of oral fluid intake after coronary computed tomography angiography: an observational study. EuropeanJournal of Radiology 2009;77(2011):118–122 [1872-7727 (electronic)].

S.I. Moos et al. / European Journal of Radiology 82 (2013) e387–e399

e399

Appendix 3. Other risk factors and CIN incidence Risk factor

Race (n = 2)a Claudication (n = 2) Coronary disease (n = 2)a Cerebrovascular disease (n = 1) Vascular (either coronary, cerebro or peripheral)(n = 1) Diuretics (n = 2) ACE inhibitors (n = 2) Angiotensin receptor blocker (n = 1) Calcium antagonist (n = 1) Repeat contrast or contrast > 250 mL (n = 2) Hypotension (n = 2) Proteinuria (n = 2) Single kidney or kidney transplant (n = 2) Liver disease (n = 1) Chemotherapy (n = 1)

Study

Mitchell (crit 3) (2010) Abujudeh (crit 1) (2009) Balemans (crit 1) (2012) Kooiman (crit 3) (2010) Balemans (crit 1) (2012) Kooiman (crit 3) (2010) Kim (crit 3) (2010) Mitchell (crit 3) (2010) Balemans (crit 1) (2012) Kim (crit 3) (2010) Balemans (crit 1) (2012) Kim (crit 3) (2010) Balemans (crit 1) (2012) Balemans (crit 1) (2012) Balemans (crit 1) (2012) Matsushima (crit 3) (2011) Dittrich (crit 3) (2007) Matsushima (crit 3) (2011) Bruce (crit 2) (2009) Pahade (crit 3) (2011) El-Hajjar (crit 2) (2008) Kim (crit 3) (2010) Kim (crit 3) (2010) Balemans (crit 1) (2012)

Risk factor present

Risk factor absent

CIN

No. of patients

CIN

33 3 10 0 3 2 0 13 12 2 7 9 3 3 16 21 0 7 7 0 0 0 1 3

303 38 273 15 246 26 55 63 362 32 229 156 200 159 429 315 7 42 72 17 5 15 79 120

37 18 13 21 20 19 13 57 11 11 16 4 20 20 7 57 3 71 245 1 7 13 12 20

Odds ratio

No. of patients 330 126 671 216 698 225 465 570 582 488 715 355 744 785 515 869 155 1142 5718 30 395 505 440 824

0.97 [0.59, 1.59] 0.51 [0.14, 1.85] 1.92 [0.83, 4.44] 0.29 [0.02, 5.08] 0.42 [0.12, 1.42] 0.90 [0.20, 4.12] 0.30 [0.02, 5.15] 2.34 [1.20, 4.57] 1.78 [0.78, 4.08] 2.89 [0.61, 13.64] 1.38 [0.56, 3.39] 5.37 [1.63, 17.72] 0.55 [0.16, 1.87] 0.74 [0.22, 2.51] 2.81 [1.15, 6.90] 1.02 [0.61, 1.71] 2.90 [0.14, 61.49] 3.02 [1.29, 7.03] 2.41 [1.09, 5.30] 0.56 [0.02, 14.56] 4.71 [0.24, 93.07] 1.18 [0.07, 20.71] 0.46 [0.06, 3.57] 1.03 [0.30, 3.52]

Crit 1: CIN defined as increase in serum creatinine ≥ 25%. Crit 2: CIN defined as increase in serum creatinine ≥ 0.5 mg/dL. Crit 3: CIN defined as increase in serum creatinine of ≥25% or ≥0.5 mg/dL. a Race = Afro-American as risk factor.

References [1] Nash K, Hafeez A, Hou S. Hospital-acquired renal insufficiency. American Journal of Kidney Diseases: The Official Journal of the National Kidney Foundation 2002;39(5):930–6. [2] Rihal C. Incidence and prognostic importance of acute renal failure after percutaneous coronary intervention. Circulation 2002;105(19):2259–64. [3] Stacul F, Van der Molen A, Reimer P, et al. Contrast induced nephropathy: updated ESUR contrast media safety committee guidelines. European Radiology 2011;21(12):2527–41. [4] Nederlandse Vereniging Voor Radiologie. Voorzorgsmaatregelen bij jodiumhoudende contrastmiddelen. CBO richtlijnen; 2007. [5] Benko A, Fraser-hill M, Magner P, et al. Canadian association of radiologists: consensus guidelines for the prevention of contrast-induced nephropathy. Journal l’Association canadienne des radiologistes 2007;58(2):79–88. [6] American College of Radiology. American College of Radiology Manual on Contrast Media Version 8 ACR Manual on Contrast Media; 2012. [7] Katzberg R, Newhouse J. Intravenous contrast medium-induced nephrotoxicity: is the medical risk really as great as we have come to believe? Radiology 2010;256(1):21–8. [8] Baumgarten D, Ellis J. Contrast-induced nephropathy: contrast material not required? American Journal of Roentgenology 2008;191(2):383–6. [9] Bruce R, Djamali A, Shinki K, et al. Background fluctuation of kidney function versus contrast-induced nephrotoxicity. American Journal of Roentgenology 2009;192(3):711–8. [10] Katzberg R, Lamba R. Contrast-induced nephropathy after intravenous administration: fact or fiction? Radiologic Clinics of North America 2009;47(5):789–800. [11] Nyman U, Almén T, Jacobsson B, et al. Are intravenous injections of contrast media really less nephrotoxic than intra-arterial injections? European Radiology 2012;22(6):1366–71. [12] Kooiman J, Pasha S, Zondag W, et al. Meta-analysis: serum creatinine changes following contrast enhanced CT imaging. European Journal of Radiology 2011;81(10):2554–61. [13] Liberati A, Altman D, Tetzlaff J, et al. The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate health care interventions: explanation and elaboration. PLoS Medicine 2009;6(7):e1000100.

[14] Verhagen A, De Vet H, De Bie R, et al. The Delphi list: a criteria list for quality assessment of randomized clinical trials for conducting systematic reviews developed by Delphi consensus. Journal of Clinical Epidemiology 1998;51(12):1235–41. [15] Whiting P, Rutjes A, Westwood M, et al. QUADAS-2: a revised tool for the quality assesment of diagnositc accuracy studies. Annals of Emergency Medicine 2011;155(4):529–36. [16] Higgins J, Thompson S, Deeks J, et al. Measuring inconsistency in meta-analyses. British Medical Journal (Clinical Research ed) 2003;327(7414):557–60. [17] Cely C, Schein R, Quartin A. Risk of contrast induced nephropathy in the critically ill: a prospective, case matched study. Critical Care (London, England) 2012;16(2):R67. [18] Kuhn M, Chen N, Sahani D, et al. The PREDICT study: a randomized doubleblind comparison of contrast-induced nephropathy after low- or isoosmolar contrast agent exposure. American Journal of Roentgenology 2008;191(1): 151–7. [19] Barrett B, Katzberg R, Thomsen H, et al. Contrast-induced nephropathy in patients with chronic kidney disease undergoing computed tomography: a double-blind comparison of iodixanol and iopamidol. Investigative Radiology 2006;41:815–21 [020-9996 (print)]. [20] Dillman J, Al-Hawary M, Ellis J, et al. Comparative investigation of i.v. iohexol and iopamidol: effect on renal function in low-risk outpatients undergoing CT. American Journal of Roentgenology 2012;198(2):392–7. [21] Gurm H, Dixon S, Smith D, et al. Renal function-based contrast dosing to define safe limits of radiographic contrast media in patients undergoing percutaneous coronary interventions. Journal of the American College of Cardiology 2011;58(9):907–14. [22] Nguyen U, Suranyi S, Ravenel P, et al. Iso-osmolality versus low-osmolality iodinated contrast medium at intravenous contrast-enhanced CT: effect on kidney function. Radiology 2008;248(1):97–105. [23] Thomsen H, Morcos S, Erley C, et al. The ACTIVE trial: comparison of the effects on renal function of iomeprol-400 and iodixanol-320 in patients with chronic kidney disease undergoing abdominal computed tomography. Investigative Radiology 2008;43(3):170–8. [24] Von Elm E, Altman D, Egger M, et al. The strengthening the reporting of observational studies in epidemiology (STROBE) statement: guidelines for reporting ˜ de salud pública 2007;82(3):251–9. observational studies. Revista espanola