Percutaneous Coronary Intervention During the Shortage of Iodinated Contrast

© 2024 HMP Global. All Rights Reserved.
Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of the Journal of Invasive Cardiology or HMP Global, their employees, and affiliates. 

Watch the accompanying author interview.

Objectives: A recent coronavirus-related factory shutdown led to a global shortage of iodinated contrast. The authors evaluated how the contrast shortage impacted percutaneous coronary interventions (PCI).

Methods: Using a statewide database incorporating CathPCI registry data from 19 hospitals, the authors evaluated 2 time periods: pre-shortage (May 1, 2021-April 30, 2022) and during the shortage (May 1, 2022-October 31, 2022). They compared procedure volumes, patient and procedure characteristics, and short-term outcomes, including acute kidney injury (AKI). Of primary interest was the difference in contrast volume per PCI and the incidence of AKI between periods.

Results: There were 8980 patients treated pre-shortage and 4046 during the shortage. Procedure volumes per hospital remained similar, as did patient characteristics. Multivessel procedures declined during the shortage (45.3% vs 42.8%, P = .007). There was a significant decline in contrast per procedure (149.9 ± 68.1 mL to 137.5 ± 62.4 mL per case, P < .0001) that began at the start of the shortage and continued throughout. There were no differences in patient outcomes, including AKI (7.9% vs 7.4%, P = .40), between study periods. When limited to patients at increased risk of AKI, there remained no difference in AKI between the study periods, despite a similar decrease in contrast volume in that cohort. Multivariable analysis showed a strong correlation between baseline risk of AKI and subsequent AKI (P < .0001), but no impact of procedure characteristics or time period.

Conclusions: The global shortage of iodinated contrast led to a significant decline in contrast use during PCI, with no impact on patient outcomes.

In addition to its direct impact on patient health, the coronavirus pandemic also caused shortages of medical supplies. Some of the shortages can be attributed to the sudden increase in demand for certain types of material, such as personal protective equipment, but other shortages are due to disruptions in the supply chain, either due to lack of workers, site shutdowns, or delivery problems.¹ A recent shutdown of a GE Healthcare factory in Shanghai, China, which produced the majority of GE Healthcare’s iodinated contrast media, led to a global shortage that lasted many months.² Iodinated contrast is used in many types of imaging, including coronary angiography, that are necessary for percutaneous coronary intervention (PCI). Whether and how the global contrast shortage impacted PCI procedures, specifically the number and type of procedures performed, has not been fully elucidated.

While required for essentially all PCIs, iodinated contrast can also cause harm, most frequently in the form of contrast-induced acute kidney injury (AKI), with the risk of AKI rising in correlation with increasing contrast volume.³ We hypothesized that if the shortage of iodinated contrast media led to less contrast use per procedure, it may have impacted the subsequent occurrence of AKI. The purpose of the present study was therefore twofold: (1) to evaluate the impact of the global iodinated contrast shortage on PCI procedures, including procedure volumes and contrast usage; and (2) to evaluate whether those changes impacted patient outcomes.

The Virginia Cardiac Services Quality Initiative (VCSQI) is a statewide consortium representing 25 hospitals in the state of Virginia. National Cardiovascular Data Registry (NCDR) CathPCI data is prospectively collected at each institution, 19 of which pool their data for analyses. Since the consortium data is deidentified, with the removal of all Health Insurance Portability and Accountability Act patient identifiers, the current study is exempt from Institutional Review Board review. Informed consent was not required by the authors’ institutions, as patient information was already deidentified.

All participating hospitals were queried regarding their experience of the contrast shortage and asked to answer 2 brief questions to identify their primary vendor for iodinated contrast and confirm whether they had taken any steps to reduce contrast use during the shortage. Although our initial intention was to compare sites that did vs did not change behaviors during the shortage (assuming only those supplied by GE Healthcare would experience the shortage), because every site reported taking at least some steps to reduce contrast use during the shortage, we chose to include all the sites in 1 shortage cohort.

Patient selection

All patients undergoing PCI at participating hospitals were included in the analysis, which included patient characteristics, procedural characteristics, and patient outcomes. For the assessment of AKI, we excluded patients missing pre- or post-procedure creatinine, those on dialysis at baseline, and those with multiple interventions during the same hospitalization. For patients with more than 1 PCI during separate hospitalizations, only the first PCI was included.

Study definitions and endpoints

We divided the patients into 2 time periods, before and during the contrast shortage, using May 1, 2022 as the date the shortage began; this date was based on the timing of when GE Healthcare informed its customers of the shortage (mid-April 2022). We used October 31, 2022 as the end date for the shortage, because supply had increased substantially by that point. The pre-shortage period was the year leading up to the shortage (May 1, 2021-April 30, 2022). Our co-primary outcomes of interest were the average volume of iodinated contrast used per PCI procedure during the 2 time periods and the incidence of AKI during those time periods, with an a priori plan for subgroup analyses focused on patients at increased risk of AKI. Secondary analyses included patient and procedure characteristics, and other PCI outcomes.

AKI was defined as an at least 50% increase in serum creatinine or an absolute increase of at least 0.3 mg/dL within 48 hours after PCI. We identified patients at an increased risk of AKI by using a model recently published by Mehran et al using a cut point of moderate or greater risk.³ All other outcomes were as previously defined.⁴

Statistical plan

Patient demographics, baseline characteristics, procedural characteristics, procedural outcomes, and renal outcomes were compared between patients treated during the pre-shortage period and those treated during the shortage period (as defined above). Baseline characteristics and outcomes were summarized with means and SDs or medians and interquartile range for continuous measures, and with proportions for categorical variables. Between the study groups, variables were compared using the Student’s t-test or Mann-Whitney test for continuous measures and the chi-squared or Fisher’s exact test for categorical variables. Nonlinear change in contrast volume over time was modeled with cubic spline analysis. A multivariable logistic regression model was used to examine the relationship between the shortage and AKI. The variables included in the model — pre-procedure AKI risk, contrast volume, slow or no flow post-procedure, and complex anatomy — were selected based on the study mentioned above.³ All P-values were 2-tailed and a P-value of less than 0.05 was considered significant for all analyses. Statistical analysis was performed using SAS version 9.4 (SAS Institute).

A total of 13 026 patients were included in the analysis: 8980 were treated in the year prior to the contrast shortage and 4046 during the contrast shortage. Overall, the average number of PCIs performed per month at each hospital did not change during the shortage (Figure 1). Patient characteristics are shown in Table 1. Overall, the cohorts were very similar. There was a slight increase in the percentage of patients at moderate or greater risk of AKI during the contrast shortage.

Procedure characteristics are shown in Table 2. There was a slight decrease in procedure length during the contrast shortage (64.5 ± 40.4 vs 62.3 ± 41.7 minutes, P = .004) and multivessel PCI became slightly less common (45.3% vs 42.8%, P = .007). The utilization of intravascular imaging (intravascular ultrasound [IVUS] or optical coherence tomography) did not change.

Contrast use declined significantly during the shortage, from 149.9 ± 68.1 mL per case to 137.5 ± 62.4 mL per case (P < .0001; Table 2, Figure 2A). This was reflected in an increase in the percentage of procedures performed with less than 100 mL of contrast, and a decline in those using 200 to 299 mL and 300 mL or more (P < .001 for all comparisons). The change was abrupt and non-linear, as demonstrated by cubic spline analysis (Figure 2B). In addition, pairwise comparisons of mean contrast use showed no differences between months within each time period, but did demonstrate an abrupt and significant decrease between the last month of the pre-shortage period (April 2022) and the first month of the shortage period (May 2022) (Supplemental Table). The reduction in average contrast volume was similar in patients at an increased risk of AKI (146.3 ± 66.3 mL to 136.9 ± 61.3 mL [P < .0001]).

There were 4974 patients in whom AKI could not be assessed (Supplemental Figure), leaving 5500 patients in the pre-shortage cohort and 2552 in the shortage cohort in whom it could be assessed. Kidney injury outcomes are shown in Table 3. There was a numerical decline in the rate of any AKI from the pre-shortage to shortage period, which did not reach statistical significance (7.9% vs 7.4%, P = .40). This was consistent in the subgroup of patients at increased risk of AKI (9.7% pre-shortage vs 9.6% during the shortage, P = .89). In the entire cohort, there was a trend toward fewer cases of Stage 3 AKI (3.7% vs 2.8%, P = .054) and a numerical decline in the proportion of patients with new requirements for dialysis that did not reach statistical significance (0.5% vs 0.3%, P = .16). When limited to those patients at moderate or greater AKI risk, there were no differences in any individual AKI outcomes.

Treatment during the shortage period was not associated with decreased risk of AKI in a univariate model or following adjustment for pre-procedure risk and procedural variables including contrast volume, slow flow or no flow post-procedure, or complex anatomy. In fact, the only variable significantly associated with increased risk of AKI in the fully adjusted model was a pre-procedure AKI risk of moderate or greater (Table 4). Other patient outcomes, for the entire cohort, are shown in Table 5. There was no change in any of the outcomes measured.

We analyzed statewide CathPCI Registry data to evaluate the impact of the global iodinated contrast shortage on PCIs. Our study has several noteworthy findings: (1) the number of procedures did not change during the contrast shortage, nor did the types of patients undergoing PCI; (2) there was a significant reduction in contrast use per procedure during the shortage; and (3) the reduction in contrast use did not result in a detectable decrease in the incidence of AKI, either in the entire cohort or in the subgroup of patients at increased AKI risk.

The shutdown of the GE Healthcare factory in Shanghai that produces iodinated contrast (iohexol) touched off a global shortage, as other factories and vendors were not able to ramp production up sufficiently to meet the acute drop in supply. This was one of many supply chain shocks to result from the coronavirus pandemic.¹ Statements from the American College of Radiology and the Society for Cardiovascular Angiography and Interventions recommended strategies to reduce contrast waste, use alternative studies where possible, and reduce intraprocedural contrast.^2,5 We did not find clear evidence of a reduction in PCI volume during the contrast shortage, although there was a reduction in multivessel interventions, possibly suggesting a desire to limit contrast use.

During the contrast shortage, there was an average reduction of over 12 mL of contrast per PCI, a significant reduction but smaller than what has been accomplished in dedicated trials. In a meta-analysis of 10 studies using an automated contrast injection device, the average reduction in contrast per case was 45 mL.⁶ An interventional approach incorporating automated injectors was recommended during the shortage,⁵ but because that data is not routinely collected as part of CathPCI, we were unable to determine its use, either before or during the shortage. A device designed to divert contrast during manual injections showed a reduction of 26 mL of contrast per case in a small, randomized trial of patients undergoing diagnostic angiography⁷ and 31 mL in a larger randomized study of patients experiencing acute coronary syndrome.⁸ While such a device reduces contrast delivered to the patient, the reduced contrast is diverted into a waste syringe, so it would not have been expected to reduce overall contrast use during the shortage.

A heavier reliance on invasive imaging techniques, such as IVUS, could limit the need for contrast. A small, randomized trial comparing a prescribed reliance on IVUS to routine contrast use demonstrated a reduction of approximately 45 mL of contrast per PCI.⁹ Perhaps surprisingly, we did not find any increased reliance on IVUS during the contrast shortage. The routine use of invasive coronary physiology may also reduce contrast use. Compared with usual care, a strategy of routine fractional flow reserve assessment in patients with multivessel disease resulted in an average reduction of 30 mL of contrast per procedure.¹⁰ A similar strategy may have been used during the shortage, which might explain the decline in multivessel interventions. We also noted a reduction in procedure time, yet similar use of radiation. It is possible that physicians relied more heavily on cineangiograms, rather than fluoroscopy, to make better use of the contrast that was used.

Despite a significant reduction in the average contrast volume during PCI, we did not see a decrease in the incidence of AKI. This was true despite the finding that contrast dose was similarly reduced in patients at increased risk of AKI, meaning the intervention impacted those most likely to benefit. One possible explanation is that the reduction in contrast was not sufficient to impact renal outcomes. In the automated injector meta-analysis, there was a 15% reduction in AKI in response to the 45 mL contrast reduction.⁶ The diversion device resulted in an absolute reduction in AKI of 11%, from 19% to 8% with a 31 mL drop in contrast use.⁸ But these results from a small sample size should be interpreted with caution, considering the very high incidence of AKI in the control group. For comparison, even in our population at moderate or greater risk, the incidence of AKI was less than 10%. While our results showed only a modest reduction in overall contrast use compared to these dedicated studies, it does not mean that larger reductions are impossible, or even that they will require novel therapies. One center reported its experience of aggressive contrast use reduction in response to the shortage and showed a reduction of over 50 mL per case spanning a variety of cardiovascular procedures.¹¹

The lack of clinical impact of the significant reduction in contrast in our study highlights the surprisingly minor role of contrast in AKI. In an earlier study of CathPCI data that included over 662 000 patients, the contrast dose in those who developed AKI was less than 9 mL higher than the cohort who did not.¹² In a recent analysis by Mehran et al, the predictive model including only pre-procedural variables was nearly the same in terms of predictive ability as one that incorporated procedural characteristics, including contrast dose.³ In our analysis, we found a univariate correlation between 2 procedural variables, complex anatomy and reduced flow at the end of the procedure, and subsequent AKI (the contrast dose did not correlate). However, after controlling for baseline AKI risk using the Mehran et al pre-procedural model, both of those procedural variables were no longer significant. In a single-center study of over 2000 patients with myocardial infarction, whether or not a patient underwent coronary angiography (and thus received contrast) had no correlation with subsequent AKI after controlling for other variables.¹³ Despite its nephrotoxicity, advances in contrast formulation and strategies to reduce its use over many years appear to have resulted in contrast playing only a small role in AKI.

Limitations

Our study has limitations. It represents the experience of 19 hospitals across 1 state and may not be generalizable to other geographies. We did not inventory the individual sites to determine which contrast mitigation strategies were used, both in terms of case selection and case performance. While contrast use clearly declined during the shortage period, we cannot know which interventions were the most beneficial. The intensity of implementation at every site was also likely variable, as many sites had access to contrast from other vendors or may have had larger inventories of contrast when the shortage began. AKI could only be assessed in about 60% of patients, primarily due to missing post-procedure creatinine. This is not uncommon in registry studies, as same-day discharge (~30% of our population) precludes post-procedure lab draws. Lastly, we are limited by the inherent difficulties of retrospectively analyzing registry data, even though CathPCI is a highly regarded, quality-controlled database.

During the global iodinated contrast shortage, there was a reduction in the average contrast use during PCIs. Despite the reduction, there was no change in the occurrence of AKI.

Zachary M. Gertz, MD;¹ Brian K. Mitchell, MD;¹ Michael C. Kontos, MD;¹ Nicholas R. Teman, MD;² Anthony Norman, MD;² Raza Ahmad, MD;² Raymond J. Strobel, MD;² Abdulla A. Damluji, MD, PhD;² Robert A. Shor, MD;³ Alan Speir, MD;³ Mohammed A. Qauder, MD⁴

From the ¹Division of Cardiology, Department of Medicine, Virginia Commonwealth University, Richmond, Virginia; ²Division of Cardiac Surgery, Department of Surgery, University of Virginia, Charlottesville, Virginia; ³Heart and Vascular Institute, Inova Fairfax Hospital, Fairfax, Virginia; ⁴Division of Cardiothoracic Surgery, Department of Surgery, Virginia Commonwealth University, Richmond, Virginia.

Disclosures: The authors report no financial relationships or conflicts of interest regarding the content herein.

Acknowledgments: The authors thank Maria Alu, MS, for assistance with manuscript preparation.

Address for correspondence: Zachary Gertz, MD, VCU Medical Center Main Hospital, 1250 E. Marshall Street, Richmond, VA 23219, USA. Email: zachary.gertz@vcuhealth.org; X: @VCUHealthHeart

Back to May 2025 Table of Contents