Strategies for assessing renal function prior to outpatient contrast-enhanced CT: a UK survey

Prevention of post-contrast kidney injury in patients with cancer

Post-contrast acute kidney injury is defined as a nephropathy with an increase in serum creatinine of >0.3 mg/dL (or >26.5 μmol/L) or >1.5-times the baseline within 48-72 h of intravascular administration of a contrast medium. Patients with cancer have an increased risk of post-contrast acute kidney injury not only related to the frequent use of contrast medium for computed tomography scans but also to other factors, including the type of tumour, age, oncological therapies, use of other nephrotoxic agents and dehydration. Preventive strategies were developed and may be applied to different risk profiles. Patients at risk may be detected by recently published risk scores.

Should the Radiologist Always Request a Blood Test Before an Emergency CT Scan in Children

Renal failure is relatively common in children presenting to the emergency department, suggesting that the assumption of normal renal function is not always valid. Although some computed tomography (CT) scans necessitate the use of intravenous contrast, one should probably consider whether a blood test is necessary to assess the patient's renal function and possibly consider other imaging modalities before proceeding. With no pediatric-specific guidelines and no validated pediatric prevention strategies, further research is needed to establish clear recommendations for contrast-enhanced exams in stable and unstable pediatric patients with unknown renal function.

Evaluation of the Accuracy of Cr and BUN Using the ABL90 FLEX PLUS Blood Gas Analyzer and the Equivalence of Candidate Specimens for Assessment of Renal Function

BACKGROUND

The ABL90 FLEX PLUS (Radiometer) is a blood gas analyzer that also provides creatinine (Cr) and blood urea nitrogen (BUN) results. We assessed the accuracy of the ABL90 FLEX PLUS to measure Cr and BUN and find suitable candidate specimens against primary specimens (heparinized whole-blood (H-WB)).

METHODS

Paired H-WB, serum, and sodium-citrated whole-blood (C-WB) samples (105) were collected. The Cr and BUN levels in the H-WB using the ABL90 FLEX PLUS were compared with those of the serum using four automated chemistry analyzers. The suitability of the candidate specimens was assessed at each medical decision level according to the CLSI guideline EP35-ED1.

RESULTS

The respective mean differences of the ABL90 FLEX PLUS for the Cr and BUN were below -0.10 and -3.51 mg/dL compared to the other analyzers. The systematic differences between the serum and the H-WB at the low, medium, and high medical decision levels were all 0% for Cr, but those of the C-WB were -12.96%, -11.81%, and -11.30%, respectively. Regarding imprecision, the SD_serum/SD_H-WB ratios at each level were 0.14, 1.41, and 0.68, whereas the SD_C-WB/SD_H-WB ratios were 0.35, 2.00, and 0.73, respectively.

CONCLUSIONS

The ABL90 FLEX PLUS provided Cr and BUN results comparable with the four widely used analyzers. Among the candidates, the serum was suitable for Cr testing using the ABL90 FLEX PLUS, while the C-WB did not satisfy the acceptance criteria.

Radiology departmental policy compliance with Swedish guidelines regarding post-contrast acute kidney injury for examinations with iodinated contrast media

INTRODUCTION

Guidelines concerning intravenous iodinated contrast media (CM) during computed tomography (CT) examinations are important to follow to minimize the risk for post-contrast acute kidney injury (PC-AKI). The purpose of this study was to investigate the radiology departmental policy compliance with Swedish guidelines concerning PC-AKI.

METHODS

In February 2020, an electronic survey was distributed to the responsible radiographer at 41 radiology departments in all university hospitals and medium-sized hospitals in Sweden. The questions focused on routines around renal functional tests, individualized contrast administration and handling of patients with diabetes mellitus taking metformin.

RESULTS

The response rate was 83%. Seventy-six percent (n = 26) of radiology departments calculated estimated glomerular filtration rate (eGFR) from serum creatinine prior to CM administration, but only 24% (n = 8) followed the recommendation to calculate eGFR from both serum creatinine and cystatin C. For acute/inpatients, 55% (n = 18) followed the recommendation that renal functional tests should be performed within 12 h before CM administration. For elective patients, 97% (n = 33) followed the recommendation to have eGFR newer than three months which is acceptable for patients with no history of disease that may have affected renal function. Approximately 80% of the radiology departments followed the recommendation that CM dose always should be individually adjusted to patient eGFR. Seventy-six percent (n = 26) followed the recommendation to continue with metformin at eGFR ≥ 45 ml/min.

CONCLUSION

Compliance with the national guidelines was high regarding routines around renal functional tests, dose adjustment of CM and metformin discontinuation. Improvements can be made in using both cystatin C and serum creatinine for eGFR calculations as well as ensuring renal function tests within 12 h for acute/inpatients with acute disease that may affect renal function.

IMPLICATIONS FOR PRACTICE

This study raises awareness of the importance of adhering to guidelines in healthcare. To have knowledge about the current level of compliance regarding PCI-AKI is important to maintain and develop effective clinical implementation of guidelines. The variation in practice seen in this study emphasizes the need of more effective implementation strategies to ensure adherence with best practice.

Utility of measuring serum creatinine to detect renal compromise in ED patients receiving IV contrast-enhanced CT scan

OBJECTIVE

The objectives of this study are to determine the efficacy of a roster of clinical factors in identifying risk for renal insufficiency in emergency department (ED) patients requiring intravenous contrast-enhanced CT scan (IVCE-CT) and to help mitigate potential for developing contrast-induced nephropathy (CIN).

METHODS

A review was conducted of consecutive ED patients who received IVCE-CT during a 4-month period in our urban ED. The values of ED serum creatinine (SCr) performed were tabulated. The medical records of all patients with an elevated SCr (> 1.4 mg/dL) were reviewed to determine and correlate the presence of clinical risk factors for underlying renal insufficiency.

RESULTS

During the 4-month study period, there were 2260 consecutive cases who received IVCE-CT; of these, 2250 (99.6%) had concomitant measurement of SCr. Elevated SCr occurred in 141 patients (6.2%); of these, 75 had a SCr > 2 mg/dL. In all, 139/141 (98.6%) with an elevated SCr had an underlying chronic or acute medical condition identified by medical record review which potentially compromised renal function, including chronic renal disease, diabetes mellitus, HIV infection, cancer, hypertension, congestive heart failure, sepsis/septic shock, chronic alcoholism, and sickle cell disease. Two patients with no identified risk factor each had (mildly) elevated SCr; both had a normal SCr measured post-CT scan. The total cost of performing serum basic metabolic panel to measure SCr in all patients during the 4-month study period was $94,500.

CONCLUSIONS

Elevated SCr is rarely present in ED patients without recognized risk factors who receive IVCE-CT scan. The vast majority with underlying renal insufficiency are readily identified by a review of the patient's medical history and/or clinical findings. Routine SCr measurement on all ED patients regardless of risk stratification prior to IVCE imaging is neither time nor cost-effective.

Evaluation of a new point-of-care testing for creatinine and urea measurement

Point of care testing makes it possible to obtain results in an extremely short time. Recently, radiometer has expanded the panel of tests available on its ABL90 FLEX PLUS blood gas analyzer (ABL90) by adding urea and creatinine. The aim of this study was to verify the performance of these new parameters. This included assessment of imprecision, linearity, accuracy by comparison with central laboratory standard assays and interferences. In addition, clinical utility in a dialysis center was evaluated. Within-lab coefficients of variation were close to 2%. The mean and limits of agreement (mean ± 1.96 SD) of the difference between ABL90 and Roche enzymatic assays on cobas 8000 were 0.5 (from -1.4 to 2.3) mmol/L and -0.9 (from -19.5 to 17.8) µmol/L for urea and creatinine, respectively. The ABL90 enzymatic urea and creatinine assays met the acceptance criteria based on biological variation for imprecision and showed good agreement with central laboratory. The two assays were unaffected by hematocrit variation between 20 and 70%, hemolysis and icterus interferences. It should be noted that the relationship between lab methods and ABL90 was conserved even for high pre-dialysis values allowing easy access to dialysis adequacy parameters (Kt/V) and muscle mass evaluation (creatinine index). Rapid measurement of creatinine and urea using whole blood specimens on ABL90 appears as a fast and convenient method. Analytical performances were in accordance with our expectations without any significant interferences by hemolysis or icterus.

Administration of iodinated contrast: What is the risk in cancer patients?

OBJECTIVE

Despite the association between intravenous contrast and kidney injury, few studies exist in oncology. Our objective was to estimate kidney outcomes following iodinated contrast-enhanced computed tomography (CECT) in cancer patients, and to evaluate whether self-assessment questionnaires can identify kidney injury risk factors.

METHODS

This prospective observational study included 289 patients who underwent a CECT scan between March and May 2017 in a hospital setting. All patients completed the modified European Society of Urogenital Radiology (ESUR) questionnaire and had an estimated glomerular filtration rate (eGFR) >30 ml/min/1.73 m² on the day of the examination. Outcomes were followed for 4 months. Univariate and logistic regression analyses were carried out.

RESULTS

In the logistic regression analysis, the only variables statistically associated with deterioration in the eGFR were age, (odds ratio (OR) = 1.091, p = 0.003), female sex, (OR 0.22, p = 0.020) and arterial hypertension (AH), (OR = 3.57, p = 0.019). Regarding exitus, only the group with a worse eGFR was close to predictive statistical significance (OR = 2.48, p = 0.09).

CONCLUSIONS

The administration of iodinated contrast in cancer patients was not associated with an increase in kidney outcomes. Risk factors in these patients were age, sex and AH.

Kidney function testing prior to contrast-enhanced CT: a comparative cost analysis of a personalised risk-stratified pathway versus a test all approach

AIM

To map current contrast-enhanced computed tomography (CT) pathways, develop a risk-stratified pathway, and model associated costs and resource use.

MATERIALS AND METHODS

Phase 1 comprised multicentre mapping of current practice and development of an alternative pathway, replacing pre-assessment of estimated glomerular filtration rate (eGFR) with a scan-day screening questionnaire for risk stratification and point of care (PoC) creatinine. Phase 2 measured resource use and analysis of routinely collected data, used to populate a model comparing the costs of current and risk-stratified pathways in Phase 3.

RESULTS

Site variation across a range of processes within the clinical care pathway was identified. Data from a single centre suggested that 78% (n=347/447) could have avoided their pre-scan laboratory test as they did not have post-contrast acute kidney injury (AKI) risk factors. Only 24% of outpatients who underwent computed tomography (CT) would have identified risk factors, which would have prompted a scan-day PoC test. There was a 94% probability that the risk-stratified pathway was cost-saving, with an estimated 5-year potential cost saving of £69,620 (95% CI: -£13,295-£154,603). Although the cost of a laboratory serum creatinine test is cheaper than the PoC equivalent (£5.29 versus £5.96), the screening questionnaire ruled out the need for a large majority of the eGFR measurements specifically for the CT examination.

CONCLUSION

The present study proposes an alternative pathway, which has the potential to improve the efficiency of the current CT pathway. A multicentre appraisal is required to demonstrate the impact of embedding this new pathway on a wider NHS level, particularly in light of new diagnostic guidance (DG37) published by NICE.

Point-of-care creatinine tests to assess kidney function for outpatients requiring contrast-enhanced CT imaging: systematic reviews and economic evaluation

BACKGROUND

Patients with low estimated glomerular filtration rates may be at higher risk of post-contrast acute kidney injury following contrast-enhanced computed tomography imaging. Point-of-care devices allow rapid measurement of estimated glomerular filtration rates for patients referred without a recent estimated glomerular filtration rate result.

OBJECTIVES

To assess the clinical effectiveness and cost-effectiveness of point-of-care creatinine tests for outpatients without a recent estimated glomerular filtration rate measurement who need contrast-enhanced computed tomography imaging.

METHODS

Three systematic reviews of test accuracy, implementation and clinical outcomes, and economic analyses were carried out. Bibliographic databases were searched from inception to November 2018. Studies comparing the accuracy of point-of-care creatinine tests with laboratory reference tests to assess kidney function in adults in a non-emergency setting and studies reporting implementation and clinical outcomes were included. Risk of bias of diagnostic accuracy studies was assessed using a modified version of the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADAS-2) tool. Probabilities of individuals having their estimated glomerular filtration rates correctly classified were estimated within a Bayesian framework and pooled using a fixed-effects model. A de novo probabilistic decision tree cohort model was developed to characterise the decision problem from an NHS and a Personal Social Services perspective. A range of alternative point-of-care testing approaches were considered. Scenario analyses were conducted.

RESULTS

Fifty-four studies were included in the clinical reviews. Twelve studies reported diagnostic accuracy for estimated glomerular filtration rates; half were rated as being at low risk of bias, but there were applicability concerns for most. i-STAT (Abbott Point of Care, Inc., Princeton, NJ, USA) and ABL (Radiometer Ltd, Crawley, UK) devices had higher probabilities of correctly classifying individuals in the same estimated glomerular filtration rate categories as the reference laboratory test than StatSensor® devices (Nova Biomedical, Runcorn, UK). There was limited evidence for epoc® (Siemens Healthineers AG, Erlangen, Germany) and Piccolo Xpress® (Abaxis, Inc., Union City, CA, USA) devices and no studies of DRI-CHEM NX 500 (Fujifilm Corporation, Tokyo, Japan). The review of implementation and clinical outcomes included six studies showing practice variation in the management decisions when a point-of-care device indicated an abnormal estimated glomerular filtration rate. The review of cost-effectiveness evidence identified no relevant studies. The de novo decision model that was developed included a total of 14 strategies. Owing to limited data, the model included only i-STAT, ABL800 FLEX and StatSensor. In the base-case analysis, the cost-effective strategy appeared to be a three-step testing sequence involving initially screening all individuals for risk factors, point-of-care testing for those individuals with at least one risk factor, and including a final confirmatory laboratory test for individuals with a point-of-care-positive test result. Within this testing approach, the specific point-of-care device with the highest net benefit was i-STAT, although differences in net benefit with StatSensor were very small.

LIMITATIONS

There was insufficient evidence for patients with estimated glomerular filtration rates < 30 ml/minute/1.73 m2, and on the full potential health impact of delayed or rescheduled computed tomography scans or the use of alternative imaging modalities.

CONCLUSIONS

A three-step testing sequence combining a risk factor questionnaire with a point-of-care test and confirmatory laboratory testing appears to be a cost-effective use of NHS resources compared with current practice. The risk of contrast causing acute kidney injury to patients with an estimated glomerular filtration rate of < 30 ml/minute/1.73 m2 is uncertain. Cost-effectiveness of point-of-care testing appears largely driven by the potential of point-of-care tests to minimise delays within the current computed tomography pathway.

FUTURE WORK

Studies evaluating the impact of risk-stratifying questionnaires on workflow outcomes in computed tomography patients without recent estimated glomerular filtration rate results are needed.

STUDY REGISTRATION

This study is registered as PROSPERO CRD42018115818.

FUNDING

This project was funded by the National Institute for Health Research (NIHR) Health Technology Assessment programme and will be published in full in Health Technology Assessment; Vol. 24, No. 39. See the NIHR Journals Library website for further project information.

Point of care creatinine testing in diagnostic imaging: A feasibility study within the outpatient computed tomography setting

INTRODUCTION

Although the risks associated with iodinated contrast administration are acknowledged to be very low, screening of kidney function prior to administration is still standard practice in many hospitals. This study has evaluated the feasibility of implementing a screening form in conjunction with point of care (PoC) creatinine testing as a method to manage the risks of post contrast acute kidney injury (PC-AKI) within the CT imaging pathway.

METHOD

Over an eight-week period 300 adult outpatients attending a UK CT department for contrast-enhanced scans were approached. Participants completed a screening questionnaire for co-morbidities linked to kidney dysfunction and consented to have a PoC and laboratory creatinine tests. Comparison was made against with previous baseline blood tests obtained within the preceding 3 months, as required by the study site. Participants were also invited to attend for follow up PoC and laboratory bloods tests at 48-72 h.

RESULTS

14 patients (4.7%) had a scan-day eGFR below 45mL/min/1.73m2, all identified through screening. The majority of patients (n=281/300; 93.7%) fell in the same risk category based on previous and scan-day blood results. Six PoC test failures were recorded on the scan day. The constant error between the Abbott i-STAT PoC scan-day measurements and the laboratory scan-day measurements was -3.71 (95% CI: -6.41 to -0.50). Five patients had an elevated creatinine (≥25% from baseline) post contrast administration, but no instances of PC-AKI (≥50% from baseline) were identified.

CONCLUSION

PoC creatinine testing is a practical method of ensuring renal function and is feasible in the radiology environment.

Point-of-care creatinine testing for kidney function measurement prior to contrast-enhanced diagnostic imaging: evaluation of the performance of three systems for clinical utility

Background:

Acute kidney injury (AKI) can occur rarely in patients exposed to iodinated contrast and result in contrast-induced AKI (CI-AKI). A key risk factor is the presence of preexisting chronic kidney disease (CKD); therefore, it is important to assess patient risk and obtain kidney function measurement prior to administration. Point-of-care (PoC) testing provides an alternative strategy but there remains uncertainty, with respect to diagnostic accuracy and clinical utility.

Methods:

A device study compared three PoC analysers (Nova StatSensor, Abbott i-STAT and Radiometer ABL800 FLEX) with a reference laboratory standard (Roche Cobas 8000 series, enzymatic creatinine). Three hundred adult patients attending a UK hospital phlebotomy department were recruited to have additional blood samples for analysis on the PoC devices.

Results:

The ABL800 FLEX had the strongest concordance with laboratory measured serum creatinine (mean bias=−0.86, 95% limits of agreement=−9.6 to 7.9) followed by the i-STAT (average bias=3.88, 95% limits of agreement=−8.8 to 16.6) and StatSensor (average bias=3.56, 95% limits of agreement=−27.7 to 34.8). In risk classification, the ABL800 FLEX and i-STAT identified all patients with an eGFR≤30, whereas the StatSensor resulted in a small number of missed high-risk cases (n=4/13) and also operated outside of the established performance goals.

Conclusions:

The screening of patients at risk of CI-AKI may be feasible with PoC technology. However, in this study, it was identified that the analyser concordance with the laboratory reference varies. It is proposed that further research exploring PoC implementation in imaging department pathways is needed.