Eur J Anaesthesiol 2015; 32:37–43
ORIGINAL ARTICLE
Doppler renal resistive index for early detection of acute kidney injury after major orthopaedic surgery A prospective observational study Philippe Marty, Simon Szatjnic, Fabrice Ferre, Jean-Marie Conil, Nicolas Mayeur, Olivier Fourcade, Stein Silva and Vincent Minville BACKGROUND Postoperative acute kidney injury (AKI) is a cause of morbidity and mortality. Its diagnosis requires better markers than variations in diuresis or postoperative serum creatinine. OBJECTIVES The aim of this study was to evaluate the accuracy of Doppler renal resistive index for early detection of AKI after hip or knee arthroplasty. DESIGN A prospective observational study. SETTING A single-centre study in a university hospital. PATIENTS Fifty men and women older than 65 years, requiring hip or knee replacement with at least two perioperative AKI risk factors, including diabetes, arteritis, chronic heart or renal dysfunction, and prescription of angiotensin-converting enzyme (ACE) inhibitors. Exclusion criteria were poor abdominal echogenicity, arrhythmia, respiratory failure or agitation. INTERVENTION Renal resistive index was measured preoperatively and in the postanaesthesia care unit.
RESULTS Sixteen patients presented with AKI in the postoperative period. Resistive index was increased in this group in both the preoperative [0.72 (0.69 to 0.73) vs. 0.66 (0.58 to 0.71); P ¼ 0.01] and postoperative periods [0.75 (0.71 to 0.75) vs. 0.67 (0.62 to 0.72); P ¼ 0.0001]. Resistive index evaluated by ROC curves and AUC to detect AKI was 0.862 [95% confidence interval (95% CI) 0.735 to 0.943]. The most accurate cut-off value was a postoperative resistive index of 0.705 (sensitivity ¼ 94%, specificity ¼ 71%, LRR ¼ 3.19 and LR– ¼ 0.09). The grey area between 0.705 and 0.73, corresponding to the inconclusive zone, included 26% (13/50) of all the patients. CONCLUSION Postoperative resistive index appears to be effective for early detection of AKI after major orthopaedic surgery. Resistive index can be measured in the postoperative care unit in patients at risk of AKI. TRIAL REGISTRATION NUMBER 29-0512. Published online 10 July 2014
Introduction Detecting acute kidney injury (AKI) in the perioperative period is of major interest in the management of high-risk surgical patients1,2 because of its impact on both morbidity3,4 and mortality.5 Although various tests are available, despite numerous studies, an early, accurate and easy-touse marker of AKI in clinical practice is still lacking. Serum creatinine concentration (sCr) remains a popular marker despite its known limitations.6 Another AKI marker, the urine output, lacks specificity. Numerous biomarkers (Ngal,7 cystatin C8) are available, but in the context of postoperative AKI, their clinical significance
remains to be clarified. Among the predictors of AKI, Doppler renal resistive index (RI) is of interest. It can be measured quickly, with low intra- and interobserver variability.9,10 Bossard et al.11 showed that RI measurement in the immediate postoperative period after cardiac surgery predicted delayed AKI. Its usefulness in other clinical settings should now be clarified. Major orthopaedic surgery is known to be haemorrhagic12 and might precipitate AKI in patients with perioperative risk factors. The aim of this study was to assess the usefulness of preoperative resistive index, postoperative resistive
From the De´partement Anesthe´sie-Re´animation, CHU Toulouse, Universite´ Toulouse III Paul Sabatier, Faculte´ de Me´decine Toulouse-Rangueil, EA 4564-MATN, Institut Louis Bugnard (IFR 150), Toulouse, France Correspondence to Philippe Marty, De´partement Anesthe´sie-Re´animation, CHU Toulouse, 31059 Toulouse, France E-mail: [email protected] 0265-0215 ß 2014 Copyright European Society of Anaesthesiology
DOI:10.1097/EJA.0000000000000120
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
38 Marty et al.
index and resistive index variation before and after surgery (%resistive index) for early AKI detection in major orthopaedic surgery.
Material and methods This prospective and observational study was conducted in the Department of Orthopaedics and Traumatology of a tertiary university hospital between November 2011 and November 2012. It was performed according to the Declaration of Helsinki. The institutional review board of the hospital gave its approval (registration no. 29-0512) in July 2011. Consent to participate in the study and consent to publish were obtained from all patients. Patients
All patients undergoing total hip or total knee arthroplasties (primary or revision) were screened. Biochemical and haematological tests undertaken were part of routine preoperative screening. Those more than 65 years old exhibiting at least two of the following known risk factor for AKI were included: diabetes, arteriopathy defined by severe lower limb arteriopathy or carotid stenosis more than 50%, chronic renal dysfunction with estimated glomerular filtration rate less than 60 ml min1 1.73 m2, chronic heart failure and receiving angiotensin-converting enzyme (ACE) inhibitors.13 Cardiac arrhythmia, poor echogenicity, asymmetrical kidney disease, postoperative agitation or confusion, postoperative tachypnoea more than 35 breaths min1 or respiratory failure were considered as exclusion criteria. Resistive index measurement
All echographic measurements were made with HD11XE equipment (Philips Medical System; Bothell, Washington, USA) by one of two investigators (P.M. and S.S.) using a 5 MHz pulsed wave Doppler probe. The kidney was visualised using bidimensional echography. Interlobular artery locations were guided by colour Doppler. Renal blood flow was obtained by pulsed Doppler imaging along a longitudinal section.14 Velocity in interlobular arteries was assessed three to five times in one of the two kidneys.15 Renal RI was calculated as RI ¼ (peak systolic velocity – end-diastolic velocity)/peak systolic velocity as the average of all measured values. %RI was calculated as %RI ¼ (postoperative resistive index – preoperative resistive index)/preoperative resistive index. The variability of resistive index measurements was calculated. Resistive index was tested three times in 10 patients by the same observer (S.S.) in order to evaluate intraobserver variability and by a second observer (P.M.) for interobserver variability. Study design
The following data were recorded after inclusion in the protocol: age, sex, presence of diabetes, hypertension, arteritis, cardiac failure, use of ACE inhibitors and sCr. Glomerular filtration rate (GFR) was estimated using the
Chronic Kidney Disease Epidemiology Collaboration equation taking into account sCr, sex and ethnicity.16 Resistive index was assessed before arrival in the operating room by one of the two investigators blinded to individual characteristics. The following data were recorded during surgery: type and length of surgery, type of anaesthesia, volume and type of fluid infusion, blood transfusion volume, blood loss and number of hypotensive episodes defined by mean arterial pressure under 80% of preoperative value for more than 5 min. Use of perioperative nephrotoxic medications were also reported (e.g. nonsteroidal anti-inflammatory drugs, vancomycin). Postoperative resistive index measurement was performed in the postanaesthesia care unit. The following criteria were needed before resistive index measurement: end of local bleeding defined by a bleeding loss of less than 100 ml over 30 min, absence of haemodynamic instability defined by a mean arterial pressure of less than 65 mmHg or a heart rate more than 120 bpm, visual analogue pain scale less than 3/10, absence of confusion or agitation and oxygen saturation more than 96%. Oxygen therapy was introduced if the saturation objective was not attained. sCr measurement in plasma was performed by routine hospital immuno-enzymatic technique and was assessed before surgery (reference value) and daily each morning until the level fell. Urine output was measured at 6-h intervals over 48 h. AKI was defined according to the Acute Kidney Injury Network classification scheme17 either as an increase in sCr greater than or equal to 26.4 mmol l1 (0.3 mg dl1) or greater than or equal to 150% from baseline or as urine output less than 0.5 ml kg1 h1 for 6 h or more. Statistical analysis
Data are expressed as mean SD, median (range) or number (percentage), as appropriate. Comparisons between non-AKI and AKI groups were performed using Fisher test for categorical variables and nonparametric Wilcoxon test for continuous variables. Because preoperative and postoperative resistive index are numerical data, receiver-operating characteristics (ROCs) curves and area under the curve (AUC) were computed. Cutoff values for preoperative resistive index, postoperative resistive index and resistive index variation before and after surgery (%RI) were chosen to correspond to the best respective Youden’s index calculated as follows: Youden’s index ¼ sensitivity þ specificity – 1.18 Positive and negative predictive values (PPV and NPV), sensitivity, specificity, positive and negative likelihood ratios (LRþ and LR–), and Kappa coefficient were calculated according to these best cut-off values. The ‘grey zone’ determination was determined using a two-step procedure.19,20 The first step consisted of the determination of the best threshold according to Youden’s index. The 95% CI of the best threshold was defined. The second step was conducted to determine the value for which the variables
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 39
of interest did not provide conclusive information. It corresponds to a range of values for which formal conclusions could not be obtained. We defined inconclusive responses for values presenting with either sensitivity lower than 90% or specificity lower than 90% (diagnosis tolerance of 10%). The grey zone was then defined as the values of the variables that did not permit a diagnosis tolerance of 10%. However, if the characteristics of the study population produce a 95% CI of the best thresholds larger than the inconclusive zone, the values obtained during the first step were retained as the grey zone. This two-step procedure allows us to provide robust results not influenced by potential outliers. A P value of less than 0.05 was considered statistically significant. Statistical analysis was performed using Stata SE 11.1 software.
Among the individuals undergoing hip or knee arthroplasty, 72 met the inclusion criteria and were eligible for inclusion, and 50 individuals completed the study. Twenty-two patients were excluded for poor echogenicity, cardiac arrhythmia, agitation or tachypnoea. Patients with AKI and without AKI were similar except for basal GFR and age (Table 1). Sixteen individuals met the criteria for AKI that was not long-lasting and did not require dialysis. According to the AKI Network (AKIN) classification, 11 patients were classified as AKIN stage 1, and five patients were classified as AKIN stage 2. In the AKI group, peak sCr occurred within the first 4 postoperative days in all individuals: it was observed in 38% of individuals on the day after surgery (day 1), 50% on day 2, 6% on day 3 and 6% on day 4.
Results
Perioperative period
Patient characteristics
Perioperative characteristics were similar between AKI and non-AKI groups (Table 2). The incidence of AKI did
Chart flow and study design are shown in Fig. 1. Fig. 1
72 patients elected for inclusion
22 exclusions : • • • •
Poor insonation = 6 (27%) Cardiac arrhythmia = 8 (36%) Agitation/confusion = 4 (18%) Tachypnoea = 4 (18%)
50 patients included
RI measurements before and after surgery
No AKI : 34 (68%)
Surgery
AKI : 16 (32%)
Study design. AKI, acute kidney injury; RI, resistive index.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
40 Marty et al.
Table 1
Individual characteristics according to postoperative renal status
Male, n (%) Age, years; median (25th to 75th percentiles) Chronic heart failure, n (%) Diabetes, n (%) HTA, n (%) ACE inhibitor, n (%) Arteritis, n (%) GRF (Mean SD) ml min1 1.73 m – 2; median (25th to 75th percentiles)
All subjects (50)
Non-AKI (34)
AKI (16)
P
21 (42) 72 (67 to 74)
14 (41) 69 (67 to 73)
7 (44) 77 (69 to 81)
0.86 0.005
7 (14) 17 (34) 38 (76) 8 (16) 20 (40) 92 (66 to 124)
3 (9) 12 (35) 23 (68) 6 (18) 14 (41) 96 (66 to 124)
4 (25) 5 (31) 13 (81) 2 (12) 6 (37) 79 (59 to 114)
0.12 0.77 0.07 0.64 0.45 0.038
Values in italics indicate significant P values. ACE, angiotensin-converting enzyme; AKI, acute kidney injury; GFR, glomerular filtration rate; HTA, hypertension.
not differ between hip and knee arthroplasty, nor between general and spinal anaesthesia. No significant difference was found in terms of duration of surgery, blood loss, intraoperative fluid management and haemodynamic stability. The incidence of AKI did not differ between septic and nonseptic surgery. Nephrotoxic medication was recorded and was not associated with postoperative AKI.
Renal Doppler measurements
Intraobserver and interobserver variabilities for resistive index were 2.1% 2.6 and 2.9% 2.7%, respectively. Distribution of preoperative and postoperative resistive index among non-AKI and AKI groups is represented in Fig. 2. A significant difference was found between nonAKI and AKI groups for preoperative resistive index [0.66 (0.58 to 0.71) vs. 0.72 (0.69 to 0.73) respectively; P ¼ 0.01], postoperative resistive index [0.67 (0.62 to 0.72) vs. 0.75 (0.71 to 0.75); P ¼ 0.0001] and %RI [2.3% (1.1 to 3.8) vs. 4.4% (3.1 to 5.2); P ¼ 0.0002]. Resistive index assessed by ROC curves and AUC to distinguish non-AKI and AKI groups was 0.862 [95% confidence interval (95% CI 0.735 to 0.943] for postoperative resistive index, 0.726 (95% CI 0.582 to 0.843) for preoperative resistive index and 0.829 (95% CI 0.696 to 0.921) for resistive index variation (Fig. 3). The most accurate cut-off value to distinguish Table 2
non-AKI and AKI groups was a postoperative resistive index of 0.705. At this threshold, sensitivity was 94%, specificity was 71%, LRþ was 3.19 and LR– was 0.09. For postoperative resistive index, the grey zone was between 0.705 and 0.73 (Fig. 4). Among all the patients, 26% (13/50) had a resistive index in this inconclusive zone. Univariate analysis showed that AKI correlated significantly with preoperative resistive index, postoperative resistive index, variation of resistive index, estimated GFR and age (Tables 1 and 3). After including the five most significant variables in multivariate analysis and allowing the model to only retain the two variables with the greater association level, postoperative resistive index and, to a lesser extent, resistive index variation were the most significant predictive variables (Table 4). AKI was defined according to the Acute Kidney Injury Network classification scheme.17 Postoperative resistive index allowed a prediction of AKI 24 h before AKIN classification (median value).
Discussion In this study, the incidence of postoperative AKI was 32%. No patient required dialysis. This incidence is higher than generally expected after hip or knee
Intraoperative data according to postoperative renal status
Primary and revision THA n (%) Septic surgery n (%) General anaesthesia n (%) Length of anaesthesia, min median (25th to 75th percentiles) Intraoperative bleeding, ml median (25th to 75th percentiles) Intraoperative fluid loading, ml median (25th to 75th percentiles) Intraoperative blood transfusion, n RBC median (25th to 75th percentiles) Number of intraoperative hypotensive episodes median (25th to 75th percentiles) Perioperative nephrotoxic medication n (%) NSAID Vancomycin
All
Non-AKI
AKI
P
29 (58) 7 (18) 43 (86) 144 (121 to 166)
21 (62) 3 (9) 30 (88) 152 (130 to 176)
8 (50) 4 (25) 13 (81) 136 (118 to 158)
0.56 0.12 0.83 0.25
455 (373 to 528)
492 (419 to 548)
390 (336 to 498)
0.42
1790 (1130 to 2543)
1911 (1210 to 2730)
1531 (980 to 2120)
0.12
1 (0 to 4)
1 (0 to 5)
1 (0 to 4)
0.37
2 (0 to 4)
2 (0 to 3)
3 (0 to 5)
0.16
1 (2) 8 (16)
1 (3) 4 (12)
0 (0) 4 (25)
0.88 0.25
AKI, acute kidney injury; NSAID, nonsteroidal anti-inflammatory drug; RBC, red blood cell; THA, total hip arthroplasty.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 41
Fig. 2
two infections were due to propionibacterium acnes and three to coagulase negative staphylococcus, which are known to provoke a weak inflammatory response.
P < 0.0001
ns
RI 0.9 0.8 0.7 0.6 0.5 Preoperative
Postoperative
Preoperative
No AKI
Postoperative
AKI
Preoperative and postoperative resistive index according to renal status. AKI, acute kidney injury.
replacement.13 The difference results from the decision to select a population at risk, corresponding to the one that would give concern in clinical practice. Our university hospital is a referral centre for complex hip or knee revision, and this type of surgery represents a source of increased risk for AKI. Another explanation for the difference might stem from the choice of AKIN classification for the diagnosis of AKI. It is more sensitive than other classifications such as the RIFLE classification (Risk of renal failure, Injury to kidney, Failure of kidney function, Loss of kidney function and End-stage renal failure).16 Surprisingly, septic surgery was not associated with an increased risk of AKI. This could be explained not only by insufficient septic patients but also by the lack of a systemic inflammatory response that is associated with some bacterial infections. In this study,
In this study, resistive index was measured before and after surgery. Preoperative resistive index probably reflects pre-existing renal disease and could help to identify those at risk of AKI. Postoperative measurement takes into account not only the pre-existing condition but also changes following the operation. The best variable to distinguish non-AKI and AKI was postoperative resistive index with an ROC curve area of 0.862 (95% CI 0.735 to 0.943). The most accurate cut-off value to distinguish noAKI and AKI was a postoperative resistive index of 0.705. The ‘grey zone’ approach has been proposed to avoid the binary approach of a ‘black or white’ decision that does not reflect clinical reality. Employment of these boundaries seems more clinically relevant for discrimination of patients with or without AKI. In this study, the ‘grey zone’ of postoperative resistive index was between 0.705 and 0.73. In this interval, the clinician should not predict the occurrence of AKI due to low sensitivity and specificity. In this study, only 13 (26%) patient values were placed in the ‘grey zone’. However, seven (44%) patients in the AKI group lay in the ‘grey zone’. We have shown that preoperative resistive index can predict AKI. This underlines the importance of preoperative risk factors, and the susceptibility of some patients to any form of renal aggression. However, preoperative resistive index was less effective than postoperative resistive index in this study, according to AUC of the ROC curves. Age is known to influence resistive index, as it modifies vascular properties. However, multivariate analysis, which included the five most significant variables in
Fig. 3
100
80
Sensitivity
Postoperative RI = 0.862 Preoperative RI = 0.726
60
RI variation = 0.829 SCr clearance = 0.662
40
20
0 0
20
40
60
80
100
100-Specificity Receiver operating characteristics curves representing the discriminative power of different variables to detect postoperative acute kidney injury. RI, resistive index; SCr, serum creatinine clearance.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
42 Marty et al.
Fig. 4
Our study is an observational study and the impact of renal Doppler on the management of AKI prevention must be defined in randomised trials.
P < 0.001
RI 0.9
0.8
0.7
0.6
A KI
N
o
A KI
0.5
Postoperative resistive index according to renal status. Comparison between AKI and non-AKI groups. AKI, acute kidney injury.
univariate analysis (preoperative resistive index, postoperative resistive index, variation of resistive index, estimated preoperative GFR and age), found that postoperative resistive index and, to a lesser extent, resistive index variation were the most significant predictive variables. This underlines the fact that age is not the principal determinant of postoperative AKI. As an early indicator of AKI, resistive index contrasts with other markers that use an increase in sCr either alone or in combination with urine output, such as the AKIN classification. Resistive index evaluation allowed earlier detection of AKI 1 or 2 days before the sCr increase. Advanced warning allows preventive measures such as careful fluid and haemodynamic adjustments, and avoiding the use of nephrotoxic medication. However, the impact of early recognition of AKI on renal outcome needs a larger study. Table 3
The factors determining the value of the resistive index remain unclear, but there are sufficient available data to help reduce the factors that could interfere with the value of resistive index regardless of kidney injury and renal prognosis. Accordingly, we sought to achieve haemodynamic stability and adequate analgesia with a visual analogue pain score less than 3/10 to limit the influence of pain on the mean arterial pressure.21 Facial oxygen was given after surgery to maintain saturation greater than 96%,22 and the influence of paCO2 was also limited preoperatively and postoperatively, as patients were woken breathing spontaneously, and in the event of respiratory failure were excluded. Heart rate was not considered a significant influence on resistive index, given the low level of evidence, and the absence of any clinically important adverse report.23 The delay between the onset of kidney injury and changes in resistive index remains unclear. Our results favour a rapid response of this index secondary to kidney injury and this is supported by studies in other clinical situations. In cardiac surgery, resistive index variation during surgery correlated with postoperative AKI,11 and resistive index improved rapidly in response to catecholamine infusion during septic shock.24 The speed of change in resistive index following kidney injury is vital to its acceptance for the early diagnosis of AKI, and for the implementation of measures to preserve renal function. A main limitation of this study is its monocentric design. Ultrasonography is known to be operator dependent and our results should be confirmed by other operators in other centres. We have deliberately chosen to study a relatively small group of 50 patients at risk of AKI, and
Haemodynamic and renal data according to postoperative renal status All individuals
HR, bpm median (25th to 75th percentiles) Preoperative 89 (78 to 112) Postoperative 94 (69 to 114) MAP, mmHg median (25th to 75th percentiles) Preoperative 75 (65 to 86) Postoperative 71 (64 to 83) Number of hypotensive episodes median (25th to 75th percentiles) Intraoperative 2 (0 to 4) RI median (25th to 75th percentiles) Preoperative 0.68 (0.6 to 0.71) Postoperative 0.70 (0.64 to 0.73)M RI variation median (25th to 75th percentiles) 2.7 (1.6 to 4.4) GFR, ml min1 1.73 m2 median (25th to 75th percentiles) Pre 92 (66 to 124) Peak value 88 (68 to 106)
Non-AKI
AKI
P
92 (76 to 118) 93 (67 to 116)
88 (77 to 111) 96 (70 to 118)
0.45 0.13
74 (62 to 84) 72 (63 to 86)
76 (64 to 87) 68 (63 to 81)
0.54 0.33
2 (0 to 3)
3 (0 to 5)
0.16
0.66 (0.58 to 0.71) 0.67 (0.62 to 0.72) 2.3 (1.1 to 3.8)
0.72 (0.69 to 0.73) 0.75 (0.71 to 0.75)M 4.4 (3.1 to 5.2)
0.01 0.0001 0.0002
96 (66 to 124) 86 (64 to 118)
79 (59 to 114) 62 (36 to 98)M
0.038 0.015
Values in italics indicate significant P values between AKI and non-AKI groups. AKI, acute kidney injury; bpm, beats per minute; GFR, glomerular filtration rate; HR, heart rate; MAP, mean arterial pressure; RI, Doppler renal resistive index; SCr, serum creatinine value. MSignificant P values between preoperative and postoperative period.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 43
Table 4 Multivariate forward stepwise analysis (entering the three variables that best correlated with acute kidney injury in univariate analysis)
3
4 RI post % RI
P
OR
95% inf
95% sup
0.0031 0.48
2.339E13 12.57
31535.36 0.01
1.735E22 14957.097
RI post, postoperative renal index; % RI, renal index variation.
5 6 7
our results may not be applicable to patients with a different level of risk. This should be assessed in a larger sample, in a variety of patient groups following different surgical procedures. Intraobserver and interobserver variabilities for resistive index were 2.1% 2.6 and 2.9% 2.7%, respectively. Even if this variability is acceptable, it is close to resistive index variation before and after surgery. Consequently, resistive index variation should be interpreted with caution and all measurements should be made by the same operator. Our choice of the AKIN classification to define postoperative AKI is also debatable. The decision was based on its clinical significance, stage 1 having shown an impact on patient morbidity and mortality,2 and on its sensitivity.16
Conclusion Postoperative resistive index can detect early AKI after haemorrhagic orthopaedic surgery. Resistive index should be performed in the postoperative care unit in those at risk of AKI. Larger trials are needed to confirm these data in other groups and other clinical scenarios.
Acknowledgements relating to this article Assistance with the study: none. Financial support and sponsorship: this work was supported by the Department of Anaesthesiology, CHU Toulouse, Universite´ Toulouse III Paul Sabatier, Faculte´ de Me´decine Toulouse-Rangueil, EA 4564-MATN, Institut Louis Bugnard (IFR 150), Toulouse F-31000, France. Conflicts of interest: none.
8 9
10
11
12 13
14
15
16 17
18 19
20
21
22
Presentation: none.
References 1
2
Uchino S, Bellomo R, Goldsmith D, Bates S, Ronco C. An assessment of the RIFLE criteria for acute renal failure in hospitalized patients. Crit Care Med 2006; 34:1913–1917. Praught ML, Shlipak MG. Are small changes in serum creatinine an important risk factor? Curr Opin Nephrol Hypertens 2005; 14:265–270.
23
24
Fischer MJ, Brimhall BB, Parikh CR. Uncomplicated acute renal failure and posthospital care: a not so uncomplicated illness. Am J Nephrol 2008; 28:523–530. Coca SG, Singanamala S, Parikh CR. Chronic kidney disease after acute kidney injury: a systematic review and meta-analysis. Kidney Int 2012; 81:442–448. Lafrance JP, Miller DR. Acute kidney injury associates with increased longterm mortality. J Am Soc Nephrol 2010; 21:345–352. Waring WS, Moonie A. Earlier recognition of nephrotoxicity using novel biomarkers of acute kidney injury. Clin Toxicol 2011; 49:720–728. Hjortrup PB, Haase N, Wetterslev M, Perner A. Clinical review: predictive value of neutrophil gelatinase-associated lipocalin for acute kidney injury in intensive care patients. Crit Care 2013; 17:211. Obermuller N, Geiger H, Weipert C, Urbschat A. Current developments in early diagnosis of acute kidney injury. Int Urol Nephrol 2014; 46:1–7. Rud O, Moersler J, Peter J, et al. Prospective evaluation of interobserver variability of the hydronephrosis index and the renal resistive index as sonographic examination methods for the evaluation of acute hydronephrosis. BJU Int 2012; 110:E350–E356. Tublin ME, Bude RO, Platt JF. Review. The resistive index in renal Doppler sonography: where do we stand? AJR Am J Roentgenol 2003; 180:885– 892. Bossard G, Bourgoin P, Corbeau JJ, Huntzinger J, Beydon L. Early detection of postoperative acute kidney injury by Doppler renal resistive index in cardiac surgery with cardiopulmonary bypass. Br J Anaesth 2011; 107:891–898. Oberweis BS, Nukala S, Rosenberg A, et al. Thrombotic and bleeding complications after orthopedic surgery. Am Heart J 2013; 165:427–433. Aveline C, Leroux A, Vautier P, Cognet F, Le Hetet H, Bonnet F. [Risk factors for renal dysfunction after total hip arthroplasty]. Ann Fr Anesth Reanim 2009; 28:728–734. Darmon M, Schortgen F, Vargas F, et al. Diagnostic accuracy of Doppler renal resistive index for reversibility of acute kidney injury in critically ill patients. Intensive Care Med 2011; 37:68–76. Ansarin K, Bavil AS, Ghabili K, et al. Are Doppler ultrasonography parameters symmetric between the right and left kidney? Int J Gen Med 2010; 3:371–373. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150:604–612. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007; 11:R31. Youden WJ. Index for rating diagnostic tests. Cancer 1950; 3:32–35. Cannesson M, Le Manach Y, Hofer CK, et al. Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a ‘gray zone’ approach. Anesthesiology 2011; 115:231– 241. Le Manach Y, Hofer CK, Lehot JJ, et al. Can changes in arterial pressure be used to detect changes in cardiac output during volume expansion in the perioperative period? Anesthesiology 2012; 117:1165–1174. Dewitte A, Coquin J, Meyssignac B, et al. Doppler resistive index to reflect regulation of renal vascular tone during sepsis and acute kidney injury. Crit Care 2012; 16:R165. Darmon M, Schortgen F, Leon R, et al. Impact of mild hypoxemia on renal function and renal resistive index during mechanical ventilation. Intensive care Med 2009; 35:1031–1038. Mostbeck GH, Gossinger HD, Mallek R, Siostrzonek P, Schneider B, Tscholakoff D. Effect of heart rate on Doppler measurements of resistive index in renal arteries. Radiology 1990; 175:511–513. Deruddre S, Cheisson G, Mazoit JX, Vicaut E, Benhamou D, Duranteau J. Renal arterial resistance in septic shock: effects of increasing mean arterial pressure with norepinephrine on the renal resistive index assessed with Doppler ultrasonography. Intensive Care Med 2007; 33:1557–1562.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
ORIGINAL ARTICLE
Doppler renal resistive index for early detection of acute kidney injury after major orthopaedic surgery A prospective observational study Philippe Marty, Simon Szatjnic, Fabrice Ferre, Jean-Marie Conil, Nicolas Mayeur, Olivier Fourcade, Stein Silva and Vincent Minville BACKGROUND Postoperative acute kidney injury (AKI) is a cause of morbidity and mortality. Its diagnosis requires better markers than variations in diuresis or postoperative serum creatinine. OBJECTIVES The aim of this study was to evaluate the accuracy of Doppler renal resistive index for early detection of AKI after hip or knee arthroplasty. DESIGN A prospective observational study. SETTING A single-centre study in a university hospital. PATIENTS Fifty men and women older than 65 years, requiring hip or knee replacement with at least two perioperative AKI risk factors, including diabetes, arteritis, chronic heart or renal dysfunction, and prescription of angiotensin-converting enzyme (ACE) inhibitors. Exclusion criteria were poor abdominal echogenicity, arrhythmia, respiratory failure or agitation. INTERVENTION Renal resistive index was measured preoperatively and in the postanaesthesia care unit.
RESULTS Sixteen patients presented with AKI in the postoperative period. Resistive index was increased in this group in both the preoperative [0.72 (0.69 to 0.73) vs. 0.66 (0.58 to 0.71); P ¼ 0.01] and postoperative periods [0.75 (0.71 to 0.75) vs. 0.67 (0.62 to 0.72); P ¼ 0.0001]. Resistive index evaluated by ROC curves and AUC to detect AKI was 0.862 [95% confidence interval (95% CI) 0.735 to 0.943]. The most accurate cut-off value was a postoperative resistive index of 0.705 (sensitivity ¼ 94%, specificity ¼ 71%, LRR ¼ 3.19 and LR– ¼ 0.09). The grey area between 0.705 and 0.73, corresponding to the inconclusive zone, included 26% (13/50) of all the patients. CONCLUSION Postoperative resistive index appears to be effective for early detection of AKI after major orthopaedic surgery. Resistive index can be measured in the postoperative care unit in patients at risk of AKI. TRIAL REGISTRATION NUMBER 29-0512. Published online 10 July 2014
Introduction Detecting acute kidney injury (AKI) in the perioperative period is of major interest in the management of high-risk surgical patients1,2 because of its impact on both morbidity3,4 and mortality.5 Although various tests are available, despite numerous studies, an early, accurate and easy-touse marker of AKI in clinical practice is still lacking. Serum creatinine concentration (sCr) remains a popular marker despite its known limitations.6 Another AKI marker, the urine output, lacks specificity. Numerous biomarkers (Ngal,7 cystatin C8) are available, but in the context of postoperative AKI, their clinical significance
remains to be clarified. Among the predictors of AKI, Doppler renal resistive index (RI) is of interest. It can be measured quickly, with low intra- and interobserver variability.9,10 Bossard et al.11 showed that RI measurement in the immediate postoperative period after cardiac surgery predicted delayed AKI. Its usefulness in other clinical settings should now be clarified. Major orthopaedic surgery is known to be haemorrhagic12 and might precipitate AKI in patients with perioperative risk factors. The aim of this study was to assess the usefulness of preoperative resistive index, postoperative resistive
From the De´partement Anesthe´sie-Re´animation, CHU Toulouse, Universite´ Toulouse III Paul Sabatier, Faculte´ de Me´decine Toulouse-Rangueil, EA 4564-MATN, Institut Louis Bugnard (IFR 150), Toulouse, France Correspondence to Philippe Marty, De´partement Anesthe´sie-Re´animation, CHU Toulouse, 31059 Toulouse, France E-mail: [email protected] 0265-0215 ß 2014 Copyright European Society of Anaesthesiology
DOI:10.1097/EJA.0000000000000120
Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
38 Marty et al.
index and resistive index variation before and after surgery (%resistive index) for early AKI detection in major orthopaedic surgery.
Material and methods This prospective and observational study was conducted in the Department of Orthopaedics and Traumatology of a tertiary university hospital between November 2011 and November 2012. It was performed according to the Declaration of Helsinki. The institutional review board of the hospital gave its approval (registration no. 29-0512) in July 2011. Consent to participate in the study and consent to publish were obtained from all patients. Patients
All patients undergoing total hip or total knee arthroplasties (primary or revision) were screened. Biochemical and haematological tests undertaken were part of routine preoperative screening. Those more than 65 years old exhibiting at least two of the following known risk factor for AKI were included: diabetes, arteriopathy defined by severe lower limb arteriopathy or carotid stenosis more than 50%, chronic renal dysfunction with estimated glomerular filtration rate less than 60 ml min1 1.73 m2, chronic heart failure and receiving angiotensin-converting enzyme (ACE) inhibitors.13 Cardiac arrhythmia, poor echogenicity, asymmetrical kidney disease, postoperative agitation or confusion, postoperative tachypnoea more than 35 breaths min1 or respiratory failure were considered as exclusion criteria. Resistive index measurement
All echographic measurements were made with HD11XE equipment (Philips Medical System; Bothell, Washington, USA) by one of two investigators (P.M. and S.S.) using a 5 MHz pulsed wave Doppler probe. The kidney was visualised using bidimensional echography. Interlobular artery locations were guided by colour Doppler. Renal blood flow was obtained by pulsed Doppler imaging along a longitudinal section.14 Velocity in interlobular arteries was assessed three to five times in one of the two kidneys.15 Renal RI was calculated as RI ¼ (peak systolic velocity – end-diastolic velocity)/peak systolic velocity as the average of all measured values. %RI was calculated as %RI ¼ (postoperative resistive index – preoperative resistive index)/preoperative resistive index. The variability of resistive index measurements was calculated. Resistive index was tested three times in 10 patients by the same observer (S.S.) in order to evaluate intraobserver variability and by a second observer (P.M.) for interobserver variability. Study design
The following data were recorded after inclusion in the protocol: age, sex, presence of diabetes, hypertension, arteritis, cardiac failure, use of ACE inhibitors and sCr. Glomerular filtration rate (GFR) was estimated using the
Chronic Kidney Disease Epidemiology Collaboration equation taking into account sCr, sex and ethnicity.16 Resistive index was assessed before arrival in the operating room by one of the two investigators blinded to individual characteristics. The following data were recorded during surgery: type and length of surgery, type of anaesthesia, volume and type of fluid infusion, blood transfusion volume, blood loss and number of hypotensive episodes defined by mean arterial pressure under 80% of preoperative value for more than 5 min. Use of perioperative nephrotoxic medications were also reported (e.g. nonsteroidal anti-inflammatory drugs, vancomycin). Postoperative resistive index measurement was performed in the postanaesthesia care unit. The following criteria were needed before resistive index measurement: end of local bleeding defined by a bleeding loss of less than 100 ml over 30 min, absence of haemodynamic instability defined by a mean arterial pressure of less than 65 mmHg or a heart rate more than 120 bpm, visual analogue pain scale less than 3/10, absence of confusion or agitation and oxygen saturation more than 96%. Oxygen therapy was introduced if the saturation objective was not attained. sCr measurement in plasma was performed by routine hospital immuno-enzymatic technique and was assessed before surgery (reference value) and daily each morning until the level fell. Urine output was measured at 6-h intervals over 48 h. AKI was defined according to the Acute Kidney Injury Network classification scheme17 either as an increase in sCr greater than or equal to 26.4 mmol l1 (0.3 mg dl1) or greater than or equal to 150% from baseline or as urine output less than 0.5 ml kg1 h1 for 6 h or more. Statistical analysis
Data are expressed as mean SD, median (range) or number (percentage), as appropriate. Comparisons between non-AKI and AKI groups were performed using Fisher test for categorical variables and nonparametric Wilcoxon test for continuous variables. Because preoperative and postoperative resistive index are numerical data, receiver-operating characteristics (ROCs) curves and area under the curve (AUC) were computed. Cutoff values for preoperative resistive index, postoperative resistive index and resistive index variation before and after surgery (%RI) were chosen to correspond to the best respective Youden’s index calculated as follows: Youden’s index ¼ sensitivity þ specificity – 1.18 Positive and negative predictive values (PPV and NPV), sensitivity, specificity, positive and negative likelihood ratios (LRþ and LR–), and Kappa coefficient were calculated according to these best cut-off values. The ‘grey zone’ determination was determined using a two-step procedure.19,20 The first step consisted of the determination of the best threshold according to Youden’s index. The 95% CI of the best threshold was defined. The second step was conducted to determine the value for which the variables
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 39
of interest did not provide conclusive information. It corresponds to a range of values for which formal conclusions could not be obtained. We defined inconclusive responses for values presenting with either sensitivity lower than 90% or specificity lower than 90% (diagnosis tolerance of 10%). The grey zone was then defined as the values of the variables that did not permit a diagnosis tolerance of 10%. However, if the characteristics of the study population produce a 95% CI of the best thresholds larger than the inconclusive zone, the values obtained during the first step were retained as the grey zone. This two-step procedure allows us to provide robust results not influenced by potential outliers. A P value of less than 0.05 was considered statistically significant. Statistical analysis was performed using Stata SE 11.1 software.
Among the individuals undergoing hip or knee arthroplasty, 72 met the inclusion criteria and were eligible for inclusion, and 50 individuals completed the study. Twenty-two patients were excluded for poor echogenicity, cardiac arrhythmia, agitation or tachypnoea. Patients with AKI and without AKI were similar except for basal GFR and age (Table 1). Sixteen individuals met the criteria for AKI that was not long-lasting and did not require dialysis. According to the AKI Network (AKIN) classification, 11 patients were classified as AKIN stage 1, and five patients were classified as AKIN stage 2. In the AKI group, peak sCr occurred within the first 4 postoperative days in all individuals: it was observed in 38% of individuals on the day after surgery (day 1), 50% on day 2, 6% on day 3 and 6% on day 4.
Results
Perioperative period
Patient characteristics
Perioperative characteristics were similar between AKI and non-AKI groups (Table 2). The incidence of AKI did
Chart flow and study design are shown in Fig. 1. Fig. 1
72 patients elected for inclusion
22 exclusions : • • • •
Poor insonation = 6 (27%) Cardiac arrhythmia = 8 (36%) Agitation/confusion = 4 (18%) Tachypnoea = 4 (18%)
50 patients included
RI measurements before and after surgery
No AKI : 34 (68%)
Surgery
AKI : 16 (32%)
Study design. AKI, acute kidney injury; RI, resistive index.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
40 Marty et al.
Table 1
Individual characteristics according to postoperative renal status
Male, n (%) Age, years; median (25th to 75th percentiles) Chronic heart failure, n (%) Diabetes, n (%) HTA, n (%) ACE inhibitor, n (%) Arteritis, n (%) GRF (Mean SD) ml min1 1.73 m – 2; median (25th to 75th percentiles)
All subjects (50)
Non-AKI (34)
AKI (16)
P
21 (42) 72 (67 to 74)
14 (41) 69 (67 to 73)
7 (44) 77 (69 to 81)
0.86 0.005
7 (14) 17 (34) 38 (76) 8 (16) 20 (40) 92 (66 to 124)
3 (9) 12 (35) 23 (68) 6 (18) 14 (41) 96 (66 to 124)
4 (25) 5 (31) 13 (81) 2 (12) 6 (37) 79 (59 to 114)
0.12 0.77 0.07 0.64 0.45 0.038
Values in italics indicate significant P values. ACE, angiotensin-converting enzyme; AKI, acute kidney injury; GFR, glomerular filtration rate; HTA, hypertension.
not differ between hip and knee arthroplasty, nor between general and spinal anaesthesia. No significant difference was found in terms of duration of surgery, blood loss, intraoperative fluid management and haemodynamic stability. The incidence of AKI did not differ between septic and nonseptic surgery. Nephrotoxic medication was recorded and was not associated with postoperative AKI.
Renal Doppler measurements
Intraobserver and interobserver variabilities for resistive index were 2.1% 2.6 and 2.9% 2.7%, respectively. Distribution of preoperative and postoperative resistive index among non-AKI and AKI groups is represented in Fig. 2. A significant difference was found between nonAKI and AKI groups for preoperative resistive index [0.66 (0.58 to 0.71) vs. 0.72 (0.69 to 0.73) respectively; P ¼ 0.01], postoperative resistive index [0.67 (0.62 to 0.72) vs. 0.75 (0.71 to 0.75); P ¼ 0.0001] and %RI [2.3% (1.1 to 3.8) vs. 4.4% (3.1 to 5.2); P ¼ 0.0002]. Resistive index assessed by ROC curves and AUC to distinguish non-AKI and AKI groups was 0.862 [95% confidence interval (95% CI 0.735 to 0.943] for postoperative resistive index, 0.726 (95% CI 0.582 to 0.843) for preoperative resistive index and 0.829 (95% CI 0.696 to 0.921) for resistive index variation (Fig. 3). The most accurate cut-off value to distinguish Table 2
non-AKI and AKI groups was a postoperative resistive index of 0.705. At this threshold, sensitivity was 94%, specificity was 71%, LRþ was 3.19 and LR– was 0.09. For postoperative resistive index, the grey zone was between 0.705 and 0.73 (Fig. 4). Among all the patients, 26% (13/50) had a resistive index in this inconclusive zone. Univariate analysis showed that AKI correlated significantly with preoperative resistive index, postoperative resistive index, variation of resistive index, estimated GFR and age (Tables 1 and 3). After including the five most significant variables in multivariate analysis and allowing the model to only retain the two variables with the greater association level, postoperative resistive index and, to a lesser extent, resistive index variation were the most significant predictive variables (Table 4). AKI was defined according to the Acute Kidney Injury Network classification scheme.17 Postoperative resistive index allowed a prediction of AKI 24 h before AKIN classification (median value).
Discussion In this study, the incidence of postoperative AKI was 32%. No patient required dialysis. This incidence is higher than generally expected after hip or knee
Intraoperative data according to postoperative renal status
Primary and revision THA n (%) Septic surgery n (%) General anaesthesia n (%) Length of anaesthesia, min median (25th to 75th percentiles) Intraoperative bleeding, ml median (25th to 75th percentiles) Intraoperative fluid loading, ml median (25th to 75th percentiles) Intraoperative blood transfusion, n RBC median (25th to 75th percentiles) Number of intraoperative hypotensive episodes median (25th to 75th percentiles) Perioperative nephrotoxic medication n (%) NSAID Vancomycin
All
Non-AKI
AKI
P
29 (58) 7 (18) 43 (86) 144 (121 to 166)
21 (62) 3 (9) 30 (88) 152 (130 to 176)
8 (50) 4 (25) 13 (81) 136 (118 to 158)
0.56 0.12 0.83 0.25
455 (373 to 528)
492 (419 to 548)
390 (336 to 498)
0.42
1790 (1130 to 2543)
1911 (1210 to 2730)
1531 (980 to 2120)
0.12
1 (0 to 4)
1 (0 to 5)
1 (0 to 4)
0.37
2 (0 to 4)
2 (0 to 3)
3 (0 to 5)
0.16
1 (2) 8 (16)
1 (3) 4 (12)
0 (0) 4 (25)
0.88 0.25
AKI, acute kidney injury; NSAID, nonsteroidal anti-inflammatory drug; RBC, red blood cell; THA, total hip arthroplasty.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 41
Fig. 2
two infections were due to propionibacterium acnes and three to coagulase negative staphylococcus, which are known to provoke a weak inflammatory response.
P < 0.0001
ns
RI 0.9 0.8 0.7 0.6 0.5 Preoperative
Postoperative
Preoperative
No AKI
Postoperative
AKI
Preoperative and postoperative resistive index according to renal status. AKI, acute kidney injury.
replacement.13 The difference results from the decision to select a population at risk, corresponding to the one that would give concern in clinical practice. Our university hospital is a referral centre for complex hip or knee revision, and this type of surgery represents a source of increased risk for AKI. Another explanation for the difference might stem from the choice of AKIN classification for the diagnosis of AKI. It is more sensitive than other classifications such as the RIFLE classification (Risk of renal failure, Injury to kidney, Failure of kidney function, Loss of kidney function and End-stage renal failure).16 Surprisingly, septic surgery was not associated with an increased risk of AKI. This could be explained not only by insufficient septic patients but also by the lack of a systemic inflammatory response that is associated with some bacterial infections. In this study,
In this study, resistive index was measured before and after surgery. Preoperative resistive index probably reflects pre-existing renal disease and could help to identify those at risk of AKI. Postoperative measurement takes into account not only the pre-existing condition but also changes following the operation. The best variable to distinguish non-AKI and AKI was postoperative resistive index with an ROC curve area of 0.862 (95% CI 0.735 to 0.943). The most accurate cut-off value to distinguish noAKI and AKI was a postoperative resistive index of 0.705. The ‘grey zone’ approach has been proposed to avoid the binary approach of a ‘black or white’ decision that does not reflect clinical reality. Employment of these boundaries seems more clinically relevant for discrimination of patients with or without AKI. In this study, the ‘grey zone’ of postoperative resistive index was between 0.705 and 0.73. In this interval, the clinician should not predict the occurrence of AKI due to low sensitivity and specificity. In this study, only 13 (26%) patient values were placed in the ‘grey zone’. However, seven (44%) patients in the AKI group lay in the ‘grey zone’. We have shown that preoperative resistive index can predict AKI. This underlines the importance of preoperative risk factors, and the susceptibility of some patients to any form of renal aggression. However, preoperative resistive index was less effective than postoperative resistive index in this study, according to AUC of the ROC curves. Age is known to influence resistive index, as it modifies vascular properties. However, multivariate analysis, which included the five most significant variables in
Fig. 3
100
80
Sensitivity
Postoperative RI = 0.862 Preoperative RI = 0.726
60
RI variation = 0.829 SCr clearance = 0.662
40
20
0 0
20
40
60
80
100
100-Specificity Receiver operating characteristics curves representing the discriminative power of different variables to detect postoperative acute kidney injury. RI, resistive index; SCr, serum creatinine clearance.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
42 Marty et al.
Fig. 4
Our study is an observational study and the impact of renal Doppler on the management of AKI prevention must be defined in randomised trials.
P < 0.001
RI 0.9
0.8
0.7
0.6
A KI
N
o
A KI
0.5
Postoperative resistive index according to renal status. Comparison between AKI and non-AKI groups. AKI, acute kidney injury.
univariate analysis (preoperative resistive index, postoperative resistive index, variation of resistive index, estimated preoperative GFR and age), found that postoperative resistive index and, to a lesser extent, resistive index variation were the most significant predictive variables. This underlines the fact that age is not the principal determinant of postoperative AKI. As an early indicator of AKI, resistive index contrasts with other markers that use an increase in sCr either alone or in combination with urine output, such as the AKIN classification. Resistive index evaluation allowed earlier detection of AKI 1 or 2 days before the sCr increase. Advanced warning allows preventive measures such as careful fluid and haemodynamic adjustments, and avoiding the use of nephrotoxic medication. However, the impact of early recognition of AKI on renal outcome needs a larger study. Table 3
The factors determining the value of the resistive index remain unclear, but there are sufficient available data to help reduce the factors that could interfere with the value of resistive index regardless of kidney injury and renal prognosis. Accordingly, we sought to achieve haemodynamic stability and adequate analgesia with a visual analogue pain score less than 3/10 to limit the influence of pain on the mean arterial pressure.21 Facial oxygen was given after surgery to maintain saturation greater than 96%,22 and the influence of paCO2 was also limited preoperatively and postoperatively, as patients were woken breathing spontaneously, and in the event of respiratory failure were excluded. Heart rate was not considered a significant influence on resistive index, given the low level of evidence, and the absence of any clinically important adverse report.23 The delay between the onset of kidney injury and changes in resistive index remains unclear. Our results favour a rapid response of this index secondary to kidney injury and this is supported by studies in other clinical situations. In cardiac surgery, resistive index variation during surgery correlated with postoperative AKI,11 and resistive index improved rapidly in response to catecholamine infusion during septic shock.24 The speed of change in resistive index following kidney injury is vital to its acceptance for the early diagnosis of AKI, and for the implementation of measures to preserve renal function. A main limitation of this study is its monocentric design. Ultrasonography is known to be operator dependent and our results should be confirmed by other operators in other centres. We have deliberately chosen to study a relatively small group of 50 patients at risk of AKI, and
Haemodynamic and renal data according to postoperative renal status All individuals
HR, bpm median (25th to 75th percentiles) Preoperative 89 (78 to 112) Postoperative 94 (69 to 114) MAP, mmHg median (25th to 75th percentiles) Preoperative 75 (65 to 86) Postoperative 71 (64 to 83) Number of hypotensive episodes median (25th to 75th percentiles) Intraoperative 2 (0 to 4) RI median (25th to 75th percentiles) Preoperative 0.68 (0.6 to 0.71) Postoperative 0.70 (0.64 to 0.73)M RI variation median (25th to 75th percentiles) 2.7 (1.6 to 4.4) GFR, ml min1 1.73 m2 median (25th to 75th percentiles) Pre 92 (66 to 124) Peak value 88 (68 to 106)
Non-AKI
AKI
P
92 (76 to 118) 93 (67 to 116)
88 (77 to 111) 96 (70 to 118)
0.45 0.13
74 (62 to 84) 72 (63 to 86)
76 (64 to 87) 68 (63 to 81)
0.54 0.33
2 (0 to 3)
3 (0 to 5)
0.16
0.66 (0.58 to 0.71) 0.67 (0.62 to 0.72) 2.3 (1.1 to 3.8)
0.72 (0.69 to 0.73) 0.75 (0.71 to 0.75)M 4.4 (3.1 to 5.2)
0.01 0.0001 0.0002
96 (66 to 124) 86 (64 to 118)
79 (59 to 114) 62 (36 to 98)M
0.038 0.015
Values in italics indicate significant P values between AKI and non-AKI groups. AKI, acute kidney injury; bpm, beats per minute; GFR, glomerular filtration rate; HR, heart rate; MAP, mean arterial pressure; RI, Doppler renal resistive index; SCr, serum creatinine value. MSignificant P values between preoperative and postoperative period.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
Renal resistive index for kidney injury detection after surgery 43
Table 4 Multivariate forward stepwise analysis (entering the three variables that best correlated with acute kidney injury in univariate analysis)
3
4 RI post % RI
P
OR
95% inf
95% sup
0.0031 0.48
2.339E13 12.57
31535.36 0.01
1.735E22 14957.097
RI post, postoperative renal index; % RI, renal index variation.
5 6 7
our results may not be applicable to patients with a different level of risk. This should be assessed in a larger sample, in a variety of patient groups following different surgical procedures. Intraobserver and interobserver variabilities for resistive index were 2.1% 2.6 and 2.9% 2.7%, respectively. Even if this variability is acceptable, it is close to resistive index variation before and after surgery. Consequently, resistive index variation should be interpreted with caution and all measurements should be made by the same operator. Our choice of the AKIN classification to define postoperative AKI is also debatable. The decision was based on its clinical significance, stage 1 having shown an impact on patient morbidity and mortality,2 and on its sensitivity.16
Conclusion Postoperative resistive index can detect early AKI after haemorrhagic orthopaedic surgery. Resistive index should be performed in the postoperative care unit in those at risk of AKI. Larger trials are needed to confirm these data in other groups and other clinical scenarios.
Acknowledgements relating to this article Assistance with the study: none. Financial support and sponsorship: this work was supported by the Department of Anaesthesiology, CHU Toulouse, Universite´ Toulouse III Paul Sabatier, Faculte´ de Me´decine Toulouse-Rangueil, EA 4564-MATN, Institut Louis Bugnard (IFR 150), Toulouse F-31000, France. Conflicts of interest: none.
8 9
10
11
12 13
14
15
16 17
18 19
20
21
22
Presentation: none.
References 1
2
Uchino S, Bellomo R, Goldsmith D, Bates S, Ronco C. An assessment of the RIFLE criteria for acute renal failure in hospitalized patients. Crit Care Med 2006; 34:1913–1917. Praught ML, Shlipak MG. Are small changes in serum creatinine an important risk factor? Curr Opin Nephrol Hypertens 2005; 14:265–270.
23
24
Fischer MJ, Brimhall BB, Parikh CR. Uncomplicated acute renal failure and posthospital care: a not so uncomplicated illness. Am J Nephrol 2008; 28:523–530. Coca SG, Singanamala S, Parikh CR. Chronic kidney disease after acute kidney injury: a systematic review and meta-analysis. Kidney Int 2012; 81:442–448. Lafrance JP, Miller DR. Acute kidney injury associates with increased longterm mortality. J Am Soc Nephrol 2010; 21:345–352. Waring WS, Moonie A. Earlier recognition of nephrotoxicity using novel biomarkers of acute kidney injury. Clin Toxicol 2011; 49:720–728. Hjortrup PB, Haase N, Wetterslev M, Perner A. Clinical review: predictive value of neutrophil gelatinase-associated lipocalin for acute kidney injury in intensive care patients. Crit Care 2013; 17:211. Obermuller N, Geiger H, Weipert C, Urbschat A. Current developments in early diagnosis of acute kidney injury. Int Urol Nephrol 2014; 46:1–7. Rud O, Moersler J, Peter J, et al. Prospective evaluation of interobserver variability of the hydronephrosis index and the renal resistive index as sonographic examination methods for the evaluation of acute hydronephrosis. BJU Int 2012; 110:E350–E356. Tublin ME, Bude RO, Platt JF. Review. The resistive index in renal Doppler sonography: where do we stand? AJR Am J Roentgenol 2003; 180:885– 892. Bossard G, Bourgoin P, Corbeau JJ, Huntzinger J, Beydon L. Early detection of postoperative acute kidney injury by Doppler renal resistive index in cardiac surgery with cardiopulmonary bypass. Br J Anaesth 2011; 107:891–898. Oberweis BS, Nukala S, Rosenberg A, et al. Thrombotic and bleeding complications after orthopedic surgery. Am Heart J 2013; 165:427–433. Aveline C, Leroux A, Vautier P, Cognet F, Le Hetet H, Bonnet F. [Risk factors for renal dysfunction after total hip arthroplasty]. Ann Fr Anesth Reanim 2009; 28:728–734. Darmon M, Schortgen F, Vargas F, et al. Diagnostic accuracy of Doppler renal resistive index for reversibility of acute kidney injury in critically ill patients. Intensive Care Med 2011; 37:68–76. Ansarin K, Bavil AS, Ghabili K, et al. Are Doppler ultrasonography parameters symmetric between the right and left kidney? Int J Gen Med 2010; 3:371–373. Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009; 150:604–612. Mehta RL, Kellum JA, Shah SV, et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007; 11:R31. Youden WJ. Index for rating diagnostic tests. Cancer 1950; 3:32–35. Cannesson M, Le Manach Y, Hofer CK, et al. Assessing the diagnostic accuracy of pulse pressure variations for the prediction of fluid responsiveness: a ‘gray zone’ approach. Anesthesiology 2011; 115:231– 241. Le Manach Y, Hofer CK, Lehot JJ, et al. Can changes in arterial pressure be used to detect changes in cardiac output during volume expansion in the perioperative period? Anesthesiology 2012; 117:1165–1174. Dewitte A, Coquin J, Meyssignac B, et al. Doppler resistive index to reflect regulation of renal vascular tone during sepsis and acute kidney injury. Crit Care 2012; 16:R165. Darmon M, Schortgen F, Leon R, et al. Impact of mild hypoxemia on renal function and renal resistive index during mechanical ventilation. Intensive care Med 2009; 35:1031–1038. Mostbeck GH, Gossinger HD, Mallek R, Siostrzonek P, Schneider B, Tscholakoff D. Effect of heart rate on Doppler measurements of resistive index in renal arteries. Radiology 1990; 175:511–513. Deruddre S, Cheisson G, Mazoit JX, Vicaut E, Benhamou D, Duranteau J. Renal arterial resistance in septic shock: effects of increasing mean arterial pressure with norepinephrine on the renal resistive index assessed with Doppler ultrasonography. Intensive Care Med 2007; 33:1557–1562.
Eur J Anaesthesiol 2015; 32:37–43 Copyright © European Society of Anaesthesiology. Unauthorized reproduction of this article is prohibited.
