Risk score
Formula
Glasgow aneurysm score
Age + 17 for shock + 7 for myocardial disease + 10 for cerebrovascular disease + 14 for renal disease
Hardman index
Score from 1 to 5 depending on number of 5 risk factors present
Risk factors: age >76, ECG ischemia, Cr > 0.19 mmol/L, LOC, Hgb <9
Vancouver score
Ex/(1+ Ex) where x = (−3.44) + [sum of coefficients of significant variables]
Variable
Coefficient
Age
0.062 × age
Reduced consciousness: yes
1.14
Reduced consciousness: no
−1.14
Cardiac arrest: yes
.6
Cardiac arrest: no
−.6
Edinburg ruptured aneurysm score
Score from 0 to 3 summing number of 3 risk factors present
Risk factors: Hob <9, preoperative GCS <15, preop BP <90
Vascular study group of New England RAAA risk score
Score summing integer weights of 4 risk factors:
Age >76: 2
Cardiac arrest: 2
Loss of consciousness: 1
Suprarenal clamp: 1
Total VSGNE RAAA score = 0 − 6
University of Washington RAAA risk score
Score from 0 to 4 summing number of risk factors present
Risk factors: preoperative SBP ever <70, pH < 7.2, age >76, creatinine >2 g/dL
Glasgow Aneurysm Score (GAS)
The Glasgow score identifies age, “shock,” myocardial disease, cerebrovascular disease, and renal disease as the primary predictors of inpatient mortality after open repair of RAAA. The score is additive, with increasing points assigned for the presence of the preceding risk factors (Table 8.1). Application of the GAS resulted in mortality estimates ranging from 0 % for scores below “70 GaAs” to 80 % for scores greater than “95 GaAs” [13].
Hardman Index
The Hardman index identified age >76, electrocardiographic ischemia, creatinine greater than 190 μmoles/liter, loss of consciousness, and hemoglobin less than 9 g/dL as predictive of mortality (Table 8.1). Patients who had three or more factors present had a 100 % 30-day mortality. It must be noted that only eight patients in the data for this study had three or four factors present and no patients had all five factors present [8].
Vancouver Score
The Vancouver score identified age, preoperative unconsciousness, and cardiac arrest as predictive of mortality. The authors created an equation for predicting mortality based upon the coefficients for each independent predictor and derived the variable logistic regression model. Based on application of this equation, 100 % mortality could not be predicted. However, for patients who had sustained cardiac arrest and who were unconscious, the predicted mortality was 88 %, 93 %, 96 %, and 99 % for patients 60, 70, 80, and 90 years of age, respectively [9].
Edinburg Ruptured Aneurysm Score
Developed in 2007, the Edinburg Ruptured Aneurysm Score was developed which predicted mortality based on the presence or absence of three preoperative variables available on admission including hemoglobin less than 9 g/dL, preoperative Glasgow Coma Scale less than 15, and preoperative systolic blood pressure less than 90 mmHg. Tambyraja and colleagues found that there was an 80 % mortality rate in patients who presented with all three factors, although all three factors were only present in five out of 105 patients [10].
VSGNE RAAA Risk Score
This VSGNE RAAA Risk Score, the first derived in a United States cohort, identified age greater than 76, cardiac arrest, loss of consciousness, and the use of a suprarenal clamp as the primary independent predictors of in-hospital mortality on logistic regression analysis. Based on the presence or absence of these variables, an integer score ranging from 0 to 6 can be derived (Table 8.2). This score was demonstrated to accurately predict in-hospital mortality after open repair of RAAA with excellent calibration (X 2 = 1.96, P = .85) and discrimination (C-statistic = .79) in 242 patients from the Vascular Study Group of New England between 2003 and 2009. A VSGNE RAAA Risk Score of four was associated with an 80 % mortality, while a score ≥5 was associated with an 87 % mortality (Fig. 8.1). No score was associated with 100 % mortality. The VSGNE Risk Score was subsequently validated in the larger cohort of 1165 patients undergoing open and endovascular repair in the Vascular Quality Initiative (VQI) [4].
Table 8.2
Preoperative and intraoperative variables and in-hospital survival
Creatinine | Clamp site | Urine output | Probability survival | 95 % range |
|---|---|---|---|---|
≤1.3 | Infrarenal | ≥200 | 90 | 76–96 |
≤1.3 | Infrarenal | 1–199 | 76 | 52–90 |
>1.3 | Infrarenal | ≥200 | 71 | 53–85 |
≤1.3 | Suprarenal | ≥200 | 65 | 33–87 |
≤1.3 | Infrarenal | 0 | 52 | 21–82 |
>1.3 | Infrarenal | 1–199 | 46 | 32–61 |
≤1.3 | Suprarenal | 1–199 | 39 | 15–69 |
>1.3 | Suprarenal | ≥200 | 33 | 9–71 |
>1.3 | Infrarenal | 0 | 23 | 9–46 |
≤1.3 | Suprarenal | 0 | 18 | 4–50 |
>1.3 | Suprarenal | 1–199 | 15 | 4–53 |
>1.3 | Suprarenal | 0 | 6 | 1–24 |
Fig. 8.1
In-hospital mortality rate after open RAAA repair according to VSGNE RAAA Risk Score (From Robinson et al. [11])
The University of Washington RAAA Score
A new score was recently proposed and presented by Garland, Starnes, and colleagues from University of Washington [12]. Age greater than 76 years, pH less than 7.2, systolic blood less than 70 mmHg, and preoperative creatinine >2.0 mg/dL were found to be most predictive of 30-day mortality. One point each is assigned to the presence of each variable for a total score ranging between 0 and 4. This score is unique in that it predicted 100 % mortality in the cohort from which it was derived. In 303 patients, a score of four was associated with 100 % mortality irrespective of whether the patients underwent open or endovascular repair.
Variables Consistently Associated with High Mortality
Age
It is clear from all the existing prediction models that age is directly related to risk of mortality after repair of ruptured AAA. While this is intuitive, no study has identified a specific age threshold which can predict certain death. In fact, on multivariable analysis controlling for other predictors of mortality, our group found that patients older than 85 had lower mortality than patients aged 80–85 [11]. One must recognize that, the majority of current risk scores utilize a specific age cutoff is a dichotomous variables in their risk prediction models. The Hardman Index, the VSGNE RAAA Risk Score, and the University of Washington RAAA Risk Score all determined age >76 to be significant predictors of death. The VSGNE RAAA authors performed a threshold analysis to determine the optimal age for accurately predicting mortality and determined that age >76 was in fact the optimal threshold for analysis. A variety of other series have identified advanced age as predictive of death. Remarkably, all of these studies identify an age threshold ranging between 70 and 80 years of age [15–17]. It is clear that the physiologic reserve required to survive a ruptured abdominal aortic aneurysm diminishes with age. While it is impossible to identify an absolute age above which repair should not be offered in all instances, it is clear that patients in the eighth or ninth decade of life who have significant premorbid conditions or present in shock have little chance of survival.
Indices of Shock
As expected, all prediction models have identified indices of severe shock including hypotension, cardiac arrest, and loss of consciousness as predictive of mortality. Often the symptoms and signs coexist, making identification of the most accurate predictor of mortality difficult.
Severe Hypotension
In addition to the Edinburg Ruptured Aneurysm Score and the University of Washington RAAA Score, a variety of other studies have identified hypotension as a significant predictor of mortality [18]. In virtually all studies, hypotension was defined as a systolic blood pressure less than 70–90 mmHg. It must be noted that the majority of studies investigating predictors of mortality were conducted before the use of permissive hypotension (“hypotensive hemostasis”) was recommended and frequently utilized. This strategy, while not exhaustively investigated, has been increasingly recommended over the last decade [19]. In contrast to a traditional approach to preoperative resuscitation which emphasized aggressive fluid resuscitation to achieve normal attention, permissive hypotension involves maintenance of a systolic blood pressure between 80 and 100 mmHg as long as there is evidence of end-organ perfusion such as maintained consciousness. Until aortic control can be obtained, permissive hypotension is thought to prevent ongoing blood loss secondary to dilution, coagulopathy, and clot disruption that may occur with aggressive fluid resuscitation. Based on the most current evidence, short periods of moderate hypotension (systolic blood pressure less than 80 mmHg) are likely not associated with increased mortality as long as the patient is mentating which is indicative of adequate organ perfusion [20, 21].
Cardiac Arrest
Both the VSGNE RAAA Risk Score and the Vancouver Score identified preoperative cardiac arrest as strong independent predictors of mortality. Similarly, the Hardman Index identifies electrocardiographic ischemia as significant predictor mortality. In addition to these models, multiple additional studies have likewise identified cardiac arrest as predictive of mortality [16, 22, 23].
Loss of Consciousness
The Hardman Index, Vancouver Score, VSGNE RAAA Risk Score, and the Edinburg Ruptured Aneurysm Score all identify diminished consciousness on presentation as predictive of mortality. Other studies have likewise identified loss of consciousness as predictive of mortality [24, 25]. Loss of consciousness represents end-organ malperfusion that will remain an important predictive sign for the clinician, especially when permissive hypotension is employed.
Renal Insufficiency or Increased Creatinine
A number of studies have demonstrated that increased creatinine is associated with mortality, including the Glasgow Aneurysm Score, the Hardman Index, and the University of Washington RAAA Score. Based on these studies and others, a creatinine greater than 180–190 μmol/l (2–2.1 mg/dL) appears to be the threshold associated with increased mortality [15, 26].The detrimental impact of preoperative renal insufficiency is not surprising considering the lethal nature of postoperative renal failure to which these patients are strongly predisposed.
Severe Anemia
Severe blood loss and subsequent disruption of oxygen delivery to vital organs intuitively would predict poor outcome. Accordingly, low hematocrit or hemoglobin (defined in most studies as either as Hgb < 9 g/dL or Hgb < 10 g/dL) has been shown to be another factor predictive of mortality in the Hardman Index, the Edinburg Ruptured Aneurysm Score, and other studies [24, 27]. No studies have investigated whether a hemoglobin less than nine can accurately predict certain death.
Intraoperative Variables
Some studies have analyzed the impact of intraoperative variables on the ability to predict survival [14, 18]. Johnston and colleagues reported that the site of the aorta cross-clamping (suprarenal versus infrarenal), the volume of blood transfusion, and a drop in urine output help predict early survival. For example, a patient who required a suprarenal cross-clamping, was transfused more than 3500 mL of blood, and had no intraoperative urine output was predicted to have a 3 % survival. On multivariable model including preoperative and intraoperative variables, preoperative serum creatinine, cross-clamp site, and intraoperative urine output would predict survival [14] (Table 8.2). A number of other studies, including the VSGNE RAAA Risk Score, also identified the need for a suprarenal clamp as predictive of mortality [11, 28]. In some instances, the need for a suprarenal clamp is not known until operation is undertaken and the abdomen is explored. The need is then based on whether or not patient has hemodynamic collapse and the location of the retroperitoneal hematoma. However, in modern algorithms for RAAA management, 78–93 % of patients who undergo RAAA repair have a preoperative computed tomographic scan [19, 29].The need for a suprarenal clamp can often be readily determined preoperatively. The impact of suprarenal clamp on mortality is an especially important consideration in the current management of RAAA. Increasingly, patients with a suitable infrarenal neck are being offered endovascular repair [4, 5]. Therefore, as time goes on, almost all patients requiring open repair may be unsuitable for endovascular repair and may require a suprarenal clamp. Increased blood loss has also been identified as independently predictive of mortality in an analysis by Panneton et al. and in the derivation of the ruptured AAA POSSUM (RAAA POSSUM) score, which combines both preoperative physiologic and operative variables in predicting mortality [18, 30].
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
