Abstract

Background

Current approaches to risk assessment before major surgery have important limitations. The aim of this pilot study was to compare predictive accuracy of preoperative scoring systems, plasma biomarkers, and cardiopulmonary exercise testing (CPET) for complications after major non-cardiac surgery.

Methods

Single-centre, observational study of patients aged ≥40 yr undergoing major elective non-cardiac surgery. Before surgery, risk scores were calculated and blood samples collected for measurement of plasma biomarkers. Patients underwent CPET for measurement of anaerobic threshold (AT) and peak oxygen consumption (forumla peak). After surgery, patients were followed for 28 days to evaluate complications and major adverse cardiac events (MACE). Data are presented as area under the receiver operating characteristic curve (AUROC) with 95% confidence intervals.

Results

A total of 100 patients were recruited between April 2009 and October 2010; 17 of whom did not proceed to surgery. CPET variables suggested good predictive accuracy for MACE [AT: AUROC 0.83 (0.69–0.96); forumla peak AUROC 0.81 (0.69–0.96)] and poor predictive accuracy for all complications [AT: AUROC 0.64 (0.52–0.77); forumla peak AUROC 0.64 (0.52–0.77)]. There was a trend towards predictive accuracy of the plasma biomarkers B-type natriuretic peptide and estimated glomerular filtration rate (calculated from serum creatinine) for MACE but not all complications. C-reactive protein, ASA score, and revised cardiac risk index had little or no predictive value.

Conclusions

These pilot data suggest that CPET and plasma biomarkers may improve risk assessment before surgery. Only large clinical studies can confirm this observation and define the optimal use of these tests in clinical practice.

Editor's key points

  • The ability of cardiopulmonary exercise testing (CPET) and plasma biomarkers to predict complications after major surgery is uncertain.

  • This small pilot study found that CPET and some biomarkers might be useful in predicting major adverse cardiac events within 28 days.

  • However, the study was not powered to detect differences between the scoring systems, biomarkers, or other measures used.

  • Larger studies are required.

Postoperative complications represent an important cause of avoidable morbidity and mortality.1–4 Approximately 234 million patients undergo surgery worldwide each year, around 25% of whom will develop complications.4–6 Estimates of hospital mortality rates for in-patient non-cardiac surgery may be as high as 2–4%.1–6 Long-term survival is significantly reduced after surgery for those patients who develop complications but survive to leave hospital.5,7 Importantly, around 80% of deaths occur among a subgroup of patients who can be identified as being at high risk before surgery is performed.2,3

Typical risk factors for postoperative complications include advanced age, co-morbid disease, and major surgery. However, evidence suggests that most patients who meet these high-risk criteria do not receive additional care to improve the prospect of good postoperative outcomes.1–3 One reason for this may be the subjective nature of the current approach to clinical assessment of risk which often requires a doctor to make the case for additional care for individual high-risk patients. An alternative approach may be the use of objective measures of risk including scoring systems,8,9 plasma biomarkers,10–21 and cardiopulmonary exercise testing (CPET).22–26 Each of these approaches has strengths and limitations in the assessment of individual patient risk. Issues of feasibility, accuracy, and cost have led to debate over their relative merits and utility in clinical practice. However, few data are available to provide direct comparisons between the most promising candidate methods. Our aim was to undertake a pilot study to compare the prognostic accuracy of CPET and the plasma biomarkers B-type natriuretic peptide (BNP), estimated glomerular filtration rate (eGFR; calculated from serum creatinine), and C-reactive protein (CRP) to the most commonly used clinical scoring systems to predict complications among patients undergoing elective major non-cardiac surgery.

Methods

Patients

This was a single-centre observational study of adult patients aged 40 yr and over, undergoing major elective non-cardiac surgery at the Royal London Hospital, London, UK. The study was approved prospectively by the East London and City Research ethics committee (reference 09/H0704/23) and written informed consent was given by all patients before enrolment. Exclusion criteria were unsuitability for CPET based on the American Thoracic Society guidelines,27 pregnancy, and refusal of consent. Potential participants were screened by local investigators having been identified in a consultant-led preoperative assessment clinic. The decision to offer surgery and allocation of postoperative critical care were made on the basis of established clinical protocols which did not include consideration of CPET data. However, CPET data were made available to anaesthetic staff on specific request. Plasma biomarker data were not made available to clinical staff. ASA physical status score was documented on the anaesthetic chart for all patients. Revised cardiac risk index (RCRI) is not routinely calculated by clinical staff.

Preoperative data

Baseline clinical data were collected allowing calculation of RCRI and the Portsmouth Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity (P-POSSUM).8,28 The ASA score (assessed by the attending anaesthetist) was also noted. Patients were then asked to lie quietly in the supine position before venepuncture. Within 30 min of collection, blood samples were centrifuged at 4300 rpm for 10 min and stored at −80°C for subsequent analysis. BNP was analysed by immunoassay (Architect i2000SR, Abbott Diagnostics, USA). The measuring interval was 10–5000 pg ml−1, the average analytical coefficient of variation was <5%, and the reference interval was <135 pg ml−1. CRP was analysed using an immunoturbidimetric method (Architect c16000, Abbott Diagnostics) with a measuring interval of 0.1–160 mg litre−1. The average analytical coefficient of variation was <2%, and the reference interval was <7 mg litre−1. eGFR was calculated from age and serum creatinine with adjustment for ethnicity using the modification of diet in renal disease equation.29 Creatinine was analysed using a kinetic Jaffe reaction assay (Roche Diagnostics, Switzerland). On the day of exercise testing, patients were fasted for 2 h and refrained from caffeine in accordance with American Thoracic Society guidelines.27 They were seated on a computer-controlled, electromagnetically braked cycle ergometer and cycled at a cadence of 55–65 rpm. Subjects wore a tight facemask to permit continuous measurements of ventilation, oxygen consumption (forumla), and carbon dioxide production (forumla) in expired gas. These measurements allowed estimation of peak oxygen consumption (forumla peak) and anaerobic threshold (AT) using a modified V-slope method. Arterial pressure, 12-lead ECG, and arterial pulse oximetry readings were taken at rest. During exercise, arterial pressure was measured every 3 min while pulse oximetry and ECG rhythm were monitored continuously. Subjects performed an incremental ramp test to the limit of tolerance. A ramp slope, based on gender, age, and estimated physical fitness of the subject, was set in order to obtain a test of ∼10 min duration. The limit of tolerance (maximum) was defined as the point at which the subject could not maintain a pedalling cadence of 55–65 rpm despite encouragement. All tests were supervised by a physician trained in the analysis of these tests. Criteria for stopping or preventing a subject from being tested were also based on American Thoracic Society guidelines.27

Postoperative data

After surgery, patients were followed up for 28 days by a research assistant who was not present at the exercise test. Data were collected describing pre-defined postoperative complications (Supplementary material) including major adverse cardiac events (MACE) (defined as myocardial infarction, cardiogenic pulmonary oedema, cardiac arrest, or complete heart block),8 duration of hospital stay, and mortality. The P-POSSUM score was completed for each patient immediately after surgery.

Statistical analysis

This was a pilot study to evaluate the feasibility and inform the design of a larger observational study. No formal sample size calculation was made. Instead, we planned to recruit 100 patients undergoing major surgery. Patients who were enrolled but did not subsequently undergo surgery were prospectively excluded from the analysis. Receiver operating characteristic (ROC) curves were constructed by first tabulating and then plotting the sensitivity and specificity of the test result at various cut-offs, then calculating the area under these curves (AUROC) to quantify overall prognostic discrimination for MACE and all complications. Categorical variables were tested with the Fisher's exact test. AUROC is presented with 95% confidence intervals. Optimal threshold values and associated sensitivities and specificities are presented for variables with AUROC >0.8 for the corresponding outcome (good predictive accuracy). Other data are presented as mean (sd) where normally distributed or median (IQR) where not normally distributed. Statistical analysis was performed using GraphPad Prism version 4.0 (GraphPad Software, USA).

Results

One hundred patients were recruited between April 2009 and October 2010. After preoperative assessment, 17 patients did not proceed to surgery (Fig. 1 and Supplementary Table S1). A total of 83 patients who underwent CPET proceeded to surgery all of whom were followed up for 28 days. The median operating time was 270 min (198–378). All patients underwent general anaesthesia, 64 (77%) in combination with epidural analgesia. Patient characteristics are presented in Table 1. Forty patients developed complications (48%); nine of whom developed MACE (11%) (Table 2). There were three deaths, all of which followed MACE. The median duration of hospital stay was 12 days (7–22) and three patients died. For those patients alive at day 28, the median number of critical care free days was 25 (21–26). Clinicians made few requests for CPET data, but unfortunately, we did not document the frequency of these requests and did not record any qualitative data on how they influenced patient care.

Table 1

Patient characteristics. Some patients experienced both MACE and non-MACE complications. Data presented as mean (sd), median (IQR), n (%), and mean (range) for age. P-POSSUM data presented as probability (%) of patient experiencing morbidity with the corresponding sd. MACE, major adverse cardiac events; P-POSSUM, Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity

 No complications (n=43) Complications (n=40) No MACE (n=74) MACE (n=9) 
Baseline risk factors 
 Age (yr) 67 (60–75) 69 (63–75) 68 (62–75) 71 (67–75) 
 Male sex (%) 30 (70%) 30 (75%) 53 (72%) 7 (78%) 
 P-POSSUM morbidity probability (%) 50% (23) 55% (27) 50% (24) 71% (24) 
 Ischaemic heart disease (%) 10 (23%) 13 (33%) 18 (24%) 5 (56%) 
 Heart failure (%) 8 (19%) 2 (5%) 9 (12%) 1 (11%) 
 Diabetes (%) 5 (12%) 9 (23%) 11 (15%) 3 (33%) 
 Chronic lung disease (%) 7 (16%) 10 (25%) 14 (19%) 3 (33%) 
 Chronic kidney disease (%) 0 (0%) 4 (10%) 1 (1%) 3 (33%) 
Surgical procedure category 
 Vascular surgery 14 (33%) 14 (35%) 24 (32%) 4 (44%) 
 Upper gastrointestinal surgery 5 (12%) 7 (18%) 11 (15%) 1 (11%) 
 Hepato-biliary surgery 18 (42%) 10 (25%) 26 (35%) 2 (22%) 
 Lower gastrointestinal surgery 6 (14%) 9 (23%) 13 (18%) 2 (22%) 
Survival 
 Deaths (%) 0 (0%) 3 (8%) 0 (0%) 3 (33%) 
 No complications (n=43) Complications (n=40) No MACE (n=74) MACE (n=9) 
Baseline risk factors 
 Age (yr) 67 (60–75) 69 (63–75) 68 (62–75) 71 (67–75) 
 Male sex (%) 30 (70%) 30 (75%) 53 (72%) 7 (78%) 
 P-POSSUM morbidity probability (%) 50% (23) 55% (27) 50% (24) 71% (24) 
 Ischaemic heart disease (%) 10 (23%) 13 (33%) 18 (24%) 5 (56%) 
 Heart failure (%) 8 (19%) 2 (5%) 9 (12%) 1 (11%) 
 Diabetes (%) 5 (12%) 9 (23%) 11 (15%) 3 (33%) 
 Chronic lung disease (%) 7 (16%) 10 (25%) 14 (19%) 3 (33%) 
 Chronic kidney disease (%) 0 (0%) 4 (10%) 1 (1%) 3 (33%) 
Surgical procedure category 
 Vascular surgery 14 (33%) 14 (35%) 24 (32%) 4 (44%) 
 Upper gastrointestinal surgery 5 (12%) 7 (18%) 11 (15%) 1 (11%) 
 Hepato-biliary surgery 18 (42%) 10 (25%) 26 (35%) 2 (22%) 
 Lower gastrointestinal surgery 6 (14%) 9 (23%) 13 (18%) 2 (22%) 
Survival 
 Deaths (%) 0 (0%) 3 (8%) 0 (0%) 3 (33%) 
Table 2

Incidence of individual pre-defined complications (Supplementary material) occurring during 30 days after surgery. Some patients experienced more than one complication. Data presented as n (%)

Complication Number of patients (n = 83) 
Pneumonia 17 (21) 
Wound infection 17 (21) 
Paralytic ileus 9 (11) 
Acute kidney injury 8 (10) 
Myocardial infarction 6 (7) 
Anastomotic leak 5 (6) 
Cardiogenic pulmonary oedema 4 (5) 
Haemorrhage 3 (4) 
Limb ischaemia 3 (4) 
Urinary tract infection 2 (2) 
Stroke/transient ischaemic attack 2 (2) 
Cardiac arrest 2 (2) 
Other 2 (2) 
Complication Number of patients (n = 83) 
Pneumonia 17 (21) 
Wound infection 17 (21) 
Paralytic ileus 9 (11) 
Acute kidney injury 8 (10) 
Myocardial infarction 6 (7) 
Anastomotic leak 5 (6) 
Cardiogenic pulmonary oedema 4 (5) 
Haemorrhage 3 (4) 
Limb ischaemia 3 (4) 
Urinary tract infection 2 (2) 
Stroke/transient ischaemic attack 2 (2) 
Cardiac arrest 2 (2) 
Other 2 (2) 
Fig 1

Patient recruitment and follow-up.

Fig 1

Patient recruitment and follow-up.

Univariate analysis of the association with outcomes for the candidate prognostic markers is presented in Table 3. For preoperative scoring systems, both ASA score [AUROC 0.60 (0.48–0.72) for all complications; 0.68 (0.49–0.87) for MACE] and RCRI [0.53 (0.40–0.65) for all complications; 0.68 (0.49–0.86) for MACE] were poor predictors of MACE and not predictive of all complications. For plasma biomarkers, CRP was not predictive of clinical outcome while BNP and eGFR had fair predictive accuracy for MACE but were not predictive of all complications (Fig. 2). The predictive value of CPET-derived variables was greater than that of scoring systems and plasma biomarkers. Both AT (sensitivity 75%, specificity 85% for an optimal threshold value of 10.6 ml min−1 kg−1) and forumla peak (sensitivity 88%, specificity 69% for an optimal threshold value of 14.0 ml min−1 kg−1) were good predictors of MACE and poor predictors of all complications (Fig. 3).

Table 3

Univariate analysis of preoperative risk assessment markers for all complications and MACE. Data presented as median (IQR), mean (sd), or n (%). RCRI, revised cardiac risk index; BNP, B-type natriuretic peptide; CRP, C-reactive protein; eGFR, estimated glomerular filtration rate; AT, anaerobic threshold

 No complications (n=43) Complications (n=40) P-value No MACE (n=74) MACE (n=9) P-value 
ASA I and II 18 (45%) 10 (25%) 0.16 27 (36%) 1 (11%) 0.26 
ASA III and IV 25 (55%) 30 (75%)  47 (56%) 8 (89%)  
RCRI I and II 26 (60%) 23 (58%) 0.83 47 (56%) 2 (33%) 0.03 
RCRI III and IV 17 (40%) 17 (43%)  27 (36%) 7 (78%)  
BNP (pg ml−126 (16–71) 48 (21–106) 0.15 28 (17–79) 89 (43–345) 0.0001 
CRP (mg litre−14.9 (2.4–12.1) 2.8 (1.7–7.0) 0.28 3.5 (1.9–10.1) 5.0 (2.8–17.1) 0.87 
eGFR (ml min−179 (19) 77 (28) 0.59 80 (23) 61 (21) 0.02 
AT (ml min−1 kg−113.4 (2.9) 12.0 (2.8) 0.03 13.1 (2.9) 10.1 (1.8) 0.006 
forumla peak (ml min−1 kg−117.3 (4.5) 14.9 (3.8) 0.02 16.6 (4.3) 12.3 (2.3) 0.007 
 No complications (n=43) Complications (n=40) P-value No MACE (n=74) MACE (n=9) P-value 
ASA I and II 18 (45%) 10 (25%) 0.16 27 (36%) 1 (11%) 0.26 
ASA III and IV 25 (55%) 30 (75%)  47 (56%) 8 (89%)  
RCRI I and II 26 (60%) 23 (58%) 0.83 47 (56%) 2 (33%) 0.03 
RCRI III and IV 17 (40%) 17 (43%)  27 (36%) 7 (78%)  
BNP (pg ml−126 (16–71) 48 (21–106) 0.15 28 (17–79) 89 (43–345) 0.0001 
CRP (mg litre−14.9 (2.4–12.1) 2.8 (1.7–7.0) 0.28 3.5 (1.9–10.1) 5.0 (2.8–17.1) 0.87 
eGFR (ml min−179 (19) 77 (28) 0.59 80 (23) 61 (21) 0.02 
AT (ml min−1 kg−113.4 (2.9) 12.0 (2.8) 0.03 13.1 (2.9) 10.1 (1.8) 0.006 
forumla peak (ml min−1 kg−117.3 (4.5) 14.9 (3.8) 0.02 16.6 (4.3) 12.3 (2.3) 0.007 
Fig 2

ROC curves for eGFR and the plasma biomarkers BNP and CRP in the preoperative prediction of all complications and MACE. Area under ROC curves: eGFR 0.58 (0.45–0.71) for complications and 0.72 (0.54–0.90) for MACE, BNP 0.60 (0.48–0.73) for all complications and 0.75 (0.59–092) for MACE; CRP 0.61 (0.48–0.74) for all complications and 0.56 (0.36–0.76) for MACE.

Fig 2

ROC curves for eGFR and the plasma biomarkers BNP and CRP in the preoperative prediction of all complications and MACE. Area under ROC curves: eGFR 0.58 (0.45–0.71) for complications and 0.72 (0.54–0.90) for MACE, BNP 0.60 (0.48–0.73) for all complications and 0.75 (0.59–092) for MACE; CRP 0.61 (0.48–0.74) for all complications and 0.56 (0.36–0.76) for MACE.

Fig 3

ROC curves for the CPET indices AT and forumla peak in the preoperative prediction of all complications and MACE. Area under ROC curves: AT 0.64 (0.52–077) for all complications and 0.83 (0.69–0.96) for MACE; forumla peak 0.64 (0.52–0.77) for all complications and 0.81 (0.68–0.93) for MACE.

Fig 3

ROC curves for the CPET indices AT and forumla peak in the preoperative prediction of all complications and MACE. Area under ROC curves: AT 0.64 (0.52–077) for all complications and 0.83 (0.69–0.96) for MACE; forumla peak 0.64 (0.52–0.77) for all complications and 0.81 (0.68–0.93) for MACE.

Discussion

The principal findings of this small pilot study are that CPET and plasma biomarkers may have adequate prognostic accuracy in predicting which patients will develop complications after major non-cardiac surgery. CPET and biomarkers appeared to be more effective in the prediction of MACE than for unselected complications. Of the plasma biomarkers, BNP appeared to be the strongest prognostic marker followed by eGFR, while CRP was not predictive of any clinical outcomes. ASA score and RCRI were not predictive of overall complications in this small cohort and only poorly predictive of MACE. Our findings do not confirm the clinical utility of these tests but do suggest that a multi-centre study to confirm and extend these findings would be both feasible and worthwhile. These data may help in the estimation of event rates, and the rates of drop-out before and after surgery.

There is a growing evidence base describing the use of various tests for the prediction of postoperative outcomes. However, in many cases, this consists of small, single-centre case series with methodological limitations. In the case of CPET, there are few reports of the predictive accuracy of this test for complications after surgery.22–26 Two systematic reviews have suggested that CPET-derived variables are valid predictors of complications after non-cardiac surgery which may outperform alternative methods.30,31 Recent evidence suggests that CPET-derived variables may also be predictive of long-term postoperative mortality.32 However, there is substantial heterogeneity in the literature. Two studies, both notable for being clinician blinded, provide data describing the accuracy of CPET in predicting postoperative complications.33,34 In the first study, AUROC of only 0.59 for AT and 0.66 for forumla peak are described,33 while investigators in the more recent study report AUROC of 0.85 for AT (AUROC for forumla peak not reported).34 Our findings are consistent with those of published studies suggesting a role for BNP in the preoperative assessment of risk.11–19 The findings of a systematic review also suggest a role for this biomarker, in particular in the prediction of MACE.35 There are less published data describing the potential for eGFR in risk assessment before non-cardiac surgery patients but the available data are encouraging,20,21 and it is interesting to note that preoperative serum creatinine was an independent risk factor for MACE in the work which defined RCRI.8 We identified two publications exploring risk assessment with CRP and it is not possible to confirm the utility of this biomarker at the current time.10,19 The findings of one study suggest some predictive value of preoperative CRP measurement,10 while those of a more recent study suggested a weak relationship in keeping with our findings in the current study.19 Recent data have demonstrated the value of the ASA score in describing the severity of co-morbid disease in a large patient cohort.1 However, the limited number of categories limits the predictive value of the ASA score. This is especially the case because the five ASA categories were collapsed together into two categories to improve prognostic accuracy. We are not aware of any previous evaluation of the predictive value of the ASA score for postoperative complications. The current literature for preoperative prediction of postoperative complications is characterized by a number of common limitations for all the methods described here; most importantly, small sample size, a lack of clinician blinding, and retrospective methodology.36 None of the published studies has compared all the candidate risk assessment tools in a single large patient cohort.

The strength of the current study is that it is, to our knowledge, the first comparison of CPET, plasma biomarkers, and scoring systems in the evaluation of the risk of postoperative complications before non-cardiac surgery. However, this work also has a number of important limitations. The sample size was small and susceptible to selection bias. Patients were already considered at sufficient risk to require referral to a consultant-led preoperative assessment clinic. Unfortunately, we did not document the frequency of clinician requests to view CPET data or how this may have influenced patient care. The profile of patients who developed MACE appeared to favour prediction of this outcome whereas the poor predictive accuracy for all complications may reflect the wider presence of risk factors for this outcome. Even among those patients at high risk, not every patient will develop complications because this outcome is also influenced by factors which occur during and after surgery. The high event rate for all complications suggests a uniformly high level of risk across all patients which might reduce the opportunity to discriminate between patients in terms of the preoperative characteristics associated with poor postoperative outcome. We did not perform a multivariate analysis because of the small sample size and anticipated low event rates which limit the number of variables which can be included in such models. These problems can only be resolved in a much larger clinical study with a wider distribution of perioperative risk to provide more robust estimates of predictive value and the associated discriminatory thresholds.

Conclusions

These pilot data suggest that CPET-derived variables and plasma biomarkers have the potential to predict complications, in particular cardiac complications, within 28 days of surgery. Large multi-centre studies are now required to confirm the clinical utility of these prognostic tests. This would allow evaluation of each category of prognostic test (scoring systems, plasma biomarkers, and CPET biomarkers) and define how and if multiple tests may be used effectively in an integrated assessment of risk of MACE, overall complications, and postoperative death. Importantly, an integrated approach has the potential for low cost risk screening for all patients allowing both an individual risk assessment for each patient and risk-adjusted evaluation of outcomes using multivariate models, facilitating effective large-scale clinical audit and benchmarking between centres. These pilot data confirm the feasibility of large clinical studies of this type.

Supplementary material

Supplementary material is available at British Journal of Anaesthesia online.

Authors’ contributions

R.M.P. formulated the hypothesis and developed the protocol. Data collection was performed by S.J. and A.S. Biochemical analysis and data interpretation was performed by M.F. and G.O'B. S.J., S. Jh., and R.M.P. performed the data analysis and drafted the manuscript for critical revision by all authors and approval of the final version.

Declaration of interest

None declared.

Funding

This work was supported by departmental research funds. R.M.P. is a National Institute for Health Research (UK) Clinician Scientist and a Royal College of Anaesthetists Career Development Fellow.

References

1
Pearse
RM
Moreno
RP
Bauer
P
, et al.  . 
Mortality after surgery in Europe: a 7 day cohort study
Lancet
 , 
2012
, vol. 
380
 (pg. 
1059
-
65
)
2
Pearse
RM
Harrison
DA
James
P
, et al.  . 
Identification and characterisation of the high-risk surgical population in the United Kingdom
Crit Care
 , 
2006
, vol. 
10
 pg. 
R81
 
3
Jhanji
S
Thomas
B
Ely
A
Watson
D
Hinds
CJ
Pearse
RM
Mortality and utilisation of critical care resources amongst high-risk surgical patients in a large NHS trust
Anaesthesia
 , 
2008
, vol. 
63
 (pg. 
695
-
700
)
4
Ghaferi
AA
Birkmeyer
JD
Dimick
JB
Variation in hospital mortality associated with inpatient surgery
N Engl J Med
 , 
2009
, vol. 
361
 (pg. 
1368
-
75
)
5
Jencks
SF
Williams
MV
Coleman
EA
Rehospitalizations among patients in the Medicare fee-for-service program
N Engl J Med
 , 
2009
, vol. 
360
 (pg. 
1418
-
28
)
6
Weiser
TG
Regenbogen
SE
Thompson
KD
, et al.  . 
An estimation of the global volume of surgery: a modelling strategy based on available data
Lancet
 , 
2008
, vol. 
372
 (pg. 
139
-
44
)
7
Khuri
SF
Henderson
WG
DePalma
RG
Mosca
C
Healey
NA
Kumbhani
DJ
Determinants of long-term survival after major surgery and the adverse effect of postoperative complications
Ann Surg
 , 
2005
, vol. 
242
 (pg. 
326
-
41
)
8
Lee
TH
Marcantonio
ER
Mangione
CM
, et al.  . 
Derivation and prospective validation of a simple index for prediction of cardiac risk of major noncardiac surgery
Circulation
 , 
1999
, vol. 
100
 (pg. 
1043
-
9
)
9
Saklad
M
Grading of patients for surgical procedures
Anesthesiology
 , 
1941
, vol. 
2
 (pg. 
281
-
4
)
10
Ackland
GL
Scollay
JM
Parks
RW
de Beaux
I
Mythen
MG
Pre-operative high sensitivity C-reactive protein and postoperative outcome in patients undergoing elective orthopaedic surgery
Anaesthesia
 , 
2007
, vol. 
62
 (pg. 
888
-
94
)
11
Rajagopalan
S
Croal
BL
Bachoo
P
Hillis
GS
Cuthbertson
BH
Brittenden
J
N-terminal pro B-type natriuretic peptide is an independent predictor of postoperative myocardial injury in patients undergoing major vascular surgery
J Vasc Surg
 , 
2008
, vol. 
48
 (pg. 
912
-
7
)
12
Gibson
SC
Payne
CJ
Byrne
DS
Berry
C
Dargie
HJ
Kingsmore
DB
B-type natriuretic peptide predicts cardiac morbidity and mortality after major surgery
Br J Surg
 , 
2007
, vol. 
94
 (pg. 
903
-
9
)
13
Cuthbertson
BH
Amiri
AR
Croal
BL
, et al.  . 
Utility of B-type natriuretic peptide in predicting perioperative cardiac events in patients undergoing major non-cardiac surgery
Br J Anaesth
 , 
2007
, vol. 
99
 (pg. 
170
-
6
)
14
Dernellis
J
Panaretou
M
Assessment of cardiac risk before non-cardiac surgery: brain natriuretic peptide in 1590 patients
Heart
 , 
2006
, vol. 
92
 (pg. 
1645
-
50
)
15
Berry
C
Kingsmore
D
Gibson
S
, et al.  . 
Predictive value of plasma brain natriuretic peptide for cardiac outcome after vascular surgery
Heart
 , 
2006
, vol. 
92
 (pg. 
401
-
2
)
16
Feringa
HH
Bax
JJ
Elhendy
A
, et al.  . 
Association of plasma N-terminal pro-B-type natriuretic peptide with postoperative cardiac events in patients undergoing surgery for abdominal aortic aneurysm or leg bypass
Am J Cardiol
 , 
2006
, vol. 
98
 (pg. 
111
-
5
)
17
Mahla
E
Baumann
A
Rehak
P
, et al.  . 
N-terminal pro-brain natriuretic peptide identifies patients at high risk for adverse cardiac outcome after vascular surgery
Anesthesiology
 , 
2007
, vol. 
106
 (pg. 
1088
-
95
)
18
Yeh
HM
Lau
HP
Lin
JM
Sun
WZ
Wang
MJ
Lai
LP
Preoperative plasma N-terminal pro-brain natriuretic peptide as a marker of cardiac risk in patients undergoing elective non-cardiac surgery
Br J Surg
 , 
2005
, vol. 
92
 (pg. 
1041
-
5
)
19
Biccard
BM
Naidoo
P
de Vasconcellos
K
What is the best pre-operative risk stratification tool for major adverse cardiac events following elective vascular surgery? A prospective observational cohort study evaluating pre-operative myocardial ischaemia monitoring and biomarker analysis
Anaesthesia
 , 
2012
, vol. 
67
 (pg. 
389
-
95
)
20
Kertai
MD
Boersma
E
Bax
JJ
, et al.  . 
Comparison between serum creatinine and creatinine clearance for the prediction of postoperative mortality in patients undergoing major vascular surgery
Clin Nephrol
 , 
2003
, vol. 
59
 (pg. 
17
-
23
)
21
Ackland
GL
Moran
N
Cone
S
Grocott
MP
Mythen
MG
Chronic kidney disease and postoperative morbidity after elective orthopedic surgery
Anesth Analg
 , 
2011
, vol. 
112
 (pg. 
1375
-
81
)
22
Older
P
Hall
A
Hader
R
Cardiopulmonary exercise testing as a screening test for perioperative management of major surgery in the elderly
Chest
 , 
1999
, vol. 
116
 (pg. 
355
-
62
)
23
Carlisle
J
Swart
M
Mid-term survival after abdominal aortic aneurysm surgery predicted by cardiopulmonary exercise testing
Br J Surg
 , 
2007
, vol. 
94
 (pg. 
966
-
9
)
24
Nugent
AM
Riley
M
Megarry
J
O'Reilly
MJ
MacMahon
J
Lowry
R
Cardiopulmonary exercise testing in the pre-operative assessment of patients for repair of abdominal aortic aneurysm
Ir J Med Sci
 , 
1998
, vol. 
167
 (pg. 
238
-
41
)
25
McCullough
PA
Gallagher
MJ
Dejong
AT
, et al.  . 
Cardiorespiratory fitness and short-term complications after bariatric surgery
Chest
 , 
2006
, vol. 
130
 (pg. 
517
-
25
)
26
Nagamatsu
Y
Shima
I
Yamana
H
Fujita
H
Shirouzu
K
Ishitake
T
Preoperative evaluation of cardiopulmonary reserve with the use of expired gas analysis during exercise testing in patients with squamous cell carcinoma of the thoracic esophagus
J Thorac Cardiovasc Surg
 , 
2001
, vol. 
121
 (pg. 
1064
-
8
)
27
ATS/ACCP Statement on cardiopulmonary exercise testing
Am J Respir Crit Care Med
 , 
2003
, vol. 
167
 (pg. 
211
-
77
)
28
Prytherch
DR
Whiteley
MS
Higgins
B
Weaver
PC
Prout
WG
Powell
SJ
POSSUM and Portsmouth POSSUM for predicting mortality. Physiological and Operative Severity Score for the enUmeration of Mortality and morbidity
Br J Surg
 , 
1998
, vol. 
85
 (pg. 
1217
-
20
)
29
Levey
AS
Bosch
JP
Lewis
JB
Greene
T
Rogers
N
Roth
D
A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group
Ann Intern Med
 , 
1999
, vol. 
130
 (pg. 
461
-
70
)
30
Smith
TB
Stonell
C
Purkayastha
S
Paraskevas
P
Cardiopulmonary exercise testing as a risk assessment method in non cardio-pulmonary surgery: a systematic review
Anaesthesia
 , 
2009
, vol. 
64
 (pg. 
883
-
93
)
31
Hennis
PJ
Meale
PM
Grocott
MP
Cardiopulmonary exercise testing for the evaluation of perioperative risk in non-cardiopulmonary surgery
Postgrad Med J
 , 
2011
, vol. 
87
 (pg. 
550
-
7
)
32
Colson
M
Baglin
J
Bolsin
S
Grocott
MP
Cardiopulmonary exercise testing predicts 5 yr survival after major surgery
Br J Anaesth
 , 
2012
, vol. 
109
 (pg. 
735
-
41
)
33
Forshaw
MJ
Strauss
DC
Davies
AR
, et al.  . 
Is cardiopulmonary exercise testing a useful test before esophagectomy?
Ann Thorac Surg
 , 
2008
, vol. 
85
 (pg. 
294
-
9
)
34
Snowden
CP
Prentis
JM
Anderson
HL
, et al.  . 
Submaximal cardiopulmonary exercise testing predicts complications and hospital length of stay in patients undergoing major elective surgery
Ann Surg
 , 
2010
, vol. 
251
 (pg. 
535
-
41
)
35
Rodseth
RN
Padayachee
L
Biccard
BM
A meta-analysis of the utility of pre-operative brain natriuretic peptide in predicting early and intermediate-term mortality and major adverse cardiac events in vascular surgical patients
Anaesthesia
 , 
2008
, vol. 
63
 (pg. 
1226
-
33
)
36
Grocott
MP
Pearse
RM
Prognostic studies of perioperative risk: robust methodology is needed
Br J Anaesth
 , 
2010
, vol. 
105
 (pg. 
243
-
5
)

Supplementary data

Comments

0 Comments