Umbrella review and meta-analysis of antiplatelet therapy for peripheral artery disease.

BACKGROUND: The literature on antiplatelet therapy for peripheral artery disease has historically been summarized inconsistently, leading to conflict between international guidelines. An umbrella review and meta-analysis was performed to summarize the literature, allow assessment of competing safety risks and clinical benefits, and identify weak areas for future research. METHODS: MEDLINE, Embase, DARE, PROSPERO and Cochrane databases were searched from inception until January 2019. All meta-analyses of antiplatelet therapy in peripheral artery disease were included. Quality was assessed using AMSTAR scores, and GRADE analysis was used to quantify the strength of evidence. Data were pooled using random-effects models. RESULTS: Twenty-eight meta-analyses were included. Thirty-three clinical outcomes and 41 antiplatelet comparisons in 72 181 patients were analysed. High-quality evidence showed that antiplatelet monotherapy reduced non-fatal strokes (3 (95 per cent c.i. 0 to 6) fewer per 1000 patients), In symptomatic patients, it reduced cardiovascular deaths (8 (0 to 16) fewer per 1000 patients), but increased the risk of major bleeding (7 (3 to 14) more events per 1000). In asymptomatic patients, monotherapy reduced non-fatal strokes (5 (0 to 8) fewer per 1000), but had no other clinical benefit. Dual antiplatelet therapy caused more major bleeding after intervention than monotherapy (37 (8 to 102) more events per 1000), with very low-quality evidence of improved endovascular patency (risk ratio 4·00, 95 per cent c.i. 0·91 to 17·68). CONCLUSION: Antiplatelet monotherapy has minimal clinical benefit for asymptomatic peripheral artery disease, and limited benefit for symptomatic disease, with a clear risk of major bleeding. There is a lack of evidence to guide antiplatelet prescribing after peripheral endovascular intervention.

ANTECEDENTES: Históricamente, la literatura del tratamiento antiplaquetario en la enfermedad arterial periférica se ha sintetizado inconsistentemente, lo que ha dado lugar a divergencias entre las guías internacionales. Se efectuó una amplia revision con metaanálisis para sintetizar claramente la literatura, permitiendo evaluar los riesgos competitivos de seguridad y los beneficios clínicos, e identificar áreas poco claras susceptibles de futuras investigaciones. MÉTODOS: La búsqueda se realizó en las bases de datos MEDLINE, EMBASE, DARE, PROSPERO y Cochrane desde su inicio hasta enero de 2019. Se incluyeron todos los metaanálisis del tratamiento antiplaquetario en la enfermedad arterial periférica. Se estimó su calidad utilizando la puntuación Amstar y la consistencia de su evidencia mediante el sistema GRADE. Los datos se agruparon utilizando modelos de efectos aleatorios. RESULTADOS: Se incluyeron 28 metaanálisis. Se analizaron 33 resultados clínicos y 41 comparaciones antiplaquetarias en 72.181 pacientes. Una evidencia de alta calidad demostró que la monoterapia antiplaquetaria reducía los accidentes cerebrovasculares no mortales y la muerte cardiovascular en pacientes sintomáticos (3 y 8 veces menos por 1.000 pacientes, respectivamente, i.c. del 95% 0-6 y 0-16), pero aumentó el riesgo de hemorragia grave (7 veces más por 1.000, i.c. del 95% 3-14). En pacientes asintomáticos, la monoterapia redujo los accidentes cerebrovasculares no mortales (5 veces menos por 1.000, i.c. del 95% 0-8) sin otro beneficio clínico. El doble tratamiento antiplaquetario causó más hemorragias graves después de cualquier intervención que la monoterapia (37 veces más por 1.000, i.c. del 95% 8-102), con una evidencia de muy baja calidad acerca de la mejoría de la permeabilidad endovascular (riesgo relativo 4,00, i.c. del 95% 0,91-17,68). CONCLUSIÓN: La monoterapia antiplaquetaria tiene un beneficio clínico mínimo en la enfermedad arterial periférica asintomática y un beneficio limitado en la sintomática, con un claro riesgo de hemorragia grave. No existe evidencia para recomendar la prescripción de antiagregantes plaquetarios después de una intervención endovascular periférica, situación que debería abordarse en ensayos aleatorizados con una potencia estadística adecuada.

Angiosome Specific Revascularisation: Does the Evidence Support It?

OBJECTIVE: To explain the angiosome concept and explore the practical application of the angiosome literature to a clinical scenario, in this case a tibial angioplasty for critical ischaemia. METHODS: Clinical vignette with explanation of the decisions made and subsequent clinical results based on the theory of the angiosome concept and the literature on angiosomal revascularisation; in this case the results of our group's recent update to a systematic review and meta-analysis. RESULTS: Endovascular combined or direct angiosomal revascularisation if superior to indirect revascularisation. This was borne out in the clinical scenario, where an indirect peroneal reperfusion of the AT angiosome resulted in major amputation. Open surgery is less dependent on the angiosome concept. The presence of adequate collateralisation into a foot arch seems to be the most important factor predicting success of indirect revascularisation. The evidence for both suffers from selection bias and many of the findings in the literature are wholly due to selection bias. CONCLUSION: The angiosome concept is useful during both open and endovascular tibial revascularisation. However, the runoff in the foot is critical to success and may not follow the 'classic' angiosome model in diabetes.

Systematic Review and Meta-analysis of the Effect of Internal Iliac Artery Exclusion for Patients Undergoing EVAR.

OBJECTIVE: Endovascular abdominal aortic aneurysm repair (EVAR) sometimes requires internal iliac artery (IIA) coverage to achieve a landing zone in the external iliac artery. The aim of this study was to determine complication rates following IIA exclusion. MATERIALS AND METHODS: A systematic review of key journals was undertaken from January 1980 to April 2016. Studies detailing occlusion (using coils or plugs) or coverage of the IIA with outcome data were included. Weighted means were calculated for continuous variables. Meta-analysis was performed when comparative data were available. Quality was assessed using the GRADE system. RESULTS: Sixty-one non-randomised studies (2671 patients; 2748 IIAs) were analysed. Fifteen per cent of EVARs require IIA sacrifice. Buttock claudication (BC) occurred in 27.9% of patients, although 48.0% resolved after 21.8 months. BC rates were 32.6% with coils, 23.8% with plugs, and 12.9% with coverage alone, and less with unilateral (vs. bilateral) IIA treatment (OR 0.57, 95% CI 0.36-0.91). More proximal coil placement resulted in lower rates of BC (OR 0.12, 95% CI 0.03-0.48). Erectile dysfunction occurred in 10.2% of males, with higher rates after coiling. Type II endoleaks were more frequent after covering alone; however re-interventions were rare. Significant ischaemic events (bowel/gluteal/spinal ischaemia) were very rare. Plugs were quicker to place and required less radiation (p < .001) than coils. GRADE scoring was very low for all outcomes. CONCLUSION: Overall the quality of reported data on IIA sacrifice is poor. Buttock claudication and erectile dysfunction occurred frequently after IIA sacrifice. Where both options are technically possible, plugs could be considered preferential to coils, and placed as proximally in the IIA as possible.

Systematic review and meta-analysis of the efficacy of perineural local anaesthetic catheters after major lower limb amputation.

OBJECTIVE: The aim of this systematic review and meta-analysis was to evaluate the effects of using an intraoperatively placed perineural catheter (PNC) with a postoperative local anaesthetic infusion on immediate and long-term outcomes after lower limb amputation. METHODS: A systematic review of key electronic journal databases was undertaken from inception to January 2015. Studies comparing PNC use with either a control, or no PNC, were included. Meta-analysis was performed for postoperative opioid use, pain scores, mortality, and long-term incidence of stump and phantom limb pain. Sensitivity analysis was performed for opioid use. Quality of evidence was assessed using the GRADE system. RESULTS: Seven studies reporting on 416 patients undergoing lower limb amputation with PNC usage (n = 199) or not (n = 217) were included. Approximately 60% were transtibial amputations PNC use reduced postoperative opioid consumption (standardised mean difference: -0.59, 95% CI -1.10 to -0.07, p = .03), maintained on sensitivity analysis for large (p = .03) and high-quality (p = .003) studies, but was marginally lost (p = .06) on studies enrolling patients with peripheral arterial disease only. PNC treatment did not affect postoperative pain scores (p = .48), in-hospital mortality (p = .77), phantom limb pain (p = .28) or stump pain (p = .37). GRADE quality of evidence for all outcomes was very low. CONCLUSION: There is poor-quality evidence that PNC usage significantly reduces opioid consumption following lower limb amputation, without affecting other short- or long-term outcomes. Well-performed randomised studies are required.

Benefits of remote ischaemic preconditioning in vascular surgery.

OBJECTIVES: Remote ischaemic preconditioning (RIPC) is a physiological mechanism to protect against ischaemia-reperfusion injury. It is a technique in which short pre-emptive periods of ischaemia and reperfusion are thought to protect against ischaemia-reperfusion injury during procedures requiring longer periods of ischaemia. Discovered in the 1980s, its clinical application has been investigated heavily since the first human study in 2006. The aim of this paper was to provide a review of this rapidly expanding subject. METHODS: This study consists of a narrative review of the literature focusing on previous meta-analyses and randomised control trials. RESULTS: Five small randomised trials have been published on the effects of RIPC in vascular surgery. Several randomised trials have been published in cardiac surgery and percutaneous coronary intervention. Meta-analysis shows a significant reduction in troponin levels and biomarkers of renal dysfunction in RIPC patients, but as yet no convincing clinical benefit. The largest powered randomised trial in cardiac surgery showed no benefit to RIPC. CONCLUSIONS: Current trials and therefore meta-analyses are generally underpowered. The technique is physiologically sound but remains lacking in clear clinical benefit.

Systematic review and meta-analysis of direct versus indirect angiosomal revascularisation of infrapopliteal arteries.

OBJECTIVE: The aim of this systematic review was to evaluate outcomes of direct revascularisation (DR) versus indirect revascularisation (IR) of infrapopliteal arteries to the affected angiosome for critical limb ischaemia. Both open and endovascular techniques were included. METHODS: A systematic review of key electronic journal databases was undertaken from inception to 22 March 2014. Studies comparing DR versus IR in patients with localised tissue loss were included. Meta-analysis was performed for wound healing, limb salvage, mortality, and re-intervention rates, with numerous sensitivity analyses. Quality of evidence was assessed using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system. RESULTS: Fifteen cohort studies reporting on 1,868 individual limbs were included (endovascular revascularisation, 1,284 limbs; surgical revascularisation, 508 limbs; both methods, 76 limbs). GRADE quality of evidence was low or very low for all outcomes. DR resulted in improved wound healing rates compared with IR (odds ratio [OR] 0.40, 95% confidence interval [CI] 0.29-0.54) and improved limb salvage rates (OR 0.24, 95% CI 0.13-0.45), although this latter effect was lost on high-quality study sensitivity analysis. Wound healing and limb salvage was improved for both open and endovascular intervention. There was no effect on mortality (OR 0.77, 95% CI 0.50-1.19) or reintervention rates (OR: 0.44, 95% CI 0.10-1.88). CONCLUSION: DR of the tibial vessels appears to result in improved wound healing and limb salvage rates compared with IR, with no effect on mortality or reintervention rates. However, the quality of evidence on which these conclusions are based on is low.

Survival following ruptured abdominal aortic aneurysm before and during the IMPROVE Trial: a single-centre series.

OBJECTIVES: The first large-scale randomised trial (Immediate Management of the Patient with Rupture: Open Versus Endovascular repair [IMPROVE]) for endovascular repair of ruptured abdominal aortic aneurysm (rEVAR) has recently finished recruiting patients. The aim of this study was to examine the impact on survival after rEVAR when the IMPROVE protocol was initiated in a high volume abdominal aortic aneurysm (AAA) centre previously performing rEVAR. METHODS: One hundred and sixty-nine patients requiring emergency infrarenal AAA repair from January 2006 to April 2013 were included. Eighty-four patients were treated before (38 rEVAR, 46 open) and 85 (31 rEVAR, 54 open) were treated during the trial period. A retrospective analysis was performed. RESULTS: Before the trial, there was a significant survival benefit for rEVAR over open repair (90-day mortality 13% vs. 30%, p = .04, difference remained significant up to 2 years postoperatively). This survival benefit was lost after starting randomisation (90-day mortality 35% vs. 33%, p = .93). There was an increase in overall 30-day mortality from 15% to 31% (p = .02), while there was no change for open repair (p = .438). There was a significant decrease in general anaesthetic use (p = .002) for patients treated during the trial. Randomised patients had shorter hospital and intensive treatment unit stays (p = .006 and p = .03 respectively). CONCLUSIONS: The change in survival seen during the IMROVE trial highlights the need for randomised rather than cohort data to eliminate selection bias. These results from a single centre reinforce those recently reported in IMPROVE.

Systematic review and meta-analysis of the retroperitoneal versus the transperitoneal approach to the abdominal aorta.

OBJECTIVES: The aim was to systematically review and meta-analyse the differences between the retroperitoneal (RP) and the transperitoneal (TP) approach to the infrarenal abdominal aorta. DESIGN: Systematic review and meta-analysis. METHODS: PubMed, the Cochrane library, Embase and ClinicalTrials.gov were searched for all studies on differences in clinical outcomes between the RP and TP approach. Outcomes were selected based on inclusion in two or more studies: Operative (length of procedure, intraoperative blood loss); Post operative complications (paralytic ileus, pneumonia, myocardial infarction (MI), renal failure and wound hernia); Mortality (30 day, 1 year); Post-operative changes in respiratory function (forced expiratory volume in 1 second, forced vital capacity); Length of hospital and Intensive care unit (ICU) stay and Cost. The data were pooled by outcome. RESULTS: Eight randomised and 21 cohort studies involving 3035 patients were included. Meta-analysis showed significantly lower rates of postoperative ileus (Odds ratio (OR) 0.17[95% CI 0.10, 0.32] p < 0.00001), pneumonia (OR 0.42[95% CI 0.26, 0.68] p = 0.0004), ICU stay (standardised mean difference (SMD) 0.67[95% CI 1.28, 0.06] p = 0.03), total hospital stay (SMD 0.88[95% CI 1.32, 0.44] p < 0.0001) and cost (SMD 1.15[95% CI 2.11, 0.19] p = 0.02) for patients undergoing a RP approach. Study quality was generally low, with conflicting results and concerns over publication bias in some cohort studies. CONCLUSIONS: The RP approach for open aortic surgery is associated with lower rates of postoperative ileus and pneumonia when compared to the TP approach.

Systematic review and meta-analysis of histopathological factors influencing the risk of lymph node metastasis in early colorectal cancer.

AIM: Lymph node (LN) metastases are present in up to 17% of early colorectal cancers (pT1). Identification of associated histopathological factors would enable counselling of patients regarding this risk. METHOD: Pubmed and Embase were employed utilizing the terms 'early colorectal cancer', 'lymph node metastasis', 'submucosal invasion', 'lymphovascular invasion', 'tumour budding' and 'histological differentiation'. Analysis was performed using REVIEW MANAGER 5.1. RESULTS: Twenty-three cohort studies including 4510 patients were analysed. There was a significantly higher risk of LN metastasis with a depth of submucosal invasion > 1 mm than with lesser degrees of penetration (OR 3.87, 95% CI 1.50-10.00, P = 0.005). Lymphovascular invasion was significantly associated with LN metastasis (OR 4.81, 95% CI 3.14-7.37, P < 0.00001). Poorly differentiated tumours had a higher risk of LN metastasis compared with well or moderately differentiated tumours (OR 5.60, 95% CI 2.90-10.82, P < 0.00001). Tumour budding was found to be significantly associated with LN metastasis (OR 7.74, 95% CI 4.47-13.39, P < 0.001). CONCLUSION: Meta-analysis of the current literature demonstrates that in early colorectal cancer a depth of submucosal invasion by the primary tumour of > 1 mm, lymphovascular invasion, poor differentiation and tumour budding are significantly associated with LN metastasis.

Systematic review and meta-analysis of vein cuffs for below-knee synthetic bypass.

BACKGROUND: The aim was to investigate the possible benefit of vein cuffs for femoral to below-knee popliteal and femorodistal vessel synthetic bypass grafts. METHODS: PubMed, the Cochrane library, Embase and ClinicalTrials.gov were searched for all studies on any clinical effect of vein cuffs on synthetic grafts. Outcomes were selected based on inclusion in two or more studies: primary patency and limb survival. The data were subjected to meta-analysis by outcome. RESULTS: Three cohort and two randomized studies were selected for inclusion, involving 885 patients. Meta-analysis of five studies examining below-knee popliteal bypass showed a significant improvement for primary patency in cuffed grafts at 2 years, but not at 1 or 3 years (odds ratio at 2 years 0·46, 95 per cent confidence interval 0·22 to 0·97; P = 0·04). Limb salvage was significantly improved in cuffed grafts up to 2 years. Limb survival was also improved for cuffed distal grafts at 2 years (odds ratio 0·29, 0·11 to 0·75; P = 0·01) but showed no difference at any other time interval. Study quality was generally poor, with conflicting results. CONCLUSION: There was a small but significant benefit for vein cuffs on synthetic grafts used for femoral to below-knee popliteal anastomoses, but little benefit for femorodistal anastomoses.

A scoring system (DISTAL) for predicting failure of snuffbox arteriovenous fistulas.

OBJECTIVES: A first fistula failing will lead to serious morbidity in a proportion of patients. Snuffbox fistulas have the advantage of proximal vessel preservation, and although several factors have been associated with failure, the relative importance of these factors combined and their clinical applicability is unknown. The aim of this study was to determine the relative importance of risk factors for snuffbox fistula failure and create a simple scoring system to aid fistula placement decision making. METHODS: 218 consecutive patients were examined using Cox regression analysis to determine risk factors for failure. Primary patency was used as the endpoint. RESULTS: Diabetes, IHD, Stroke, Two snuffbox procedures, Age > 70 and Less than 2.0 mm vein (DISTAL, maximum score 6) were significant predictors of primary patency failure. There was a clear decrease in primary patency with increasing DISTAL score (log rank χ(2) = 30.3, DF = 5, P < 0.001). Performing snuffbox procedures on patients with a score ≤ 3 would give a 23% reduction in the number of failures within two months for a 12% reduction in the number of patients offered snuffbox procedures. CONCLUSION: The DISTAL scoring system could give large improvements in primary patency for the snuffbox fistula if the results can be validated in other datasets.

Surgical factors in the prevention of infection following major lower limb amputation.

OBJECTIVES: Infection following major lower limb amputation is common but surgical influences on the rates of infection are not known. We aim to assess the influence of peri-operative surgical factors on outcome. DESIGN AND METHODS: Review of a prospective database included all patients undergoing a major lower limb amputation from March 2008 to July 2010. Infection was classified using Centre for Disease Control criteria and multivariate analysis performed to identify significant risk factors. RESULTS: 127 patients, median age 78 yrs (31-98) were included. 34.6% of patients developed a wound infection following surgery; 47.7% of which were classed as superficial incisional surgical site infections, with 52.3% being deep incisional surgical site infections. There was a higher infection rate in below knee than above knee amputations (p < 0.001). There was no relationship between the grade of the operating surgeon (p = 0.829), peri-operative antibiotics (p = 0.933), length of operation (p = 0.651), use of nerve catheter (0.267) and the post-operative presence of infection. There was a higher rate of infection with the use of suction drains (p < 0.05). The use of skin clips rather than sutures was associated with an increased rate of infection (p < 0.05). There was an increased need for revision surgery with the use of skin clips, although this was not significant (p = 0.07). CONCLUSIONS: Skin clips and surgical drains adversely influence the risk of infection in major limb amputation and their use should be avoided.

Systematic review and meta-analysis of the effects of statin therapy on abdominal aortic aneurysms.

BACKGROUND: The aim was to investigate the effects of statin therapy on abdominal aortic aneurysm (AAA) disease. METHODS: PubMed, the Cochrane Library, Embase and ClinicalTrials.gov were searched for all studies on any clinical effect of statin therapy on AAA. Outcomes were selected based on their inclusion in two or more studies: AAA expansion rate, 30-day mortality, and short- and long-term postoperative mortality. The data were subjected to meta-analysis by outcome. RESULTS: Twelve cohort studies were selected for inclusion involving 11 933 individuals. Meta-analysis of four studies examining all-cause postoperative mortality showed a significant improvement with statin therapy at 1, 2 and 5 years (odds ratio (OR) at 5 years 0·57, 95 per cent confidence interval (c.i.) 0·42 to 0·79; P < 0·001) with minimal heterogeneity between the four included studies. There was no significant difference in 30-day mortality after AAA treatment in patients on statin therapy (OR 0·22, 0·02 to 2·90; P = 0·25). Sensitivity analysis including four high-quality studies examining AAA expansion rates showed no significant difference with statin therapy: standardized mean difference -0·14 (95 per cent c.i. -0·33 to -0·05) mm/year (P = 0·16). CONCLUSION: The claim of a reduction in AAA expansion rate with statin therapy is based on low-quality evidence and was not significant on meta-analysis. However, statin therapy did appear to improve all-cause survival after AAA repair.

Prevalence and outcome of esophagogastric anastomotic leak after esophagectomy in a UK regional cancer network.

The aim of this study was to determine the contemporary prevalence, outcome, and survival after esophagogastric anastomotic leakage (EGAL) following esophagectomy by a regional upper gastrointestinal cancer network and to investigate etiological factors. Two hundred forty consecutive patients underwent esophagectomy over a 10-year period (median age 61 [31-79] years, 147 transthoracic and 93 transhiatal esophagectomy, 105 neoadjuvant chemotherapy, 49 chemoradiotherapy). The primary outcome measures were the development of EGAL and survival. Twenty patients developed EGAL (8.3%, 15 managed conservatively, 5 reoperation). Overall operative mortality was 2% (5 patients in total, 1 after EGAL). Median, 1 and 2-year survival was 22 months, 73% and 50%, in patients after EGAL, compared with 31 months, 80% and 56%, in patients who did not suffer EGAL (P= 0.314). On multivariate analysis, low body mass indices (hazard ratio [HR] 0.29, 95% confidence interval [CI] 0.11-0.79, P= 0.016), individual surgeon (HR 1.21, 95% CI 1.02-1.43, P= 0.02), and neoadjuvant chemotherapy (HR 3.28, 95% CI 1.16-9.22, P= 0.024) were significantly associated with the development of EGAL. EGAL following esophagectomy remained common, but associated mortality was less common than reported in earlier Western series and long-term survival was unaffected.

Prognostic significance of the endoscopic ultrasound defined lymph node metastasis count in esophageal cancer.

The key prognostic factor which predicts outcome after esophagectomy for cancer is the number of malignant lymph node metastases, but data regarding the accuracy of endoscopic ultrasound (EUS) in determining and predicting the metastatic lymph node count preoperatively are limited. The aim of this study was to assess the prognostic significance of EUS defined lymph node metastasis count (eLNMC) in patients diagnosed with esophageal cancer. Two hundred and sixty-seven consecutive patients (median age 63 years, 187 months) underwent specialist EUS followed by stage directed multidisciplinary treatment (183 esophagectomy [64 neoadjuvant chemotherapy, 19 neoadjuvant chemoradiotherapy], 79 definitive chemoradiotherapy, and 5 palliative therapy). The eLNMC was subdivided into four groups (0, 1, 2 to 4, >4) and the primary measure of outcome was survival. Survival was related to EUS tumor (T) stage (P < 0.0001), EUS node (N) stage (P < 0.0001), EUS tumor length (p < 0.0001), and eLNMC (P < 0.0001). Multivariable analysis revealed EUS tumor length (hazard ratio [HR] 1.071, 95% CI 1.008-1.138, P= 0.027) and eLNMC (HR 1.302, 95% CI 1.133-1.496, P= 0.0001) to be significantly and independently associated with survival. Median and 2-year survival for patients with 0, 1, 2-4, and >4 lymph node metastases were: 44 months and 71%, 36 months and 59%, 24 months and 50%, and 17 months and 32%, respectively. The total number of EUS defined lymph node metastases was an important and significant prognostic indicator.

Systematic review and narrative synthesis of surgeons' perception of postoperative outcomes and risk.

BACKGROUND: The accuracy with which surgeons can predict outcomes following surgery has not been explored in a systematic way. The aim of this review was to determine how accurately a surgeon's 'gut feeling' or perception of risk correlates with patient outcomes and available risk scoring systems. METHODS: A systematic review was undertaken in accordance with PRISMA guidelines. A narrative synthesis was performed in accordance with the Guidance on the Conduct of Narrative Synthesis In Systematic Reviews. Studies comparing surgeons' preoperative or postoperative assessment of patient outcomes were included. Studies that made comparisons with risk scoring tools were also included. Outcomes evaluated were postoperative mortality, general and operation-specific morbidity and long-term outcomes. RESULTS: Twenty-seven studies comprising 20 898 patients undergoing general, gastrointestinal, cardiothoracic, orthopaedic, vascular, urology, endocrine and neurosurgical operations were included. Surgeons consistently overpredicted mortality rates and were outperformed by existing risk scoring tools in six of seven studies comparing area under receiver operating characteristic (ROC) curves (AUC). Surgeons' prediction of general morbidity was good, and was equivalent to, or better than, pre-existing risk prediction models. Long-term outcomes were poorly predicted by surgeons, with AUC values ranging from 0·51 to 0·75. Four of five studies found postoperative risk estimates to be more accurate than those made before surgery. CONCLUSION: Surgeons consistently overestimate mortality risk and are outperformed by pre-existing tools; prediction of longer-term outcomes is also poor. Surgeons should consider the use of risk prediction tools when available to inform clinical decision-making.

ANTECEDENTES: La precisión con la cual los cirujanos pueden predecir los resultados de la cirugía no se ha estudiado de forma sistemática. El objetivo de esta revisión fue determinar con qué precisión la intuición de un cirujano o su percepción del riesgo se correlacionaba con los resultados del paciente y con los sistemas de puntuación del riesgo disponibles. MÉTODOS: Se efectuó una revisión sistemática siguiendo las directrices PRISMA. Se realizó una síntesis narrativa de acuerdo con la guía para la realización de síntesis narrativas en revisiones sistemáticas. Se incluyeron los estudios que comparaban las evaluaciones preoperatorias o postoperatorias de los cirujanos respecto a los resultados de los pacientes. También se incluyeron aquellos estudios en los que se hacían comparaciones con herramientas de puntuación de riesgo. Se evaluaron la mortalidad postoperatoria, la morbilidad global y la morbilidad específica de las intervenciones, y los resultados a largo plazo. RESULTADOS: Se incluyeron 27 estudios con 20.898 pacientes en los que se realizaron procedimientos de cirugía general, digestiva, cardiotorácica, ortopédica, vascular, urológica, endocrina y neurocirugía. Los cirujanos predijeron consistentemente mayores tasas de mortalidad, siendo superados en precisión por los sistemas de estimación del riesgo existentes en seis de los siete estudios que utilizaron el área bajo la curva (area under curve, AUC) operativa del receptor. La predicción de la morbilidad general por parte de los cirujanos fue buena y era equivalente, incluso mejor, que los modelos de predicción de riesgos preexistentes. La capacidad de los cirujanos para predecir los resultados a largo plazo fue pobre, con una AUC que oscilaba entre 0,51 y 0,75. Cuatro de cinco estudios encontraron que las estimaciones de riesgo postoperatorias fueron más precisas que las realizadas preoperatoriamente. CONCLUSIÓN: Los cirujanos sobrestiman consistentemente el riesgo de mortalidad, siendo superados en precisión por las herramientas preexistentes. La predicción de resultados a largo plazo también es muy pobre. Los cirujanos deberían considerar el uso de herramientas de predicción de riesgo cuando estén disponibles para informar en el proceso de decisión clínica.