Clinical Decision Analysis and Markov Modeling for Surgeons: An Introductory Overview.

This study addresses the use of decision analysis and Markov models to make contemplated decisions for surgical problems. Decision analysis and decision modeling in surgical research are increasing, but many surgeons are unfamiliar with the techniques and are skeptical of the results. The goal of this review is to familiarize surgeons with techniques and terminology used in decision analytic papers, to provide the reader a practical guide to read these papers, and to ensure that surgeons can critically appraise the quality of published clinical decision models and draw well founded conclusions from such reports.First, a brief explanation of decision analysis and Markov models is presented in simple steps, followed by an overview of the components of a decision and Markov model. Subsequently, commonly used terms and definitions are described and explained, including quality-adjusted life-years, disability-adjusted life-years, discounting, half-cycle correction, cycle length, probabilistic sensitivity analysis, incremental cost-effectiveness ratio, and the willingness-to-pay threshold.Finally, the advantages and limitations of research with Markov models are described, and new modeling techniques and future perspectives are discussed. It is important that surgeons are able to understand conclusions from decision analytic studies and are familiar with the specific definitions of the terminology used in the field to keep up with surgical research. Decision analysis can guide treatment strategies when complex clinical questions need to be answered and is a necessary and useful addition to the surgical research armamentarium.

Cost-effectiveness of endovascular repair, open repair, and conservative management of splenic artery aneurysms.

OBJECTIVE: Open repair (OPEN) and conservative management (CONS) have been the treatments of choice for splenic artery aneurysms (SAAs) for many years. Endovascular repair (EV) has been increasingly used with good short-term results. In this study, we evaluated the cost-effectiveness of OPEN, EV, and CONS for the treatment of SAAs. METHODS: A decision analysis model was developed using TreeAge Pro 2013 software (TreeAge Inc, Williamstown, Mass) to evaluate the cost-effectiveness of the different treatments for SAAs. A hypothetical cohort of 10,000 55-year-old female patients with SAAs was assessed in the reference-case analysis. Perioperative mortality, disease-specific mortality rates, complications, rupture risks, and reinterventions were retrieved from a recent and extensive meta-analysis. Costs were analyzed with the 2014 Medicare database. The willingness to pay was set to $60,000/quality-adjusted life years (QALYs). Outcomes evaluated were QALYs, costs from the health care perspective, and the incremental cost-effectiveness ratio (ICER). Extensive sensitivity analyses were performed and different clinical scenarios evaluated. Probabilistic sensitivity analysis was performed to include the uncertainty around the variables. A flowchart for clinical decision-making was developed. RESULTS: For a 55-year-old female patient with a SAA, EV has the highest QALYs (11.32; 95% credibility interval [CI], 9.52-13.17), followed by OPEN (10.48; 95% CI, 8.75-12.25) and CONS (10.39; 95% CI, 8.96-11.87). The difference in effect for 55-year-old female patients between EV and OPEN is 0.84 QALY (95% CI, 0.42-1.34), comparable with 10 months in perfect health. EV is more effective and less costly than OPEN and more effective and more expensive compared with CONS, with an ICER of $17,154/QALY. Moreover, OPEN, with an ICER of $223,166/QALY, is not cost-effective compared with CONS. In elderly individuals (age >78 years), the ICER of EV vs CONS is $60,503/QALY and increases further with age, making EV no longer cost-effective. Very elderly patients (age >93 years) have higher QALYs and lower costs when treated with CONS. The EV group has the highest number of expected reinterventions, followed by CONS and OPEN, and the number of expected reinterventions decreases with age. CONCLUSIONS: EV is the most cost-effective treatment for most patient groups with SAAs, independent of the sex and risk profile of the patient. EV is superior to OPEN, being both cost-saving and more effective in all age groups. Elderly patients should be considered for CONS, based on the high costs in relation to the very small gain in health when treated with EV. The very elderly should be treated with CONS.

A comparison of open and endovascular revascularization for chronic mesenteric ischemia in a clinical decision model.

OBJECTIVE: Open revascularization (OR) has been the treatment of choice for chronic mesenteric ischemia (CMI) for many years, but endovascular revascularization (EV) has been increasingly used with good short-term results. In this study, we evaluated the comparative effectiveness and cost-effectiveness of EV and OR in patients with CMI refractory to conservative management. METHODS: A Markov-state transition model was developed using TreeAge Pro 2012 (TreeAge Inc, Williamstown, Mass) to simulate a hypothetical cohort of 10,000 65-year-old female patients with CMI requiring treatment with either OR or EV. Data for the model, including perioperative and long-term overall mortality risks, disease-specific mortality risks, complications, and reintervention and patency rates, were retrieved from original studies and systematic reviews about CMI. Costs were analyzed with the 2013 Medicare database. Outcomes evaluated were quality-adjusted life-years (QALYs), costs from the health care perspective, and the incremental cost-effectiveness ratio. Extensive sensitivity analyses were performed and different clinical scenarios evaluated. Probabilistic sensitivity analysis was performed to assess robustness of the model. RESULTS: For a reference-case 65-year-old female patient with CMI and an average risk for operation, EV is preferred with 10.03 QALYs (95% credibility interval [CI], 9.76-10.29) vs 9.59 after OR (95% CI, 9.29-9.87). The difference is comparable to 5 months in perfect health: 0.44 QALY (95% CI, 0.13-0.76). For 65-year-old men, this was 8.71 QALYs (95% CI, 8.48-8.94) for EV vs 8.42 (95% CI, 8.14-8.63) for OR. Sensitivity analysis showed that for younger patients, EV results in a higher increase in QALYs compared with older patients. Total expected reinterventions per patient are 1.70 for EV vs 0.30 for OR. Total expected health care costs for the reference-case patient were $39,942 (95% CI, $28,509-$53,380) for OR and $38.217 (95% CI, $29,329-$48,309) for EV. For men, this was $39,375 (95% CI, $28,092-$52,853) for OR and $35,903 (95% CI, $27,685-$45,597) for EV. For patients younger than 60 years, EV is a more expensive treatment strategy compared with OR, but with an incremental cost-effectiveness ratio for EV of less than $60,000/QALY. For patients 60 years and older, EV dominated OR as preferential treatment because effectiveness was higher than for OR and costs were lower. CONCLUSIONS: The results of this decision analysis model suggest that EV is favored over OR for patients with CMI in all age groups. Although EV is associated with more expected reinterventions, EV appears to be cost-effective for all age groups.

A clinical decision model for selecting the most appropriate therapy for uncomplicated chronic dissections of the descending aorta.

OBJECTIVE: The optimal treatment for patients with uncomplicated chronic Stanford type B aortic dissections (chTBADs) is still matter of debate. The purpose of this study was to design a decision tool to guide the surgeon in determining the preferred treatment option. METHODS: A Markov decision-analysis model compared chTBAD patients treated with initial open surgical repair (OSR), thoracic endovascular aortic repair (TEVAR), and optimal medical therapy (OMT), followed during follow-up by OSR (OMT-OSR) or TEVAR (OMT-TEVAR), if indicated. Procedural risks, aortic growth and rupture rates, outcomes, and quality of life values were derived from the best available evidence in the literature. A chTBAD treatment strategy decision tool was developed, including the four key variables of age, sex, surgical risk, and maximum initial aortic diameter. Primary outcome was quality-adjusted life-years (QALYs). RESULTS: For the reference patient cohort, 55-year-old men with chTBAD with a maximum aortic diameter of 5.0 cm, medium risk for surgery, and a threshold for surgery of 6.0 cm during follow-up, OSR yielded higher QALYs, with 10.06 QALYs (95% credibility interval [CI], 9.52-10.56 QALYs) vs 9.92 QALYs (95% CI, 9.23-10.58 QALYs) after TEVAR and 9.64 QALYs (95% CI, 9.38-9.88 QALYs) and 9.40 QALYs (95% CI, 9.11-9.69 QALYs) for OMT-OSR and OMT-TEVAR. The difference between OSR and OMT-OSR was 0.42 QALYs (95% CI, 0.01-0.81 QALYs) and between TEVAR and OMT-TEVAR was 0.52 QALYs (95% CI, 0.04-0.68 QALYs). This showed that intervention is preferred over OMT. A change of the four variables resulted in a change of preferred treatment. In general, OSR was the preferred treatment in younger patients with a larger aortic diameter and in low-risk patients. TEVAR was preferred in elderly patients with large aortic diameter and if the aortic diameter threshold for repair decreased. OMT was the optimal therapy in high-risk patients, elderly patients, or in patients with small aortic diameters. CONCLUSIONS: This decision-analysis model shows that there is no "one-size-fits-all" treatment for uncomplicated chTBADs. For the reference patient cohort, intervention is preferred over OMT. Age is the most important deciding factor, followed by initial aortic diameter. Immediate OSR is the preferred treatment option in younger patients with a large initial aortic diameter and in low-risk patients. Immediate TEVAR is preferred in elderly patients with a large initial aortic diameter and in patients with a lower threshold for OSR. OMT should be considered in high-risk patients, in patients with small initial aortic diameters, and in patients aged >80 years, unless their initial aortic diameter is >5.5 cm. However, the differences in some patient groups are clinically insignificant, allowing a major role for patient preferences and hospital-specific considerations. This clinical decision model may guide chTBAD treatment.

Decision analysis model of open repair versus endovascular treatment in patients with asymptomatic popliteal artery aneurysms.

OBJECTIVE: Repair is indicated of asymptomatic popliteal artery aneurysms (aPAAs) that are >2 cm. Endovascular PAA repair with covered stents (stenting) is increasingly used. It is, however, unclear when an endovascular approach is preferred to traditional open repair with great saphenous vein bypass (GSVB). The goal of this study was to assess the treatment options for aPAAs using decision analysis. METHODS: A Markov model was developed and a hypothetic cohort of patients with aPAAs was analyzed. GSVB, stenting, and nonoperative management with optimal medical treatment (OMT) were compared. Operative mortality, patency rates, quality-of-life values, and costs were determined by comprehensive review of the best available evidence. The main outcome was quality-adjusted life-years (QALYs). Secondary outcomes were cost-effectiveness and number of reinterventions. RESULTS: For a 65-year-old male patient with a 2.0-cm aPAA and without significant comorbidities, probabilistic sensitivity analysis shows that intervention is preferred over OMT (5.77 QALYs, 95% credibility interval [CI], 5.43-6.11; OMT). GSVB treatment for this patient results in slightly higher QALYs than stent placement, with a predicted 8.43 QALYs (GSVB: 95% CI, 8.21-8.64) vs 8.07 QALYs (stenting: 95% CI, 7.84-8.29), a difference of 0.36 QALYs (95% CI, 0.14-0.58). Furthermore, costs are higher for stenting ($40,464; 95% CI, $34,814-$46,242) vs GSVB ($21,618; 95% CI, $15,932-$28,070), and more reinterventions are required after stenting (1.03 per patient) vs GSVB (0.52 per patient), making GSVB the preferred strategy for all outcomes considered. Stenting is preferred in patients who are at high risk for open repair (>6% 30-day mortality) or if the 5-year primary patency rates of stenting increase to 80%. For very old patients (>95 years) and patients with a very short life expectancy (<1.5 years), OMT yields higher QALYs. CONCLUSIONS: GSVB is the preferred treatment in 65-year-old patients with aPAAs for all outcomes considered. However, patients at high risk for open repair or without suitable vein should be considered as candidates for endovascular repair. Very elderly patients and patients with a short life expectancy are best treated with OMT. Further improvement of endovascular techniques that increase patency rates of endovascular stents could make this the preferred therapy for more patients in the future.

Open repair, endovascular repair, and conservative management of true splenic artery aneurysms.

OBJECTIVE: True splenic artery aneurysms (SAAs) are a rare but potentially fatal pathology. For many years, open repair (OPEN) and conservative management (CONS) were the treatments of choice, but throughout the last decade endovascular repair (EV) has become increasingly used. The purpose of the present study was to perform a systematic review and meta-analysis evaluating the outcomes of the three major treatment modalities (OPEN, EV, and CONS) for the management of SAAs. METHODS: A systematic review of all studies describing the outcomes of SAAs treated with OPEN, EV, or CONS was performed using seven large medical databases. The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines were followed to ensure a high-quality review. All articles were subject to critical appraisal for relevance, validity, and availability of data regarding characteristics and outcomes. All data were systematically pooled, and meta-analyses were performed on several outcomes, including early and late mortality, complications, and number of reinterventions. RESULTS: Original data of 1321 patients with true SAAs were identified in 47 articles. OPEN contained 511 patients (38.7%) in 31 articles, followed by 425 patients (32.2%) in CONS in 16 articles and 385 patients (29.1%) in EV in 33 articles. The CONS group had fewer symptomatic patients (9.5% vs 28.7% in OPEN and 28.8% in EV; P < .001) and fewer ruptured aneurysms (0.2% vs 18.4% in OPEN and 8.8% in EV; P < .001), but no significant differences were found in existing comorbidities. CONS patients were usually older and had smaller-sized aneurysms than patients in the OPEN and EV groups. The only identified difference in baseline characteristics between OPEN and EV was the number of ruptured aneurysms (18.4% vs 8.8%; P < .001). OPEN had a higher 30-day mortality than EV (5.1% vs 0.6%; P < .001), whereas minor complications occurred in a larger number of the EV patients. EV required more reinterventions per year (3.2%) compared with OPEN (0.5%) and CONS (1.2%; P < .001). The late mortality rate was higher in patients treated with CONS (4.9% vs 2.1% in OPEN and 1.4% in EV; P = .04). CONCLUSIONS: EV of SAA has better short-term results compared with OPEN, including significantly lower perioperative mortality. OPEN is associated with fewer late complications and fewer reinterventions during follow-up. Patients treated with CONS showed a higher late mortality rate. Ruptured SAAs are predictors of a significantly higher perioperative mortality compared with nonruptured SAAs in the OPEN and EV groups.

Endovascular vs. open repair of complicated acute type B aortic dissections.

Purpose : To assess the comparative effectiveness of thoracic endovascular aortic repair (TEVAR) vs. open surgical repair (OSR) of complicated acute type B aortic dissections (cABAD) using decision analysis. Methods : A decision analysis comparing TEVAR and OSR for cABAD included variables extracted from the best-available evidence. Main outcomes were quality-adjusted life years (QALYs), presented with the 95% credibility intervals (CI), and number of reinterventions over the remaining lifetime. Different clinical scenarios, including age, gender, and risk profile were analyzed. Parameter uncertainty was analyzed using probabilistic sensitivity analysis. Results : In the reference case, a cohort of 55-year-old men, TEVAR was preferred over OSR: 7.07 QALYs (95% CI 6.77 to 7.38) vs. 6.34 QALYs (95% CI 6.04 to 6.66) for OSR. The difference of 0.73 QALYs (95% CI 0.29 to 1.17) is equal to 8.5 months in perfect health. TEVAR was more effective in all analyzed cases and age groups. Perioperative mortality was the most important variable affecting the difference between OSR and TEVAR, followed by the relative risk and percentage of aortic-related complications. Total expected reinterventions were 0.43/patient (TEVAR) and 0.35/patient (OSR). Conclusion : The results of this decision model for the treatment of cABAD suggest that TEVAR is preferred over OSR. Although a higher number of reinterventions is expected, the total effectiveness of TEVAR is higher for all age groups. OSR should be reserved for patients whose aortic anatomy is unsuitable for endovascular repair.

Successful treatment of a proximal type I endoleak with HeliFX EndoAnchors.

We report the use of Aptus HeliFX EndoAnchors for endovascular treatment of a proximal type I endoleak after previous endovascular aneurysm repair (EVAR) of a ruptured abdominal aortic aneurysm. An 81-year-old man had been treated with EVAR after a ruptured 12 × 11 cm abdominal aortic aneurysm. Standard computed tomographic angiography follow-up demonstrated a proximal type I endoleak. Because of the highly angulated neck and close position of the endograft to the renal arteries, placement of a proximal extension cuff was prohibited; therefore, the endoleak was treated with an alternative approach using the Aptus HeliFX EndoAnchors. Nine EndoAnchors were successfully placed circumferentially on the proximal site of the endograft. This successfully treated the endoleak by excluding the aneurysm sac from the circulation. Computed tomographic angiography follow-up after 3 months showed no residual type I endoleak. This case shows that placement of EndoAnchors can serve as a viable treatment option for proximal type I endoleaks after failed EVAR.

Surgical and anesthetic considerations for the endovascular treatment of ruptured descending thoracic aortic aneurysms.

PURPOSE OF REVIEW: Ruptured descending thoracic aortic aneurysm (rDTAA) is a life-threatening disease. In the last decade, thoracic endovascular aortic repair (TEVAR) has evolved as a viable option and is now considered the preferred treatment for rDTAAs. New opportunities as well as new challenges are faced by both the surgeon and the anesthesiologist. This review describes the impact of current developments and new modalities for the surgical and anesthetic management of rDTAAs. RECENT FINDINGS: A collaborative approach between the anesthesiologist and surgeon during critical moments such as induction, moment of aortic occlusion and placement of the aortic stent-graft is mandatory. Important issues to consider on preoperative imaging evaluation are correct sizing of the aortic stent-graft and localization of the artery of Adamkiewicz. Emergency TEVAR should preferentially be started under local anesthesia and could be switched to general anesthesia after stent placement. Patients should be kept in permissive hypotension preoperatively and during the intervention before stent-graft deployment and relative hypertension after deployment. The use of a proactive spinal cord protection protocol could decrease the risk of spinal cord ischemia and/or paraplegia and consists of permissive hypertension after stent deployment, cerebrospinal fluid drainage to maintain adequate spinal cord perfusion, relative hypothermia and possibly use of mannitol. SUMMARY: In order to improve outcomes of TEVAR for rDTAA, a close communication between the anesthesiologist and the surgeon and a thorough understanding of the events during the procedure is mandatory. The use of a proactive spinal cord protection protocol may decrease the rates of devastating spinal cord ischemia.

Thoracic endovascular aortic repair with the chimney graft technique.

OBJECTIVE: This study was conducted to provide insight into the safety, applicability, and outcomes of thoracic endovascular aortic repair (TEVAR) with the chimney graft technique. METHODS: Original data regarding the chimney technique in TEVAR in the emergent and elective setting were collected from MEDLINE, Embase, and Scopus databases. All variables were systematically extracted and included in a database. Patient and procedural characteristics, details, and outcomes were analyzed. RESULTS: In total, 94 patients with 101 chimney-stented aortic arch branches were analyzed, consisting of the brachiocephalic artery in 20, the left common carotid artery in 48, and the left subclavian artery in 33. Balloon-expandable stents were used in 36% and self-expandable stents in 64% for the aortic side branch. The interventions were elective in 72% and emergent in 28%. Technical success was achieved in 98% in elective and emergent settings combined. Endoleaks were described in 18%; with type Ia being most frequently reported in 6.4% overall and in 6.5% in the elective setting. Stroke was reported in 5.3% of the patients, of which 40% were fatal. The overall perioperative mortality was 3.2%. Median follow-up time was 11 months, and chimney stents remained patent in all patients. CONCLUSIONS: TEVAR with the chimney technique is a viable treatment option and may expand treatment strategies for patients with challenging thoracic aortic pathology and anatomy in the emergent and elective setting. Patency of the thoracic chimney stents appears to be good during short-term follow-up. Other complications, such as endoleak and stroke, deserve attention by future research to further improve treatment strategies and the prognosis of these patients.

[A male with painful blue discoloration of the legs]. / Een man met blauw verkleurde, pijnlijke benen.

A 79 year old male was presented with progressive blue discoloration of both legs. A ruptured aneurysm with fistula into the inferior vena cava was diagnosed. During subsequent emergency EVAR the aneurysm was treated. Directly after surgery the discoloration and pain disappeared.

A Giant Superior Mesenteric Artery Aneurysm Mimicking an Abdominal Aortic Aneurysm.

Superior mesenteric artery aneurysms are a rare clinical finding but can present with a wide range of symptoms. They are associated with a high risk of rupture, as well as high morbidity and mortality. We present a case of a 38-year-old female who presented with acute abdominal pain and a pulsatile palpable mass in the right epigastric region without other signs or symptoms.

Treatment of penetrating trauma of the extremities: ten years' experience at a Dutch level 1 trauma center.

BACKGROUND: A selective non-operative management (SNOM) has found to be an adequate and safe strategy to assess and treat patients suffering from penetrating trauma of the extremities (PTE). With this SNOM comes a strategy in which adjuvant investigations or interventions are not routinely performed, but based on physical examination only. METHODS: All subsequent patients presented with PTE at a Dutch level I trauma center from October 2000 to June 2011 were included in this study. In-hospital and long-term outcome was analysed in the light of assessment of these patients according to the SNOM protocol. RESULTS: A total of 668 patients (88.2% male; 33.8% gunshot wounds) with PTE presented at the Emergency Department of a level 1 traumacenter, of whom 156 were admitted for surgical treatment or observation. Overall, 22 (14%) patients that were admitted underwent exploration of the extremity for vascular injury. After conservative observation, two (1.5%) patients needed an intervention to treat (late onset) vascular complications. Other long-term extremity related complications were loss of function or other deformity (n = 9) due to missed nerve injury, including 2 patients with peroneal nerve injury caused by delayed compartment syndrome treatment. CONCLUSION: A SNOM protocol for initial assessment and treatment of PTE is feasible and safe. Clinical examination of the injured extremity is a reliable diagnostic 'tool' for excluding vascular injury. Repeated assessments for nerve injuries are important as these are the ones that are frequently missed and result in long-term disability. LEVEL OF EVIDENCE: II / III, retrospective prognostic observational cohort study Key words Penetrating trauma, extremity, vascular injury, complications.