Efficacy and safety of prostate artery embolization for benign prostatic hyperplasia: an observational study and propensity-matched comparison with transurethral resection of the prostate (the UK-ROPE study).

OBJECTIVES: To assess the efficacy and safety of prostate artery embolization (PAE) for lower urinary tract symptoms (LUTS) secondary to benign prostatic hyperplasia (BPH) and to conduct an indirect comparison of PAE with transurethral resection of the prostate (TURP). PATIENTS AND METHODS: As a joint initiative between the British Society of Interventional Radiologists, the British Association of Urological Surgeons and the National Institute for Health and Care Excellence, we conducted the UK Register of Prostate Embolization (UK-ROPE) study, which recruited 305 patients across 17 UK urological/interventional radiology centres, 216 of whom underwent PAE and 89 of whom underwent TURP. The primary outcomes were International Prostate Symptom Score (IPSS) improvement in the PAE group at 12 months post-procedure, and complication data post-PAE. We also aimed to compare IPSS score improvements between the PAE and TURP groups, using non-inferiority analysis on propensity-score-matched patient pairs. The clinical results and urological measurements were performed at clinical sites. IPSS and other questionnaire-based results were mailed by patients directly to the trial unit managing the study. All data were uploaded centrally to the UK-ROPE study database. RESULTS: The results showed that PAE was clinically effective, producing a median 10-point IPSS improvement from baseline at 12 months post-procedure. PAE did not appear to be as effective as TURP, which produced a median 15-point IPSS score improvement at 12 months post-procedure. These findings are further supported by the propensity score analysis, in which we formed 65 closely matched pairs of patients who underwent PAE and patients who underwent TURP. In terms of IPSS and quality-of-life (QoL) improvement, there was no evidence of PAE being non-inferior to TURP. Patients in the PAE group had a statistically significant improvement in maximum urinary flow rate and prostate volume reduction at 12 months post-procedure. PAE had a reoperation rate of 5% before 12 months and 15% after 12 months (20% total rate), and a low complication rate. Of 216 patients, one had sepsis, one required a blood transfusion, four had local arterial dissection and four had a groin haematoma. Two patients had non-target embolization that presented as self-limiting penile ulcers. Additional patient-reported outcomes, pain levels and return to normal activities were very encouraging for PAE. Seventy-one percent of PAE cases were performed as outpatient or day cases. In contrast, 80% of TURP cases required at least 1 night of hospital stay, and the majority required 2 nights. CONCLUSION: Our results indicate that PAE provides a clinically and statistically significant improvement in symptoms and QoL, although some of these improvements were greater in the TURP arm. The safety profile and quicker return to normal activities may be seen as highly beneficial by patients considering PAE as an alternative treatment to TURP, with the concomitant advantages of reduced length of hospital stay and need for admission after PAE. PAE is an advanced embolization technique demanding a high level of expertise, and should be performed by experienced interventional radiologists who have been trained and proctored appropriately. The use of cone-beam computed tomography is encouraged to improve operator confidence and minimize non-target embolizations. The place of PAE in the care pathway is between that of drugs and surgery, allowing the clinician to tailor treatment to individual patients' symptoms, requirements and anatomical variation.

Decision theory for comparing institutions.

Various forms of performance assessment are applied to public service institutions, such as hospitals, schools, police units, and local authorities. Difficulties arise in the interpretation of the results presented in some established formats because they require a good understanding and appreciation of the uncertainties involved. Usually the results have to be adapted to the perspectives of the users-managers of the assessed units, a consumer, or a central authority (a watchdog) that dispenses awards and sanctions. We present a decision-theoretical approach to these and related problems in which the perspectives are integrated in the analysis and its results are choices from a finite list of options (alternative courses of action).

Variation of the rates of necrotising enterocolitis in the neonatal networks in England.

Necrotising enterocolitis is an oft-fatal disease of the intestinal tract in neonates born prematurely and with low birthweight. We study the variation of its rates across the neonatal networks (groups of hospital-based neonatal care units) in England. We illustrate the problematic nature of hypothesis testing for a key decision, which an analysis is meant to inform, and apply an approach based on decision theory. We emphasise the role of sensitivity analysis in dealing with the ambiguity encountered in the process of eliciting information about the perspective of the client or sponsor for whom the analysis is conducted. In the application based on a fiducial distribution, the likelihood is approximated by a linearising transformation of the score function. Copyright © 2017 John Wiley & Sons, Ltd.

Comparing two treatments by decision theory.

Decision theory is applied to the general problem of comparing two treatments in an experiment with subjects assigned to the treatments at random. The inferential agenda covers collection of evidence about superiority, non-inferiority and average bioequivalence of the treatments. The proposed approach requires defining the terms 'small' and 'large' to qualify the magnitude of the treatment effect and specifying the losses (or loss functions) that quantify the consequences of the incorrect conclusions. We argue that any analysis that ignores these two inputs is deficient, and so is any ad hoc way of taking them into account. Sample size calculation for studies intended to be analysed by this approach is also discussed. Copyright © 2016 John Wiley & Sons, Ltd.

A decision-theoretical alternative to testing many hypotheses.

Testing a large number of hypotheses is a key problem in the analysis of microarray experiments and in other studies in which many elementary experiments are conducted, and the exceptions among them, for which a particular hypothesis does not hold, have to be identified. A class of approaches to this problem is based on controlling the false discovery rate, even though failure to discover should also be considered. We develop a decision-theoretical approach in which errors (misclassifications) of the two kinds are associated with uneven losses, and the total expected loss in the collection of the classifications (decisions made or options selected) is minimized.

Classification in two-stage screening.

Decision theory is applied to the problem of setting thresholds in medical screening when it is organised in two stages. In the first stage that involves a less expensive procedure that can be applied on a mass scale, an individual is classified as a negative or a likely positive. In the second stage, the likely positives are subjected to another test that classifies them as (definite) positives or negatives. The second-stage test is more accurate, but also more expensive and more involved, and so there are incentives to restrict its application. Robustness of the method with respect to the parameters, some of which have to be set by elicitation, is assessed by sensitivity analysis.

Survey of the clinical assessment and utility of near-infrared cerebral oximetry in cardiac surgery.

OBJECTIVES: Near-infrared cerebral oximetry increasingly is used for monitoring during cardiac surgery. Nonetheless, the scientific basis for incorporating this technology into clinical practice, the indications for when to do so, and standard diagnostic and treatment algorithms for defining abnormal values are yet to be rigorously defined. The authors hypothesized that there would be (1) variation in clinical use and practices for near-infrared spectroscopy (NIRS), and (2) variation in management of patients when clinicians are provided with NIRS information. In order to test this hypothesis, they sought to assess the nature and strength of response heterogeneity among anesthesiologists and cardiac perfusionists when provided with cardiac surgery patient scenarios and cerebral oximetry data. DESIGN: A prospectively collected survey. SETTING: A hospital-based, multi-institutional, multinational study. PARTICIPANTS: By e-mail, the authors surveyed the membership of the Society of Cardiovascular Anesthesiologists and the online Cardiovascular Perfusion Forum. INTERVENTIONS: This survey was focused on ascertaining what actions clinicians would take in each scenario, given case information and cerebral oximetry tracings. Questions were based on 11 patient scenarios selected to represent small, large, symmetric, or asymmetric decreases in measured regional cerebral oxygen saturation (rScO2) encountered during cardiac surgery. Information on the respondents' (n = 796; 73% anesthesiologists) clinical practice, demography, and cerebral oximetry utilization was collected. An index of dispersion was used to assess response heterogeneity overall and within demographic subgroups. MEASUREMENTS AND MAIN RESULTS: The majority of respondents indicated that cerebral oximetry monitoring was either useful or an essential monitor, especially perfusionists and clinicians who used cerebral oximetry most frequently. There were marked differences in responses between perfusionists and anesthesiologists for 4 of the 6 scenarios (p<0.005 for each of these 4 scenarios) occurring during cardiopulmonary bypass. Scenarios having greatest rScO2 reduction or asymmetry in rScO2 were associated with the highest dispersion, indicating least agreement in management. Scenarios with mild or moderate rScO2 reduction were associated with the lowest dispersion, indicating greater agreement in management. CONCLUSIONS: Although experimental data gradually are accumulating to support the role for cerebral oximetry monitoring during cardiac surgery, the results of the present survey support the view that its role remains poorly defined, and consensus for its appropriate use is lacking. Importantly, the authors observed marked variation in the use, perceived utility, and management of patients for 4 of the 6 CPB scenarios between perfusionists and anesthesiologists who share the management of CPB. These findings support the need for well-designed, adequately-powered clinical trials examining the value of this technology.

Screening as an application of decision theory.

We develop a decision-theoretical approach to setting the threshold for a screening procedure that declares each examined subject as a positive or a negative. It is fundamentally different from maximising the Youden index. The method incorporates the consequences of the two kinds of bad decisions (false positives and false negatives) by means of a set of plausible loss functions elicited from a subject-matter expert or committee. We present details for several classes of loss functions and within-group distributions of the outcomes. We outline extensions related to mixture distributions and compositions of loss functions. We illustrate the method on simulated examples and apply it to real datasets.

Handling the limit of detection by extrapolation.

A general method of estimation with a variable observed subject to a limit of detection is introduced. It is based on extrapolation of the estimates obtained by increasing the limit of detection. Theoretical arguments support the method in some special cases, and it is explored by simulations. Several examples are presented and adaptations for other kinds of censoring explored.

Feeding during neonatal therapeutic hypothermia, assessed using routinely collected National Neonatal Research Database data: a retrospective, UK population-based cohort study.

BACKGROUND: Therapeutic hypothermia is standard of care in high-income countries for babies born with signs of hypoxic ischaemic encephalopathy, but optimal feeding during treatment is uncertain and practice is variable. This study aimed to assess the association between feeding during therapeutic hypothermia and clinically important outcomes. METHODS: We did a population-level retrospective cohort study using the UK National Neonatal Research Database. We included all babies admitted to National Health Service neonatal units in England, Scotland, and Wales between Jan 1, 2010, and Dec 31, 2017, who received therapeutic hypothermia for 72 h or died during this period. For analysis, we created matched groups using propensity scores and compared outcomes in babies who were fed versus unfed enterally during therapeutic hypothermia. The primary outcome was severe necrotising enterocolitis, either confirmed at surgery or causing death. Secondary outcomes include pragmatically defined necrotising enterocolitis (a recorded diagnosis of necrotising enterocolitis in babies who received at least 5 consecutive days of antibiotics while also nil by mouth during their neonatal unit stay), late-onset infection (pragmatically defined as 5 consecutive days of antibiotic treatment commencing after day 3), survival to discharge, measures of breastmilk feeding, and length of stay in neonatal unit. FINDINGS: 6030 babies received therapeutic hypothermia, of whom 1873 (31·1%) were fed during treatment. Seven (0·1%) babies were diagnosed with severe necrotising enterocolitis and the number was too small for further analyses. We selected 3236 (53·7%) babies for the matched feeding analysis (1618 pairs), achieving a good balance for all recorded background variables. Pragmatically defined necrotising enterocolitis was rare in both groups (incidence 0·5%, 95% CI 0·2-0·9] in the fed group vs 1·1% [0·7-1·4] in the unfed group). The enterally fed group had fewer pragmatically defined late-onset infections (difference -11·6% [95% CI -14·0 to -9·3]; p<0·0001), higher survival to discharge (5·2% [3·9-6·6]; p<0·0001), higher proportion of breastfeeding at discharge (8·0% [5·1-10·8]; p<0·0001), and shorter neonatal unit stays (-2·2 [-3·0 to -1·2] days; p<0·0001) compared with the unfed group. INTERPRETATION: Necrotising enterocolitis is rare in babies receiving therapeutic hypothermia. Enteral feeding during hypothermia is safe and associated with beneficial outcomes compared with not feeding, although residual confounding could not be completely ruled out. Our findings support starting milk feeds during therapeutic hypothermia. FUNDING: UK National Institute for Health Research Health Technology Assessment programme 16/79/13.

Estimation of the effect size in meta-analysis with few studies.

Meta-analysis is often conducted with only a small number of studies. Adjustments of the (restricted) maximum likelihood estimator of the effect size are derived and their gains in efficiency are explored. The proposed estimators are applied to three sets of studies.

Should the dilated ascending aorta be repaired at the time of bicuspid aortic valve replacement?

OBJECTIVES: Bicuspid aortic valve (BAV) is the most common congenital valvular abnormality and frequently presents with accelerated calcific aortic valve disease, requiring aortic valve replacement (AVR) and thoracic aortic aneurysm and dissection. Supporting evidence for Association Guidelines of aortic dimensions for aortic resection is sparse. We sought to determine whether concurrent repair of dilated or aneurysmal aortic disease during AVR in patients with BAV substantially improves morbidity and mortality outcomes. METHODS: Mortality and reoperation outcomes of 1301 adults with BAV and dilated aorta undergoing AVR-only surgery were compared to patients undergoing AVR with aortic resection (AVR-AR) using Cox proportional hazards modelling and patient matching. RESULTS: Clinically important differences in patient characteristics, aortic valve function and aortic dimensions were identified between cohorts. Event rates were low, with rates of reoperation and death within 1 year of only 1.8% and 5.4%, respectively, and no aortic dissection observed during follow-up. There were no significant differences in reoperation or mortality outcomes between the AVR-only and AVR-AR cohorts. Age, aortic dimension or a combination thereof was not associated with better or worse outcomes after each AVR-AR compared with AVR. CONCLUSIONS: We conclude AVR-only and AVR-AR surgery have low morbidity and mortality and have utility over a wide range of age and aortic sizes. Our results do not provide support for the 45-mm aortic dimension recommended in the current guidelines for aortic resection while performing AVR or any other specific dimension.

Duration of Postoperative Atrial Fibrillation After Cardiac Surgery Is Associated With Worsened Long-Term Survival.

BACKGROUND: Studies of the effects of postoperative atrial fibrillation (poAF) on long-term survival are conflicting, likely because of comorbidities that occur with poAF and the patient populations studied. Furthermore, the effects of poAF duration on long-term survival are poorly understood. METHODS: We utilized a prospectively collected database on outcomes of cardiac surgery at a large tertiary care institution between August 2001 and December 2010 with survival follow-up through June 2015 to analyze long-term survival of patients with poAF. In addition, we identified patient- and procedure-related variables associated with poAF, and estimated overall comorbidity burden using the Elixhauser comorbidity index. Survival was compared between patients with poAF (n = 513) and a propensity score matched control cohort, both for all patients and separately for subgroups of patients with poAF lasting less than 2 days (n = 218) and patients with prolonged poAF (n = 265). RESULTS: Patients with poAF were older and had a higher burden of comorbidities. Survival was significantly worse for patients with poAF than for the matched control group (hazard ratio 1.43, 95% confidence interval: 1.11 to 1.86). That was driven by decreased survival among patients with prolonged poAF (hazard ratio 1.97, 95% confidence interval: 1.37 to 2.80), whereas survival of patients with poAF for less than 2 days was not significantly different from that of matched controls (hazard ratio 0.91, 95% confidence interval: 0.60 to 1.39). CONCLUSIONS: After close matching based on comorbidity burden, prolonged poAF is still associated with decreased survival. Therefore, vigilance is warranted in monitoring and treating patients with prolonged poAF after cardiac surgery.

On selection and composition in small area and mapping problems.

Maps in which small areas, such as districts, are represented by colours, shades or symbols with sizes determined by the values of estimates are regarded as an indispensable graphical output of analyses concerned with the geographical detail of economic, social, ecological and epidemiological phenomena. The distortion of the distribution of the district specific quantities in such maps, due to misrepresentation of the uncertainty about the estimated values, is discussed, and an alternative based on drawing so-called plausible maps is described. We highlight the pervasive and nonignorable nature of the selection process that identifies the quantity (target) to be estimated. A problem specific to disease mapping is what action, often one of a discrete set, to take in response to the results of an analysis. We argue that the costs (values) associated with correct and incorrect decisions should be integrated in the analysis and, when an analytical treatment is not feasible, plausible scenarios played out by simulations.