Christopher Blauth

Data, data, data…

‘Data, data, data. I must have more data’. One wonders what Sherlock Holmes would have made of the digital age and whether he would have admired our data obsessed society. The compulsion to measure and compare the intangible and transient is particularly evident in healthcare. In 2007, the big issue dominating healthcare agendas in rich and poor nations alike is that of controlling ever escalating costs. Cost pressures are now the major source of conflict between reimbursement agencies and healthcare providers and between managers and professionals at all levels. In this complex struggle all parties are increasingly coming to rely heavily on data. The temptation to use data for support rather than illumination is strong, and both sides collect data which is meaningful to them and with which they are familiar. The important battles of healthcare are now being fought not between ideologies, but in the more mundane field of competing datasets. In this issue of the journal, Park et al. (1) examine and compare the ability of clinical and administrative datasets to predict the clinical outcomes and thereby measure quality of care. Quality in healthcare is essentially personal, subjective and largely emotional and therefore very difficult to measure. Assessments can be made of the process of care in individual patients but are too cumbersome to be applied universally. If comparisons are to be made between hospitals, objective outcome measures are needed which may seem crude but have the merit of being easily measurable. The simplest of these is survival, a very appropriate consideration for coronary artery disease in the clinical settings of acute myocardial infarction and coronary artery bypass surgery which are the subject of the study by Park et al. But crude survival rates can be misleading and a system of risk stratification is needed to correct for age, comorbidity, and other factors. As Park points out collecting and analysing clinical data is time consuming and labour intensive. In contrast, administrative data such as disease and procedure codes from reimbursement claims are readily available. Clinical datasets such as the minimum cardiac surgery dataset of the Society of Thoracic Surgeons in the USA and the Society of Cardiothoracic Surgeons in the UK which provides good risk stratification are relatively unwieldy and require customised software to interrogate and analyse. Nevertheless, there is a balance to be struck between accuracy and simplicity and as Park demonstrates clinical datasets are much more accurate. Doctors will intuitively recognise that derived data can gloss over crucial differences. The prime example identified by Park is that administrative data based on disease and procedure codes cannot reliably distinguish between conditions which contribute to pre-existing comorbidity and conditions which arise as complications of acute myocardial infarction or coronary surgery. The former are descriptors of the case-mix presenting for treatment and the latter are outcome measures of the treatment itself, and inability to distinguish the two tends to defeat the intention of providing comparisons of treatment quality. These results from Korea echo previous findings from the USA, Canada and Italy. Attempts have been made to improve the predictive power of administrative data by adding selected clinical variables, laboratory values and combining both. The drivers for these strenuous analytical efforts include public health research and the need for quality assurance controls with the latter aiming to identify hospitals with poor outcomes. But the tools of public health research have been developed to investigate populations often numbering tens of thousands and may not be appropriate when transferred to the task of obtaining meaningful and reliable comparisons between hospitals treating mere hundreds of patients in a given time frame, usually one year. To do this properly we must, like Sherlock Holmes, not only demand more data, but also the right data.

Bovine pericardial versus porcine stented replacement aortic valves: Early results of a randomized comparison of the Perimount and the Mosaic valves

OBJECTIVE A stented bovine pericardial valve might be less obstructive than a stented porcine valve. This study compared early hemodynamic function in a prospective series of 99 patients randomized to receive either a Mosaic or Perimount replacement aortic valve. METHODS Echocardiography was performed early after surgery and at 1 year after surgery. Patients also filled in psychologic questionnaires and underwent a 6-minute walk. RESULTS The groups were matched demographically. The Perimount valve was significantly less obstructive in terms of mean pressure difference (11 +/- 5 vs 17 +/- 7 mm Hg; P < .0001), with a trend in favor of a larger effective orifice area (1.47 +/- 0.45 vs 1.28 +/- 0.46 cm(2); P = .05) postoperatively. There were no differences in left ventricular mass regression, aortic regurgitation, 6-minute walk, psychologic questionnaires, or mortality and clinical events. CONCLUSION The stented bovine pericardial valve was less obstructive than the stented porcine valve. Both valves were associated with similar and significant improvements in quality of life, exercise ability, and regression of left ventricular mass.

Which patients might be suitable for a septal occluder device closure of postinfarction ventricular septal rupture rather than immediate surgery

A best evidence topic in cardiac surgery was written according to a structured protocol. The question addressed was: which patients might be suitable for a transcatheter closure (TCC) of their postinfarction ventricular septal rupture (VSR) rather than immediate surgery? Thirty papers were found using the reported search, five of which represented the best evidence to answer the clinical question. The main limitations were the paucity of level 1 evidence on this topic. Only one study provided a prospective series as part of a registry; the remaining were retrospective cohort analyses. Although multivariate analysis may adequately control for measurable biases, unmeasured bias may still exist and influence the results. All studies agreed that timing of intervention is critical in determining the outcome. Patients preoperative haemodynamic status was a major determinant of postoperative survival. Successful outcome after TCC was in patients with simple defects, <15 mm in diameter that were in the sub-acute (>3.5 weeks) or chronic stage following acute myocardial infarct (AMI). Procedural success rate varied from 73.6% to 91%. Three of five studies looking at TCC closure concluded that procedural success does not necessarily translate to improved outcome. TCC in the acute setting (within four weeks of AMI) led to a high mortality (18%-65%) and increased incidence of complications (up to 41%). These included device embolization, major residual shunting, left ventricular rupture and malignant arrhythmias. One recent study correlated mortality to residual VSR [odds ratio (OR) 0.071, P=0.02], increased time from myocardial infarction to VSR diagnosis (OR 0.757, P=0.04) and increased time from VSR diagnosis to treatment (OR 0.758, P=0.04). The overwhelming recommendations were for immediate surgical intervention in cases of large VSR≥15 mm or where another indication for surgery exists. Three of five studies commented on a lack of a suitable device for PIVSR closure, with highest technical success with the Amplatzer device. In conclusion, small or medium PIVSR (<15 mm) can be treated definitively with Amplatzer closure in the sub-acute and chronic setting with comparable mortality to surgery. In the acute setting where surgery is deemed prohibitive TCC may provide a bridge to stabilize patients with shunt reduction, prior to surgery. The standard of treatment in large defects and PIVSRs in the acute setting remains early surgery.

Trends, indications and outcomes of cardiac implantable device system extraction: a single UK centre experience over the last decade

Background:  The rising number of device implantation has seen a parallel in the rising numbers of lead extraction. Herein we have analysed our experience in cardiac device and lead extraction in a single tertiary centre over the last decade.

Nominal size in six bileaflet mechanical aortic valves: A comparison of orifice size and biologic equivalence

OBJECTIVES Nominal size remains the standard by which valves are compared, but its relationship with orifice area and the patient tissue annulus diameter may differ according to valve design. The aims of this study were to measure the orifice size and compare biologic equivalence in six bileaflet mechanical heart valve designs. METHODS The inflow aspect of each of 29 valves was photographed then digitized, and the maximum internal diameter and orifice area were calculated. Biologic equivalence was assessed with a series of machined polypropylene blocks. RESULTS The orifice area ranged between 159 and 222 mm(2) for the six size 19 valves. The internal diameter ranged from 1.6 to 4.6 mm less than the manufacturers nominal size. Biologic equivalence assessed from an estimate of tissue annulus diameter with machined blocks ranged from 1.0 and 3.5 mm larger than nominal size for the intra-annular valves. This diameter ranged from 3.5 mm smaller to 1.5 mm larger than nominal size for the supra-annular valves. CONCLUSION There are major differences between nominal size and biologic equivalence. This may lead to confusion when attempting to make comparisons between different valve designs with the same nominal size. A clearer sizing nomenclature is required and could be based on in vitro assessment of tissue annulus diameter or an alphanumeric code.

Percutaneous lead and system extraction in patients with cardiac resynchronization therapy (CRT) devices and coronary sinus leads.

Background:  Cardiac resynchronization therapy (CRT) device and coronary sinus (CS) lead extraction is required due to the occurrence of system infection, malfunction, or upgrade. Published series of CS lead extraction are limited by small sample sizes. We present a 10‐year experience of CRT device and CS lead extraction.

Interventional and surgical management of aortic stenosis and coarctation

Two patients, aged 42 and 44 years old, presented with the combination of aortic valve stenosis and coarctation. To avoid potential problems associated with one stage repair, both patients underwent successful endovascular stent implantation for coarctation followed within 2 weeks by aortic valve replacement using pulmonary autografts (Ross operation). Spiral thoracic computed tomographic scans were performed within 1 week of stent implantation to ensure the absence of aneurysm formation related to the stent before cardiopulmonary bypass.

Transcatheter mitral valve replacement in mitral annulus calcification – “The art of computer simulation”

There is considerable interest in transcatheter prosthetic valve treatment for mitral valve disease in high-risk individuals. Although the presence of mitral annular calcium (MAC) may provide an anchoring zone for such devices, results to date have been modest with reported technical failure rates approaching 30% in specialist centers. This in part relates to the risk of left ventricular outflow tract obstruction and device dislodgment but also to the lack of specific imaging guidelines to plan for such procedures. We present the use of finite element analysis and computer simulation based on cardiac CT in three patients with severe MAC in whom transcatheter devices were considered. In the first two cases, the computer simulations were performed after the clinical procedure and were concordant with the clinical outcome. For the third case, computer simulation was performed prior to the clinical procedure. This indicated unsuitability for transcatheter device deployment and a subsequent medical management was adopted. Overall, our initial results suggest that computer simulation may have the potential to improve patient selection for transcatheter mitral valve replacement in the presence of significant MAC.

Massive thrombi on an implantable cardioverter-defibrillator lead in a patient with the antiphospholipid syndrome

A patient with the antiphospholipid syndrome, who was suboptimally anticoagulated, presented with two large thrombi attached to her implantable cardioverter-defibrillator lead. Anticoagulation was unsuccessful so, in view of the risk from embolization, the system was removed surgically and an epicardial system implanted to reduce the risk of future device-related thrombosis.

The diagnostic and therapeutic challenges of infective endocarditis presenting as acute stroke

An 81-year-old man presented with sudden-onset slurred speech and right-sided weakness commencing 2 hours previously. Initial National Institute of Health Stroke Scale score was 13. Non-contrast CT brain scan excluded haemorrhage, and he received intravenous thrombolysis. Medical history included tissue aortic valve replacement. On arrival, his temperature was 38°C. Two hours after thrombolysis he had two seizures, and his Glasgow Coma Scale score dropped from 15/15 to 6/15 (breakdown: eyes 1, voice 1, motor 4). Repeat CT demonstrated a left middle cerebral artery territory infarct, and he was transferred to the intensive care unit (figure 1 …