Samuel J. Hessel

The Diagnostic Accuracy and Complications of Closed Lung Biopsies

The diagnostic yields and complication rates of closed lung biopsies were determined by responses to a questionnaire. Of 5,255 procedures reported, 30% were aspiration biopsies (AB), 7% tissue core biopsies (CB), 2% trephine biopsies (TR), 37% bronchial brush biopsies (BB), and 23% transbronchial biopsies (TB). Of these biopsies, 70% were done for suspected neoplasms. The diagnostic accuracy rates were: AB=82%, CB=84%, TR=86%, BB=61%, and TB=56%. Mortality rates were: AB=0.1%, CB=0.3%, TR=2.9%, BB=0.0%, and TB=0.3%. Transthoracic needle biopsies (AB, CB) had the highest overall diagnostic accuracy, with an associated low mortality and moderate morbidity.

Improving Diagnostic Accuracy: A Comparison of Interactive and Delphi Consultations

Consultation among physicians on difficult diagnostic problems is commonly used to improve the accuracy of medical decisions. Such consultation is most often informal and interactive. Nevertheless, behavioral studies suggest that non-interactive techniques may be more effective problem solving methods. Of these the Delphi approach, involving pooling and feedback of anonymously contributed information, has generated particular interest. To assess the relative effectiveness of independent decision making, interactive group consultation, and Delphi techniques in a clinical setting we compared the diagnostic accuracy of 17 radiologists interpreting radiologic examinations in these settings. Interactive consultation improved performance by 69% compared to radiologists interpreting the studies individually. In addition, two Delphi strategies each produced an additional 20% mean improvement in accuracy over interactive consultation. Whereas interactive consultation improved the accuracy of the best individual readers by only 6%, a Delphi model improved their performance by 25%. Thus, Delphi was an effective, easily applied method of clinical consultation whose usefulness in other clinical setting should be evaluated.

Differing attenuation coefficients of normal and infarcted myocardium

There are significant differences in attenuation coefficients between normal and infarcted myocardium measurable with a computerized transaxial tomographic scanner. Additionally, iodinated contrast material administered prior to killing the tests animals resulted in excellent visualization of the blood-myocardial interface at a time when standard radiographs detected no differences between the ventricular cavity and the myocardial wall. These natural and induced changes in attenuation coefficients offer a new approach to evaluating and understanding the processes of tissue injury and death. Their clinical relevance lies in application to the twin problems of myocardial infarction and the structure and function of the cardiac wall.

Heavy Metal Particulate Contrast Materials for Computed Tomography of the Liver

Silver iodide colloid was used as a model of a particulate hepatic contrast agent for computed tomography (CT). Following intravenous administration to rabbits, approximately 90% of the injected dose was phagocytized by the livers reticuloendothelial system, resulting in a four- to five-fold increase in the liver CT number. Suspensions of CeO2. Dy2O3. and Gd2O3, were prepared using stabilizers to prevent clumping. Particles of appropriate size for reticuloendothelial cell uptake were selected by centrifugation. Intravenous injections of the three suspensions increased the CT number of rabbit liver by approximately 30 Hounsfield units (HU: 1,000 scale) for each milligram of contrast material per gram of liver. An injected dose of 1 g of each experimental contrast agent resulted in a minimum addition of 250 HU to the liver CT number. A linear relationship was found between the CT number and its contrast material concentration. This relationship was also tested In vitro by ashing samples of livers containing various amounts of contrast material. Standard curves of CT number versus contrast material concentration in the liver were plotted, which predicted the amount of liver enhancement obtained after contrast material injection. There may be a use for heavy metal-containing particulate materials as hepatic contrast agents, since they opacify the liver more selectively, to a higher degree, and for longer periods than the conventional biliary and urographic iodinated contrast materials.

Hepatic contrast agents for computed tomography: high atomic number particulate material

We used a stepwise approach to identify, design, synthesize, and test new high atomic number particulate contrast agents that would be especially well suited for use with computed tomography (CT). Our goal was to produce extremely radiopaque compounds with highly selective biodistribution to the normal liver. In this way, dose requirements could be lessened and toxicity minimized. Suspensions of cerium, gadolinium, and dysprosium oxide particles and silver iodide colloid were tested and compared with standard agents. All four experimental agents were selectively concentrated in the reticuloendothelial systems of rats and rabbits. These compounds produced greater and longer opacification of normal livers and larger liver-to-tumor differences in rabbits with hepatic tumors than did equivalent amounts of standard, iodinated agents. Lesions as small as 5 mm were visible with CT. These experimental materials have favorable characteristics as hepatic contrast agents, but their toxicity and long term retention may limit clinical use.

Detection of myocardial ischemia in vitro by computed tomography.

The applicability of CT in the delineation of ischemic myocardium is defined. Twenty-six dogs had their left anterior descending coronary artery occluded, and radiolabeled microspheres were injected into the left atrium to measure myocardial blood flow. From 30 minutes to 7 weeks after coronary occlusion the hearts were excised and scanned in a CT head scanner. An unbiased observer divided selected CT scans slices into normal, abnormal, and borderline areas of attenuation coefficients; these were correlated with blood flow measurements and a progressive decrease in flow from normal to borderline and borderline to abnormal segments was found. As early as 2 hours after coronary arterial occlusion, areas of reduced attenuation coefficient, corresponding to regions of reduced blood flow, were seen. Areas of increased attenuation were seen in regions containing microcalcifications and fibrosis in several longer term experiments. Areas of reduced blood flow were visible immediately after coronary arterial occlusion with intravenous iodinated contrast material (1 ml/kg). CT detection of regions of reduced blood flow defined by radiolabeled microsphere is a promising means of detecting and sizing myocardial infarcts.

Renal collateral blood supply after acute unilateral renal artery occlusion.

Preformed arterial collaterals are critical to renal parenchymal survival after acute total renal artery occlusion. This study was designed to delineate and quantify preformed collaterals and assess their response to vasodilators. A Swan-Ganz catheter induced a sudden, total occlusion of a renal artery sufficient to reduce distal arterial pressure to near zero and prevent perfusion through the renal artery. Arteriography assessed the effectiveness of the occlusion and delineated the collateral arterial pathways. Strontium, cerium-, and chromium-labeled microspheres measured renal blood flow and cardiac output 1, 60, and 120 minutes after occlusion. In two additional series of experiments either contralateral nephrectomy was performed 5 to 8 days before the study, or dibenzylene, dopamine, or glucagon were administered in an attempt to increase blood flow through the collaterals. Collateral renal blood flow was demonstrated in all dogs. Mean blood flow to the occluded kidneys ranged from 0.13 +/- 0.05 cm3/minute/g to 0.22 +/- 0.08 cm3/minute/g, about 5% of control values. Neither prior contralateral nephrectomy nor vasodilator agents increased the flow to the obstructed kidneys. In the dogs with intact contralateral kidneys, however, there was a progressive decrease in cardiac output during the experiment, which was not found in uninephrectomized animals. We concluded that preformed arterial channels are available to maintain a small, but probably critical level of perfusion following sudden total occlusion of the renal artery. Neither hypertrophy due to prior contralateral nephrectomy nor active vasodilators modify flow through the preformed channels. It is likely that total renal ischemia provides a maximal stimulus for vasodilatation. The pattern of hind limb collaterals differed strikingly from those of the kidney, with maintenance of a greater portion of a normal flow and rapid increase in flow within 1 hour after femoral artery occlusion. Thus, data concerning collateral circulation cannot be generalized from one vascular bed to another even in the same species.

Surgery and the Progression of the Occlusive Process in Patients with Peripheral Vascular Disease

The angiograms and clinical records of 42 patients with arteriosclerosis obliterans who underwent repeat angiography were analyzed in order to correlate the effects of surgery with progression of the occlusive process in native vessels. Occlusive disease progressed significantly faster in operated limbs (77%) than in nonoperated limbs (44%). When progression occurred, it was more likely to take the form of occlusion in operated limbs (85%) than in nonoperated limbs (61%). Graft closure was associated with a 93% incidence of disease progression, but even limbs with patent grafts had a more rapid progression than the nonoperated limbs (62% vs. 44%). There was a good correlation between the presence of symptoms and the angiographic progression.

The composition of the radiologic report

In order to understand and evaluate fully the composition of the radiologic report, an experiment was performed in which eight readers each read the same 100 chest radiographs. The reports were dissected into individual statements, each of which was analyzed for accuracy, statement type (descriptive or etiologic), confidence, and specificity. In addition, each report was evaluated for stylistic factors of succinctness, orderliness, and the use of supplemental comments. Detailed analysis showed the radiologic report to be a series of largely descriptive statements in which confidence level, specificity, and orderliness were associated with the presence of errors in film interpretation. These factors can serve as important markers for identifying error-laden films. The use of specific, etiologic diagnoses in these film readings was limited. The composition of the report was not related to the length of radiologic training beyond the first year of residency. The implications of these findings concerning report composition in light of present day radiologic practice and new, automated radiologic reporting systems were discussed.

TRUE AND FALSE EXTERNAL CAROTID STEALS

Proximal external carotid artery occlusion, resulting in rerouting of flow from the vertebral to the external carotid artery is a distinct entity called external carotid steal . Diversion of flow from the vertebrobasilar to the external carotid system was induced technically in a patient, without significant vertebral basilar disease. Thus, demonstration of the vertebro-occipital anastomosis per se is insufficient evidence to diagnose occlusive vascular disease. Rate of filling, flow, run-off, site of catheter tip placement, and haemodynamic status of the patient must also be considered in ascertaining that significant opacification of the external carotid artery on vertebral injection represents a true external carotid steal. For corroboration, common carotid arteriography will define the external carotid artery proximal to the origin of the occipital artery. In patients undergoing angiography in the investigation of vertebrobasilar insufficiency, external carotid artery occlusion requires consideration in the differential diagnosis of vascular steals.