Definitions of remission in psoriasis: a systematic literature review from the National Psoriasis Foundation.

Psoriasis studies increasingly employ outcomes that indicate complete disease resolution, yet remission and cure are poorly defined for psoriasis. We conducted a systematic literature review to identify definitions of psoriasis remission and cure reported in the literature. Medline, EMBASE, and The Cochrane Central Register of Controlled Trials databases were searched on July 22, 2020, for full-text studies providing definitions for psoriasis remission/cure. Definitions were analysed descriptively for endpoint, time-frame, on/off treatment, patient-reported outcomes, and disease domains. We identified 106 studies that provided 41 unique remission definitions. Most definitions included endpoints based on Psoriasis Area and Severity Index (PASI), such as PASI75 (n = 16 studies), PASI90 (n = 10), PASI100 (n = 10), and PASI of 0 (n = 3), and descriptive endpoints related to 'skin clearance' (n = 18). Few definitions specified time-frame, on/off treatment or other psoriasis-related disease domains. One small consensus-initiative defined drug-free remission for plaque psoriasis by BSA of 0 without any therapy for at least 12 months. While there is no cure for psoriasis, seven studies defined psoriasis cure using similar endpoints to those used to define remission. We identified a variety of definitions of psoriasis remission. These results will inform the development of consensus-based definitions for psoriasis remission to support efforts to improve research and clinical outcomes.

[Surgical reconstruction for radiation-induced extracranial vertebral artery stenosis: a case report].

We report a case of symptomatic extracranial vertebral artery stenosis after radiation therapy. This 49-year-old female received radiation therapy to the neck for nasopharyngeal carcinoma 11 years earlier, was admitted because of continuous dizziness and a floating sensation. Magnetic resonanse imaging showed no abnormalities, but an aortography demonstrated complete occlusion of the right common carotid artery as well as occlusion of the right vertebral artery and severe stenosis of the left vertebral artery at its origin, which was presumed to be the result of previous radiation therapy. Percutaneous transluminal angioplasty (PTA) for the left vertebral artery was performed using conventional balloon treatment, which resulted in wall dissection. Because of this, she underwent end-to-side vertebral artery to subclavian artery transposition, and she has had no further ischemic events science that time. PTA has been successfully performed as the first treatment of choice for vertebral artery stenosis, but surgical reconstruction can be a therapeutic management of choice for cases of failed PTA.

[Therapeutic outcome of spasmodic torticollis].

We investigated 117 patients with spasmodic torticollis who had visited us to seek for appropriate treatment in these 14 years. They were 71 men and 46 women, aged 44 +/- 14 (mean +/- SD) years, and suffered from this disorder during 4 +/- 5 years, maximum 26 years. Involuntary abnormal head positions, not only torticollis but also laterocollis and antero- or retrocollis, were contained in this study. Most of them were torticollis due to idiopathic focal dystonia. One or more courses of alcoholization therapy was accomplished in 82 patients who wished to be done. This therapy course consisted of about ten times totally of 99% ethanol injection to the motor point of two most hypertonic neck muscles, either side of the sternocleidomastoideus and the opposite side of the splenius in most cases, repeated every 2 or 3 weeks. One patient received as many as 98 times of this injection and resolved completely. Training to reinforce antagonistic muscles was also instructed. Twenty-one patients (26%) were resolved completely after this treatment. Fifty-four patients (66%) were ameliorated and satisfied partially, but 18 of them relapsed in 1 to 4 years after the treatment and were obliged to repeat one more course of this treatment. On the other hand, in five patients their torticollis improved under certain drug therapy alone. Sixteen patients (14%) gave up to continue the treatment within two months, and 14 patients (12%) dropped out before starting the therapy. This alcoholization therapy resulted in amelioration of torticollis in about 90% of the patients with a long effective period. Nevertheless, this alcohol injection is painful, and requires 5 to 6 months to be completed. In 2 patients who had already received many times of this injection, sudden hoarseness occurred one day immediately after the alcohol injection to the sternocleidomastoideus. This complication was presumably brought about by the unexpected infiltration of alcohol to the laryngeal area, located posterior to that muscle. They recovered in two months, but careful attention should be paid to the adverse effects. If botulinum toxin be available also in our country, we will be able to have another choice of therapy and the treatment of this disorder will become easier.

[Study on the latency difference between compound muscle and sensory nerve action potentials].

In motor nerve conduction studies compound muscle action potentials (CMAPs) appear later than sensory nerve action potentials (SNAPs). This time lag originates from the conduction delay at the distal motor axon, neuromuscular transmission time and muscle action potential induction time. To investigate the latency difference between CMAPs and SNAPs we studied 46 healthy individuals, 46 patients with diabetes mellitus and 33 patients with carpal tunnel syndrome, using the lumbrical and interossei recording method. In this method the recording active electrode was placed on the 2nd lumbrical muscle and the reference electrode on the proximal palmar aspect of the index finger. Supramaximal stimulation was given to the median or ulnar nerve trunk at 9-cm proximal to the recording active electrode. The CMAP from the 2nd lumbrical muscle (L) and the SNAP from the digital nerve (N) were recorded after median nerve stimulation, and the CMAP from the 2nd interossei muscles (I) was recorded after ulnar nerve stimulation. The residual latency, which is arbitrary defined as the latency difference (L-N) in this study, was 1.38 +/- 0.15 (mean +/- SD) msec in healthy individuals. About 1 msec of the residual latency is regarded as the time for neuromuscular transmission and the time to evoke muscle activities. Thus, the conduction delay at the distal motor axon was calculated as about 0.4 msec in healthy individuals. The residual latency was relatively constant in 29 diabetic patients without conduction delay across the carpal tunnel, which was defined by the latency difference (L-I) < or = 0.4 msec. Their sensory nerve conduction velocities (calculated from N latency) were always above 40 m/sec. On the other hand in diabetic patients with conduction delay across the carpal tunnel, which was defined by the latency difference (L-I) > 0.4 msec, the residual latency gradually increased as the sensory nerve conduction velocity decreased. Their sensory nerve conduction velocities were mostly less than 40 m/sec. The similar relationship was observed in patients with carpal tunnel syndrome without diabetes mellitus. We consider that the diabetic neuropathy alone doesn't cause the increase of the residual latency. Instead, severe conduction delay across the carpal tunnel decreases the N velocity and increases the residual latency. We can also regard the relationship between the latency difference (L-N) and N velocity as being in inverse proportion. Perhaps the increase of the residual latency was simply caused by the proportional decrease in the conduction velocity at the distal motor axon, not by the special mechanism concerning to the carpal tunnel syndrome. This paper presented the electrophysiological changes seen in the distal segment secondary to the proximal entrapment.

[Innervation pattern to the extensor digitorum brevis by deep peroneal nerve and accessory deep peroneal nerve].

On the electrophysiological basis, extensor digitorum brevis(EDB) muscle is innervated electrophysiologically not only by deep peroneal nerve(DPN) but also by accessory deep peroneal nerve(ADPN), an anomalous branch of superficial peroneal nerve, with a prevalence of 17-28%. We investigated 23 patients who had both DPN and sufficient ADPN innervation to the EDB on the intramuscular distribution of DPN and ADPN innervation to the medial and lateral side of the EDB. Recording electrodes were placed on the medial and lateral edges of the EDB with a supramaximal stimulation to the anterior or lateral ankle, compound muscle action potential (CMAP) of DPN or ADPN innervation was recorded. In 19 patients (83%) the DPN innervation was larger than or equal to the ADPN innervation. Only in 4 patients (17%) the ADPN innervation obviously exceeded the DPN innervation. DPN enters to the EDB from the medial side, and ADPN from the lateral side of the EDB. In 16 patients(70%) the DPN innervation was relatively large and the ADPN innervation was relatively small at the medial side of the EDB, and vice versa at the lateral side of the EDB. These distributions were almost uniform in 5 patients(22%). This study clarified that a biased larger DPN innervation and smaller ADPN innervation to the medial side of the EDB, and vice versa to the lateral side of the EDB in the majority cases. In some cases diffuse innervation to the EDB was found.

[Spontaneous dissection of the anterior cerebral artery presenting subarachnoid hemorrhage and cerebral infarction: a case report].

A case is reported of anterior cerebral artery dissecting aneurysm presenting with subarachnoid hemorrhage and cerebral infarction. A 50-year-old man presented with sudden onset of weakness of the left lower limb was admitted to our hospital. CT scan on admission showed a subarachnoid hemorrhage in the interhemispheric fissure and CT on the 6th day demonstrated a cerebral infarction on the right medial frontal lobe. A carotid angiogram 12 hours after the onset showed no aneurysmal lesion, but, the angiogram repeated 11 days after the onset revealed an aneurysmal dilatation with distal narrowing at the right A2-A3 segment. To prevent rebleeding, we performed a wrapping procedure through the interhemispheric route on the 18th day after onset. The postoperative course was uneventful. We reviewed 27 previously reported cases with symptomatic dissecting aneurysm confined to the anterior cerebral artery.

[Polyneuropathy index-revised in the evaluation of diabetic neuropathy].

The polyneuropathy index-revised(PNI-R), based on 8 electrophysiological parameters(conduction velocities and F-latencies), was constructed to obtain an overall estimation of peripheral nerve conduction in diabetic patients, taking PNI as a model. PNI was calculated as a mean percentage of the normal on 12 velocity or latency parameters on motor nerve conduction studies. PNI-R is composed of 8 parameters; motor nerve conduction velocities in the forearm or leg segment and F-wave latencies after wrist or ankle stimulation concerning to the median, ulnar, peroneal and posterior tibial nerves. F-wave latencies were adjusted to 160 cm height and used reciprocals to compare with the normal values. Subjects were 101 patients with diabetes mellitus. Correlation of PNI-R or PNI with other parameters or indices on conventional sensory and intrafascicular conduction studies or items concerning to the diabetes mellitus were studied. Coefficient of correlation between PNI-R and PNI was as high as 0.97. The mean value of PNI-R was 0.6% smaller than PNI. This was presumably due to the greater influence of the peroneal parameters, weighted more in PNI-R than in PNI. Peroneal nerve is known to be sensitive to various neuropathies, and is often damaged independently. Each parameter composing PNI-R had a close relationship with PNI-R itself. Mutual independence between 8 parameters was considered to be enough. Among neuropathic signs Achilles tendon reflex in particular, and among diabetic complications retinopathy in particular, had a high degree of correlation with PNI-R. These results were identical both with PNI-R and PNI. We can save 20-30% of time in measuring PNI-R as compared to measure PNI, and the usefulness of PNI-R was as well as PNI. Therefore, using PNI-R as substitute for PNI is considered to be appropriate in the evaluation of diabetic polyneuropathy. Between parameters concerning to the median nerve F-wave latency correlated less with PNI-R than motor nerve conduction velocity in the forearm segment. Presumably this was owing to an unrecognized subclinical carpal tunnel syndrome, often observed in patients with diabetes mellitus. PNI-R will be an excellent index to express the function of peripheral nerve conduction, which can be retarded by the axonal degeneration in diabetes mellitus.

[Prevalence of Martin-Gruber anastomosis on motor nerve conduction studies].

Prevalence of median to ulnar anastomosis in the forearm(Martin-Gruber anastomosis; MGA) to the first dorsal interosseous(FDI), abductor digiti quinti (ADQ) and adductor pollicis(AP) was investigated. Subjects contained 106 patients with normal nerve conduction or patients with various neuropathies. Recording electrodes were placed on the motor point of FDI, ADQ and AP. Supramaximal stimulations were given to the median and ulnar nerves at the wrist or above the elbow. The diagnosis of MGA was made by the following criteria; amplitude of compound muscle action potential(CMAP) increased after elbow stimulation as compared with the wrist stimulation in median nerve conduction studies. The corresponding decrease in CMAP amplitude was found after above elbow stimulation as compared with the wrist stimulation in ulnar nerve conduction studies. No MGA was found in 80(75%) out of 106 patients. MGA to FDI was found in all 26 patients who had MGA. MGA to ADQ and AP was found in 11% and 10% of the patients, respectively. Only 8 out of 26 patients had MGA to all 3 muscles. In the presence of MGA median motor nerve conduction studies demonstrate larger CMAP, with a small initial positivity, after elbow stimulation than after wrist stimulation. And moreover, ulnar motor nerve conduction studies reveal a conduction block-like finding in the forearm. In this study MGA was found in 25% of the patient to FDI, in 11% to ADQ and in 10% to AP. Although a very small MGA might be overlooked in our method, such a small MGA doesn't mislead us into erroneous interpretation of motor nerve conduction studies.

Evaluation of distal and proximal axonal degeneration in patients with carpal tunnel syndrome.

In patients with carpal tunnel syndrome, varying degrees of demyelination and axonal degeneration occur in the median nerve. Only a few studies have examined axonal degeneration produced at proximal to the lesion. In this study proximal axonal degeneration was evaluated and compared with other parameters. In 40 consecutive CTS patient hands, distal latency (DL), compound muscle action potential amplitude (CMAP) and motor conduction velocity (MCV) were analyzed by conventional motor nerve conduction studies. Intrafascicular compound nerve action potential amplitude (N-CNAP) at the elbow after wrist simulation and its nerve conduction velocity (NCV) between wrist and elbow were also analyzed. The negative correlation of DL with CMAP was statistically significant (r = 0.577, p < 0.001). CMAP was correlated with either MCV (r = 0.537, p < 0.001) or N-CMAP (r = 0.710, p < 0.001). A significant correlation of MCV with NCV (r = 0.517, p < 0.001) was also indicated. There were no any other significant correlation among the parameters. In CTS the degree of demyelination and axonal degeneration influence the prognosis for nerve recovery after decompressive surgery. DL is mainly influenced by demyelination that results in conduction block and slowing at the carpal tunnel. CMAP and N-CNAP indicate the degree of axonal degeneration at distal and proximal to the compression site. As in electrophysiologic evaluation of mononeuropathies, proximal axonal degeneration is best assessed by both stimulation and recording electrode locationing proximal to the lesion. Recording of intrafascicular nerve action potential was a little invasive method, but it provided important informations. The negative correlation between DL and CMAP implies that distal axonal degeneration can occur in proportion to the conduction disturbance. Moreover, N-CNAP had a higher correlation with CMAP. The greater the distal axonal degeneration, the more the proximal axonal degeneration. Conduction velocity represents the velocity of the fastest conduction fiber, not the degree of axonal degeneration.

[Nerve conduction velocity of repeater F-waves is identical to that of M-waves].

F-wave normally varies in latency and waveform from one response to the next. But the number of identical responses in a series of F-waves may be increased with neurogenic atrophy consistent with a decreased number of motoneurons capable of responding to antidromic stimulation. They are called "repeater F-waves". We herein demonstrate some repeater F-waves observed in three patients with moderate or slight diabetic polyneuropathy. In their motor nerve conduction studies on the peroneal nerve the maximum conduction velocity was 33 m/sec in patient 1, 36 m/sec in patient 2 and 48 m/sec in patient 3. A total of 6 delayed indirect potentials were repeatedly evoked after nerve trunk stimulation. They fulfilled the characteristics of F-wave. Their conduction velocities in the leg segment were 27, 26, 23 m/sec in patient 1, 34, 33 m/sec in patient 2 and 46 m/sec in patient 3. Repeater F-waves are occasionally observed in patients with amyotrophic lateral sclerosis, cervical spondylosis or entrapment neuropathies, in which the number of motoneuron is decreased. In diabetic polyneuropathy some repeater F-waves were also observed in patients not only with moderate to severe neuropathy but also with normal nerve conduction. F-waves are generated by an antidromic backfiring of motor neurons, and they occur preferentially in large motor neurons. Larger motor neurons inhibit smaller axons through the activation of Renshaw cells. In our 3 patients conduction velocities of the repeated F-waves were all identical to the main component of M-wave. These observations reconfirmed the hypothesis that relatively large motor neurons generating F-waves are preferentially activated also in repeater F-waves.

[Comparison between Dyck's criteria and the polyneuropathy index-revised (PNI-R) in the electrophysiologic evaluation of diabetic neuropathy].

In Rochester diabetic neuropathy research by Dyck et al., abnormal value in two or more nerves was introduced into the nerve conduction criteria of diabetic neuropathy. Polyneuropathy index-revised(PNI-R) is calculated as the mean percentage of the normal of 8 parameters on the motor nerve conduction studies. They were motor nerve conduction velocities in the forearm or leg segment and F-wave latencies after wrist or ankle stimulation concerning to the median, ulnar, peroneal and posterior tibial nerves. F-wave latencies were adjusted to 160 cm height and used reciprocals in comparison with normal values. To compare these two indices, first we obtained the normal limit(1st or 99th percentile value) of each parameter from the data of 62 healthy individuals. Then in 78 patients with diabetes mellitus number of abnormal nerves and the PNI-R were investigated. Abnormal values were frequently observed in the categories of motor nerve conduction velocities and F-wave latencies. Amplitude of compound muscle action potential (CMAP) or sensory nerve action potential(SNAP) in each nerve had a large standard deviation. In such parameters abnormal rate was extremely low, because the lower limit of normal being very small. Nevertheless, sigma CMAP which means the summation of amplitudes of 3 CMAPs had as high as 53% of abnormal rate. The coefficient of correlation between number of abnormal nerves and the value of PNI-R mounted up to -0.87. Instead, the coefficient of correlation of sigma CMAP or sigma SNAP, which means the summation of amplitudes of ulnar and sural SNAPs, with PNI-R were 0.65 and 0.79, respectively. In 14 patients PNI-R was normal and the number of abnormal nerves was 0 or 1. In 59 both categories were abnormal, and only in 5 they were not coincide. As to the clinical signs PNI-R had better correlation than number of abnormal nerves with vibration threshold or degree of Achilles tendon reflex. sigma CMAP is a convenient index to detect the existence and the degree of neuropathy. This index expresses the degree of neurogenic muscular atrophy, though it doesn't always advance parallel to the decrease in number of motor nerves. sigma SNAP had higher coefficient of correlation with PNI-R or number of abnormal nerves than sigma CMAP. In conclusion, abnormal PNI-R and abnormal value in two or more nerves are both useful and coincide with each other in the detection of diabetic neuropathy. The PNI-R is an excellent quantitative index, and the PNI-R corresponds well with the number of abnormal nerves. These observations indicate that the number of nerves with abnormal value is also available as a simple and semi-quantitative index of diabetic neuropathy.

[Isolated extensor digitorum brevis involvement in the population of normal systemic nerve conduction velocities].

To investigate the characteristics of isolated atrophy in extensor digitorum brevis(EDB), we analysed 262 patients whose polyneuropathy index(PNI) was 90% or more than the normal mean value. The PNI was determined as the mean percentage of the normal in 12 indices concerning to the velocity or long distance latency in motor nerve conduction studies. Latencies were adjusted to 160 cm height. Amplitude of compound muscle action potential (CMAP) in EDB showed no correlation with the patient's age. Similar result was obtained as well when studied in 115 patients whose PNI level was 95% or more than the mean normal value. In 18(7%) out of 262 patients CMAP amplitude in EDB was 1 mV or less; larger prevalence(p < 0.05) in women(10%) than in men(4%). The number of motor units which innervate EDB decreases along with the age, but this age-related change could be compensated by the magnification of each motor unit. In Western reports isolated EDB palsy has a predilection for emaciated men. Instead, our results showed the predominance in women. We may have some factors other than in Western countries, for example customs to sit directly on the mat for a long time, in the occurrence of isolated EDB palsy. In conclusion, amplitude reduction in EDB CMAP may reflect the following two factors; neuropathy-related factor and another factor independent of age or neuropathy.

Embolization of posterior circulation aneurysms with detachable coils.

SUMMARY: We evaluated the usefulness of endovascular treatment of posterior circulation aneurysms with GDCs and IDCs. Five cases were treated with IDCs, and 15 cases were treated with GDCs. In this study, 8 aneurysms were identified at the basilar bifurcation, 3 at the PI segment of the posterior cerebral artery, 1 at the origin of the superior cerebellar artery, 2 at the vertebrobasilar junction, 1 at the origin of the posterior inferior cerebellar artery, 1 at the distal anterior inferior cerebellar artery, and 4 dissecting aneurysms at the vertebral artery. Thirteen of the aneurysms were small (< 12 mm), 5 were large (13-24 mm), and 2 were giant (> 25 mm). Of the 20 patients, 14 patients returned to their previous occupation. Patients with permanent deficits included 2 patients with infarction caused by thromboembolic complications during the embolization procedure, and 2 with infarction caused by vasospasm. There were 2 deaths. The outcomes of the patients seemed favorable. However, long-term followup is necessary to determine the usefulness of detachable coils.

Wrap-clipping with a Dacron mesh silastic sheet. Technical note.

A technique combining wrapping and clipping using a Silastic sheet coated with Dacron mesh is described for treatment of fusiform or broad-based cerebral aneurysms. This sheet is easily tailored to wrapping the aneurysm base while avoiding involvement of the cranial nerves or branching vessels. The sheet is semitransparent so that the caliber of the newly constructed parent artery is easily adjusted during wrap-clipping. After the aneurysm and the parent artery have been circumferentially wrapped with the sheet, aneurysm clips are applied on the sheet so that the base of the aneurysm is clipped between the two leaves of the sheet. This wrap-clipping technique avoids the risks involved in extracting the aneurysm from the parent artery. The Dacron mesh coating the outer surface and sufficient clip closing pressure are both helpful in preventing the clip blades from sliding. Similar previously reported techniques are reviewed and discussed in detail.

Cranio-nasal median splitting for radical resection of craniopharyngioma.

A new surgical approach for radical resection of craniopharyngioma is presented. This approach (cranio-nasal median splitting) involves craniotomy in the centre of the frontal bone, removal of the median portion of the supraorbital bar that incorporates the nasal bone, and detachment of the medial canthal ligaments. The frontal lobes, the cribriform plates, the planum sphenoidale, and the upper nasal cavities are split in the midline. The extraventricular surface of the hypothalamus, the pituitary stalk, and the posterior portion of the Willis' arterial ring are well visualized through the midline infrachiasmatic route. The intraventricular surface of the hypothalamus is also visible in the same operative field through the lamina terminalis and/or the anterior portion of the corpus callosum. This excellent visualization is quite helpful for minimizing operative injury to the hypothalamus and the pituitary stalk whichever surface of the third ventricular floor the tumour is situated upon. Three cases of craniopharyngioma operated upon by this approach are presented. Discussions are focused not only on the indication, but on the advantages and disadvantages of this approach. The surgical techniques for reconstruction of the cranial base are also described, together with some precautions that should be taken to prevent possible postoperative complications.

[Posterior fossa hemorrhage 11 years after the use of silastic dural substitute: case report].

A case of posterior fossa hemorrhage is reported. The hemorrhage occurred 11 years after posterior fossa craniotomy with closure using a graft of silastic dural substitute. This 14 year-old girl underwent suboccipital craniotomy at the age of 3 years because of traumatic posterior fossa hemorrhage. The dura mater was repaired with a piece of dural substitute. Eleven years later, she developed severe headache and vomiting. Computerized tomography scanning revealed a high density area in the right posterior fossa. At exploration, a fresh clot was evacuated between the dural graft and the fibrous scar tissue. The resected fibrous scar tissue was 1cm thick, and was adhering to the duro glial scar tissue. Her postoperative course was uneventful, and she was discharged 2 weeks after surgery. Microscopic examination of the fibrous scar tissue revealed an outer layer of loose connective tissue with numerous capillaries. The inner layer consisted of dense connective tissue. It was speculated that fragile capillaries of fibrous scar tissue caused this hemorrhage. Surgeons should not forget this complication although silastic dural substitute is less used than it used to be.

A new intracranial Silastic encircling clip for hemostasis. Technical note.

A new encircling clip made of a silicone tube has been designed for treating accidentally injured cerebral vessels. No special holders are necessary. This clip can be tailored depending on the shape of the injured vessel. The clip is a simple and effective tool for achieving complete hemostasis.

[Comparison of five modes of dialysis in neurosurgical patients with renal failure].

In neurosurgical patients with renal failure, dialysis entails specific problems, chief of which is increased intracranial pressure and progressive brain edema as a result of rapid lowering of the serum osmolality. Another major problem is a tendency to hemorrhage, in response to either systemic heparinization or insufficient dialysis. The authors describe the results obtained with hemodialysis (HD), continuous arteriovenous hemofiltration (CAVH), continuous ambulatory peritoneal dialysis (CAPD), continuous peritoneal dialysis (CPD), and intermittent peritoneal dialysis (IPD). Nine patients were treated with HD, one with CAVH, five with CAPD or CPD, and two with IPD. Three of the six patients treated with continuous dialysis (CAVH, CAPD, and CPD) died, whereas intermittent dialysis (HD and IPD) carried an 82% mortality rate (nine of 11 patients). The causes of death were progressive brain edema in three cases, intracranial hemorrhage in three, gastrointestinal bleeding in three, overhydration due to insufficient dialysis in one, septicemia in one, and rupture of a cerebral aneurysm in one. Continuous dialysis appeared to be superior to intermittent dialysis in these neurosurgical patients in that it produced less brain edema and was less often associated with hemorrhage due to insufficient dialysis. In HD and CAVH, systemic heparinization was also thought to account for the high incidence of hemorrhage. However, CAVH with short half-life anticoagulants may be useful in patients who have abdominal complications and are therefore not suitable candidates for peritoneal dialysis.