Scaling of suction-induced flows in bluegill: morphological and kinematic predictors for the ontogeny of feeding performance

During ontogeny, animals undergo changes in size and shape that result in shifts in performance, behavior and resource use. These ontogenetic changes provide an opportunity to test hypotheses about how the growth of structures affects biological functions. In the present study, we ask how ontogenetic changes in skull biomechanics affect the ability of bluegill sunfish, a high-performance suction feeder, to produce flow speeds and accelerations during suction feeding. The flow of water in front of the mouth was measured directly for fish ranging from young-of-year to large adults, using digital particle imaging velocimetry (DPIV). As bluegill size increased, the magnitude of peak flow speed they produced increased, and the effective suction distance increased because of increasing mouth size. However, throughout the size range, the timing of peak fluid speed remained unchanged, and flow was constrained to approximately one gape distance from the mouth. The observed scaling relationships between standard length and peak flow speed conformed to expectations derived from two biomechanical models, one based on morphological potential to produce suction pressure (the Suction Index model) and the other derived from a combination of morphological and kinematic variables (the Expanding Cone model). The success of these models in qualitatively predicting the observed allometry of induced flow speed reveals that the scaling of cranial morphology underlies the scaling of suction performance in bluegill.

Timing is everything: coordination of strike kinematics affects the force exerted by suction feeding fish on attached prey

During aquatic suction feeding, the predator opens its mouth and rapidly expands its buccal cavity, generating a flow field external to the mouth. The rapid expansion of the buccal cavity produces high fluid velocities and accelerations that extend only a short distance from the mouth (about half of one mouth diameter), and only persist for several milliseconds. Therefore, the predator must precisely time its strike to locate the prey within the narrow region of high flow, during the brief period when flow is at its peak. With flow being the agent for transferring force to the prey, the predator may enhance these forces by producing higher water velocities and faster acceleration at the mouth, but also through increasing the strike's accuracy,i.e. locating the prey closer to the mouth at the instant of peak flow speed. The objectives of this study were to directly measure forces exerted by bluegill Lepomis macrochirus on their prey and to determine how bluegill modify force output. Bluegill were offered ghost shrimp tethered to a load cell that recorded force at 5000 Hz, and feeding sequences were synchronously recorded using 500 Hz video. Peak forces exerted on attached 20 mm shrimp ranged from 0.005 N to 0.506 N. In accordance with the short duration of the strikes (average time to peak gape of ∼13 ms), the forces recorded were brief (∼12 ms from initiation to peak force), and force magnitude declined rapidly after peak force. Statistical analysis indicated that rate of buccal expansion, and prey size, but not strike initiation distance, significantly affected peak force. These observed variables were used with results from flow visualization studies to estimate the flow at the prey's location, which allowed the calculation of drag, pressure gradient force and acceleration reaction force. The relationship between these calculated forces and the measured forces was strong, indicating that the model can be used to estimate forces from strike kinematics. This model was then used to study the effects of strike initiation distance on peak force and on the rate of increasing force. Comparisons of model output to empirical results indicated that bluegill time their strike so as to exert an average of∼70% of the peak possible force on the prey, and that the observed strike initiation distance corresponded to the distance that maximized modeled force on an attached prey. Our results highlight the ability of bluegill to produce high forces on their prey, and indicate that precision and visual acuity play important roles in prey acquisition, beyond their recognized role in prey detection.

The forces exerted by aquatic suction feeders on their prey

Successful prey capture by aquatic suction feeders depends on the ability of the predator to generate a flow of water external to the mouth that overcomes any movements and forces that the prey uses to resist the suction flow. Elucidating the nature and magnitude of these forces is a key to understanding what limits suction feeding performance. We identify three potential forces produced by the suction flow field: drag, acceleration reaction and the fluid pressure gradient. Using a mathematical model parametrized with empirical data from feeding bluegill, Lepomis macrochirus, we explore the relative magnitude of these forces under three encounter scenarios with a 5mm diameter, spherical prey: an immobile mid-water prey; a similar prey that executes an escape response; and a prey item that grips a substratum. Contrary to the almost exclusive emphasis on drag in the suction feeding literature, it made a minor contribution to the total forces in all three cases. In all three scenarios, the pressure gradient is the largest of the three forces. These results are important because previous researchers have emphasized drag and have not explicitly recognized a role for the pressure gradient force in suction feeding. The simulations suggest previously unrecognized mechanisms that suction feeders can use to enhance the forces that they exert, by increasing the steepness of the pressure gradient that the prey item is exposed to. This can be accomplished either by increasing the rate of increase in fluid velocity or by restricting the size of the mouth aperture, which creates a steeper spatial gradient in pressure.

The pressures of suction feeding: the relation between buccal pressure and induced fluid speed in centrarchid fishes

Suction feeding fish rapidly expand their oral cavity, resulting in a flow of water directed towards the mouth that is accompanied by a drop in pressure inside the buccal cavity. Pressure inside the mouth and fluid speed external to the mouth are understood to be mechanically linked but the relationship between them has never been empirically determined in any suction feeder. We present the first simultaneous measurements of fluid speed and buccal pressure during suction feeding in fishes. Digital particle image velocimetry (DPIV)and high-speed video were used to measure the maximum fluid speed in front of the mouth of four largemouth bass and three bluegill sunfish by positioning a vertical laser sheet on the mid-sagittal plane of the fish. Peak magnitude of pressure inside the buccal cavity was quantified using a transducer positioned within a catheter that opened into the dorsal wall of the buccal cavity. In both species the time of peak pressure preceded the time of peak fluid speed by as much as 42 ms, indicating a role for unsteady flow effects in shaping this relation. We parameterized an existing model of suction feeding to determine whether the relationship between peak pressures and fluid speeds that we observed could be predicted using just a few kinematic variables. The model predicted much higher fluid speeds than we measured at all values of peak pressure and gave a scaling exponent between them (0.51) that was higher than observed (0.36 for largemouth bass, 0.38 for bluegill). The scaling between peak buccal pressure and peak fluid speed at the mouth aperture differed in the two species, supporting the recent conclusion that species morphology affects this relation such that a general pattern may not hold.

Multidimensional analysis of suction feeding performance in fishes: fluid speed, acceleration, strike accuracy and the ingested volume of water

Suction feeding fish draw prey into the mouth using a flow field that they generate external to the head. In this paper we present a multidimensional perspective on suction feeding performance that we illustrate in a comparative analysis of suction feeding ability in two members of Centrarchidae, the largemouth bass (Micropterus salmoides) and bluegill sunfish(Lepomis macrochirus). We present the first direct measurements of maximum fluid speed capacity, and we use this to calculate local fluid acceleration and volumetric flow rate. We also calculated the ingested volume and a novel metric of strike accuracy. In addition, we quantified for each species the effects of gape magnitude, time to peak gape, and swimming speed on features of the ingested volume of water. Digital particle image velocimetry (DPIV) and high-speed video were used to measure the flow in front of the mouths of three fish from each species in conjunction with a vertical laser sheet positioned on the mid-sagittal plane of the fish. From this we quantified the maximum fluid speed (in the earthbound and fish's frame of reference), acceleration and ingested volume. Our method for determining strike accuracy involved quantifying the location of the prey relative to the center of the parcel of ingested water. Bluegill sunfish generated higher fluid speeds in the earthbound frame of reference, accelerated the fluid faster, and were more accurate than largemouth bass. However, largemouth bass ingested a larger volume of water and generated a higher volumetric flow rate than bluegill sunfish. In addition, because largemouth bass swam faster during prey capture, they generated higher fluid speeds in the fish's frame of reference. Thus, while bluegill can exert higher drag forces on stationary prey items, largemouth bass more quickly close the distance between themselves and prey. The ingested volume and volumetric flow rate significantly increased as gape increased for both species, while time to peak gape had little effect on the volume. However, peak gape distance did not affect the maximum fluid speed entering the mouth for either species. We suggest that species that generate high fluid speeds in the earthbound frame of reference will commonly exhibit small mouths and a high capacity to deliver force to buccal expansion,while species that ingest a large volume of water and generate high volumetric flow rates will have larger buccal cavities and cranial expansion linkage systems that favor displacement over force delivery.

Spatial and temporal patterns of water flow generated by suction-feeding bluegill sunfish Lepomis macrochirus resolved by Particle Image Velocimetry

The suction-feeding fish generates a flow field external to its head in order to draw prey into the mouth. To date there are very few empirical measurements that characterize the fluid mechanics of suction feeding, particularly the temporal and spatial patterns of water velocity in front of the fish. To characterize the flow in front of suction-feeding bluegill sunfish Lepomis macrochirus, measurements with high spatial (<1 mm) and temporal (500 Hz) resolution were taken using Particle Image Velocimetry (PIV). In an analysis separate from the PIV, high-speed video sequences were used for a novel method of visually tracking every seed particle for the duration of each feeding in order to determine directly the total parcel of water that the fish ingests. PIV measurements and particle tracking show that water is drawn from all around the mouth. Fluid velocity decreases rapidly with distance from the mouth and is only significant (>5% of speed at the mouth) within roughly 1 mouth diameter of the fish. Suction feeders gain little in terms of extending this flow field by even substantial increases in the fluid speed at the mouth opening. Instead, the chief advantage of increased flow speed at the mouth may be the increased magnitude of generated forces within the space very close to the mouth. After scaling of the velocity field based on size of the mouth opening and the measured fluid speed at a fixed position, the measured velocity profiles for all feedings are very similar to one another, so that a functional relationship for the magnitude of fluid speed as a function of distance from the predator mouth is presented and shown to be accurate over the range of kinematic variables tested. This relationship describes the velocity field both along the centerline of the fish and along transects lying at an angle to the centerline within both the mid-sagittal and frontal planes. Comparison of the time-resolved fluid velocity measurements to gape kinematics demonstrate that peak fluid speed occurs simultaneously with 95% of peak gape, showing that the bluegill maximizes nearly simultaneously both the generated forces and size of the region over which these forces act. The magnitude of peak fluid speed during each strike decreases as a function of increasing time to peak gape (r2 = 0.87), demonstrating a strong relationship between the rate of buccal cavity expansion and maximum generated flow speed.

PIV measurements of flow in a centrifugal blood pump: time-varying flow

Measurements of the time-varying flow in a centrifugal blood pump operating as a left ventricular assist device (LVAD) are presented. This includes changes in both the pump flow rate as a function of the left ventricle contraction and the interaction of the rotating impeller and fixed exit volute. When operating with a pulsing ventricle, the flow rate through the LVAD varies from 0-11 L/min during each cycle of the heartbeat. Phase-averaged measurements of mean velocity and some turbulence statistics within several regions of the pump, including the inlet, blade passage, exit volute, and diffuser, are reported at 20 phases of the cardiac cycle. The transient flow fields are compared to the constant flow rate condition that was reported previously in order to investigate the transient effects within the pump. It is shown that the quasi-steady assumption is a fair treatment of the time varying flow field in all regions of this representative pump, which greatly simplifies the comprehension and modeling of this flow field. The measurements are further interpreted to identify the effects that the transient nature of the flow field will have on blood damage. Although regions of recirculation and stagnant flow exist at some phases of the cardiac cycle, there is no location where flow is stagnant during the entire heartbeat.

PIV measurements of flow in a centrifugal blood pump: steady flow

Magnetically suspended left ventricular assist devices have only one moving part, the impeller. The impeller has absolutely no contact with any of the fixed parts, thus greatly reducing the regions of stagnant or high shear stress that surround a mechanical or fluid bearing. Measurements of the mean flow patterns as well as viscous and turbulent stresses were made in a shaft-driven prototype of a magnetically suspended centrifugal blood pump at several constant flow rates (3-9 L/min) using particle image velocimetry (PIV). The chosen range of flow rates is representative of the range over which the pump may operate while implanted. Measurements on a three-dimensional measurement grid within several regions of the pump, including the inlet, blade passage, exit volute, and diffuser are reported. The measurements are used to identify regions of potential blood damage due to high shear stress and/or stagnation of the blood, both of which have been associated with blood damage within artificial heart valves and diaphragm-type pumps. Levels of turbulence intensity and Reynolds stresses that are comparable to those in artificial heart valves are reported. At the design flow rate (6 L/min), the flow is generally well behaved (no recirculation or stagnant flow) and stress levels are below levels that would be expected to contribute to hemolysis or thrombosis. The flow at both high (9 L/min) and low (3 L/min) flow rates introduces anomalies into the flow, such as recirculation, stagnation, and high stress regions. Levels of viscous and Reynolds shear stresses everywhere within the pump are below reported threshold values for damage to red cells over the entire range of flow rates investigated; however, at both high and low flow rate conditions, the flow field may promote activation of the clotting cascade due to regions of elevated shear stress adjacent to separated or stagnant flow.

Sucking while swimming: evaluating the effects of ram speed on suction generation in bluegill sunfishLepomis macrochirususing digital particle image velocimetry

It is well established that suction feeding fish use a variable amount of swimming (ram) during prey capture. However, the fluid mechanical effects of ram on suction feeding are not well established. In this study we quantified the effects of ram on the maximum fluid speed of the water entering the mouth during feeding as well as the spatial patterns of flow entering the mouth of suction-feeding bluegill sunfish Lepomis macrochirus. Using Digital Particle Image Velocimetry (DPIV) and high-speed video, we observed the flow in front of the mouth of three fish using a vertical laser sheet positioned on the mid-sagittal plane of the fish. From this we quantified the maximum fluid speed (measured at a distance in front of the mouth equal to one half of the maximum mouth diameter), the degree of focusing of water flow entering the mouth, and the shape of the ingested volume of water. Ram speed in 41 feeding sequences, measured at the time of maximum gape, ranged between 0 and 25 cm s–1, and the ratio of ram speed to fluid speed ranged from 0.1% to 19.1%. In a regression ram speed did not significantly affect peak fluid speed, but with an increase in ram speed the degree of focusing of water entering the mouth increased significantly, and the shape of the ingested volume of water became more elongate and narrow. The implications of these findings are that (1) suction feeders that employ ram of between 0% and 20% of fluid speed sacrifice little in terms of the fluid speeds they generate and(2) ram speed enhances the total body closing speed of the predator.

Design and transient computational fluid dynamics study of a continuous axial flow ventricular assist device

A ventricular assist device (VAD), which is a miniaturized axial flow pump from the point of view of mechanism, has been designed and studied in this report. It consists of an inducer, an impeller, and a diffuser. The main design objective of this VAD is to produce an axial pump with a streamlined, idealized, and nonobstructing blood flow path. The magnetic bearings are adapted so that the impeller is completely magnetically levitated. The VAD operates under transient conditions because of the spinning movement of the impeller and the pulsatile inlet flow rate. The design method, procedure, and iterations are presented. The VAD's performance under transient conditions is investigated by means of computational fluid dynamics (CFD). Two reference frames, rotational and stationary, are implemented in the CFD simulations. The inlet and outlet surfaces of the impeller, which are connected to the inducer and diffuser respectively, are allowed to rotate and slide during the calculation to simulate the realistic spinning motion of the impeller. The flow head curves are determined, and the variation of pressure distribution during a cardiac cycle (including systole and diastole) is given. The axial oscillation of impeller is also estimated for the magnetic bearing design. The transient CFD simulation, which requires more computer resources and calculation efforts than the steady simulation, provides a range rather than only a point for the VAD's performance. Because of pulsatile flow phenomena and virtual spinning movement of the impeller, the transient simulation, which is realistically correlated with the in vivo implant scenarios of a VAD, is essential to ensure an effective and reliable VAD design.

Studies of Turbulence Models in a Computational Fluid Dynamics Model of a Blood Pump

Computational fluid dynamics (CFD) is used widely in design of rotary blood pumps. The choice of turbulence model is not obvious and plays an important role on the accuracy of CFD predictions. TASCflow (ANSYS Inc., Canonsburg, PA, U.S.A.) has been used to perform CFD simulations of blood flow in a centrifugal left ventricular assist device; a k-epsilon model with near-wall functions was used in the initial numerical calculation. To improve the simulation, local grids with special distribution to ensure the k-omega model were used. Iterations have been performed to optimize the grid distribution and turbulence modeling and to predict flow performance more accurately comparing to experimental data. A comparison of k-omega model and experimental measurements of the flow field obtained by particle image velocimetry shows better agreement than k-epsilon model does, especially in the near-wall regions.

Computational fluid dynamics modeling of impeller designs for the HeartQuest left ventricular assist device

To finalize the design of the next generation of the HeartQuest left ventricular assist device, a suitable impeller had to be designed and tested. The new prototype was based on calculations and test results of previous designs, but required several changes to decrease the size. For most pump designs, this is a simple matter of altering impeller geometry and rotational speed to achieve the desired pressure rise and flow rate. However, this particular pump was limited by housing geometry and the magnetic bearings that support the impeller. Without much freedom in the overall impeller size, the only parameters open to the designers were the blade profiles and the rotating speed. Rather than build several candidates and test them in a rig at enormous cost, computational models of several designs were tested and analyzed. This not only saved money, but also sped up the development time for the project. The computer models were developed in TASCflow, a computational fluid dynamics software package from AEA Technologies. This paper analyzes the data from several of the selected models, paying close attention to pumping performance and general trends from specific design changes.

A prototype HeartQuest ventricular assist device for particle image velocimetry measurements

The objective of this study is to fully characterize the flow within the HeartQuest ventricular assist device (VAD), a magnetically levitated centrifugal VAD, using particle image velocimetry (PIV) to identify regions of potential high shear or stagnation and validate and refine computational models of the flow. An acrylic model of the pump was designed and constructed to allow optical access into all interior regions of the pump. The geometry of the exterior housing and the use of a novel working fluid make quantitative measurements of velocity within the exit volute, blade passage, cut-water, blade tip clearance, and pump inlet possible. Highly accurate velocity measurements using particle PIV have been made in one region (the inlet elbow), and measurements in the other critical regions of the pump will be made. These measurements are used for investigation of regions with potential for hemolysis resulting from high shear stress or with potential for thrombosis caused by recirculation or stagnation. Quantitative velocity data are also needed for comparison with computational fluid dynamics (CFD) models of the VAD. In this study, experiments have again proven to be an essential complement to CFD for thorough investigations of the flow inside the pump.

The application of quantitative oil streaking to the HeartQuest left ventricular assist device

Methods of flow visualization using oil streaking are established techniques for investigating surface shear and near wall flow patterns. Recent studies have used an array of oil dots on a surface which form streaks when exposed to shear forces. This method is generally qualitative, but it is possible to make quantitative measurements of the shear if the oil streaks have been calibrated. This paper presents the application of a quantitative oil streak method to the HeartQuest left ventricular assist device (LVAD). An array of dots was applied to the top housing of the pump, yielding quantitative values for the shear and qualitative patterns of the near wall flow in that region. The results were used to locate regions likely to promote thrombosis, such as stagnation points or recirculation regions. Regions of high shear, where hemolysis might occur, also can be identified with this method. In addition to being an important design technique, quantitative oil streaking assisted in the verification of computational fluid dynamics results within the HeartQuest LVAD.

Particle image velocimetry measurements of blood velocity in a continuous flow ventricular assist device

The third prototype of a continuous flow ventricular assist device (CFVAD3) is being developed and tested for implantation in humans. The blood in the pump flows through a fully shrouded four-bladed impeller (supported by magnetic bearings) and through small clearance regions on either side of the impeller. Measurements of velocities using particle image velocimetry of a fluid with the same viscosity as blood have been made in one of these clearance regions. Particle image velocimetry is a technique that measures the instantaneous velocity field within an illuminated plane of the fluid field by scattering light from particles added to the fluid. These measurements have been used to improve understanding of the fluid dynamics within these critical regions, which are possible locations of both high shear and stagnation, both of which are to be avoided in a blood pump. Computational models of the pump exist and these models are currently being used to aid in the design of future prototypes. Among other things, these models are used to predict the potential for hemolysis and thrombosis. Measurements of steady flow at two operating speeds and flow rates are presented. The measurements are compared with the computed solutions to validate and refine, where necessary, the existing computational models.

Successful newborn sickle cell trait counseling program using health department nurses

Every year more than 50,000 infants with sickle cell trait are identified through newborn screening programs. An infant with trait provides a "genetic window" into a family that may be at risk for having a child with sickle cell disease. No national standards have been established for newborn trait follow-up and family counseling. In the United States, counseling to families having an infant with trait is usually very limited in scope and, in many cases, nonexistent. In West Tennessee, the Mid-South Sickle Cell Center (MSSCC) and the Department of Health have jointly developed an effective, practical, and affordable trait counseling program. A five-step counseling protocol, an accompanying counseling manual, a documentation checklist, and a fact sheet for parents were developed to support trait counselling. By training and using health department nurses, this program has provided counseling to a large percentage of families in which an infant tests positive for sickle cell trait.

Recurrent Streptococcus pneumoniae sepsis in children with sickle cell disease

Streptococcus pneumoniae sepsis is the most common invasive infection among patients with sickle cell disease. The risk of a recurrent episode of sepsis and subsequent death in those patients who have had a previous septic event is much higher. Patients with sickle disease who have had pneumococcal sepsis should continue penicillin prophylaxis indefinitely and should not be candidates for out-patient management of febrile episodes.

Outpatient therapy with ceftriaxone and oral cefixime for selected febrile children with sickle cell disease

PURPOSE

Children with sickle cell disease are at increased risk for bacterial sepsis and, when febrile, are usually hospitalized for intravenous antibiotic therapy pending results of blood cultures. In this study, we prospectively identified a group of febrile patients with sickle cell disease who were at low risk for sepsis and treated them with outpatient therapy.

PATIENTS AND METHODS

Children identified as low risk for sepsis were treated with an initial dose of intravenous ceftriaxone, followed by outpatient therapy with oral cefixime, and were monitored for 14 days after the initial visit. Compliance was assessed by phone calls to parents and by analysis of urine samples.

RESULTS

In 107 eligible febrile episodes (80 patients) over a 21-month period, no patient developed sepsis. One child developed bacteremia 3 days after completing the course of cefixime, and one had splenic sequestration on the fourth study day. Both patients did well. Side effects of cefixime were modest, and overall compliance was excellent (approximately 95%), although urine samples were returned by only 56% of parents.

CONCLUSION

We conclude that outpatient therapy is safe and effective in febrile patients with sickle cell disease who meet the criteria for a low risk of sepsis.

Penicillin- and cephalosporin-resistant strains of Streptococcus pneumoniae causing sepsis and meningitis in children with sickle cell disease

OBJECTIVE

We investigated the possibility that antimicrobial-resistant pneumococci were causing invasive disease in children with sickle-cell disease (SCD).

STUDY DESIGN

Records of all children with SCD observed at the Mid-South Sickle Cell Center (MSSCC) at LeBonheur Children's Medical Center were reviewed from January 1990 to June 1994. Children with SCD and pneumococcal sepsis were identified. The Streptococcus pneumoniae isolates from these children were examined for serotype and antimicrobial susceptibilities. Two additional children not observed in the MSSCC had pneumococcal sepsis caused by penicillin-resistant isolates and were also included.

RESULTS

Antimicrobial susceptibility testing of the six penicillin-resistant isolates revealed that four were resistant to trimethoprim-sulfamethoxazole, two to erythromycin, and one to clindamycin. The two isolates that were resistant to ceftriaxone also were multiply resistant. From the MSSCC, 26 children had pneumococcal sepsis during the 4 1/2-year period studied. Five of these children (19%) died. Four (15%), including one who died, were infected with penicillin-resistant strains.

CONCLUSION

Pneumococcal sepsis, meningitis, and infections of other foci in children with SCD may be caused by S. pneumoniae that is resistant to one or more antimicrobial agents, including penicillin. The addition of vancomycin to the antibiotics currently used for initial management should be considered in areas where the antibiotic resistance of S. pneumoniae is prevalent.

A randomized study of outpatient treatment with ceftriaxone for selected febrile children with sickle cell disease

BACKGROUND

Because of their susceptibility to pneumococcal sepsis, children with sickle cell disease and fever are usually hospitalized for antibiotic therapy. Outpatient treatment may be a safe and less expensive alternative for selected patients.

METHODS

After evaluation in the emergency room, children ranging from 6 months to 12 years of age who had sickle hemoglobinopathies and temperatures exceeding 38.5 degrees C were randomly assigned to treatment as either inpatients or outpatients. We excluded from randomization children at higher risk of sepsis (as defined by specific criteria, including temperature above 40 degrees C, white-cell count below 5000 per cubic millimeter or above 30,000 per cubic millimeter, and the presence of pulmonary infiltrates) or with complications of sickle cell disease (such as a hemoglobin level below 5 g per deciliter, dehydration, or severe pain); these children were treated as inpatients. All patients received an initial intravenous dose of ceftriaxone (50 mg per kilogram of body weight). Those treated as outpatients returned 24 hours later for a second dose of ceftriaxone, whereas the in patients were treated as directed by their physicians.

RESULTS

None of the 86 patients (with a total of 98 febrile episodes) in the randomized groups had sepsis, as compared with 6 of the 70 patients (7 of 86 episodes) excluded because of higher risk (P = 0.004). Among the 44 children (50 episodes) assigned to outpatient treatment, there were 11 hospitalizations (22 percent of episodes) within two weeks after treatment (95 percent confidence interval, 12 to 36 percent), whereas after inpatient care only a single patient (2 percent of episodes) was rehospitalized. When the randomized groups were compared, outpatient treatment saved a mean of $1,195 per febrile episode. The median hospital stay was 3 days (range, 1 to 6) for the children randomly assigned to inpatient care and 4 days (range, 1 to 18) for the higher-risk children treated as inpatients (P < 0.001).

CONCLUSIONS

With the use of conservative eligibility criteria, at least half the febrile episodes in children with sickle cell disease can be treated safely on an outpatient basis, with substantial reductions in cost.

High risk of recurrent stroke after discontinuance of five to twelve years of transfusion therapy in patients with sickle cell disease

Although long-term transfusion therapy is at least 90% effective in preventing recurrent strokes after an initial cerebrovascular accident in patients with sickle cell disease, it is unknown how long transfusion therapy should be continued. To address this question, we prospectively discontinued transfusions in 10 patients with sickle cell disease whose median duration of transfusion therapy after an initial stroke was 9 1/2 years (range 5 to 12 years). Before the transfusions were discontinued, patients were examined by cerebral angiography, magnetic resonance imaging of the head, neuropsychologic testing, electroencephalography, and a complete neurologic examination. Within 12 months after transfusion therapy was stopped, 5 of 10 patients had had an ischemic event. Three events caused relatively mild deficits in the same areas as those originally affected. Two were associated with massive intracranial hemorrhage, including one on the contralateral side of original involvement. An additional patient died suddenly of unknown causes. Of the four remaining patients, three declined to resume transfusion and are relatively well at greater than or equal to 18 months after therapy was stopped. The studies performed before transfusions were stopped were not predictive of recurrent stroke. The risk of recurrent cerebrovascular accident in this group was significantly greater than the estimated risk of 10% in patients who are receiving long-term transfusion therapy (p = 0.002). This adverse outcome suggests that patients with sickle cell disease who have had a stroke must receive long-term transfusion indefinitely or a suitable therapeutic alternative must be devised.