Fluid delivery system

Abstract

A fluid delivery system 100 comprising a fluid source 130, a pump 120 arranged to deliver fluid from the fluid source 130 to a patient 140, a sensor 150 configured to measure a physiological parameter

Description

Title — Fluid delivery system The present invention relates to a fluid delivery system, and more particularly to an intravenous fluid delivery system for delivering intravenous fluid therapy to a patient. Patients in emergency care, such as in intensive care units, that are suffering from critical illnesses generally have increased oxygen requirements, but also suffer from reduced oxygen delivery, as a result of a variety of issues such as blood loss, dehydration, or dysfunction of the heart and blood vessels. This physiological state imbalance of supply and demand is generally referred to as “shock” within this technical field. Shock has a mortality rate of up to 40%, causes in excess of 10 million deaths each year globally, and even survivors can be left with life-changing organ injury (chronic renal failure), cognitive impairment, or psychological sequalae. The consensus in the field is that a key treatment for shock is to deliver intravenous fluids to the patient. However, it is also well recognised within the field that intravenous fluid delivery is linked to increased hospital length of stay, complications, and mortality. The benefits of fluid delivery must therefore be balanced against the potential to cause harm. This relationship is known in the field as the Starling curve, which is demonstrated in Figure 1, which indicates a patient's cardiac output response to different volumes of intravenous fluid delivery. In a first region 10, the patient exhibits a high fluid responsiveness, and the continuous delivery of fluid causes a large increase in cardiac output. In a second region 20, the patient exhibits a partial fluid responsiveness, and the continuous delivery of fluid initially causes an increase in cardiac output (albeit a smaller increase than in the first region 10), but after a certain volume of fluid has been delivered, the delivery of further fluid no longer causes an increase in cardiac output. In a third region 30, the patient exhibits a low fluid responsiveness, and the continuous delivery of fluid causes little or no increase in cardiac output at all. Historically, there has been no effective way of determining where on this illustrated curve a patient is at any given time, and therefore no effective way of determining the patient's likely responsiveness to fluid delivery. Thus, a clinician could not effectively determine whether fluid should be delivered at all, never mind how much fluid should be delivered. Hence, a standard dose of 500mls of fluid would typically be delivered at first, with the success or failure of that delivery then determining whether a further 500mis should be delivered. More recently, new methods have been developed that enable a clinician to estimate fluid responsiveness in advance of delivery. However, these methods have their own limitations in that they are time consuming, require movement of the patient, eg the passive leg raise test, or require ventilator settings that are not viable for all patients. Perhaps most importantly, these methods have been developed to predict whether a full 500ml dose will cause a an increase in cardiac output of more than 15%. Hence, fluids may be denied where a smaller volume could still be beneficial to the patient, or the full 500ml may be delivered where a smaller dosage would provide the same benefits whilst minimising the potential to cause harm. There has now been devised an improved fluid delivery system and an improved method of delivering fluid to a patient, which overcome or substantially mitigate the abovementioned and / or other disadvantages associated with the prior art. According to a first aspect of the invention, there is provided a fluid delivery system comprising: a fluid source; a pump arranged to deliver fluid from the fluid source to a patient; a sensor configured to measure a physiological parameter associated with the patient during fluid delivery; and a controller configured to control operation of the pump by: receiving from the sensor a plurality of measurements of the physiological parameter, generating curve data that is dependent on the plurality of measurements of the physiological parameter, comparing the curve data with predetermined curve data, generating a control signal based on said comparison, and outputting the control signal to the pump to control the delivery of fluid from the fluid source to the patient. According to a second aspect of the invention there may be provided a method of controlling fluid delivery to a patient, the method comprising the steps of: (i) receiving from a sensor a plurality of measurements of the physiological parameter; (ii) generating curve data that is dependent on the plurality of measurements of the physiological parameter; (iii) comparing the curve data with predetermined curve data; (iv) generating a control signal based on said comparison; and (v) outputting the control signal to the pump to control the delivery of fluid from a fluid source to the patient. The fluid delivery system and method of controlling fluid delivery to a patient according to these aspects of the invention may be advantageous in that the system is able to monitor the effects of fluid delivery on the patient's physiological parameter in real-time, and by comparing these effects with predetermined data, predict the exact time at which the delivery of further fluids would no longer prove beneficial to the patient. This ensures that excess fluid is not delivered to the patient, which could cause them more harm than good. This in turn leads to a reduction in the time and cost of intravenous fluid therapy treatment itself, and since one side effect of delivering too much fluid is an increased length of stay in hospital, also leads to a reduction in the time and cost of overall treatment. The fluid delivery system may be an intravenous fluid delivery system. The fluid delivery system may be configured to deliver intravenous fluid therapy to a patient, eg via an intravenous fluid drip. The fluid delivery system may comprise a patient interface for delivering fluid to the patient. The patient interface may deliver fluid directly into the bloodstream of the patient. The patient interface may be a catheter, for example. The fluid source may be an intravenous fluid source. The fluid source may comprise blood or blood components, and / or starch-based fluids such as colloids, and / or electrolyte fluids, eg saline-based fluids. The physiological parameter may be cardiac output. The physiological parameter may be a parameter indicative of and / or dependent on cardiac output. The physiological parameter may be indicative of and / or dependent on oxygen delivery and / or utilisation of oxygen by the patient's organs. The physiological parameter may be any of, or any combination of: heart rate, oxygen consumption, oxygen content of arterial blood, oxygen content of venous blood, stroke volume variation, plasma volume, hydration, volume of blood loss, blood pressure, and systemic vascular resistance. Alternatively, the physiological parameter may be a parameter indicative and / or dependent on any of, or any combination of: oxygen consumption, oxygen content of arterial blood, oxygen content of venous blood, stroke volume variation, plasma volume, hydration, volume of blood loss, blood pressure, and systemic vascular resistance. The controller may be receiving the plurality of measurements of the physiological parameter from the sensor via an intermediary monitoring device. The monitoring device may be arranged to monitor and display the plurality of measurements. The controller may receive the plurality of measurements via an electrical connection or wireless transmission. Alternatively, the controller may comprise an optical sensor, such as a camera, arranged to image the display of the monitoring device. This may be advantageous in that it enables the controller to be retrospectively fitted to existing fluid delivery systems. The plurality of measurements of the physiological parameter may be taken sequentially during fluid delivery, eg during a single fluid delivery event or instance. For example, a first measurement of the physiological parameter may be taken at the start of the fluid delivery or instance, and an additional measurement of the physiological parameter may be taken every 1 second, every 2 seconds, every 5 seconds, or every 10 seconds thereafter. The controller may update the curve data after each new measurement of the physiological parameter is taken, ie to take account of the additional measurement. The controller may make a new comparison of the curve data with predetermined curve data after updating the curve data. The controller may generate and output a new control signal after making the new comparison. Generating curve data may comprise determining or plotting the plurality of measurements of the physiological parameter against a second parameter. The second parameter may be representative of the time at which each of the physiological parameter measurements were taken. The second parameter may be representative of the amount of fluid that had been delivered to the patient at the time at which each of the physiological parameter measurements were taken. The amount of fluid delivered may refer to the amount of fluid delivered during the current fluid delivery event or instance. The plot may define a trace, line, path or curve representative of the patient's cardiac output response to fluid delivery over time. The predetermined curve data may comprise one or more predetermined plot. The one or more predetermined plot may define a trace, line, path or curve representative of a typical cardiac output response to fluid delivery over time. The typical cardiac output response to fluid delivery over time may be defined based on personal and / or group historical data. The predetermined curve data may therefore be dependent on previously recorded measurements of a physiological parameter associated with the individual patient during previous fluid delivery events or instances. Additionally or alternatively, the predetermined curve data may therefore be dependent on previously recorded measurements of a physiological parameter associated with a plurality of patients during previous fluid delivery events or instances. The comparison of the curve data with predetermined curve data may comprise comparing the shape of the plot to the shape of the one or more predetermined plots. The comparison of the curve data with predetermined curve data may comprise comparing the shape of the trace, line, path or curve representative of the patient's cardiac output response to fluid delivery over time with the trace, line, path or curve representative of a typical cardiac output response to fluid delivery over time. A mathematical analysis / model, eg comprising a statistical and / or pattern recognition analysis, may be used to identify differences or similarities between the shape of the plots and / or the shape of the traces, lines, paths or curves during said comparison. Generating curve data may further comprise analysing the plurality of measurements of the physiological parameter and / or the plotted data to calculate curve data that is representative of the magnitude of change between any two of the plurality of measurements of the physiological parameter. The predetermined curve data may be representative of the magnitude of change between any two measurements of the physiological parameter in a typical cardiac output response to fluid delivery over time. Additionally or alternatively, generating curve data may further comprise analysing the plurality of measurements of the physiological parameter and / or the plotted data to calculate curve data that is representative of the rate of change, ie the gradient, between any two of the plurality of measurements of the physiological parameter. The predetermined curve data may be representative of the rate of change between any two measurements of the physiological parameter in a typical cardiac output response to fluid delivery over time. The controller may be further configured to determine a fluid responsiveness dependent on the comparison of the curve data with predetermined curve data. Fluid responsiveness may be an indication or measure of the likelihood of the physiological parameter associated with the patient reacting positively to the delivery of more fluid. Thus, the controller may be configured to determine whether the continued delivery of fluid from the fluid source to the patient will have a positive or negative effect on the physiological parameter. Where the controller determines a positive fluid responsiveness, the controller may generate a positive control signal. The positive control signal may have no effect on the delivery of fluid from the fluid source to the patient, ie the positive control signal may instruct the pump to continue delivering fluid. Where the controller determines a negative, or less positive, fluid responsiveness, the control signal may generate a negative control signal. The negative control signal may initiate a change in the delivery of fluid from the fluid source to the patient. For example, the negative control signal may be a stop signal which instructs the pump to stop the delivery of fluid from the fluid source to the patient. Alternatively, the negative control signal may instruct the pump to change the rate of delivery of fluid from the fluid source to the patient. The controller may generate a negative control signal in response to identifying a reduction in the rate of change between any two of the plurality of measurements of the physiological parameter, and / or in response to identifying a zero or negative rate of change between any two of the plurality of measurements of the physiological parameter. This may be advantageous in that as soon as the delivery of fluids is having a limited effect, or is no longer having a positive effect, on the physiological parameter being measured, the delivery of fluids is stopped. In this regard, the controller may define a control factor. The control factor may be a predetermined setting associated with the fluid delivery system. Alternatively, the control factor may be selected by an operator from a plurality of possible control factors of the fluid delivery system. The control factor may determine when a negative control signal is generated. The control factor may determine whether a negative control signal is generated in response to identifying a reduction in the rate of change between any two of the plurality of measurements of the physiological parameter, or in response to identifying a zero or negative rate of change between any two of the plurality of measurements of the physiological parameter. The control factor may define a control threshold. The control threshold may correspond to a rate of change between any two of the plurality of measurements of the physiological parameter. The controller may generate a negative control signal in response to the rate of change between any two of the plurality of measurements of the physiological parameter being below the control threshold. The sensor may be configured to measure a second physiological parameter. Alternatively, the system may comprise a second sensor configured to measure a second physiological parameter. The controller may be configured to control operation of the pump by receiving from the sensor the plurality of measurements of the second physiological parameter. The controller may be configured to generate second curve data that is dependent on the plurality of measurements of the second physiological parameter. The controller may be configured to compare the second curve data with second predetermined curve data. The controller may be configured to generate a second control signal based on said comparison. The controller may be further configured to determine a fluid responsiveness dependent on the comparison of the second curve data with the second predetermined curve data. An error in the plurality of measurements of the first physiological parameter may be determined based on the plurality of measurements of the second physiological parameter. For example, where only one of the first and second control signals is a negative control signal, eg where only one of the comparisons determines a negative fluid responsiveness, the controller may determine an error in the plurality of measurements of the first parameter. In response to determining the error, the controller may be configured to output a positive control signal to the pump to control the delivery of fluid from the fluid source to the patient. Where both of the first and second control signals are a negative control signal, eg where both of the comparisons determine a negative fluid responsiveness, the controller may be configured to output a negative control signal to the pump to control the delivery of fluid from the fluid source to the patient. The system may comprise any number of sensors configured to measure any number of physiological parameters, some or all of which may be used to determine an error in measurements of some or all of the other physiological parameters. This may be advantageous in that it ensures that a negative control signal, eg to stop fluid delivery from the fluid source, is generated because a negative fluid responsiveness has been validly and / or accurately determined, rather than because an erroneous measurement of a physiological parameter has been taken. Additionally, or alternatively, the controller may define an error factor. The error factor may be a predetermined setting associated with the fluid delivery system, or a setting that is selected by an operator from a plurality of possible error factors of the fluid delivery system. The error factor may determine when a negative control signal is generated. The error factor may define an error threshold. The error threshold may correspond to a rate of change between any two of the plurality of measurements of the physiological parameter. The controller may generate a negative control signal in response to the rate of change between any two of the plurality of measurements of the physiological parameter being below the error threshold. Additionally, or alternatively, the error factor may be dependent on personal and / or group historical data. For example, the error factor may be dependent on previously recorded measurements of a physiological parameter associated with the individual patient prior to the fluid delivery event. Additionally, or alternatively, the error factor may be dependent on previously recorded measurements of a physiological parameter associated with a plurality of patients prior to fluid delivery events or instances. For example, the error factor may be dependent on a variation in measurements of a physiological parameter prior to the fluid delivery event. This may be advantageous in that it accounts for natural variations in measurements of the patient's physiological parameters, which may occur as a result of sensor inaccuracies or signal noise, eg when the patient moves or coughs during measurement. This in turn ensures that a negative control signal, eg to stop fluid delivery from the fluid source, is generated because fluid responsiveness has been validly and / or accurately determined, rather than because an erroneous measurement of a physiological parameter has been taken. According to aspects of the invention there is provided one or more processors comprising machine readable instructions for performing or controlling any, any combination of, or all of, the method steps described. According to further aspects of the invention there is provided a data carrier or data storage medium comprising machine readable instructions for controlling one or more processor to perform said method step(s). Practicable embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, of which: Figure 1 is a graph illustrating the effects of intravenous fluid delivery dosage on a patient's cardiac output; Figure 2 is a schematic diagram of a fluid delivery system according to an embodiment of the invention; Figure 3 is a graph illustrating the effects of intravenous fluid delivery on a patient's cardiac output; Figure 4 is a graph illustrating a first example of a noise and / or error detection mechanism; Figure 5 is a graph illustrating a second example of a noise and / or error detection mechanism; and Figure 6 is a graph illustrating the potential effects of using a fluid delivery system according to the invention on the volume of intravenous fluid delivery. Figure 2 illustrates a fluid delivery system 110 comprising an infusion pump 120 for delivering fluid from a fluid bag 130 to a patient 140. The system 110 further comprises at least one sensor 150 arranged to measure physiological parameters associated with the patient 140, including cardiac output, and a monitoring device 160 for monitoring and displaying said physiological parameters. Finally, the system 110 comprises a controller 170 arranged to receive physiological data from the monitoring device 160, and generate a control signal 180 based on that physiological data. The control signal 180 is then sent to infusion pump 120 to control the infusion of fluid to the patient 140. Figure 3 illustrates a graph plotting volume of fluid delivered to the patient (or time) on the x-axis, and cardiac output of the patient on the y-axis. This graph is indicative of the method by which the controller 170 controls the delivery of fluid to a patient. The at least one sensor 150 takes a measurement of cardiac output every one second, ie at points 200, 210, 220, 230, 240 and 250. These measurements are sent to the monitoring device 160, and are in turn received by controller 170. Upon receiving each measurement, controller 170 plots the cardiac output measurement against the known volume of fluid delivered to the patient, to produce a graph as indicated in Figure 3. Controller 170 then calculates the gradient or the derivative of the plotted curve between the most recent measurement and one or more of the previous measurements, and compares this gradient or derivative with a gradient or derivative calculated from historic data, so as to effectively determine where the patient's instantaneous fluid responsiveness is on the curve illustrated in Figure 1. This in turn enables the controller 170 to determine whether further fluid should be delivered, or if fluid delivery should now stop. In particular, the gradient or derivative is compared with historic gradient data and / or historic curve data to match the current curve against a known curve, and thus determine whether the delivery of further fluid is likely to have a positive effect on cardiac output. Upon receiving the first cardiac output measurement 200, since no curve data is available, controller 170 generates a control signal 250 indicating that fluid should be delivered. This control signal 250 is sent to infusion pump 120, and fluid is delivered from fluid bag 130 to the patient 140. One second later, upon receiving the second cardiac output measurement 210, controller 170 determines the gradient of the curve between points 210 and 200. This curve gradient is compared to known curve gradient data, and it is predicted that further fluid delivery would have a positive effect on cardiac output. Thus, controller 170 generates a control signal 260 indicating that further fluid should be delivered. This control signal 260 is sent to infusion pump 120, and more fluid is delivered from fluid bag 130 to the patient 140. Another second later, upon receiving the third cardiac output measurement 220, controller 170 determines the gradient of the curve between points 220 and 210, or the derivative of the curve between points 220 and 200. This curve gradient or curve derivative is compared to known data, and it is predicted that further fluid delivery would have a positive effect on cardiac output. Thus, controller 170 generates a control signal 270 indicating that further fluid should be delivered. This control signal 270 is sent to infusion pump 120, and more fluid is delivered from fluid bag 130 to the patient 140. Another second later, upon receiving the fourth cardiac output measurement 230, controller 170 determines the gradient of the curve between points 230 and 220, or the derivative of the curve between points 230 and 210, or between points 230 and 200. This curve gradient or curve derivative is compared to known data, and it is predicted that further fluid delivery would have a positive effect on cardiac output. Thus, controller 170 generates a control signal 280 indicating that further fluid should be delivered. This control signal 280 is sent to infusion pump 120, and more fluid is delivered from fluid bag 130 to the patient 140. Another second later, upon receiving the fifth cardiac output measurement 240, controller 170 determines the gradient of the curve between points 240 and 230, or the derivative of the curve between points 240 and 220, or between points 240 and 210, or between points 240 and 200. This curve gradient or curve derivative is compared to known data, and it is predicted that further fluid delivery would have a limited or negative effect on cardiac output. Thus, controller 170 generates a control signal 290 indicating that further fluid should not be delivered. This control signal 290 is sent to infusion pump 120, and fluid delivery is stopped. In particular, from comparison of the gradient of the curve between points 240 and 230, or the derivative of the curve between points 240 and 220, 240 and 210, or 240 and 200, with known data, controller 170 determines that if further fluid were to be delivered, the gradient of the curve would be equal to or less than zero, as indicated in region 300 of Figure 3. Thus, where 500ml of fluid would have been delivered to get to point 310 on the curve using the standard method known in the art, by using the method according to the invention, less fluid is delivered to get to point 240 on the curve, which results in the same (if not higher) cardiac output benefit. That is, by monitoring cardiac output in real-time, and predicting fluid responsiveness with greater accuracy, unnecessary fluid delivery is minimised. It is anticipated that to ensure fluid delivery is not stopped erroneously, one or more noise and / or error detection mechanisms may be required. A first example of such a mechanism is illustrated in Figure 4. Figure 4 illustrates a graph plotting volume of fluid delivered to the patient (or time) on the x-axis, and cardiac output of the patient on the y-axis. According to the method described in relation to Figure 3, at point 410, if the controller 170 were to determine an instantaneous gradient or derivative of the curve, then the controller would determine that the gradient / derivative of the curve is equal to or less than zero, and therefore generate a control signal indicating that further fluid should not be delivered. This control signal would be sent to infusion pump 120, and fluid delivery would be stopped. However, it can be seen in Figure 4 that this negative gradient is only temporary, and that further fluid delivery does continue to have a positive effect on the cardiac output of the patient, suggesting that the temporary negative effect on cardiac output at point 410 is likely due to an erroneous measurement and / or noise associated with the measurements. To prevent this temporary negative gradient from initiating a stop signal, instead of determining an instantaneous gradient or derivative of the curve at point 410, the method determines the gradient or the derivative over a time period 420. In doing so, because the gradient / derivative remain positive over the time period, a control signal indicating that further fluid should not be delivered is not erroneously initiated. A second example of such a mechanism is illustrated in Figure 3. Figure 5 illustrates a graph plotting volume of fluid delivered to the patient (or time) on the x-axis, and cardiac output of the patient on the y-axis. Figure 5 illustrates a jagged curve (darker line) that represents the actual measurements of cardiac output taken over time, and a smooth curve (lighter line) that represents a line of best fit through those cardiac output measurements. Point 510 on the curve represents the time at which fluid is first delivered to the patient, and cardiac output measurements before point 510 represent the patient's baseline cardiac output. According to the method described in relation to Figure 3, at point 520, if the controller 170 were to determine an instantaneous gradient or derivative of the curve, then the controller would determine that the gradient / derivative of the curve is equal to or less than zero, and therefore generate a control signal indicating that further fluid should not be delivered. This control signal would be sent to infusion pump 120, and fluid delivery would be stopped. However, it can be seen in Figure 5 that this negative gradient is only temporary, and that further fluid delivery does continue to have a positive effect on the cardiac output of the patient, suggesting that the temporary negative effect on cardiac output at point 520 is likely due to an erroneous measurement and / or noise associated with the measurements. To prevent this temporary negative gradient from initiating a stop signal, the instantaneous gradient or derivative of the curve at point 520 is compared with a baseline deviation to determine whether the negative gradient or derivative is the result of an error or noise, or is the result of negative fluid responsiveness. In this example, the baseline deviation is determined based on the noise associated with the patient's baseline cardiac output measurements. This noise may be created due to measurement inaccuracy, which may be caused, for example, by a patient moving or coughing during measurement. At point 530, a maximum negative deviation from the line of best fit is taken, and recorded as the baseline deviation. During fluid delivery, where a negative gradient or derivative is determined, the instantaneous deviation of that measurement from the line of best fit is determined. If that instantaneous deviation is less than the baseline deviation, then it is determined that the negative gradient or derivative is the result of an error or noise, and fluid delivery is continued. If that instantaneous deviation is more than the baseline deviation, then it is determined that the negative gradient or derivative is the result of negative fluid responsiveness, and fluid delivery is stopped. Other indicators of baseline deviation will be anticipated by the skilled person, such as a mean or median deviation from the line of best fit over the patient's baseline cardiac output measurements, or a standard deviation of the deviation from the line of best fit over the patient's baseline cardiac output measurements. A further example of such a mechanism is described without reference to a figure. Thus far the effects of fluid delivery have been described in relation to cardiac output. However, there are a number of physiological parameters that can be measured to determine the effects of fluid delivery, and many of those physiological parameters are linked. Thus, where fluid delivery causes a change in cardiac output, a change in a second physiological parameter would also be expected, making it possible to detect erroneous measurements by monitoring more than one of those physiological parameters. Where a negative gradient or derivative is detected in cardiac output, the method may check for a corresponding change in the second parameter. If no such change is detected in the second parameter, then it is determined that the negative gradient or derivative is the result of an error or noise, and fluid delivery is continued. If the same change is detected in the second parameter, then it is determined that the negative gradient or derivative is the result of negative fluid responsiveness, and fluid delivery is stopped. The corresponding change may be a corresponding and / or equally negative gradient or derivative in the second parameter. Alternatively, where the second parameter has a negative correlation with cardiac output, the corresponding change may be a corresponding and / or equally positive gradient or derivative in the second parameter. It is anticipated that any of, or any combination of, the above-described noise and / or error detection mechanisms may be implemented, along with any other standard mechanisms known in the field. It is also anticipated that an operator of the fluid delivery system 110 may be able to select from a plurality of available noise and / or error detection mechanisms, and / or control the thresholds for how / when a measurement is determined as being erroneous or not erroneous. The advantages of the above-described method are further illustrated in Figure 6, which represents data taken from fluid delivered to a sample of patients, and compares the results of delivering 500ml of fluid to each of the patients as standard practice against the point at which the method according to the invention would have stopped fluid delivery. In particular, Figure 6 plots the amount of unnecessary fluid given per patient on the y-axis, against the potential benefits to the cardiac output of the patient that were denied by not giving more fluid to the patient on the x-axis, for both the method according to the invention (in red), and the standard method used in the art (in blue). In comparing the method according to the invention with standard practice across the sample of patients, the inventors found that, on average, the method according to the invention delivered just under 50ml of unnecessary fluid, whereas the standard method known in the art delivered just under 150ml of unnecessary fluid (as indicated by the comparative arrow illustrated in Figure 6). As a result, on average, around 100ml of fluid was saved per patient by using the method according to the invention rather than the standard method known in the art. Figure 6 also shows that in doing so, on average, only 2% of potential cardiac output was denied to the patient. Similarly, it can be seen throughout the graph of Figure 6 that for any given percentage of denied potential increase in cardiac output, the method according to the invention delivers less unnecessary fluid than the standard method known in the art.

Claims

1. A fluid delivery system comprising: a fluid source;5 a pump arranged to deliver fluid from the fluid source to a patient;a sensor configured to measure a physiological parameter associated with the patient during fluid delivery; and a controller configured to control operation of the pump by: receiving from the sensor a plurality of measurements of the physiological parameter, generating a curve that is dependent on the plurality of 10 measurements of the physiological parameter, determining a series of curve data measurements associated with the generated curve, comparing the series of curve data measurements with predetermined curve data measurements, generating a control signal based on said comparison, and outputting the control signal to the pump to control the delivery of fluid from the fluid source to the patient.

152. A fluid delivery system according to Claim 1, wherein the plurality of measurements of the physiological parameter are taken sequentially during fluid delivery, and the controller is configured to update the series of curve data measurements after each new measurement of the physiological parameter is 20 taken, and compare the updated series of curve data measurements with predetermined curve data measurements.

3. A fluid delivery system according to Claim 2, wherein the controller is configured to generate a new control signal and to output the new control signal to 25 the pump following comparison of the updated series of curve data measurements with the predetermined curve data measurements.

4. A fluid delivery system according to any preceding claim, wherein the controller is configured to predict a fluid responsiveness dependent on the 30 comparison of the series of curve data measurements with the predetermined curve data measurements, and the control signal is dependent on the predicted fluid responsiveness.13 05 255. A fluid delivery system according to any preceding claim, wherein generating the curve comprises determining or plotting the plurality of measurements of the physiological parameter against a second parameter, and the predetermined curve data measurements are a series of measurements associated with one or more 5 predetermined plot of the physiological parameter against the second parameter.

6. A fluid delivery system according to Claim 5, wherein the curve defines a trace, line, path or curve representative of the patient’s cardiac output response to fluid delivery over time, and the one or more predetermined plot defines a trace, 10 line, path or curve representative of a typical cardiac output response to fluid delivery over time.

7. A fluid delivery system according to Claim 5 or Claim 6, wherein the comparison of the series of curve data measurements with the predetermined 15 curve data measurements comprises comparing the shape of the curve to the shape of the one or more predetermined plots.

8. A fluid delivery system according to Claim 6 or Claim 7, wherein the comparison of the series of curve data measurements with the predetermined 20 curve data measurements comprises comparing the shape of the trace, line, path or curve representative of the patient’s cardiac output response to fluid delivery over time with the trace, line, path or curve representative of a typical cardiac output response to fluid delivery over time.25 9. A fluid delivery system according to any of Claims 5-8, wherein the secondparameter is representative of the time at which each of the physiological parameter measurements were taken, and / or the amount of fluid that had been delivered to the patient at the time at which each of the physiological parameter measurements were taken.3010. A fluid delivery system according to any preceding claim, wherein the determining of the series of curve data measurements further comprises analysing the plurality of measurements of the physiological parameter and / or the curve to13 05 25calculate the series of curve data measurements, each of the curve data measurements being representative of the magnitude of change between any two of the plurality of measurements of the physiological parameter.5 11. A fluid delivery system according to Claim 10, wherein the predeterminedcurve data measurements are representative of the magnitude of change between any two measurements of the physiological parameter in a typical cardiac output response to fluid delivery over time.10 12. A fluid delivery system according to any preceding claim, wherein thedetermining of the series of curve data measurements further comprises analysing the plurality of measurements of the physiological parameter and / or the curve to calculate the series of curve data measurements, each of the curve data measurements being representative of the rate of change between any two of the15 plurality of measurements of the physiological parameter.

13. A fluid delivery system according to Claim 12, wherein the predetermined curve data measurements are representative of the rate of change between any two measurements of the physiological parameter in a typical cardiac output20 response to fluid delivery over time.

14. A fluid delivery system according to Claim 12 or Claim 13, wherein the calculation of the series of curve data measurements comprises calculating a series of gradients or derivatives of the curve between a series of measurements25 of the physiological parameter.

15. A fluid delivery system according to any of Claims 6-14, wherein the typical cardiac output response to fluid delivery over time is defined based on personal and / or group historical data.3016. A fluid delivery system according to any preceding claim, wherein the physiological parameter is cardiac output, or a physiological parameter indicative of and / or dependent on cardiac output.

Images