Non-invasive blood pressure measurement

Abstract

A method for adaptively scheduling time intervals for non-invasive blood pressure measurements based on calculating, for example, using a risk model, a patient's risk of suffering from one or more predefined adverse clinical events within a predefined time window, and based on a risk assessment by the patient's clinician.

Description

[Technical Field]

[0001] The present invention relates to the field of non-invasive blood pressure monitoring. [Background technology]

[0002] US2018042493A1 relates to an apparatus and method for providing a control signal for a blood pressure measuring device. Furthermore, the present invention relates to a system for monitoring a patient.

[0003] WO2021062292A1 relates to a system and method for risk-based patient monitoring.

[0004] Arterial blood pressure (ABP) is an important physiological parameter relevant for medical diagnosis, prevention and therapeutic guidance.

[0005] The gold standard for blood pressure measurement is invasive measurement, which uses intra-arterial cannulation. This provides the most accurate readings and allows for continuous measurement. However, it requires trained medical staff and is therefore usually only utilized in the highest acuity settings, requiring very careful monitoring and immediate alarms.

[0006] Most blood pressure measurements are performed noninvasively. The most common noninvasive blood pressure (NIBP) technique in patient monitoring is standard cuff-based ABP measurement using automated oscillometry. This provides intermittent measurements. Typically, the intervals are set at fixed or standardized time intervals depending on the patient's condition. For example, a typical time interval is every 15 minutes in an ICU room. However, this leaves a significant risk that a change in the patient's condition could suddenly occur between one of these intervals.

[0007] For example, FIG. 1 shows an example of a patient's real-time systolic blood pressure (y-axis) as a function of time (line 22) compared to blood pressure measurements recorded at regular intervals by a measurement device (line 24). In this example, the interval time is 15 minutes. This example shows that a sudden drop in blood pressure during the interval between measurements may go unnoticed for a significant amount of time until the next measurement is taken. This is evident, for example, in the interval from about 16:15 to about 16:30, when the patient's real-time blood pressure dropped significantly, but because this occurred within the interval between measurements, approximately 15 minutes passed before this change in condition was addressed by a clinician.

[0008] In the current state of the art, to address the fact that the appropriate measurement interval varies depending on the patient's condition, the clinician observing the patient manually adjusts the measurement interval depending on the clinician's judgment of the patient's level of acuity and the risk of the condition rapidly deteriorating. The clinician uses standard protocols combined with clinical experience to set the best measurement interval. Summary of the Invention [Problem to be solved by the invention]

[0009] It would be beneficial to provide a means for more objectively determining the best measurement time interval for a patient, taking into account the patient's condition. [Means for solving the problem]

[0010] The invention is defined by the claims.

[0011] According to an example according to one aspect of the present invention, there is provided a computer-implemented method, the method comprising receiving biometric data of a patient, the method further comprising applying a patient risk model to calculate a risk parameter p indicative of a risk level of at least one predefined adverse clinical event, wherein at least one input to the model is the biometric data, the method further comprising obtaining a clinician-defined patient risk parameter A, the method being based on A and p and satisfying the relationship:

number

[0012] The method further includes controlling the non-invasive blood pressure measurement device to obtain blood pressure measurements at a frequency determined by the determined time interval T.

[0013] Thus, embodiments of the present invention provide a method for adaptively setting timing intervals between automatically activated non-invasive blood pressure measurements based on a risk model that takes into account the current value of at least one biological parameter of the patient and, advantageously, also takes into account a risk assessment by a clinician. Thus, the clinician's prescriptive medical opinion regarding the condition still influences the timing, but at the same time, the time intervals can be set in a partially objective manner based on a risk model (configured to calculate the risk of at least one particular serious adverse event, such as hypotension or hypertension).

[0014] The relationship defined for T keeps constant the product of the model-based probability p for a clinical event and the time between blood pressure measurements.

[0015] With respect to the clinician-defined patient risk parameter A, it represents the clinician's assessment of the risk level of said at least one predefined adverse clinical event. In other words, it may represent the clinician's (subjective) judgment regarding the patient's risk level. In other words, it may represent the clinician's own assessment of the patient's risk.

[0016] With regard to the variable parameter α, this allows the weight given to machine-derived risk assessments relative to human-derived risk assessments to be adjusted according to preference or according to the situation.

[0017] In some embodiments, the value of α is obtained or retrieved as a step in the method.

[0018] For example, the value of α may be obtained or retrieved from memory or a register, or any other data source.

[0019] The value of a may be defined by a user, or may be received from a user interface, i.e. the method may comprise receiving or determining the value of a, for example during a setup phase or at least as a preliminary step, and optionally storing the value of a in a memory or register for subsequent retrieval during the execution of the remaining steps of the method (i.e. during calculation of the timing interval T).

[0020] In some embodiments, the method may include calculating or determining the value of α. For example, the value may be determined using a predefined lookup table or mapping that associates different patient or clinician conditions or classes (e.g., patient acuity levels) with different values ​​of α. The patient or clinician conditions or classes may be obtained or retrieved from a data store or other data source. For example, a lower value of α may be preferable for more acuity patients because (for values ​​of p ranging between 0 and 1) this has the effect of shortening the time interval between blood pressure measurements if the risk parameter p increases during the time period in which the clinician is updating their risk assessment A. Thus, in these situations, a higher value of α will weight more careful attention. As a further example, for a more experienced clinician who trusts their own assessment, the value of α may be set lower compared to a less experienced clinician. This has the effect of giving a higher relative weight to the clinician-defined parameter A compared to a lower value of α (assuming values ​​of p ranging between 0 and 1).

[0021] It will be appreciated that patients are typically at risk for more than one particular adverse clinical event, and thus, in some embodiments, the model is designed or trained to output a combined risk level for the patient of suffering from any of a series of different adverse events. Additionally or alternatively, the model can output a risk associated with only one adverse clinical event, although the method may include running multiple versions of the model, each configured to predict the probability of a different type of adverse event, and selecting one of the determined probability parameters based on predefined criteria or input from a clinician.

[0022] By way of example, the at least one adverse or clinical event may include hypotension and / or hypertension in the patient, shock in the patient, cardiac arrest, or any cardiac event.

[0023] In some embodiments, in the model, p is the probability that an event occurs in a predefined period of time.

[0024] In some embodiments, the biometric data includes real-time sensor data or data derived therefrom.

[0025] In some embodiments, the biometric data includes past biometric data of the patient retrieved from a data store, which may include, by way of example, heart rate or other vital signs, previous blood pressure measurements, pathology tests, or any other biological parameters.

[0026] In some embodiments, the risk parameter p is recalculated iteratively over the patient's monitoring session, and the biometric data is preferably updated with each recalculation.

[0027] In some embodiments, the risk model is a Bayesian model.

[0028] In some embodiments, the Bayesian model is a risk model personalized to the patient and / or the clinician treating the patient, hi some embodiments, the Bayesian model is pre-configured according to prior information including one or more of the patient's medical history, the severity / urgency of the patient's condition, the training level of the clinician treating the patient, and a measure of the clinician's speed of response to changes in condition.

[0029] In some embodiments, the method further comprises iteratively adjusting or updating the risk model based on a patient monitoring database having monitoring data of multiple patients. Thus, a self-learning model can be implemented that can learn over time, for example, through retrospective analysis of monitored patients. This can be based on monitoring data acquired at the same medical institution or medical department and / or more broadly, for example, based on monitoring data acquired across an entire population across multiple different medical institutions.

[0030] In some embodiments, the biometric data includes data from one or more of an electrocardiogram (ECG) sensing device, a photoplethysmogram (PPG) sensing device, a capnography measuring device, and a bioimpedance measuring device.

[0031] The method may be implemented in the form of software.

[0032] Therefore, another aspect of this method is a computer program product having code means configured to, when executed on a processor, cause the processor to perform a method according to any embodiment or example in the present disclosure.

[0033] The present invention may also be implemented in the form of hardware. Thus, another aspect of the method is a processing unit having an input / output and one or more processors, said one or more processors: receiving patient biometric data at said input / output; applying a patient risk model to calculate a risk parameter p indicative of a risk level of at least one predefined adverse clinical event, wherein at least one input to the model is biometric data; obtaining a clinician-defined patient risk parameter A; Based on A and p, the following relationship

number

[0034] Another aspect of the present invention is a system having a processing unit as generally described above or according to any embodiment described in this disclosure, and a non-invasive blood pressure measurement device operatively coupled to the processing unit.

[0035] In some embodiments, the system further comprises one or more biological parameter sensing devices operably coupled to the processing unit for acquiring biological measurement data.

[0036] In some embodiments, the one or more biological parameter sensing devices include a PPG sensor integrated as part of a non-invasive blood pressure measurement device.

[0037] These and other aspects of the invention will be apparent from and elucidated with reference to the embodiments described hereinafter. [Brief explanation of the drawings]

[0038] For a better understanding of the present invention, and to show more clearly how the same may be carried into effect, reference will now be made, by way of example only, to the accompanying drawings in which: [Figure 1] FIG. 1 shows a patient's continuous blood pressure as a function of time compared to intermittent blood pressure measurements. [Figure 2] FIG. 2 outlines the steps of an exemplary method according to one or more embodiments of the present invention. [Figure 3]FIG. 3 illustrates generally the components and process flow of an exemplary processing arrangement in accordance with one or more embodiments of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0039] The present invention will now be described with reference to the drawings.

[0040] While the detailed description and specific examples indicate exemplary embodiments of the devices, systems, and methods, it should be understood that they are intended for purposes of illustration only and are not intended to limit the scope of the invention. These and other features, aspects, and advantages of the devices, systems, and methods of the present invention will become better understood from the following description, the appended claims, and the accompanying drawings. It should be understood that the drawings are merely schematic and not drawn to scale. It should also be understood that the same reference numerals are used throughout the drawings to indicate the same or similar parts.

[0041] The present invention provides a method for adaptively scheduling non-invasive blood pressure measurements at determined time intervals based on using a risk model to calculate a patient's risk of suffering from one or more predefined serious adverse events, for example within a predefined time window, and also based on a risk assessment by the patient's clinician.

[0042] Blood pressure is typically measured by intermittent cuff non-invasive blood pressure (NIBP) measurements. Currently, medical staff must determine what blood pressure monitoring interval is appropriate based on an assessment of the patient's hemodynamic stability.

[0043] Because a patient's health status may change over time, the NIBP measurement interval ideally requires continuous adaptation to reflect the patient's changing risk level. However, in practical situations, this is not done, which can lead to several possible adverse effects, including: Missing important events (hypotensive and hypertensive periods); unnecessary NIBP measurements that reduce patient comfort and cause sleep disturbances; Incorrect representation of the patient's response to treatment (e.g., medication), resulting in subsequent suboptimal therapy decisions

[0044] Previously, an example of NIBP trend monitoring over stages of instability and stability was discussed with reference to FIG. 1. FIG. 1 shows that when NIBP monitoring is performed using intermittent measurements with a fixed measurement time interval (line 24), many episodes of large BP changes (line 22) are missed. The clinician's selection of the measurement time interval was inappropriate for this patient. Note that the graph shown in FIG. 1 depicts a blood pressure trend line using a zero-order hold representation (e.g., in which a discrete-time blood pressure signal is represented as a continuous-time signal by holding each sample value for one sample interval).

[0045] Blood pressure can be used as an indicator of a wide variety of adverse medical events in a patient's body.Two adverse events that are generally of concern to clinicians who monitor patients are hypertensive and hypotensive.Hypertensive events are typically understood to correspond to a systolic blood pressure of more than about 140 mm / Hg, and are extremely high blood pressure events.Hypotension events are typically understood to correspond to a systolic blood pressure of less than about 90 mm / Hg, and are extremely low blood pressure events.

[0046] Depending on the stage of treatment, the probability of death or long-term health problems for patients increases significantly with the duration of hypotensive periods. For example, the postoperative stage of treatment carries a significant risk of myocardial infarction and death, even when the duration of hypotensive periods is short, particularly due to low cardiac and cerebral perfusion (ischemia). See, for example, "McEvoy, M. et al., Perioperative Quality Initiative consensus statement on postoperative blood pressure, risk, and outcomes for elective surgery," British Journal of Anaesthesia, Volume 122, Issue 5, 2019, Pages 575-586. Referring to data presented on page 578, this paper explains that for every 10-minute episode of hypotension on postoperative day 0, the risk of myocardial infarction (MI) and death increases by 3%, and that episodes of hypotension on postoperative days 1 to 4 nearly double the risk of MI and death.

[0047] When a patient is in a hypertensive or hypotensive phase, the clinician must act quickly through the administration of drugs or other interventions to return the patient to a normal hemodynamic state.

[0048] A hypotensive event begins the moment a patient's actual blood pressure value falls below some predetermined critical threshold. The duration of a hypotensive blood pressure event can be defined as the interval time (in minutes) between the moment the actual blood pressure falls below the critical blood pressure threshold and the moment the blood pressure rises above the predetermined critical threshold.

[0049] The severity of hypotension is related to its depth and duration. Depth refers to the lowest blood pressure value reached during a hypotensive episode. By generalizing this concept, the depth and duration dose metric (DDD) for a hypotensive episode per patient is:

number

[0050] Regarding the critical hypotension threshold, there are various ways in which this threshold can be defined. Regardless of this definition, studies have shown that in various populations (e.g., noncardiac surgery patients and traumatic brain injury patients), patient outcomes are strongly dependent on the duration and depth of hypotension episodes (meaning how much blood pressure drops, e.g., to the minimum value). This is because damage accumulates over successive hypotension episodes. For example, by quantifying the hypotension dose (depth of hypotension integrated over exposure time) in 7,521 traumatic brain injury patients, one study (ibid.) showed that hypotension increases mortality by 20%. Furthermore, evidence demonstrates a significant association between the degree and duration of intraoperative and postoperative hypotension, on the one hand, and myocardial injury and acute kidney injury (AKI), on the other. See, for example, the article "EJ Mascha et al., "Intraoperative Mean Arterial Pressure Variability and 30-day Mortality in Patients Having Noncardiac Surgery," Anesthesiology, vol. 123, no. 1, pp. 79-91, 2015."

[0051] Further evidence exists demonstrating that postoperative hypotensive events correlate with poor outcomes. One study showed that SBP <90 mmHg was the most common cause (25%) of emergency team activation in orthopedic or general surgery wards. For this study, see S. Mohammed Iddrisu et al., "Frequency, nature, and timing of clinical deterioration in the early postoperative period," J. Clin. Nurs., vol. 27, no. 19-20, pp. 3544-3553, 2018.

[0052] A further study, which collected and reviewed data from 90 hospitals in the UK, Australia, and New Zealand, found that the most common antecedent of patient deterioration was a systolic blood pressure (SBP) of <90 mmHg. This study is detailed in the paper "J. Kause et al., "A comparison of antecedents to cardiac arrests, deaths, and emergency intensive care admissions in Australia and New Zealand, and the United Kingdom - The ACADEMIA study," Resuscitation, vol. 62, no. 3, pp. 275-282, 2004."

[0053] As a result of these problems, patient outcomes are impaired, for example, with longer hospital stays, higher morbidity, and higher mortality. Therefore, improved blood pressure control helps improve patient outcomes and reduce complications, and therefore also increases hospital efficiency.

[0054] Embodiments of the present invention provide a system and method for estimating and automatically setting optimal / appropriate blood pressure measurement intervals that are personalized to patient risk profile and clinician opinion.

[0055] 1 outlines in block diagram form the steps of an exemplary method according to one or more embodiments, which steps will be described in summary form before being further described in the form of exemplary embodiments.

[0056] The method includes receiving (12) biometric data of a patient.

[0057] The method further comprises applying (14) a patient risk model to calculate a risk parameter p indicative of a risk level of at least one predefined adverse clinical event, wherein at least one input to the model is biometric data.

[0058] The method further comprises the step of obtaining (16) a patient risk parameter A defined by a clinician.

[0059] The method is based on A and p and

number

[0060] The method further comprises the step of controlling (20) the non-invasive blood pressure measurement device to obtain blood pressure measurements at a frequency determined by the determined time interval T.

[0061] As mentioned above, the method may also be implemented in the form of hardware, for example in the form of a processing unit configured to perform the method according to any example or embodiment described in this document or according to any claim of the present application.

[0062] To further aid understanding, FIG. 3 shows a schematic diagram of an exemplary processing unit 32 configured to perform methods according to one or more embodiments of the present invention, and also shows a more detailed schematic of the process flow.

[0063] The processing unit 32 has an input / output 34 and one or more processors 36. The one or more processors perform the following steps: receiving patient biometric data at the input / output 34, for example from one or more biometric devices or sensors 44; applying a patient risk model to calculate a risk parameter p indicative of a risk level of at least one predefined adverse clinical event, wherein at least one input to the model is biometric data; obtaining a clinician-defined patient risk parameter A based on input from a user interface 46, for example, received at input / output 34; Based on A and p,

number

[0064] Another aspect of the present invention is a system 30 having a processing unit 32. FIG. 3 illustrates the processing unit 32 operating within such a system. The system 30 may further include a non-invasive blood pressure measurement device 42, e.g., a cuff-based oscillometric blood pressure measurement device, operably coupled to the processing unit. The system 30 may further include one or more biological parameter sensing devices 44 operably coupled to the processing unit 32 for acquiring biological measurement data. By way of example, the one or more biological parameter sensing devices may include a PPG sensor. The PPG sensor may be used to acquire not only blood oxygen levels but also pulse or heart rate. It is also possible to infer a patient's respiratory rate from the PPG sensor signal. In some examples, the PPG sensor is advantageously incorporated as part of the non-invasive blood pressure measurement device.

[0065] The system 30 further includes a user interface 46 for use in receiving user input indicating, for example, a clinician-defined patient risk parameter A, and / or for displaying a representation of the determined blood pressure measurement interval, the current value of the calculated risk parameter, one or more biological measurements of the patient, or any combination thereof.

[0066] For further clarity, a simple example workflow of the method in operation is outlined.

[0067] Patients in a medical facility are monitored with a series of one or more measurement devices 42, e.g., sensors, each configured to measure at least one biological parameter, e.g., a physiological parameter. Data output from the series of one or more measurement devices is received at a processing unit 32. This data forms a patient's biological measurement dataset. In some examples, the processing unit may be a processing unit of a patient monitoring system. In addition to or instead of real-time data from the measurement devices, the biological measurement dataset may include, for example, historical values ​​of one or more biological parameters or data retrieved from a data store containing variables related to the patient's medical history. The data store may include parameters that cannot be measured using sensors, such as pathological test results.

[0068] The processing unit 32 is further adapted to obtain an indication of a clinician-defined risk parameter A. In some embodiments, this parameter is obtained based on user input received from the user interface device 46. For example, a clinician responsible for monitoring the patient can enter the parameter. In another example, the patient's parameter A may be determined based on a lookup table. For example, the clinician may have predetermined a set of relationships between different patient status classifications and values ​​associated with parameter A, which may be encoded in the stored lookup table.

[0069] The processing unit 32 is further adapted to retrieve a patient risk model from the data store. The patient risk model is adapted to receive input values ​​of one or more variables corresponding to one or more respective biological measurements. These may include some or all of the values ​​forming the dataset of received biological measurements. The patient risk model is configured to generate as output a value of a parameter p indicative of a risk level of at least one predefined adverse clinical event, such as hypotension, hypertension, or a cardiac event. The value of p may be a probability value and may range from 0 to 1. The model may be a general model for any patient, as described in more detail below, or may be personalized to the patient. In some cases, the model may additionally or alternatively be personalized to the clinician monitoring the patient, as also described in more detail below.

[0070] Having retrieved the model, processing unit 32 applies the model by providing the relevant biological measurements as inputs to the model, thereby generating an output risk parameter value p.

[0071] The processing unit 32 then first α Calculate the value of

number

[0072] In some embodiments, T is simply expressed by the formula T=A / p α It is determined by evaluating

[0073] In some embodiments, T is calculated using the formula T=β(A / p α), where β may be, for example, a constant, or β may be, for example, a predefined transfer function. β may be, for example, a function of any or all of A, p, and α.

[0074] The purpose of the parameter β is to α The purpose of this is to provide a transfer or conversion of the value of A / p into a value appropriate for T in the correct units. The parameter β is the relationship between T and the desired ratio A / p α can be predefined to provide a specific mapping between

[0075] The parameter α provides an additional degree of freedom in this calculation to further refine the calculation of the time interval. In effect, the parameter defines the relative weighting to be applied to the computer-determined risk parameter p compared to the clinician-defined risk parameter A. As the value of α increases, the value of p as a function of p decreases. α It will be appreciated that the scaling of the value of α increases with increasing steepness, meaning that higher values ​​of α place more relative weight on the computationally defined risk parameter p compared to the clinician-defined risk parameter A.

[0076] In some examples, the parameter α may be freely determined by a user. In some examples, the parameter α may be determined using a predefined lookup table or mapping that associates different patient or clinician conditions or classes with different values ​​of α. For example, if the risk parameter p increases during the time period in which the clinician updates their risk assessment A (for values ​​of p varying between 0 and 1), it may be preferable to set a lower value of α for more urgent patients, as this has the effect of decreasing the time interval between blood pressure measurements. Thus, in these situations, higher values ​​of α are weighted toward greater attention. As a further example, for more experienced clinicians who trust their assessment, the value of α is set lower than for less experienced clinicians. This has the effect of giving a higher relative weight to the clinician-defined parameter A compared to lower values ​​of α (assuming values ​​of p vary between 0 and 1).

[0077] The processing unit 32 may store in memory a record of the newly calculated value of the calculated interval time T, which may be updated each time T is recalculated. A log of the calculated values ​​of T over time may be kept. T may be recalculated iteratively over a patient monitoring session. The risk parameter p may be recalculated iteratively over the patient monitoring session, with the biometric data being updated with each recalculation. The newly calculated value for p may be used by the processing unit to recalculate T.

[0078] The processing unit 32 is further configured to control the NIBP measurement device to obtain BP measurements at a frequency determined by the determined time interval T. The actual control of the NIBP measurement device may be performed by the processing unit 32 itself or by a subsystem to which the processing unit is operatively coupled. In the latter case, controlling the NIBP measurement device to obtain BP measurements at a frequency determined by the determined time interval may simply comprise communicating a control command to the control subsystem of the NIBP measurement device to cause said subsystem to obtain measurements at the calculated interval T. Alternatively, if the processing unit is adapted to regulate the control of the NIBP measurement device itself, the processing unit may simply update the value of the parameter T in a local register, whereupon the processing unit executes a control routine or program to operate the BP measurement at a time interval determined by the current value of T in that register.

[0079] As noted, the value of T is preferably recalculated repeatedly. With respect to the timing of the recalculation, this can be at fixed time intervals, at intervals set by the clinician, or at automatically determined or acted upon intervals. For example, one approach is to recalculate T every time a new BP measurement is taken. In other words, the value of T itself is updated at a frequency determined by the current value of T, so that at the end of each measurement interval, a new blood pressure measurement is taken and a new value for the interval time T is calculated. To calculate a new value of T, new values ​​for each of the one or more biological measurements are sampled, and these are used as inputs to re-run the patient risk model to generate new values ​​for the parameter p, after which a new value of T is calculated using the relationships already outlined above.

[0080] It should be noted that a patient risk model is inherently trained or programmed to model a patient's risk for a particular adverse clinical event, and thus in some embodiments there are multiple versions of the model, each configured to predict the probability of a different type of adverse event.

[0081] In some embodiments, these multiple versions of the risk model are run to generate a set of initial risk parameters p, each of which indicates the probability of occurrence of a respective adverse clinical event. i These initial risk parameters can then be further processed to derive an overall risk parameter p that indicates the risk of one or more adverse clinical events. In some embodiments, the initial risk parameter p that corresponds to the most likely adverse event or has the most severe outcome for the patient is i Only one of the initial risk parameters p is selected. This selection may be based on user input at a user interface. For example, the user interface may display a set of calculated initial risk parameters, each with a corresponding indication of an adverse clinical event, and the user may select one to be used as the overall patient risk parameter. In another example, some or all of the initial risk parameters p may be combined to form an overall patient risk parameter p. For example, the probabilities may be simply added together to determine the cumulative probability of each adverse clinical event occurring. Alternatively, an average may be calculated.

[0082] Further details regarding one or more of the features briefly outlined above will now be provided.

[0083] With respect to the biometric data acquired as part of the method, it is preferred that this data be received continuously or in real time from one or more measurement devices or sensors deployed to monitor the patient. In this way, the patient risk model comprises up-to-date information for the patient. An alternative approach is to use archived patient biometric data.

[0084] By way of non-limiting example, the biometric data may include one or more of the following: Electrocardiogram data Photoplethysmogram data (reflectance or transmission) Pulse oximetry / blood oxygen saturation data Capnography measurement data Bioimpedance measurement data Any other vital signs Pathological examination results

[0085] In some examples, a photoplethysmogram (PPG) sensor is provided as an auxiliary sensor separate from the blood pressure measurement device, for example, a finger PPG sensor. In some examples, a PPG sensor may be provided that is integrated into the blood pressure measurement device, for example, integrated into a blood pressure cuff and positioned such that an optical region of the PPG sensor is in optical communication with the surface of the patient's skin when the cuff is attached to a body part to measure blood pressure.

[0086] These data may be received in real time from one or more biological parameter sensors. Additionally or alternatively, the biological measurement data may include historical biological measurement data for the patient retrieved from a data store.

[0087] With respect to the non-invasive blood pressure (NIBP) measuring device controlled as part of the method, the device preferably includes a cuff-based measuring device. In use, the cuff is attached to the patient's arm or wrist. A typical example of such a device operates by actively pressurizing the cuff, for example, using a pump or pressure vessel, and measuring the cuff pressure using a pressure sensor. Other methods for operating a cuff-based NIBP measuring device also exist and are known to those skilled in the art. Any other type of NIBP measuring device can also be used as an alternative to a cuff-based system. NIBP measuring devices are very well understood in the art, and those skilled in the art will understand how to implement this feature.

[0088] The guiding insight behind the approach proposed by the inventors herein is the concept that for a constant value of the clinician-defined parameter A, the product of the probability p of a clinical event (potentially adjusted by a power law) and the NIBP interval time T remains constant. As a result, during any interval in which A remains unchanged (i.e., the clinician does not provide an update to their assessment of the patient's condition or risk), an increase in the patient's risk level, as calculated by the model, will result in a shorter NIBP interval.

[0089] As mentioned above, in accordance with this principle, in some embodiments, the value of the measurement time interval T is:

number

[0090] T is the recommended measurement interval time. A is a clinician-defined parameter that represents the clinician's assessment of the patient's risk level of a particular (predefined) adverse or clinical event (e.g., hypotension, hypertension) for the next monitoring period. The clinician is expected to make this choice based on their own judgment, using their knowledge and experience to assess this risk level. Parameter p is the probability of the occurrence of the same predefined adverse event within a defined time window, as calculated based on the patient risk model. The exponent / power parameter α can be adjusted to give more or less weight to the computer's risk parameter p compared to the clinician's risk parameter A.

[0091] In some instances, a calibration multiplier or factor β may be added to the above formula to obtain A / p α can provide a precise mapping between the numerical value resulting from the evaluation of and the time units given by T. For example, this formula can be written as T = β(A / p α) In some examples, β is a single-valued constant. In some examples, β may be a function of any or all of A, p, and α, such that β is a function of A / p α provides a transfer function between the value of A / p and the appropriate timing value T of the measurement interval. α Reference to using only β can be understood to optionally include the coefficient or transfer function β as a further multiplier.

[0092] Alternatively, this calibration may already incorporate ranges of acceptable values ​​for A, p, and α. For example, each of these parameters may take on values ​​across a respective range and / or at respective intervals across that range, and these values ​​may be expressed as A / p α is constructed so that evaluation of automatically results in a value of T with the correct units.

[0093] As an example, a typical 15-minute interval is estimated for a clinical scenario. Assume that hospital policy allows an average of one event per 1000 patients, A=0.001, as the level of risk. The probability of an event per minute is p=6.67*10 -5 / min, resulting in a recommended interval time T=15 min.

[0094] By way of example only, other different combinations of resulting interval times are shown in the table below, where the parameter α is set to α=1. [Table 1]

[0095] This formula has the effect that an increasing patient risk level of potentially suffering from a clinical event results in a decreasing interval between measurements.

[0096] Parameter A can be specified by a particular clinician treating a patient, and this parameter represents the clinician's assessment of the patient's risk level. Additionally or alternatively, parameter A can be set based on standardized guidelines in a particular medical department or based on general guidelines, for example, based on a look-up table that associates different classifications of patient status with different values ​​of parameter A.

[0097] The time-dependent estimation of the probability p of a clinical event allows the interval time T to be adjusted in real time.

[0098] In some embodiments, in addition to determining the value of T, the method further comprises determining an estimate of the uncertainty of T. To this end, an uncertainty parameter ΔT is calculated.

[0099] A simple means for estimating uncertainty is to define uncertainty interval values ​​ΔA and Δp, where Δp can be estimated from a Bayesian model. The Bayesian model essentially outputs a probability distribution with an uncertainty dp, which can be used for Δp. The value of ΔA can be predefined, for example using a lookup table, or can be defined according to hospital policy depending on the patient's condition or other factors.

[0100] Using ΔA and Δp, the uncertainty parameter ΔT estimated for T is

number

[0101] The calculated value of the uncertainty parameter ΔT is presented to the user (clinician) on the user interface display along with the current calculated value of T. The user is provided via this user interface with the option of accepting T as the current measurement interval or not. The value of the uncertainty parameter provides additional information to assist the user in making this decision. If the uncertainty is high, the user (clinician) may err on the side of caution and decide to set the interval at a time shorter than the current value of T.

[0102] In some examples, the user is provided with a set of three preset options for the measurement interval time: T, T+ΔT, or T−ΔT. For example, if the uncertainty is relatively high, the user may select T−ΔT.

[0103] In the above formula, the value of ΔA can be set to zero or estimated from historical data or guidelines.

[0104] Part of the method is the use of a patient risk model to calculate a risk parameter p indicative of a risk level of at least one predefined adverse clinical event, details regarding patient risk models are described in more detail herein.

[0105] In operation, the patient risk model is configured to receive inputs having at least values ​​for each of the patient's one or more biometric parameters (as described above), and to generate as an output a value p indicative of the probability of a predetermined adverse clinical event occurring within a particular time interval.

[0106] A particularly well-suited technique for estimating such probabilities is to use Bayesian statistical models.

[0107] A statistical model using Bayes' principle can determine a posterior probability distribution of a particular outcome based on one or more input variables, where the model is constructed based on one or more prior probability distributions for one or more events.

[0108] Thus, using Bayes' principle to derive an estimated value of risk based on the current values of one or more parameters and the prior distribution of these parameters is well known to those skilled in the art.

[0109] As an example, a Bayesian model can be configured to provide a value of p indicating, for example, that the estimated risk of a blood pressure decrease is below a value x for y minutes, where x and y are defined values set to correspond to a hypotensive event of interest. For example, x is 90 mm / Hg and y is 15 minutes. For this scenario, the prior distribution can be determined a priori as a function of the variation of the measured biological parameter d(t). When multiple biological parameters are measured and used as inputs to the model, the prior distribution for p is calculated as a function of the multivariate space of d(t), where d(t) is a vector representing all biological parameters (and / or other parameters) used as inputs to the model.

[0110] Using standard Bayes' theorem, the probability p that blood pressure (BP) is below x for y minutes is

Equation

[0111] Using standard Bayes notation, the above formula for T becomes

number

[0112] As mentioned above, the probability p can preferably be estimated iteratively, for example continuously or quasi-continuously, from acquired real-time biometric data and / or available historical data.

[0113] Using Bayesian methods, a wide variety of different factors and variables can be considered, as long as appropriate prior distributions are obtained. For example, in addition to the measured biometric data d(t), parameters such as: The training level of the patient's clinicians The rate at which clinicians respond over a given period of time Patient history Severity of the patient can also be included in the probability p output from the model.

[0114] Each of these variables can be organized according to its respective standardized score metric, so that the prior distribution of p can be measured over a range of values ​​of each.

[0115] By taking these various factors into account, models can be made to effectively mimic human mental processes in light of new data.

[0116] In some embodiments, the method can further include updating the patient risk model based on monitoring data of patients in a medical institution. For example, the monitoring data can include blood pressure measurement data and biometric data, such that relationships between these parameters can be used to train the model. Thus, the model can learn over time through retrospective analysis of monitored patients.

[0117] Additionally, changes in the training level of clinicians can be monitored and reflected in updates to the prior information used to train the model.

[0118] To update the model, the method may further include a model updating step, which may be performed once or at intervals that are the same as or different from the intervals for updating any of the parameters p and T. For example, this updating step may include running an algorithm to obtain and verify appropriate updated model priors and update the model with the updated priors.

[0119] It should be noted that while the above-described embodiments employ the use of a Bayesian model for the patient risk model, the risk model may be implemented using other types of models. For example, in a further example, a machine learning model is used that has one or more machine learning algorithms trained to receive at least the patient's biometric data as input and to generate as output a risk parameter p indicative of the risk level of at least one predefined adverse clinical event. By way of non-limiting example, suitable machine learning algorithms include multivariate regression models or artificial neural networks, such as convolutional neural networks.

[0120] It should be noted that although the above-described embodiments relate to blood pressure measurement intervals, the concepts of the present invention are more generally applicable to determining time intervals for any other biological measurements, e.g., other physiological parameter measurements, e.g., other vital sign measurements. The method is most applicable to measurement modalities that cannot or preferably cannot be performed continuously.

[0121] As mentioned above, the present invention can be embodied in software form. Thus, another aspect of the present invention is a computer program product having code means configured to, when executed on a processor, cause the processor to perform a method according to any example or embodiment of the invention described herein or according to any claim of the present patent application.

[0122] It should be noted that for any embodiment of the present invention, the method may be performed by a processing unit in any of a variety of different locations, for example, the method may be performed by a processing unit of a patient monitoring device at the bedside, by a patient monitoring system, by a separate computing system, or by a cloud-based computer system.

[0123] The above-described embodiments of the present invention use a processing unit. This processing unit typically has a single processor or multiple processors. The processing unit may be located within a single storage device, structure, or unit, or may be distributed among several different devices, structures, or units. Thus, reference to a processing unit being adapted or configured to perform a particular step or task corresponds to the step or task being performed by any one or more of multiple processing components, alone or in combination. Those skilled in the art will understand how such a distributed processing device can be implemented. The processing unit may include a communication module or input / output for receiving data and outputting data to further components.

[0124] The one or more processors of the processing unit can be implemented in various ways using software and / or hardware to perform the various functions required. The processor typically uses one or more microprocessors that are programmed using software (e.g., microcode) to perform the required functions. The processor may also be implemented as a combination of dedicated hardware to perform some functions and one or more programmed microprocessors and associated circuitry to perform other functions.

[0125] Examples of circuitry that may be used in various embodiments of the present disclosure include, but are not limited to, conventional microprocessors, application specific integrated circuits (ASICs), and field programmable gate arrays (FPGAs).

[0126] In various implementations, the processor may be associated with one or more storage media, e.g., volatile and non-volatile computer memory such as RAM, PROM, EPROM, and EEPROM. The storage media may be encoded with one or more programs that, when executed on the one or more processors and / or controllers, perform the required functions. The various storage media may be mounted within the processor or controller, or may be transportable such that one or more programs stored on the storage media can be loaded into the processor.

[0127] Variations to the disclosed embodiments can be understood and effected by those skilled in the art in practicing the claimed invention, from a study of the drawings, the disclosure, and the appended claims. In the claims, the word "comprising" does not exclude other elements or steps, nor does it exclude a plurality if a plurality is not stated.

[0128] A single processor or other unit may fulfill the functions of several items recited in the claims.

[0129] The mere fact that certain measures are recited in mutually different dependent claims does not indicate that a combination of these measures cannot be used to advantage.

[0130] The computer program may be stored / distributed on a suitable medium, for example an optical storage medium or a solid-state medium, supplied together with or as part of other hardware, but may also be distributed in other forms, for example via the Internet or other wired or wireless telecommunications systems.

[0131] When the term "adapted for" is used in the claims or the specification, it is meant to be equivalent to the term "configured to."

[0132] Any reference signs in the claims should not be construed as limiting the scope.

Claims

1. A computer implementation method, The steps include receiving the patient's biological measurement data, A step of applying a patient risk model to calculate a risk parameter p that indicates a risk level of at least one predefined adverse clinical event, wherein at least one input to the patient risk model is the biological measurement data, A step of obtaining a clinician-defined patient risk parameter A that indicates the clinician's assessment of the risk level for at least one predefined adverse clinical event, Based on A and p, and the following relationships [Math 11] The steps include: calculating the measurement time interval T for continuous non-invasive blood pressure measurement of a patient, wherein α is a predefined parameter; The steps include controlling a non-invasive blood pressure measuring device to acquire blood pressure measurements at a frequency determined by the time interval T, and A method having

2. The method according to claim 1, wherein the at least one adverse event or clinical event includes hypotension or hypertension in the patient.

3. The method according to claim 1 or 2, wherein the risk parameter p is the probability that the event occurs during a defined period.

4. The method according to claim 1, wherein the biological measurement data includes real-time sensor data or data derived from the sensor data.

5. The method according to claim 1, wherein the biological measurement data includes past biological measurement data of the patient retrieved from a data store.

6. The method according to claim 1, wherein the risk parameter p is iteratively recalculated over the patient's monitoring sessions, and the biological measurement data is updated each time it is recalculated.

7. The method according to claim 1, wherein the risk model is a Bayesian model.

8. The Bayesian model is a risk model that is personalized to the patient and / or the clinician treating the patient. Patient's medical history, Severity of the patient's condition, The training level of clinicians treating patients, and A measure of the speed at which clinicians respond to changes in a patient's condition. The method according to claim 7, which is set in advance according to prior information including one or more of the following.

9. The method according to claim 1, further comprising the step of iteratively adjusting or updating the risk model based on a patient monitoring database having monitoring data of multiple patients.

10. The aforementioned biological measurement data Electrocardiogram (ECG) detection device, Photoplethysmogram (PPG) detection device, Capnography measuring device, and Bioelectrical impedance measurement device The method according to claim 1, comprising data from one or more of the following.

11. A computer program product having code means configured to cause the processor to perform the method described in claim 1 when executed on the processor.

12. Input / Output, One or more processors and A processing unit having, wherein one or more processors In the aforementioned input / output, the patient's biological measurement data is received, Applying a patient risk model to calculate a risk parameter p that indicates the risk level of at least one predefined adverse clinical event, wherein at least one input to the patient risk model is biological measurement data, Obtain a clinician-defined patient risk parameter A, which indicates the clinician's assessment of the risk level for at least one predefined adverse clinical event. Based on A and p, and the following relationships 【Number 12】 The method involves calculating the measurement time interval T for continuous non-invasive blood pressure measurement of a patient, where α is a predefined parameter, and The non-invasive blood pressure measurement device is controlled to acquire blood pressure measurements at a frequency determined by the time interval T by generating a control signal for output at the input / output. A processing unit adapted to the environment.

13. The processing unit according to claim 12, A non-invasive blood pressure measuring device operably coupled to the processing unit. A system that has

14. The system according to claim 13, further comprising one or more biological parameter sensing devices for acquiring biological measurement data, operably coupled with the processing unit.

15. The system according to claim 14, wherein the one or more biological parameter sensing devices include a PPG sensor incorporated as part of a non-invasive blood pressure measurement device.

Images