JPH08103416A - Patient alarm detecting method using tendency fluctuation vector analysis - Google Patents

Abstract

PURPOSE: To generate an alarm at a more appropriate time in correspondence with a patient's symptom. CONSTITUTION: Physiological parameters of a patient are detected by an ECG sensor 12, a pressure sensor 14, an Sp O2 sensor 16 and a temperature sensor 18 attached to the patient, then input to a system processor 20 and the patient's state is monitored. An operator interface 28 is connected to determine whether the physiological parameter values of the patient are within a limit of safety scope. When the physiological parameter values are out of the limit of safety scope, the calculation of the tendency variable vectors which is the variable functions of the physiological parameter values starts. Then the tendency variable vectors are compared with the alarm limit functions. An alarm is generated when the tendency variable vectors are beyond the alarm limit functions.

Description

Detailed Description of the Invention

[0001]

FIELD OF THE INVENTION The present invention relates to medical monitoring of patients and, more particularly, to a method of patient alarm detection using improved trend vector analysis for alarm detection in patient monitors.

[0002]

2. Description of the Related Art Generally, a patient monitoring apparatus is used, for example, in a coronary care unit and an intensive care unit (int.
It is used to monitor the patient's condition in an intensive care unit). Such devices are typically, for example,
A bedside monitor with one or more sensors attached to the patient, such as ECG (electrocardiograph) sensors, blood pressure sensors, and temperature sensors.
or) is included. These sensors measure various physiological parameters of the patient. The measured parameters can be processed by the system processor and displayed on a video display screen or recorded for subsequent analysis. The patient's physiological information obtained from several bedside monitors can be sent to a central station located, for example, at the nursing station.

These bedside monitors and central stations waveform and / or physiological parameters.
Alternatively, it can be displayed as a numerical value. Another important function of the patient monitor is to raise an alarm when one or more physiological parameters indicate that the patient needs attention. Such an alert is necessary because the display screen of a patient monitoring system cannot always be observed. Alerts are typically appealed to both visual and audible alerts.

[0004]

In the conventional alarm reference designating method shown in FIG. 7, an upper threshold value 102 and a lower threshold value 104 regarding a measured value such as a heart rate must be set. When the measured value exceeds the upper threshold 102 or falls below the lower threshold 104, an alarm is triggered.
These thresholds are fixed irrespective of the transition of time, except for exclusion ranges 106 and 108 in which a warning is not issued even if a measured value appears outside both the thresholds for a short time. Selecting thresholds that prevent excessive false alarms while producing clinically acceptable results can be difficult. False alarms may frequently occur when the threshold is close to the desired value, and clinically significant alarm conditions may not be detected when the threshold is far from the desired value. In some cases, the user of the patient monitor may wish to alert the monitor when there is a sudden change in the patient's condition or when there is a gradual gradual deterioration in the patient's condition. Traditional fixed thresholds do not properly detect such conditions. Among other conventional methods, there are methods that combine other methods such as fixed delay, hysteresis, and refractory period with the threshold value. The problem with methods that use a combination of these approaches is that the additional complexity added to the process dramatically increases the complexity of the method and makes it costly to implement new functionality. It means that it will be impossible to see. Therefore, it is desirable to be able to describe complex timing conditions and thresholds in a single way. Various more sophisticated alarm criteria have been proposed. For example, JS Gravenstein et al. “Essential Noninvasive”
JH Philip's paper entitled "Thoughtful Alarms" in "Monitoring in Anesthesia" (1980) (pp.191-201).
And JH Philip's paper "Overview; Creating Practical Alarms for the Futur."
e (Summary: Creation of practical alarms for the future) ”(1989). It can be said that such devices are too complex to achieve relatively easy patient monitoring requirements.

[0005]

According to a first aspect of the present invention, a method for detecting a patient alert is provided. The method includes patient monitoring including at least one sensor for measuring a value representative of a physiological parameter, and a processor for processing the measured parameter value measured by the sensor to provide information representative of the measured parameter value. Used in equipment. In a patient alert detection method using trend vector analysis according to the present invention, a patient monitoring device measures a value representative of a physiological parameter and determines whether the measured parameter value is within a safe range limit. When the measured parameter value goes out of the safe range limit, a step of starting the calculation of a trend fluctuation vector (trend vector) which is a function of a change in the measured parameter value and time, and the trend fluctuation vector as an alarm limit function The step of comparing and the step of issuing an alarm when the trend variation vector exceeds the alarm limit function are executed.

It is preferable that the alarm limit function has a change limit that changes with the passage of time after the calculation of the trend change vector is started. Various alarm limit functions can be used for positive and negative trend vectors.
An alarm may only be triggered when the trend vector is outside a predefined exclusion range, which typically limits the range of changes that are too short to be nonsensical or that cannot occur physiologically. It is possible. This trend vector is calculated for each new measured parameter value that is outside the safe range limit. When the measured parameter value returns within the safe range limit, the calculation of the trend vector is stopped.

According to another aspect of the present invention, a patient monitor comprises a sensor for measuring a value representative of a physiological parameter and a processor for processing said physiological parameter value. The processor includes means for determining whether the physiological parameter value is within a safe range limit, and for initiating a trend vector calculation when the physiological parameter value is outside the safe range limit. Means for comparing the trend fluctuation vector with the alarm limit function, and means for raising an alarm when the trend fluctuation vector exceeds the alarm limit function.

The patient monitoring device preferably includes a display device. In this case, the processor comprises means for displaying on a display device information representative of the measured value of the physiological parameter. The processor includes means for displaying the trend vector on a display device as an arrow having a direction indicating the polarity of the change in the parameter value and having a length indicating the magnitude of the change in the parameter value. It is possible. Optionally, this trend arrow may be sized in color, or color may be used to represent the alarm condition of the parameter.

According to yet another aspect of the invention, a method for displaying information representative of physiological parameters is provided. The method is used in a patient monitor including at least one sensor for measuring a value representative of a physiological parameter, a processor for processing this measurement, and a display device. This patient monitor
Performing the steps of measuring a value representative of a physiological parameter, measuring a trend change of the parameter value over a specific time period, and displaying the trend change on the display device.

[0010]

DETAILED DESCRIPTION OF THE INVENTION A block diagram of a patient monitor suitable for implementing the patient alarm detection method using the trend vector analysis of the present invention is shown in FIG. A bedside monitor 10 as a clinical monitoring device typically includes one or more transducers or sensors located at the patient's bedside and attached to the patient. The transducer can include an ECG sensor 12, a blood pressure sensor 14, a S _P O ₂ sensor 16, and a temperature sensor 18. The number and type of such sensors is arbitrary. These sensors sense various physiological parameters of interest.

The physiological parameter measurements obtained by the above sensors are sent to the system processor 20. Typically, the analog sensor output signal is amplified and converted to digital data by an AD converter (not shown). This digital data representing the sensor signal is sent to the system processor 20. The system processor 20 includes a system memory 22 for monitoring the condition of the patient and providing information to the user.
And a display device 24 (typically a video display
Screen), the display controller 26,
Works in conjunction with the operator interface 28. The system processor 20 may be, for example, a Motorola 68.
It is also possible to include a 0x0 microprocessor.

The information displayed on the display device 24 may include a waveform of one or more physiological parameters, a numerical value of the one or more physiological parameters, and an alarm condition indicating that the patient needs attention. It is possible. The physiological parameter information obtained by the sensor can be stored in system memory 22 for later analysis. Information about the patient's condition can also be provided to the central station. One example of a bedside monitor of the type shown in FIG. 1 is the Model M1176A manufactured and sold by Hewlett-Packard Company.

The "vector intercept method" according to the present invention
An alarm detection method called "vector intercept method" will be described with reference to FIGS. As shown in FIG. 2, safe range limits are set for each physiological parameter. The physiological parameter may be a directly measured parameter or a parameter calculated from one or more measurements. Examples of calculated parameters include heart rate and cardiac index. The safe range limit is an upper limit 40 higher than the nominal parameter value 42 and a lower limit 44 lower than the nominal parameter value 42.
Defined by As will become apparent below, the safe range limits are upper 40 and lower 44 when appropriate.
Can be closer to the nominal parameter value than is the case with the thresholds used in conventional alarm methods without significantly increasing the false alarm rate. If the measured value is between the upper limit 40 and the lower limit 44 of the safe range limit, no alarm can be triggered. As will be described later, when the measured value exceeds the upper limit 40 or falls below the lower limit 44, the trend variation vector analysis is started.

When a measured value of a physiological parameter falls outside the safe range limit of that physiological parameter, a trend vector analysis is initiated, as shown in FIG. The trend variation vector 50 depends on the change in the measured value from the start of the trend variation vector analysis and the time elapsed from the start of the trend variation vector analysis. Both magnitude and polarity are associated with measurement changes. Typically, the measured value is, for example, 1
It is measured in periodic intervals such as seconds intervals. After the trend vector analysis is started, the trend vector is preferably calculated for each measured value of the physiological parameter until the trend vector analysis is completed as described later. For example, if you have a heart rate of 80 to 9 during a time of 20 seconds
If changing to 0, the trend vector can be represented by 10 measurement changes in a time of 20 seconds.

The trend vector 50 is an approximate value of the measured value change of the physiological parameter. After the physiological parameter value is outside the safe range limit (at zero time), measurements continue to be taken at periodic intervals. From these measured values, the measured value change is measured in proportion to the value measured when the physiological parameter value is outside the safe range limit. The values of the measured value changes are shown as points 60, 62, 64 in FIG.
And so on. The trend variation vector 50 is preferably a linear approximation of the measured value change value. In the preferred embodiment, conventional least squares approximation is used to calculate the trend vector 50. The trend variation vector 50 is calculated for each new measurement value.

Each of the calculated trend vector 50 is compared to an alarm limit function 52 or 54. The alarm limit functions 52 and 54 typically change according to the transition of time, starting from the time when the trend vector analysis is started. In one embodiment, the alarm limit function is a straight line connecting the maximum allowable change in the shortest interval and the minimum allowable change in the longest interval. Alternatively, the alarm limit functions 52, 54 can be obtained by graphing clinically significant changes as a function of time. In general, the alarm limit functions 52, 54 can be any function of time. The alarm limit function may be the same or different for positive and negative measurement change. As long as the trend vector 50 does not exceed the appropriate alarm limit function, no alarm will be generated and monitoring of the measured value will continue. An alarm is generated when the trend vector 50 exceeds one of the alarm limit functions. It should be appreciated that the alarm limit function may have a different value at each time interval after the trend vector analysis begins. Thus, for example, the alarm limit may be reduced over time.

The trend vector analysis is arbitrarily excluded in range 5.
6, 58 can be included. Trend vector 5
When 0 is included in one of the exclusion ranges 56 or 58,
No alarm is triggered. Exclusion ranges 56, 58 represent changes that are too short to be meaningless, physiologically improbable, or otherwise not warranted. For example, blood pressure cannot double within 3 seconds, so a blood pressure alert in this condition would not be appropriate. The exclusion ranges 56, 58 can be arbitrarily defined and can be different for positive and negative measurement changes.

The trend vector analysis continues until it is terminated. This termination can occur in either of two forms. When an alarm is generated as described above, the trend vector analysis is terminated. Alternatively, if the measured value of the physiological parameter returns to within the safe range limit as shown in FIG. 2, the trend vector analysis is terminated and it is determined whether the measured parameter value is within the safe range limit. Therefore, the measured parameter values are monitored as usual.

A flow chart of the alarm detection method of the present invention is shown in FIGS. In one of the preferred embodiments, the alarm detection method of the present invention includes a system processor 20 (FIG. 1).
It is implemented as a software program running on. The present invention can be implemented, for example, in the form of a C / C ++ programming language.

In step 70, the measured parameter values are read by system processor 20. As mentioned above, the output of each sensor is converted into digital data at periodic intervals and this digital data is provided to the system processor 20. In step 72, the measured parameter values are compared to the upper and lower safe limits 40 and 44 to determine if the measured parameter values are within safe range limits. No action is required when the measured parameter value is within the safe range limits,
The process returns to step 70 to read the next measured parameter value.

If the measured parameter value is outside the safe range limit, trend vector analysis is initiated as shown in FIG. The next measured parameter value is read in step 74. In step 76, the new measured parameter value is compared to the safe range limits upper 40 and lower 44 limits. If this new measured parameter value is outside the safe range limit, a trend vector analysis is calculated at step 78. As described above, the least squares fit calculation (least s
quares fit calculation) can be used. After this, in step 80, the trend vector is compared to the exclusion range. No action is required when the trend vector is within the exclusion range and the process returns to step 74 to read the next measured parameter value. Note that the exclusion range is arbitrary and may not be used in some applications. If the trend vector is outside the exclusion range, step 8
At 2, this trend vector is compared to the appropriate alarm limit function. As mentioned above, the alarm limit function typically changes over time. If the trend vector does not exceed the alarm limit function at step 82, no action is required and the process returns to step 74 to read the next measured parameter value. If the trend vector exceeds the alarm limit function, step 8
An alarm is issued at 4. After this, the trend vector analysis is terminated.

Referring again to step 76, if any of the measured parameter values are within the safe range limits, the trend vector analysis is terminated and normal measurement parameter values for the upper and lower safe range limits 40 and 44 are determined. The process returns to step 70 (FIG. 4) for comparison.

The trend variation vector can be displayed on the display device 24 as an arrow having a direction indicating the polarity of the change in the measurement parameter value and a length indicating the magnitude of the change in the measurement parameter value. Optionally, the trend arrows may be sized in color, or
Color can be used to indicate the alarm status of the measurement parameter.

An example of a display of a patient monitor including trend vector is shown in FIG. This display shows waveform 1
20, 122 and currently measured parameter values 126, 128,
It may include any required information such as 130, 132, 134. An indication of trends in parameter values is associated with one or more currently measured parameter values. In the example of FIG. 6, the trend change is represented by the upward arrow 140 when the measurement parameter value is increasing, and the downward arrow 1 when the measurement parameter value is decreasing.
The trend variation is represented by 42. A constant value can be displayed by the bar 144. Or "up"
An indication of the required trend can be indicated by words such as (up) or "down", or by any other trend indicator.
As noted above, the length and / or color of the trend indicators, such as up arrow 140, down arrow 142, may be used to indicate other phases of the trend, including magnitude and alarm condition. It is possible. In one embodiment, the trend indicator may be displayed only when the measured parameter value exceeds the safe range limit as described above. In another embodiment, the trend indicator is always displayed. The advantage of trend indicators is that not only is it possible for the clinician to look at the monitor to see what the current measured parameter value is, but also the measured parameter over a specific time period (eg, the last 15 minutes). It is possible to know what the values were.

Next, some examples are given to illustrate the advantages of the patient alert detection method using the trend vector analysis of the present invention. In the first example, as the patient becomes debilitated, the patient's respiration rate gradually decreases. In the case of the prior art, it is possible to have more than one hour to cross the low respiration rate alarm limit. Therefore, the patient may be hopeless by the time the alarm limit is exceeded. In most cases the above alarm limits may be adequate overall, but not in this case. The alarm detection method described above allows the condition to be detected by lowering the threshold as trends appear and by issuing an alarm earlier, so that the present invention provides the appropriate results. . If the alarm limit was initially set at the level at which the alarm was raised, then many false alarms would have occurred.

In the second example, the anesthetized patient may have low systolic pressure, which is not a significant problem. After a period of time, the response to the drug can cause the patient's blood pressure to rise rapidly. This condition is temporarily acceptable, but if the condition persists for a few minutes, which is the time it takes to reach the limits of the prior art systolic pressure alarms, the patient may become serious. May be damaged. The disclosed alarm detection method of the present invention detects the above condition by lowering the threshold in response to an increase in blood pressure and adjusting the trend vector to result in an earlier alarm. It

In the third example, the zero heart rate measurement is
Indicates that the patient has ventricular fibrillation, or
A series of transient motion arti
fact). With the disclosed alarm detection method of the present invention, the trend vector is first calculated to be within the exclusion range and no alarm will be issued. In a real emergency, subsequent values of the trend vector will exceed the alarm limit function and give rise to an asystole alarm in time for physician intervention.

While what is presently considered to be the preferred embodiments of the present invention have been illustrated and shown, without departing from the scope of the present invention as limited by the claims, It will be apparent to those skilled in the art that various modifications and variations can be made to the examples.

Although the respective embodiments of the present invention have been described in detail above, in order to facilitate understanding of the respective embodiments, the respective embodiments are summarized and listed below.

(1). At least one sensor (12, 14, 1) for measuring a value representative of a physiological parameter
6, 18) and a processor (20) for processing the measured values to obtain information representative of the parameter values, the method of detecting a patient alarm in a patient monitoring device 10, the step of measuring a value representative of a physiological parameter. (70),
Determining (72) whether the parameter value is within a safe range limit (40, 44), and as a function of the change in the parameter value and time when the parameter value falls outside the safe range limit. A certain trend vector (5
0) starting the calculation (78), comparing the trend variation vector (50) with the alarm limit function (52, 54) (82), and when the trend variation vector exceeds the alarm limit function. A method of detecting a patient alarm using trend vector analysis, including:

(2). 2. The patient alarm detection method using the trend fluctuation vector analysis according to 1, wherein the alarm limit function has a change limit that changes in accordance with the transition of time after the parameter value goes out of the safe range limit.

(3). The patient monitoring device includes a display device, and the method is the patient alert detection method using the trend vector analysis according to the above 1, further comprising displaying the trend vector on the display device.

(4). A patient monitoring device comprising a sensor (12, 14, 16, 18) for measuring a value representative of a physiological parameter and a processor (20) for processing said parameter value, said processor comprising: Means (72) for determining whether the parameter value is within the safe range limit, and a trend vector which is a function of the change in the parameter value and the time when the parameter value goes out of the safe range limit. Means (78) for starting the calculation of (50), and the trend vector (5)
Means (82) for comparing 0) to the alarm limit function;
If the trend vector (50) is the alarm limit function (5
2, 54) and means (84) for raising an alarm.

(5). The means for comparing the trend fluctuation vector with the alarm limit function includes means for comparing the trend fluctuation vector with a limit that changes according to the transition of time after starting the calculation of the trend fluctuation vector. The patient monitoring apparatus according to the above item 4.

(6). The patient monitoring device further includes a display device, and the processor further includes on the display device as an arrow having a direction indicating a polarity of a change in the parameter value and a length indicating a magnitude of the change in the parameter value. And a means for displaying the trend vector in
The patient monitoring device according to 1.

(7). The patient monitor further includes a display, and the processor further includes an indicator on the display as an arrow having a direction indicating a polarity of the change in the parameter value and a color indicating a magnitude of the change in the parameter value. 5. The patient monitoring apparatus according to 4 above, which also includes means for displaying a trend vector.

(8). The patient monitor further includes a display, and the processor further includes a trend on the display as an arrow having a direction indicating the polarity of the change in the parameter value and having a color indicating an alarm condition of the physiological parameter. The above-mentioned 4 which also includes means for displaying the variation vector
The patient monitoring device according to 1.

(9). At least one sensor (12, 14, 1) for measuring a value representative of a physiological parameter
6, 18), a processor (20) for processing the measurements, and a display device (24) for displaying information representative of the physiological parameter in a patient monitor. The patient monitor measuring a value representative of a physiological parameter (70), determining a trend of the parameter value over a particular time period (78), and the display (24). B) displaying the trend variation above; and displaying information representing a physiological parameter.

(10). A method for displaying information representing a physiological parameter, such as limited to the above item 9, wherein the step of indicating the trend change is performed only when the parameter value is outside the safe range limit.

[0040]

As described above, according to the present invention, it is determined whether or not the measured physiological parameter value of the patient is within the safe range limit, and the result of the determination is that the physiological parameter value is within the safe range. Beyond the limits, we begin to calculate a trend vector that is a function of changes in physiological parameter values,
By comparing the trend vector with the alarm limit function, and when the trend vector exceeds the alarm limit function, an alarm is generated, so that appropriate treatment according to the patient's condition can be performed, and the patient's serious condition It is possible to prevent the accident.

[Brief description of drawings]

FIG. 1 is a block diagram of a patient monitoring system suitable for implementing the present invention.

FIG. 2 is a graph of measured value change as a function of time showing the safe range limits used by the present invention.

FIG. 3 is a graph of measured value change as a function of time showing the trend vector used by the present invention.

FIG. 4 is a flowchart of a patient alert detection method using trend vector analysis according to the present invention.

FIG. 5 is another flow chart of a patient alert detection method using trend vector analysis according to the present invention.

FIG. 6 is an explanatory diagram showing an example of a display of a patient monitoring apparatus that displays trend changes in physiological parameters.

FIG. 7 is a graph of measured values as a function of time showing the alarm limits used in the prior art.

[Explanation of symbols]

10 Side Bed monitor 12 ECG sensor 14 pressure sensor 16 S _P O ₂ sensor 18 Temperature sensor 20 system processor 22 system memory 24 display 26 upper 42 nominal parameter values 44 of the display controller 28 operator interface 40 safety range limits Lower limit of safe range limit 50 Trend vector 52, 54 Alarm limit function 56, 58 Exclusion range

Claims (1)

[Claims] 1. At least one sensor (12, 14, 16, 1) for measuring a value representative of a physiological parameter.
8) and a processor (20) for processing the measured values measured by the sensor to obtain information representative of parameter values, the method for detecting patient alarms in a patient monitoring device 10 comprising the steps of: Measuring step (7
0) and determining (72) whether the parameter value is within the safe range limit (40, 44), and when the parameter value is outside the safe range limit, the change and time of the parameter value. A step (78) for starting the calculation of the trend fluctuation vector (50) which is a function of the warning fluctuation function (52,
54) A method of detecting a patient alarm using trend vector analysis, comprising: comparing (82) with (54); and issuing an alarm when the trend vector exceeds the alarm limit function.