« PreviousContinue »
FAULT DETECTION APPARATUS AND METHOD FOR PARENTERAL INFUSION SYSTEM
BACKGROUND OF THE INVENTION 5
This invention relates generally to systems for administering parenteral fluids to a patient, and, more particularly, to systems of this type having an infusion apparatus for infusing the fluid into the patient's vascular system. 10
Systems of this particular type have enjoyed widespread usage in hospitals for administering parenteral fluids at precise rates. The systems are useful for both venous and arterial infusions and typically include an infusion pump and an associated controlling device for 15 pumping the parenteral fluid through a fluid tube and needle to the patient's vein or artery.
One drawback to conventional infusion pump systems of this type is that the needle can sometimes become dislodged from the patient's vein or artery. This 2^ will normally cause an increase in back pressure, but the pump will nevertheless continue to pump fluid at substantially the same fixed rate. The fluid therefore can infiltrate into the patient's body tissue and cause severe damage. Similarly, the needle can sometimes become 25 dislodged from the patient entirely, yet the pump will continue to pump the fluid at the same fixed rate.
One known prior technique for detecting fluid infiltrations is to monitor the patient's skin temperature in the vicinity of the needle. Since the parenteral fluid is 30 ordinarily cooler than the patient's body temperature, and since the fluid is not carried away as rapidly when an infiltration occurs, an infiltration will ordinarily create a temperature drop in the vicinity of the needle. Thus, whenever a drop in skin temperature is detected, 35 it is deduced that an infiltration is occurring. This technique is not believed to have proven completely satisfactorily in all circumstances, such as, for example, when the parenteral fluid has a temperature substantially the same as that of the patient's blood. 40
Other known prior techniques for detecting an infiltration of a parenteral fluid into a patient's body tissue involve intervention by hospital personnel. In one such technique, an attendant visually inspects the region around the needle, to detect any swelling that might 45 indicate an infiltration. In another technique, useful only when the fluid is being administered from a bottle under the force of gravity, the attendant periodically lowers the bottle to an elevation below the needle such that fluid flows outwardly from the patient. If when this 50 is done the patient's blood does not appear in the fluid tube, it can be deduced that the needle is not in fluid communication with a vein or artery. Neither of these techniques has proven to be entirely satisfactory, one reason being that they both require the presence of 55 trained hospital personnel and cannot be performed automatically.
Still another prior technique for detecting infiltrations and other fault conditions is used in a parenteral administration system that regulates flow rate using a 60 pinch valve located in the fluid tube, between a drop chamber and the patient. In particular, the pinch valve is controllably adjusted in order to maintain the frequency of fluid drops into the drop chamber at a selected value. If the limits of the pinch valve are ex- 65 ceeded in attempting to maintain the selected drop frequency, it is deduced that a fault condition is present. Operator intervention is still required, however, in
order to determine the particular type of fault condition, e.g., an infiltration, that is present.
A complete dislodging of the needle from the patient, such that the fluid is directed onto his skin, bedding, etc., is ordinarily detected only through a visual inspection by a hospital attendant. Such active participation by hospital personnel is not believed to be an entirely satisfactory solution to this problem.
It should be appreciated from the foregoing that there still is a need for an effective method and apparatus for automatically detecting faults such as infiltrations or an open line in a parenteral administration system of the type having an infusion device. The present invention fulfills this need.
SUMMARY OF THE INVENTION
The present invention is embodied in a fault detection apparatus, and related method, for use with a parenteral administration system of the type having an infusion device for infusing a parenteral fluid through a fluid tube and needle to a patient's vascular system. The apparatus includes pressure transducer means for monitoring the pressure of the fluid in the fluid tube and producing a corresponding pressure signal. In accordance with the invention, the apparatus further includes automatic fault detection means for qualitatively evaluating the pressure signal to determine when the fluid tube is not in proper fluid communication with the patient's vascular system, and for producing a corresponding alarm signal. This frees hospital personnel to perform other tasks without the need for repeatedly monitoring the status of the parenteral administration system.
One embodiment of the fault detection means is adapted for use with a parenteral administration system that includes a pulsing-type infusion device for administering the parenteral fluid to a patient's venous system. The fault detection means produces the alarm signal whenever it detects an infiltration of the fluid into body tissue separate from the venous system. In this embodiment, the fault detection means analyzes the pressure signal following each infusion pulse to detect impedance changes distal to the needle. Specifically, the fault detection means detects an infiltration by determining if the pressure signal ever fails to return to its steady state level within a predetermined time duration following each infusion pulse. In particular, the fault detection means high-pass filters the pressure signal and compares the filtered pressure signal with a prescribed threshold. The alarm signal is produced whenever the filtered pressure signal exceeds the threshold for longer than a prescribed time period following each infusion pulse.
Another embodiment of the fault detection means detects infiltrations whenever the infusion device is infusing fluid at a relatively high rate, e.g., 40 ml per hour. In this embodiment, the fault detection means includes pressure change means for determining if the pressure signal ever increases by more than a particular amount during a predetermined time duration. This pressure change means preferably includes means for sampling the pressure signal at spaced intervals of time, along with means for comparing the current pressure signal sample with with previous pressure signal sample, to determine if the pressure signal has increased by more than the prescribed amount during the interval between samples. The pressure change means is preferaBRIEF DESCRIPTION OF THE DRAWINGS
bly enabled only after pressure derivative means has determined that the instantaneous rate of change of the pressure signal exceeds a prescribed level.
Two other embodiments of the fault detection means of the invention are adapted for use when the parenteral 5 administration system administers the parenteral fluid to a patient's arterial system. In one such embodiment, the fault detection means low-pass filters the pressure signal to remove the effects of the patient's heartbeats and compares the filtered pressure signal with a predeter- 10 mined threshold. An alarm signal is produced whenever the signal drops below the threshold. In another such embodiment, the fault detection means high-pass filters the pressure signal to pass only the signal components attributable to the patient's heartbeats. An alarm signal 15 is produced whenever a dropout in the heartbeat pulses is detected.
Yet another embodiment of the fault detection means detects an open line or air bubble in the fluid tube connection between the pulsing-type infusion device and the patient. When such a condition occurs, the configuration is underdamped and the pressure signal has a ringing characteristic following each infusion pulse. The fault detection means detects this ringing by ana- ^ lyzing the pressure signal following each infusion pulse, to measure impedance changes distal to the needle. Specifically, the fault detection means determines, after each infusion pulse, if the ac pressure signal (i.e., highpass filtered pressure signal) drops below a prescribed ^Q negative pressure threshold. This threshold is preferably a prescribed negative fraction of the peak positive ac pressure signal occurring immediately after each infusion pulse.
In the preferred embodiment of the apparatus, the -j5 fault detection means includes several of the embodiments described above, for detecting venous and arterial infiltrations and open lines. Appropriate switching circuitry enables operation of either the venous or the arterial infiltration circuit, depending on the use to 40 which the system is being made. The alarm is actuated when any of the various circuits detects a fault.
Other aspects and advantages of the present invention should become apparent from the following description of the preferred embodiments, taken in con- 45 junction with the accompanying drawings, which illustrate, by way of example, the principles of the invention.
FIG. 1 is a block diagram of a parenteral administration system having circuitry for detecting venous infiltrations, arterial infiltrations and open fluid lines;
FIG. 2 is a simplified schematic diagram of a low infusion rate venous infiltration detector included in the 55 system of FIG. 1;
FIGS. 3(a)-(/) illustrate a series of waveforms that can be present in the venous infiltration detector of FIG. 2;
FIG. 4 is a simplified schematic diagram of a high 60 infusion rate venous infiltration detector included in the system of FIG. 1;
FIG. 5 is a simplified schematic diagram of one embodiment of an arterial infiltration detector suitable for use in the system of FIG. 1; 65
FIG. 6 is a simplified schematic diagram of an alternative arterial infiltration detector suitable for use in the system of FIG. 1;
FIG. 7 is a simplified schematic diagram of an open line detector included in the system of FIG. 1; and
FIGS. 8(a)-(c) illustrate several waveforms that can be present in the open line detector of FIG. 7.
DESCRIPTION OF THE PREFERRED
Referring now to the drawings, and particularly to FIG. 1, there is shown fault detection circuitry 10 for use in a system for administering a parenteral fluid to the vascular system of a patient 11. The system includes a conventional infusion pump 13 and associated pump controlling device 15 for pumping the parenteral fluid through a fluid tube 17 and needle 19 to the patient. The pump is preferably of the peristaltic type, which pumps the fluid in a cyclic fashion. One such suitable pump and an associated controller for controlling its speed are described in a copending and commonly-assigned application for U.S. Patent, Ser. No. 06/281,848, filed July 9, 1981, in the names of Stephen H. O'Leary et al. and entitled "Method and Apparatus for Fluid Flow Control."
The pump controlling device 15 outputs a motor step signal for coupling on line 21 to the infusion pump 13. The signal is a sequence of pulses, each of which increments the pump by one step, to infuse a predetermined volume of parenteral fluid to the patient 11.
The parenteral administration system further includes a pressure transducer 23 and an associated amplifier 25 for monitoring the fluid pressure in the fluid tube 17 and producing a corresponding pressure signal for output on line 27.
In accordance with the invention, the fault detection circuitry 10 evaluates the pressure signal on line 27 to detect certain characteristic patterns indicative of an improper fluid communication between the fluid tube 17 and the patient's vascular system. An alarm 29 is actuated if the circuitry detects such a condition. Such fault conditions include infiltrations of the fluid into body tissue other than the patient's vascular system, as well as a complete dislodging of the needle 19 from the patient 11 or a leak or air bubble in the fluid tube. In this way, a proper administration of the parenteral fluid can be ensured without the need for frequent monitoring or testing by hospital personnel.
More particularly, the fault detection circuitry 10 includes a low infusion rate venous infiltration detector 31 for detecting infiltrations when the infusion pump 13 is pumping parenteral fluid into the patient's venous system at a relatively low rate, and a high infusion rate venous infiltration detector 33 for detecting infiltrations when the pump is pumping fluid into the venous system at a relatively high rate. The fault detection circuitry further includes an arterial infiltraton detector 35 for detecting infiltrations when the pump is pumping fluid into the patient's arterial system, and an open line detector 37 for detecting when there is a leak of some kind or an air bubble in the fluid connection between the pump and the patient 11.
The system further includes a mode switch 39 for indicating whether the system is intended to administrater parenteral fluid to the patient's venous system or arterial system. The system also includes an infusion rate detector circuit 41, operable whenever the switch indicates that the system is pumping fluid into the patient's venous system to indicate whether the fluid is being pumped at a relatively high rate or a relatively low rate. The switch 39 and detector circuit 41 are used
to enable operation of the appropriate venous infiltration detector circuit 31 or 33 or arterial infiltration detector circuit 35, depending on the system's operating mode.
More particularly, the mode switch 39 is a single- 5 pole, double-throw switch having its middle terminal connected directly to ground and its two remaining terminals connected through separate resistors 43 to a positive voltage. The binary signals present on these two terminals are therefore opposite in phase to each 10 other. One such signal is defined to be an arterial enable signal and the other is defined to be a venous enable signal.
The arterial enable signal is coupled on line 45 directly to the arterial infiltration detector circuit 35, and 15 the venous enable signal is coupled on line 47 to the infusion rate detector 41. The infusion rate detector circuit relays the venous enable signal to either the low infusion rate venous infiltration detector 31 or the high infusion rate venous infiltration detector 33, depending 20 on the infusion rate being effected by the infusion pump 13.
The infusion rate detector circuit 41 includes a frequency discriminator 49 for monitoring the motor step signal present on line 21 and producing an output signal 25 having a voltage level generally proportional to the motor step signal's frequency. This output signal is coupled on line 51 to the positive input terminal of a comparator 53, which compares it with a selected reference level coupled to its negative input terminal. The 30 reference level is supplied on line 55 from the wiper of a potentiometer 57, whose remaining two terminals are connected between ground and a positive supply voltage. If the discriminator output signal exceeds the threshold, indicating that the infusion pump 13 is pump- 35 ing at a relatively high rate (e.g., above about 40 ml per hour), the comparator outputs a positive voltage level. On the other hand, if the discriminator output signal does not exceed the threshold, indicating that the pump is pumping at a relatively low rate, the comparator 40 outputs a low voltage level signal.
The signal output by the comparator 53 is coupled on line 59 to a first AND gate 61 for ANDing with the venous enable signal supplied on line 47 from the mode switch 39. This produces a high infusion rate venous 45 enable signal for coupling on line 63 to the high infusion rate venous infiltration detector 33. The detector 33 is thereby enabled to detect infiltrations whenever a venous infusion is selected by the mode switch and the infusion rate exceeds the prescribed threshold. 50
The signal output by the comparator 53 of the infusion rate detector 41 is also coupled on line 59 to a NOT gate 65, for inversion and coupling in turn on line 67 to a second AND gate 69, where it is ANDed with the same venous enable signal present on line 47. The result- 55 ing low infusion rate venous enable signal is coupled on line 71 to the low infusion rate venous infiltration detector 31. This detector 31 is thereby enabled whenever a venous infusion is selected by the mode switch 39 and the infusion rate does not exceed the prescribed thresh- 60 old.
Referring now to FIG. 2, there is shown a simplified schematic diagram of the low infusion rate venous infiltration detector 31. This circuit monitors the pressure signal supplied on line 27 from the amplifier 25, to de- 65 tect a characteristic pattern indicative of an occlusion or infiltration of fluid into the patient's body tissue separate from his venous system. When the circuit detects
such a condition, it outputs an alarm signal for coupling on line 73 to the alarm 29. Basically, the circuit determines that an occlusion or an infiltration has occurred whenever the pressure signal fails to return to its nominal value within a prescribed time duration following each of the successive pumping pulses of the infusion pump 13.
The low infusion rate venous infiltration detector 31 includes a 0.1 Hz high-pass filter 75 for filtering the pressure signal supplied on line 27, to produce a filtered pressure signal for output on line 77. The pressure signal, which is depicted in FIG. 3(c), normally includes a series of positive pressure pulses, each with an exponentially-decaying tail, as a result of the successive motor steps of the infusion pump 13.
A comparator 79 compares the filtered pressure signal with a selected positive voltage threshold supplied on line 81 from the wiper of a potentiometer 83. The other two terminals of the potentiometer are connected between ground and a positive potential. If the filtered pressure signal has a voltage level exceeding the threshold, the comparator outputs a positive signal on line 85 as shown in FIG. 3(d). The threshold is preferably selected to be about 30 percent of the pulse's peak value. Typically, the threshold is about four to six cm H2O.
The detector 31 further includes a monostable multivibrator or one-shot 87 and a flip-flop 89 for sampling the signal output by the comparator 79 about 0.2 seconds after each pulse of the motor step signal (FIG. 3(a)). In particular, the motor step signal is supplied on line 21 to the one-shot, which produces a corresponding sequence of pulses (FIG. 3(b)), each having a duration of about 0.2 seconds. This one shot signal is coupled on line 91 to the clock input terminal of the flip-flop, which samples the pressure derivative signal at the trailing edge of each pulse. If the comparator output signal is still a positive value at this time, the flip-flop outputs a positive signal as well.
The signal output by the flip-flop 89 is coupled on line 93 to one input terminal of an AND gate 94, which ANDs the signal with the low infusion rate venous enable signal supplied on line 71. It will be recalled that this enable signal indicates that the parenteral administration system is intended to be administering fluid to the patient's venous system at a relatively low rate. If both inputs to the AND gate are a positive value, it is deduced that an infiltration is occurring and the AND gate outputs a trigger signal for coupling on line 95 to a latch 96, which, in turn, produces an alarm signal for coupling on line 73 to the alarm 29 (FIG. 1).
During normal situations, when the needle 19 is properly inserted into the patient's vein, the pressure signal (FIG. 3(c)) returns to its nominal value relatively quickly after each pulse of the motor step signal (FIG. 3(a)). As a result, the pressure signal does not exceed the threshold and the comparator output signal (FIG. 3(d)) is at a low level at the successive sample times. The detector circuit 31 therefore does not produce an alarm signal.
On the other hand, during abnormal situations, when the needle 19 is not in proper communication with the patient's venous system and an infiltration is occurring, the pressure signal (FIG. 3(e)) has a relatively long decay time. This apparently occurs because of an inability of the body tissue to dissipate rapidly each infusion of parenteral fluid. Because of the long decay time, the comparator output signal (FIG. 3(f)) is still at a high