Atlas of Procedures in Neonatology, 4th Edition

Physiologic Monitoring


Blood Pressure Monitoring

  1. Kabir Abubakar

Accurate monitoring of blood pressure is essential for the optimal management of ill infants. Blood pressure is one of the most frequently measured physiologic parameters to evaluate clinical stability in critical care. The recognition and treatment of abnormal blood pressure states has significant prognostic implications in neonatal intensive care (1).

Noninvasive (Indirect) Methods

There are two types of noninvasive blood pressure measurements: auscultatory measurement (manual noninvasive) and oscillatory arterial blood pressure measurement (automatic noninvasive).

Auscultatory Measurement (Manual Noninvasive)

Manual noninvasive auscultatory measurement employs the sphygmomanometer, which is simple, inexpensive, and allows for intermittent blood pressure measurements.

  1. Background
  2. This technique employs a blood pressure cuff, insufflator, manometer, and stethoscope.
  3. The sphygmomanometer uses a pneumatic cuff to encircle the upper arm or leg and a pressure gauge (manometer) to register the pressure in the cuff.
  4. There are two types of manometers:
  5. Mercury (mercury column)
  6. Aneroid (mechanical air gauge)
  7. The encircling pneumatic cuff is inflated to a pressure higher than the systolic pressure in the underlying artery. The cuff pressure compresses the artery and stops blood flow.
  8. A stethoscope placed distal to the cuff, over the occluded artery, will pick up the Korotkoff sounds as the pressure of the cuff is decreased to the point at which blood flow resumes through the artery.

Korotkoff sounds are the noise generated by blood spurting from the compressed artery, producing turbulence and vibration within the vessel.

  1. As the pressure in the cuff is reduced from above the systolic pressure, five characteristic Korotkoff sounds are heard:
  2. Korotkoff I is a sharp thud.
  3. Korotkoff II is a loud blowing sound.
  4. Korotkoff III is a soft thud.
  5. Korotkoff IV is a soft blowing sound.
  6. Korotkoff V is silence.
  7. An 8- to 9-MHz Doppler device can be used in place of a stethoscope. This device will detect only systolic blood pressure levels.
  8. Indications
  9. Measurement of blood pressure in stable infants or when invasive blood pressure measurement is not required or unavailable
  10. When only intermittent blood pressure measurements are required
  11. Contraindications
  12. Profound edema in the limb to be measured; will affect result
  13. Decreased perfusion, ischemia, or infiltrate in limb
  14. Peripheral venous/arterial line in limb
  15. Limitations
  16. Provides only intermittent blood pressure measurements
  17. Manual measurement cumbersome in small infants
  18. Accuracy depends on ability to recognize Korotkoff sounds and may be user-dependent
  19. Pressure may not be detectable in low perfusion state or shock. Do not assume that it is simply an equipment problem; use clinical correlation.


  1. Pressure is not detectable or is inaccurate in neonates experiencing convulsions or tremors.
  2. Measures only systolic and diastolic blood pressure; mean blood pressure measurement not available
  3. Can be used only to measure pressure in the upper arm or thigh
  4. Inaccurate measurements (Table 8.1)
  5. Defective manometer
  6. Air leaks at the tubing connections or from dry, rotted tubing
  7. Inappropriate cuff size
  8. Cuff applied loosely
  9. Rapid deflation of cuff
  10. Active or agitated patient

TABLE 8.1 Sources of Error in Indirect Blood Pressure Measurements


Effect on Blood Pressure


Defective manometer

Falsely low values

1. Check level of mercury at zero cuff

1. Air leaks


2. Improper valve function

2. Check for cleared definition of meniscus

3. Dry, rotted tubing

3. Verify that pressure holds when tightened

4. Loss of mercury

Inappropriate cuff size

Verify appropriately sized cuff

1. Too narrow

1. Falsely high values

2. Too wide

2. Falsely low values

Cuff applied loosely

Falsely high values owing to ballooning of bag and narrowing of effective surface

Apply cuff snugly

Cuff applied too tightly

Inaccurate reading owing to impedance of flow through artery

Apply cuff snugly without undue pressure

Rapid deflation of cuff

1. Falsely low values owing to inaccurate detection of beginning of sounds or

Deflate cuff at rate of 2–3 mm Hg/s


2. Falsely high values owing to inadequate equilibration between cuffpressure and manometer pressure


Active or agitated patient


Recheck when patient is quiet

TABLE 8.2 Neonatal Cuff

Cuff No. (Size)

Limb Circumference (cm)











From American Academy of Pediatrics Task Force Pressure Control: Report. Pediatrics. 1977;59:797, with permission.

  1. Equipment
  2. Neonatal cuff (Table 8.2). Select a cuff that will fit comfortably around the upper arm or thigh; the inflatable bladder should completely encircle the extremity without overlapping. The width should be 90% of the limb circumference at the midpoint (2).
  3. Mercury manometer or aneroid-type gauge
  4. Appropriate size stethoscope with diaphragm or Doppler system
  5. Precautions (Table 8.1)
  6. Carefully select the appropriate cuff size, because incorrect size can significantly alter the blood pressure recorded (3).
  7. Cuff too small, blood pressure will be higher; cuff too large, blood pressure will be lower
  8. Check functional integrity of manometer
  9. Check integrity of cuff for leaks
  10. Check speed of cuff deflation: If deflation is too rapid, accuracy may be compromised.
  11. Patient must be quiet and still during measurements.
  12. For optimal infection control, use disposable cuff issued to the patient.
  13. Technique
  14. Place the infant supine, with the limb fully extended level with the heart.



  1. Measure the circumference and select the appropriate size cuff for the limb.
  2. Neonatal cuffs are marked with the size range (Fig. 8.1).
  3. When the cuff is wrapped around the limb, the end of the cuff should line up with the range mark.
  4. If the end of the cuff falls short of the range mark, the cuff size is too small.
  5. If the end of the cuff falls beyond the range mark, the cuff size is too large.
  6. Apply the cuff snugly to the bare limb, above the elbow or knee joint.
  7. Place the stethoscope or Doppler over the brachial artery for the upper arm or above the popliteal artery for the thigh.
  8. Inflate the cuff rapidly to a pressure 15 mm Hg above the point at which the brachial pulse disappears.
  9. Deflate the cuff slowly.
  10. The pressure at which a sound is first heard is the systolic pressure (Korotkoff I). The pressure at which silence begins corresponds to the diastolic pressure (Korotkoff V). The pressure should be measured to the nearest 2 mm Hg.

In patients in whom the sounds do not disappear, the point at which the sounds change abruptly to a muffled tone can be accepted as an approximation of the diastolic pressure but will be slightly higher than true diastolic pressure.


FIG. 8.1. Properly sized infant blood pressure cuff. Arrow indicates cuff range mark.

  1. Complications
  2. Perfusion in the limb may be compromised if the cuff is not completely deflated.
  3. Nosocomial infection may result from using the same cuff for more than one patient.
  4. Prolonged or repeated cuff inflation has been associated with ischemia, purpura, and/or neuropathy.
  5. Cuff inflation will interfere with pulse oximetry measurement in the same limb.

Oscillometric Measurement of Arterial Blood Pressure (Automatic Noninvasive)

  1. Background

The oscillometric technique offers a means for measuring all arterial blood pressure parameters (systolic, diastolic, mean, heart rate) (4,5,6,7,8 and 9). This technique is referred to as NIBP (noninvasive blood pressure).

  1. This technique employs a blood pressure cuff interfaced to a computerized blood pressure monitor.
  2. The pneumatic cuff is used in the same fashion as with the auscultatory technique.
  3. The monitor employs a miniature computer-controlled air pump and a bleed valve to control inflation and deflation of the cuff.
  4. A pressure transducer interfaced to the cuff tubing senses the pressure pulsations transmitted to the cuff by the underlying artery and also the inflation pressure of the cuff.
  5. The system will inflate the cuff to a level above the point at which no pulsations are detected.
  6. As the cuff is being deflated to the level of the systolic pressure, arterial pulses are transmitted to the cuff.

The systolic pressure is assigned the value of the cuff pressure at the time pulsations were initially detected.

  1. With most systems, the mean pressure value is determined by the highest pulsation level detected at the lowest cuff pressure.
  2. The diastolic value is determined by the lowest cuff pressure before baseline arterial pulsations are detected (Fig. 8.2).
  3. Heart rate values are calculated by computing the mean value of the time interval between pulsations.
  4. Higher detection sensitivity allows this technique to be used on parts of the extremities where auscultatory methods are not possible (i.e., distal arm and lower leg).




FIG. 8.2. Determination sequence for oscillometric measurement.

  1. Indications
  2. Measurement of blood pressure in stable infants or when invasive blood pressure measurement is not required or unavailable
  3. When only intermittent blood pressure measurements are required
  4. Contraindications
  5. Profound edema in the limb to be measured; will affect result
  6. Decreased perfusion, ischemia, or infiltrate in limb
  7. Peripheral venous/arterial line in place in limb
  8. Limitations
  9. Provides only intermittent blood pressure measurements
  10. Pressure may not be detectable in low perfusion state or shock. Do not assume that it is simply an equipment problem; use clinical correlation.
  11. Pressure is not detectable or may be inaccurate in neonates experiencing convulsions or tremors.
  12. Inaccurate measurements (Table 8.1)
  13. Air leaks at the tubing connections or from dry, rotted tubing
  14. Aborted determination cycle due to air leaks in the cuff, air hoses, or connection points
  15. Inappropriate cuff size
  16. Cuff applied loosely
  17. Active or agitated patient
  18. Equipment
  19. Neonatal noninvasive blood pressure monitor—display should include systolic, diastolic, mean, and heart rate values (Fig. 8.3)
  20. Neonatal cuff (designed for use with the specific monitor)—cuff may be single-tube or double-tube type, provided the appropriate adapter is used. Neonatal cuff sizes range from 1 to 5 (Table 8.2).

FIG. 8.3. Oscillometric blood pressure monitor. (Courtesy of GE Healthcare.)



  1. Precautions
  2. Carefully select appropriate cuff size, because incorrect size can significantly alter the blood pressure value obtained.
  3. Oversized cuff will yield lower blood pressure values; undersized cuff will produce higher pressure values.
  4. Patient must be still during measurements.
  5. For optimal infection control, cuffs should be for single-patient use.
  6. Caution should be exercised when used for preterm, very low-birthweight infants in a hypotensive state (10).
  7. Technique
  8. Become familiar with the monitor and the equipment to be used. Be aware of the normal blood pressure changes with gestational and postnatal age (11).
  9. Measure the circumference of the extremity where the cuff is to be applied. Select the appropriately sized cuff for the limb (Fig. 8.4).
  10. Apply the cuff snugly to the limb. The cuff can be applied over a thin layer of clothing if necessary; however, a bare limb is recommended.
  11. Attach the monitor air hoses to the cuff. The limb from which pressure is to be measured should be level with the heart.
  12. Turn the monitor on and ensure that it passes the power-on self-test before proceeding.
  13. Press the appropriate button to start a blood pressure determination cycle.
  14. If the values obtained from the initial cycle are questionable, repeat step 6.

Multiple readings with similar values yield the optimal assurance of accuracy.

  1. If, after repeating the cycle, readings are still questionable, reposition the cuff and repeat the measurement.
  2. Periodic inspection of the cuff and extremity is critical to avoid problems such as cuff detachment or shift in extremity position.
  3. Most NIBP systems can be programmed by the user to measure blood pressure automatically at user-determined intervals.

The interval between measurements should be as long as possible to ensure adequate circulation and minimize trauma to the limb and skin distal to the cuff.

  1. In infants with suspected congenital heart disease, blood pressure should be measured in all four extremities or at least the right arm and right leg for pre- and postductal comparison.

FIG. 8.4. Cuff of correct size applied to upper arm.

  1. Complications
  2. Perfusion in the limb may be compromised if the cuff is not completely deflated.
  3. Nosocomial infection may arise from using the same cuff for more than one patient.
  4. Repeated continuous cycling may cause ischemia, purpura, and/or neuropathy.
  5. Cuff inflation will interfere with pulse oximetry measurement and IV infusion in the same limb.

Continuous Blood Pressure Monitoring (Invasive)

  1. Purpose

Accurate blood pressure measurements are essential to guide the level of circulatory support provided to critically ill infants. Direct reading through an intravascular catheter provides the most accurate measurement and is the “gold standard” in measuring blood pressure.

  1. Background
  2. Blood pressure measurement is obtained from the vascular system via a catheter that has been introduced into a vein or artery. The catheter is coupled to a pressure transducer outside the body by a fluid-filled conduit.



  1. Because fluid is noncompressible, it is the medium used to transmit the pressure from the vessel to the transducer.
  2. The tubing used to couple the catheter to the transducer is a special high-pressure tubing with low compliance that is capable of withstanding high pressures without breakage and with very little stretching.
  3. A blood pressure transducer is a device that converts mechanical forces (pressure) to electrical signals. There are two major types of transducers:
  4. Strain gauge pressure transducer:composed of metal strands or foil that is either stretched or released by the applied pressure on the diaphragm

Applied pressure causes a proportional and linear change in electrical resistance. Problems associated with strain gauges include drift due to temperature changes (departure from the real signal value), fragility, and cost.

  1. Solid-state pressure transducer (semiconductor):composed of a silicone chip that undergoes electrical resistance changes because of the applied pressure.

This technology offers a superior alternative to the strain gauge.

  1. Because of its low cost, it has made possible inexpensive, accurate, disposable transducers.
  2. Because of the miniature integration on the silicone chip, the circuitry necessary to minimize temperature drift is incorporated in the device.
  3. Pressure transmitted to the transducer causes a micromechanical displacement of the diaphragm (receptor of mechanical energy linked to the transducer element) or directly to the transducer.
  4. This mechanical displacement causes a proportional change in the electrical resistance of the transducer.
  5. Excitation voltage applied to the transducer is responsible for the signal produced in relation to the pressure applied.
  6. The standard medical blood pressure transducer output rating is 5 mV/V/cm Hg.
  7. Miniature transducer-tipped catheters are available that do not depend on fluid-filled lines for the transmission of pressure.

Microtransducer catheters in general have better fidelity characteristics, but at a much higher cost than conventional fluid-filled systems.

  1. The pressure monitor processes the electrical signal generated by the transducer and converts it to blood pressure units in either millimeters of mercury or kilopascals. Part of this processing involves the detection of systolic, diastolic, and mean values.
  2. Indications
  3. To monitor intravascular pressure continuously
  4. In very small or unstable infants, particularly those with severe hypotension (shock)
  5. During major procedures that could cause or exacerbate intravascular instability
  6. To monitor infants on aggressive ventilator support or extracorporeal membrane oxygenation
  7. Contraindications

None absolute, except for those specific to catheter placement

  1. Limitations
  2. The pulse pressure waveform measured in the periphery is narrower and taller than that in the proximal aorta. Thus, systolic blood pressure in the peripheral arteries can be higher than that in proximal aorta. This amplification is greater in patients with increased vascular tone or on inotropic therapy.
  3. Very–small-diameter catheters may result in underreading of systolic blood pressure.
  4. Equipment
  5. Neonatal physiologic monitor (multiparameter monitoring system)
  6. Minimum configuration should have the capability of displaying systolic, diastolic, and mean pressures.
  7. It should have provision for high and low alarm settings.
  8. Mechanical infusion device (infusion pump)

Pressurized intravenous bag should never be used.

  1. Fluid to be infused via the pressure line
  2. Appropriately sized catheter and supplies
  3. Pressure-monitoring kit with integrated disposable transducer and continuous flush device (Fig. 8.5)
  4. Pressure-monitoring tubing should not exceed 48 inches from the transducer to the patient connection.
  5. The distal end of the kit should have a stopcock or an arterial extension no longer than 12 inches that can be used for drawing blood samples.
  6. Some disposable pressure-monitoring kits offer closed-loop systems for sampling.
  7. The system employs a mechanism for aspirating and holding a fixed amount of blood in the pressure tubing rather than in a syringe.



  1. The distal end is equipped with a small chamber with a rubber septum that allows a self-guiding short blunt syringe adapter to penetrate and aspirate blood for the sample.

The initial volume pulled back is sufficient to ensure that the blood drawn into the sample chamber is greater than the catheter/distal tubing volume and is not diluted by the fluid being infused. The absence of stopcocks at the distal end eliminates a possible site for contamination. In addition, the blood pulled back is conserved, and the amount of fluid used to flush the sample line is reduced.


FIG. 8.5. Representative disposable blood pressure transducer setup. (1) Pressure transducer; (2) integral continuous flush device; (3) infusion port (connects to infusion pump); (4) high-pressure tubing.

  1. Technique

For catheter placement, see Section 5 of the atlas, “Vascular Access”


FIG. 8.6. Arterial pressure waveforms: normal arterial waveform (top); dampened arterial waveform (middle); arterial waveform with spike caused by catheter whip or inappropriate tubing (bottom). (Note that figure demonstrates waveform appearance only and not actual pressure values.)

  1. Familiarize yourself with the bedside monitor and the pressure zero/calibration procedure.
  2. If using discrete components, assemble the pressure-monitoring circuit, maintaining the sterile integrity.
  3. A basic circuit configuration will consist of a transducer dome, flush device, stopcock, pressure tubing, and an optional arterial extension set (short length of pressure tubing, less than 12 in in length, inserted between the catheter and the pressure tubing).
  4. If using a pressure-monitoring kit, ensure that all the Luer-lock connections are tight and free of any defects.
  5. If possible, avoid the use of intravenous tubing components in the pressure-monitoring circuit.
  6. Set up the infusion pump that will be used for the continuous infusion through the flush device. Continuous flush devices limit flow rates to 3 or 30 mL/h, depending on the model (12). For neonatal arterial lines, the infusion pump supplying the flush device should be set to 0.5 to 3 mL/h and should never exceed the flow rating of the flush device. When pump flow exceeds the flush device rating, it will cause an occlusion alarm in most intravenous pumps. A pump flow rate of 1 mL/h is recommended for most arterial lines.
  7. For circuit priming, use the solution that will be used for the continuous infusion. Prime the circuit slowly, in order to avoid trapping air bubbles in the flush device inlet. Ensure that the entire circuit and all the ports are filled and bubble-free.
  8. If using disposable transducers, connect the reusable interface cable to the transducer and to the monitor. Turn the monitor on.
  9. Secure the transducer at the patient's reference level, defined as the midaxillary line (heart level). If using


transducer holders, level the reference mark on the holder at the patient's reference level.

  1. Connect the distal end of the circuit to the patient's catheter, ensuring that the catheter hub is filled with fluid and is bubble-free.
  2. Start the infusion pump. The pump rate cannot exceed the flow rate of the flush device.
  3. Open the stopcock connected to the transducer to air (shut off to the patient, open to atmosphere).
  4. Zero/calibrate the monitor according to the manufacturer's instructions.
  5. Close the stopcock connected to the transducer (open to the patient).
  6. Set the monitor pressure waveform scale to one that accommodates the entire pressure wave.
  7. Observe the waveform obtained. If the wave appears to be dampened (flattened, poorly defined, with slow rise time), check the circuit for air bubbles starting at the distal end. If no air bubbles are detected, then gently flush the catheter (Fig. 8.6). (For central venous waveforms, see Chapter 29.)
  8. Once a stable pressure reading is obtained, set the alarm limits. Mean pressure value is optimally used to set alarm limits (Fig. 8.7).
  9. Zero the transducer every shift or every 8 hours.
  10. When blood samples are drawn from the line, flushing should be done gently with a syringe using a minimal amount of heparinized saline solution.

FIG. 8.7. Pressures obtained by direct measurement through umbilical artery catheter in healthy newborn infants during first 12 hours of life. Broken lines represent linear regressions; solid lines represent 95% confidence limits. A: Systolic pressure (top) and diastolic pressure (bottom)B: Mean aortic pressure (top) and pulse pressure (systolic–diastolic pressure amplitude) (bottom). (From Versmold HT, Kitterman JA, Phibbs RH, et al. Aortic blood pressure during the first 12 hours of life in infants with birth weight 610 to 4,220 grams. Pediatrics. 1981;67:611, with permission of American Academy of Pediatrics.)

  1. Complications (Table 8.3)
  2. Defective transducer
  3. Cracked Luer-lock connections, causing leaks, low pressure readings, or blood to back up in the line
  4. Air in the line
  5. Malfunctioning infusion pump not providing continuous flush, causing the line to clot off



  1. Defective reusable transducer interface cable (disposable transducer system)
  2. Erroneous readings caused by the transducer not being properly set at the patient's reference level or if it slips from the preset level

Lower readings occur when the transducer is high; higher readings occur when the transducer is low.

  1. Problems associated with catheters
  2. Tip of the catheter lodging against the wall of the vessel (will cause the pressure wave to flatten and the pressure to rise slowly as a result of the continuous infusion)
  3. Transducer not zeroed to atmosphere (static pressure trapped by stopcock valve and a syringe stuck in the port that should be opened to air)

Will cause lower or negative pressure readings

  1. Loss of blood if stopcock is left opened to the patient and air
  2. Fluid overload if a pressurized intravenous bag is used instead of an infusion pump and the fast flush mode is used to clear the line (11)

TABLE 8.3 Trouble-shooting for Intravascular Pressure Monitoring





Damped pressure tracing

Catheter tip against vessel wall

Usually unavoidable

Reposition catheter while observing waveform


Partial occlusion of tip by clot

Use continuous infusion of 5% dextrose in water or add heparin (1 U/mL IV fluid)

Aspirate clot with syringe and flush with heparinized saline


Clotting in stopcock or transducer, or blood in system

Flush catheter carefully after blood withdrawal and reestablish IV drip; back-flush stopcocks to remove blood

Flush stopcock and transducer; if no improvement, change components

Abnormally high or low readings

Change in transducer level

Maintain patient in same position for serial pressure measurements

Recheck patient and transducer positions


Leaks in transducer system

Assemble transducer carefully, ensuring that dome is attached snugly; use Luer-lock fittings and disposable stopcocks

Check all fittings, transducer dome, and stopcock connections


External vascular compression

Secure catheter firmly without putting tape circumferentially on extremity

Loosen tape, securing catheter in place


Strained transducer

Attention to stopcocks when aspirating to module

Replace transducer


Loose dome on transducer producing low blood pressure without leak or change in waveforma

Attach transducer dome securely to module

Vent system and retighten dome


High intrathoracic pressure secondary to ventilation; particularly influences central venous pressure

Be aware of problem

Take pressure readings off ventilation or in expiratory phase

Damped pressure without improvement after flushing

Air bubbles in transducer connector tubing

Flush transducer and tubing carefully when setting up system and attaching to catheter; handle system carefully

Check system, rapid flush, attach syringe to transducer, and aspirate bubble

No pressure available

Transducer not open to catheter or settings on monitor amplifiers incorrect—still on zero, cal, or off

Follow routine, systematic steps for setting up system and pressure measurements

Check system—stopcocks, monitor, and amplifier setup

aData from Weindling AM. Blood pressure monitoring in the newborn. Arch Dis Child. 1989;64:444.


  1. Nuntnarumit P, Yang W, Bada-Ellzey HS.Blood pressure measurements in the newborn. Clin Perinatol. 1999;26(4): 981.
  2. Bickley LS, Hoekelman RA.In: Bickley LS, ed. Bates' Guide to Physical Examination and History Taking. 7th ed. Philadelphia: Lippincott Williams & Wilkins; 1999.
  3. Moss AJ.Indirect methods of blood pressure measurement. Pediatr Clin North Am. 1978;25:3.
  4. Kirkendall WM, Feinleib M, Freis ED, et al. Recommendation for human blood pressure determination by sphygmomanometers: Subcommittee on the American Heart Association Postgraduate Education Committee. Circulation.1980;62:1145A.
  5. Sadove MS, Schmidt G, Wu H-H, et al. Indirect blood pressure measurement in infants: a comparison of four methods in four limbs.Anesth Analg.1973;52:682.
  6. Ramsey M III.Automatic oscillometric noninvasive blood pressure: theory and practice. In: Meyer-Sabellak W, ed. Blood Pressure Measurements. Darmstadt: Steinkopff Verlag; 1990:15.



  1. Kimble KJ, Darnall RA Jr, Yelderman M, et al. An automated oscillometric technique for estimating mean arterial pressure in critically ill newborns. Anesthesiology.1981; 54:423.
  2. Friesen RH, Lichtor JL.Indirect measurement of blood pressure in neonates and infants utilizing an automatic noninvasive oscillometric monitor. Anesth Analg. 1981;60:742.
  3. Park MK, Menard SM.Accuracy of blood pressure measurement by the Dinamap monitor in infants and children. Pediatrics.1987;79:907.
  4. Weindling AM.Blood pressure monitoring in the newborn. Arch Dis Child. 1989;64:444.
  5. Lee J, Rajadurai VS, Tank KW.Blood pressure standards for very low birth weight infants during the first days of life. Arch Dis Child Fetal Neonat Ed. 1999;81:F168.
  6. Morray J, Todd S.A hazard of continuous flush systems for vascular pressure monitoring in infants. Anesthesiology. 1983;38:187.