Inspiratory Flow Dynamics during Mechanical Ventilation in Patients with Respiratory Failure 1,2

DAVID J. PREZANT,3 THOMAS K. ALDRICH, JILL P. KARPEL, and SUNG S. PARK

Introduction

The human bronchial tree is a complex system of asymmetric tubes that can be subjected to a wide range of flow rates and pressure gradients. The purpose of this paper is to describe resistance-flow relationships in mechanically ventilated acute respiratory failure (ARF) patients over the range of inspiratory flow rates currently in clinical use (0.66 to 2 Lis). Our analysis is based on the measurement of respiratory resistance under conditions of constant flow inflation (1,2). ARF patients with acute respiratory failure from obstructive airway disease (OAD) and status asthmaticus are analyzed separately to determine whether pre-existing airway flow limitations influence the resistance-flow relationship. If the human bronchial tree behaves as a laminar flow model, then varying the inspiratory flow rates will not affect respiratory resistance; the pressure-flow relationship will be linear, and the resistance-flow relationship will have a slope of zero. Alternatively, an increase in the rate of constant flow inflation may be expected to increase airflow resistance due to turbulence. If inspiratory airflow follows a turbulent flow model, then the pressure-flow relationship will be curvilinear and the resistance-flow relationship will be linear with a positive slope. Significant turbulence may be especially prominent in obstructive airway diseases. However, because the bronchial tree is not a rigid tube system, inspiratory flow dynamics should also be influenced by nonaerodynamic variables, such as tissue resistance, stress relaxation, and possibly airway dilatation from the increasing airway pressures that necessarily accompany rapid flow rates. Each of these factors would produce a flat or even negative resistance-flow slope. Methods We evaluated 15 patients who required mechanical ventilation. Exclusion criteria included ventilator dysynchrony, tachypnea (> 20 breaths/min), excessive secretions, or the clin1284

SUMMARY We studied the effect of Inspiratory flow rate on respiratory resistance during mechanical ventilation In 15 patients with acute respiratory failure (ARF). Resistance was measured by both constant flow Inflation and occlusion methods as Inspiratory flow rates ware increased from 0.66 to 2.0 Us. Endotracheal tube resistance was subtracted from total resistance to obtain respiratory resistance. in contrast to the flow-dependent Increase In endotracheal tube resistance, respiratory resistance decreased continuously as flow rate and airway pressure Increased, except In four of six patients with asthma In whom respiratory resistance Increased as flow Increased. Mechanical airway dilatation, tissue resistance, stress relaxation, and tlme-constant inequalities may contribute to the decrease In respiratory resistance. In status asthmatlcus, however, the effects of turbulence, noncompliant airways, and/or "reflex" bronchoconstrlctlon may be sufficient to cause a flowAM REV RESPIR DIS 1990; 142:1284-1287 depandent Increase in resistance.

ical use of positive end-expiratory pressure (PEEP), intermittent mandatory ventilation, or neuromuscular blocking agents. Most patients were already sedated, but all had spontaneous respiratory efforts. Patients were ventilated with constant-flow, volume-cycled respirators. Tidal volume (8 to 14 ml/kg) and respiratory rate (12 to 16 breaths/min) were selected by the patient's own physician. Our protocol was to vary inspiratory flow rates, in approximately O.l7-L/s increments, across a range from 0.66 to 2.0 L/s. Because of the ventilator's mechanical limitations, the measured flow rate was often lower than the machine's flow rate setting. In each case only the measured flow rate was used for resistance calculations. After 5 min at a given inspiratory flow, airway pressure and inspiratory flow rate were measured proximal to the endotracheal tube during a ventilator-initiated, nonassisted breath using a differential pressure transducer (Model MP45-1; Validyne Corp., Northridge, CA), heated pneumotachometer (Collins, Braintree, MA), and oscillographic recorder (Model VR6; Electronics for Medicine, Westchester, NY). Tidal volume was quantified by digital planimetry (Numonics Corp., Lansdale, PA) of the inspiratory flow tracing. At peak inspiratory pressure (PIP) the expiratory port was occluded for 1.5 s, or until a plateau was reached, and end-inspiratory plateau pressure (EIPP) was measured. Both a stable EIPP, representing elastic recoil pressure of the total respiratory system (1), and a constant slope of the inspiratory pressure ramp during constant flow inflation were evidence for adequate relaxation of the respiratory muscles (2). Inflation pressure (lP) was measured by extrapolating the slope of the constant-flow inflation pressure recording

back to the point at which inspiration was initiated. At each flow rate all pressures were corrected for the level of auto-PEEP as previously described (3, 4). The resistance of the total system (equipment and patient) was determined at constant inspiratory flow by both the occlusion (Ro) and inflation (Ri) methods (1, 2). Inflation resistance was determined by dividing IP by the measured flow rate. The occlusion resistance was the difference between PIP and EIPP divided by flow. Resistance was measured by both inflation and occlusion methods to verify that our overall conclusions were not influenced by measurement technique. Resistive-flow relationships were analyzed separately for each patient by linear regression: R = K, + K2 V, where V is flow and K, and K2 (Rohrer's resistance-flow constants) are the y intercept and slope, respectively (5). 'IWo components contributed to the flow resistance of the respiratory equipment: the pneumotachometer, which displayed a nearly constant resistance of 0.99 cm H 2 0/L/s, and the endotracheal tubes, which displayed

(Received in original form April 25, 1990 and in revised form May 17, 1990) 1 From the Department of Medicine, Pulmonary Division, Montefiore Medical Center, and the Department of Physiology and Biophysics, Albert Einstein College of Medicine, Bronx, New York. 2 Correspondence and requests for reprints should be addressed to David J. Prezant, M.D., Montefiore Medical Center Pulmonary Division, Centennial Building, 4th Floor, Bronx, NY 10467. 3 Recipient of Clinical Investigator Award No. lK08-HL0216S-Q1 from the National Heart, Lung, and Blood Institute.

1285

EFFECT OF INSPIRATORY FLOW RATE ON RESISTANCE

20 lG

'0

18

4'

17

40

18

"

I~ U>

:J 0

'"

I

E .!!o c: Ui

I. 13

.. ... ... -.

'0

08

x x

12

...'"

"

a:

x

•

x x x

6 6

~

6

•

0

3

+ c

2

•

0

• 0+

+ a

6

0

+ + a a

0

0

0

+

•

0

6

•

•

• •

x x

8 8 7 8

X

x

'.2

11 ID

"

Fig. 1. Resistance-flow relationships for endotracheal tubes 24 cm long, (ex marks) 7, (triangles) 7.5, (diamonds) 8, (plus signs) 8.5, and (open squares) 9 mm 10, as measured when connected t,) an artificial trachea (12 cm long, 2.5 cm 10). As tube size increased from 7 to 9 mm 10, K, was equal to 1.14,0.71, 0.77,0.54, and 0.59 cm H2 0/Us; K2 was equal to 8.73, 6.44, 4.76, 3.62, and 2.95 cm H20/L21s2 • For all endotracheal tubes studied the pressure·flow relationship was curvilinear, the resistance-flow relationship was linear, and resistance was inversely related to tube 10. When a Iinear resistor was connected to the endotracheal tube-(circles) 7 mm 10, (closed squares) 9 mm 10-resistance remained essentially constant as flow increased.

000°0°0°0

+

+

+

0

0

0

+ 0

0

I D D

u

D.8

1.2

1.8

2

Flow (Liters/second)

K, and K2 for the total model system were equal to the sum of the individual components. Resistance was also measured with the endotracheal tube-artificial trachea model connected in series to a linear resistor (8) equal to the highest resistance found in our patients (figure 1). Statistical analysis was performed by linear regression on experimental values relating resistance to flow for each individual patient. All values are reported as the mean ± standard error of the mean (SEM). Comparison between groups was performed using an unpaired t test (p < 0.05).

curvilinear pressure-flow relationships or linear resistance-flow relationships with slopes dependent on the tube's internal diameter (ID). Pressure gradients were measured in five endotracheal tubes, 24 cm long, 7 to 9 mm ID (figure 1). To account for kinetic energy losses at the transition from tube to trachea, this measurement was made with the endotracheal tube inserted (cuff inflated) into an artificial trachea constructed from ventilator tubing (Airlife, Valencia, CA) 12 cm long, 2.5 cm ID (6, 7). The Rohrer constants K, and K 2 were derived by linear regression analysis on experimental values of resistance versus flow (figure 1). The resistance of the respiratory equipment was then subtracted from the total system to yield respiratory resistance. In all instances the linear regression constants

Results

Over the entire range of flow rates mea-

TABLE 1 K, AND K, VALUES Total Resistance Patient Non-OAD 1

2 3 4 5 6 Non-OAD Mean ± SEM OAD-asthma

7 8 9

10 11

12

Tube

Vt(ml)

Rate

8.0 8.0 8.0 8.0 7.0 8.0

850 800 750 800 500 450

14 14 16 15 15 12

Respiratory Resistance

K,

K,

K,

K,

26.77 21.92 6.25 14.14 24.36 7.35

-2.11 -0.55 5.40 0.95 4.78 1.70

26.00 21.15 5.48 13.37 23.22 6.58

-6.87 -5.31 0.64 -3.81 -3.95 -3.06

16.80 3.29

1.69 1.10

15.97 3.27

-3.73 0.94

0.945 0.949 0.306 0.943 0.800 0.699

7.5 7.5 8.0 8.5 8.5 8.5

550 700 700 600 600 600

14 12 12 12 12 10

-1.56 4.04 34.37 -3.40 10.12 21.98

12.97 17.52 -3.87 39.02 5.74 0.88

-2.27 3.33 33.60 -4.11 9.58 21.44

6.53 11.09 -8.63 32.58 2.12 -2.74

0.770 0.838 0.906 0.705 0.746 0.872

7.5 8.0 8.0

700 800 500

15 14 12

28.48 16.23 15.19

2.37 3.00 0.97

27.77 15.46 14.43

-4.06 -1.76 -5.73

0.602 0.476 0.983

13.94 4.09

8.73 4.12

13.25 4.09

3.27 3.96

OAD-COPD

13 14 15 OAD Mean ± SEM

Definition of abbreviations: K, = em H,O/Us; K, = em H,O/L'/s'.

sured (0.64 to 2.35 Lis), the mean PIP increased by 97 ± 9010 but the mean EIPP increased by only 0.3 ± 3.3% (NS). Stability in both the pressure ramp tracing and the EIPP with occlusion demonstrates that our patients were adequately synchronized with the mechanical ventilator during the course of this study (1, 2). Tidal volume and auto-PEEP also remained relatively constant, and therefore respiratory compliance did not significantly change with increasing flow rates. Of 15 patients, 10 (Patients 7 through 15), had ARF resulting from OAD. Correlation between R i and Ro measurements is shown by the following regression equations: Ri

= 0.954Ro - 0.756 for all patients.

r

= 0.793

Ri

= 0.990Ro - 0.834 for OAD patients.

r

=

0.845

Differences between Ri and Ro were present, mostly due to the variability in Ri that results from extrapolation of IP from the slope of the pressure recording during constant-flow inflation. In contrast, R o was highly reproducible. As the outcome of our analysis remained the same regardless of which resistance measurement technique was used, we chose R o measurements for this presentation. For each patient Rohrer's K I and K 2 constants for the total system and respiratory system are listed in table 1. The scatter in K, reflects the wide range of respiratory resistance found in ARF patients. The slope (K2 ) of the total system's flow-resistance relationship averaged 1.69 ± 1.10 for patients without OAD. After subtracting endotracheal tube resistance, respiratory K 2 averaged - 3.73 ± 0.94. The negative slope of the respiratory resistans;e;;'flow relationship for non-OAD patients (figure 2A) was in sharp contrast to the positive slope of the endotracheal tube resistance-flow relationship, and because of these opposing forces the total system resistance remained relatively constant as flow increased from 0.66 to 2.0 Lis. We separately analyzed patients with acute respiratory failure as a result of OAD (table 1). Patients with chronic obstructive COPD and status asthmaticus are shown in figure 2B and C, respectively. After correction for the endotracheal tube, respiratory K2 averaged 3.27 :c 3.96. However, evaluation of each patient individually revealed that a positive respiratory K2 constant, reflecting increasing resistance over this range of inspiratory flow rates, was present in only four of

PREZANT, ALDRICH, KARPEL, AND PARK

1286 B

A 30 2G 22 --'

i r

B 0