Sustained Improvement in Gas Exchange After Negative Pressure Ventilation for 8 Hours Per Day on 2 Successive Days in Chronic Airflow Limitation 1 , 2
ENRIQUE FERNANDEZ, PALTIEL WEINER, EPHRAIM MELTZER, MARY M. WTZ, DAVID B. BADISH, and REUBEN M. CHERNIACK
Introduction regular application of negative pressure ventilation (NPV) on a daily basis has been shown to result in improvement in arterial blood gases and symptomatology in patients with chronic respiratory insufficiency secondary to neuromuscular disease (1) or kyphoscoliosis (2-4) and hypoventilation syndromes (5). Similarly, daily NPV has been reported to improve gas exchange in patients with chronic airflow limitation (6-13). Improvement in arterial blood gases was seen after short periods of artificial hyperventilation (14)or mechanical ventilation for 3 to 6 h/day for 3 days (10), for 4 to 10 h daily (9, 11),and after weekly administration for 8 h/day for 7 wk (12), but the duration of improvement has not been well studied. In most studies, the improvement in gas exchange has been accompanied by an increase in respiratory strength, so that the improvement has been attributed to rest of the respiratory muscles. On the other hand, several investigators have reported a lack of significant improvement in gas exchange or respiratory muscle strength after mechanical ventilation (15-18). The purpose of this article is to report studies indicating that the application of NPV for 8 h/day on 2 consecutive days in patients with chronic airflow limitation who are in respiratory failure would lead to sustained improvement in gas exchange. In addition, studies are reported attempting to elucidate the mechanisms underlying the sustained improvement.
Methods NPV was applied for 8 h/day for 2 consecutive days in 13 patients with severe chronic airflow limitation who were clinically stable and had been on the same medication schedule for a least 3 months (table 1). Informed consent was obtained from the subjects, and the protocol was reviewed and approved by 390
SUMMARY Negative pressure ventilation (NPV)was applied for 6 to 8 h/day for 2 consecutive days In 13 patients with severe airflow limitation and chronic respiratory failure. After cessation of NPV, the mean arterial blood gases were Improved In 10 patients, and this Improvement was sustained for the next 2 days In eight patients, for 3 days In seven patients, and was stili present In four patients on the fourth day. Respiratory muscle strength Improved In all patients, but there was no relationship between the Increase In strength and sustained Improvement In gas exchange. Ventilation and respiratory pattern were unchanged In all patients, but the mean VDNT fell and VA rose while the V0 2 and VC0 2 fell. The ventilatory responses to hypoxia and hypercapnia Increased In patients who demonstrated sustained Improvement In blood gases. The mechanism underlying the sustained Improvement In gas exchange following NPV Is not clear but Is likely multifactorial. AM REV RESPIR DIS 1991; 144:390-394
the Institutional Review Board of the National Jewish Center. As is shown in table 1, the mean FEV 1 (± standard error of the mean [SEMD was 20.8 ± 1.5070 of predicted; the mean volume of thoracic gas (Vtg) was 205.1 ± 8.7% of predicted, and mean inspiratory muscle strength (PImax) was 43.2 ± 2.7 em HzO. The mean ± SEM arterial oxygen and carbon dioxide tensions while breathing an enriched gas mixture (PIoz = 0.28) on 3 consecutive days before initiation of NPV were 61.8 ± 3.2 and 55.8 ± 1.8 mm Hg, respectively. All patients were admitted to the hospital for the application of NPV and assessment of respiratory function and gas exchange before and after the NPV. NPV was applied utilizing a body pneumowrap ventilator (either a Thompson MV Maxivent'" [PuritanBennett Corp., Boulder, CO] or an Emerson Model 33-CRE [J. H. Emerson Co., Cambridge, MAD while in the recumbent position (about 30° from the horizontal plane). The patients were first acclimatized to the ventilator for 2 to 4 h on 1 day, and NPV was then instituted for 8 h/day on the following 2 days. The ventilator was set to deliver a negative pressure of - 20 to - 35em HzO, with a tidal volume 100to 200 ml greater and a respiratory frequency 2 to 4 cycles/min higher than those of the patient. Arterial blood gases were measured 30 min after the initiation of NPV to ensure good oxygenation without marked respiratory alkalosis, and oxygen saturation was monitored continuously with a Biox~ oximeter (Bioxymetry Technology, Inc., Boulder, CO) to ensure a saturation above 92%
throughout. A fall in the end-tidal Pco, of more than 4 mm Hg during NPV and elimination of diaphragm surface electromyographic electrical activity, determined by the PN22-powered resting unit (TECA Corp., Pleasantville, NJ) for at least 85% of the breaths, were considered indications of control of ventilation by NPV and rest of the inspiratory muscles. Thoracic gas volume, FVC, and FEV 1 were performed while the subjects were seated in a volumedisplacement body plethysmograph. The mean of three measurements of Vtg and the maximum of at least three determinations of FVCand FEV 1 wereutilized and compared to expected values (19).The diffusing capacity for carbon monoxide (DLcO) was determined by the single-breath technique (20) and compared with predicted values (21).
Gas Exchange Gas exchange was assessed for 3 consecutive days before and for 4 days after termination (Received in original form June 29, 1990 and in revised form January 11, 1991) 1 From the Division of Pulmonary Sciences, Departments of Medicine, University of Colorado Health Sciences Center and National Jewish Center for Immunology and RespiratoryMedicine,Denver, Colorado. 2 Correspondence and requests for reprints should be addressed to Enrique Fernandez, M.D., Department of Medicine, National Jewish Center for Immunology and Respiratory Medicine, 1400 Jackson Street, Denver, CO 80206.
NPV AND SUSTAINED IMPROVEMENT IN BLOOD GASESIN CHRONIC AIRFLOW LIMITATION
TABLE 1 CHARACTERISTICS AND RESPIRATORY FUNCTION OF SUBJECTS Subject (No./Sex)
11M 21M 31M 4/M 5/F 6/M 7/M 8/F 9/F 10/F 111M 121M 131M Mean ± SEM
OLeo (% of predicted)
FEV, (% of predicted)
FEV,/FVC (% of predicted)
64 66 58 57 64 67 46 57 70 55 67 55 52
97.1 63.5 55.3 68 52.2 60.8 68 65.5 72.7 55.2 62.3 74.8 66.8
2.10 1.74 1.68 1.83 1.57 1.71 1.73 1.65 1.74 1.57 2.76 1.90 1.77
233 165 210 208 192 215 142 249 214 220 240 166 212
30 14 15 42 41 25 76 52 46 40 48 30 28
14 21 25 14 27 24 17 23 16 33 18 17 21
73 38 30 57 55 50 59 43 48 24 33 33
51 59 60 48 63 62 50 53 88 69 60 79 61
± ± ± ± ± ± ± ± ± ± ± ± ±
0.66 0.35 1.10 0.37 0.75 1.10 0.87 0.57 0.57 1.00 0.33 0.35 0.33
64 63 53 56 54 48 65 53 52 44 64 55 54
± ± ± ± ± ± ± ± ± ± ± ± ±
0.93 1.10 0.58 0.83 1.10 0.69 1.30 0.33 1.70 0.57 0.33 0.35 0.30
37 45 48 62
40 57 45 35 40 48 38 44 23 43.2 2.7
Definitionof abbreviation: BSA = body surface area. • Mean ± SEM for 3 consecutive days while breathing an F102 of 0.28.
of NPV. On each day, the arterial blood and expired gas were collected while breathing an enriched gas mixture (FI02 = 0.28) after the subject had fasted for at least 8 h and had been seated in a comfortable chair for 30min. Tidal volume (VT), respiratory frequency (f), dead space (Vn), dead space/tidal volume ratio (VO/VT), alveolar ventilation (VA), oxygen consumption (V02), CO 2 production (Vc02), respiratory quotient (R), and the alveolar-arterial oxygen tension gradient (AaP02) were calculated. Respiratory Muscle Function Inspiratory muscle strength was assessed in all patients before and 2 days after NPV utilizing the technique of Black and Hyatt (22). The greatest of at least three measurements of maximal inspiratory pressure determined at residual volume was reported. The efficiency of the respiratory muscles for handling extra work was assessed in six patients by measuring the additional oxygen consumption associated with inspiring against incremental inspiratory threshold loads (23-26) using a modification of the apparatus described by Nickerson and Keers (27). Studies werecarried out after the subjects had fasted for at least 8 h and been seated in a comfortable chair for 30 min. Subjects breathed through the apparatus for 15 min, and VT, minute ventilation (VE), and mean expired O2concentration were determined for 30 severy 3 min during the final 9 min. The mean of the three measurements was reported as the resting value. Inspiratory work was then increased incrementally by the progressive addition of 25 to 100g weights to the plunger at 2-min intervals, the end-tidal CO 2 concentration being kept constant by the addition ofCO 2as necessary to the inspired air. Ventilation and oxygen consumption were determined over the last 30 s while inspiring against each of the added loads. In all subjects, measurements were carried out at four to six work loads that
generated peak mouth pressures between 20 and 60010 of Pl max, where efficiency has been shown to be relatively constant (23). The amount of added mechanical work performed while inspiring against each load was calculated by multiplying the VE (liters) by the increment in mouth pressure (em H 20) and was expressed as kilogram/meters per minute. The additional oxygen consumption associated with each work load was calculated by determining the difference between the oxygen consumption at rest and that while breathing against the added load. This was converted to its energy equivalent (assuming a respiratory quotient of 0.82) and expressed as kilogram/meters per minute. The efficiency of the respiratory muscles for handling the added work was calculated by determining the ratio ofthe added mechanical work and the added energy requirement. The mean of the four to six determinations was reported in each subject.
Respiratory Responses to Hypoxia and Hypercapnia The respiratory responses to carbon dioxide and hypoxia were assessed in seven patients before and 2 days after NPV. For both determinations, subjects breathed through a pneumotachograph (Fleisch No.3; Fleisch, Lausanne, Switzerland) connected to a lowresistance (1.3 em H 20/L/s) rebreathing circuit partitioned into separate inspiratory and expiratory lines. A silent vibration-free balloon apparatus located in the inspiratory line was inflated at random intervals (approximately every eight breaths) during expiration and the airway occlusion pressure at 100 ms of the succeeding inspiration (P O•I ) determined. Pressure was sensed by a differential pressure transducer Model PM 5 E (Statham Instruments, Hato Rey, PR). Volume was derived by integration of the airflow signal, and VE was calculated from the three breaths preceding each occlusion.
The response to carbon dioxide was assessed while progressive normoxic hypercapnia was induced by the method of Read (28). The subjects rebreathed for 4 to 6 min from a bag that initially contained a mixture of 6010 CO 2, 40010 O 2, and 54010 N 2. Ear oxygen saturation (Sa02) was monitored continuously with a Bioxill II-A oximeter (Bioxymetry) and maintained at greater than 92010 throughout the determination. The response to hypoxemia was assessed while progressive isocapnic hypoxemia was induced by rebreathing from a bag that initially contained a volume of room air that was 1 L greater than the patient's vital capacity. End-tidal CO 2 was monitored continuously using a capnometer Model 2200 (Sensors, Inc., Saline, MI) and maintained within ± 2 mm Hg by means of a CO 2 absorber attached to the expiratory limb of the rebreathing circuit.
Data Analysis Data for each variable was analyzed over time by a repeated-measures analysis of variance. When the time effect between baseline and follow-up measurements was significant a Dunnett's multiple-comparison procedure at a significance level of 0.05 was used to compare the means at each postintervention time to the preintervention data. For all the variablesthat weremeasured once before and once after the application of NPV, data were compared using the paired Student's t test. Differences were considered significant if the p value was < 0.05.
Results During the application of NPV, the mean Pacoz fell from 55.8 ± 1.8 (mean ± SEM) to 47.4 ± 1.7 mm Hg (p < 0.001) and the mean Pao1 rose from 61.8 ± 3.2 to 70.9 ± 2.8 mm Hg (p < 0.001). In addition, the diaphragm surface electrical
FERNANDEZ, WEINER, MELTZER, WTZ, BADISH, AND CHERNlACK
392 TABLE 2 EFFECT OF NPV ON GAS EXCHANGE (MEAN ± SEM)· After NPV Parameter Pao2, mm Hg Paco 2, mm Hg AaPo2, mm Hg
61.7 ± 3.2 55.8 ± 1.8 36.56 ± 2.30
66.2 ± 3.2t 50.1 ± 1.6t 36.38 ± 2.28
65.6 ± 2.6t 51.0 ± 1.4t 36.25 ± 2.16
65.8 ± 2.7t 51.5 ± 1.4t 36.82 ± 2.11
61.3 ± 3.5 53.1 ± 2.1t 37.56 ± 1.69
• While breathing an F102 of 0.28. Significant using Dunnett's multiple-comparison procedure at a significance level of 0.05.
cantly (p < 0.05) after NPV, as indicated in table 5. However, the responses were once again variable, the P O• I response to CO 2increased in all subjects and the ventilatory response fell in Patients 1 and 8. The P 0.1 response to hypoxia increased in all but Patient 8, and the ventilatory response to hypoxia increased in all but Patient 10.
activity disappeared for at least 850/0 of the breaths in all patients. Of note, most of the patients fell asleep as soon as NPV was instituted. Only one patient developed obvious clinical upper airway obstruction during NPV, and this was alleviated by reducing the amount of negative pressure applied and increasing the rate of the ventilator. Most patients claimed a reduction in dyspnea after cessation of NPV. The changes in arterial blood gases over the 4-day period are shown in table 2. The mean Pao, and Pac02 were significantly improved for the 3 days after cessation of NPV (p = 0.(01), but only the Pac02was significantly different from the preintervention days on the fourth day. There was no change in the AaP02 gradient during this time period. The change in Pac02 following NPV in each of the 13patients is shown in figure 1. It can be seen that the Pac02 was essentially unchanged in three patients (No.3, 7, and 8) and was reduced on the first day following NPV in 10 patients, on the second day in 8 patients, on the third day in 7 patients, and was still sustained on the fourth day in 4 patients. Measurements of ventilatory parameters over the 4-day period are presented in table 3. There was no change in mean minute ventilation, VT, or respiratory fre-
quency on the 4 days after NPV. On the other hand, there was a significant fall in mean VD/VT ratio and increase in alveolar ventilation for 3 days after cessation of NPV. As is also seen, the meanVc02and V02remained significantly lower for the 2 days following cessation of NPV. The effect of NPV on respiratory muscle function is shown in table 4. Respiratory muscle strength, as reflected by the PImax, increased significantly from a mean of 43.15 ± 2.72 to 53.0 ± 3.09 em H 20 (p < 0.001). However, it will be noted that PImax failed to increase in Patients 4, 11, and 12, who demonstrated improvement in gas exchange, and increased in Patients 3 and 7, who failed to demonstrate improvement. Also indicated in table 4 is that the efficiency of the respiratory muscles increased (p < 0.05) after NPV in the patients who were studied. Once again, however, there was no relationship between improvement in respiratory muscle function and sustained improvement in blood gases; the efficiency increased in Patient 8, who failed to demonstrate improvement in gas exchange and was essentially unchanged in Patients 12 and 13, who showed sustained improvement. The respiratory response to carbon dioxide and to hypoxia improved signifi-
Fig. 1. The changesin Paco 2(mm Hg) after cessation of NPV for 8 h/dayon 2 consecutive daysin the 13 patients.The dotted line represents a fall in Paco2 of 2.5 mm Hg.