INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Nitric oxide (NO) and nitrogen dioxide (NO₂) produced by automobiles and industrial combustion are air pollutants found in the ambient atmosphere of large industrial cities (1). Epidemiological studies have suggested that air pollution caused by oxides of nitrogen and ozone is associated with impaired respiratory function and exacerbation of previous respiratory disease. Although largely influenced by outdoor pollution, indoor concentrations of NO and NO₂ also depend on specific domestic sources such as gas stoves and kerosene-fueled space heaters (2). Recently, NO administered by the inhalation route has been shown to improve arterial oxygenation in critically ill patients with acute respiratory distress syndrome (3). In most countries of the European Economic Community, inhaled NO is currently administered in medical and surgical intensive care units (ICUs) and could be an additional source of contamination of the indoor environment. Because of the potential risk of toxicity for medical and paramedical staff working in ICUs, it has been recommended to scavenge exhaust gases coming out from the ventilators of critically ill patients receiving inhaled NO (10). The aim of this study, performed in a Parisian surgical ICU, was to assess whether the therapeutic use of inhaled NO represents a significant source of contamination of the indoor environment in which medical and paramedical staff are working.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

From October 1, 1994 to September 30, 1995, indoor and outdoor concentrations of NO and NO₂ were regularly measured in the ambient atmosphere of the surgical ICU of la Pitié-Salpétrière Hospital in Paris, France. La Pitié-Salpétrière Hospital is a 2,500-bed teaching hospital (University of Paris VI) located in the southeastern part of the city. The surgical ICU is a 14-bed unit with single rooms and open doors. There is no air conditioning in the unit. The ventilation system consists of an air circulator interconnecting the different rooms, with an air intake located inside the internal courtyard of the hospital. The rudimentary system was built in the early 1970s and is supposed to renew the ambient atmosphere of a given room twice an hour. Among the 159 patients admitted during the study period, 103 were mechanically ventilated for more than 3 d (65%) and 72 (45%) fulfilled the criteria for acute respiratory failure (ARF) (11).

NO was released from a tank of nitrogen that had a NO concentration of 900 ppm and a NO₂ concentration < 10 ppm measured using chemiluminescence (Air Liquide, Antony, France). NO was delivered into the inspiratory limb of the ventilator via an electromagnetic flowmeter at continuous flows ranging between 50 and 120 ml/min. Ventilators were free of any scavenging system. A positive response to inhaled NO was defined as an increase in Pa_O2 of at least 40 mm Hg using an inspired oxygen fraction (FI_O2) of 1, and NO was continued until a Pa_O2 of 70 mm Hg could be obtained without NO at an FI_O2 of 0.5. Among the 72 patients with ARF admitted during the study period, 42 were NO responders and received inhaled NO at a mean inspiratory concentration of 4.3 ± 0.7 ppm during a mean period of 7 ± 6 d (range 12 h-34 d). The concentrations of inhaled NO used during the study period were based on three recent studies that demonstrated that the optimal therapeutic concentrations of NO administered to patients with ARF are most often less than 5 ppm (5, 12, 13). In 19 patients, inhaled NO was combined with intravenous almitrine in order to amplify the improvement in arterial oxygenation (13, 14).

During the study period, 100 simultaneous measurements of indoor and outdoor concentrations of NO and NO₂ were performed on 100 different days. Indoor concentrations were measured in the main corridor of the ICU, whereas outdoor concentrations were measured by passing a 1-m long tubing through the windows of the ICU that open onto an inside square of the hospital, located 250 m away from Boulevard de l'Hôpital, one of the main streets of Paris characterized by dense and continuous automobile traffic. In periods during which inhaled NO was administered to one or several patients, NO concentrations measured at the bedside of a patient receiving NO were compared with NO concentrations measured at the opposite end of the main corridor of the ICU. Nineteen of these comparative measurements were performed during the study period. On three different occasions, NO₂ and NO concentrations were measured outside the ICU in the ambient atmosphere of the circular highway of Paris at 6:00 P.M., the rush hour. Measurements were performed on February 8, July 5, and July 20, 1995, with the 2 d in July being considered as "red days" for pollution in Paris. Using the chemiluminescence apparatus plugged into a power generating set, measurements were performed on the sidewalks of the bridge, La Porte Dorée, in presence of backed-up traffic.

Indoor and outdoor concentrations of NO and NO₂ were measured using the same chemiluminescence apparatus used for monitoring NO and NO₂ concentrations in the tracheobronchial tree of patients receiving inhaled NO (NOX 4000, Sérès, Aix-en-Provence, France). Calibrations were performed twice a week using a reference NO reservoir tank of 100 ppb (Air Liquide).

At the end of the study period, atmospheric pressure (in mbar), cloud mass (in % of the sky), and wind speed (in mph), corresponding to the place where comparative measurements of indoor and outdoor concentrations of NO and NO₂ were performed, were obtained from the meterological service of Paris (Airparif, city of Paris, France).

All data are presented as mean ± SD. Unpaired t test was used to evaluate: (1) differences in indoor NO concentrations measured over periods of time with and without NO administration in the ICU; (2) differences in outdoor NO concentrations according to cloud mass (> or < 40% of the sky); and (3) differences in outdoor NO₂ concentrations according to high (> 1,020 mbar) or low to normal atmospheric pressure (< 1,020 mbar). A one-way analysis of variance was used to evaluate differences in outdoor NO concentrations according to three different speeds of wind (< 5 mph, 5-15 mph, and > 15 mph). Linear regression analysis was used to compare indoor concentrations of NO₂ and NO, and indoor and outdoor concentrations of NO. Relationships between outdoor concentrations of NO and atmospheric pressure were analyzed using a curve fitter program (Slide Write Plus version 6.0, Carlsbad, NM). Factors influencing indoor NO and NO₂ concentrations were determined using a stepwise logistic regression analysis. For indoor NO concentrations, the following factors were analyzed: (1) administration of inhaled NO to one or several patients; (2) outdoor NO concentration; and (3) atmospheric pressure. For indoor NO₂ concentrations, the following factors were analyzed: (1) administration of inhaled NO to one or several patients; (2) outdoor NO₂ and NO concentrations; and (3) atmospheric pressure.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The stepwise logistic regression analysis identified increased outdoor NO concentrations as an independent factor associated with an increase in indoor NO concentrations. As shown in Figure 1, administration of inhaled NO to critically ill patients did not affect ambient concentrations of NO within the ICU: 247 ± 137 ppb during periods of NO administration versus 289 ± 147 ppb during periods without NO administration. During periods of NO administration, no statistically significant difference could be found between NO concentrations measured within the ambient atmosphere of the room where NO was administered and NO concentrations measured within the ambient atmosphere of the main corridor of the ICU at a distance of 20 m away from the site of NO administration. Whether inhaled NO was administered or not in the surgical ICU, a significant relationship was found between indoor and outdoor concentrations of NO, suggesting that NO concentrations measured inside the ICU were predominantly influenced by outdoor air pollution (Figure 2).

View larger version (32K): [in this window] [in a new window]   Figure 1.   Distribution of indoor NO concentrations in the surgical intensive care unit of la Pitié Hospital during 52 d without therapeutic NO administration and during 48 d where one or several patients received inhaled NO at inspiratory concentrations  5 ppm. Each bar represents one measurement performed during 1 d; ppb = parts per billion.

View larger version (16K): [in this window] [in a new window]   Figure 2.   Relationships between indoor (NOIN) and outdoor NO concentrations (NOOUT) measured inside and outside the surgical intensive care unit during 100 different days (ppb = parts per billion; dashed line = identity line; dotted line = regressions line; r = coefficient correlation).

As shown in Figure 3, an exponential relationship was found between indoor NO concentrations and atmospheric pressure. As shown in Figure 4, a significant decrease in outdoor NO concentration was observed during windy periods. Similarly, the presence of cloudy weather was associated with lower concentrations of NO: 159 ± 140 ppb during periods with cloud mass > 40% versus 273 ± 166 ppb during periods with cloud mass < 40%, p < 0.01. Ambient concentrations of NO measured along the circular highway of Paris at rush hour and in the presence of backed-up traffic are shown in Table 1. Although very similar conditions of automobile traffic were observed during the 3 d when measurements were performed, NO and NO₂ concentrations appeared markedly influenced by climatic conditions.

View larger version (17K): [in this window] [in a new window]   Figure 3.   Relationships between indoor NO concentrations (NOIN) and atmospheric pressures (Patm) measured during 100 different days between October 1994 and September 1995. The lowest atmospheric pressure was measured at 998 mb in December 1994 (Patm = 0) and the highest atmospheric pressure was measured at 1,042 mb in July 1995 (Patm = 40 mb) (ppb = parts per billion).

View larger version (16K): [in this window] [in a new window]   Figure 4.   Values of outdoor concentrations of NO measured outside the surgical intensive care unit (NOOUT) for 3 different wind speeds: < 5 mph, > 15 mph, and between 5 and 15 mph. NOOUT is inversely proportional to the speed of the wind (p < 0.05).

The stepwise logistic regression analysis identified elevated outdoor NO₂ concentrations as an independent factor associated with an increase in indoor NO₂ concentrations. As shown in Figure 5, outdoor NO₂ concentrations (NO_2OUT) significantly increased when atmospheric pressures above 1,020 mbar were observed. Outdoor NO₂ concentrations significantly decreased in the presence of cloudy and windy weather and were not dependent on outdoor NO concentrations. As shown in Table 1, concentrations of NO₂ measured along the circular highway of Paris nearly reached 1 ppm on July 20, 1995. 

View larger version (17K): [in this window] [in a new window]   Figure 5.   Values of outdoor concentrations of nitrogen dioxide measured outside the surgical intensive care unit (NO2OUT) for normal and low atmospheric pressures ( 1,020 mb) and high atmospheric pressures (> 1,020 mb). NO2OUT is significantly lower at normal or low atmospheric pressures than at high atmospheric pressures (p = 0.04).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In a Parisian intensive care unit where critically ill patients with acute respiratory failure are routinely treated with inhaled NO, we found that indoor NO and NO₂ concentrations were mainly influenced by outdoor pollution and did not increase with therapeutic NO administration. Elevated indoor NO and NO₂ concentrations were almost always associated with climatic factors such as high atmospheric pressure, absence of wind, and clear sky and never with administration of NO to one or several patients in the ICU.

Before extending these results to all critical care units, some specific characteristics of the ICU where the study was performed should be outlined: First, its geographic situation within the extended urban area of Paris characterized by a population of around 12 million inhabitants, a dense industrial network, heavy and quasi-continuous automobile traffic, and therefore, a high level of outdoor pollution. In this context, NO administered by the inhalation route to mechanically ventilated patients did not appear to be a significant additional exogenous source of NO and NO₂ adding to nitrogen oxides produced by industry, gasoline engines, and domestic devices such a gas stoves, kerosene-fueled space heaters, cooking ranges, and pilot lights (15). Second, the data were obtained in an ICU where several prospective studies on NO dose response clearly demonstrated that the minimum inspiratory NO concentration required for improving gas exchange in critically ill adult patients is generally less than 5 ppm (5, 12, 13). Therefore, NO was administered at inspiratory concentration 5 ppm in contrast to many ICUs where, during the same period, higher doses ranging between 10 and 40 ppm were routinely administered. If exhaled NO coming out from the ventilator is a significant source of indoor NO pollution, then the expired concentration should play a determinant role. In patients with acute respiratory failure, pulmonary absorption of NO is reduced because the alveolar surface available for gas exchange is markedly decreased. In healthy humans, expired NO represents 2-30% of the inspired concentration, whereas in patients with ARF, expired NO represents 40-60% of the inspired concentration (13). In our surgical ICU with the same system of administration, NO concentrations of 2.8, 6.8, and 17 ppm were found in expired gases of patients with acute respiratory failure receiving inspiratory NO concentrations of 4.5, 15, and 45 ppm, respectively (13). Our results may not apply to ICUs where higher concentrations of NO are routinely administered (3, 7, 8, 16, 17) and/or located far away from urban polluted areas. The mode of NO administration may also influence contamination of the ambient atmosphere of the ICU. Continuous administration of NO during the entire respiratory cycle is likely to produce more contamination of the environment than sequential administration limited to the inspiratory phase. In pediatric ICUs, where pressure generator ventilators based on continuous high flow sweeping the respiratory tubings are used, the NO concentration of the exhaust gas is very close to the inspiratory concentration. Because higher doses of NO are used in infants and neonates (16, 17), the contamination of the ambient atmosphere of the ICU may be significant. Only studies measuring indoor concentrations of NO and NO₂ in the presence or absence of NO administration in pediatric ICUs will indicate whether medical and paramedical teams are exposed to higher ambient concentrations of NO and NO₂. In an adult unit, we never found indoor NO and NO₂ concentrations exceeding 0.7 ppm, which is far away from the accepted limit for NO and NO₂ exposures in the workplace, i.e., 25 ppm over an 8-h period (18). On July 20, 1995 we measured outdoor NO and NO₂ concentrations of 1.7 and 0.995 ppm, respectively, along the circular highway of Paris in the presence of backed-up traffic. In the ICU, corresponding indoor concentrations of NO and NO₂ were found at 0.652 and 0.22 ppm, respectively.

Inhabitants of big cities continuously inhale small quantities of NO and NO₂ issued from outdoor and domestic pollution. Therapeutic administration of NO does not appear to be a significant additional source of nitrogen oxides. Clear and sunny weather characterized by high atmospheric pressure, heavy automobile traffic, and lack of wind movement are generally associated with the highest outdoor and indoor NO concentrations. In an ICU where therapeutic NO concentrations  5 ppm are used, protection of nurses and medical staff against incidental inhalation of nitrogen oxides should start with general attempts at reducing urban pollution rather than with specific scavenging of NO coming out from the ventilators.