CJA
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Résumé de cet Article
Right arrow Full Text (PDF)
Right arrow An erratum has been published
Right arrow Submit a scholarly reply
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gagnon, C.
Right arrow Articles by Donati, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gagnon, C.
Right arrow Articles by Donati, F.
Canadian Journal of Anesthesia 53:86-91 (2006)
© Canadian Anesthesiologists' Society, 2006
Cardiothoracic Anesthesia, Respiration and Airway

When a leak is unavoidable, preoxygenation is equally ineffective with vital capacity or tidal volume breathing

[Quand une fuite est inévitable, la préoxygénation n’est pas plus efficace avec des manoeuvres de capacité vitale qu’avec une respiration normale]

Caroline Gagnon, MD, Louis-Philippe Fortier, MSc MD FRCPC and François Donati, PhD MD FRCPC

From the Department of Anesthesiology, Maisonneuve-Rosemont Hospital, University of Montreal, Montreal, Quebec, Canada.

Address correspondence to: Dr. François Donati, Department of Anesthesiology, Maisonneuve-Rosemont Hospital, 5415, l’Assomption blvd, Montreal, Quebec H1T 2M4, Canada. Phone: 514-252-3426; Fax: 514-252-3542; E-mail: francois.donati{at}umontreal.ca


   Abstract
TOP
 Abstract
Introduction
 Methods
 Results
 Discussion
 References
 
Purpose: Ideally, preoxygenation is performed using a tight fitting mask either by breathing normally for three to five minutes or with four to eight vital capacity (VC) breaths in 0.5 to one minute, but in practice leaks are frequent and sometimes unavoidable. This study was designed to determine which breathing method provided the best oxygenation in the presence of leak.

Methods: Twenty volunteers were instructed to breathe from a circle circuit supplied with 6 L·min–1 of fresh oxygen. Each subject was tested under four situations selected in random order: 1) normal breathing for three minutes without leak; 2) normal breathing for three minutes with a leak; 3) four VCs in 30 sec without a leak; and 4) four VCs in 30 sec with a leak. The leak was created by a piece of size 18 French nasogastric tube, 5 cm long, taped under the face mask. Inspired and expired O2 and CO2 were sampled at the nostrils.

Results: In the absence of a leak, the end-tidal oxygen fraction (FEO2) was greater after three minutes of tidal breathing (89 ± 3%; mean ± SD) in comparison with the response to four VCs (76 ± 7%; P < 0.001). Introduction of a leak decreased the FEO2 significantly (P < 0.001). With a leak, the FEO2 was similar with normal breathing (61 ± 8%) and after four VCs (59 ± 11%).

Conclusion: Preoxygenation with tidal volume breathing for three minutes yields higher FEO2 in comparison to four VCs. If a small leak (4 mm internal diameter) is introduced, the FEO2 decreases significantly with both breathing methods to approximately 60%.


   Introduction
TOP
 Abstract
 Introduction
Methods
 Results
 Discussion
 References
 
PREOXYGENATION with 100% oxygen is performed routinely before induction of anesthesia.1,2 Its goal is to increase the body’s oxygen stores by replacing nitrogen in the lungs by an equivalent volume of oxygen, thus delaying the onset of arterial desaturation and hypoxemia during the apneic period following induction of anesthesia.3

Preoxygenation was described and shown to be effective as early as 1955,4 and several techniques have been used since. Most often, the patient is required to breathe 100% oxygen for three to five minutes.2,3,5,6 To shorten preoxygenation time, a technique where the patient is asked to take four vital capacity breaths (VC) in 30 sec has been proposed, and claimed to be as effective as normal breathing for three to five minutes.7,8 However, these preoxygenation techniques require a tightly fitting face mask, and do not account for the leaks that occur frequently in practice.9–12 Leaks have been reported in 10 to 11.5% of subjects with no facial anomalies and normal dentition.13,14 In real life, leaks probably occur even more frequently since we often preoxygenate patients who are edentulous, wear beards or mustaches, have a nasogastric tube in place, or cooperate poorly. When a leak is unavoidable, an imperfect preoxygenation is presumably better than none at all, but it is unclear which breathing method then yields the best results.

A leak is expected to allow air entrainment into the circuit and dilute the oxygen present. The flow through a leak is turbulent and depends upon the square root of the pressure difference across it.15 Therefore, the relative importance of the leak is expected to decrease when a large pressure, created by a VC maneuver, is generated. However, a leak might cause the reservoir bag of the circuit to empty partially, limiting the amount of oxygen available for a VC breath. It is unclear which of these two opposing effects predominates. Therefore, the purpose of the study was to compare the effectiveness of preoxygenation using tidal volume breathing vs VC breathing, with and without a leak.


   Methods
TOP
 Abstract
 Introduction
 Methods
Results
 Discussion
 References
 
The study was approved by the Research Ethics Committee of the Maisonneuve-Rosemont Hospital, and informed consent was obtained from all volunteers prior to their enrollment in the study. Twenty healthy volunteers of both sexes were recruited. They were between the ages of 18 and 65 yr. All subjects were within ± 30% of ideal body weight. None of the subjects had a history of lung disease or any active airway infection. Subjects were not recruited if they had a mustache or a beard, if they were edentulous, or if they had a known allergy to latex.

The same Modulus II Ohmeda (Ohmeda, Madison, WI, USA) anesthesia machine was used throughout the study. The circle system consisted of an absorber with baralyme 1.5-L, two disposable 72-inch corrugated breathing tubes and a 2-L capacity breathing bag. Gases were analyzed with the Siemens MultiGas+ system (Siemens Medical Systems, Danvers, MA, USA) and calibration with known gas mixtures (air, 95% O2 and 5% CO2) was carried out according to the manufacturer’s specifications.

After being informed of the different steps in the study, subjects were asked to lie in a supine position on a stretcher. They were instructed to breathe through their nose at all times. A pulse oximeter was applied to a finger tip to measure oxygen saturation and heart rate. A nasal cannula was fixed at the nostrils to sample expired gases at a rate of 200 mL·min–1. The cannula exited the circuit through a sealed connector between the mask and the circuit and was connected to the gas analyzer (Figure 1Go). Inspired and end-tidal oxygen concentrations were measured. End-tidal carbon dioxide concentrations were also recorded. To create a standardized leak, a 5-cm long piece of #18 nasogastric tube (Levin type) was fixed to the upper crux of the facemask, so it lay on the bridge of the nose. A Rush facemask (#2, #3 or #4) was chosen to achieve the best fit for each subject. Each subject acted as his/her own control by going through each of the different phases in a random order, separated by rest periods of at least five minutes breathing room air.


Figure 1
View larger version (81K):
[in this window]
[in a new window]
  		FIGURE 1 Experimental setup. Expiratory gases were sampled via a nasal cannula (arrow 1). The tubing went through the inside of mask, out through the circuit connector (arrow 2), and was connected to the gas analyzer. On the right, the sealed connection between the gas analyzer and the nasal cannula (arrow 2) is shown. In both pictures the leak (arrow 3) was created with a #18 piece of nasogastric tube, 5 cm in length.

 
Randomization was achieved by the use of sealed envelopes. Before each step, the reservoir bag was fully inflated with 100% oxygen, using the oxygen flush while occluding the mask. The reservoir bag was emptied three times to expel the air and replace it with 100% oxygen. Preoxygenation was performed with 100% oxygen using a fresh gas flow of 6 L·min–1. There were four different techniques, all using 100% oxygen: 1) tidal volume breathing for three minutes with no leak; 2) tidal volume breathing for three minutes with leak; 3) four VC breaths in 30 sec with no leak; 4) four VC breaths in 30 sec with leak.

Data were recorded by taking a picture of the anesthesia monitor with a digital camera every 7.5 sec for the first 30 sec, then every 15 sec for three minutes. The primary outcome measurement in this study was the end-expiratory oxygen fraction (FEO2) reached at the end of each intervention technique. Comparisons were made at the end of the preoxygenation period between the tidal volume breathing method and the four VCs in 30 sec. Similar comparisons were made between end-expiratory values using the same breathing techniques, with or without leaks. A student t test with four Bonferroni corrections for multiple comparisons was used with a P value of < 0.05/4 = 0.0125 indicating a significant difference. Statistical analysis was performed using the S-plus statistical software, version 4.5 (Insightful, Seattle, WA, USA). The number of subjects (20) was chosen to detect a 10% difference in end-tidal oxygen values between techniques, with a two-sided test, assuming a type I ({alpha}) error of 0.05 and a power (ß) of 0.8. A 10% difference was considered to be of clinical significance. The standard deviation with a leak was expected to be 10%, based on the results of McGowan et al.12


   Results
TOP
 Abstract
 Introduction
 Methods
 Results
Discussion
 References
 
All volunteers were non smokers. Their demographic data are presented in Table IGo. Each volunteer completed all steps in the study.


View this table:
[in this window]
[in a new window]
  		TABLE I Demographic data
 
In the absence of a leak, inspired oxygen fraction (FIO2) increased rapidly within 30 sec, irrespective of the breathing method (Figure 2Go). The FIO2 at the end of the preoxygenation period was significantly greater (P = 0.004) with tidal volume breathing (mean 97 ± 2%, range: 89–98%) compared with values at the end of VC breathing (mean 88 ± 13%, range: 57–98%), (Figure 2Go, Panels A and B).


Figure 2
View larger version (20K):
[in this window]
[in a new window]
  		FIGURE 2 (Panel A) Inspired fraction of oxygen (FI02) during tidal volume breathing preoxygenation for three minutes, with or without leak vs time. (Panel B) FI02 during vital capacity breathing preoxygenation for 0.5 min, with or without leak vs time. (Panel C) End tidal fraction of oxygen (FE02) during tidal volume breathing preoxygenation for three minutes, with or without leak vs time. (Panel D) FE02 during vital capacity breathing preoxygenation for 0.5 min, with or without leak vs time. At the end of the measurement period, FIO2 and FEO2 are statistically less in all panels with a leak than without a leak (P < 0.001). There is a statistically significant difference between Panel A and Panel B (P < 0.01) for no leak. There is no difference between Panel A and Panel B (NS) with the leak present. Similarly, differences between Panel C and Panel D are observed for no leak only (P < 0.01), but not for leak present (NS).

 
In the absence of a leak, the FEO2 for tidal volume breathing for three minutes was greater (89 ± 3%; range 79–91%) than with four VCs (76 ± 7%; range 55–84%); P < 0.001 (Figure 2Go, Panels C and D). An FEO2 > 85% was reached in 90% of subjects with tidal volume breathing, while this was achieved in none of the subjects breathing deeply four times (Table IIGo).


View this table:
[in this window]
[in a new window]
  		TABLE II Number (percentage) of patients with end FEO2 > 90, 85, 80%
 
In the presence of a leak, the FIO2 reached a plateau within 0.5 min with both breathing techniques. The FIO2 reached 61 ± 14% (range 36–89%) with tidal volume breathing and 65 ± 17% (range 31–98%) with VC breathing (Figure 2Go, Panels A–D).

In the presence of a leak, there was no significant difference in FEO2 values between the two breathing methods 61 ± 8% (range 39–78%) for tidal volume and 59 ± 11% (range 34–77%) for VC breaths.

Both the FIO2 and the FEO2 had a similar time course whether or not a leak was present, but reached a lower plateau with a leak (Figure 2Go). After three minutes of tidal volume breathing with a leak the mean FEO2 was 61%, a 28% decrease compared with no leak (FEO2 = 89%), (P < 0.001). A leak also produced a decrease in with the four VC breath technique (P < 0.001), but the difference was only 17%, from 76 to 59%.

The end expiratory carbon dioxide tension was similar in the presence or absence of a leak. When a leak was present it was 33 ± 3 mmHg (range 29–38) after four VCs, compared with 42 ± 4 mmHg (range: 36–50) after three minutes of tidal breathing.


   Discussion
TOP
 Abstract
 Introduction
 Methods
 Results
 Discussion
References
 
The present study shows that during preoxygenation, if a leak is unavoidable, breathing normally for three minutes or taking four rapid VC breaths will yield similar, but suboptimal, FEO2 values. However, without a leak, normal breathing for three minutes provides better oxygenation than four VC breaths.

The few studies which have examined the effect of a leak on preoxygenation used several different techniques to create a leak. We used a nasogastric tube #18, which has an internal diameter of 4 mm, and placed it between the face mask and the bridge of the nose. Others created a leak by cutting a 9.5-mm diameter hole in the mask,1 which we thought to be too large to go undetected. McGowan et al.12 chose to either hold the mask over the subject’s face by gravity alone, or to place it over two small 1-cm3 cubes inserted between the mask and the face. The leak created by a nasogastric tube in the present study is likely to be more reproducible, uniform, and less sensitive to slight changes in pressure applied by the holder of the mask. Whether all the air went through the lumen of the tube, or some of it escaped through leaks on the outside is of little practical importance, as the same method was used to produce a leak in all subjects. The FEO2 obtained in this study after tidal volume breathing (61%) is comparable to that measured while leaving the mask on the face by gravity alone (70%) or putting it 1 cm away from the face (55%).12 Thus, the nasogastric tube is likely to simulate the situation created by a poorly fitting mask.

A leak as small as 4 mm in diameter (nasogastric tube #18) was large enough to prevent high FEO2 values to be obtained, and hence, hindered preoxygenation significantly. The results show a similar FEO2 after both methods with a leak present. However, when compared with a perfect seal, introducing a leak had more effect with tidal volume breathing (a 28% decrease, from 89 to 61%), than with a VC technique (a 17% change, from 76 to 59%). It was expected that VC breathing might have two opposite effects when a leak was introduced. First, the 2-L reservoir bag could empty partially through the leak, causing air to enter the circuit if the volume remaining in the bag was insufficient during a VC maneuver. This alone would make VC breathing worse than tidal volume breathing in the presence of a leak. However, the flow through the leak is likely to be more turbulent than in the rest of the circuit. When flow is laminar, it is proportional to the pressure gradient. When it becomes turbulent, it is proportional to the square root of the pressure gradient.15 Thus, when a large negative pressure is created, as occurs when taking a VC breath, it is expected that the flow might be proportionately less through the leak, that is, through the device that produces turbulent flow. For example, if pressure inside the circuit doubles, laminar flow (from the circuit) doubles, while turbulent flow (through the leak) increases by a factor of {surd}2, or 1.41. This turbulent flow effect, which alone would make the leak less important with VC breaths, was probably greater than the empty reservoir bag effect, because the decrease in FEO2 with the introduction of a leak was less with four VC breaths (17%) than with tidal volume breathing (28%).

The fresh gas flow (FGF) in the present study was set at 6 L·min–1, as most preoxygenation studies were performed with flows in the range of 5 to 10 L·min–1.3 The results obtained in the present study without a leak are similar to those of previous investigations. Nimmagadda et al.3 measured, after three minutes of tidal volume breathing, FEO2 values of 86.2 and 88.1% with FGF rates of 5 and 7 L·min–1, respectively. In comparison we observed an FEO2 of with 88.7% in the present study at a FGF rate of 6 L·min–1. Following four VCs, mean FEO2 values of 75 and 77% were obtained with FGF rates of 5 and 7 L·min–1 of oxygen, respectively.3 This compares well with the value of 76% observed in the present study, at 6 L·min–1 FGF.

In the present study, gas was sampled at the nostrils, not in the circuit. Although both sites might be considered equivalent if no leak is present, differences are likely to exist if gas can escape. We chose to sample oxygen by means of a modified nasal cannula, which allowed non-invasive collection of end-tidal gases without being altered by a leak. With this setup, nose breathing was mandatory to obtain adequate measurements, but facial morphology was not altered, and a complete seal could be obtained when needed. McGowan et al.12 used mouth breathing with a Guedel airway and a nose clip in place. This setup measures end-tidal values well, but modifies the shape of the face and might increase the risk of creating an unwanted leak.

The role of a leak between the patient’s face and the mask in reducing the effect of preoxygenation has been recognized since 1955.10 Berthoud et al. realized that allowing even one breath contaminated with air increased end-tidal nitrogen concentration markedly.9 Drummond et al.1 studied the effect of a fixed leak on oxygen saturation after one minute of preoxygenation. They observed a significant difference in saturation at three minutes postinduction, between the group preoxygenated with (mean saturation 93.6%) and without (mean saturation 96.8%) a 9.5-mm diameter leak cut into the facial mask.1 Leaks during preoxygenation occur in 10 to 15% of individuals with no facial anomalies and normal dentition.14 This probably occurs more frequently in a population of unselected patients.

Several end-points have been used to evaluate the effectiveness of preoxygenation, such as time to onset of oxygen desaturation, partial pressure of oxygen in arterial blood (PaO2), end-tidal nitrogen, and end-tidal oxygen. We chose FEO2 since it has been shown to be an accurate, continuous and non-invasive measure of alveolar oxygen.3,13,16 Furthermore, using FEO2 as an endpoint value in awake volunteers allowed us to study the subjects with the different preoxygenation techniques, with each subject serving as his/her own control.

Various techniques have been advocated to accomplish preoxygenation. Tidal volume breathing of oxygen for three to five minutes is used most often.4,10 We used three minutes of preoxygenation because extending the period to five minutes gives only a marginally greater FEO2.2,3,17,18 Gold et al. questioned the need for three-minute periods of tidal volume breathing by demonstrating that four VCs in 30 sec was equally effective in increasing PaO2.7 The choice of four VCs was also made to minimize preoxygenation time and to improve compliance. This technique was recognized to be effective in earlier studies in American Society of Anesthesiologists (ASA) I–II patients,1,7 in pregnant women5 and in obese individuals8 using PaO2 as an end-point. However, Gambee et al., in ASA I–II patients, found that it took 8.9 ± 1.0 min for patients to desaturate to 90% if preoxygenated with three minutes of tidal volume breathing compared with only 6.8 ± 1.8 min in subjects preoxygenated with four VCs.17 Similar results were reported in the elderly, where time to desaturation was 6.75 ± 1.25 min and 3.5 ±1.5 min respectively.6,19 Thus, it appears that, as was found in the present study, four VCs are not as effective as three minutes of normal breathing in most individuals. Possible exceptions are patients with reduced functional residual capacity, such as obese subjects and pregnant women.

A leak was associated with a smaller decrease in FEO2 if the four deep breath technique was used (17%) compared to three minutes of tidal volume breathing (28%). However, both techniques are equally ineffective if a leak is present because the four VC method is not optimal even without a leak. It is possible that increasing the number of VCs and/or increasing the FGF rate could correct the situation, but this hypothesis was not tested here. A more cautious recommendation is that whenever a leak is detected, everything should be done to correct it. Other alternatives have been proposed, such as delivering oxygen via a mouth piece20 or using a very high FGF (42 L·min–1).21 Peres has also suggested using a shortened endotracheal tube between the lips, for preoxygenation while gently pinching the patient’s nose in cases of facial injuries.22 Our study supports that any leak, even as small as 4 mm of internal diameter, significantly decreases the effectiveness of preoxygenation. Changing the preoxygenation technique does not alter the end result. In the presence of a leak, either three minutes of normal breathing or 4 VCs in 30 sec, using 6 L·min–1 of oxygen, is better than no preoxygenation, but will result in similar and suboptimal FEO2 values.


   Footnotes
 
This study was supported by internal departmental funds. Presented at the 78th IARS congress in Tampa Bay, March 27-31, 2004.

Accepted for publication March 29, 2005. Revision accepted August 23, 2005.


   References
TOP
 Abstract
 Introduction
 Methods
 Results
 Discussion
 References
 
1 Drummond GB, Park GR. Arterial oxygen saturation before intubation of the trachea. An assessment of oxygenation techniques. Br J Anaesth 1984; 56: 987–93.[Abstract/Free Full Text]

2 Baraka AS, Taha SK, Aouad MT, El-Khatib MF, Kawkabani NI. Preoxygenation. Comparison of maximal breathing and tidal volume breathing techniques. Anesthesiology 1999; 91: 612–6.[Medline]

3 Nimmagadda U, Chiravuri SD, Salem MR, et al. Preoxygenation with tidal volume and deep breathing techniques: the impact of duration of breathing and fresh gas flow. Anesth Analg 2001; 92: 1337–41.[Abstract/Free Full Text]

4 Dillon JB, Darsie ML. Oxygen for acute respiratory depression due to administration of thiopental sodium. JAMA 1955; 12: 1114–6.

5 Norris MC, Dewan DM. Preoxygenation for cesarean section: a comparison of two techniques. Anesthesiology 1985; 62: 827–9.[Medline]

6 Valentine SJ, Marjot R, Monk CR. Preoxygenation in the elderly: a comparison of the four-maximal-breath and three-minute techniques. Anesth Analg 1990; 71: 516–9.[Abstract/Free Full Text]

7 Gold MI, Duarte I, Muravchick S. Arterial oxygenation in conscious patients after 5 minutes and after 30 seconds of oxygen breathing. Anesth Analg 1981; 60: 313–5.[Abstract/Free Full Text]

8 Goldberg ME, Norris MC, Larijani GE, Marr AT, Seltzer JL. Preoxygenation in the morbidly obese: a comparison of two techniques. Anesth Analg 1989; 68: 520–2.[Free Full Text]

9 Berthoud M, Read DH, Norman J. Pre-oxygenation--how long? Anaesthesia 1983; 38: 96–102.[Medline]

10 Hamilton WK, Eastwood DW. A study of denitrogenation with some inhalation anesthetic systems. Anesthesiology 1955; 16: 861–7.[Medline]

11 Russell GN, Smith CL, Snowdon SL, Bryson TH. Preoxygenation and the parturient patient. Anaesthesia 1987; 42: 346–51.[Medline]

12 McGowan P, Skinner A. Preoxygenation--the importance of a good face mask seal. Br J Anaesth 1995; 75: 777–8.[Abstract/Free Full Text]

13 Berry CB, Myles PS. Preoxygenation in healthy volunteers: a graph of oxygen "washin" using end-tidal oxygraphy. Br J Anaesth 1994; 72: 116–8.[Abstract/Free Full Text]

14 Machlin HA, Myles PS, Berry CB, Butler PJ, Story DA, Heath BJ. End-tidal oxygen measurement compared with patient factor assessment for determining preoxygenation time. Anaesth Intensive Care 1993; 21: 409–13.[Medline]

15 Grippi MA. Pulmonary Pathophysiology. Philadelphia: J.B. Lippincott Company; 1995.

16 Campbell IT, Beatty PC. Monitoring preoxygenation (Editorial). Br J Anaesth 1994; 72: 3–4.[Free Full Text]

17 Gambee AM, Hertzka RE, Fisher DM. Preoxygenation techniques: comparison of three minutes and four breaths. Anesth Analg 1987; 66: 468–70.[Free Full Text]

18 McCrory JW, Matthews JN. Comparison of four methods of preoxygenation. Br J Anaesth 1990; 64: 571–6.[Abstract/Free Full Text]

19 McCarthy G, Elliott P, Mirakhur RK, McLoughlin C. A comparison of different pre-oxygenation techniques in the elderly. Anaesthesia 1991; 46: 824–7.[Medline]

20 Hirsch J, Fuhrer I, Kuhly P, Schaffartzik W. Preoxygenation: a comparison of three different breathing systems. Br J Anaesth 2001; 87: 928–31.[Abstract/Free Full Text]

21 Ooi R, Pattison J, Joshi P, Chung R, Soni N. Pre-oxygenation: the Hudson mask as an alternative technique. Anaesthesia 1992; 47: 974–6.[Medline]

22 Peres P. Preoxygenation and facial injuries (Letter). Anaesth Intensive Care 1990; 18: 580–1.[Medline]




This article has been cited by other articles:


Home page
 Canadian J. AnesthesiaHome page
S. Taha, M. El-Khatib, S. Siddik-Sayyid, C. Dagher, J.-M. Chehade, and A. Baraka
Preoxygenation with the Mapleson D system requires higher oxygen flows than Mapleson A or circle systems: [La preoxygenation avec le systeme Mapleson D requiert un debit d'oxygene plus eleve que les systemes Mapleson A ou en cercle]
Can J Anesth, February 1, 2007; 54(2): 141 - 145.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Résumé de cet Article
Right arrow Full Text (PDF)
Right arrow An erratum has been published
Right arrow Submit a scholarly reply
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Gagnon, C.
Right arrow Articles by Donati, F.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Gagnon, C.
Right arrow Articles by Donati, F.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS