- Email: [email protected]
Effectiveness of respirator masks for healthcare workers, in France
Disponible en ligne sur
www.sciencedirect.com Médecine et maladies infectieuses 42 (2012) 264–269
Original article
Effectiveness of respirator masks for healthcare workers, in France夽 Efficacité des masques protecteurs respiratoires pour les soignants en France C. Ciotti a,∗ , G. Pellissier a , C. Rabaud a , J.-C. Lucet b , D. Abiteboul a , E. Bouvet a , GERES a
Groupe d’étude sur le risque d’exposition des soignants aux agents infectieux (GERES), Faculté de Médecine Xavier-Bichat, 16, rue Henri-Huchard, 75018 Paris, France b Unité d’hygiène et de lutte contre les infections nosocomiales, GH Bichat-Claude-Bernard, 46, rue Henri-Huchard, 75018 Paris, France Received 16 November 2011; received in revised form 9 February 2012; accepted 1st May 2012 Available online 9 June 2012
Abstract Objective. – The authors had for objective to evaluate the air-tightness of FFP2 respirator masks used by healthcare workers, with a quantitative fit-test protocol. Materials and methods. – This test measures the number of ambient particles inside and outside the respirator mask. The ratio between both is called fit-factor. The fit-test is successful for an FFP2 respirator mask when the fit-factor is equal or superior to 100. The tests were performed in three hospitals. Nine types of FFP2 respirator masks were fit-tested, classified in three groups: hard shell, duckbill, and flat-fold respirator masks. Results. – One hundred and eighty fit-tests were performed. Less than a third of the fit-tests were successful (35/130). The rate of successful tests was higher with flat-fold (57.5%, 23/40) than with duckbill (18.3%, 11/60), or hard shell respirator masks (3.3%, 1/30), (P < 0.05). Zero to 60% of healthcare workers had a successful fit-test with the respirator masks used in each hospital. This percentage increased with the number of tested respirator masks. No 100% success rate was ever reached in any hospital with the three tested respirator masks. Conclusion. – Duckbill, and flat-fold respirator masks seem to be better adapted for healthcare workers than hard shell respirator masks. It seems necessary to implement new recommendations for respiratory protection in France. At least two types of respirator masks with various sizes and shape should be available and fitting controls should be performed with respirator masks that are worn by healthcare workers exposed to infectious risks. © 2012 Published by Elsevier Masson SAS. Keywords: Air transmission; FFP2; Fit-test; Fitting; Half-face respirator mask
Résumé Objectif. – Évaluer l’étanchéité de masques de protection respiratoire FFP2 chez des soignants à l’aide de fit-tests quantitatifs. Matériels et méthodes. – Ce test mesure la concentration en particules à l’intérieur et à l’extérieur du masque. Le rapport entre les deux est appelé fit-factor. Pour un masque FFP2, quand le fit-factor est supérieur ou égal à 100, le fit-test est réussi. Les tests ont été conduits dans trois hôpitaux. Neuf masques FFP2 ont été testés, classés en trois catégories: à coquille dure, bec de canard, à plis. Résultats. – Cent trente fit-tests ont été réalisés. Moins d’un tiers des fit-tests étaient réussis (35/130). Le taux de fit-tests réussis était plus élevé avec les masques à plis (57,5 %, 23/40) qu’avec les bec de canard (18,3 %, 11/60) ou à coquille dure (3,3 %, 1/30), (p < 0,05). De zéro à 60 % des soignants ont obtenu un fit-test réussi avec le masque utilisé dans chaque établissement. Ce pourcentage augmentait avec le nombre de masques testés. Le taux de 100 % de réussite n’a jamais été atteint avec les trois masques testés dans aucun établissement.
夽 Dates and place of congress where the study was presented: 12e Journées nationales d’Infectiologie, 9 juin 2011, Toulouse; 19e Journée annuelle du GERES, 18 septembre 2009, Paris; XXe congrès de la SFHH, 4 juin 2009, Nice. ∗ Corresponding author. E-mail address: [email protected] (C. Ciotti).
0399-077X/$ – see front matter © 2012 Published by Elsevier Masson SAS. http://dx.doi.org/10.1016/j.medmal.2012.05.001
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
265
Conclusion. – Les masques à plis et bec de canard semblent être plus adaptés au milieu de soins, que les masques à coquille dure. Il paraît nécessaire de faire évoluer les recommandations franc¸aises en matière de protection respiratoire afin de mettre à disposition des soignants au moins deux types de masques, de formes et/ou de tailles différentes, et de préconiser des contrôles préalables de l’étanchéité des masques chez les professionnels à risque. © 2012 Publié par Elsevier Masson SAS. Mots clés : FFP2 ; Fit-test ; Fuite au visage ; Masques de protection respiratoire ; Transmission aérienne
1. Introduction The National Public Health Council (Conseil supérieur d’hygiène publique de France, [CSHPF]) recommends carrying protective respirator masks in healthcare institutions to prevent infections transmitted by aerosols, and especially tuberculosis [1,2]. The respirator mask used must answer FFP1 (Filtering Facepiece Particles) requirements at least (maximal facial leak of 22% according to standard EN 149: 2001) [3]. They must answer FFP2 requirements (maximal facial leak of 8%) in specific cases such as tuberculosis resistant, bronchial fibroscopy, or intubation. A lot of healthcare institutions have chosen FFP2 respirator masks in every case, to facilitate their use. Protective respirator masks are now well known and widely distributed in French healthcare institutions, as reported in a GERES study [4]. Several shapes of protective respirator masks are available on the market and do not fit all types of faces in the same way. The respirator mask should be correctly adjusted for an adequate protection, so as to ensure an airtight seal with the face and prevent contaminated air to penetrate the respirator mask and be inhaled. If the respirator mask does not adhere well to an individual’s face, its effectiveness against aerosols is compromised, even if its performances were tested and complied with standard EN 149: 2001 [3]. If the user does not obtain the expected airtightness with his respirator mask, he should try another size or another model. In the first GERES study on protective respirator masks [4], 493 healthcare providers had made fit-checks1 , after an investigator had checked the respirator mask placement. A lack of air-tightness was observed in one-quarter of healthcare providers with the respirator mask model available in their institution. Other tests are available to assess the air-tightness of a respirator mask, in a more accurate way; they are called fit-tests2 (qualitative or quantitative). Derick et al. [5] discussed the interest of making fit-checks without having previously checked a respirator mask’s air-tightness, with a fit-test which, if it is more demanding, appears to be more reliable, in any situation (routine or epidemic). This is why, to complete the results of its first study, the GERES decided to test the air-tightness of a panel of protective FFP2 respirator masks in institutions, i.e. with healthcare
1 The fit-check is a test allowing checking how well a respirator mask is adjusted; it is recommended to perform it every time a respirator mask is used. If there are no leaks with forced inspiration, with a plastic sheet or hands wrapped around the respirator mask, the respirator mask is slightly depressed on the face. 2 cf. Materials and Methods for definitions.
providers who regularly use a respirator mask, with an objective, quantitative, and validated test such as the fit-test [6]. 2. Materials and methods 2.1. Fit-test The fit-test is an operating mode allowing checking the airtightness of a protective respirator mask on an individual: • qualitatively: test based on the perception of sweet or sour taste through the respirator mask while performing a few exercises. Any leak on the face is revealed by the perception of taste; • quantitatively: test measuring the concentration of particles inside and outside the respirator mask while performing a few exercises. Any leak on the face is revealed by an increased concentration of particles inside the respirator mask. The quantitative fit-tests were performed with a Portacount® Pro+ model 8038 (TSI), following the protocol issued by the Occupational Safety and Health Administration (OSHA) [6]. The Portacount® measures the concentration of particles inside and outside the respirator mask while the wearer of the respirator mask is performing a few exercises, and calculates the ratio between these two concentrations called adequation factor or fit-factor [7]. The more the respirator mask leaks, the higher the concentration of particles inside the respirator mask will be, and the lower the fit-factor. If the fit-factor is superior or equal to 100 [6], the fit-test for a protective respirator mask FFP2 is successful and the respirator mask is considered as airtight. A model 8026 (TSI) particle generator was used to ensure an adequate concentration of particles in the ambient air. 2.2. Performing the tests The fit-tests were performed and controlled by a GERES technician previously trained in the use of measuring devices and software. The technician explained how to wear and adjust the respirator mask to healthcare providers. The adequate adjustment of the respirator mask, before beginning the fit-test, was ensured by measuring the fit-factor in real time (visualization of the fit-factor value in real time) and the respirator mask was readjusted if necessary. Testing lasted 10 minutes by respirator mask. During the test, the technician inspected visually the presence of leaks (visible space between the chin and the respirator mask, presence of fogging on the glasses, etc.).
266
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
2.3. Respirator mask tested Nine models of FFP2 protective respirator masks were tested. These respirator masks were the most frequently used in France, according to the GERES survey on respiratory protection in healthcare institutions [4]. These respirator masks were classified according to their shape for the survey. Three categories were defined: hard-shell respirator masks, duckbill respirator masks, and those with alternative shapes called flat fold respirator masks.
Table 1 Leaks observed by the technician according to the type of respirator mask for failed fit-test. Fuites observées par le technicien en fonction du type de masques pour les tests s’étant conclus par un échec. Leaks
Nose Chin Cheeks
Respirator masks n (%) Hard shell (n = 29)
Duckbill (n = 49)
Flat fold (n = 17)
27 (93.1) 28 (96.6) 15 (51.7)
43 (87.6) 19 (38.8) 2 (4.1)
11 (64.7) 11 (64.7) 0
2.4. Performing the study The measures were performed in three hospitals. Ten volunteer healthcare providers were recruited in each center, from units in which protective respirator masks were commonly used (infectious diseases unit, sanatorium, pneumology unit). Nine different protective respirator masks were tested 10 times each: three hard-shell respirator masks, four duckbill respirator masks, and two flat-fold respirator masks. In two institutions, the participating healthcare providers tested a hard-shell respirator mask, a duckbill respirator mask, and a flat fold respirator mask. In the third institution, the tests were performed with a hard shell respirator mask and two duckbill respirator masks because only two flat-fold respirator masks were available during the study period. In each institution, one of the tested respirator mask was the one routinely used. The results needed confirmation for four respirator masks. A second series of tests was performed for: • two duckbill models having obtained discordant results even though these respirator masks were identical except for the respirator mask edge joint; • two flat fold models with different brand and design whose first test results were similar. Twenty new healthcare providers were recruited, 10 to test the two duckbill respirator masks and 10 for the two flat-fold respirator masks.
Ninety-five tests failed (fit-factor inferior to 100; [minimum: 2–maximum: 99]). Leaks were observed by the technician at the nose level for 81 tests, the chin for 58 tests, and cheeks for 17 tests. The total was higher than 95 because during a single test, the respirator mask could leak in several places (examples: nose and chin, nose and cheeks, etc.). The leaks observed for each type of respirator mask are listed in Table 1. The leaks were located mainly on the nose, chin, and cheeks for hard-shell respirator masks, on the nose and chin for flat-fold respirator masks, and on the nose for duckbill respirator masks. No fit-test failed without the technician having observed visually facial leaks. Fig. 1 illustrates the distribution of fit-factors according to the type of respirator mask tested, during the first phase of tests. The distribution profiles varied according to the types of protective respirator mask. The fit-factors for hard-shell respirator masks were inferior to 25, except for one test equal to 200. The fit-factors for the two flat-fold respirator masks were regularly widely distributed between 0 and 200 (median: 106.5 vs. 119.0; p25: 73.3 vs. 43.5; p75: 183.0 vs. 146.8; 60% vs. 55% of successful fit-tests). The duckbill respirator masks were less widely distributed. The distribution of fit-factors varied according to the brand of tested duckbill respirator masks (median: 15.5 vs. 13.5 vs. 29.0 vs. 51.5). Thus, two major distribution profiles of
200
2.5. Data collection and processing
3. Results Finally, 130 fit-tests were performed with 50 healthcare providers (90 with 30 healthcare providers, completed by 40 tests with 20 healthcare providers). 35 tests out of 130 (26.9%) were successful (fit-factor higher or equal to 100; [minimum: 101–maximum: 200]). The rate of successful fit-tests was higher with flat-fold respirator masks (23/40, 57.5%), than with duckbill respirator masks (11/60, 18.3%), and hard shell respirator masks (1/30, 3.3%) (P < 0.05).
100
Fit-factor
150
The data was collected and processed with the Epi Info software version 6.04d. The 2 test was used to compare results among the different types of respirator masks. A value of P < 0.05 was considered as statistically significant.
50
0
Hard shell
Duckbill
Flat fold
Type of respirator mask Fig. 1. Distribution of fit-factors according to respirator mask shape. Distribution of fit-factors; median; + percentiles 25 and 75. Distribution des fit-factors en fonction de la forme de masques. Distribution des fit-factors; médiane; + percentiles 25 et 75.
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
fit-factors were observed for the nine respirator masks tested. A first profile concerned hard shell respirator masks (n = 3) and two duckbill respirator masks. For this profile, most fit-factors, 92%, were inferior or equal to 25 (46/50). A second profile was identified for flat-fold respirator masks (n = 2) and a duckbill respirator masks, with a comparable number of fit-factors inferior or superior to 100 equally distributed among these three respirator masks. The last duckbill respirator mask had a fit-factor distribution profile considered as intermediate. Table 2 lists, by institution, the number of individuals for whom at least one fit-test was successful in the consecutive trial of three types of respirator masks, including the respirator masks available in the institution. The latter systematically failed the fit-test in two institutions (a duckbill respirator mask and a hardshell respirator mask). In the third institution (duckbill respirator mask), the rate of failure was still 40% (4/10). The number of healthcare providers for whom a fit-test was successful increased with the number of respirator masks tested. Thus, after having tested three different types of respirator masks, the cumulative percentage of successful fit-tests was 70, 30, and 90 respectively in the three institutions. No 100% success rate was reached in any institution with the three tested respirator masks. 4. Discussion In Europe, the global effectiveness of a protective respirator mask must be assessed before it can be marketed. It must comply with standard EN 149: 2001 [3]. This standard includes assessing the filter’s effectiveness and total inner leak. The measures are performed with a panel of chosen individuals exposed to an aerosol while performing a series of exercises. Protective respirator masks have been used only recently in French healthcare institutions. Carrying a protective respirator mask was first recommended by the CSHPF, in 1993, for healthcare providers in contact with contagious patients presenting with tuberculosis [1]. Marketing of a class N95 protective respirator mask in the USA, with a filtration capacity equivalent to a class FFP2respirator, is possible after the single assessment of filter material effectiveness according to the American standard [8]. Adherence to the face is not mentioned in this standard. Thus, performing fit-tests is strongly recommended to choose a model adapted to the user, both for the industry and for hospitals [9], because of the lack of facial leak detection test in the standard. The National Institute of Occupational Safety and Health (NIOSH) requires a mandatory fit-test for any individual exposed to tuberculosis [10]. Furthermore, respirator masks with various shape and size may be available for healthcare providers. They can thus be given a respirator mask corresponding to their facial morphology. Protective respirator masks in French healthcare institutions are required to comply with standard EN 149: 2001 [3] (FFP1, FFP2, or FFP3). Nevertheless, “true” effectiveness criteria (adherence to the face) are not taken into account on the site. This is why, with the “tuberculosis” recommendations and the emergence of new respiratory risks (SRAS, flu pandemics, etc.), it seemed necessary to assess as objectively as possible,
267
the true effectiveness of respirator masks on the site, for facial air-tightness. Our results prove that flat-fold and duckbill respirator masks are better adapted for healthcare providers than hard-shell respirator masks. They are soft, and thus more easily adjustable to the face of healthcare providers, mostly women, who generally have finer features. But all the duckbill respirator masks are not equivalent for air-tightness. Indeed, they may have comparable results with flat-fold respirator masks or hard-shell respirator masks, partly because of the respirator mask design. All the duckbill respirator masks do have the same shape, but not the same quality of joint, of rubber bands, of nasal clip, etc. The hard-shell respirator masks, do not seem adapted to healthcare settings because they are too wide and rigid to be adjustable. A smaller size could be better adapted. A respirator mask may perfectly fit some healthcare providers and not be airtight for others because of the different morphology of faces. This is the case for hard shell respirator masks tested in our study which do not seem adapted (with fit-tests inferior of 25), except in one case (with a fit-factor of 200). This is why the shape and size of the respirator mask as well as the facial morphology of the user should always be taken into account when choosing a protective respirator mask. This adequacy is the condition for the respirator mask’s air-tightness and thus for an optimal protection of the carrier. Given the diversity of facial morphology in healthcare institution personnel, a single type of protective respirator mask is not sufficient to adequately protect all the personnel. Indeed, no fit-test was successful in two institutions where only one type of respirator mask was available. It was a hard shell respirator mask in one and a duckbill respirator mask in the other. In the third institution (duckbill respirator mask), a little less than half of the tests failed. After having tested three respirator masks, the percentage of individuals with at least one successful fit-test rose respectively from 0 to 30%, 0 to 90%, and 10 to 70% in the three institutions. The results of another study made in the SaintPierre University Hospital in Brussels in which 417 quantitative fit-tests were performed [11] correlates with our results; 67% of the personnel was protected by the respirator mask available in the institution. A better adjustment of the respirator mask could have made the test successful, whereas a number of fit-tests failed, especially for fit-factors close to 100 (six fit-factors ranged between 82 and 99, out of the 100 tests performed with flat-fold respirator masks and duckbill respirator masks). This better adjustment could be obtained with training sessions for respirator mask carriage. But this approach cannot improve the test results if the respirator mask is not adapted to facial morphology; in this case leaks will persist. Various authors [11,12] reported that healthcare providers were never given any specific instructions, 56% to 64% did not place the respirator mask adequately, for either the quantitative or the qualitative fit-test; and a respirator mask adapted to a morphology, chosen with a fit-test, is effective only if it is carried correctly. It thus seems mandatory to teach healthcare providers how to wear and adjust the respirator mask. These training sessions should be renewed regularly, especially for the most exposed personnel [12,13]. Using a fit-test for personnel
268
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
Table 2 Number of individuals for whom at least one fit-test was successful during the consecutive use of three types of respirator masks. Nombre de personnes pour lesquelles au moins un fit-test était réussi lors de l’essai successif de trois types de masques. Number of respirator masks tested
Individuals with at least one successful fit-test n (%) Institution 1
Institution 2
Institution 3
1 respirator mask (= respirator mask available in the institution) 2 respirator masks 3 respirator masks
0b
0a
6/10 (60)c 7/10 (70)a
1/10 (10)b 3/10 (30)b
6/10 (60)b 9/10 (90)c 9/10 (90)a
a b c
Four different hard shell respirator masks. Four different duckbill respirator masks. Two different flat fold respirator masks.
who use a protective respirator mask occasionally may be discussed [12]. But, the fit-test is a much more accurate quantitative and objective test than visual observation which may be observer dependent. Healthcare providers should perform at least one fit-check every time they use a respirator mask, to ensure its adequate placement and air-tightness [4]. Our study had limitations. The fit-tests were not performed in real conditions. Indeed, the measured particles were not infectious. The tests lasted only 10 minutes whereas a healthcare provider may have to wear a protective respirator mask longer during caregiving. But the fit-test protocol was approved by the Occupational Safety and Health Administration OSHA [6] and has been applied since the late 90 s in the USA. The study sample may be considered weak (130 fit-tests), but this sample is larger than the one used for standard EN 149: 2001 (10 tests per respirator mask). Furthermore, the respirator masks were tested with healthcare providers and not on dummies or individuals with a chosen morphology as was the case in the European standard. Performing fit-tests in uncommon in French hospitals. But these tests may help answer a number of questions about the airtightness of respirator masks, questions that were often raised during the last flu pandemic.
Acknowledgements
5. Conclusion
References
This survey revealed that a single model of respirator mask is not sufficient to protect healthcare providers in a healthcare institution especially if its shape and/or size are not adapted. The technical training for protective respirator mask placement is essential but contributive only if the respirator mask is adjustable enough to fit the face. The flat-fold and duckbill respirator masks seem to be better adapted to healthcare settings than hard-shell respirator masks. Indeed, the hard shell respirator masks currently available in healthcare institutions are not safe enough because they are too large; they should not be used for caregiving when there is a risk of airborne transmission such as in tuberculosis. In a context of emerging viral risk of respiratory transmission, and of multiresistant tuberculosis, it is necessary to update French recommendations concerning respiratory protection. The need to have at least two types of respirator masks, with different shapes and/or sizes, should be discussed as well as the need for training and checking the chosen respirator mask’s
air-tightness: fit-test (at least initially when choosing the respirator mask) and/or at least a fit-check every time a respirator mask is used. Disclosure of interest The authors declare that they have no conflict of interest concerning this article.
The GERES would like to thank the Direction Générale de la Santé for its financial support, the firm TSI and the 3M Santé laboratories, and more especially Franck Ajuelos, for loaning the devices needed to perform the fit-tests, the various suppliers of respirator masks, as well as the participating institutions (GH Bichat-Claude Bernard, Paris–GH La PitiéSalpêtrière, Paris–Centre Médical de Bligny, Briis-sous-Forges) and the personnel who tested the respirator masks. Financing: Direction Générale de la Santé.
[1] Conseil Supérieur d’Hygiène Publique de France, section maladies transmissibles. Avis du CSHPF du 14 mars 2003 relatif au choix d’un masque de protection contre la tuberculose en milieu de soins. http://www.sante.gouv.fr/IMG/pdf/Avis du CSHPF du 14 mars 2003 relatif au choix d un masque de protection contre la tuberculose en milieu de soins-.pdf. (Accès 16 juin 2011). [2] Surveiller et prévenir les infections associées aux soins. HygièneS. 2010; XVIII (4). http://www.sf2h.net/publications-SF2H/SF2H surveilleret-prevenir-les-IAS-2010.pdf. (Accès 16 juin 2011). [3] Respiratory protective devices–Filtering half masks to protect against particles–Requirements, testing, marking. S 76-014. AFNOR 2001. [4] Ciotti C, Bouvet E, Abiteboul D, le GERES et l’INRS. Utilisation des masques de protection respiratoire chez les soignants. Med Mal Infect 2008;38:452–6. [5] Derrick JL, Chan YF, Gomersall CD, Lui SF. Predictive value of the user seal check in determining half-face respirator mask fit. J Hosp Infect 2005;59:152–5. [6] Occupational Safety and Health Administration (OSHA). Appendix A to §1910.134 Fit Testing Procedures (Mandatory). Washington DC: Office of the Federal Register, National Archives and Records Administration; 1998. [7] TSI, Inc. Portacount® Pro 8030 and Portacount® Pro + 8038 manual. St. Paul, Minnesota: TSI, Inc, 2010.
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269 [8] National Institute for Occupational Safety and Health (NIOSH). 42 CFR 84 Respiratory protective devices: final rules and notice. Federal Register 60:110. US Centers for Disease Control and Prevention, NIOSH; 1997. [9] National Institute for Occupational Safety and Health (NIOSH) 29 CFR Part 1910.134. Code of Federal Regulations. Occupational Safety and Health Administration: Respiratory Protection. Washington DC: Office of the Federal Register National Archives and Records Administration; 1997. [10] National Institute for Occupational Safety and Health (NIOSH). TB respiratory protection program in health care facilities. Morgantown, WV: US
269
Department of Health and Human Services, Public Health Service, Centers for Disease Control, NIOSH; 1999. [11] Gérard M, Suys M. Fit-testing des masques respiratoires FFP2. Une expérience de terrain. Noso-Info 2006:2–3. [12] Lee MC, Takaya S, Long R, Joffe M. Respirator mask -fit testing: does it ensure the protection of healthcare workers against respirable particles carrying pathogens? Infect Control Hosp Epidemiol 2008;29:1149–56. [13] Kelly L, Clark K. The effectiveness of training and taste testing when using respirator mask. J Hosp Infect 2004;58:240–1.
www.sciencedirect.com Médecine et maladies infectieuses 42 (2012) 264–269
Original article
Effectiveness of respirator masks for healthcare workers, in France夽 Efficacité des masques protecteurs respiratoires pour les soignants en France C. Ciotti a,∗ , G. Pellissier a , C. Rabaud a , J.-C. Lucet b , D. Abiteboul a , E. Bouvet a , GERES a
Groupe d’étude sur le risque d’exposition des soignants aux agents infectieux (GERES), Faculté de Médecine Xavier-Bichat, 16, rue Henri-Huchard, 75018 Paris, France b Unité d’hygiène et de lutte contre les infections nosocomiales, GH Bichat-Claude-Bernard, 46, rue Henri-Huchard, 75018 Paris, France Received 16 November 2011; received in revised form 9 February 2012; accepted 1st May 2012 Available online 9 June 2012
Abstract Objective. – The authors had for objective to evaluate the air-tightness of FFP2 respirator masks used by healthcare workers, with a quantitative fit-test protocol. Materials and methods. – This test measures the number of ambient particles inside and outside the respirator mask. The ratio between both is called fit-factor. The fit-test is successful for an FFP2 respirator mask when the fit-factor is equal or superior to 100. The tests were performed in three hospitals. Nine types of FFP2 respirator masks were fit-tested, classified in three groups: hard shell, duckbill, and flat-fold respirator masks. Results. – One hundred and eighty fit-tests were performed. Less than a third of the fit-tests were successful (35/130). The rate of successful tests was higher with flat-fold (57.5%, 23/40) than with duckbill (18.3%, 11/60), or hard shell respirator masks (3.3%, 1/30), (P < 0.05). Zero to 60% of healthcare workers had a successful fit-test with the respirator masks used in each hospital. This percentage increased with the number of tested respirator masks. No 100% success rate was ever reached in any hospital with the three tested respirator masks. Conclusion. – Duckbill, and flat-fold respirator masks seem to be better adapted for healthcare workers than hard shell respirator masks. It seems necessary to implement new recommendations for respiratory protection in France. At least two types of respirator masks with various sizes and shape should be available and fitting controls should be performed with respirator masks that are worn by healthcare workers exposed to infectious risks. © 2012 Published by Elsevier Masson SAS. Keywords: Air transmission; FFP2; Fit-test; Fitting; Half-face respirator mask
Résumé Objectif. – Évaluer l’étanchéité de masques de protection respiratoire FFP2 chez des soignants à l’aide de fit-tests quantitatifs. Matériels et méthodes. – Ce test mesure la concentration en particules à l’intérieur et à l’extérieur du masque. Le rapport entre les deux est appelé fit-factor. Pour un masque FFP2, quand le fit-factor est supérieur ou égal à 100, le fit-test est réussi. Les tests ont été conduits dans trois hôpitaux. Neuf masques FFP2 ont été testés, classés en trois catégories: à coquille dure, bec de canard, à plis. Résultats. – Cent trente fit-tests ont été réalisés. Moins d’un tiers des fit-tests étaient réussis (35/130). Le taux de fit-tests réussis était plus élevé avec les masques à plis (57,5 %, 23/40) qu’avec les bec de canard (18,3 %, 11/60) ou à coquille dure (3,3 %, 1/30), (p < 0,05). De zéro à 60 % des soignants ont obtenu un fit-test réussi avec le masque utilisé dans chaque établissement. Ce pourcentage augmentait avec le nombre de masques testés. Le taux de 100 % de réussite n’a jamais été atteint avec les trois masques testés dans aucun établissement.
夽 Dates and place of congress where the study was presented: 12e Journées nationales d’Infectiologie, 9 juin 2011, Toulouse; 19e Journée annuelle du GERES, 18 septembre 2009, Paris; XXe congrès de la SFHH, 4 juin 2009, Nice. ∗ Corresponding author. E-mail address: [email protected] (C. Ciotti).
0399-077X/$ – see front matter © 2012 Published by Elsevier Masson SAS. http://dx.doi.org/10.1016/j.medmal.2012.05.001
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
265
Conclusion. – Les masques à plis et bec de canard semblent être plus adaptés au milieu de soins, que les masques à coquille dure. Il paraît nécessaire de faire évoluer les recommandations franc¸aises en matière de protection respiratoire afin de mettre à disposition des soignants au moins deux types de masques, de formes et/ou de tailles différentes, et de préconiser des contrôles préalables de l’étanchéité des masques chez les professionnels à risque. © 2012 Publié par Elsevier Masson SAS. Mots clés : FFP2 ; Fit-test ; Fuite au visage ; Masques de protection respiratoire ; Transmission aérienne
1. Introduction The National Public Health Council (Conseil supérieur d’hygiène publique de France, [CSHPF]) recommends carrying protective respirator masks in healthcare institutions to prevent infections transmitted by aerosols, and especially tuberculosis [1,2]. The respirator mask used must answer FFP1 (Filtering Facepiece Particles) requirements at least (maximal facial leak of 22% according to standard EN 149: 2001) [3]. They must answer FFP2 requirements (maximal facial leak of 8%) in specific cases such as tuberculosis resistant, bronchial fibroscopy, or intubation. A lot of healthcare institutions have chosen FFP2 respirator masks in every case, to facilitate their use. Protective respirator masks are now well known and widely distributed in French healthcare institutions, as reported in a GERES study [4]. Several shapes of protective respirator masks are available on the market and do not fit all types of faces in the same way. The respirator mask should be correctly adjusted for an adequate protection, so as to ensure an airtight seal with the face and prevent contaminated air to penetrate the respirator mask and be inhaled. If the respirator mask does not adhere well to an individual’s face, its effectiveness against aerosols is compromised, even if its performances were tested and complied with standard EN 149: 2001 [3]. If the user does not obtain the expected airtightness with his respirator mask, he should try another size or another model. In the first GERES study on protective respirator masks [4], 493 healthcare providers had made fit-checks1 , after an investigator had checked the respirator mask placement. A lack of air-tightness was observed in one-quarter of healthcare providers with the respirator mask model available in their institution. Other tests are available to assess the air-tightness of a respirator mask, in a more accurate way; they are called fit-tests2 (qualitative or quantitative). Derick et al. [5] discussed the interest of making fit-checks without having previously checked a respirator mask’s air-tightness, with a fit-test which, if it is more demanding, appears to be more reliable, in any situation (routine or epidemic). This is why, to complete the results of its first study, the GERES decided to test the air-tightness of a panel of protective FFP2 respirator masks in institutions, i.e. with healthcare
1 The fit-check is a test allowing checking how well a respirator mask is adjusted; it is recommended to perform it every time a respirator mask is used. If there are no leaks with forced inspiration, with a plastic sheet or hands wrapped around the respirator mask, the respirator mask is slightly depressed on the face. 2 cf. Materials and Methods for definitions.
providers who regularly use a respirator mask, with an objective, quantitative, and validated test such as the fit-test [6]. 2. Materials and methods 2.1. Fit-test The fit-test is an operating mode allowing checking the airtightness of a protective respirator mask on an individual: • qualitatively: test based on the perception of sweet or sour taste through the respirator mask while performing a few exercises. Any leak on the face is revealed by the perception of taste; • quantitatively: test measuring the concentration of particles inside and outside the respirator mask while performing a few exercises. Any leak on the face is revealed by an increased concentration of particles inside the respirator mask. The quantitative fit-tests were performed with a Portacount® Pro+ model 8038 (TSI), following the protocol issued by the Occupational Safety and Health Administration (OSHA) [6]. The Portacount® measures the concentration of particles inside and outside the respirator mask while the wearer of the respirator mask is performing a few exercises, and calculates the ratio between these two concentrations called adequation factor or fit-factor [7]. The more the respirator mask leaks, the higher the concentration of particles inside the respirator mask will be, and the lower the fit-factor. If the fit-factor is superior or equal to 100 [6], the fit-test for a protective respirator mask FFP2 is successful and the respirator mask is considered as airtight. A model 8026 (TSI) particle generator was used to ensure an adequate concentration of particles in the ambient air. 2.2. Performing the tests The fit-tests were performed and controlled by a GERES technician previously trained in the use of measuring devices and software. The technician explained how to wear and adjust the respirator mask to healthcare providers. The adequate adjustment of the respirator mask, before beginning the fit-test, was ensured by measuring the fit-factor in real time (visualization of the fit-factor value in real time) and the respirator mask was readjusted if necessary. Testing lasted 10 minutes by respirator mask. During the test, the technician inspected visually the presence of leaks (visible space between the chin and the respirator mask, presence of fogging on the glasses, etc.).
266
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
2.3. Respirator mask tested Nine models of FFP2 protective respirator masks were tested. These respirator masks were the most frequently used in France, according to the GERES survey on respiratory protection in healthcare institutions [4]. These respirator masks were classified according to their shape for the survey. Three categories were defined: hard-shell respirator masks, duckbill respirator masks, and those with alternative shapes called flat fold respirator masks.
Table 1 Leaks observed by the technician according to the type of respirator mask for failed fit-test. Fuites observées par le technicien en fonction du type de masques pour les tests s’étant conclus par un échec. Leaks
Nose Chin Cheeks
Respirator masks n (%) Hard shell (n = 29)
Duckbill (n = 49)
Flat fold (n = 17)
27 (93.1) 28 (96.6) 15 (51.7)
43 (87.6) 19 (38.8) 2 (4.1)
11 (64.7) 11 (64.7) 0
2.4. Performing the study The measures were performed in three hospitals. Ten volunteer healthcare providers were recruited in each center, from units in which protective respirator masks were commonly used (infectious diseases unit, sanatorium, pneumology unit). Nine different protective respirator masks were tested 10 times each: three hard-shell respirator masks, four duckbill respirator masks, and two flat-fold respirator masks. In two institutions, the participating healthcare providers tested a hard-shell respirator mask, a duckbill respirator mask, and a flat fold respirator mask. In the third institution, the tests were performed with a hard shell respirator mask and two duckbill respirator masks because only two flat-fold respirator masks were available during the study period. In each institution, one of the tested respirator mask was the one routinely used. The results needed confirmation for four respirator masks. A second series of tests was performed for: • two duckbill models having obtained discordant results even though these respirator masks were identical except for the respirator mask edge joint; • two flat fold models with different brand and design whose first test results were similar. Twenty new healthcare providers were recruited, 10 to test the two duckbill respirator masks and 10 for the two flat-fold respirator masks.
Ninety-five tests failed (fit-factor inferior to 100; [minimum: 2–maximum: 99]). Leaks were observed by the technician at the nose level for 81 tests, the chin for 58 tests, and cheeks for 17 tests. The total was higher than 95 because during a single test, the respirator mask could leak in several places (examples: nose and chin, nose and cheeks, etc.). The leaks observed for each type of respirator mask are listed in Table 1. The leaks were located mainly on the nose, chin, and cheeks for hard-shell respirator masks, on the nose and chin for flat-fold respirator masks, and on the nose for duckbill respirator masks. No fit-test failed without the technician having observed visually facial leaks. Fig. 1 illustrates the distribution of fit-factors according to the type of respirator mask tested, during the first phase of tests. The distribution profiles varied according to the types of protective respirator mask. The fit-factors for hard-shell respirator masks were inferior to 25, except for one test equal to 200. The fit-factors for the two flat-fold respirator masks were regularly widely distributed between 0 and 200 (median: 106.5 vs. 119.0; p25: 73.3 vs. 43.5; p75: 183.0 vs. 146.8; 60% vs. 55% of successful fit-tests). The duckbill respirator masks were less widely distributed. The distribution of fit-factors varied according to the brand of tested duckbill respirator masks (median: 15.5 vs. 13.5 vs. 29.0 vs. 51.5). Thus, two major distribution profiles of
200
2.5. Data collection and processing
3. Results Finally, 130 fit-tests were performed with 50 healthcare providers (90 with 30 healthcare providers, completed by 40 tests with 20 healthcare providers). 35 tests out of 130 (26.9%) were successful (fit-factor higher or equal to 100; [minimum: 101–maximum: 200]). The rate of successful fit-tests was higher with flat-fold respirator masks (23/40, 57.5%), than with duckbill respirator masks (11/60, 18.3%), and hard shell respirator masks (1/30, 3.3%) (P < 0.05).
100
Fit-factor
150
The data was collected and processed with the Epi Info software version 6.04d. The 2 test was used to compare results among the different types of respirator masks. A value of P < 0.05 was considered as statistically significant.
50
0
Hard shell
Duckbill
Flat fold
Type of respirator mask Fig. 1. Distribution of fit-factors according to respirator mask shape. Distribution of fit-factors; median; + percentiles 25 and 75. Distribution des fit-factors en fonction de la forme de masques. Distribution des fit-factors; médiane; + percentiles 25 et 75.
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
fit-factors were observed for the nine respirator masks tested. A first profile concerned hard shell respirator masks (n = 3) and two duckbill respirator masks. For this profile, most fit-factors, 92%, were inferior or equal to 25 (46/50). A second profile was identified for flat-fold respirator masks (n = 2) and a duckbill respirator masks, with a comparable number of fit-factors inferior or superior to 100 equally distributed among these three respirator masks. The last duckbill respirator mask had a fit-factor distribution profile considered as intermediate. Table 2 lists, by institution, the number of individuals for whom at least one fit-test was successful in the consecutive trial of three types of respirator masks, including the respirator masks available in the institution. The latter systematically failed the fit-test in two institutions (a duckbill respirator mask and a hardshell respirator mask). In the third institution (duckbill respirator mask), the rate of failure was still 40% (4/10). The number of healthcare providers for whom a fit-test was successful increased with the number of respirator masks tested. Thus, after having tested three different types of respirator masks, the cumulative percentage of successful fit-tests was 70, 30, and 90 respectively in the three institutions. No 100% success rate was reached in any institution with the three tested respirator masks. 4. Discussion In Europe, the global effectiveness of a protective respirator mask must be assessed before it can be marketed. It must comply with standard EN 149: 2001 [3]. This standard includes assessing the filter’s effectiveness and total inner leak. The measures are performed with a panel of chosen individuals exposed to an aerosol while performing a series of exercises. Protective respirator masks have been used only recently in French healthcare institutions. Carrying a protective respirator mask was first recommended by the CSHPF, in 1993, for healthcare providers in contact with contagious patients presenting with tuberculosis [1]. Marketing of a class N95 protective respirator mask in the USA, with a filtration capacity equivalent to a class FFP2respirator, is possible after the single assessment of filter material effectiveness according to the American standard [8]. Adherence to the face is not mentioned in this standard. Thus, performing fit-tests is strongly recommended to choose a model adapted to the user, both for the industry and for hospitals [9], because of the lack of facial leak detection test in the standard. The National Institute of Occupational Safety and Health (NIOSH) requires a mandatory fit-test for any individual exposed to tuberculosis [10]. Furthermore, respirator masks with various shape and size may be available for healthcare providers. They can thus be given a respirator mask corresponding to their facial morphology. Protective respirator masks in French healthcare institutions are required to comply with standard EN 149: 2001 [3] (FFP1, FFP2, or FFP3). Nevertheless, “true” effectiveness criteria (adherence to the face) are not taken into account on the site. This is why, with the “tuberculosis” recommendations and the emergence of new respiratory risks (SRAS, flu pandemics, etc.), it seemed necessary to assess as objectively as possible,
267
the true effectiveness of respirator masks on the site, for facial air-tightness. Our results prove that flat-fold and duckbill respirator masks are better adapted for healthcare providers than hard-shell respirator masks. They are soft, and thus more easily adjustable to the face of healthcare providers, mostly women, who generally have finer features. But all the duckbill respirator masks are not equivalent for air-tightness. Indeed, they may have comparable results with flat-fold respirator masks or hard-shell respirator masks, partly because of the respirator mask design. All the duckbill respirator masks do have the same shape, but not the same quality of joint, of rubber bands, of nasal clip, etc. The hard-shell respirator masks, do not seem adapted to healthcare settings because they are too wide and rigid to be adjustable. A smaller size could be better adapted. A respirator mask may perfectly fit some healthcare providers and not be airtight for others because of the different morphology of faces. This is the case for hard shell respirator masks tested in our study which do not seem adapted (with fit-tests inferior of 25), except in one case (with a fit-factor of 200). This is why the shape and size of the respirator mask as well as the facial morphology of the user should always be taken into account when choosing a protective respirator mask. This adequacy is the condition for the respirator mask’s air-tightness and thus for an optimal protection of the carrier. Given the diversity of facial morphology in healthcare institution personnel, a single type of protective respirator mask is not sufficient to adequately protect all the personnel. Indeed, no fit-test was successful in two institutions where only one type of respirator mask was available. It was a hard shell respirator mask in one and a duckbill respirator mask in the other. In the third institution (duckbill respirator mask), a little less than half of the tests failed. After having tested three respirator masks, the percentage of individuals with at least one successful fit-test rose respectively from 0 to 30%, 0 to 90%, and 10 to 70% in the three institutions. The results of another study made in the SaintPierre University Hospital in Brussels in which 417 quantitative fit-tests were performed [11] correlates with our results; 67% of the personnel was protected by the respirator mask available in the institution. A better adjustment of the respirator mask could have made the test successful, whereas a number of fit-tests failed, especially for fit-factors close to 100 (six fit-factors ranged between 82 and 99, out of the 100 tests performed with flat-fold respirator masks and duckbill respirator masks). This better adjustment could be obtained with training sessions for respirator mask carriage. But this approach cannot improve the test results if the respirator mask is not adapted to facial morphology; in this case leaks will persist. Various authors [11,12] reported that healthcare providers were never given any specific instructions, 56% to 64% did not place the respirator mask adequately, for either the quantitative or the qualitative fit-test; and a respirator mask adapted to a morphology, chosen with a fit-test, is effective only if it is carried correctly. It thus seems mandatory to teach healthcare providers how to wear and adjust the respirator mask. These training sessions should be renewed regularly, especially for the most exposed personnel [12,13]. Using a fit-test for personnel
268
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269
Table 2 Number of individuals for whom at least one fit-test was successful during the consecutive use of three types of respirator masks. Nombre de personnes pour lesquelles au moins un fit-test était réussi lors de l’essai successif de trois types de masques. Number of respirator masks tested
Individuals with at least one successful fit-test n (%) Institution 1
Institution 2
Institution 3
1 respirator mask (= respirator mask available in the institution) 2 respirator masks 3 respirator masks
0b
0a
6/10 (60)c 7/10 (70)a
1/10 (10)b 3/10 (30)b
6/10 (60)b 9/10 (90)c 9/10 (90)a
a b c
Four different hard shell respirator masks. Four different duckbill respirator masks. Two different flat fold respirator masks.
who use a protective respirator mask occasionally may be discussed [12]. But, the fit-test is a much more accurate quantitative and objective test than visual observation which may be observer dependent. Healthcare providers should perform at least one fit-check every time they use a respirator mask, to ensure its adequate placement and air-tightness [4]. Our study had limitations. The fit-tests were not performed in real conditions. Indeed, the measured particles were not infectious. The tests lasted only 10 minutes whereas a healthcare provider may have to wear a protective respirator mask longer during caregiving. But the fit-test protocol was approved by the Occupational Safety and Health Administration OSHA [6] and has been applied since the late 90 s in the USA. The study sample may be considered weak (130 fit-tests), but this sample is larger than the one used for standard EN 149: 2001 (10 tests per respirator mask). Furthermore, the respirator masks were tested with healthcare providers and not on dummies or individuals with a chosen morphology as was the case in the European standard. Performing fit-tests in uncommon in French hospitals. But these tests may help answer a number of questions about the airtightness of respirator masks, questions that were often raised during the last flu pandemic.
Acknowledgements
5. Conclusion
References
This survey revealed that a single model of respirator mask is not sufficient to protect healthcare providers in a healthcare institution especially if its shape and/or size are not adapted. The technical training for protective respirator mask placement is essential but contributive only if the respirator mask is adjustable enough to fit the face. The flat-fold and duckbill respirator masks seem to be better adapted to healthcare settings than hard-shell respirator masks. Indeed, the hard shell respirator masks currently available in healthcare institutions are not safe enough because they are too large; they should not be used for caregiving when there is a risk of airborne transmission such as in tuberculosis. In a context of emerging viral risk of respiratory transmission, and of multiresistant tuberculosis, it is necessary to update French recommendations concerning respiratory protection. The need to have at least two types of respirator masks, with different shapes and/or sizes, should be discussed as well as the need for training and checking the chosen respirator mask’s
air-tightness: fit-test (at least initially when choosing the respirator mask) and/or at least a fit-check every time a respirator mask is used. Disclosure of interest The authors declare that they have no conflict of interest concerning this article.
The GERES would like to thank the Direction Générale de la Santé for its financial support, the firm TSI and the 3M Santé laboratories, and more especially Franck Ajuelos, for loaning the devices needed to perform the fit-tests, the various suppliers of respirator masks, as well as the participating institutions (GH Bichat-Claude Bernard, Paris–GH La PitiéSalpêtrière, Paris–Centre Médical de Bligny, Briis-sous-Forges) and the personnel who tested the respirator masks. Financing: Direction Générale de la Santé.
[1] Conseil Supérieur d’Hygiène Publique de France, section maladies transmissibles. Avis du CSHPF du 14 mars 2003 relatif au choix d’un masque de protection contre la tuberculose en milieu de soins. http://www.sante.gouv.fr/IMG/pdf/Avis du CSHPF du 14 mars 2003 relatif au choix d un masque de protection contre la tuberculose en milieu de soins-.pdf. (Accès 16 juin 2011). [2] Surveiller et prévenir les infections associées aux soins. HygièneS. 2010; XVIII (4). http://www.sf2h.net/publications-SF2H/SF2H surveilleret-prevenir-les-IAS-2010.pdf. (Accès 16 juin 2011). [3] Respiratory protective devices–Filtering half masks to protect against particles–Requirements, testing, marking. S 76-014. AFNOR 2001. [4] Ciotti C, Bouvet E, Abiteboul D, le GERES et l’INRS. Utilisation des masques de protection respiratoire chez les soignants. Med Mal Infect 2008;38:452–6. [5] Derrick JL, Chan YF, Gomersall CD, Lui SF. Predictive value of the user seal check in determining half-face respirator mask fit. J Hosp Infect 2005;59:152–5. [6] Occupational Safety and Health Administration (OSHA). Appendix A to §1910.134 Fit Testing Procedures (Mandatory). Washington DC: Office of the Federal Register, National Archives and Records Administration; 1998. [7] TSI, Inc. Portacount® Pro 8030 and Portacount® Pro + 8038 manual. St. Paul, Minnesota: TSI, Inc, 2010.
C. Ciotti et al. / Médecine et maladies infectieuses 42 (2012) 264–269 [8] National Institute for Occupational Safety and Health (NIOSH). 42 CFR 84 Respiratory protective devices: final rules and notice. Federal Register 60:110. US Centers for Disease Control and Prevention, NIOSH; 1997. [9] National Institute for Occupational Safety and Health (NIOSH) 29 CFR Part 1910.134. Code of Federal Regulations. Occupational Safety and Health Administration: Respiratory Protection. Washington DC: Office of the Federal Register National Archives and Records Administration; 1997. [10] National Institute for Occupational Safety and Health (NIOSH). TB respiratory protection program in health care facilities. Morgantown, WV: US
269
Department of Health and Human Services, Public Health Service, Centers for Disease Control, NIOSH; 1999. [11] Gérard M, Suys M. Fit-testing des masques respiratoires FFP2. Une expérience de terrain. Noso-Info 2006:2–3. [12] Lee MC, Takaya S, Long R, Joffe M. Respirator mask -fit testing: does it ensure the protection of healthcare workers against respirable particles carrying pathogens? Infect Control Hosp Epidemiol 2008;29:1149–56. [13] Kelly L, Clark K. The effectiveness of training and taste testing when using respirator mask. J Hosp Infect 2004;58:240–1.