Strategies for minimizing nosocomial measles transmission.

ReviewslAnalyses Strategies for minimizing nosocomial measles transmission* R.J. Biellik1 & C.J. Clements2 As a result of the highly contagious nature of measles before the onset of rash, nosocomial transmission will remain a threat until the disease is eradicated. However, a number of strategies can minimize its nosocomial spread. It is therefore vital to maximize awareness among health care staff that an individual with measles can enter a health facility at any time and that a continual risk of the nosocomial transmission of measles exists. The present review makes two groups of recommendations: those which are generally applicable to all countries, and certain additional recommendations which may be suitable only for industrialized countries. Introduction In the first decade of the 21st century, measles may well be the next disease after smallpox, poliomyeli- tis, and dracunculiasis to be targeted for global eradi- cation. However, until eradicated, measles remains a threat, especially to children admitted to health care facilities. The contribution of nosocomial transmission to the incidence of measles varies considerably, accord- ing to the prevailing measles epidemiology in an area. Several studies have demonstrated that nosocomial transmission is important in industrial- ized countries and in urban settings in developing countries. However, such transmission probably does not contribute significantly to measles inci- dence in rural communities in developing countries, where immunization coverage remains low to mod- erate and access to health facilities or their use by community members for diseases such as measles is limited. * A previous version of this article was issued as Measles control in the 1990s: minimizing nosocomial transmission. Geneva, World Health Organization, 1994 (unpublished document WHO/EPI/ GEN/94.6, available on request from Expanded Programme on Immunization, World Health Organization, 1211 Geneva 27, Switzerland). 1 EPI Advisor, World Health Organization, Harare, Zimbabwe. 2 Medical Officer, Expanded Programme on Immunization, Global Programme for Vaccines and Immunization, World Health Organi- zation, Geneva, 1211 Geneva 27, Switzerland. Requests for re- prints should be sent to this author. Reprint No. 5790 Until measles is eradicated it will be difficult to eliminate its nosocomial transmission because of cer- tain limiting factors, including the highly contagious nature of the disease in the incubation phase. How- ever, several strategies can minimize such transmis- sion. It is therefore vital to raise awareness among health care staff that an individual with measles can enter a health facility at any time, so that personnel remain alert to the risk of spread to non-immune persons. Measles is caused by one of the most infectious human pathogens, and is certainly the most con- tagious vaccine-preventable disease. This is the result of the virus's ability to remain viable for extended periods of time in small droplets expelled by infected individuals on coughing (1-3). Measles transmission is therefore facilitated by the collection of susceptible individuals in confined spaces where such infectious aerosols linger. Outside the home, high-risk settings include educational facilities, sports halls, religious institu- tions, orphanages, refugee camps, and health facili- ties (4, 5). Health facilities are of particular concern because of the risk of exposure to measles for those who may already be seriously ill. The use of a sterile syringe for every immunization or therapeutic injec- tion will ensure that the infection is not spread by this route, but such precautions do nothing to pre- vent aerosol transmission. This article reviews relevant literature, where possible quantifies the contribution of nosocomial transmission to the overall incidence of measles, and provides practical recommendations for strategies to minimize nosocomial transmission. Bulletin of the World Health Organization, 1997, 75 (4): 367-375 © World Health Organization 1997 367 R.J. Biellik & C.J. Clements Epidemiology of nosocomial measles Where measles virus circulates in a community, it is likely to be a common pathogen in health facilities, and a significant proportion of individuals may ac- quire the infection while visiting a health facility. A study of nosocomial infection conducted during 1946-47 in 26 children's wards in 14 representative hospitals throughout the United Kingdom found that 12% of all nosocomial infections were measles (6). A similar finding was reported among black children of low socioeconomic class admitted to a major teaching hospital in South Africa in 1987 (7). These studies may in fact have underestimated the proportion of nosocomial infections due to mea- sles since post-discharge surveillance was not always performed. The relative contribution of nosocomial trans- mission to the overall incidence of measles appears to vary in accordance with prevailing measles epi- demiology (Table 1). Analysis of published data suggests an increasing trend for the proportion of nosocomially acquired cases in certain settings. Published national surveillance data are avail- able for the USA for 1980-90, where high vaccina- tion coverage had reduced measles incidence to historically low levels and the majority of reported cases were associated with isolated outbreaks (8). Information on the probable source of exposure was, on average, available for half of the reported cases. Of these, <10% were reported as infected in health Table 1: Relative contribution of nosocomial transmission to measles incidence in industrialized and developing countries % cases associated with Country Year nosocomial spread Author, publication year (ref.) Industrialized countries National surveillance USA 1980 0.4 CDC,a 1981 (41) USA 1981 1.1 Davis et al., 1986 (40) USA 1982 1.2 Davis et al., 1986 (40) USA 1983 3.7 Davis et al., 1986 (40) USA 1984 2.5 Davis et al., 1986 (40) USA 1985 4.7 CDC, 1986 (42) USA 1987 5.8 CDC, 1989 (43) USA 1989-90 9.0 Gindler et al., 1992 (5) Outbreak investigations (community-based) USA (OK) 1981-85 27 Istre et al., 1987 (11) USA (CA) 1983 29 Dales et al., 1985 (10) USA (HI) 1984 15 CDC, 1984 (9) Outbreak investigations (health facility-based) France 1983-84 16 Foulon et al., 1986 (38) USA (Ml) 1985 57 Sienko et al., 1987 (15) USA (NM) 1986 43 CDC, 1987 (13) South Africa 1985-86 25 Reynolds et al., 1987 (14) USA (Ml) 1990 48 McGrath et al., 1992 (16) USA (CA) 1990 72 Mason et al., 1993 (17) Case-control USA (CA) 1988-89 23b Farizo et al., 1991 (18) USA (TX) 1988-89 41b Farizo et al., 1991 (18) Developing countries Outbreak investigations (community-based) Brazil 1983 21 PAHO,c 1984 (19) Outbreak investigations (health facility-based) Cameroon 1975 35 Guyer, 1976 (20) Mozambique 1980 33 Rabelo, 1982 (45) Guinea-Bissau 1980-82 16 Aaby et al., 1985 (21) Case-control C6te d'lvoire 1985 71b Klein-Zabban et al., 1987 (23) Burundi 1988 32 Chen, 1990 (24) a CDC = Centers for Disease Control and Prevention. b Significantly more cases than controls (46 of 64 compared with none of 21; P < 0.1) had visited a health facility during the 2 weeks prior to the onset of rash among cases. c PAHO = Pan American Health Organization. WHO Bulletin OMS. Vol 75 1997368 Minimizing nosocomial measles transmission facility settings during this period. However, our re- view of published data suggests that nosocomial transmission in the USA, as a proportion of all trans- mission, increased by a factor of 20, in a roughly linear fashion, across the same period. Although these data are potentially biased (a large proportion of cases had no reported source of exposure), it is unlikely that bias alone would explain the marked trend ob- served in the proportion of nosocomial spread. Data derived from measles outbreak investiga- tions in three industrialized countries (France, South Africa, and the USA), show the proportion of cases with exposure in health facility settings to range from 15% to 72%. This proportion was generally lower in community-based studies (9-11) than in studies of urban health-facility-based outbreaks or hospital- ized cases (12-17). This is as expected, since facility- based studies tend to over-represent community members who make greater use of health facilities, and therefore tend to overestimate the community infection rate. Case-control studies performed in the USA of measles transmission in Los Angeles and Houston showed that attendance at a hospital emergency room 10-18 days prior to the onset of measles rash was a significant risk factor (18). Among data from measles outbreak investiga- tions in six developing countries (Brazil, Burundi, Cameroon, C6te d'Ivoire, Guinea-Bissau, and Mo- zambique), the proportion of measles cases with ex- posure in health facility settings ranged from 21% to 71%. The proportion of cases associated with nosocomial spread was generally similar in a com- munity-based study (19) in Brazil and in studies of urban health-facility-based outbreaks or hospital- ized patients (20, 21). A case-control study per- formed in Abidjan showed that attendance at a hospital emergency room up to 2 weeks previously (the approximate incubation period for measles) was a significant risk factor for acquiring measles (22,23). However, a study in a rural area of Burundi found no difference in the proportion of cases and controls visiting a health facility during the previous month (24). A community-based study in Burundi found that only 1% of those with measles had visited a health facility during the previous month. In sum, these studies suggest that nosocomial transmission is important in industrialized countries and in urban settings in developing countries. Fur- ther, data from the USA suggest that the relative contribution of nosocomial transmission will in- crease as high vaccination coverage rates are achieved in more countries (25) and endemic (non- outbreak) measles declines in importance. These studies also suggest that nosocomial transmission does not contribute significantly to measles inci- dence in rural communities in developing countries where vaccination coverage remains low to moder- ate and where access to health facilities or their uti- lization by community members for diseases like measles is limited. Factors influencing the control of nosocomial measles Characteristics of transmission Measles is an acute, highly contagious disease trans- mitted by direct contact with infectious droplets (26). Immunity from the disease is lifelong; seroconver- sion following successful immunization is also for all practical purposes lifelong. The virus is secreted in the nasopharynx of sick individuals and expelled in aerosolized droplets on coughing. Several studies have demonstrated that infection can occur without face-to-face contact with a contagious individual. Measles transmission has been documented even when the contagious person has left the room up to 2 hours before the arrival of those subsequently acquiring infection (1, 3, 9,13,15, 27-29). In a U.S. high school, one child with a vigor- ous cough caused an explosive single-generation measles epidemic among 69 non-immune children, many of whom had no direct contact with the index case but shared the same corridors between classes (29). Since measles patients are contagious from 3-5 days before the onset of rash (1-2 days before the onset of fever and other symptoms), it can be diffi- cult to identify infectious individuals, many of whom exhibit non-specific symptoms. Such patients facili- tate nosocomial transmission and complicate its control. Age of patients and disease severity Measles transmission is facilitated by human proxim- ity and crowding (30), such as occurs in busy health clinics. In addition, there is an increased risk of se- vere measles in crowded conditions, such as occur in developing countries (31-36). Indeed, crowding (which is a proxy for the intensity of the viral dose) appears to be more important than nutritional status (a proxy for host resistance) in determining the se- verity of disease (33, 35). Other studies have shown that complication rates and case-fatality rates were significantly higher, and recovery times longer, in infants and young children acquiring measles nosocomially, compared with those acquiring it in the community (14, 37-39). In the USA during 1983-84, 20% of measles cases occurred in infants and children too young to WHO Bulletin OMS. Vol 75 1997 369 R.J. Biellik & C.J. Clements have been eligible for vaccination, whereas 59% of those with nosocomially acquired measles were in- eligible (40). Infants and young children may be at greater risk of nosocomial acquisition because of more frequent visits to medical facilities, for example for immunization and well-baby clinics. Characteristics of health facilities Health facilities play an important role in measles transmission, not only within their walls but also to the whole community, in both industrialized coun- tries (1, 3, 5, 9,11,13-18, 28, 38, 40-43) and develop- ing countries (7, 19-21, 23, 24, 38, 44, 45). A 5-year retrospective study in the USA during the early 1980s implicated all types of health care settings; the greatest proportions of nosocomially acquired mea- sles infections were in hospitals (47%), physicians' offices (35%), and outpatient clinics (15%) (40). During community-wide measles outbreaks in two cities in the USA, visiting a hospital emergency room 10-18 days prior to the onset of measles rash was a significant risk factor (19). Measles transmission also occurred in pharmacies, nonhospital emergency centres, a drug rehabilitation centre, and a labora- tory. A total of 45% of measles cases evaluated in another study in the USA resulted either directly or indirectly from exposure in medical waiting rooms (11). In developing countries, maternal and child health care clinics have been implicated in the nosocomial spread of measles (20, 38). Studies of measles outbreaks in the USA revealed that low relative humidity and a lack of fresh-air circulation in waiting rooms seem to facilitate measles transmis- sion (3). The role of health care staff While most nosocomially spread outbreaks of measles involve non-immune patients and visitors, non-immune health workers are often involved, es- pecially in countries where measles has not been endemic for many years (46-51). The above-men- tioned 5-year retrospective study in the USA during the early 1980s revealed that, among nosocomially transmitted measles cases, 50% were transmitted patient-to-patient, 37% patient-to-staff, 8% patient- to-visitor, 3% staff-to-staff, and 2% staff-to-patient (40). Staff acquiring measles most frequently were those who had contact with patients or laboratory specimens, including nurses, physicians, laboratory technicians, and clerk/receptionists. In contrast to industrialized countries, many health care workers in developing countries grew up before measles vacci- nation was widespread, are therefore likely to be immune from exposure (usually in childhood), and thus do not contribute to nosocomial transmission. Prevention through vaccination If all infants and children were immune to measles before exposure to a contagious individual in a medi- cal setting, nosocomial transmission could not occur. However, routine measles vaccination of infants 9 months of age or older will not protect susceptible infants under 9 months of age. Individuals who re- ceive measles vaccine but fail to respond with pro- tective levels of antibody will also not be protected. In addition, the ascertainment of measles vaccina- tion status among infants and children, especially in developing countries, can be complicated by low rates of vaccination-card retention. The prevention of nosocomially acquired measles through pre- exposure vaccination therefore requires additional vaccination strategies. Several studies have demonstrated that measles vaccine is effective in preventing the development of clinical measles in exposed individuals if vaccination is performed within 72 hours of exposure (52). The rate of protection varied between 68% and 100%. Four studies of household or institutional (including nosocomial) exposure reported protection rates of 94% or greater (59-62). However, the effectiveness of post-exposure prophylaxis among children under 1 year of age, particularly those less than 9 months of age, has not been studied. Nevertheless, these findings lend support to recommendations to vacci- nate individuals against measles within 72 hours of exposure (26). Recommendations for control Since measles is highly contagious during the pre- rash phase, it will be extremely difficult to eliminate its nosocomial transmission until the disease itself is eradicated. However, a number of strategies can minimize nosocomial spread. The following recom- mendations are divided into those which are appli- cable to all countries, and certain recommenda- tions which may be suitable only for industrialized countries. Measures applicable to all countries Maintain a high level of awareness among staff. It is vital to maximize awareness among health facility staff that an individual with measles can enter the facility at any time and that a continuous risk of the nosocomial spread of measles to non-immune per- sons exists. Maintain high coverage in the community. The accu- mulation of susceptible individuals in the community will be minimized by maintaining high rates of mea- WHO Bulletin OMS. Vol 75 1997370 Minimizing nosocomial measles transmission sles vaccination coverage. Several strategies for achieving high coverage have been described (25). A combination may be required, including routine vaccination, special actions for high-risk groups or areas, and mass campaigns. Since there are virtually no contraindications to measles vaccination, measles vaccine should be ad- ministered regardless of the patient's health status. Measles vaccination is particularly important for malnourished children and for those with chronic illnesses, as they are at increased risk of complica- tions due to measles. An exception to this recom- mendation are children who, on admission, are so ill that they are at serious risk of dying. Although ad- ministration of measles vaccine is not dangerous in such cases, parents may incorrectly attribute a death to the vaccination. Reduce vaccination age during outbreaks. WHO rec- ommends that the age for administration of measles vaccine should be lowered to 6 months during out- breaks where high attack rates are anticipated. Vac- cine should be administered any time after 6 months of age and, for infants vaccinated between 6 and 9 months of age, a second dose should be administered as soon as possible after 9 months of age, provided that at least 4 weeks have elapsed since the last dose of vaccine. The second dose after 9 months of age is important, as the serological response to vaccine given before 9 months of age may be significantly lower, resulting in lower levels of protection. Parents should receive appropriate instructions when an early dose is administered. (The recommendation for the re- vaccination of children vaccinated before 9 months of age also applies to the sections that follow.) Eliminate missed vaccination opportunities. Missed opportunities for vaccination should be eliminated by routinely checking the measles vaccination status of all children attending any health facility for any reason. This should be implemented at all immuniza- tion clinics, physicians' offices, public and private clinics, health centres, and hospital emergency and outpatient wards. Even children with a history of measles disease should be vaccinated, as physicians, health care workers, mothers, and other caretakers occasionally mistake other fever and rash illnesses for measles. In addition, a verbal history of previous measles vaccination is insufficient to contraindicate vaccination-only vaccination card documentation should be accepted. In developing countries, where vaccination cards or clinic records are often missing, local condi- tions should be evaluated when determining criteria for administering measles vaccine to children whose vaccination status cannot be verified. Where indi- cated by epidemiological analysis, including the assessment of case-fatality rate, the response to mea- sles outbreaks may include the vaccination of all children aged 6 or 9 months to 9 years of age, regard- less of measles vaccination history. In the absence of an outbreak, it is recommended that all children aged 9 months to 2 years, at a minimum, without documented evidence of vaccination should receive measles vaccine. A simple method is available for assessing the extent of missed vaccination opportunities (53). Systematic efforts should be made to eliminate all missed opportunities. Any child eligible for measles vaccine (or other antigens, and any woman eligible for tetanus toxoid) should receive any missing dose(s) before leaving a health facility. If a non- immune person is exposed to measles through di- rect or indirect contact with a contagious individual, measles vaccine will usually provide protection if administered within 72 hours of exposure. Ensure vaccination among hospitalized patients. The vaccination status of all children admitted to hospital should be rigorously checked. WHO recommends that, in developing countries, a dose of measles vac- cine should be given to all infants aged 6 months to 9 years (the age range to be adjusted in light of local conditions) upon admission to hospital, even if there is documented evidence of previous measles vaccina- tion. In most industrialized countries, only patients who lack vaccination documentation need be vacci- nated upon admission. At best, measles vaccination is associated with an efficacy of 80-90% in developing countries (19, 54-57). Thus, an extra dose of vaccine upon admis- sion to hospital can significantly reduce the likeli- hood of nosocomial measles transmission. The vaccination status of hospitalized children should also be checked before they are discharged from hospital. Vaccination of those without docu- mentation of previous measles vaccination reduces the chances of a child returning home while incubat- ing a nosocomially acquired measles infection. Fail- ure to do this could result in the infection of other children in the community. Non-immune persons over 6 months of age who are exposed to measles patients in the hospital, such as patients sharing the same ward and visitors, should receive a dose of measles vaccine, wherever possible, within 72 hours of exposure. For patients who are not immunocompromised, the use of hyper- immune measles gamma globulin (IMGG) may be less effective and is more costly than measles vaccine. Isolate individuals with fever and rash. To reduce the chance of exposure to measles, individuals with fever WHO Bulletin OMS. Vol 75 1997 371 R.J. Biellik & C.J. Clements and rash should preferably not enter the common waiting areas of a health facility. Wherever feasible, such patients should be fitted with a mask and taken directly to a room reserved for those with diseases subject to respiratory isolation. Moreover, if pos- sible, waiting and treatment areas should be well ventilated, and, to the extent possible, care should be taken to ensure that sick and well children do not sequentially use the same room and are not dealt with by the same staff for weighings, clinical exami- nations, vaccinations, or other consultations, as this could clearly provide a vehicle for measles transmis- sion. Where female literacy is common, a sign can be mounted outside health facilities instructing parents or guardians accompanying a child with fever and rash to wait outside and to ask another attendee to inform the health care staff that a sick child has arrived. To reduce the severity of measles and the risk of subsequent complications, vitamin A supplements (200000IU on two successive days for those >1 year of age) should be administered to all hospitalized children with suspected measles.a Inform health authorities. Measles is a reportable disease in almost all countries. Cases of measles should be reported promptly to the district health authorities in accordance with accepted local proce- dures. In addition, nosocomially acquired cases of measles should be reported immediately to the hospital's infection control authorities for investiga- tion and response such as increased isolation precau- tions for sick children. Control measures for industrialized countries Ensure vaccination of health facility staff. All health facility staff should be immune to measles. Most adults in developing countries will already have natural immunity. Particularly in industrialized countries, where some adults may not be immune, any staff member without documentary proof of measles vaccination or adequate measles antibody titres at the time of employment should receive mea- sles vaccination, after screening for contraindica- tions such as pregnancy or immune suppression. Administer gamma globulin to selected patients. Since there is a risk of overwhelming viraemia, live viral vaccines such as measles vaccine are a See: Using immunization contacts as the gateway to eliminating vitamin A deficiency: a policy document. Geneva, World Health Organization, 1994 (unpublished document WHO/EPI/GEN/94.9 Rev. 1, available on request from Expanded Programme on Immunization, World Health Organization, 1211 Geneva 27, Switzerland). contraindicated in individuals with congenital im- mune function disorders or those receiving immunosuppressive therapy. Hence, immunocom- promised individuals coming into contact with measles patients should receive IMGG as soon as possible after exposure. However, persons sero- positive for human immunodeficiency virus (HIV) or with suspected or confirmed acquired immuno- deficiency syndrome (AIDS) can receive live mea- sles vaccine (58). Conclusions Until measles is eradicated, its nosocomial transmis- sion remains a threat. Reducing through immu- nization the amount of virus circulating in the community remains the strategy of choice for com- bating measles. However, additional measures are available that will help reduce transmission of mea- sles in the health care setting. These strategies in- clude admission and discharge vaccination checks, vaccination of staff, and isolation of patients with fever and rash, particularly in epidemic situations. Strategies such as the triage of possible measles cases in hospital and health clinic waiting areas and engi- neering modifications to air flows, are potentially beneficial but are more difficult to implement and largely untested. Acknowledgements The authors acknowledge comments and suggestions on the manuscript provided by Dr F. Cufts, Dr J. Gindler, and Dr. J.-M. Olive. Resume Strategies de reduction du risque de transmission nosocomiale de la rougeole Due a plusieurs facteurs comme la grande conta- giosite de la maladie pendant la phase d'incubation - avant I'apparition des sympt6mes - la transmis- sion nosocomiale de la rougeole constituera une menace tant que cette maladie n'aura pas e 6limin6e ou 6radiqu6e. 11 est cependant possible, grace a diverses strat6gies, de reduire sa propaga- tion nosocomiale. 11 est indispensable que le per- sonnel de sante soit conscient qu'un malade atteint de rougeole peut se presenter dans un 6tablisse- ment de soins a tout moment et qu'il existe un risque permanent de transmission nosocomiale de cette maladie aux personnes non immunes. DNs que le virus de la rougeole circule dans une communaut6, il est probable qu'il devienne un agent pathogene courant dans les 6tablissements 372 WHO Bulletin OMS. Vol 75 1997 Minimizing nosocomial measles transmission de soins, et en fait une proportion notable des cas sont probalement contract6s lors d'une visite dans un tel 6tablissement. Bien que, globalement, moins de 10% des cas de rougeole pour lesquels une source probable d'exposition a ete identifiee aient 6te declar6s comme contract6s dans un 6tablisse- ment de soins aux Etats-Unis d'Am6rique entre 1980 et 1990, 1'examen des donn6es publi6es indique que la proportion de cas attribuables a une transmission nosocomiale a 6t6 multipliee par vingt, de fa,on pratiquement lin6aire, pendant cette meme periode. Une 6tude retrospective de 5 ans realis6e aux Etats-Unis d'Amerique au d6but des ann6es 80 a montr6 que tous les types d'6tablissements de soins 6taient impliqu6s, la plus forte proportion de rougeoles nosocomiales 6tant contractee a l'h6pital (47%), dans un cabinet medical (35%) ou dans un dispensaire (15%). Lors des flamb6es d'ampleur communautaire, les agents de sant6 non immuns sont souvent en cause, surtout dans les pays ou la rougeole n'a plus ete endemique depuis de nombreuses ann6es. Les donn6es rapportees indiquent donc que le contact nosocomial avec le virus est un important mecanisme de transmission de la rougeole dans les pays industrialises et dans les secteurs urbains des pays en d6veloppement. En revanche, ce mode de transmission ne contribue peut-etre pas de fa,on significative a l'incidence de la rougeole dans les communautes rurales des pays en d6veloppement ou la couverture vaccinale est encore faible a moyenne et ou l'acces aux 6tablissements de soins ou leur utilisation par les membres de la com- munaut6 pour des maladies telles que la rougeole sont limites. Les donnees am6ricaines suggerent neanmoins que la contribution relative de la trans- mission nosocomiale a la propagation g6n6rale de la rougeole ira en augmentant a l'echelle mondiale, au fur et a mesure de l'am6lioration de la couverture vaccinale dans de nombreux pays et du d6clin de la rougeole endemique (survenant en dehors des 6pid6mies). Les recommandations relatives a la lutte contre la transmission nosocomiale de la rougeole se divisent en deux groupes, celles qui sont appli- cables dans tous les pays et certaines recom- mandations compl6mentaires qui peuvent n'etre applicables que dans les pays industrialises. Mesures de lutte applicables dans tous les pays * Maintenir un taux 6leve de couverture vaccinale antirougeoleuse dans la communaute * Eliminer syst6matiquement les occasions man- quees de vaccination antirougeoleuse. * Veiller a ce que les agents de sant6 aient par- faitement conscience du risque de transmission nosocomiale. * Isoler des leur arrivee les personnes qui se pr6sentent a l'etablissement de soins avec de la fievre et une eruption cutan6e. * Verifier l'6tat vaccinal de tous les malades, a I'admission et a la sortie de l'h6pital. * Abaisser l'age de la vaccination a 6 mois lors des epid6mies ou des taux d'atteinte elev6s sont pr6vus. * Declarer les cas de rougeole aux autorites sanitaires. Mesures de lutte applicables uniquement dans les pays industrialises * Veiller a ce que tous les agents de sant6 soient vaccin6s contre la rougeole. * Administrer des gammaglobulines antirougeo- leuses hyperimmunes aux personnes immuno- d6prim6es expos6es a la rougeole. Tant que la rougeole ne sera pas 6radiquee, la transmission nosocomiale restera une menace. Face a cette situation, la strat6gie de choix consiste a r6duire, grace a la vaccination, la quantit6 de virus en circulation dans la communaut6. Des mesures complementaires peuvent cependant contribuer a r6duire le risque de transmission de la maladie dans les 6tablissements de soins. On peut ainsi vacciner les patients hospitalises a I'admission et a la sortie, vacciner le personnel, et isoler les malades, en particulier en periode d'epid6mie. D'autres mesures, comme I'adoption de salles d'attente separees pour les patients suspects de rougeole, et des amenagements des systemes de circulation et de filtration de l'air, bien que d'un grand int6ret potentiel, sont plus difficiles a mettre en oeuvre et doivent encore etre test6es. References 1. Bloch AB. Measles in a private physician's office. In: Proceedings of the 16th Immunization Conference. Atlanta, GA, Centers for Disease Control and Preven- tion, 1981: 59-63. 2. Langmuir AD. Changing concepts of airborne infec- tion of acute contagious diseases: a reconsidera- tion of classical epidemiologic theories. Annals of the New York Academy of Sciences, 1980, 353: 35- 44. 3. Remington PL et al. Airborne transmission of mea- sles in a physician's office. Joumal of the American Medical Association, 1985, 253: 1574-1577. WHO Bulletin OMS. Vol 75 1997 373 R.J. Biellik & C.J. Clements 4. Markowitz LE et al. Patterns of transmission in mea- sles outbreaks in the United States, 1985-86. New England journal of medicine, 1989, 320: 75-81. 5. Gindler JS et al. Epidemiology of measles in the United States in 1989 and 1990. Pediatric infectious disease joumal, 1992, 11: 841-846. 6. Watkins AG, Lewis-Faning E. Incidence of cross- infection in children's wards. British medical joumal, 1949, 2: 616-619. 7. Cotton MF et al. Nosocomial infections in black South African children. Pediatric infectious disease joumal, 1989, 8: 676-683. 8. Measles control in the 1990s: principles for the next decade. Geneva, World Health Organization, 1990 (unpublished document WHO/EPI/GEN/90.2, avail- able on request from Expanded Programme on Immu- nization, World Health Organization, 1211 Geneva 27, Switzerland). 9. Measles - Hawaii. Morbidity and mortality weekly report, 1984, 33(50): 702, 707-711. 10. Dales LG, Kizer KW. Measles transmission in medi- cal facilities. Westem journal of medicine, 1985, 142: 415-416. 11. Istre GR et al. Measles spread in medical settings: an important focus of disease transmission? Pediatrics, 1987, 79: 356-358. 12. Foulon G et al. Preventing the spread of measles in children's clinics. Lancet, 1983, 2: 1498-1499. 13. Measles transmitted in a medical office building New Mexico, 1986. Mortality and morbidity weekly report, 1987, 36(2): 25-27. 14. Reynolds LG, Klein M. The hospital as a vector of measles in the community. South African medical journal, 1987, 71: 637-638. 15. Sienko DG et al. A measles outbreak at university medical settings involving health care providers. American journal of public health, 1987, 77: 1222- 1224. 16. McGrath D et al. Analysis of a measles outbreak in Kent County, Michigan in 1990. Pediatric infectious disease journal, 1992, 11: 385-389. 17. Mason WH et al. Epidemic measles in the post vac- cine era: evaluation of epidemiology, clinical presen- tation and complications during an urban outbreak. Pediatric infectious disease joumal, 1993, 12: 42-48. 18. Farizo KM et al. Paediatric emergency room visits: a risk factor for acquiring measles. Pediatrics, 1991, 87: 74-79. 19. PAHO. Investigation of a measles outbreak in Brazil's Federal District. Bulletin of the Pan American Health Organization, 1984, 18: 397-400. 20. Guyer B. Transmission de la rougeole a Yaounde en 1975. M6decine tropicale, 1976, 36: 451-454. 21. Aaby P et al. Introduction of measles into a highly immunized West African community: the role of health care institutions. Journal of epidemiology and commu- nity health, 1985, 39: 113-116. 22. Expanded Programme on Immunization: nosocomial measles. Weekly epidemiological record, 1986, 61(44): 338-340. 23. Klein-Zabban ML et al. Frequence des rougeoles nosocomiales dans un centre de protection maternelle et infantile d'Abidjan. Bulletin of the World Health Organization, 1987, 65: 197-201. 24. Chen RT. Report on epidemiologic studies conducted for the 1988-89 measles outbreak, Muyinga Health Sector, Burundi. Unpublished report, Centers for Dis- ease Control and Prevention, Atlanta, GA, 1990. 25. Revised plan of action for global measles control. Geneva, World Health Organization 1993 (unpub- lished WHO document EPI/GAG/93/WP.10, available upon request from Expanded Programme on Immuni- zation, World Health Organization, 1211 Geneva 27, Switzerland). 26. Report of the Committee on Infectious Diseases, 22nd ed. Elk Grove Village, IL, American Academy of Pediatrics, 1991. 27. Interstate importation of measles following transmis- sion in an airport - California, Washington, 1982. Morbidity and mortality weekly report, 1983, 32(16): 210, 215-216. 28. Imported measles with subsequent airborne transmis- sion in a pediatrician's office - Michigan. Morbidity and mortality weekly report, 1983, 32(31): 401-403. 29. Chen RT et al. An explosive point-source measles outbreak in a highly vaccinated population. Journal of epidemiology, 1989, 129: 173-182. 30. Greenwood M. Epidemics and crowd-diseases: mea- sles. Reviews of infectious diseases, 1988, 10: 492- 499. 31. Sharma RS et al. An epidemiological investigation of a measles outbreak in Alwar-Rajastan. Journal of communicable diseases, 1984, 16: 299-302. 32. Aaby P et al. Overcrowding and intensive exposure as determinants of measles mortality. American jour- nal of epidemiology, 1984, 120: 49-63. 33. Aaby P et al. High measles mortality in infancy re- lated to intensity of exposure. Joumal of pediatrics, 1986,109:40-44. 34. Aaby P et al. Further community studies on the role of overcrowding and intensive exposure on measles mortality. Reviews of infectious diseases, 1988, 10: 474-491. 35. Lamb WH. Epidemic measles in a highly immunized rural West African (Gambian) village. Reviews of in- fectious diseases, 1988, 10: 457-462. 36. Garenne M, Aaby P. Pattern of exposure and mea- sles mortality in Senegal. Journal of infectious dis- eases, 1990, 161: 1088-1094. 37. Glyn-Jones R. Measles vaccine and gamma globulin in the prevention of cross infection with measles in an acute paediatric ward. Central African joumal of medi- cine, 1972, 18: 4-9. 38. Foulon G et al. Transmission and severity of measles acquired in medical settings. Journal of the American Medical Association, 1986, 256: 1135-1136. 39. Hussey G, Simpson J. Nosocomial bacteraemia in measles. Pediatric infectious disease journal, 1990, 9: 715-717. 40. Davis RM et al. Transmission of measles in medical settings: 1980 through 1984. Journal of the American Medical Association, 1986, 255: 1295-1298. 41. Measles in medical settings. Morbidity and mortality weekly report, 1981, 30(11):125-126. 374 WHO Bulletin OMS. Vol 75 1997 Minimizing nosocomial measles transmission 42. Measles - United States, 1985. Morbidity and mor- tality weekly report, 1986, 35(22): 366-370. 43. Measles - United States, 1988. Morbidity and mor- tality weekly report, 1989, 38(35): 601-605. 44. Gao JP, Malison MD. The epidemiology of a measles outbreak on a remote offshore island near Taiwan. Intemational joumal of epidemiology, 1988, 17: 894- 898. 45. Rabelo AL. [Observations on an outbreak of mea- sles in a peripheral area of the city of Maputo]. Revista medica de Mocambique, 1982, 1(1): 9-13 (in Portuguese). 46. Enguidanos R, Mascola L, Frederick P. A survey of hospital infection control policies and employee mea- sles cases during Los Angeles County's measles epi- demic, 1987 to 1989. American joumal of infection control, 1992, 20: 301-304. 47. Gurevich I, Barganza RA, Cunha BA. Measles: les- sons from an outbreak. American journal of infection control, 1992, 20: 319-325. 48. Navarrete-Navarro S et al. [Nosocomial measles: proposals for control in hospitals]. Boletin medico del hospital infantil de Mexico, 1990, 47: 495-499 (in Spanish). 49. Rivera ME et al. Nosocomial measles infection in a pediatric hospital during a community-wide epidemic. Journal of pediatrics, 1991, 119: 183-186. 50. Schwarcz S, McCaw B, Fukushima P. Prevalence of measles susceptibility in hospital staff. Archives of internal medicine, 1992, 152: 1481-1483. 51. Subbarao EK, Amin S, Kumar ML. Pre-vaccination serologic screening for measles in health care workers. Journal of infectious diseases, 1991, 163: 876-878. 52. Post-exposure prophylaxis. Geneva, World Health Organization, 1993 (unpublished document EPI/ TECHCOM/WP/93.19, available on request from Ex- panded Programme on Immunization, World Health Organization, 1211 Geneva 27, Switzerland). 53. Protocol for the assessment of missed opportunities for immunization. Geneva, World Health Organiza- tion, 1986 (unpublished document WHO/EPI/GEN/ 86.6, available on request from Expanded Pro- gramme on Immunization, World Health Organization, 1211 Geneva 27, Switzerland). 54. Hull HF, Williams PJ, Oldfield F. Measles mortality and vaccine efficacy in rural West Africa. Lancet, 1983, 1: 972-975. 55. Expanded Programme on Immunization: measles vaccine efficacy. Weekly epidemiological record, 1984, 59(17): 127-128. 56. Dabis F et al. The epidemiology of measles in a partially vaccinated population in an African city: impli- cations for immunization programs. American journal of epidemiology, 1988, 127: 171-178. 57. Thapa AB. Measles outbreak in a partially vaccinated population and measles vacine efficacy, South Lalitpur District, Nepal, 1990. Joumal of the Nepal Medical Association, 1992, 30: 154-158. 58. Clements CJ, von Reyn CF, Mann JM. HIV infec- tions and routine childhood immunization. Bulletin of the World Health Organization, 1987, 65: 905- 911. 59. Watson GI. Protection after exposure to measles by attenuated vaccine without gamma-globulin. British medical joumal, 1963, 1: 860-861. 60. Berkovich A, Starr S. Use of live-measles-virus vaccine to abort an expected outbreak of measles within a closed population New England journal of medicine, 1963, 269: 75-77. 61. Ruuskanen 0, Salmi TT, Halonen P. Measles vacci- nation after exposure to natural measles. Journal of pediatrics, 1978, 93: 43-45. 62. Addis DG, Berg JL, Davis JP. Revaccination of pre- viously vaccinated siblings of children with measles during an outbreak. Journal of infectious diseases, 1988, 157: 610-611. WHO Bulletin OMS. Vol 75 1997 375