Measles is a vaccine-preventable disease. It is transmitted via respiratory droplets and is highly infectious.1 Although it presents with nonspecific symptoms, it is highly virulent and can potentially cause severe complications in almost any organ such as pneumonitis, acute respiratory distress syndrome, encephalitis, postinfectious encephalomyelitis, subacute sclerosing panencephalitis and death.1,2
The Singapore national immunization schedule for measles, mumps and rubella (MMR) vaccination is for the first dose to be given at 12 months of age with a booster dose given between 15 and 18 months of age. Despite a highly successful vaccination program in Singapore with high vaccination rates for measles,3 the number of pediatric patients hospitalized for measles has been increasing in recent years.4 Vaccine refusal or nonvaccination has been associated with increased incidence and outbreaks in different parts of the world.5–10 Furthermore, Singapore has well-travelled population and hosts a high volume of international tourists and foreign workers with incomplete or unknown vaccination history.
The Children's Emergency Department (ED) at KK Women's and Children's Hospital has an annual attendance of >150,000 patients. Pediatric patients who are suspected to have highly infectious diseases are ideally identified at triage and isolated to the negative-pressure consultation rooms in a separate annex.
There has been no study known to date that looks at early identification and appropriate isolation of community-acquired measles cases presenting to our ED. This study aims to describe the epidemiologic trends of microbiologically proven measles cases that failed isolation at triage and subsequently on admission. We also aim to identify risk factors that can lead to early and appropriate isolation of suspected cases at triage to minimize risk of nosocomial transmission and spread.
This is a retrospective observational cohort study. Retrospective chart review was done for patients who were admitted from our ED from January 2010 to December 2016, with the International Classification of Diseases codes for measles, confirmed by microbiologic laboratory studies with immunofluorescence or polymerase chain reaction from nasopharyngeal swab specimen.
Patients who were immunocompromised (congenital or acquired) were excluded from this study because they may have atypical presentations and would be isolated if febrile. We collected data on demographics, clinical features, isolation history (isolation initiated at triage, at ED consult, on admission or missed cases which were identified only during inpatient), length of stay, highest level of care, reattendance rates at ED, day of illness at presentation, history of sick contact and immunization history.
All analyses were carried out using IBM Statistical Package for Social Sciences version 25 (SPSS Inc, Chicago, IL). Descriptive statistics for numerical variables are presented as mean (SD), range or median (interquartile range) when appropriate and n (%) for categorical variables. The differences in day of illness presentation and duration of admission between isolation at triage or admission were analyzed using t test when normality and homogeneity assumptions were satisfied, otherwise, Mann-Whitney U test was used. Sensitivity, specificity and positive and negative predictive values were specified for rash and koplik's spots on isolation. Statistical significance was set at P < 0.05.
The study was approved by the hospital ethics board (SingHealth Centralised Institutional Review Board).
In the 7-year study period, 277 patients were admitted for measles infection (Table 1). The patients' age ranged from 12 days of age to 15 years of age. The median age was 13 months (interquartile range, 9–24 months). One hundred ninety-one (69.0%) patients were 18 months of age or younger. One hundred fifty (54.2%) were males. Majority of patients were Chinese (43.3%), followed by Malays (34.0%), Indians (7.9%) and the rest (14.8%) from other ethnic groups. One hundred forty-three (51.6%) of patients or their parents reported poor feeding.
Of the 277 cases who presented to our ED, 177 (63.9%) patients were not isolated at triage (Table 1). Five patients with measles who were initially isolated at triage were subsequently de-isolated upon admission (Table 1). A review of their case records showed that their rash was felt to be atypical of measles although all 5 patients were infants with incomplete MMR vaccination. Incomplete vaccination was defined as having received <2 doses of MMR vaccination at the time of presentation.
The average day of illness at presentation was 4.3 days (Table 1). Those who were not isolated on admission presented earlier, at an average of 3.9 days of illness compared with those were isolated on admission, who presented at an average of 4.5 days of illness (P = 0.034). For 75 (27.1%) patients who had no rash at their initial ED presentation (Table 2), the average day of illness was 3.1 compared with an average of 4.8 days for those who had a rash at presentation (p < 0.001). The main reason for hospitalization for these 75 patients without rash at ED presentation was poor feeding with concern for dehydration. The median age for those with poor feeding was 12.5 months (interquartile range, 9–21.5 months).
Eighty-one (29.2%) patients reattended the ED before getting admitted. After both initial and repeated ED consultations, a total of 92 (33.2%) patients failed to be isolated on admission (Table 1). These 92 patients had no known prior measles contact. Exposure surveillance and contact tracings for all exposed healthcare workers and potential patient exposure were performed on laboratory confirmation of all these cases.
The average duration of hospitalization was 3.4 days (Table 1). Those who were not isolated on admission had a longer duration of hospitalization of average 4.2 days compared with an average hospitalization of 3.0 days for those who are isolated on admission (P < 0.001).
Six patients required admission to high-dependency unit for respiratory or circulatory support and monitoring. None of the patients required admission to the intensive care unit, and we observed no mortality.
One hundred eighty-five (66.8%) reported no known contact with persons having measles infection. On contact tracing after the patients were admitted, before laboratory confirmation of disease, 39 (14.1%) patients had contact with siblings with measles, 7.3% had contact with a caregiver with measles and 5.8% were exposed to a known relative with measles. Of these known relatives with measles, 13 were adult contacts and 3 were children. Seven (2.5%) patients were exposed to measles at a shelter home. Ten (3.6%) patients had positive contact in their preceding hospitalization.
Of all 277 admitted cases, only 1 (0.4%) patient had completed 2 doses of measles vaccination as per Singapore's immunization schedule (Table 2). Of those who had no or incomplete vaccination, 208 (75.1%) patients were 24 months of age or younger.
At initial presentation to the ED, 75 (27.1%) patients did not present with rash, of which 55 patients were 2 years of age or younger (Table 2). Two hundred and three (73.3%) patients had no conjunctivitis. Koplik's spots were not found in most patients (253 [91.3%]), at their initial ED presentation. Only 71 (21.6%) patients had the classical triad of fever, rash and conjunctivitis at first presentation to the ED.
In this retrospective study, we found that, as expected, incomplete immunization is a key risk factor in acquiring measles infection with 99.6% of admitted patients having no prior or incomplete measles vaccination. This potentially places children <18 months of age at higher risk for measles infection despite the local high uptake in vaccination and strict vaccination policies. The Singapore ministry of health reported an increase of first dose measles immunization in children by 2 years of age to have increased from 93.2% coverage in 2003 to 95.1% coverage in 2015.3,11 However, second dose of measles vaccination in children by 2 years of age was 89.5% in 2015, still short of the World Health Organization target of 95% coverage.3,12
The ethnic distribution of patients with measles was different from with the composition of residents in Singapore where Chinese make up 74.3%, Malays 13.4%, Indians 9.0% and other ethnicities 3.2% as reported by the Singapore Department of Statistics in September 2017.13
More than a quarter of patients reattending ED before admission could be due to initial attendance at very early stage of illness. Other factors that may have triggered the need for repeat consultation include, but are not limited to, persistent fever, onset of new symptoms, poor feeding and caregiver concerns regarding child's well-being.
One of the main study findings was that over a quarter of patients had no rash at initial ED presentation. This made successful isolation of measles patients extremely challenging. Of these, 3 had fever with conjunctivitis only at their initial ED presentation (Table 2). A significant proportion of these patients was younger (<24 months) and presented earlier in the course of their illness (mean, 3.09 days) and were hospitalized for main concerns of poor feeding. This could also account for the longer duration of hospitalization for those who were not isolated initially as they tend to be younger with concerns of feeding and presented earlier (with no rash) in the course of their infection.
Koplik's spots, while pathognomonic of measles, may be difficult to identify. They are transient and are only present for 3–4 days, appearing 1–4 days before the widespread rash.1 In our study, only 8.7% were observed to have koplik's spots with excellent specificity and positive predictive values (100% for both) but poor sensitivity and negative predictive value (13.0% and 36.4%, respectively). It would be challenging, however, for ED triage staff to identify this classical clinical feature, especially for the younger patients who have yet to complete their MMR vaccination (<15–24 months of age in our local population).
Despite the lack of initial isolation in 92 (33.2%) of the admitted patients, there was no large outbreak of measles infection in the ED or inpatient setting. This may be due to strict cross-infection and hygiene practices, mandatory documented healthcare worker measles immunity status and herd immunity.
Immunization is vital in the battle against measles. Due to unsubstantiated and nonevidence-based community beliefs on the adverse effects attributed to MMR vaccination, year on year, there are increasing number of patients infected with measles globally. Even Singapore, which has an extremely high uptake of MMR vaccination, has not been spared of the effects of this global increase in measles cases. This is especially concerning for patients who have not had their second booster of MMR before 24 months of age or were unable to receive MMR for medical reasons.
Vigilance, early identification and isolation are still needed to minimize risk of acquiring measles in the ED and hospital, but with a significant number of measles patients presenting without a rash and earlier in their illness, these strategies may be insufficient. Physicians should have a high index of suspicion in patients presenting with fever with rash or conjunctivitis and incomplete measles immunization history. We hope that this study will help heighten awareness and global need for MMR immunization. The substantial socioeconomic cost in the event of an outbreak has been well described in the literature.14–18 The importance of healthcare workers being fully immunized and global community herd immunity is also emphasized, given the findings of this study.
The study should be interpreted based on its limitations, primarily those that are inherent with a single-center, retrospective analysis. The data collected include information from the hospital infectious disease surveillance and screening. The data collected were, however, limited to only patients confirmed to have measles infection. Active data capturing with the use of a case investigation surveillance form would have improved the collection of symptom data.
We were unable to include patients who were clinically well and were discharged from the ED as microbiologic testing to confirm presence of measles was not available. Measles is a mandatory notifiable disease in Singapore, and if any cases revisited other hospitals or clinics, these cases would have been flagged up and contact tracing would be done including close contacts in healthcare facilities.
As it was a retrospective study, we were unable to probe further into the factors contributing to incomplete immunization for patients above 24 months of age for whom MMR was not contraindicated such as socioeconomic factors, parental education levels, vaccine knowledge and acceptance levels, accessibility to healthcare facilities that provide immunization or medical conditions preventing timely immunization. These are important concerns to address, especially for the older children who have missed their scheduled childhood immunization and require catch-up vaccinations.
The applicability of this study for other countries may be limited due to the high uptake of MMR in our local population and the differences in the childhood immunization schedule.
Our study found that a significant number of patients with measles present with nonspecific symptoms at an early phase of illness. Majority of measles patients were <24 months of age who had incomplete vaccination as they were not due for their MMR vaccination. It is challenging to successfully isolate all potential suspected measles patients in the ED setting even with strict isolation policy.
1. Perry RT, Halsey NA. The clinical significance of measles
: a review. J Infect Dis. 2004;189(suppl 1):S4–S16.
2. Rafat C, Klouche K, Ricard JD, et al. Severe measles
infection: the spectrum of disease in 36 critically ill adult patients. Medicine (Baltimore). 2013;92:257–272.
4. Vemula VN, Li L, Thoon KC, et al. Risk factors and clinical profile of measles
infection in children in Singapore. Infect Dis Health. 2016;21:192–196.
5. Goh D, Chew F, Khor S, et al. Resurgence of measles
in Singapore: profile of hospital cases. J Paediatr Child Health. 1999;35:493–496.
6. Porretta A, Quattrone F, Aquino F, et al. A nosocomial measles
outbreak in Italy, February-April 2017. Euro Surveill. 2017;22:30597.
7. Six C, Blanes de Canecaude J, Duponchel J, et al. Spotlight on measles
outbreak in the Provence-Alpes-Côte d'Azur region, France, January to November 2010 - substantial underreporting of cases: measles
outbreak in the Provence-Alpes-Cote d Azur region, France, January to November 2010 - substantial underreporting of cases. Euro Surveill. 2010;15:19754.
8. Parent du Châtelet I, Antona D, Freymuth F, et al. Spotlight on measles
2010: update on the ongoing measles
outbreak in France, 2008-2010. Euro Surveill. 2010;15:19656–XXX.
10. Brown KF, Kroll JS, Hudson MJ, et al. Factors underlying parental decisions about combination childhood vaccinations including MMR: a systematic review. Vaccine. 2010;28:4235–4248.
14. Coughlin M, Beck A, Bankamp B, et al. Perspective on global measles
epidemiology and control and the role of novel vaccination strategies. Viruses. 2017;9:11.
15. Yip WC, Vellayappan K, Wong HB. Measles
: mortality and morbidity in Singapore children. J Singapore Paediatr Soc. 1979;21:191–200.
16. Ghebrehewet S, Thorrington D, Farmer S, et al. The economic cost of measles
: healthcare, public health and societal costs of the 2012-13 outbreak in Merseyside, UK. Vaccine. 2016;34:1823–1831.
17. Fiebelkorn AP, Seward JF, Orenstein WA. A global perspective of vaccination of healthcare personnel against measles
: systematic review. Vaccine. 2014;32:4823–4839.
18. Botelho-Nevers E, Cassir N, Minodier P, et al. Measles
among healthcare workers: a potential for nosocomial outbreaks. Euro Surveill. 2011;16:19764.