eMJA: Lambert et al, Enhanced measles surveillance during an interepidemic period in Victoria

Public Health

Enhanced measles surveillance during an interepidemic period in Victoria

Stephen B Lambert, Heath A Kelly, Ross M Andrews, Mike C Catton, Pauline A Lynch, Jennie A Leydon, Debbie K Gercovich, Geoffrey G Hogg, Melissa L Morgan and Rosemary A Lester

MJA 2000; 172: 114-118
For related article see McIntyre et al

→ Other articles have cited this article

Abstract - Methods - Results - Discussion - Acknowledgements - References - Authors' details
Make a comment - Register to be notified of new articles by e-mail - Current contents list - More articles on Infectious diseases and parasitology

Abstract

Objective: To describe results of the first two years of enhanced measles surveillance in Victoria.
Design: Case series identified through enhanced measles surveillance.
Participants and setting: All measles cases notified to the Disease Control Section, Department of Human Services, Victoria, in 1997 and 1998.
Main outcome measures: Proportion of notified cases laboratory confirmed as measles, rubella, or human parvovirus infection; identification of clusters (two or more linked cases of measles); and utility of the National Health and Medical Research Council clinical case definition for suspected measles.
Results: Rates of laboratory testing of notified cases improved after introduction of a paediatric phlebotomy service in July 1997, from 21 of 90 notified patients (23%) in the preceding six months, to 258 of 317 notified patients (81%) between July 1997 and December 1998. Of the 317, only 19 (6%) were laboratory confirmed with measles, while a further 26 (8%) were laboratory confirmed with human parvovirus infection (18) or rubella (8). Three clusters of measles, involving 11 cases, were identified during 1998. Use of the NHMRC case definition did not greatly improve the positive predictive value for diagnosis of measles above that of notification alone (14% versus 8%).
Conclusions: Circulation of measles virus in Victoria in 1997 and 1998 appeared minimal. In this interepidemic period most notified cases of measles were not measles; to identify true cases, surveillance during an interepidemic period must include laboratory testing of notified cases.

Cases of classical measles are uncommon in countries with successful measles control programs, making clincal diagnosis less reliable. To ensure the validity of clinical notifications, it is increasingly important to confirm the diagnosis in every sporadic case of measles and in at least one case in every chain of transmission in such countries.¹

Australia suffered a nationwide outbreak of measles in 1993-1994.² Since 1994, a two-dose measles-mumps-rubella (MMR) vaccination program has been implemented,³ and in 1998 a national campaign targeting primary school-aged children was conducted.⁴ The country was free of any substantial outbreak until early 1999, when importation of the disease from Bali resulted in measles cases, mainly among young adults in Victoria.⁵

To monitor the success of the measles control program, the State of Victoria began a state-based enhanced surveillance program in 1997. This program concentrates on confirming the diagnosis of measles for all notifications received by the Disease Control Section of the Victorian Department of Human Services.⁶ We report the results of the first two (interepidemic) years of this enhanced measles surveillance program and make recommendations for the investigation of notified cases of measles.

Methods

The enhanced measles surveillance strategy adopted by Victoria has been reported elsewhere.⁶ In brief, all notifications to the Department of Human Services in 1997 and 1998 were followed up by a structured telephone interview with the patient or, if the patient was a child, with the parent or guardian. Demographic data, clinical symptoms, and measles vaccination history were recorded. The parent/guardian was asked to read the date of vaccination from the personal vaccination record when available. We attempted to identify a possible source of infection, as well as contacts who required advice about immunoglobulin or MMR vaccination. Suspected preceding or subsequent cases were followed up in a similar manner to identify clusters of infection (defined as two or more epidemiologically linked cases⁷). A sporadic case was one that could not be linked to another case.

Serological testing

A serum specimen was sought from each notified patient for laboratory confirmation of the clinical diagnosis. From July 1997, this specimen was obtained by a paediatric phlebotomist in the patient's home. Some patients provided a combined throat and nose swab and a urine specimen for viral culture or isolation of genetic material by polymerase chain reaction, and subsequent virus genotyping.⁸

Sera were tested for measles IgM and IgG at the Victorian Infectious Diseases Reference Laboratory (VIDRL) or, if original testing was performed elsewhere, the testing laboratory was asked to forward remaining sera from measles IgM-positive specimens to VIDRL for confirmatory testing. Testing at VIDRL used a commercial enzyme immunoassay (Dade Behring Enzygnost, Marburg, Germany). The manufacturer reports the measles IgM assay as having a sensitivity of 100% and specificity of 98%.

Sera that were negative for measles IgM at VIDRL were assayed for human parvovirus IgM and IgG (Biotrin Parvovirus B19 Enzyme Immunoassay, Dublin, Ireland), rubella IgM (DiaSorin ETI-RUBEK-M reverse PLUS, Saluggia, Italy) and rubella IgG (Panbio Rubella IgG ELISA Test, Brisbane, Australia).

Analyses

Using a defined algorithm,⁶ each notified case was classified as confirmed measles or otherwise according to the criteria in Box 1. These included serological and other results, as well as concordance with the clinical case definition for suspected measles¹¹ recommended by the National Health and Medical Research Council (NHMRC) -- morbilliform rash, fever present at rash onset, and cough.¹²

Analysis was performed using Epi Info version 6.04.¹³ Significance of differences between categorical data was tested by the Fisher's exact or ² test.

Results

In the first six months of surveillance (January to June 1997), sera were collected from 21 of 90 notified patients (23%). After employment of a paediatric phlebotomist to collect samples in the patient's home, collection rates improved progressively -- sera were collected from 258 of 317 notified patients (81%) between July 1997 and December 1998, including from 107/120 (89%) in the second half of 1998.⁶

Because of the lower rate of specimen collection in the first six months of surveillance, we analysed data for July 1997 to December 1998 only. In this period, only 19/317 notifications (6%) were classified as laboratory confirmed (Box 2). The remainder were laboratory rejected (229; 72%), clinically compatible (12; 4%), not clinically compatible (41; 13%) and not classifiable (16; 5%). All epidemiologically linked cases were able to be laboratory confirmed. Of the 229 cases that were laboratory rejected as measles, 18 had human parvovirus infection (8%), and eight had rubella (3%).

Box 3 shows serological results by age group. Serum collection rates did not differ significantly between age groups (P = 0.4), but laboratory confirmation was significantly more likely among patients aged 10 years or over than among younger children (P = 0.0002).

Clusters of measles

Three clusters of measles, involving 11 patients, were identified, all in 1998. The first, involving four people, began in January 1998. A 19-year-old man from New South Wales visited Melbourne soon after illness onset on 10 January. Three other people were infected: his 22-year-old brother (onset, 18 January), six-month-old nephew (onset, 1 February), and a 23-year-old male household contact (onset, 3 February). None of the Victorian patients in this cluster reported previous measles vaccination; all required hospital admission.

In the second cluster, the index patient was a two-year-old girl (onset, 1 February). Although she lived within a kilometre of the household of the first cluster, no clear epidemiological link could be established with any of the earlier cases. Three other children, aged 10 months to three years, and an 18-year-old woman were infected (onset, 12 February-13 March); all attended the same small church group as the index patient. The index patient's parent reported she had been vaccinated against measles in New Zealand at the age of one year, but did not have a record to confirm this. No other patients in the cluster had been vaccinated against measles.

In the third cluster, the index patient was an 18-year-old woman who had returned from Bali on 4 December and became ill seven days later. Her brother developed prodromal symptoms 12 days later. Neither had been vaccinated against measles.

Measles vaccination history

Vaccination histories of the 317 notified patients are shown in Box 4. More than half those notified (55%) reported having been vaccinated, more than half of whom provided a vaccination date from a personal vaccination record.

Reported measles vaccination status was compared with the presence of measles IgG for those with serological results available. Only 7% of those who reported prior vaccination lacked measles IgG. In contrast, 67% of patients who were aged over one year (and therefore eligible for vaccination) and did not report being vaccinated lacked measles IgG (P < 0.001). Among patients who reported vaccination, those who provided a vaccination date were no more likely to have measles IgG detected than those who did not provide a date (P = 0.76).

Prior measles vaccination was reported by 141 patients (62%) who were classified as laboratory rejected, compared with six (32%) who were classified as laboratory confirmed (P = 0.01). Among patients with laboratory-confirmed measles, sporadic cases were more likely to give a history of vaccination (5/8) than those who were part of a cluster (1/11) (Fisher's exact test, P = 0.04).

Reference laboratory testing

Of the 19 patients classified with laboratory-confirmed measles, 16 were positive for measles IgM on testing at VIDRL, two after initial positive results elsewhere. The 16 comprised all 11 cluster cases and five sporadic cases. Another three sporadic cases were positive for measles IgM on testing at other laboratories but had insufficient serum available for retesting at VIDRL. These cases were still classified as "laboratory confirmed". A further three patients were positive for measles IgM on testing at other laboratories but were negative on retesting at VIDRL and were classified as "laboratory rejected".

Evaluation of NHMRC clinical case definition for suspected measles

There was sufficient clinical information to classify 275 notified patients (87%) according to the NHMRC clinical case definition for suspected measles: 92 (33%) met the definition, and 183 (67%) did not. To examine the utility of the NHMRC case definition, we analysed cases that were able to be classified both in this way and according to serological results -- either laboratory confirmed (18) or rejected (202) as measles.

Results are shown in Box 5. Sensitivity of the NHMRC case definition was 61% and specificity was 66%, while positive and negative predictive values were 14% and 95%, respectively. When non-index cases from clusters were excluded (to test the utility of the definition in identifying cases with no epidemiological link to a confirmed measles case), sensitivity and positive predictive value fell to 40% and 5%, respectively, while specificity and negative predictive value remained almost unchanged. Cases from clusters were more likely than sporadic cases to satisfy the NHMRC case definition (10/11 [91%] versus 1/7 [14%]; Fisher's exact test, P = 0.002).

The relationship between notification and laboratory measles diagnosis was also examined: the positive predictive value of notification was 8% (18/220), dropping to 5% (10/212) when non-index cases were excluded.

Discussion

We found that, during the interepidemic period of July 1997 to December 1998 in Victoria, a clinical diagnosis of measles had a low positive predictive value. Despite an 81% rate of serological testing, only 6% of all measles notifications were laboratory confirmed (8% of those that could be classified on the basis of serological results). Laboratory diagnoses of human parvovirus or rubella infections accounted for a further 8% of measles notifications, similar to experience in other countries that have conducted enhanced surveillance.¹⁴

These results highlight the critical importance of laboratory confirmation as part of enhanced measles surveillance. They also highlight the low utility of the NHMRC clinical case definition for suspected measles. As only 33% of notified cases met this definition, it does not seem widely used as the basis for notification. Furthermore, it was neither sensitive (40%) nor highly predictive of true measles (5%) during this interepidemic period. Therefore, rather than the NHMRC clinical case definition for suspected measles, we advocate a definition similar to that used by the Pan American Health Organization of all cases in which a health worker suspects measles.¹⁵

Our findings do not mean that those responsible for measles surveillance, investigation and control can ignore measles notifications. The Disease Control Section now relies on urgent serological testing performed by VIDRL to inform public health action and improve the quality of the surveillance dataset. In Victoria, clinical specimens can often be collected within 24 hours of notification, with a laboratory result available on the next testing day.⁶ During the interepidemic period, when measles was rare, if public health action were to involve excluding contacts of a notified case from a school or childcare centre, we attempted to arrange urgent serological testing. No action was taken until the result was available. If serological testing was not possible, we treated the case as though it were measles regardless of whether it met the NHMRC case definition.

Based on our experience, and drawing on elements from the National Measles Surveillance Strategy,⁷ we have refined recommendations for follow-up of measles notifications in a region with good disease control during an interepidemic period (Box 6). We believe these recommendations will allow identification of clusters of disease and minimise unnecessary public health action. We have maximised the sensitivity of the passive surveillance system by following up notifications from any source. By using laboratory testing to identify cases that are not measles, we have minimised the likelihood that our surveillance dataset will consist largely of false-positive notifications.

Because no IgM antibody test is 100% specific, even laboratory-confirmed cases may not be measles. We found that three of five laboratory diagnoses of measles made in non-reference laboratories could not be confirmed at VIDRL. Sporadic cases were less likely to be confirmed at VIDRL than cluster cases and were also less likely to meet the NHMRC case definition, but were more likely to report prior measles vaccination. As prior measles vaccination correlates well with measles immunity, we believe that at least some of the sporadic cases classified as laboratory confirmed were not true measles. This reinforces the important role of reference laboratories as we approach national measles elimination and global eradication.⁷

We suggest that local transmission of measles within Victoria during this interepidemic period was minimal. We base this belief on the small number of sporadic cases identified, along with the possibility that some of these cases were not true measles, and the fact that identified clusters of infection involved few people and were self-limiting. Specimen collection for genotyping is already under way and will provide further evidence of the interruption of indigenous transmission in Victoria.^16,17

The findings of the enhanced surveillance program, along with those from investigation of the 1999 measles outbreak in Victoria,⁵ lead us to believe that the two-dose MMR vaccination policy and the 1998 measles control campaign have dramatically reduced circulation of measles virus in the targeted age groups. We have demonstrated that, when measles is rare, enhanced surveillance relying on laboratory confirmation is essential to identify true cases of measles promptly and to ensure that surveillance datasets do not largely comprise false positive notifications.

Acknowledgements

The Victorian Enhanced Measles Surveillance Working Party appreciates the cooperation of the patients who agreed to be interviewed and provided serum samples for enhanced surveillance. We also gratefully acknowledge the nursing staff, clinicians, and pathology collection centres who collected serum specimens during the study period. Enhanced surveillance and public health intervention would not be possible without notification of cases by clinicians and laboratories.

References

World Health Organization. Expanded programme on immunization (EPI). Meeting on advances in measles elimination: conclusions and recommendations. Wkly Epidemiol Rec 1996; 71: 305-309.
Lambert S. Measles in Victoria 1992 to 1996: the importance of laboratory confirmation. Comm Dis Intell 1998; 22: 17-22.
National Health and Medical Research Council. The Australian immunisation handbook. Canberra: AGPS, 1997.
National Centre for Disease Control. Immunise Australia program: measles control campaign. Comm Dis Intell 1998; 22: 156.
Lambert S, Lynch P, Morgan M, et al. Measles outbreak -- young adults at high risk. Victorian Infectious Diseases Bulletin 1999; 2: 21-22.
The Enhanced Measles Surveillance Working Party. Implementing a system of enhanced surveillance for measles in Victoria. Commun Dis Intell 1999; 23: 51-54.
Heath T, Burgess M, McIntyre P, Catton M. A national measles surveillance strategy. Commun Dis Intell 1999; 23: 41-49.
Jenkin GA, Chibo D, Kelly HA, et al. What is the cause of a rash after measles-mumps-rubella vaccination? Med J Aust 1999; 171: 194-195.
Centers for Disease Control and Prevention. Measles, mumps, and rubella -- vaccine use and strategies for elimination of measles, rubella, and congenital rubella syndrome and control of mumps: recommendations of the Advisory Committee on Immunisation Practices (ACIP). MMWR Morb Mortal Wkly Rep 1998; 47 (RR-8): 1-58.
Helfand R, Heath J, Anderson L, et al. Diagnosis of measles with an IgM capture EIA: the optimal timing of specimen collection after rash onset. J Infect Dis 1997; 175: 195-199.
Ferson M, Young L, Robertson P, Whybin L. Difficulties in clinical diagnosis of measles: proposal for modified clinical case definition. Med J Aust 1995; 163: 364-366.
National Health and Medical Research Council. Measles: guidelines for the control of outbreaks in Australia. Canberra: AGPS, 1996.
Dean A, Dean J, Coulombier D, et al. Epi Info, version 6: a word processing database, and statistics program for public health on IBM-compatible microcomputers. Atlanta, Ga: Centers for Disease Control and Prevention, 1995.
Brown D, Ramsay M, Richards A, Miller E. Salivary diagnosis of measles: a study of notified cases in the United Kingdom, 1991-3. BMJ 1994; 308: 1015-1017.
Centers for Disease Control and Prevention. Measles eradication: recommendations from a meeting cosponsored by the World Health Organization, the Pan American Health Organization, and CDC. MMWR Morb Mortal Wkly Rep 1997; 46 (RR-11): 1-20.
Rota JS, Heath JL, Rota PA, et al. Molecular epidemiology of measles virus: identification of pathways of transmission and implications for measles elimination. J Infect Dis 1996; 173: 32-37.
Chibo D, Birch C, Rota P, Catton M. Genetic characterisation of measles viruses isolated in Victoria, Australia 1973-1998. Immunisation beyond 2000. 6th National Public Health Association Immunisation Conference; 1998 Nov 4-5; Melbourne. Canberra: Public Health Association of Australia, 1998.

(Received 30 Jun, accepted 27 Oct, 1999)

Authors' details

Department of Human Services, Melbourne, VIC
Stephen B Lambert, FAFPHM, Public Health Physician;
Ross M Andrews, MPH, MAppEpid, Epidemiologist;
Pauline A Lynch, Public Health Nurse;
Debbie K Gercovich, Paediatric Phlebotomist;
Melissa A Morgan, MB BS, Immunisation Coordinator;
Rosemary A Lester, FAFPHM, Public Health Physician.

Victorian Infectious Diseases Reference Laboratory, Melbourne, VIC
Heath A Kelly, FAFPHM, Head, Epidemiology Division;
Mike C Catton, FRCPA, Head, Virology Division;
Jennie A Leydon, BAppSci, Senior Scientist. Microbiological Diagnostic Unit, University of Melbourne, Melbourne, VIC.
Geoffrey G Hogg, FRACP, FRCPA, Director.

Reprints will not be available from the authors.
Correspondence: Dr H A Kelly, Victorian Infectious Diseases Reference Laboratory, Locked Bag 815, Carlton South, VIC 3053.
heath.kellyATnwhcn.org.au

Other articles have cited this article: