Skip Navigation

Completeness of Measles Case Reporting: Review of Estimates for the United States

  1. Walter A. Orenstein, Section Editor and
  2. Rafael Harpaz
  1. Epidemiology and Surveillance Division, National Immunization Program, Centers for Disease Control andPrevention, Atlanta Georgia
  1. Reprints or correspondence: Dr. Rafael Harpaz, National Immunization Program, Mailstop E-61, Centers for Disease Control and Prevention, 1600 Clifton Rd., Atlanta, GA 30333 (rzh6{at}cdc.gov)
 
Next Section

Abstract

Measles surveillance is complex: the patient must seek health care, the diagnosis must be recognized by the physician, and the case must be reported to health departments. The portion of total (incident) measles cases⩾ that is reported to health departments is termed “completeness of reporting.” Few studies describe this measure of the quality of surveillance in the United States; these studies use different methods, but they are all limited because the actual number of measles cases needed to derive completeness of reporting could not be determined. Estimates of completeness of reporting from the 1980s and 1990s vary widely, from 3% to 58%. One study suggests that 85% of patients with measles sought health care, the proportion of compatible illnesses for which measles was considered varied from 13% to 75%, and the proportion of; suspected cases that were reported varied from 22% to 67%. Few cases were laboratory-confirmed, but all were reported. Surveillance in the United States is responsive, and its sensitivity likely increases when measles is circulating. Continued efforts to reinforce the clinical recognition and reporting of measles cases are warranted.

In its most common form, passive measles surveillance in the United States consists of several steps: The person with measles must seek health care, the health care provider must consider the diagnosis of measles and test the patient accordingly, and the health care provider must report the case to the local or state health department. The sensitivity of measles surveillance depends on each of these components.

All disease-reporting requirements derive from state laws and regulations. Although specific reporting requirements vary, measles is a reportable condition in all states. A clinical case definition has been adopted by the Council of State and Territorial Epidemiologists (CSTE), along with national criteria for classifying possible measles cases into categories of suspected, probable, and confirmed cases on the basis of the likelihood that the patient actually has disease due to measles virus infection [1]. The following case definition measles was adopted by the CSTE: an illness characterized by a generalized rash lasting for ⩾3 days; a temperature of ⩾38.3°C; and cough, coryza, or conjunctivitis. The following case classifications were added: suspected, a case that includes febrile illness accompanied by generalized maculopapular rash; probable, a case that meets the clinical case definition, has noncontributory or no serological or virological testing, and is not epidemio-logically linked to a confirmed case; confirmed, a case that is laboratory-confirmed or that meets the clinical case definition and is epidemiologically linked to a confirmed case (a laboratory-confirmed case does not need to meet the clinical case definition).

However, state requirements are not consistent with regard to which of these categories is reportable and when reporting should occur during the diagnostic process. In particular, state regulations often do not define whether notification is required on the basis of the clinical manifestations of the illness or, instead, on an often unspecified degree of clinical suspicion of measles by the health care provider (i.e., independent of symptoms and case definitions).

All states require physicians to report measles to public health authorities, and all but 5 states require laboratories to report confirmed cases as well [2, 3]. Some states require hospital infection-control nurses, school nurses, day care staff, or even the general public to report measles cases. It is clear, however, that physicians play the key role in measles reporting, because notification-whether by the physicians themselves, by laboratory personnel, or by infection-control staff-is possible only if the physicians possess the clinical judgment to first consider the diagnosis of measles and to order confirmatory laboratory testing [3].

In general, states do not impose penalties for failure to report. Furthermore, notification is a passive process initiated by the reporter. Reporting of communicable diseases has thus often been found to be incomplete, with the degree of underreporting varying by disease on the basis of awareness of reporting requirements, disease severity, perceived public health importance, associated social stigma, and other factors [4–10]. Little information exists regarding notification practices for measles in the United States. Evaluating completeness of reporting of measles cases is difficult because the true number of measles cases in the population is not known. Although multiple methods of estimating the true number of cases have been used [11–14], it is difficult to assess the accuracy of these estimates, and often these methods do not provide sufficient information to properly classify the cases. In this article, available data relating to the completeness of reporting of measles are reviewed. These data are summarized in table 1.

Table 1.
View larger version:
Table 1.

Data relating to completeness of measles reporting to public health authorities. United States.

Previous SectionNext Section

METHODS

To identify relevant literature, the MEDLINE database for 1966 through 2002 was searched for the terms “surveillance,” “notification,” “reporting,” and “case finding,” together with the term “measles.” References from the studies thereby identified were also searched to find additional relevant articles. Data from a study published elsewhere in this supplement issue were also included [21]. Finally, a managed care data set was analyzed to obtain additional information regarding completeness of measles reporting.

Previous SectionNext Section

RESULTS

National surveillance compared with inferred prevaccine ⩾incidence. Before licensure of measles vaccine in 1963, multiple surveys indicated that virtually all persons experienced measles by adulthood [22]. In steady-state conditions, ∼1 birth cohort should therefore have had measles annually during the immediate prevaccine era. This would have been ∼4 million cases per year in the decade preceding measles vaccine licensure. However, 400,000–500,000 cases of measles were reported to the Centers for Disease Control and Prevention (CDC) annually during these years, suggesting that completeness of reporting was ∼10% [11]. Before vaccine licensure, health care seeking and disease reporting for measles were possibly low because of the common nature of the illness and the limited treatment or prevention options available.

Community-based study, Maryland, 1922–1923. During 1922–1923, community-based canvassing that included questions about measles was conducted every 6–8 weeks in Hag-erstown, Maryland, a city of 30,000 persons served by 37 general practitioners [15, 16]. Medical record reviews were not conducted, and the case definition for measles was simple parental diagnosis. Respondents noted 568 episodes of measles among 8587 persons during 16,517 person-years of observation; 64% of these patients sought medical attention, of which 40% (i.e., 26% of all children with measles, based on parental diagnosis and recall) were reported to the health department.

Community-based study, St. Louis, 1970–1971. During a measles epidemic in St. Louis, physicians were contacted and encouraged to report cases [17]. There were 745 measles cases with 6 deaths reported by physicians during the outbreak. Questionnaires were also sent to parents of 10,000 children enrolled in 56 city and county schools and day care centers inquiring about measles history. Response rate was 43%. By extrapolating attack rate among respondents to the entire population, the authors estimated that 10,000 children in the metropolitan area experienced measles during the outbreak. Thus, despite physician prompting, completeness of reporting was 7% (745/10,000). The study contained potential sampling biases: Parents of children with measles may have preferentially responded to the survey; furthermore, the authors extrapolated from school and day care settings, where measles is typically most likely to circulate.

Community-based study, Los Angeles, 1990–1991. A community-based survey was conducted in Los Angeles during the 1989–1991 measles epidemic that swept through the United States [18]. Cluster-sample techniques were used to select parents of 947 children <5 years of age. The parents were visited and asked whether their children had ever been ill with measles; they were asked details regarding dates and symptoms of these episodes and whether their children received medical care. On the basis of parental responses, providers were contacted for those children with illnesses meeting CSTE criteria for probable measles.

Public health authorities were notified of 10 (29%) of 35 children with probable measles. Information regarding health-seeking behavior was available for 33 of the 35 children: 5 (15%) children did not obtain medical care and 1 (20%) of these 5 cases was reported to the health department by a school nurse; of the 28 children (85%) who did obtain medical care, 9 (32%) were reported to the health department by their health care providers. Medical records were not available for review for 7 children obtaining medical care; 1 of these cases had been reported to the health department. Of the 21 children with available records, 5 had presented to their private physicians before rash onset, and none of these cases were reported to the health department. A diagnosis of measles was documented for 12 (75%) of the 16 children presenting for care after rash onset (11 [92%] of 12 for patients seen in hospitals and 1 [25%] of 4 for patients seen by private physicians). Three (25%) of these 12 case patients underwent measles IgM testing, yielding positive results, and 8 (67%), including the 3 with confirmed measles, were reported. Notification varied by source of care: In addition to the case reported by the school nurse, the 10 reported cases included both hospitalized patients, 7 of 11 case patients seen in emergency departments, 0 of 3 case patients seen in county public health clinics, 0 of 12 case patients seen by private physicians (Fisher's exact test, P = .001).

Hospital-based study, Los Angeles, 1986 and 1989. Capture-recapture analysis was used to assess measles reporting by hospitals in Los Angeles before and during the measles epidemic [19]. Data on patients discharged from area hospitals with a diagnosis of measles were matched with data on patients hospitalized whose cases were reported to the health department. Measles diagnoses were not laboratory-confirmed for the hospitalized case patients. Completeness of reporting was 58% in 1986 and 51% in 1989. There was no improvement in response to increased awareness generated by the measles epidemic in 1989.

Hospital-based study, New York, 1991. A capture-recapture analysis was conducted in New York City during the measles epidemic [20]. Twelve city hospitals were selected for study. Logs, billing records, and infection-control records were used to detect measles diagnoses among hospitalized and ambulatory (i.e., clinic and emergency room) patients <19 years of age; measles diagnoses for these cases were not laboratory-confirmed. The data were matched with cases reported to the city health department. Completeness of reporting among the sites was 45% (range, 19%-83%). Data were not presented in this study regarding the proportion of patients who were ambulatory.

Managed care - based study, Minnesota, 1994–1998. Completeness of reporting in ambulatory settings was recently evaluated in a large health maintenance organization (HMO) in Minnesota [21]. Administrative data, electronic clinic notes, and laboratory data were searched to ascertain measleslike presentations during 1994–1998; the filtered records were then reviewed by 2 physicians to determine whether symptoms recorded in the medical record met the study case definition for measles (a generalized maculopapular rash with simultaneous fever [by history or >37.7°C on examination], and either cough, coryza, or conjunctivitis). Six million patient encounters occurred during the study period, and 68 records were identified that met the case definition. A measles diagnosis was considered for 9 (13%) of 68 presentations: 2 underwent confirmatory measles IgM testing and 1 of these was confirmed; 3 were ultimately identified as reactions to recently administered measles vaccine; 2 were thought to be due to streptococcal pharyngitis; 1 was thought to be a drug reaction; and 1 had no diagnostic plan documented. The diagnoses for the remaining illnesses were based on clinical judgment and were generally listed as unspecified viral rashes. The health department was notified about 2 (3%) of the 68 cases, those being the patients who had measles IgM testing, and about an additional patient who did not meet the study case definition.

Managed care-based study, California, Oregon, and Washington, 1997–1998. To further explore reporting behavior in ambulatory, managed care settings, an evaluation was conducted during 2000 at the Vaccine Safety Datalink (VSD) sites [23]. VSD is a coalition of 4 HMOs that have formed partnership with the CDC to assess vaccine safety issues. The HMOs are located in Washington, Oregon, and California, with combined enrollment of almost 6 million persons. Administrative data at the sites were searched by use of ICD-9 (International Classification of Diseases, 9th edition) codes for measles, and the identified records were abstracted manually.

During 1997–1998, there were 24 patient encounters coded as measles. However, only 8 (33%) of these patients had a recorded diagnosis relating to measles, serving as a reminder that administrative data may require validation before they are used for surveillance. Only 2 (25%) of 8 patients with measles diagnoses met the CSTE clinical case definition for measles. Measles IgM testing was conducted for 2 (25%) of these 8 patients, including 1 who met the case definition; both test results were negative. The health department had no records that any of the 8 cases of measles had been reported; it is possible they had been reported but that the records had not been retained because measles was ruled out during investigation.

Previous SectionNext Section

DISCUSSION AND CONCLUSIONS

Measles reporting depends on effective case recognition and subsequent notification. The reviewed studies provide insights relating to these processes and to the completeness of measles reporting in the United States.

First, in the Los Angeles community-based study, 85% of children with probable measles sought medical care, suggesting that in recent times and in a relevant setting, most patients with measles are evaluated clinically, providing opportunity for diagnosis and reporting to occur. Of note, 1 case was reported by a school nurse, although the patient had not sought medical care, demonstrating that schools can play a partnership role in disease notification [24–26],

Once patients with measleslike illnesses present for medical care, the diagnostic performance of their providers is difficult to assess. In the Los Angeles community-based study, the classification of probable measles was established by parental recollection of symptoms, and the diagnostic performance of providers was evaluated on the basis of that classification. It is quite possible that parental recall was faulty. Nonetheless, providers did diagnose measles for 75% of children with probable measles who presented after rash onset. In the Minnesota study, measles was considered for 13% of cases meeting the study case definition. However, although cases were classified on the basis of findings from the medical records, the study case definition differed from the CSTE case definition for measles. Furthermore, even for cases that meet the formal CSTE case definition, it is possible that physicians consider the diagnosis of measles but rule out the possibility on the basis of clinical considerations that lead to alternative diagnoses; unfortunately, these diagnostic considerations are often not recorded in the medical chart, making it difficult to evaluate the physician's diagnostic process. State reporting requirements are generally based on the premise of suspicion of measles rather than on compatibility with a specific case definition. Nonetheless, the data reviewed here suggest that physicians may not always consider measles in the differential diagnosis when seeing patients with compatible findings: Measles is rare, and in recent years, most physicians have had little experience with the illness. The data from the Los Angeles community-based study suggest that hospital and emergency department physicians are more likely to consider the diagnosis of measles. Large volumes of patients are seen in such settings; hospital-based physicians may be more attuned to uncommon diseases circulating in the community, or they may have more familiarity with measles. Hospital-based physicians also have easier access to specialist consultation and laboratory testing.

The next step in the reporting process involves notification of cases suspected by the health care provider to be measles. Data addressing this issue are sparse. Reporting was in the range of 45%–58% in the New York and Los Angles studies of hospital-based reporting. Because most patients with measles are not hospitalized [27], monitoring of hospital-based notification would provide only ancillary information on the quality of measles surveillance. In the Los Angeles community-based study, 12 children had a diagnosis of measles in their medical records. Of these, 8 of 11 patients hospitalized or seen in the emergency department were reported, and the 1 patient given a measles diagnosis by a private provider was not reported. In the Minnesota study, physician reported 2 of 9 patients for which a measles diagnosis was considered, including the 1 laboratory-confirmed case, and alternative final diagnoses were documented for the other 7 patients. In the VSD study, 0 of 8 cases with measles diagnoses were reported. Only 2 of these cases met the clinical case definition, and neither was confirmed. Other evidence exists that a substantial proportion of suspected measles cases may not be reported: >20,000 measles IgM tests are ordered by physicians annually, presumably, for the most part, to confirm suspected cases as measles [28]. Yet, the number of suspected measles cases reported annually to health departments is -1400 [29].

Although reporting of suspected measles appeared inadequate, reassuring data from national surveillance suggest that reporting of actual measles is considerably better: Few new unreported cases of measles are detected when health departments conduct thorough investigations in response to report of an index case [30]. Three of the studies reviewed here assessed reporting of laboratory-confirmed cases. Although there were only 4 IgM-positive results, all 4 of these cases were reported. Laboratory reporting of notifiable diseases is typically more complete than that by physicians [7, 8, 31] and could serve as a backup for notification by physicians, assuming that physicians consider the diagnosis of measles and order testing in the first place.

Public health regulations usually require that any illness suspected to be measles be reported, though the degree of clinical suspicion is not generally defined. Furthermore, notification typically needs to be immediate, prior to laboratory confirmation, so that rapid intervention can prevent spread of the highly communicable disease. But with measles incidence in the United States at record low levels, the predictive value of simple clinical suspicion is correspondingly low, and few, if any, suspected measles cases are likely to be true measles. Indeed, an average of only 100 cases of measles are confirmed annually [32], despite the fact that >20,000 tests are conducted [28], directly suggesting the low predictive value of clinical suspicion alone. Although difficult to formally demonstrate, surveillance in the United States is responsive: When measles cases do occur, community awareness in increased, and diagnostic and reporting behavior are likely increased by providers. As measles incidence increases, a more sensitive, less specific system is appropriate.

The surveillance system, and by implication, the current level of measles reporting, appears adequate to detect endemic measles if it were occurring [30]. Nonetheless, public health authorities need to continue to reinforce physician awareness of measles and notification requirements for vaccine-preventable diseases in general through education, health department feedback, and other measures. Assessments of measles reporting in the United States have been sparse, and outcomes have been somewhat conflicting. Additional research would clearly be helpful to better understand provider diagnosis and reporting practices for notifiable diseases, particularly those such as measles that have been targeted for elimination.

Previous SectionNext Section

Acknowledgments

I thank Mary McCauley, Mark Papania, and Jane Seward from the National Immunization Program, CDC, for their editorial and scientific input.

Previous Section
 

References

  1. 1.↵
    1. Centers for Disease Control and Prevention
    . Case definitions for infectious conditions under public health surveillance. MMWR Morb Mortal Wkly Rep 1997;46(RR-10):1-55.
    Medline
  2. 2.↵
    1. Roush MT,
    2. Birkhead G,
    3. Koo D,
    4. Cobb A,
    5. Fleming D
    . Mandatory reporting of diseases and conditions by health care professionals and laboratories. JAMA 1999;282:164-170.
    Abstract/FREE Full Text
  3. 3.↵
    1. Güris D,
    2. Harpaz R,
    3. Redd SB,
    4. Smith NJ,
    5. Papania MJ
    . Measles surveillance in the United States: an overview. J Infect Dis 2004;189(Supplement_1):S177-S184.
    CrossRefMedline
  4. 4.↵
    1. Thacker SB,
    2. Berkelman RL
    . Public health surveillance in the United States. Epidemiol Rev 1998;10:164-190.
  5. 5.
    1. Haward RA
    . Scale of under notification of infectious diseases by general practitioners. Lancet 1973;1:873-874.
    MedlineWeb of Science
  6. 6.
    1. Thacker SB,
    2. Choi K,
    3. Brachman PS
    . The surveillance of infectious diseases. JAMA 1983;249:1181-1185.
    Abstract/FREE Full Text
  7. 7.↵
    1. Schramm M,
    2. Vogt RL,
    3. Mamolen M
    . Disease surveillance in Vermont: who reports? Public Health Rep 1991;106:95-97.
    MedlineWeb of Science
  8. 8.↵
    1. Harkess JR,
    2. Gildon BA,
    3. Archer PW,
    4. Istre GR
    . Is passive surveillance always insensitive? An evaluation of shigellosis surveillance in Oklahoma. Am J Epidemiol 1998;128:878-881.
  9. 9.
    1. Konowitz PM,
    2. Petrossian GA,
    3. Rose DN
    . The under reporting of disease and physician's knowledge of reporting requirements. Public Health Rep 1984;99:31-35.
    MedlineWeb of Science
  10. 10.↵
    1. Vogt RL,
    2. Larue D,
    3. Klauke DN,
    4. Jillson DA
    . Comparison of an active and passive surveillance system of primary care providers for hepatitis, measles, rubella, and salmonellosis in Vermont. Am J Public Health 1983;73:795-797.
    Abstract/FREE Full Text
  11. 11.↵
    1. Hinman AR,
    2. Brandling-Bennett AD,
    3. Bernier RH,
    4. et al
    . Current features of measles in the United States: feasibility of measles elimination. Epidemiol Rev 1980;2:153-170.
    FREE Full Text
  12. 12.
    1. Bean JA,
    2. Burmeister LF,
    3. Paule CL,
    4. Isaacson P
    . A comparison of national infection and immunization estimates for measles and rubella. Am J Public Health 1971;61:2304-2310.
    FREE Full Text
  13. 13.
    1. Hedrich AW
    . The corrected average attack rate from measles among city children. Am J Hyg 1930;11:576-600.
    Web of Science
  14. 14.↵
    1. Los Angeles County Department of Health Services
    . Public Health Lett. Vol. 2. 1980. Disease reporting a health professional's responsibility; p. S1-S3.
  15. 15.↵
    1. Sydenstricker E
    . A study of illness in a general population group. Hagerstown morbidity studies no. I: the method of study and general results. Public Health Rep 1926;41:2069-2088.
  16. 16.↵
    1. Sydenstricker E
    . The reporting of notifiable diseases in a typical small city Agerstown morbidity studies no. II. Public Health Rep 1926;41:2186-2191.
  17. 17.↵
    1. Cherry JD,
    2. Feigin RD,
    3. Lobes LA Jr.,
    4. et al
    . Urban measles in the vaccine era: a clinical, epidemiologic, and serologic study. J Pediatr 1972;81:217-230.
    CrossRefMedlineWeb of Science
  18. 18.↵
    1. Ewert DP,
    2. Westman S,
    3. Frederick PD,
    4. Waterman SH
    . Measles reporting completeness during a community-wide epidemic in inner-city Los Angeles. Public Health Rep 1995;110:161-165.
    MedlineWeb of Science
  19. 19.↵
    1. Ewert DP,
    2. Frederick PD,
    3. Run GHF,
    4. Mascola L
    . The reporting efficiency of measles by hospitals in Los Angeles County, 1986 and 1989. Am J Public Health 1994;84:868-869.
    FREE Full Text
  20. 20.↵
    1. Davis F,
    2. Strebel PM,
    3. Atkinson WL,
    4. et al
    . Reporting efficiency during a measles outbreak in New York City, 1991. Am J Public Health 1993;83:1011-1015.
    Abstract/FREE Full Text
  21. 21.↵
    1. Nordin JD,
    2. Harpaz R,
    3. Harper P,
    4. Rush W
    . Syndromic surveillance for measles-like illnesses in a managed care setting. J Infect Dis 2004;189(Supplement_1):S222-S226.
    CrossRefMedlineWeb of Science
  22. 22.↵
    1. Langmuir AD
    . Medical importance of measles. Am J Dis Child 1962;103:224-226.
    Abstract/FREE Full Text
  23. 23.↵
    1. Chen RT,
    2. Glasser JW,
    3. Rhodes PH,
    4. et al
    . Vaccine Safety Datalink project: a new tool for improving vaccine safety monitoring in the United States. The Vaccine Safety Datalink Team. Pediatrics 1997;99:765-773.
    Abstract/FREE Full Text
  24. 24.↵
    1. Seward JF,
    2. Watson BM,
    3. Peterson CL,
    4. et al
    . Varicella disease after introduction of varicella vaccine in the United States, 1995–2000. JAMA 2002;287:606-611.
    Abstract/FREE Full Text
  25. 25.
    1. Lenaway DD,
    2. Ambler A
    . Evaluation of a school-based influenza surveillance system. Public Health Rep 1995;110:333-337.
    MedlineWeb of Science
  26. 26.↵
    1. Davis JP,
    2. Pfeiffer JA
    . Surveillance of communicable diseases in child day care settings. Rev Infect Dis 1986;8:613-917.
    MedlineWeb of Science
  27. 27.↵
    1. Markowitz LE,
    2. Tomasi A,
    3. Sirotkin BI,
    4. et al
    . Measles hospitalizations, United States, 1977–84: comparison with national surveillance data. Am J Public Health 1987;77:866-868.
    Abstract/FREE Full Text
  28. 28.↵
    1. Harpaz R,
    2. Papania MJ
    . Can a minimum rate of investigation of measles-like illnesses serve as a standard for evaluating measles surveillance? J Infect Dis 2004;189(Supplement_1):S204-S209.
    CrossRefMedlineWeb of Science
  29. 29.↵
    1. Harpaz R,
    2. Papania MJ,
    3. Fujii KE,
    4. et al
    . Lessons learned from establishing and evaluating indicators of the quality of measles surveillance in the United States, 1996–1998. J Infect Dis 2004;189(Supplement_1):S196-S203.
    CrossRefMedlineWeb of Science
  30. 30.↵
    1. Harpaz R,
    2. Papania MJ,
    3. McCauley MM,
    4. Redd SB
    . Has surveillance been adequate to detect endemic measles in the United States? J Infect Dis 2004;189(Supplement_1):S191-S195.
    CrossRefMedlineWeb of Science
  31. 31.↵
    1. Effler P,
    2. Ching-Lee M,
    3. Bogard A,
    4. Ieong MC,
    5. Nekomoto T,
    6. Jernigan D
    . Statewide system of electronic notifiable disease reporting from clinical laboratories: comparing automated reporting with conventional methods. JAMA 1999;282:1845-1850.
    Abstract/FREE Full Text
  32. 32.↵
    1. Papania MJ,
    2. Seward JF,
    3. Redd SB,
    4. Lievano F,
    5. Harpaz R,
    6. Wharton ME
    . The epidemiology of measles in the United States, 1997–2001. J Infect Dis 2004;189(Supplement_1):S61-S68.
    CrossRefMedlineWeb of Science
| Table of Contents

This Article

  1. J Infect Dis. (2004) 189 (Supplement 1): S185-S190. doi: 10.1086/378501
  1. AbstractFree
  2. » Full Text (HTML)Free
  3. Full Text (PDF)Free

Classifications

Share

  1. Email this article

Navigate This Article

Current Issue

  1. March 1, 2012 205 (5)
  1. Journal of Infectious Diseases
  1. Alert me to new issues

Most