|
|
Can Fam Physician. 2005 January 10; 51(1): 60–65. | PMCID: PMC1479580 |
Predictive value of a history of varicella infection Candice N. Holmes, MD, MHSC, CCFP OBJECTIVE To determine whether a history of previous varicella infection provides a
reliable marker for prior infection. QUALITY OF EVIDENCE MEDLINE was searched from January 1996 to May 2002 using the MeSH headings
“varicella,” “chickenpox,” and “medical history taking.” Recommendations in
this paper are based on evidence from well designed cross-sectional
studies. MAIN MESSAGE Serologic testing is advised, rather than presumptive vaccination, for those
with a negative or uncertain history of varicella; most will be immune. For
those with a positive history of varicella, the advice given depends on the
population. For populations at higher risk of varicella infection (eg,
health care workers, pregnant women), routine serum testing is recommended.
For low-risk populations, physicians could accept a positive history of
varicella as a reliable indicator of immunity. CONCLUSION Most studies found that patients’ history of varicella had a high positive
predictive value and a low negative predictive value. These findings suggest
that a positive history of varicella is a reliable marker of disease while a
negative history does not not predict lack of immunity. OBJECTIF Déterminer si une histoire de varicelle est un marqueur fiable d’infection
antérieure. QUALITÉ DES PREUVES Une recherche a été effectuée dans MEDLINE entre janvier 1996 et mai 2002 à
l’aide des rubriques MeSH «varicella,» «chicken pox» «medical history
taking.» Les recommandations formulées dans cet article sont fondées sur des
études transversales bien structurées. PRINCIPAL MESSAGE En cas d’histoire de varicelle négative ou incertaine, il est préférable de
demander un test sérologique plutôt que de présumer une vaccination. Devant
une histoire positive de varicelle, la recommandation dépend de la
population en cause. Pour les groupes à risque élevé de contracter
l’infection (travailleurs de la santé, femmes enceintes, etc.), on
recommande un test sérologique systématique. Dans le cas de populations à
faible risque, une histoire positive de varicelle pourrait être acceptée
comme un indicateur fiable d’immunité. CONCLUSION La plupart des études ont trouvé qu’une histoire de varicelle possède une
forte valeur prédictive positive et une faible valeur prédictive négative.
Les présentes données suggèrent qu’une histoire positive de varicelle
constitue un marqueur fiable de cette maladie tandis qu’une histoire
négative n’indique pas un défaut d’immunité. EDITOR’S KEY POINTS- Can physicians trust a positive history of varicella or must they verify
immunity through antibody testing?
- This update concludes that a positive history is a reliable marker of
disease. A negative history, however, is unreliable, because most people who
have not had varicella still have antibodies in their systems.
- Clinical management varies according to risk of exposure to varicella. For
those at high risk (pregnant women, health care workers), routine
vaccination is recommended. For those at low risk, physicians can consider a
positive history of varicella as a reliable marker of disease.
|
POINTS DE REPÈRE DU RÉDACTEUR- Peut-on se fier à une histoire clinique positive de varicelle ou doit-on
vérifier l’immunité par un dosage des anticorps?
- Cet article de mise à jour conclut qu’une histoire clinique positive de
varicelle est fiable. Par contre, une histoire négative n’est pas fiable car
la plupart des individus n’ayant pas d’antécédent de varicelle ont des
anticorps.
- La conduite clinique varie selon niveau de risque d’être exposé à la
varicelle. Chez les individus à haut risque (p.ex. certains travailleurs,
femmes en âge de procréer), il est prudent de faire une sérologie de routine
même en présence d’une histoire clinique positive de varicelle. Chez les
individus à faible risque, il est raisonnable de considérer qu’une histoire
clinique positive de varicelle est fiable.
|
Varicella (chickenpox) is caused by primary infection with the varicella zoster virus
(VZV), a member of the herpes family. The virus is highly contagious and is spread
through respiratory droplets or direct contact. Varicella can infect nonimmune adults;
disease increases in severity with age, and often causes serious morbidity and absences
from work. 1 In Canada from 1987 to 1996, 70% of
reported deaths due to varicella were among adolescents and adults. 2 When acquired during pregnancy, varicella can cause serious
perinatal morbidity. 3,4 Following varicella infection, more than 95% of people develop
antibodies against varicella (VZV IgG). 5 These
antibodies can be detected by serologic testing; they indicate lifelong immunity to
varicella. Clinicians face the dilemma of whether to rely on a history of varicella to identify
susceptible patients or to do serum testing. This article examines whether a patient’s
history of previous varicella infection provides a reliable marker of prior infection. MEDLINE was searched from January 1996 to May 2002 using the key words “varicella” or
“chickenpox” for studies in which patients’ history of chickenpox exposure was
compared with VZV serology. The key words were exploded, and 2103 articles were
found. After limiting the search to human studies and English language, 1716
articles remained. Combining these articles with the key word “medical history
taking” identified five articles. When the abstracts of all 1716 articles were
reviewed, six more relevant articles were identified. Bibliographies of these 11
articles were reviewed to obtain further references. One additional article was
found. Thus, 12 articles were identified that examined the predictive ability of
self-reported history of varicella. 6-17 All 12 articles were cross-sectional studies
(level II evidence), one was a cost-benefit study, 14 and one examined cost-effectiveness. 13
Table 1,6-17 summarizes the literature review. | Table 1Summary of literature review |
Study outcomes included one or more of sensitivity, specificity, negative predictive
value (NPV), and positive predictive value (PPV) ( Table 2). Six of the 12 studies included only patients with a negative or uncertain
history of varicella 8,10-14 so only the
negative predictive value, and not the positive predictive value, could be
determined. | Table 2Two-by-two table illustrating sensitivity, specificity, and positive and
negative predictive values |
Three studies included only health care workers, 6-8 two studies included only
pregnant women, 9,10 two studies included only children, 11,18 and four studies
included only adolescents and young adults (three of these examined military
populations). 12-15 One study examined two populations: pregnant women and men
and women of all ages. 16Health care workers. Alagappan et al 6 determined the association
of self-reported history of varicella infection with varicella serology in
medical house officers. Only two of 119 house officers with a positive varicella
history were nonimmune. Only one of 10 subjects who reported previously
receiving varicella vaccine was nonimmune. Thus, a reported history of varicella
or past varicella vaccination did not ensure the presence of protective
varicella titres (PPV 98%, NPV 27%). The authors 6 recommend documenting varicella titres in all house officers. Gallagher et al 7 confirmed that a reported
history of chickenpox was not a reliable marker for immune status among health
care workers (PPV 95%, NPV 11%). The authors 7 concluded that all health care workers involved in caring for patients
should be screened with VZV serology before taking up duty. Coyle et al 8 examined seroprevalence of VZV
among female health care workers. An NPV of 6% was found. Based on extrapolation
of results, the authors recommended asking women at their first antenatal visit
whether they had had chickenpox and offering serologic testing to those who had
not. The authors 8 suggested that nonimmune
pregnant women should be counseled on the risk of exposure to patients with
active chickenpox. Studies examining health care workers 6-8 support serologic testing
for all health care workers, regardless of self-reported history of varicella.
Varicella, a recognized nosocomial infection, is an occupational hazard for
susceptible health care workers. Workers could also spread the disease to
patients under their care; immunocompromised patients are at particularly high
risk of varicella-associated morbidity. It is perhaps even more important to
determine varicella susceptibility among health care workers than in the general
population. Pregnant women. Karunajeewa and Kelly 9 found that 5% of
pregnant women who think they have had chickenpox are actually nonimmune (PPV
95%, NPV 9%). For high-risk groups, such as pregnant women, health care workers,
and household contacts of immunocompromised people, the authors recommend
serologic testing regardless of self-reported history of varicella infection.
Despite studying only pregnant women, the authors 9 have generalized their results to other high-risk groups. Silverman et al 10 studied women at their
first prenatal visit who gave either a negative or uncertain varicella history
and had VZV serologic testing. Women with negative varicella histories were
significantly less likely to be immune to varicella than those with uncertain
histories (NPV 53%, NPV 6%, respectively). A comparison group of women with
positive varicella histories were all found to be immune (PPV 100%). In contrast
to the recommendations of Karunajeewa and Kelly, 9 Silverman et al 10 concluded
that women with positive histories of varicella infection could be excluded from
prenatal screening programs. They suggested it might be prudent to screen women
with uncertain histories only as needed because most of them would be immune.
They thought women with negative VZV histories might benefit most from routine
prenatal screening because fewer than half would actually be nonimmune. The
authors 10 made no mention of the role
of varicella vaccine. Nordin et al 11 studied two series of
patients, one consisting of pregnant women with negative histories of varicella
infection. They found that the NPV decreased with increasing age. The
authors 11 extrapolated their results
to health care workers and advised serologic testing for all those with negative
histories of varicella before immunization. It is crucial to establish VZV immunity in women of childbearing age and pregnant
women because of potential adverse fetal and perinatal sequelae. 18 Two studies recommend that pregnant
women have routine serologic testing for varicella, regardless of history. 9,11
Coyle et al 8 recommend routine testing for
varicella immunity, but Silverman et al 10
say that pregnant women with a positive varicella history can be excluded from
routine VZV serologic testing. Preconception testing would be beneficial; susceptible women could be vaccinated
against varicella. Women should be advised to avoid pregnancy for 1 month
following each dose of varicella vaccine. 18 Pregnancy is a contraindication to the varicella vaccine, but
breastfeeding is not, so women nonimmune to varicella should be vaccinated as
early in the postpartum period as possible. 18 Future studies should examine the PPV and NPV of a history of
varicella among nonpregnant women of childbearing age. Children and adolescents. Lieu et al 12 described varicella
seroprevalence among children aged 7 to 12 years with a negative or uncertain
history of chickenpox. Prevalence varied from 9% to 68%, depending on age and
clinical history. More children with uncertain histories were seropositive, and
prevalence of seropositivity increased with age. The authors found that it would
be most cost effective to test serum samples before deciding about varicella
vaccination for children aged 9 to 12 with uncertain chickenpox histories. Boulianne et al 13 also examined varicella
immunity among 10-year-old children with negative or uncertain chickenpox
histories. Prevalence of seropositive results was 63%. Again, more children with
uncertain histories were seropositive, compared with children with negative
histories. The authors 13 concluded that a
negative or unknown history is not well correlated with aphylaxis. Harel et al 14 performed a retrospective
chart review of adolescents reporting a negative or uncertain chickenpox
history. They found a NPV of 20%. The authors recommended serum testing for a
previously unnoticed varicella infection and subsequent vaccination of
susceptible individuals. Studies involving children and adolescents found a low NPV overall. The NPV was
significantly lower ( P <  .001) among those
with uncertain histories than among those with negative histories. 12-14 Serum testing before vaccination is suggested for those with a
negative or uncertain varicella history. Children and adolescents with a
positive history of varicella were not studied, so clinicians are left with the
question of whether to test or presumptively vaccinate them. Young adults (military personnel). Three studies focused on military personnel only. 15-17 Wallace et al 15 evaluated young male military recruits
with varicella, who claimed to have had varicella previously, to determine
whether they had true second episodes of varicella. All prior serum samples of
those with positive varicella histories were negative. Thus, history of previous
varicella infection in adults with varicella might be unreliable; true second
episodes of varicella are rare. Two of the three studies examining military recruits found a high PPV and a low
NPV for history of varicella infection. 16,17 The authors drew
varying conclusions from these results. Jerant et al 16 found that, among army
recruits, the most cost-effective strategy was to test only those with a
negative varicella history and vaccinate seronegative recruits (PPV 99%, NPV
23%). They note, however, that this strategy led to the lowest number of cases
being prevented and that testing serum samples from all recruits regardless of
history was nearly as cost effective and led to more cases being prevented.
Burnham et al 17 found that serologic
screening of all cadets and vaccinating susceptible ones was cost effective (PPV
96%, NPV 44%). An important limitation of many studies is that the PPV was not determined. Also,
whether results of these studies can be generalized to primary care is
uncertain. This study was done to determine the predictive value of a history of varicella.
Findings suggest that a positive history of varicella is reliable, but a negative
history is not. Negative or uncertain history of varicella. Most studies in this review conclude that, for those with a negative or uncertain
history of varicella, serologic testing is advisable, rather than presumptive
vaccination, because most of these people will be immune. Positive history of varicella. Advice differs for those with a positive history of varicella, depending on the
particular risks of varicella infection among the population. Among those at
higher risk of varicella infection, such as health care workers and pregnant
patients, routine serum testing is recommended regardless of self-reported
varicella history. It is important to consider the level of risk of exposure to
varicella, however. For pregnant patients at low risk of exposure, a positive
history of varicella might be sufficient. For those at high risk of exposure or
transmission, routine serum testing for varicella is prudent. High-risk groups include people who live or work in environments where
transmission of varicella is likely (eg, teachers of young children, day-care
employees, and residents and staff in institutions); people who live and work in
environments where transmission can occur (eg, college students, inmates and
staff of correctional institutions, and military personnel); non-pregnant women
of childbearing age; adolescents and adults living in households with children;
and international travelers. 18Our findings indicate that a history of varicella has a high PPV and a low NPV for
immunity. Serologic testing is advised, rather than presumptive vaccination, for
those with a negative or uncertain history of varicella because most of these people
will be immune. For those at increased risk of varicella infection, routine serum
testing, regardless of self-reported varicella history, is recommended. For
populations at lower risk, it might be reasonable to accept a positive history of
varicella as a reliable indicator of immunity. More investigation is needed to evaluate the PPV of a self-reported history of
varicella, including further studies of primary care populations. Whether particular
vaccination strategies are appropriate, cost-effective, or cost-beneficial depends
on the unique circumstances of the population examined. Cost of serologic testing;
cost of vaccination; and the potential financial, social, and medical consequences
of developing varicella should be considered. I thank Dr Rick Glazier for his helpful comments during preparation of the
manuscript. | • | Dr Holmes is an Assistant Professor in the Department of Family and
Community Medicine at the University of Toronto and is a Staff Physician
in the Department of Family and Community Medicine at St Michael’s
Hospital in Toronto, Ont. |
1. Centers for Disease Control and Prevention. Prevention of varicella: recommendations of the Advisory
Committee on Immunization Practices (ACIP). MMWR Recomm Rep. 1996;45(RR11):1–36. 2. Institute for Clinical Evaluative Sciences. A disease most fowl: an update on the varicella-zoster virus. Informed. 1999;5(4):1. 1,3,8. 3. Health Canada. Statement on recommended use of varicella virus vaccine. Can Commun Dis Rep. 1999;25:1–16. 4. Koren G. Varicella virus vaccine before pregnancy: important breakthrough
in protecting fetuses. Can Fam Physician. 2000;46:1975–1977. [PubMed] 5. Glantz JC, Mushlin AI. Cost-effectiveness of routine antenatal varicella screening. Obstet Gynecol. 1998;91:519–528. [PubMed] 6. Alagappan K, Fu L, Strater S, Atreidis V, Auerback C. Seroprevalence of varicella antibodies among new house officers. Ann Emerg Med. 1999;33:516–519. [PubMed] 7. Gallagher J, Quaid B, Cryan B. Susceptibility to varicella zoster virus infection in health care
workers. Occup Med. 1996;46:289–292. 8. Coyle PV, McCaughey C, Wyatt DE, O’Neill HJ. Varicella vaccine in pregnancy: testing should be offered to
women without a history of chickenpox. BMJ. 1997;314:226. [PubMed] 9. Karunajeewa HA, Kelly HA. Predictive value of personal recall of chickenpox infection:
implications for the use of varicella vaccine. Med J Aust. 2001;174:153. [PubMed] 10. Silverman NS, Ewing SH, Todi N, Montgomery OC. Maternal varicella history as a predictor of varicella immune
status. J Perinatol. 1996;16:35–38. [PubMed] 11. Nordin J, Baken L, Carlson R, Hering J. Age-specific rates of serological immunity in patients with a
negative history for varicella infection. Infect Control Hosp Epidemiol. 1998;19:823–824. [PubMed] 12. Lieu TA, Black SB, Takahashi H, Ray P, Capra AM, Shinefield HR, et al. Varicella serology among school age children with a negative or
uncertain history of chickenpox. Pediatr Infect Dis J. 1998;17:120–125. [PubMed] 13. Boulianne N, Duval B, De Serres G, Deceuninck G, Masse R, Coulliard M. Most ten-year-old children with negative or unknown histories of
chickenpox are immune. Pediatr Infect Dis J. 2001;20:1087–1088. [PubMed] 14. Harel Z, Ipp L, Riggs S, Vaz R, Flanagan P. Serotesting versus presumptive varicella vaccination of
adolescents with a negative or uncertain history of chickenpox. J Adolesc Health. 2001;28:26–29. [PubMed] 15. Wallace MR, Chamberlin CJ, Zerboni L, Sawyer MH, Oldfield EC, Olson PE, et al. Reliability of a history of previous varicella infection in
adults. JAMA. 1997;278:1520–1522. [PubMed] 16. Jerant AF, Degaetano JS, Epperly TD, Hannapel AC, Miller DR, Lloyd AJ. Varicella susceptibility and vaccination strategies in young
adults. J Am Board Fam Pract. 1998;11:296–306. [PubMed] 17. Burnham BR, Wells TS, Riddle JR. A cost-benefit analysis of a routine varicella vaccination
program for United States Air Force Academy cadets. Mil Med. 1998;163:631–634. [PubMed] 18. Centers for Disease Control and Prevention. Prevention of varicella: updated recommendations of the Advisory
Committee on Immunization Practices (ACIP). MMWR Morb Mortal Wkly Rep. 1999;48:1–5.
|
