this rule to very young infants. In both studies, cases of missedbacterial meningitis were less than one year of age. AlthoughNigrovic, et al found a peripheral ANC � 103 to be anindependent predictor of bacterial source, a study of febrileinfants less than 90 days old found an inverse relationshipbetween peripheral WBC count and likelihood of bacterialmeningitis, suggesting that young infants’ physiological re-sponses to CNS infection may differ.3

Peter C. Wyer, MDColumbia University College of Physicians and Surgeons

New York, New York

REFERENCES1. Attia J, Hatala R, Cook DJ, Wong JG. Does this adult patient have acutemeningitis? JAMA. 1999;282(2):175-181.2. Nigrovic LE, Kupperman N, Malley R. Development and validation of a multiva-riable predictive model to distinguish bacterial from aseptic meningitis in children in thepost-haemophilus influenzae era. Pediatrics. 2002;110:712-719.3. Bonsu BK, Harper MB. Utility of the peripheral blood count for identifying sickyoung infants who need lumbar puncture. Ann Emerg Med. 2003;41:206-214.

School-based influenza vaccination programreduces influenza-related outcomes amonghousehold membersKing JC, Stoddard JJ, Gaglani MJ, Moore KA, Magder L,McClure E, et al. Effectiveness of school-based influenzavaccination. N Engl J Med 2006:355:2523-32.

Question How effective is a school-based influenza vacci-nation program at decreasing the incidence of influenza-likeillness?

Design Prospective study of households of elementaryschool students.

Setting Demographically similar clusters of elementaryschools in four states.

Participants 11 clusters of schools. Intervention schoolswere assigned to participate in a vaccination program. 1-2control schools were selected for each cluster.

Intervention During a predicted week of peak influenzaactivity in each state, all households with children in inter-vention and control schools were surveyed.

Outcomes Influenza vaccination rates and outcomes of in-fluenza-like illness during the previous 7 days.

Main Results In all, 47% of students in intervention schoolsreceived live attenuated influenza vaccine (LAIV). As comparedwith control-school households, intervention-school householdsreported significantly fewer influenza-like symptoms and officevisits during the recall week; rates of emergency department usedid not vary between the two groups. Paradoxically, interven-tion-school households (both children and adults) had higherrates of hospitalization per 100 persons than did control-schoolhouseholds. However, there was no difference in the overallhospitalization rates for children or adults in households withvaccinated children, as compared with those with unvaccinated

children, regardless of study-group assignment. Rates of schoolabsenteeism for any cause (based on school records) were notsignificantly different between intervention and control schools.

Conclusions Most outcomes related to influenza-like illnesswere significantly lower in intervention school householdsthan in control-school households.

Commentary Influenza vaccination rates among childrenhave historically been low,1 especially among children with high-risk conditions.2,3 King, et al illustrate the potential of a popu-lation-based program to reduce the burden of influenza-likeillness among elementary school children; they do so through aninfluenza vaccination intervention focused solely on healthy chil-dren. Although the findings presented are encouraging, there areseveral important aspects to bear in mind while interpretingthese results. This intervention was restricted to administrationof FluMist (MedImmune) live attenuated influenza vaccine(LAIV). Since children with underlying medical conditions suchas a chronic disease are not eligible for LAIV, they were excludedfrom receiving influenza vaccine through the intervention.4 Ad-ditionally, this study did not assess the reasons why childrenchose not to be vaccinated through the school-based program,nor does it control for the degree to which students may havebeen vaccinated in other settings. These aspects of the studydesign confound interpretation of the findings presented, such asincreased absenteeism among unvaccinated students – a groupthat presumably includes the entirety of the student populationwith underlying medical conditions. As a consequence, onecannot determine which characteristics of schools prior to theintervention were associated with the greatest incremental im-provements in influenza-related outcomes. In addition, the in-fluenza-like illness outcomes assessed by King et al are based onself-reported data from households, the accuracy of which maybe questionable. While the analysis presented does provide anunderstanding of the overall effect of a school-based LAIVprogram, it falls short of illustrating the full potential that mightbe realized by a more comprehensive school-based effort thatalso includes the option of trivalent inactivated influenza vaccine(TIV). This is an important consideration given the prevalenceof chronic conditions among children and the relative expense ofLAIV, which is not covered by many insurance programs.

Kevin Dombkowski, DrPH, MSUniversity of MichiganAnn Arbor, Michigan

REFERENCES1. Centers for Disease Control and Prevention. Estimated influenza vaccinationcoverage among adults and children -- United States, September, 1 2004 - January 31,2005. MMWR Morb Mortal Wkly Rep. 2005;54:304-7.2. Daley MF, Beaty BL, Barrow J, et al. Missed opportunities for influenza vaccinationin children with chronic medical conditions. Arch Pediatr Adolesc Med. 2005;159:986-91.3. Dombkowski KJ, Davis MM, Cohn LM, Clark SJ. Effect of missed opportunitieson influenza vaccination rates among children with asthma. Arch Pediatr Adolesc Med.2006;160:966-71.4. Centers for Disease Control and Prevention. Prevention and Control of Influenza:Recommendations of the Advisory Committee on Immunization Practices (ACIP).Morbidity and Mortality Weekly Report. 2006;55:1-42.

100 The Journal of Pediatrics • July 2007