Advertising Disclaimer

Published online April 1, 2008
PEDIATRICS Vol. 121 No. 4 April 2008, pp. 657-668 (doi:10.1542/peds.2007-1913)
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Meropol, S. B.
Right arrow Articles by Asch, D. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Meropol, S. B.
Right arrow Articles by Asch, D. A.
Related Collections
Right arrow Infectious Disease & Immunity
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?

ARTICLE

Age Inconsistency in the American Academy of Pediatrics Guidelines for Acute Otitis Media

Sharon B. Meropol, MD, MSCEa,b, Henry A. Glick, PhDc,d, David A. Asch, MD, MBAa,c,d,e,f

a Center for Clinical Epidemiology and Biostatistics
b Center for Education and Research on Therapeutics
c Department of Biostatistics and Epidemiology
d Division of General Internal Medicine, Department of Medicine
e Leonard Davis Institute of Health Economics, University of Pennsylvania, Philadelphia, Pennsylvania
f Center for Health Equity Research and Promotion, Philadelphia Veterans Affairs Medical Center, Philadelphia, Pennsylvania


   ABSTRACT
TOP
 ABSTRACT
METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 APPENDIX: ASSUMPTIONS
 REFERENCES
 
OBJECTIVE. The American Academy of Pediatrics acute otitis media guidelines could reduce antibiotic use. The objective was to compare strategies for diagnosing and treating otitis: (1) a commonly used, 2-criteria strategy, (2) the guidelines’ 3-criteria algorithm, and (3) initially watching without antibiotics.

METHODS. A decision analysis was performed with literature-based parameter. The target population was children presenting to primary care physicians with possible otitis media. Main outcomes were antibiotic use, sick days, mild adverse drug events, and number needed to treat/avoided sick day.

RESULTS. For children 2 to <6 months of age, compared with the 2-criteria strategy, guideline use predicted 21% less antibiotic use, 13% more sick days, and 23% fewer adverse drug events; the number needed to treat for the 2-criteria strategy versus the American Academy of Pediatrics strategy was 1.2 children per avoided sick day. For children 6 to <24 months of age, guideline use, compared with the 2-criteria strategy, predicted 26% less antibiotic use, 14% more sick days, and 28% fewer adverse drug events; the number needed to treat for the 2-criteria strategy versus the American Academy of Pediatrics strategy was 1.4 children per avoided sick day. For children >2 years of age, guideline use, compared with the 2-criteria strategy, predicted 67% less antibiotic use, 4% more sick days, and 68% fewer adverse drug events. The number needed to treat for the guideline strategy versus the watch strategy was 6.3 children per avoided sick day; that for the 2-criteria strategy versus the guideline strategy was 12.3. Guideline use for children <2 years implies that our number needed to treat to avoid a sick day is <1.4; for children >2, guideline use implies we are willing to treat at least 6.3 children to avoid a sick day. Thus, the guidelines imply a greater willingness to treat older children, compared with younger children.

CONCLUSIONS. The American Academy of Pediatrics guidelines are inconsistent in their outcomes across age groups. Guideline implementation under age 2 reduces antibiotic use but at a relatively heavy cost of sick days and parental missed work days. This trade-off may be particularly unfavorable for working parents, who might reasonably prefer greater antibiotic use.

Key Words: antibiotic use • child • cost-effectiveness • guidelines • otitis media

Abbreviations: AAP—American Academy of Pediatrics • ADE—adverse drug event • AOM—acute otitis media • ENT—ear, nose, and throat • MEE—middle ear effusion • NNT—number needed to treat • CI—confidence interval

Previous antibiotic use is a risk factor for carrying and being infected with antimicrobial agent-resistant Streptococcus pneumoniae.1–5 There is great interest in minimizing unnecessary pediatric antibiotic use, because children serve as important reservoirs for resistant organisms.4 Acute otitis media (AOM) accounts for 45% to 62% of antibiotics prescribed for US children. Most AOM in the United States is treated with antibiotics and, in the face of increasing antibiotic resistance, increasingly with broad-spectrum antibiotics. Many children diagnosed as having AOM experience improvement without antibiotics; younger and/or sicker children are less likely to experience improvement without them.6,7

The American Academy of Pediatrics (AAP) guidelines define AOM as an acute illness accompanied by both tympanic membrane inflammation and middle ear effusion (MEE). They base antibiotic treatment recommendations on the child's age, illness severity, and AOM diagnostic certainty6 (Tables 1 and 2). Of these criteria, MEE is the most difficult to ascertain8–12; this difficulty leads many clinicians to treat with antibiotics by using only the 2 criteria of illness acuity and inflammation.13,14 By imposing more-specific criteria on AOM diagnosis and treatment, the AAP guidelines could decrease antibiotic use and consequent adverse drug events (ADEs). However, the trade-off of reduced sensitivity for increased specificity might lengthen illness episodes and increase clinical complications and their associated costs. Alternatively, some authors have recommended an initial "watch and wait" AOM strategy.15–19 Our aim was to conduct a decision analysis for a target population of children presenting to primary care physicians with possible AOM, comparing the AAP guidelines with a strategy that is less restrictive in antibiotic use and with a watch and wait strategy.


View this table:
[in this window]
[in a new window]

  		TABLE 1 AAP Criteria for Diagnosis of AOM6

 

View this table:
[in this window]
[in a new window]

  		TABLE 2 AAP Criteria for Treatment of AOM6

 

   METHODS
TOP
 ABSTRACT
 METHODS
RESULTS
 DISCUSSION
 CONCLUSIONS
 APPENDIX: ASSUMPTIONS
 REFERENCES
 
Decision Analysis Model
We used 3 decision trees to evaluate the 3 strategies for each of the guidelines’ age groups, that is, 1 tree each for children 2 to <6 months, 6 to <24 months, and 2 to 12 years of age.6 The time horizon was resolution of a single acute illness episode. Children independently could have true AOM and could present with "severe" illness as defined by the AAP guidelines (moderate/severe otalgia or fever of ≥39°C).6,13 The AAP guidelines and this model exclude children with a history of recurrent AOM or conditions putting them at higher risk of persistent or invasive bacterial infection.6

Diagnostic and Treatment Strategies
In the 2-criteria strategy, children are diagnosed as having AOM if they satisfy both illness acuity and tympanic membrane inflammation criteria (otalgia or tympanic membrane erythema; criteria 1 and 2 in Table 1). All children diagnosed as having AOM are treated with amoxicillin6; those not diagnosed as having AOM are monitored for 48 hours without antibiotic treatment. In the AAP 3-criteria strategy, children are diagnosed as having AOM if they satisfy the 2-criteria strategy criteria plus the additional criterion of MEE. Antibiotic treatment for children diagnosed as having AOM follows AAP guidelines (Table 2).6 In the watch and wait strategy, all children are monitored for 48 hours without antibiotic therapy.

Clinical Management
Table 3 shows the algorithms for initial and follow-up antibiotic prescribing, and the Appendix lists our clinical assumptions. Children who show no improvement are reassessed at 48 hours, with the assumption of a single revisit, with their true disease status unchanged between visits.


View this table:
[in this window]
[in a new window]

  		TABLE 3 Initial and Revisit Antibiotic Treatment Strategies in the Principal Analysis

 
Children without improvement with true AOM can re-present with or without mastoiditis, the only suppurative complication that is modeled. Children without improvement with AOM but no mastoiditis are treated with amoxicillin if they were not treated initially and are treated with amoxicillin-clavulate if they were treated initially with amoxicillin. Children with mastoiditis are hospitalized, receive broad-spectrum antibiotics, and require 1 ear, nose, and throat (ENT) follow-up appointment. Initially untreated children without improvement without true AOM remain untreated at re-presentation. Children with MEE lasting >3 months receive ENT referral.20 There is a maximum of 1 antibiotic change. All cases ultimately resolve.

Sick days and missed parent work days are measured from the time of the initial office visit; 0.5 sick days are experienced if initial resolution occurs. Two additional sick days and 2 missed parent work days are experienced for children needing extra office visits at 48 hours, and 2 additional sick days and 5 or 4 additional missed parent work days are experienced for mastoiditis in children <2 years or 2 to 12 years of age, respectively.21

Representative examples of portions of the decision trees are shown in Figs 1 and 2. Modeling was performed by using Tree Age Pro 2005 0.2 (Tree Age Software, Williamstown, MA), and assumed a societal perspective.


Figure 1
View larger version (22K):
[in this window]
[in a new window]

  		FIGURE 1 Beginning of decision tree, 6 to <24 months.

 

Figure 2
View larger version (15K):
[in this window]
[in a new window]

  		FIGURE 2 Portion of end of decision tree, 2 to 12 years, AAP strategy, true AOM.

 
Probabilities
Model probabilities and ranges (Table 4) were derived from published studies. Age-specific probabilities were used when available. We used data from the National Health Survey22 to estimate the annual number of children who are evaluated for possible AOM and the fraction of those children who actually have AOM. We based our estimate of the annual number of visits for possible AOM on the total number of visits for "cough, cold, runny nose," "viral syndrome," "hayfever, sinusitis," and "middle ear infection." Given the tendency for overdiagnosis,23 we assumed that 75% of the visits coded as "middle ear infection" represented true AOM. The resulting probability range is consistent with published values.24,25


View this table:
[in this window]
[in a new window]

  		TABLE 4 Probabilities

 
Costs/Charges
Table 5 presents cost estimates, ranges, and their sources for antibiotics, parent work days (including 30% benefits26), and hospitalizations. Charges were used to estimate costs for primary care and ENT office visits. When 2005 data were not directly available, values were literature based and adjusted for inflation to 2005 dollars, according to the recommendations of the US Panel on Cost-effectiveness Research for Performing and Reporting Economic Analyses,27 using the general Consumer Price Index for wages or price category-specific Consumer Price Index values, when applicable.28 Costs were incurred during the acute illness year and were not discounted.


View this table:
[in this window]
[in a new window]

  		TABLE 5 Costs

 
The average wholesale price29,30 was used for antibiotic costs. For costs related to lost parent work days, we used the mean September 2004 US hourly wage adjusted to January 2005 dollars.31 In our principal analysis, we valued all parent time equally as working time.27 For extra primary care office visits, we used the average charge of estimates from private insurance and Medicaid.32,33 For outpatient ENT visits, we used the charge for a level 3, initial, extended, ENT visit.33,34 Because we did not have reliable cost data for ADE-related hospitalizations (more prevalent in children prescribed antibiotics), we did not include costs related to mastoiditis hospitalization (less prevalent in children prescribed antibiotics) in the principal analysis but evaluated their potential impact with sensitivity analysis. To estimate hospital costs, we adjusted mean age-specific hospitalization charge data from the Agency for Healthcare Research and Quality Health Care Cost and Utilization Project KIDS’ Inpatient Database,21 using the fiscal year 2005 mean operating cost/charge ratio for short-term, US, prospective payment-system hospitals.35 We included charges from admissions with a mastoiditis principal diagnosis (International Classification of Diseases, 9th Revision, Clinical Modification codes 383.00, 383.01, 383.02, and 383.9).36

Analysis
We compared the outcomes of the 2-criteria, AAP, and watch and wait strategies. Primary outcomes included overall antibiotic use, sick days, mild ADEs, and overall dollar costs. Secondary outcomes included broad-spectrum antibiotic use, mastoiditis, outpatient ENT referrals, and lost parent work days. We report the number needed to treat (NNT) with antibiotics to avoid 1 sick day. We also report the incremental cost per incremental antibiotic prescription avoided.

We evaluated whether the implicit clinical and economic thresholds for antibiotic use embodied in the AAP guidelines are consistent across children in the 3 different age groups. By consistent, we mean that the guidelines imply values for NNT to avoid 1 sick day and cost per antibiotic prescription avoided that are similar across different age groups.

Inconsistency can occur if the NNT for the recommended strategy for one age group (eg, the AAP strategy for 2-12-year-old children) differs from the NNT for the same strategy in a different age group but instead is similar to the NNT for a different strategy in the other age group. For example, suppose an analysis reveals that adoption of the AAP strategy, rather than the 2-criteria strategy, for children 2 to <6 months of age implies that we are not willing to treat >1.5 infants to avoid 1 sick day. Suppose that the same kind of analysis reveals that, for children 2 to 12 years of age, adoption of the AAP strategy over the watch and wait strategy implies that we are willing to treat ≥2.4 children to avoid 1 sick day. These 2 results would imply a stricter threshold for antibiotic use for infants than for children 2 to 12 years of age and would represent an inconsistency in the guidelines.

These consistency analyses are complicated by the uncertainty of point estimates, but the same reasoning can be applied if ranges described by confidence intervals (CIs) are compared instead of individual point estimates. In those circumstances, nonoverlapping CIs can be used to demonstrate inconsistency.

We used second-order Monte Carlo simulations to derive SEs and correlations and calculated 95% CIs for the NNTs and the incremental cost-effectiveness ratios. The latter CIs were calculated by using Fieller's Theorem method.37 We specified distributions for key variables, using normal distributions for cost variables and β distributions for probabilities.38–40

Sensitivity Analyses
We performed sensitivity analyses for 4 structural assumptions of the model. First, we evaluated the impact of adding the costs of mastoiditis hospitalization to our analysis. If we had also evaluated the impact of the costs of ADE hospitalization, they would have partially balanced mastoiditis hospitalization costs. Second, we evaluated the impact of alternative assumptions about follow-up antibiotic prescribing by modeling 2 extreme alternative follow-up assumptions, that is, (1) no children without improvement without true AOM received antibiotic treatment at the follow-up visit, and even those initially treated with antibiotics stopped antibiotic treatment, and (2) all children without improvement, even those without true AOM, received antibiotic treatment at the follow-up visit. Third, we evaluated an intermediate strategy that used 2 criteria for diagnosis but based treatment decisions on the AAP guidelines. Fourth, we performed 1-way sensitivity analysis for the assumption of 0.5 sick days experienced for improvement if resolution occurs by using a range of 0 (immediate resolution) to 1 sick day. We also performed 1-way sensitivity analysis for the NNT to avoid 1 sick day and the incremental cost per incremental avoided antibiotic prescription by using the minimum and maximum values for variable ranges listed in Tables 4 and 5, including the exclusion of the cost of lost parent work days from these ratios.


   RESULTS
TOP
 ABSTRACT
 METHODS
 RESULTS
DISCUSSION
 CONCLUSIONS
 APPENDIX: ASSUMPTIONS
 REFERENCES
 
Antibiotic Use
Table 6 reports expected outcomes for cohorts of 1000 children using alternative clinical strategies among the 3 different age groups. The 2-criteria strategy had the greatest antibiotic use. Adoption of the AAP strategy reduced antibiotic use significantly, by 21%, 26%, and 67% among children 2 to <6 months, 6 to <24 months, and 2 to 12 years of age, respectively (all P < .0001). The watch and wait strategy had the least antibiotic use and, compared with the AAP strategy, reduced use by an additional 43%, 38%, and 67% for the same age groups, respectively (all P < .0001).


View this table:
[in this window]
[in a new window]

  		TABLE 6 Clinical and Economic Outcomes Among a Cohort of 1000 Patients

 
Clinical Outcomes
The 2-criteria strategy, with the greatest antibiotic use, had the most ADEs. Compared with the 2-criteria strategy, the AAP strategy reduced mild ADEs by 23% (P = .26), 28% (P = .13), and 68% (P = .0003) among children 2 to <6 months, 6 to <24 months, and 2 to 12 years of age, respectively. The watch and wait strategy further reduced ADEs by 47%, 46%, and 72% in the same age groups, respectively (P = .02). The 2-criteria strategy had the fewest sick days, lost work days, mastoiditis, and ENT referrals. The AAP strategy increased sick days by 13%, 14%, and 4% (P < .0001 for all 3 comparisons) among children 2 to <6 months, 6 to <24 months, and 2 to 12 years of age, respectively. The watch and wait strategy increased values even more (all P < .0001).

Costs
When the cost of missed parent work days was included in the analysis, antibiotic use reduced costs. Compared with the 2-criteria strategy, the AAP strategy yielded 23% (P = .01), 24% (P = .01), and 4% (P = .75) higher overall costs for children 2 to <6 months, 6 to <24 months, and 2 to 12 years of age, respectively. The watch and wait strategy had the highest costs in all age groups.

NNT and Cost per Antibiotic Prescription Avoided
As shown in Table 6, among children 2 to <6 months of age, the 2-criteria strategy had a NNT of 1.2 (95% CI: 1.0–1.5) per avoided sick day, compared with the AAP strategy; the AAP strategy had a NNT of 0.9 (95% CI: 0.7–1.0), compared with the watch and wait strategy. Among children 6 to <24 months of age, the 2-criteria strategy had a NNT of 1.4 (95% CI: 1.1–1.8), compared with the AAP strategy, and the AAP strategy had a NNT of 0.7 (95% CI: 0.6–0.8), compared with the watch and wait strategy. Among children 2 to 12 years of age, the 2-criteria strategy had a NNT of 12.3 (95% CI: 6.6–86.0), compared with the AAP strategy, and the AAP strategy had a NNT of 6.3 (95% CI: 2.4 to being dominated), compared with the watch and wait strategy.

Among children 2 to <6 months of age, the AAP strategy had an incremental cost-effectiveness ratio of $228 (95% CI: $178–$283) per antibiotic prescription avoided, compared with the 2-criteria strategy; the watch and wait strategy had an incremental ratio of $325 (95% CI: $276–$380), compared with the AAP strategy. Among children 6 to <24 months of age, the AAP strategy had an incremental cost-effectiveness ratio of $199 (95% CI: $140–$263), compared with the 2-criteria strategy; the watch and wait strategy had an incremental ratio of $491 (95% CI: $395–$604), compared with the AAP strategy. Among children 2 to 12 years of age, the AAP strategy had an incremental cost-effectiveness ratio of $11 (the AAP strategy dominates or costs up to $34), compared with the 2-criteria strategy; the watch and wait strategy had an incremental ratio of $34 (the watch and wait strategy dominates or costs up to $111), compared with the AAP strategy.

Consistency Across Age Groups
For the 2- to 12-year group, use of the AAP strategy rather than the watch and wait strategy implies that we are willing to treat a minimum of 2.4 children (minimum of the 95% CI for the AAP strategy, compared with the watch and wait strategy) with antibiotics to avoid 1 sick day; that is because, if we were not willing to treat ≥2.4 children to avoid 1 sick day, then we would be 95% confident that the AAP strategy exceeds the number we are willing to treat. For younger age groups, however, the decision to reject the 2-criteria strategy in favor of the AAP strategy implies that we are willing to treat no more than 1.5 children (upper limit of the 95% CI for the 2-criteria strategy, compared with the AAP strategy, for children 2 to <6 months of age) or 1.8 children (upper limit of the 95% CI for the 2-criteria strategy, compared with the AAP strategy, for children 6 to <24 months of age) to avoid 1 sick day; that is because, if we were willing to treat ≥1.8 children to avoid 1 sick day, then we would be >95% confident that we should adopt the 2-criteria strategy, rather than the AAP strategy.

The same inconsistency is observed in a cross-age analysis based on the cost per antibiotic prescription avoided. For children 2 to <6 months of age, a recommendation to adopt the AAP strategy implies a willingness to pay to avoid an antibiotic prescription that is at least $178 (lower limit of the 95% CI for the AAP strategy, compared with the 2-criteria strategy); that is because, if we were not willing to pay at least $178 to avoid an antibiotic prescription, then we would be 95% confident that the AAP strategy does not represent good value. Similarly, for children 6 to <24 months of age, a recommendation to adopt the AAP strategy implies a willingness to pay that is at least $140, again because, if we were not willing to pay at least $140, then we would be 95% confident that the AAP strategy is not a good value. For children 2 to 12 years of age, a recommendation to adopt the AAP strategy over the watch and wait strategy implies a willingness to pay that cannot exceed the $111 upper limit of the 95% CI for the watch and wait strategy (otherwise, we would be 95% confident that the watch and wait strategy is a good value). Therefore, the willingness to pay to avoid an antibiotic prescription that is incorporated implicitly into the AAP guidelines is lower for older children (no more than $111) than it is for younger children (at least $178). These findings suggest that the AAP guidelines imply a greater willingness to treat older children, compared with younger children.

Sensitivity Analyses
Structural Sensitivity Analyses
Our evaluation of the structural assumption to exclude the cost of mastoiditis hospitalization led to a $3 to $8 increase in the incremental cost-effectiveness ratios, depending on age group. Modeling alternative follow-up treatment assumptions yielded incremental cost-effectiveness ratios that were $1 to $19 higher than those in our principal analysis. The alternative strategy using the 2-criteria strategy for diagnosis and the AAP criteria for treatment had a higher incremental cost-effectiveness ratio than did the 2-criteria strategy, the next most-costly strategy, and thus was weakly dominated by the 2-criteria strategy. Finally, modeling of alternative resolution sick day assumptions led to a NNT increase of only 0.3 for the watch and wait strategy, compared with the AAP strategy, for the 2- to 12-year age group. These alternative approaches did not affect our finding of inconsistent values of willingness to pay between younger and older children.

One-Way Sensitivity Analyses
Missed parent work days made up the largest component of costs for all ages. If the proportion of working parents was reduced from 100% to 0% (reflecting a move from 1 parent losing work time to no parents losing work time), then predicted costs per child for the AAP strategy decreased ~80% in all 3 age groups and the watch and wait strategy dominated the AAP strategy and the 2-criteria strategy in the 2- to 12-year age group.

When we varied each of our 17 cost and probability estimates over their ranges, we continued to observe an inconsistency in the NNT to avoid 1 illness day and in the willingness to pay to avoid an antibiotic prescription between children 2 to <6 months of age and children 2 to 12 years of age for all except 2 variables, that is, probability of improvement in 2 days without amoxicillin given AOM and sensitivity of diagnosing MEE. We continued to observe an inconsistency between children 6 to <24 months of age and those 2 to12 years of age for all except the same 2 variables plus 4 others, that is, probability of AOM, sensitivity of diagnosing AOM with 2 criteria, probability of improvement in 2 days without AOM, and probability of severe illness.


   DISCUSSION
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
CONCLUSIONS
 APPENDIX: ASSUMPTIONS
 REFERENCES
 
The AAP guidelines were developed to address increasing antibiotic resistance, growing costs of prescribed antimicrobial medications, and uncertainty regarding which children should receive antimicrobial therapy for AOM. They were meant to provide an evidence-based framework guiding primary care clinicians toward more-judicious antibiotic use.6 Unlike the current analysis, which used many of the same references considered by the guideline authors, decision models were not used in guideline development.

Although the guidelines were meant to assist in treatment decisions for childhood AOM, we recognize that clinicians may use part of the guidelines or otherwise adapt them or combine them with other strategies in ways not evaluated here. We also recognize that the 3 strategies we compared do not cover all possible diagnostic and therapeutic permutations. Instead, they represent a range of pragmatic approaches that clinicians may adopt.

Management choices for AOM yield different clinical and economic outcomes that may vary in importance to different stakeholders. For example, reduction in antibiotic use may be an important goal, but so may be reductions in missed parent work days, ADEs, or costs. How one feels about each of these strategies depends on the relative value one places on these clinical and economic outcomes.

Our study has the following findings. First, compared with a less-antibiotic-restrictive 2-criteria strategy,8,9,11 the AAP strategy reduces antibiotic use but increases sick days and also increases costs, especially for working parents. In children <2 years of age, the AAP strategy reduces antibiotic use by 21% to 26% but increases sick days by 13% to 14% and costs by $178 to $283 per each avoided antibiotic prescription. For children >2 years of age, the AAP strategy reduces antibiotic use by 67% but increases sick days by only 4% and does not increase costs. This analysis provides stronger support for the AAP criteria in this older age group than in children <2 years of age.

Second, our analysis reveals an age inconsistency across the AAP guidelines that may further undermine support of the AAP guidelines for children <2 years of age. The AAP criteria imply a lower threshold for antibiotic use for older children than for younger ones. When considering the NNT outcomes, if we are willing to treat 2.4 children >2 years of age with antibiotics to avoid 1 sick day, as adoption of the AAP strategy implies, then to be consistent we should be even more willing to treat only 1.5 to 1.8 children <2 years of age with antibiotics to avoid 1 sick day. The latter threshold is consistent with use of the 2-criteria strategy rather than the AAP strategy in children <2 years of age.

Is it problematic if the practical, easy-to-follow, AAP guidelines are inconsistent across age groups? It is if we think that we should have the same threshold for treating infants as we do for treating older children. It is even more of a problem if, because of younger children's perceived vulnerability, clinicians have greater willingness to treat younger children, compared with older children. If we use the AAP strategy in the older age group then, for consistency, we should be more lenient with antibiotics in the infant group and instead adopt the 2-criteria strategy for them. Alternatively, if we want to retain the AAP strategy for infants, then we should be more stringent with antibiotics for children 2 to 12 years of age and instead adopt the watch and wait strategy for these older children. In this context, it would not be consistent to adopt the AAP strategy for both infants and older children.

At least 4 recent clinical trials have studied the use of delayed antibiotic prescription at the initial AOM visit,15–18 similar to our watch and wait strategy. All authors concluded that their studies supported a watchful waiting approach for AOM at the initial visit. Three trials enrolled children starting at 6 months of age and 1 enrolled children >12 months of age. Three studies excluded children with high fever, and the fourth, although it excluded only "toxic" children, enrolled only 283 of 776 children presenting with AOM. In addition, most of the children enrolled in those studies were older than 2 to 3 years of age. Consequently, those studies do not address adequately outcomes for severely ill children <2 years of age and do not reflect the typical spectrum of children presenting to primary care physicians with possible AOM.

This study has some limitations. First, this study is limited by the model assumptions, the probabilities used for chance events, and the unavailability of cost data for some variables. Nevertheless, we used the best literature evidence available for probability estimates, probabilities were consistent across the different strategies, and relative preferences were reasonably stable to changes in variable estimates. The relative values of the strategies changed only at the extremes of single-variable distributions, which are unlikely.

Second, a large number of clinical and economic outcomes are relevant to decisions about the optimal management of AOM, including overall cost, use of antibiotics, avoidance of sick days, and suppurative complications. We did not attempt to make definitive conclusions on the basis of only 1 outcome, such as antibiotic use, and we did not provide a summary metric such as utility that might have put these outcomes on a single scale. Use of only 1 outcome would have simplified our results but ignored the reality that different stakeholders might feel very differently about the relative values of component outcomes. It is also difficult to quantify the potential benefit of reducing antibiotic use in slowing or reversing the development of antimicrobial agent-resistant organisms.

A decision analysis can compare treatments that are not likely to be evaluated directly in clinical trials. For example, physicians selectively enroll patients in office-based AOM trials; all of the watch and wait trials described above selectively enrolled fewer and older patients than those typically seen in the clinical setting with possible AOM.15–18 In addition, the rare AOM outcomes, such as mastoiditis hospitalizations and ADE deaths, are difficult to measure prospectively.


   CONCLUSIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
APPENDIX: ASSUMPTIONS
 REFERENCES
 
Our results provide evidence that the AAP criteria may be more cost-effective for older children than for those <2 years of age. In the younger age groups, the AAP strategy, compared with less-restrictive strategies, reduces antibiotic use but only at a relatively heavy clinical cost of increased sick days and at a relatively heavy economic cost. Working parents’ pressure to prescribe antibiotics for children <2 years of age could be viewed as rational within this context.

We also reveal an age inconsistency in the AAP guidelines. Unless we are willing to tolerate more sick days and higher economic costs sustained by children <2 years of age than by older children, either we should be more willing to adopt the AAP AOM guidelines for younger children and monitor older children without antibiotics at the initial visit or we should use AAP guidelines for older children and be less restrictive with antibiotics for children <2 years of age.

We suspect that many clinicians already take different families’ circumstances and/or values into account when making AOM treatment decisions. Results of this study should help clinicians consider these trade-offs more explicitly and discuss them with families more effectively.


   APPENDIX: ASSUMPTIONS
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 APPENDIX: ASSUMPTIONS
REFERENCES
 

  1. Probabilities are mutually independent.
  2. Severe illness is as defined in the AAP guidelines (≥39°C fever or moderate/severe otalgia).
  3. The probability of "severe" illness is the same for AOM and for non–AOM illness.
  4. Uncertainty is limited to the presence or absence of MEE.6,13
  5. Children without AOM eventually experience improvement without antibiotic.
  6. Only 1 revisit is included, 48 hours after the initial visit.
  7. True AOM/no AOM status does not change between visits.
  8. Children without improvement who were not initially treated with antibiotic are diagnosed according to their correct true AOM/no AOM status at revisit, and only those with true AOM receive antibiotic.
  9. Amoxicillin is the first antibiotic used; amoxicillin/clavulanate is prescribed at the revisit if there is no improvement after 48 hours of amoxicillin treatment. Mastoiditis patients receive a broad-spectrum antitibiotic with the same ADE risks as amoxicillin/clavulanate.
  10. Only 1 antibiotic change is included; there is eventual improvement with antibiotic treatment unless chronic effusion or mastoiditis develops.
  11. No subsequent non–AOM diagnoses require antibiotic or hospitalization.
  12. No missed work days result from antibiotic complications.
  13. Mastoiditis is the only suppurative complication modeled.
  14. If mastoiditis develops, then it is diagnosed at the first revisit; all patients with mastoiditis are hospitalized.
  15. Sick days and missed parent work days are measured from the time of the initial office visit; 0.5 sick days are experienced for improvement if resolution occurs.
  16. Two additional sick days and 2 missed parent work days are experienced for patients needing extra office visits.
  17. Two extra sick days are experienced with mastoiditis.
  18. Five additional missed parent work days are experienced with mastoiditis in <2-year-old children and 4 additional missed parent work days are experienced with mastoiditis in 2- to 12-year-old children.21
  19. One half missed work day is experienced for an ENT visit.
  20. ENT follow-up visits for patients with mastoiditis but no other long-term effects of suppurative complications (eg, hearing loss, mental retardation, or death) are modeled.
  21. ENT referral is made if MEE lasts for ≥3 months.20
  22. Probabilities of ADEs are modeled as part of terminal outcomes only and not at decision nodes, distorting results for the sake of model simplicity for this small portion of the results.
  23. If amoxicillin and broad-spectrum antibiotics are used, then the probability of mild ADE is additive but the probability of deadly ADE is not. The probability of ADE death is equal for all patients receiving any antibiotic.
  24. Dollar costs of ADEs are not included. Dollar costs of mastoiditis hospitalization are included for sensitivity analyses only.
  25. ADE deaths do not affect later outcomes, for model simplicity.
  26. In sensitivity analyses, when MEE uncertainty increases from baseline, certain MEE decreases and not no MEE.


   ACKNOWLEDGMENTS
 
This project was supported in part by an Agency for Healthcare Research and Quality Centers for Education and Research on Therapeutics cooperative agreement (grant HS10399).

We thank Andrew Briggs, University of Glasgow, for valuable suggestions regarding variable distributions.


   FOOTNOTES
 
Accepted Aug 29, 2007.

Address correspondence to Sharon B. Meropol, MD, MSCE, Center for Clinical Epidemiology and Biostatistics, University of Pennsylvania School of Medicine, 108 Blockley Hall, 423 Guardian Dr, Philadelphia, PA 19104. E-mail: meropols{at}mail.med.upenn.edu

Financial Disclosure: Dr Meropol has served as a consultant for Wyeth Pharmaceuticals and is supported in part by an unrestricted educational grant from Amgen.

This work was presented, in part, at the annual meeting of the Pediatric Academic Societies; May 2, 2006; San Francisco, CA; and the 22nd International Conference on Pharmacoepidemiology and Therapeutic Risk Management; August 27, 2006; Lisbon, Portugal.


What's Known on This Subject

Acute otitis media afflicts most children. The American Academy of Pediatrics otitis media guidelines were meant to reduce antibiotic use by being more selective about which children are diagnosed with otitis media and which of these are prescribed antibiotic treatment.

 

What This Study Adds

Implementation of the AAP guidelines under age 2 reduces antibiotic use but at a relatively heavy cost of increased sick days and parental missed work days. In addition, the guidelines are inconsistent in their clinical and economic outcomes across age groups.

 


   REFERENCES
TOP
 ABSTRACT
 METHODS
 RESULTS
 DISCUSSION
 CONCLUSIONS
 APPENDIX: ASSUMPTIONS
 REFERENCES
 
1. Gonzales R, Bartlett JG, Besser RE, et al. Principles of appropriate antibiotic use for treatment of uncomplicated acute bronchitis: background. Ann Intern Med. 2001;134 (6):521 –529[Abstract/Free Full Text]

2. Kwan-Gett TS, Davis RL, Shay DK, Black S, Shinefield H, Koepsell T. Is household antibiotic use a risk factor for antibiotic-resistant pneumococcal infection? Epidemiol Infect. 2002;129 (3):499 –505[CrossRef][Medline]

3. Hennessy TW, Petersen KM, Bruden D, et al. Changes in antibiotic-prescribing practices and carriage of penicillin-resistant Streptococcus pneumoniae: a controlled intervention trial in rural Alaska. Clin Infect Dis. 2002;34 (12):1543 –1550[CrossRef][Web of Science][Medline]

4. Klugman KP. Pneumococcal resistance to antibiotics. Clin Microbiol Rev. 1990;3 (2):171 –196[Abstract/Free Full Text]

5. Nasrin D, Collignon PJ, Roberts L, Wilson EJ, Pilotto LS, Douglas RM. Effect of β-lactam antibiotic use in children on pneumococcal resistance to penicillin: prospective cohort study. BMJ. 2002;324 (7328):28 –30[Abstract/Free Full Text]

6. American Academy of Pediatrics, Subcommittee on Management of Acute Otitis Media. Diagnosis and management of acute otitis media. Pediatrics. 2004;113 (5):1451 –1465[Abstract/Free Full Text]

7. Hotomi M, Yamanaka N, Samukawa T, et al. Treatment and outcome of severe and non-severe acute otitis media. Eur J Pediatr. 2005;164 (1):3 –8[CrossRef][Web of Science][Medline]

8. Rothman R, Owens T, Simel DL. Does this child have acute otitis media? JAMA. 2003;290 (12):1633 –1640[Abstract/Free Full Text]

9. Hayden GF. Acute suppurative otitis media in children: diversity of clinical diagnostic criteria. Clin Pediatr (Phila). 1981;20 (2):99 –104[Abstract/Free Full Text]

10. Arnold SR, To T, McIsaac WJ, Wang EE. Antibiotic prescribing for upper respiratory tract infection: the importance of diagnostic uncertainty. J Pediatr. 2005;146 (2):222 –226[CrossRef][Web of Science][Medline]

11. Froom J, Culpepper L, Grob P, et al. Diagnosis and antibiotic treatment of acute otitis media: report from International Primary Care Network. BMJ. 1990;300 (6724):582 –586[Abstract/Free Full Text]

12. Gonzalez-Vallejo C, Sorum PC, Stewart TR, Chessare JB, Mumpower JL. Physicians’ diagnostic judgments and treatment decisions for acute otitis media in children. Med Decis Making. 1998;18 (2):149 –162[Web of Science][Medline]

13. Rosenfeld RM. Diagnostic certainty for acute otitis media. Int J Pediatr Otorhinolaryngol. 2002;64 (2):89 –95[CrossRef][Web of Science][Medline]

14. Spiro DM, King WD, Arnold DH, Johnston C, Baldwin S. A randomized clinical trial to assess the effects of tympanometry on the diagnosis and treatment of acute otitis media. Pediatrics. 2004;114 (1):177 –181[Abstract/Free Full Text]

15. Little P, Gould C, Williamson I, Moore M, Warner G, Dunleavey J. Pragmatic randomised controlled trial of two prescribing strategies for childhood acute otitis media. BMJ. 2001;322 (7282):336 –342[Abstract/Free Full Text]

16. Siegel RM, Kiely M, Bien JP, et al. Treatment of otitis media with observation and a safety-net antibiotic prescription. Pediatrics. 2003;112 (3):527 –531[Abstract/Free Full Text]

17. McCormick DP, Chonmaitree T, Pittman C, et al. Nonsevere acute otitis media: a clinical trial comparing outcomes of watchful waiting versus immediate antibiotic treatment. Pediatrics. 2005;115 (6):1455 –1465[Abstract/Free Full Text]

18. Spiro DM, Tay KY, Arnold DH, Dziura JD, Baker MD, Shapiro ED. Wait-and-see prescription for the treatment of acute otitis media: a randomized controlled trial. JAMA. 2006;296 (10):1235 –1241[Abstract/Free Full Text]

19. Coco AS. Cost-effectiveness analysis of treatment options for acute otitis media. Ann Fam Med. 2007;5 (1):29 –38[Abstract/Free Full Text]

20. American Academy of Family Physicians, American Academy of Otolaryngology-Head and Neck Surgery, American Academy of Pediatrics, Subcommittee on Otitis Media With Effusion. Otitis media with effusion. Pediatrics. 2004;113 (5):1412 –1429[Abstract/Free Full Text]

21. Agency for Healthcare Research and Quality. Healthcare Cost and Utilization Project KIDS’ Inpatient Database, 2000 hospital stays. Available at: http://hcupnet.ahrq.gov/. Accessed February 17, 2005

22. Fried V, Makuc DM, Rocks RN. Ambulatory Health Care Visits by Children: Principal Diagnosis and Place of Visit. Hyattsville, MD: National Center for Health Statistics; 1998. DHHS Publication (PHS) 98–1798

23. Pichichero ME, Poole MD. Assessing diagnostic accuracy and tympanocentesis skills in the management of otitis media. Arch Pediatr Adolesc Med. 2001;155 (10):1137 –1142[Abstract/Free Full Text]

24. Winther B, Alper CM, Mandel EM, Doyle WJ, Hendley JO. Temporal relationships between colds, upper respiratory viruses detected by polymerase chain reaction, and otitis media in young children followed through a typical cold season. Pediatrics. 2007;119 (6):1069 –1075[Abstract/Free Full Text]

25. Gonzales R, Malone DC, Maselli JH, Sande MA. Excessive antibiotic use for acute respiratory infections in the United States. Clin Infect Dis. 2001;33 (6):757 –762[CrossRef][Web of Science][Medline]

26. Bureau of Labor Statistics. Wages and Salaries, All Occupations, All Civilian: Employer Costs for Employee Compensation. Washington, DC: Bureau of Labor Statistics; 2005

27. Gold MR, Siegel JE, Russell LB, Weinstein MC, eds. Cost-effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996

28. Bureau of Labor Statistics. Consumer Price Indexes. Available at: www.bls.gov/schedule/archives/cpi_nr.htm. Accessed January 19, 2005

29. Danzon PM. The US retail price versus FSS best price comparison. In: Price Comparisons for Pharmaceuticals: A Review of US and Cross-National Studies. Washington, DC: American Enterprise Institute; 1999:13

30. Murray L, ed. Red Book: Pharmacy's Fundamental Reference. Montvale, NJ: Thomson PDR; 2004

31. Bureau of Labor Statistics. Employer cost for employee compensation summary, September 2004. Available at: www.bls.gov/ncs/ect/home.htm

32. Marcy M, Takata G, Chan LS, et al. Management of Acute Otitis Media: Evidence Report/Technology Assessment 15. Rockville, MD: Agency for Healthcare Research and Quality; 2001. AHRQ Publication 01-E010

33. Bondy J, Berman S, Glazner J, Lezotte D. Direct expenditures related to otitis media diagnoses: extrapolations from a pediatric Medicaid cohort. Pediatrics. 2000;105 (6). Available at: www.pediatrics.org/cgi/content/full/105/6/e72

34. Centers for Medicare and Medicaid Services. Selected 2005 Medicare fee schedule RVUs. Available at: www.cms.hhs.gov/pfslookup/02_PFSsearch.asp. Accessed May 1, 2005

35. Centers for Medicare and Medicaid Services. Acute inpatient prospective payment system. Historical impact files for fiscal year 1994 through fiscal year 2005, fiscal year 2005 proposed impact files. Available at: www.cms.hhs.gov/AcuteInpatientPPS/HIF/itemdetail.asp?filterType=dual,%20data&filterValue=2005&filterByDID=2&sortByDID=2&sortOrder=descending&itemID=CMS026525&intNumPerPage=10. Accessed February 5, 2007

36. 2003 ICD-9-CM Professional for Physicians: International Classification of Diseases, 9th Revision, Clinical Modification. 6th ed. Reston, VA: St Anthony Publishing; 2002

37. Glick HA, Doshi JA, Sonnad SS, Polsky D. Economic Evaluation in Clinical Trials. Oxford, England: Oxford University Press; 2007:179–206

38. Briggs AH, Goeree R, Blackhouse G, O'Brien BJ. Probabilistic analysis of cost-effectiveness models: choosing between treatment strategies for gastroesophageal reflux disease. Med Decis Making. 2002;22 (4):290 –308[Abstract/Free Full Text]

39. Pignone M, Saha S, Hoerger T, Lohr KN, Teutsch S, Mandelblatt J. Challenges in systematic reviews of economic analyses. Ann Intern Med. 2005;142 (12):1073 –1079[Abstract/Free Full Text]

40. Thompson SG, Barber JA. How should cost data in pragmatic randomised trials be analysed? BMJ. 2000;320 (7243):1197 –1200[Free Full Text]

41. Burke P, Bain J, Robinson D, Dunleavey J. Acute red ear in children: controlled trial of non-antibiotic treatment in general practice. BMJ. 1991;303 (6802):558 –562[Abstract/Free Full Text]

42. Rosenfeld RM, Kay D. Natural history of untreated otitis media. Laryngoscope. 2003;113 (10):1645 –1657[CrossRef][Web of Science][Medline]

43. Bluestone CD. Clinical course, complications and sequelae of acute otitis media. Pediatr Infect Dis J. 2000;19 (5 suppl):S37 –S46[CrossRef][Web of Science][Medline]

44. Van Zuijlen DA, Schilder AG, Van Balen FA, Hoes AW. National differences in incidence of acute mastoiditis: relationship to prescribing patterns of antibiotics for acute otitis media? Pediatr Infect Dis J. 2001;20 (2):140 –144[Web of Science][Medline]

45. Damoiseaux RA, van Balen FA, Hoes AW, Verheij TJ, de Melker RA. Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children aged under 2 years. BMJ. 2000;320 (7231):350 –354[Abstract/Free Full Text]

46. Takata GS, Chan LS, Shekelle P, Morton SC, Mason W, Marcy SM. Evidence assessment of management of acute otitis media, part I: the role of antibiotics in treatment of uncomplicated acute otitis media. Pediatrics. 2001;108 (2):239 –247[Abstract/Free Full Text]

47. Schilder AG, Lok W, Rovers MM. International perspectives on management of acute otitis media: a qualitative review. Int J Pediatr Otorhinolaryngol. 2004;68 (1):29 –36[CrossRef][Web of Science][Medline]

48. Pichichero ME. Diagnostic accuracy of otitis media and tympanocentesis skills assessment among pediatricians. Eur J Clin Microbiol Infect Dis. 2003;22 (9):519 –524[CrossRef][Web of Science][Medline]

49. Karma PH, Penttila MA, Sipila MM, Kataja MJ. Otoscopic diagnosis of middle ear effusion in acute and non-acute otitis media, part I: the value of different otoscopic findings. Int J Pediatr Otorhinolaryngol. 1989;17 (1):37 –49[CrossRef][Web of Science][Medline]

50. Kaleida PH, Stool SE. Assessment of otoscopists’ accuracy regarding middle-ear effusion: otoscopic validation. Am J Dis Child. 1992;146 (4):433 –435[Abstract/Free Full Text]

51. Kaleida PH, Casselbrant ML, Rockette HE, et al. Amoxicillin or myringotomy or both for acute otitis media: results of a randomized clinical trial. Pediatrics. 1991;87 (4):466 –474[Abstract/Free Full Text]

52. Rosenfeld RM. Antibiotic use for otitis media: oral, topical, or none? Pediatr Ann. 2004;33 (12):833 –842[Web of Science][Medline]

53. Wald ER, Guerra N, Byers C. Upper respiratory tract infections in young children: duration of and frequency of complications. Pediatrics. 1991;87 (2):129 –133[Abstract/Free Full Text]

54. Hoberman A, Paradise JL, Burch DJ, et al. Equivalent efficacy and reduced occurrence of diarrhea from a new formulation of amoxicillin/clavulanate potassium (Augmentin) for treatment of acute otitis media in children. Pediatr Infect Dis J. 1997;16 (5):463 –470[CrossRef][Web of Science][Medline]

55. Bottenfield GW, Burch DJ, Hedrick JA, Schaten R, Rowinski CA, Davies JT. Safety and tolerability of a new formulation (90 mg/kg/day divided every 12 h) of amoxicillin/clavulanate (Augmentin) in the empiric treatment of pediatric acute otitis media caused by drug-resistant Streptococcus pneumoniae. Pediatr Infect Dis J. 1998;17 (10):963 –968[CrossRef][Web of Science][Medline]

56. Engelhard D, Cohen D, Strauss N, Sacks TG, Jorczak-Sarni L, Shapiro M. Randomised study of myringotomy, amoxycillin/clavulanate, or both for acute otitis media in infants. Lancet. 1989;2 (8655):141 –143[CrossRef][Web of Science][Medline]

57. Principi N. Multicentre comparative study of the efficacy and safety of azithromycin compared with amoxicillin/clavulanic acid in the treatment of paediatric patients with otitis media. Eur J Clin Microbiol Infect Dis. 1995;14 (8):669 –676[CrossRef][Web of Science][Medline]

58. Lieu TA, Fleisher GR, Schwartz JS. Cost-effectiveness of rapid latex agglutination testing and throat culture for streptococcal pharyngitis. Pediatrics. 1990;85 (3):246 –256[Abstract/Free Full Text]

59. Ehrlich JE, Demopoulos BP, Daniel KR Jr, Ricarte MC, Glied S. Cost-effectiveness of treatment options for prevention of rheumatic heart disease from group A streptococcal pharyngitis in a pediatric population. Prev Med. 2002;35 (3):250 –257[CrossRef][Web of Science][Medline]

60. Takahashi P, Trang N, Chutka D, Evans J. Antibiotic prescribing and outcomes following treatment of symptomatic urinary tract infections in older women. J Am Med Dir Assoc. 2004;5 (2 suppl):S11 –S15[Medline]

PEDIATRICS (ISSN 1098-4275). ©2008 by the American Academy of Pediatrics

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Facebook Facebook   Add to Reddit Reddit   Add to Technorati Technorati   Add to Twitter Twitter    What's this?


This article has been cited by other articles:


Home page
 J Antimicrob ChemotherHome page
E. K. Vouloumanou, D. E. Karageorgopoulos, M. S. Kazantzi, A. M. Kapaskelis, and M. E. Falagas
Antibiotics versus placebo or watchful waiting for acute otitis media: a meta-analysis of randomized controlled trials
J. Antimicrob. Chemother., July 1, 2009; 64(1): 16 - 24.
[Abstract] [Full Text] [PDF]

Home page
 PediatricsHome page
S. B. Meropol
Valuing Reduced Antibiotic Use for Pediatric Acute Otitis Media
Pediatrics, April 1, 2008; 121(4): 669 - 673.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me when eLetters are posted
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow E-mail this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My File Cabinet
Right arrow Download to citation manager
Right arrowRequest Permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via CrossRef
Right arrow Citing Articles via Web of Science (3)
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Meropol, S. B.
Right arrow Articles by Asch, D. A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Meropol, S. B.
Right arrow Articles by Asch, D. A.
Related Collections
Right arrow Infectious Disease & Immunity
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Facebook   Add to Reddit   Add to Technorati   Add to Twitter  
What's this?