strongest conclusions of this study are that rapid influenzatesting adds very little to clinical decision-making, and thatinfluenza-vaccinated children had lower costs and betterhealth outcomes than non-vaccinated children, regardless oftreatment strategy chosen. The latter result was curiouslydownplayed in this report; in my view, it is perhaps the mostimportant one!

Geoffrey A. Weinberg, MDUniversity of Rochester School of Medicine & Dentistry

Rochester, NY 14642

REFERENCES1. Poehling KA, Walker F, Copeland J, Edwards KM, Szilagyi P, Hall C,et al. Epidemiology of outpatient visits for influenza infections in youngchildren. Pediatr Res 2004;55(4 part 2):241A.2. Poehling KA, Griffin MR, Dittus RS, Tang YW, Holland K, Li H, etal. Bedside diagnosis of influenzavirus infections in hospitalized children.Pediatrics 2002;110:83-8.3. Griffin MR, Walker FJ, Iwane MK, Weinberg GA, Staat MA, ErdmanDD, et al. Epidemiology of respiratory infections in young children: insightsfrom the new vaccine surveillance network. Pediatr Infect Dis J2004:S188-92.4. Centers for Disease Control and Prevention. High levels of adamantaneresistance among influenza A (H3N2) viruses and interim guidelines for useof antiviral agents–United States, 2005-06 influenza season. MMWR MorbMortal Wkly Rep 2006;55:44-6.5. Ison MG, Gubareva LV, Atmar RL, Treanor J, Hayden FG. Recoveryof drug-resistant influenza virus from immunocompromised patients: a caseseries. J Infect Dis. 2006;193:760-4.6. Monto AS. The threat of an avian influenza pandemic. N Engl J Med2005;352:323-5.

Depot medroxyprogesterone results inincreased weight gain compared to oralcontraceptivesBonny AE, Ziegler J, Harvey R, Debanne SM, Secic M,Cromer BA. Weight gain in obese and nonobese adolescentgirls initiating depot medroxyprogesterone, oral contraceptivepills, or no hormonal contraceptive method. Arch PediatrAdolesc Med 2006;160:40-5.

Question Does the initiation of depot medroxyprogesteroneacetate (DMPA) result in significant weight gain in obese andnonobese adolescent girls, when compared with an oral con-traceptive (OC), or no hormonal contraceptive method (con-trol)?

Design Non-randomized, prospective study.

Setting Four urban health clinics.

Participants 450 adolescent girls, aged 12 to 18 years.

Intervention The adolescents were treated with DMPA,OC, or control, based on their preferences.

Outcomes Weight was assessed at 6, 12, and 18 monthsand also as a mean change over 18 months.

Results Adolescent girls who were obese at initiation ofDMPA gained significantly more weight than did obese girlsstarting OC or control (P�.001 for both). At 18 months,

mean weight gain was 9.4, 0.2, and 3.1 kg for obese girlsreceiving DMPA, receiving OC, and control, respectively.Weight gain in obese girls receiving DMPA was also greaterthan weight gain in all nonobese categories (4.0 kg, DMPA;2.8 kg, OC; 3.5 kg, control; P�.001). A significant interac-tion (P�.006) between length of time receiving DMPA andweight gain was evident for obese subjects.

Conclusions Over 18 months, DMPA use was associatedwith increasing rates of weight gain in obese subjects. Thepotential contribution to severe obesity in this population isconcerning.

Comment This is another interesting study that shows anincrease in weight gain in adolescents using hormonal con-traception. The investigators used data that had been col-lected from a 2-year prospective cohort study evaluating hor-monal contraception and bone mineral density changes inadolescents. The participants were followed for 18 months,which is longer than most studies in the field. Unfortunately,the baseline characteristics of the cohorts have significantdifferences. The control group had a younger age, lowerbaseline weight, and decreased rate of sexual activity whencompared to both treatment groups. The DMPA cohort hada higher rate of prior pregnancy than the OC group (5.1% vs2.9%). The attrition rate was high for all groups; consistentwith other studies evaluating contraception in adolescents,however it was particularly high for the OC group (46.3%OC, 37.4% DMPA, 25% controls). This is a well-designedstudy, but the baseline differences do raise some caution. Forexample, the effect of prior pregnancy on the decision tocontinue DMPA despite weight gain is unknown. Also, thehigher dropout rate in the OC group may have been second-ary to perceived weight gain. Ultimately, it is unclear whatimpact the baseline differences have had on final data inter-pretation. The excess weight gain seen in these obese adoles-cents using DMPA is concerning. However, this concernmust be balanced with the social and medical complications ofunplanned teen pregnancy. Further study with well-matchedcohorts will be of great benefit.

Pamela Davis, MDUniversity of MichiganAnn Arbor, MI 48109

Placement of ear ventilation tubes does notimprove long-term hearing threshholds andincreases scarringStenstrom R, Pless IB, Bernard P. Hearing thresholds andtympanic membrane sequelae in children managed medicallyor surgically for otitis media with effusion. Arch PediatrAdolesc Med 2005;159:1151-6.

Question Does the use of ventilation tube insertion inchildren with otitis media with effusion have an effect onhearing thresholds and tympanic membrane pathologic ab-normalities?

Clinical Research Abstracts For Pediatrics 699

Design Prospective cohort study.

Setting Tertiary care children’s hospital, otorhinolaryngol-ogy and audiology service.

Participants Patients aged 8 to 16 years who participated ina randomized controlled trial of medical vs surgical (ventila-tion tube [VT]) treatment for recurrent otitis media witheffusion at ages 2.5 to 7 years.

Methods One hundred thirteen of 125 children who hadparticipated in the trial underwent blinded audiometric, tym-panometric, otomicroscopic, and parental questionnaire eval-uation 6 to 10 years following the trial. Thirty of 56 medicalsubjects received ventilation tubes and 18 of 57 VT subjectsreceived more than 1 set of tubes. To evaluate sequelae riskassociated with ventilation tubes independent of disease se-verity, the authors compared 27 medical subjects who neverreceived ventilation tubes and 38 subjects randomized to VTwho only received 1 set of tubes.

Outcomes Hearing thresholds and tympanic membranesequelae.

Results Tympanic membrane pathologic abnormalities werepresent in 81% of VT subjects and 19% of medical subjects(relative risk, 4.4; 95% confidence interval, 2.2-9.9). Hearingthresholds were 2.1 to 8.1 dB higher in subjects treated withtubes (P�.005).

Conclusions In children who were candidates for ventilationtube insertion randomly assigned to receive medical or VTtreatment for otitis media with effusion, elevated hearingthresholds and tympanic membrane pathologic abnormalitieswere more common in VT subjects 6 to 10 years after inser-tion.

Comment This article reports the findings of a 6 to 9 yearfollow-up assessment in children enrolled in a randomizedcontrolled trial that compared the surgical insertion of a “T”type VT with antibiotic prophylaxis to manage OME. Whencomparing those children who received medical management(MM) and never had VT to 38 children who had surgicalmanagement (SM) with a VT placed in the initial trial butnever required additional VT insertions, tympanosclerosis was4.5 times more common in the SM (66%) compared to theMM children (15%). Other tympanic membrane pathologywas 9.9 times more common in the SM (37%) compared tothe MM children (4%). Children with a VT inserted hadhigher average mean hearing thresholds between 2.13-8.17dB at all tested frequencies; and the relative risk of having ahearing threshold greater than 15 dB was 3.3 times higher inSM compared to MM children. One cautionary note is thatlong term “t-tubes,” now rarely used, compared to buttontubes may cause more sequelae that impact hearing even whenremoved after 12-14 months. These findings support therecommendation that otherwise normal children with persis-tent OME for 4 months or longer only have surgery if theirhearing thresholds are 40 dB or higher.

Stephen Berman, MDUniversity of Colorado Children’s Hospital

Denver, CO 80218

An 8-week behavioral intervention can improveweight gain in children with cystic fibrosisPowers SW, Jones JS, Ferguson KS, Piazza-Waggoner C,Daines C, Acton JD. Randomized clinical trial of behavioraland nutrition treatment to improve energy intake and growthin toddlers and preschoolers with cystic fibrosis. Pediatrics2005;116:1442-50.

Question Among children with cystic fibrosis (CF), will abehavioral and nutrition intervention (BEH) result in im-proved and sustained weight gain, compared to the usual carecontrol condition (CTL)?

Design Randomized, controlled clinical trial.

Setting Cincinnati Children’s Hospital Medical Center,Ohio.

Participants 10 children (out of 14 eligible from their CFCenter), aged 18 to 48 months at the start of the trial.

Methods BEH consisted of eight weekly sessions that in-cluded nutrition counseling to increase energy intake (viatypes of foods and addables/spreadables) and child behavioralmanagement training to teach parents differential attentionand contingency management skills. CTL was consistent withthe 2002 CF Foundation Consensus Conference Guidelinesfor nutritional care. Following the first phase of the trial,those randomized to CTL, were allowed to enter the BEHintervention (5 of 6 chose to do so).

Outcomes The primary outcome was energy intake pre-treatment and post-treatment. The secondary goal was thepercentage of children who met the energy goal of 120% ofthe recommended daily allowance of energy per day orgreater.

Results BEH led to greater increases in energy intake pre-treatment to post-treatment than CTL as measured by calo-ries per day (842 kcal/day vs -131 kcal/day change). Onreceiving BEH, the change in energy intake was replicatedwith the CTL group (892 kcal/day change). At 3- and 12-month follow-up, energy intake was maintained (672 kcal/dayincrease from baseline and 750 kcal/day increase from base-line, respectively). Children in this study met or exceedednormal weight and height velocities from pre-treatment tothe 3-month follow-up (mean weight, 1.4 kg/6 months; meanheight, 5.1 cm/6 months) and from post-treatment to the12-month follow-up (mean weight, 2.5 kg/12 months; meanheight, 8.3 cm/12 months).

Conclusions Toddlers and preschoolers who have CF andreceived BEH were able to meet the energy intake recom-mendations for this disease and maintain these gains up to 12months after treatment. In addition, these children demon-strated weight and height velocities from pretreatment to

700 The Journal of Pediatrics • May 2006