Arthur Lavin

Conflict of Interest and Purpose in Bilirubin Screening

ranted criticism of our design. We used a nonequivalence design and large sample size to test a null hypothesis about expected small differences generated from known pharmacokinetic and pharmacodynamic properties of CON3 and MCD.4 Evaluation of “substitutes” would require a noninferiority rather than nonequivalence design and acceptance not rejection of the null hypothesis. Dr Adesman implies that our choice of doses was a planned methodologic flaw related to the source of support and alludes to differences in the immediate-release (IR) components of CON and MCD to support his contention. However, in our article, we clearly stated that “MCD releases 50% more IR MPH . . . than CON”2(p.e208) for near-equal daily doses, and we used this5 to formulate our null hypothesis of no difference, which we were able to reject. Dr Adesman objects to our primary hypothesis about effects during the school day, but we also evaluated a later time point and reported at the 12-hour assessment that “CON was statistically significantly better than MCD and PLA [placebo]”2(p.e210) (see figure 1 in ref 2). A biased study would have included only time points when superiority of MCD was expected and not an assessment time when superiority of CON was expected. We also performed secondary analyses of morning effects, which showed that for “the 2 dose conditions with equal 12-mg IR MPH boluses (MCD 40 and CON 54), the ESs [effect sizes] were large and indistinguishable.”2(p.e207) For the doses evaluated, our study showed no overall superiority for either product but instead documented a changing pattern of superiority throughout the 12-hour study day and explained it in terms of known pharmacokinetic/pharmacodynamic relationships. We believe that this result demonstrates scientific rigor (not bias) and is highly relevant (not irrelevant) because it provides an empirical basis for practical choices among alternative options for clinical treatment. Food may6 or may not7 affect the pharmacokinetic properties of methylphenidate, but few studies8 have evaluated food effects on efficacy. It was not feasible to double the size of an already large study to address this factor, but in our methods section we should have stated that food intake was controlled by administering the morning dose after an overnight fast and providing a standardized breakfast 1 hour later.

A survey of pediatric management of dyslipidemias in new England

SummaryWe recently surveyed physicians attending the New England Pediatric Preventive Cardiology Society. Sixteen physicians who actively evaluated children with dyslipidemia completed questionnaires; at least one representative from six of the seven medical schools in New England was included. The survey elicited responses to five hypothetical cases of childhood dyslipidemia which were representative of the types of lipid problems commonly referred to pediatric lipid specialists. Diet modification was the initial treatment of choice of all participants. For any set of lipid values, postpubertal age increased the proportion of respondents who would have prescribed medication. When pharmacologic intervention was elected, resin binders (cholestyramine or cholestipol) and niacin were most commonly prescribed. The responses of the physicians showed considerable variation in the threshold for beginning medications. In summary, this survey suggests substantial variation in the approach to pharmacologic management of pediatric dyslipidemias in the New England region.