JOURNAL CLUB

JOURNAL CLUB

COMMENTARIES BY MICHAEL G. BURKE, MD

Does knowing cost reduce tests?

Investigators set out to determine whether knowing the cost of testswould affect doctors' ordering habits or the quality of patient care ina pediatric emergency department (ED). They tracked the visits of about5,000 patients 2 months to 10 years of age to an urban, university-affiliatedpediatric ED. In two consecutive three-month periods, physicians placedorders for presenting patients from a list of 22 tests that was attachedto each chart. In the first "control" period no price informationwas provided for the tests; in the second period the list included standardhospital charges for each test. During a third, brief washout period, priceswere not included with the list of tests. Investigators interviewed patientfamilies one week after each ED visit to find out whether the child wasseen again by a health-care worker and to ascertain overall satisfactionwith the initial visit.

Charges for tests in patients who visited the ED during the interventionperiod when physicians had price information were 27% less than chargesin the control period, after controlling for triage level, vital signs,and admissions rates. The largest decrease (43%) was among nonadmitted patientsin lower-acuity triage categories. The use of inexpensive tests, such asurine dipstick, glucometer measurement, rapid streptococcal tests, and cultures,changed little. Common but more expensive tests, such as chest radiographyand serum electrolyte studies, were ordered significantly less often duringthe intervention period.

Telephone follow-up showed that patients in the intervention period wereslightly more likely than patients in the control period to make an unscheduledfollow-up visit to a health-care provider. Those who visited the ED duringthe intervention period were only slightly less likely to report that thechild was "better," however, and about the same number of familiesin each period said that they were "very satisfied" or "somewhatsatisfied" with the ED visit. Adjusted charges in the washout periodwere 15% lower than in the control period and 15% higher than in the interventionperiod, suggesting that the effects of appending prices to the charts werereduced but not entirely lost once the intervention was over (Hampers LCet al: Pediatrics 1999;103[4]:877).

Commentary: This is a good reminder. Do the test if it's needed,but remember that someone is paying the bill. And maybe it's good to knowhow much the bill is going to be.

Pacifiers and breastfeeding

Mothers who give their child a pacifier before the infant is 6 weeksold breastfeed for a significantly shorter period than mothers who don'tuse a pacifier by this age, a new study shows. Early pacifier use does notaffect whether a mother continues to breastfeed during the first three monthsof her infant's life, however; the effect is primarily on long-term breastfeeding.

Investigators interviewed 265 breastfeeding mothers about pacifier useand breastfeeding at two, six, 12, and 24 weeks' postpartum and every 90days thereafter until breastfeeding ended. By the time infants were 6 monthsof age, 74% of mothers had begun offering a pacifier. These pacifier userstended to breastfeed their infants less often than mothers who had not introduceda pacifier. At 12 weeks postpartum, mothers whose babies used pacifierswere more likely to report that breastfeeding was inconvenient and thatthey had problems with producing enough breast milk--problems consistentwith infrequent nursing. These findings suggest that women who introducepacifiers to their infants at an early age tend to breastfeed for fewermonths than women who don't introduce a pacifier at this time because theearly pacifier users breastfeed less frequently than other breastfeedingmothers (Howard CR et al: Pediatrics 1999; 103(3):e33).

Commentary: I remember well the days when I had a "Binkie"in every pocket when I was at home. Having a pacifier nearby often allowedmy wife to complete a thought or a shower before the next breastfeeding,so I don't like the idea of giving up on pacifiers completely.

As the authors point out, this was an observational study. Although itshowed that mothers who introduced pacifiers early breastfed for a shortertime than mothers who didn't, it doesn't attempt to claim cause and effect.Perhaps those mothers were less enthusiastic than others about breastfeedingto begin with, had schedules that were less conducive to feeding on demand,or introduced the pacifier after deciding to wean their babies.

Dr. Ruth Lawrence, one of the authors of this article, says in her textbookBreastfeeding, "Clearly, pacifiers are a parental decision." Ithink her statement remains true until we have further information.

Also of note

Increase iron by cooking in iron pots. Children fed food cooked in ironpots have less anemia and faster growth than children whose food is cookedin aluminum pots, according to two studies to assess the effects of typesof cookware used in Ethiopian homes.

In a laboratory study, investigators compared total and available ironin traditional Ethiopian foods cooked in iron pots with that of food cookedin clay and aluminum pots. Meat and vegetables cooked in an iron pot hadabout five times more available iron than these foods cooked in pots thatwere not iron. The type of pot made no difference in available iron forlegumes.

In the second study, a clinical trial set in a semiurban community inEthiopia, investigators gave the households of 407 children an iron potor an aluminum pot for cooking the family's food. (Clay pots were not usedin this trial.) By the end of the 12-month study, mean hemoglobin of childrenin the iron-pot group had increased 1.3 g/dL more than that of childrenin the aluminum pot group. Children in the iron pot group also grew morethan children in the aluminum-pot group (Adish AA et al: Lancet 1999;353:712).

Another therapy for neonatal immune hemolytic disease (NIHD).A studyof newborn term infants with NIHD caused by blood group and Rh factor incompatibilitiesindicates that high-dose intravenous immunoglobulin (HDIVIG) therapy reducesserum bilirubin levels and the need for exchange transfusions. The 116 infantsin the study were divided into two groups. Those in the treatment groupreceived conventional phototherapy plus HDI­VIG. The babies in the controlgrouphad conventional phototherapy alone.

After beginning treatment, eight patients in the HDIVIG group required11 exchange transfusions because of continued high serum bilirubin concentrations.This compares with 29 exchange transfusions performed in 22 babies in thecontrol group. Infants in the HDIVIG group also required significantly shorterperiods of phototherapy and hospitalization than those in the control group.HDIVIG treatment did not produce any observable side effects (Alpay F etal: Acta Paediatr 1999;88: 216).