Skip navigation

Antenatal corticosteroid therapy


"The first RCT on the effects of giving a short course of corticosteroids to women expecting to give birth prematurely was reported in 1972. The Cochrane Collaboration logo summarises the evidence that would have been revealed had a systematic review of available RCTs been performed a decade later in 1982. Those who have seen the logo will notice a diamond lying to the left of a vertical line which indicates no effect; this marks that the result produced significant benefit when the results of seven trials were combined.

The reality was that it was 1989 that a systematic review of the RCTs had been prepared and published, but the problem seems to be that uptake of knowledge is very slow."

These words appeared in Bandolier 13 in 1995, but the issue of steroids in preterm delivery had been examined in issues 1 and 2 of Bandolier . There is an historic context as well, because as long ago as 1970 Bandolier was working on tests of foetal pulmonary maturity. By 1976 they were no longer needed because of the use of steroids. What a surprise, then, in 1993 to sit in meetings discussing the promotion of their use in the same hospital in which their use was standard practice 16 years previously.

Now it's déjà vu all over again as a randomised trial is published testing mechanisms to increase uptake of their use in appropriate women [1]. The study has the merit of giving good advice about how complex organisations can make sure that the right thing is done right more often.


This was a study in which 27 hospitals were randomised to a usual dissemination intervention or an active dissemination intervention. This followed the NIH consensus panel recommendations ( Bandolier 13 ). The hospitals were all tertiary care facilities with neonatal intensive care units, those most likely to adopt the NIH recommendations about the use of preterm steroids faster.

Usual dissemination included mailing a brochure on the NIH recommendations to hospitals, universities, medical societies and obstetricians, and via the American College of Obstetricians and Gynecologists to its membership. The recommendations were also carried in JAMA and a supplement to the American Journal of Obstetrics and Gynecology.

Intervention hospitals, in addition received a staggered five-component series of actions, comprising:

  1. An influential physician and nurse coordinator at each hospital to influence institutional dissemination.
  2. A grand rounds lecture on the use of preterm steroids by a nationally respected expert.
  3. A chart reminder system to prompt physicians to consider prescribing therapy on a timely basis.
  4. Group discussions led by influential physicians discussing four case scenarios in which steroids might be used.
  5. Monitoring of care by providing feedback to physicians. Nurse coordinators kept logs of admissions and deliveries.

Medical charts from 3516 women were examined in the baseline year (to February 1994), and 3282 in the year following the NIH conference proceedings (April 1995 to July 1996). These were eligible cases including all women giving birth at 34 weeks or less, including cases of spontaneous preterm labour, premature rupture of membranes, and preterm delivery indicated for medical conditions.


In the year before the NIH conference, steroids were used in 33% of appropriate women. Usual dissemination increased this to 58% of appropriate women in the year after the conference, an increase of 75%. Active dissemination gave a significantly greater increase to 68% of appropriate women, an increase of 105% (Figure).

Figure: Steroid use in appropriate cases before and after NIH conference


At a minimum active dissemination hastened the adoption of steroid therapy by 13 women per hospital in the first year. Given the reduction in morbidity and mortality these results are important. The estimate was that cost savings of $3000 per treated neonate in intensive care resulted from the use of preterm steroids. The estimate for the costs of active implementation were between $77 and $1231 per treated neonate, depending on how costs of active dissemination were determined.

Active dissemination of good evidence benefits patients, and reduces cost burdens on institutions. The paper is worth a read because of the way it describes how active dissemination can be done, generically and not just specifically for this particular condition.

The paper also discusses considerable inter-hospital variation, both in initial use of steroids and in their uptake. Institutions are complex organisations, and the paths of righteousness are seldom straight. So whatever is attempted needs to take account of local conditions and overcome local problems.

The bottom line, though, is that the paper is one that answers the question about whether adoption of the evidence makes things better. Sure does.


  1. LC Leviton et al. Methods to encourage the use of antenatal corticosteroid therapy for fetal maturation. A randomized controlled trial. JAMA 1999 281: 46-52.
previous or next story in this issue