complacent. We obviously had not recognised the perception that we represented the regime and were its mouthpiece; that the situation was more sensitive than we had realised; that feelings of hurt and anguish ran deep; that we had in fact been resigned to accepting an atrocious situation. I apologised for this complacency, for the loss of dignity our colleagues had suffered for so long, and for all our errors of and expressed the heartfelt hope that we could look forward to a future of reconciliation and join hands in peace and friendship. It is with very great joy that in the subsequent two years I have seen such forgiveness and willingness to put the past aside, both in medical circles and in our country as a whole. J.Adno Suite 6. Park Lane Clinic Parktown Johannesburg 1. Van der Unde I. Sometimes having to say you're sorry (Izindaba). S Atr Med J 1995; 85: 715. The effectiveness and cost of exogenous pulmonary surfactant replacement therapy To the Editor: The articles"" recently published on surfactant replacement therapy (SR1) in the treatment of newborn infants with hyaline membrane disease (HMD) being ventilated in a country with limited health resources are both timeous and urgently needed. Surfactant, being an expensive drug, needs to be administered in the most cost- effective fashion and for this reason research guidelines for its administration in South Africa are needed. We are concerned that the discussions of this series of articles do not sufficiently emphasise their limitations, as well as the enormous impact of SRT on survival rates and other neonatal morbidities in newborn infants ventilated with HMD. It has been stated that SRT is one of the many recent triumphs of medical science. s The efficacy of both natural and synthetic exogenous surfactant for prevention and treatment of HMD has been confirmed in several large randomised clinical trials." The impact on mortality in the USA since its release in 1990 has been dramatic, with decreases of 15% in overall neonatal mortality and as high as 50% in high-risk groups such as infants delivered at less than 29 weeks' gestation or weighing less than 1 200 g. The costs saved are also considerable, with Soli et a/. 8 reporting a reduction of US$ 3 319 per 28-day survivor for babies who received exogenous SRT (natural surfactant) versus sham-air. Calculations that take into account both the offset of savings incurred by increasing numbers of very-Iow-birth- weight infants and the higher costs for babies who survived instead of dying, suggest that the USA saved an estimated $90 million (1985 - 1990).9 Like the local studies, these studies do not show a net reduction in the total number of days of hospitalisation!·4 Since these calculations might not be applicable to a country with limited financial resources, their use in South Africa requires investigation. The indications investigated by Ballot et al.' to prevent the unnecessary use of SRT stimulate further debate. In the group of neonates with severe HMD (Table Ill, group 2) where SRT was withheld for longer than 6 hours, the mortality rate was an unacceptable 38% (5/13). This did not significantly differ from the mortality rate (22%, 2/9) experienced in the group of babies (group 1) who received early SRT. This lack of difference is probably due to a type 2 statistical error (too few babies enrolled). In neonates with severe HMD the delayed administration led to .a 15% reduction in SRT use. The authors acknowledge that this is not justified and that SRT should be administered early to neonates with severe HMD. The study design did not allow for the inclusion of a group of neonates with moderate HMD (fractional inspired oxygen concentration (Fi0 2 ) < 0,75) who received SRT within 6 hours of birth. 'lYe consider that this is a major limitation, since we do not know whether there would have been a reduction in the morbidity (duration of ventilation, pneumothorax rate or length of hospitalisation) of these neonates when compared with the neonates who received SRT after 6 hours. These data·;are essential for formulating guidelines and calculating CQsts. A subsequent article by Davies et a/. 4 assessed and compared the cost and effectiveness of a policy of delp.yed surfactant replacement therapy (SR1) in a group of bgbies with HMD versus that of a historical control group using guidelines similar to those developed by Ballot et al. 2 The authors show that SRT (Survanta; Abbott) led to an increase in the total cost of treating a baby ventilated for HMD. A critical review of these published data is essential, as they could have far-reaching consequences for the effective care of neonates with HMD in southern Africa. The authors studied ventilated babies, selected to receive SRT according to their initial oxygen requirements (arterial/alveolar oxygen ratio), and compared them with a retrospective group of babies ventilated at their institution before the introduction of SRT. As the authors report, there are significant demographic differences between the SRT and control groups. The babies in the SRT group were of lower birth weight, received less antenatal care (72% v. 15% un booked) and included more black infants (74% v. 28%). The infants in the SRT group probably had more severe disease, as reflected by the fact that more of them needed inotropic support and paralysis, although no recognised objective evidence of the severity of the HMD such as arterial/alveolar oxygen ratio for the two groups is presented. We want to emphasise firstly the differences between the SRT and control groups and secondly the limitations of the indications used for the administration of SRT. We believe that the conclusions drawn from the data should be interpreted with caution. It is not clear to us why the use of SRT did not lead to a reduction in mortality and morbidity, as has been universally reported. Clarification for this could be found in a faulty study design or in the neonates selected to receive SRT. Ballot et al. 2 included 1 baby with a congenital heart lesion, 4 with presumed bacterial pneumonia, and, as pointed out in the article, an unknown number of severely asphyxiated babies. To date more than 35 randomised controlled trials on SRT have now been conducted involving more than 6 000 babies and have demonstrated a consistent 40% reduction in the odds of neonatal death after surfactant treatment." Treatment of established HMD offers the advantage of limiting the number of individuals treated to only those babies with definite HMD. The most appropriate time to treat manifest (rescue) HMD seems to be within the first 6 hours after birth. Thereafter various factors may be operative in 1192 Volume 85 No. 11 November 1995 SAMJ