Neonatal extracorporeal membrane oxygenation—a bridging technique See page 75 The technique developed for cardiopulmonary bypass was first adapted for use in moribund infants by Bartlett and co-workers in 1976.1 Later Bartlett showed that infants with a high likelihood of dying from respiratory failure could be salvaged by extracorporeal membrane oxygenation (ECMO).2 Given the challenge of testing this promising technology, investigators chose novel adaptive experimental designs that would allow for the fewest number of infants to be denied access to this “potentially lifesaving therapy” yet still give a statistically valid answer as to whether ECMO saved lives.3,4 In the “play the winner” strategy of Bartlett and co-workers, the only infant randomised to receive conventional management subsequently died; the remaining 11 infants received ECMO and survived.3 ECMO passed the statistical test, but failed to convince the medical community worldwide. Investigators from different countries evaluated this information and proceeded in very different directions. Whether the route taken was based on interpretation of the evidence, or on the economics of health care and research in the different countries, is hard to say. In the 1980s, ECMO centres began to proliferate in the USA. The US Neonatal ECMO Registry was started by the University of Michigan to record use and potential complications of ECMO therapy. Despite wide recognition that the potential complications from this highly invasive procedure had not been fully documented, in a society that has placed few limits on medical care few were willing to withhold life-saving therapy in order to conduct a more thorough evaluation of this technique.5 75 centres in the United States and 17 other centres worldwide contribute data to the ECMO Registry. Up to the end of 1992, centres belonging to the Extracorporeal Life Support Organization have registered the outcomes of 7647 infants receiving ECMO.6 The conditions that led to severe respiratory failure in term infants were a heterogeneous group of disorders (panel). In view of how gravely ill babies are before consideration for ECMO, the overall survival rate of 81% is impressive. Survival is highly influenced by underlying diagnosis: meconium aspiration syndrome fares best (93% survival), and congenital diaphragmatic hernia worst (59% survival). The Neonatal ECMO registry also tracks complications related to ECMO, including mechanical and patientspecific complications. The most worrying data from the registry suggests that 17% of infants experience a cranial infarct or haemorrhage, and of these 57% survive.
Centres in the United Kingdom were more cautious than their US counterparts in their approach to the introduction of ECMO. In general, severe respiratory failure in the term newborn did not seem to be as pressing a problem.7 Investigators in the United Kingdom report a substantially lower rate of meconium aspiration syndrome and persistent fetal circulation—disorders responsible for nearly half the ECMO cases in the United States. The response in the UK was to organise the UK Collaborative ECMO Trial Group, which first had to grapple with the thorny ethical problem of whether there was sufficient 70
Causes of severe respiratory failure recorded by ECMO registry Disorder Meconium-aspiration syndrome Congenital diaphragmatic hernia Pneumonia or sepsis Type II respiratory distress syndrome Persistent pulmonary hypertension Other cardiopulmonary disorders
% 37% 19% 15% 13% 12% 4%
evidence to institute therapy routinely. 51 centres evaluated the effectiveness of ECMO in term infants with severe respiratory failure from a variety of causes. The results, in this week’s Lancet, not surprisingly show that ECMO substantially decreased mortality (relative risk 0·55, 95% CI 0·39, 0·77), an effect equivalent to producing one extra survivor for every three or four infants allocated to ECMO. The UK collaborative trial also gives an estimate of severe neurodevelopmental impairment or disability at 1 year of life; the estimate suggests that the improvement in mortality is not associated with poor developmental outcome. A more detailed analysis of developmental outcome and costs will be available soon. The precise estimate of the impact on mortality as well as careful developmental follow up is the major contribution of this collaborative experience. Even as the UK group presents the most interpretable experience with ECMO, other therapies that might supplant ECMO in the treatment of respiratory failure in the term newborn are being developed.8 Nitric oxide, a mediator of a variety of physiological processes, including the regulation of pulmonary vascular tone, shows promise for the treatment of persistent pulmonary hypertension. Other modes of ventilation, especially high-frequency oscillation, are being tested in infants with respiratory failure. In meconium aspiration syndrome, natural surfactant extracts may prove to be successful in improving respiratory status. All these therapies, if sucessful, represent solutions that are less invasive and less expensive than ECMO, and universally applicable. In the trials of these interventions, a common outcome measure has been “the need for ECMO, or death”, which shows that, although ECMO is of benefit, ultimately the use of this technique is considered an undesirable outcome. ECMO is of proven benefit in preventing mortality in term infants with severe respiratory failure. The experience of the UK collaborative group allows clinicians to give specific estimates of that effect, as well as to judge the counterbalancing effect on development. How the UK group will proceed will be interesting to follow. Although effective, ECMO is clearly a bridging technique, to be supplanted by any number of treatments that would be considered superior. Having proved ECMO effective, the UK group should use the same infrastructure to test many of the other potential therapies that will lead to ECMO being a historical footnote.
Roger F Soll Department of Pediatrics, University of Vermont, Vermont 05405-0068, USA 1
Bartlett RH, Gazzaniga AB, Jeffries MR, Huxtable RF, Haidue NJ, Fong SW. Extracorporeal membrane oxygenation (ECMO) cardiopulmonary support in infancy. Trans Am Soc Artif Intern Organs 1976; 22: 80–93.
Vol 348 • July 13, 1996
Bartlett RH, Andrews AF, Toomasian JM, Haiduc NJ, Gazzaniga AB. Extracorporeal Membrane Oxygenation (ECMO) for newborn respiratory failure: 45 cases. Surgery 1982; 92: 425–33. Bartlett RH, Roloff DW, Cornell RG, Andrews AF, Dillon PW, Zwischenberger JB. Extracorporeal circulation in neonatal respiratory failure: a prospective randomized study. Pediatrics 1985; 76: 479–87. Ware JH, Epstein MF. Extracorporeal circulation in neonatal respiratory failure: a prospective randomized study. Pediatrics 1985; 76: 849–51. Greer AL. Diffusion of medical technology: the case for ECMO. Report of the workshop on diffusion of ECMO technology. US Department of Health and Human Services. 1993. NIH Publication No 93-3399. Kanto WP. A decade of experience with neonatal extracorporeal membrane oxygenation. J Pediatr 1994; 124: 335–47. Editorial. Persistent fetal circulation and extracorporeal membrane oxygenation. Lancet 1988; ii: 1289–91.
Impact of daytime sleepiness underrated Sleepiness is appropriate and desirable at the usual retiring time. However, there is also a circadian propensity to feel mildly sleepy in the early to mid afternoon (between 2.00 and 5.00 pm). Excessive daytime sleepiness is an important symptom of several chronic sleep disorders, including obstructive sleep apnoea, narcolepsy, idiopathic hypersomnia, insomnia, and periodic limb-movement disorder. It is a common but often poorly recognised health problem. The tendency to sleep is increased by sleep disruption and sleep deprivation. Sleep disruption seems more detrimental than low total sleep duration. The effects of sleep loss are cumulative—sleepiness increases progressively with the sleep debt. It is, of course, important but often difficult to separate daytime sleepiness from less specific symptoms such as tiredness or fatigue. In a recent survey of a Finnish population (11 354 adults; 66% women; aged 33–60 years), 11·0% of women and 6·7% of men reported daytime sleepiness every or almost every day.1 In this study daytime sleepiness was most commonly associated with depression, insomnia, and sleep-disordered breathing (snoring and apnoeas), and to a lesser extent with the use of hypnotics and tranquillisers and insufficient sleep. Sleep-disordered breathing was about twice as common in men with sleepiness than in women. Patients with depressive illness almost invariably complain of some form of sleep disorder—most commonly diminished and unrefreshing sleep, with resultant daytime sleepiness.2 Antidepressants produce differing acute effects on measures of sleep continuity, although resolution of the depressive disorder is usually accompanied by improvements in sleep continuity. Tertiary tricyclic antidepressants—for example, imipramine and dothiepin—commonly cause sedation and produce “hangover” effects the following day. By contrast, selective serotonin reuptake inhibitors—for example, fluoxetine—generally do not sedate but can cause sleep disruption. As with the tricyclic antidepressants, daytime somnolence may result with the selective serotonin reuptake inhibitors, perhaps in part because of insufficient sleep. These daytime sleep disturbances may affect compliance with treatment. Drug tolerability profiles are important in terms of the quality of the patient’s compliance with treatment, which in turn can affect relapse rates. A meta-analysis looking at sleep deprivation and
Vol 348 • July 13, 1996
performance has shown that sleep deprivation has a considerable effect on human functioning. 3 Sleep deprivation was strongly associated with low mood, followed by impairment of cognitive and motor performance. Interestingly, partial sleep deprivation had a greater overall effect than either short-term or long-term sleep deprivation. It is possible that the effects of partial sleep deprivation more closely resemble those of sleep fragmentation than those of total sleep deprivation. Attention has been focused in recent years on the relation btween sleepiness and accidents. Accidents were the fourth leading cause of mortality, with motor-vehicle accidents representing 51% of total deaths due to accidents in the USA.4 Sleepiness seems to be an underrated factor in accidents, not only as a primary cause but also as a co-factor related to other well-known factors, such as alcohol or drug use and hazardous driving conditions. Hypnotics may also be a contributory factor to accidents because of their residual daytime effects. It has been reported that patients taking benzodiazepines were more likely to have automobile accidents than when they were off medication.5 In the elderly, patients receiving daily doses of 125 mg amitriptyline equivalent or greater had a sixfold higher rate of car accidents than did those off medication.6 However, the peak of motorvehicle fatalities occurs among 21-year-olds. Young adults commonly have poor sleep habits (voluntarily reduced nocturnal sleep time), which result in daytime sleepiness. The Stanford Sleep Disorders Clinic statistics show that of all patients with sleep apnoeas, 15–45% have had at least one accident related to sleepiness (12–30% suffering from narcolepsy, and 2–8% suffering from insomnia). Daytime sleepiness is a common condition that can have potentially serious effects. The long-term resolution of the problem is likely to require a better understanding of the mechanisms of sleep itself.
Ann Sharpley Psychopharmacology Research Unit, Littlemore Hospital, Oxford OX4 4XN, UK 1
3 4 5 6
Hublin C, Kaprio J, Partinen M, Heikkila K, Koskenvuo M. Daytime sleepiness in an adult, Finnish population. J Intern Med 1996; 239: 417–23. Ford DE, Kamerow DB. Epidemiologic study of sleep disturbances and psychiatric disorders: an opportunity for prevention? JAMA 1989; 262: 1479–84. Pilcher JJ, Huffcutt AI. Effects of sleep deprivation on performance: a meta-analysis. Sleep 1996; 19: 318–26. Leger D. The cost of sleep-related accidents: a report for the national commission on sleep disorders research. Sleep 1994; 17: 84–93. Binnie GA. Psychotropic drugs and accidents in a general practice. BMJ 1983; 287: 1349–50. Ray WA, Thapa PB, Shorr RI. Medication and the older driver. Clin Geriatr Med 1993; 9: 413–38.
Why the resistance to diagnostic imaging in childhood urinary tract infections Some 3–5% of girls and 1–2% of boys will have at least one symptomatic urinary tract infection (UTI) before puberty.1 Reflux-associated nephropathy is the term currently used to describe renal scarring and subsequent renal damage in a child known to have vesico-ureteric reflux (VUR) and UTIs. Reflux-associated nephropathy is reported to account for 5–25% of cases of end-stage renal failure in childhood worldwide.2,3 Accordingly, it has 71