Article Text


Readmission of neonates
  1. R Scott-Jupp
  1. Correspondence to:
    Dr R Scott-Jupp
    Salisbury District Hospital, Salisbury SP2 8BJ, UK;

Statistics from

Commentary on the papers by Oddie et al (see page 119) and Escobar et al (see page 125)

Two papers coincidentally submitted to Archives from opposite sides of the Atlantic give an opportunity to point out some interesting similarities and differences in the controversial area of hospital readmission shortly after birth. Clinicians in both the UK and the USA are concerned about the knock-on effects of the increasing trend towards earlier neonatal discharge from hospital.

The studies differ significantly in their objectives and methods, so direct numerical comparisons may not be valid. Oddie et al looked at over 11 000 births in the Northern NHS region of the UK in 1998, excluding infants less than 35 weeks gestation.1 They concentrated on factors associated with early neonatal discharge, and then looked at what influenced readmission to hospital within 28 days. Escobar et al studied a population of over 33 000 using the Kaiser Permanente Medical Care Program (KPMCP) in California and Colorado, for which good data are available, in 1998–2000;2 they included all gestations and did not look at early discharge at all, but analysed in some detail factors associated with readmission within two weeks. The KPMCP, although not government run, is a managed healthcare system which has been described as being in many ways similar to the British National Health Service.3 Its membership is not restricted to the more prosperous sections of society, as may be inferred from the racial distribution of infants in some of the institutions in the study (see table 1 in Escobar et al2).

Oddie et al found early discharge to be less likely, not surprisingly, in smaller, lower gestation infants. Social deprivation was associated with earlier discharge; this may reflect economic and family pressures to return home, or possibly a tendency for the more deprived to feel relatively disempowered in negotiating with professionals about the timing of their discharge. Older maternal age was associated with longer stays once other factors were allowed for. Breast feeding led to longer stays, but the association was weak. Their findings on readmission within 28 days are perhaps counter-intuitive to many paediatricians: on multivariate analysis, there was no association between early neonatal discharge and readmission, nor did social deprivation have an influence. Indeed the only significant independent associations they found were breast feeding with reduced readmission rates, and lower gestation with increased rates. They did not study factors such as community support and postnatal visits, but there was considerable variation between the five hospitals in the study. Clinical reasons for readmission were, as expected, a broad range of postnatal problems, probably including a nebulous group of “parental anxiety” admissions with no firm diagnosis; about 9% were primarily because of jaundice.

Escobar et al looked at a different range of factors influencing readmission in the first two weeks. As in the UK study, lower gestations were more likely to be readmitted (including those that had avoided the neonatal intensive care unit), as were those who were sicker at birth. They did not look at feeding mode, but were able to determine which had received home health visits or had attended outpatients: those who had been visited at home in the first 72 hours after discharge were less likely to be readmitted, unless they had also attended outpatients, in which case readmission was more likely. Presumably this latter category reflects a more vulnerable group. Analysis of race revealed that African-American infants were less likely to be readmitted: if, as is often suggested, race in the USA is taken as a proxy for social deprivation, then this represents an interesting difference from the UK study. Asian infants were more likely to be readmitted, but this was almost entirely explained by jaundice. Again, the diagnoses leading to readmission were as expected, but jaundice was the reason in 34%, much more frequent than in the UK study. This difference is unlikely to be completely explained by the inclusion of more preterm infants in the US study. Management of early neonatal jaundice has been the subject of recent reviews, and clearly practice varies between institutions and between nations.4,5 American practice may be much more cautious in preventing high bilirubin levels. However, as discussed by Escobar et al, home phototherapy has an important role here.

A major difference between the two countries is the statutory provision in the UK of a community based midwifery and health visiting service. These two professions are largely non-existent in the USA. The American study found fewer readmissions for those who received a home visit as part of their KPMCP “package”, and this might be analogous to the British postnatal midwife’s visit. Unfortunately the UK study had no data on postnatal visits, nor on their communities’ policies in this area. It cannot be assumed that postnatal home visits will always reduce readmission rates: where the parent’s anxiety exceeds that of the health professional it may do so, but where the parent had been unconcerned and the professional detects a problem, it may have the opposite effect.

A strikingly consistent theme in both studies is the huge variation in readmission rates between neighbouring hospitals. This actually accounts for more of the variation than the “biological” factors already discussed. This is not apparently explained by differences in the degree of social deprivation in the populations served. It is difficult to identify reasons for this: associations with institutional characteristics looked for in the US study were weak and inconsistent, although availability of home phototherapy appears to have some influence. Presumably, in both countries, traditional policies and practices vary between institutions for historical reasons that may have nothing to do with their patient population profile.

From the perspective of a manager or health economist, it might appear that if all institutions could adopt the practices of those with the earliest discharge and lowest readmission rates, then major financial savings could be made. However, we should be cautious about going down this road on the basis of these data, as we know nothing about outcomes, including, possibly, rare but potentially avoidable death and disability; moreover, pressure to send newborns home early, and then to avoid readmission at all costs, may add to the turmoil and anxiety experienced by some vulnerable families that we are unable to quantify.

Commentary on the papers by Oddie et al (see page 119) and Escobar et al (see page 125)


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles