; paper+
Learning Center
Plans & pricing Sign in
Sign Out
Your Federal Quarterly Tax Payments are due April 15th Get Help Now >>



  • pg 1
                        Survival of very premature babies:
                        prospective cohort study of 1994-9 compared
                        with 2000-5
                         This month’s paper                                                    intensive care, death before discharge from neonatal intensive
                         “Survival of extremely premature babies in a geographically           care, and survival to discharge home in two time periods:
                         defined population: prospective cohort study of 1994-9                1994-9 and 2000-5 inclusive.
                         compared with 2000-5” by David J Field and colleagues (BMJ            Results—The proportion of infants dying in delivery rooms was
                         2008;336:1221-3; doi: 10.1136/bmj.39555.670718.BE).                   similar in the two periods, but a significant improvement was seen
                         You can read the paper and responses to it by going to                in the number of infants surviving to discharge (P<0.001). Of 497
                         www.student.bmj.com and clicking on the link.                         infants admitted to neonatal intensive care in 2000-5, 236 (47%)
                                                                                               survived to discharge compared with 174/490 (36%) in 1994.
                                                                                               These changes were attributable to substantial improvements
                         Objective—To assess changes in survival for infants born before 26    in the survival of infants born at 24 and 25 weeks. During the
                         completed weeks of gestation.                                         12 years of the study none of the 150 infants born at 22 weeks’
                         Design— Prospective cohort study in a geographically defined          gestation survived. Of the infants born at 23 weeks who were
                         population.                                                           admitted to intensive care, there was no significant improvement
                         Setting— Former Trent health region of the United Kingdom.            in survival to discharge in 2000-5 (12/65 (18%) in 2000-5 v
                         Subjects— All infants born at 22+0 to 25+6 weeks’ gestation to        15/81 (19%) in 1994-9).
                         mothers living in the region. Terminations were excluded but all      Conclusions—Survival of infants born at 24 and 25 weeks of
                         other births of babies alive at the onset of labour or the delivery   gestation has significantly increased. Although over half the cohort
                         process were included.                                                of infants born at 23 weeks was admitted to neonatal intensive
                         Main outcome measures— Outcome for all infants was                    care, there was no improvement in survival at this gestation. Care
                         categorised as stillbirth, death without admission to neonatal        for infants born at 22 weeks remained unsuccessful.

                        Ailbhe Burke considers a study that looked at the viability of extremely premature babies
                        and had implications for abortion law

Ailbhe Burke academic         he UK Abortion Act 1967 made abortion legal up to                survival, between cohorts that differ based on exposure to a
foundation programme          28 weeks of completed gestation. It was amended by               variable, here whether they were born in 1994-9 or in 2000-5.
doctor, Mitochondrial
Research Group,
                              the Human Fertilisation and Embryology Act 1990 to               Prospective cohort studies are the most robust observational
Newcastle University    allow abortion only up to 24 weeks except in limited cir-              method and give the strongest evidence possible from an
a.burke@ncl.ac.uk       cumstances. Since then outcomes for premature babies have              observational study. This is a good choice of design for this
                        improved, and babies born at less than 24 weeks’ gestation             study. Interventional studies provide more robust evidence
                        may now survive.                                                       and can establish cause and effect relations but would be ethi-
                           In the light of this, a House of Commons review of the UK           cally and technically difficult to carry out in this area.
                        abortion legislation in late 2007 and early 2008 considered               The authors compared the outcomes of two cohorts
                        reducing the legal limit for abortions to less than 24 weeks.          from different time periods in a geographically defined
                        The review yielded no further amendments to the Abortion               population born before 26 weeks’ gestation. They used
                        Act but highlighted a lack of published, peer reviewed                 data from a pre-existing register, the Trent neonatal
                        evidence from the United Kingdom about whether the sur-
                        vival of infants born at 23 or 24 weeks has improved.
                           Current research in this area is limited. The main source
                        for the United Kingdom is the EPICure study (www.epicure.
                        ac.uk), also known as population based studies of survival and
                        later health status in extremely premature infants, which has
                        published data on survival and longer term outcomes of babies
                        born below 26 weeks’ gestation in the UK and Ireland from a
                        10 month period in 1995. Much clinical practice in the UK is
                        based on findings from this survey, but up to date information
                        is needed. The full results from the follow-up study in 2006,
                        EPICure 2, are yet to be published and EPICure cite data
                        from the Trent neonatal survey as being the most up to date
                        figures on survival for the UK. There is some international
                        research in this area, but its usefulness is limited by questions of
                        applicability of findings to the UK healthcare system.

                        What did the authors do?
                        The authors used a prospective cohort study, in which the
                        groups (cohorts) are defined before data are collected. This
                        type of study is used to compare outcomes, in this case

        student.bmj.com                                                                                                                                        331

                                                                               % survival to discharge
paper+   survey, which continuously collects data on pregnancy, birth,
         and outcomes. The authors included data about babies from                                        60
         Trent who were cared for in other regions. Information on                                                       2000-5
         miscarriages and stillbirths and deaths without admission were                                   50
         obtained from the Confidential Enquiry into Stillbirths and
         Deaths in Infancy (now the Confidential Enquiry into Maternal
         and Child Health).                                                                               30
            The authors chose the 12 years 1994-2005 for analysis and
         divided it into two six year periods, yielding a 1994-9 cohort
         and a 2000-5 cohort. Infants were included if they were alive                                    10
         at the onset of labour; were born before 26 weeks’ gestation                                                               P=0.99                    P<0.001         P=0.016
         between 1 January 1994 and 31 December 2005; and were                                              22                         23                        24                 25
         born to a mother whose home address was in the Trent region
         of the UK. The authors obtained ethical approval from two                                                                                Gestational age (completed weeks)
         bodies.                                                               Percentage survival to discharge by gestational age for infants admitted
            The authors assessed statistical significance of differences        to neonatal intensive care, 1994-9 compared with 2000-5
         in outcomes between the two cohorts using the 2 test. This
         compares outcomes as categorical variables, which do not vary
         by amounts or degree but are mutually exclusive—for exam-             that survival of babies in Trent born after 24 weeks’ gestation
         ple, sex—between two or more independent groups. It assesses          improved, and babies born before 24 weeks did not, despite
         whether the distribution of outcomes is different in different        similar treatment. Whether this is true for the UK as a whole
         groups. In this study outcomes on a number of categorical             and, therefore, whether it can be used to inform policy and evi-
         variables were compared between the 1994-9 and 2000-5                 dence based medicine in this field depends on whether Trent is
         cohorts, and here the main variable considered was survival.          representative of the UK. If this is true for all variables, includ-
                                                                               ing demographics and healthcare provision, then these findings
         What did they find?                                                   are generalisable throughout the UK.
         The authors found a difference between babies born either side of        Secondly, the survival figures from this study broadly mirror
         24 weeks’ gestation (figure). Babies born before 24 weeks showed       those of from the EPICure 1995 study. Guidelines and clinical
         no improvement in survival, and babies born at 24 and 25 weeks        practice are still based on the 1995 study, including the 2006
         in the 2000-5 cohort were more likely to survive compared with        Nuffield Council for Bioethics guidelines, which propose not
         babies born at the same gestational age in the 1994-9 cohort. This    routinely resuscitating babies born before 23 weeks’ gestation. 1
         is in the context of an increase in the number of days in hospital    This study confirms that the 1995 data are still representative of
         or intensive care, representing an increase in bed occupancy and      clinical outcomes.
         actual cost across all groups.                                           In summary, further research is needed from other regions
             So how robust, or true, are these findings? The study reports      in the UK, and this may be provided by EPICure 2. In the
         many strengths, including use of experienced research nurses          meantime, this study is a valuable contribution to the literature
         and rigorous data checking, and the follow-up of babies who           on outcomes for premature babies in the UK and is the most up
         left the region is a strength of this study. The authors’ assertion   to date resource in this area.
         that this population is representative of the UK population
         as a whole would, if the health services are also representa-         What are the implications?
         tive, render their findings more generalisable to the UK as a          In clinical terms the current practice of supporting clinical
         whole.                                                                decision making with data from the EPICure study is broadly
             No study is perfect, and the robustness of the findings            affirmed by this study because the outcomes in both cohorts
         could be improved. Gestational age was established by                 are similar. Given the different improvements in survival
         several methods in the following hierarchy—mother certain             between the two groups it might be appropriate to have differ-
         of dates (most reliable); early dating scan; late dating scan;        ent clinical approaches to or policies about the care of babies
         and postnatal examination (least reliable). We are not told           born before and after 24 weeks’ gestation.
         how many were dated by each method in either cohort or the               From an ethical perspective there are two main implications.
         accuracy of each method. Also, the authors do not explain             The authors comment that use of resources was significantly
         how they established the relative reliabilities of each method.       more for the second cohort, so the NHS is expending more
         Inaccurate dating would make the results less interpretable.          resources on very premature babies than before. In babies born
         And we do not know if there are differences in demographics—          after 24 weeks’ gestation this may have contributed to greater
         for example, sex or parental characteristics—between any of           survival. It is important to weigh this survival gain against the
         the populations (earlier or later cohort; babies before or after      EPICure findings of poor long term outcomes in very pre-
         the 24 weeks’ gestational age) which could affect outcomes.           mature babies to establish how much benefit this increased
         Finally, it is not possible to determine from this study if the       resource use confers in the long term. In babies born before
         centre had any effect on survival.                                    24 weeks’ gestation this extra resource use has not improved
             As well as investigating potential differences in demograph-      outcomes. This study supports reconsidering whether such
         ics, further analyses could look at whether there was a trend of      increased resource use is appropriate in babies born before 24
         gradually increasing survival with time or whether there were         weeks’ gestation.
         specific times at which this improvement occurred. The former             Further, these findings may have ethical and political impli-
         would support gradually improving care for this population,           cations in terms of the abortion debate. It is important to
         and the latter would point to specific advances responsible for        recognise that findings from research may be used to inform
         this improvement.                                                     future ethical and legal debates around such matters, whatever
                                                                               the intention of the original research.
         What does the study mean?                                             Competing interests: None declared.
         Firstly, it may mean that survival has improved in babies in          Provenance and peer review: Commissioned, not externally reviewed.
         the UK born between 24 and 26 weeks’ gestation. It tells us           1                         Brazier M. How to treat premature infants. Scientist 2006;20:22-3.

         332                                                                                                                                  STUDENTBMJ | VOLUME 16 | SEPTEMBER 2008

To top