Childhood outcomes after prescription of antibiotics to pregnant women with preterm rupture of the membranes: 7-year follow-up of the ORACLE I trial

BACKGROUND
The ORACLE I trial compared the use of erythromycin and/or amoxicillin-clavulanate (co-amoxiclav) with that of placebo for women with preterm rupture of the membranes without overt signs of clinical infection, by use of a factorial randomised design. The aim of the present study--the ORACLE Children Study I--was to determine the long-term effects on children of these interventions.


METHODS
We assessed children at age 7 years born to the 4148 women who had completed the ORACLE I trial and who were eligible for follow-up with a structured parental questionnaire to assess the child's health status. Functional impairment was defined as the presence of any level of functional impairment (severe, moderate, or mild) derived from the mark III Multi-Attribute Health Status classification system. Educational outcomes were assessed with national curriculum test results for children resident in England.


FINDINGS
Outcome was determined for 3298 (75%) eligible children. There was no difference in the proportion of children with any functional impairment after prescription of erythromycin, with or without co-amoxiclav, compared with those born to mothers who received no erythromycin (594 [38.3%] of 1551 children vs 655 [40.4%] of 1620; odds ratio 0.91, 95% CI 0.79-1.05) or after prescription of co-amoxiclav, with or without erythromycin, compared with those born to mothers who received no co-amoxiclav (645 [40.6%] of 1587 vs 604 [38.1%] of 1584; 1.11, 0.96-1.28). Neither antibiotic had a significant effect on the overall level of behavioural difficulties experienced, on specific medical conditions, or on the proportions of children achieving each level in reading, writing, or mathematics at key stage one.


INTERPRETATION
The prescription of antibiotics for women with preterm rupture of the membranes seems to have little effect on the health of children at 7 years of age.


FUNDING
UK Medical Research Council.


Introduction
The sequelae of preterm birth pose major public-health challenges. Children born preterm are at risk of major disabilities-eg, cerebral palsy-with risk increasing as length of gestation at birth decreases; 1 many preterm children without disability develop important behavioural and educational diffi culties. 2,3 The prevention of preterm birth and reduction of associated disability are therefore important health priorities.
Intrauterine infection and infl ammation have been associated with an increased risk of preterm birth 4 and independently with cerebral palsy and chronic lung disease. 5 The antecedent of preterm birth which has been most strongly associated with infection and infl ammation is preterm rupture of the fetal membranes (PROM). With conventional microbiological techniques, microbial invasion of the amniotic cavity has been identifi ed in about 30% of cases of PROM; 6 more sensitive PCR techniques that detect conserved 16s rRNA genes suggest that about 70% of births are found to be colonised after PROM. 7 Against this background, there is unbiased evi dence from randomised controlled trials for anti biotic treatment of women with PROM. The ORACLE I trial 8 assessed the use of amoxicillin-clavulanate (coamoxiclav) 375 mg, or erythromycin 250 mg, or both, or placebo, four times daily for 10 days or until birth (whichever was sooner) in women presenting with PROM using a factorial randomised controlled trial design. In singleton babies born to women after PROM, erythromycin decreased the risk of the composite primary outcome (death or major cerebral abnormality or chronic neonatal lung disease). The prescription of erythromycin was also associated with prolongation of pregnancy and reductions in neonatal morbidity compared with women who did not receive erythromycin, and is now recommended practice. 9 Coamoxiclav was not associated with any change in the primary endpoint. However, although women who received co-amoxiclav had prolonged pregnancy compared with those who did not receive any coamoxiclav, it was also associated with a signifi cantly higher incidence of neonatal necrotising enterocolitis.
In this paper we report the results of the ORACLE Children Study I (OCS I), which provided a unique opportunity to determine whether the perinatal use of erythromycin and/or co-amoxiclav for women with PROM in ORACLE I resulted in diff erences in functional and educational ability in childhood.

Participants
OCS I began in 2002 and sought follow-up information for children at 7 years of age who were born to the 4809 women with PROM who completed ORACLE I. 8 The study protocol has been described in detail elsewhere. 10 Briefl y, in the original trial, women were informed of the intention to do a subsequent follow-up assessment when they gave written informed consent. Children were traced with the help of the UK Offi ce of National Statistics (ONS) and by contact with their family doctors (general practitioners; GPs). Details of deaths and families who moved out of the UK were notifi ed to us by ONS. We did not attempt to contact the families of children who had been adopted, were in foster care, or had emigrated.
Contact details of surviving children were obtained from the National Health Service (NHS) National Strategic Tracing Service (NSTS). If no response was obtained to an invitation letter, the child's GP was contacted to check details and possible reasons for non-response (eg, the child was currently in care or was a non-English speaker). Translations of all study materials were available. For those children who were not 7 years of age at the initial invitation, contact was maintained from 2001 onwards with birthday cards, newsletters, and change of address cards, and via a website.
When the child was 7 years old we confi rmed their current address with ONS and the GP. An information leafl et was sent to the parents, and 2 weeks later the study questionnaire was sent. If no response was obtained, contact details were checked with the GP and a reminder letter, including a £5 high street voucher, 11 was sent by registered post to the parents or carers of the child during the week of the child's 7th birthday. If no response was forthcoming, 3 weeks later, a fi nal letter was sent or telephone contact was made.
The West Midlands Multi-centre Research and Ethics Committee (MREC) approved the study and the University of Leicester, UK, sponsored the OCS. Oversight was provided by an independent trial steering committee and data monitoring committee, both of which met annually. Those involved in tracing and data entry remained blind to the allocated treatment. All data to assess health and educational outcomes were double entered and subject to validation and logic checks.

Data collection
Data were collected with a parent-completion postal questionnaire. This comprised the Health Utilities Index (HUI) 12 from which the Multi-Attribute Health Status (MAHS) is derived, the Strengths and Diffi culties Questionnaire, 13 and specifi c questions on respiratory symptoms, 14 hospital admissions, convulsions, other specifi c medical conditions, and demographic data.
The primary outcome was defi ned as the presence of any level of functional impairment (severe, moderate, or mild) derived from the mark III MAHS classifi cation system 12 within any of the individual attributes of vision, hearing, speech, ambulation, dexterity, emotion, cognition, and pain. Each attribute has either fi ve or six defi ned levels of impairment, ranging from normal function to severe dysfunction. 15 These have been classifi ed further into none, mild, moderate, or severe levels of severity for the individual attributes from the standard algorithms available within the HUI coding/procedure manual. The overall level of functional impairment was determined by their worst score in any attribute. Sensitivity analyses were also done based on the HUI3 multi-attribute utility scores of overall health-related quality of life, 16 which became available after the study had begun.
Secondary outcomes were the presence of three or more abnormal attributes derived from the MAHS classifi cation system and the degree of functional impairment (severe, moderate, mild, none) within the individual domains; the number of deaths between trial entry and discharge and age 7 years; overall and subscale  scores derived from the parent-completed Strengths and Diffi culties Questionnaire; the frequency of specifi c medical conditions including CNS problems (cerebral palsy, fi ts/seizures, hydrocephalus with a shunt), respiratory problems (wheezing, medication for asthma), hospital admission (both in the last year and for chest problems), diabetes, bowel disorders, and developmental problems (attention defi cit hyperactivity disorder derived from the Strengths and Diffi culties Questionnaire 17 or parent report), and other development problems.
With support from the UK Qualifi cations and Curriculum Authority (QCA), we used results from national curriculum tests (key stage one), done at 7 years of age to assess the children's educational attainment for residents in England. Such tests are not routinely done in other parts of the UK. Changes to the way in which the results of these tests are derived have occurred since 2005, when the assessment became teacher-based rather than test-based, with the results of written tests used by the teacher to inform the level awarded. 18 At key stage one, all children are awarded a level for each of reading, writing, and mathematics. These include levels three and four, which are above average, level two (the average level awarded to 60-70% of pupils, and which is also subdivided into three sublevels), and below level two, which includes children who attained level one, those who were working towards level one, or who were not entered by the teacher. For all eligible OCS I children in England, key stage one level data were provided anonymously by the UK Department for Children, Schools and Families (DCSF), categorised by treatment group.

Statistical analysis
The size of the study was predefi ned by the number of women recruited to the ORACLE I trial. The indicative power calculation in the protocol noted that about 4400 children were expected to be eligible for the followup study. About 3·1% of the children in the any erythromycin group were expected to have three or more abnormal attributes using the MAHS scale. 12 Assuming an 85% response rate, this gave 80% power (with two-tailed α=0·05) to detect a prevalence of three or more abnormal attributes in the no erythromycin group of 5% (a relative diff erence between the two groups of 38%).

Erythromycin and co-amoxiclav
Child's age at the start of the academic year they sat key stage one tests (years) Data are n (%) or median (IQR). Odds ratios with 95% CI are presented for primary and secondary outcomes in the groups receiving co-amoxiclav (any co-amoxiclav: either with or without erythromycin) and erythromycin (any erythromycin: either with or without co-amoxiclav) separately. Odds ratios approximate relative risks when the risk is low. 19 Logistic models with terms indicating allocation to co-amoxiclav and erythromycin and an interaction term, corresponding to the ORACLE I trial's factorial design, were also fi tted. 20 Since our fi ndings were generally not altered when interaction terms were included, for simplicity they have not been included in models presented in the main tables, but they are available on the internet. Multiple births were treated as being independent in the analyses, but sensitivity analyses based on randomly selecting one child were also done. 21 Data from the Strengths and Diffi culties Questionnaire were classifi ed as normal, borderline, or abnormal. Odds ratios and CI for proportions with borderline or abnormal scores are presented. Models with interaction terms were also fi tted as above.
Subgroup analyses were done as specifi ed in the protocol, relating to multiple and singleton pregnancies, and gestational age subgroups used at the time of randomisation (<32 weeks' gestation and <28 weeks' gestation).
Relative risks and 95% CI for key stage one educational data were obtained from Poisson regression models, adjusting for test year. Because of restricted information available from the anonymous summary of these data supplied by the DCSF, sensitivity and subgroup analyses were not possible for these data.
Most of the outcomes presented, including the primary outcome of the follow-up study, can only be assessed in surviving children. Thus the analyses presented are not based on the intention-to-treat principle (ie, by analysis of outcomes in all those entered into the trial). However, the absolute risk of death was low, limiting any potential bias that might be introduced by undertaking the analyses of surviving children only, as pre-specifi ed in the study protocol. There were no clear diff erences in the numbers of deaths in each of the study groups at the end of ORACLE I, but we present sensitivity analyses using a composite death or any functional impairment outcome to confi rm the limited eff ect of including deaths in the analyses.
Informal allowance for the large number of comparisons undertaken is made in interpreting the results throughout. 22 Outcomes for children not assessed may diff er from outcomes for those who were assessed. 23 To explore this, we investigated diff erences between responders and nonresponders for a variety of factors: treatment group, shortterm maternal and neonatal outcomes, ethnicity, indices of deprivation from ONS derived from post code, and for those children in England, educational attainment.

Role of the funding source
The study sponsors had no involvement in the study design; collection, analysis, and interpretation of the data; in the writing of the report; and in the decision to submit the paper for publication. DJ and KP had full access to all the data. SK had fi nal responsibility for the decision to submit for publication.

Results
Of the 4378 UK children eligible for follow-up, outcome was known for 3298 (75%). Full questionnaire data are available for 3171 (72%) children, data were collected from telephone interview or GP for 90 children (2%; included in the analysis of specifi c medical conditions); 37 (1%) children had died (fi gure 1). Women were unaware of their treatment allocation with the exception of one woman who requested this information before returning data for OCS I. Only 0·8% of data within returned questionnaires were missing. Full key stage one educational data were available for 3238 (96%) of the 3377 English children who had been entered into the tests by 2007. The characteristics of the responders, in each factorial group, were broadly similar in terms of maternal, demographic, and neonatal data to that of the main ORACLE I trial results (table 1). Children whose outcome was unknown had younger mothers and experienced less neonatal morbidity (webtable 1), and, at least for those in England, were from more deprived areas than those whose outcome was known (webtable 2). More English children whose outcome was unknown scored below level two in key stage one tests, and were less likely to be white than were those whose outcome was known (webtable 2).
There was no diff erence in the proportion of children with any functional impairment whose mothers had been prescribed any erythromycin, compared with those who received no erythromycin, or in the proportion of children with any functional impairment whose mothers had received co-amoxiclav, compared with those who received no co-amoxiclav (table 2). There was no evidence of an eff ect of either antibiotic on any functional impairment when fi tting a logistic model with terms for erythromycin, co-amoxiclav, and an interaction between the antibiotics (webtable 3); this model also showed no evidence for an interaction in eff ects between the antibiotics (webtable 3). Neither antibiotic had an eff ect on any functional impairment when missing data were assumed to correspond to a functional impairment (data not shown). Table 3 presents the results for the four treatment groups "inside the table" of the factorial trial; 24 there were no clear diff erences between the treatment groups. Sensitivity analyses based on the HUI3 multi-attribute utility function 16 showed similar results (webtable 4). The use of erythromycin or of co-amoxiclav was not associated with any change in risk of death or of a composite endpoint of death or any functional impairment (table 4; webtable 5).
Neither antibiotic had a signifi cant eff ect on the overall level of behavioural diffi culties experienced or the impact on families (table 5; webtable 6). Neither antibiotic had any eff ect on the proportions of children failing to achieve each level in reading, writing, or mathematics at key stage one (table 6). Compared with national data, a higher proportion of the children surveyed here attained a score below level two (table 6).
There were no signifi cant diff erences with either antibiotic in the proportions of children reported as having   CNS or development problems, diabetes, or total bowel disorders (table 7; webtable 7). There is some evidence of a reduction in respiratory problems in the children whose mothers received any erythromycin compared with those who received no erythromycin (table 7). Although there is no evidence of a signifi cant increase in the prevalence of total bowel problems with either antibiotic, more children were reported to have other bowel problems (eg, hospital admission for constipation, diarrhoea, or stomach troubles, or under care of doctor for bowel problems) in the any co-amoxiclav group than in the no co-amoxiclav group (table 7). However, both of these results must be interpreted with caution because of the lack of formal adjustment for multiple comparisons. Subgroup analyses for multiple/single births and for gestation above and below 28 and 32 weeks were preplanned; they generally suggest eff ects consistent with neither antibiotic having an eff ect on any functional impairment or on the presence of three or more abnormal attributes (fi gure 2). Adjustment for maternal baseline, social class, and other factors did not substantially alter any of the treatment eff ects seen here (data not shown). For further analyses, see http://www2. le.ac.uk/Members/drj/supplementary-materials-forpapers. Data are number of children failing to achieve level two or higher (%). Overall relative risks (RR) and 95% CI are from Poisson models for level achieved adjusting for test year, 2002-07. National norms for 2002-07 have been standardised by the numbers of children in the OCS I cohort each year, and suggest that we would expect the following percentages of children to fail to achieve level two or above: reading 15%, writing 18%, maths 11%. *No evidence of overdispersion when these Poisson models are fi tted. Hydrocephalus with shunt 3 (0·2%) 6 (0·4%) 0·52 (0·13-2·10) 6 (0·4%) 3 (0·2%) 2·00 (0·50-8·01)

Discussion
In this long-term follow-up of children whose mothers were entered into the ORACLE I trial of antibiotics prescribed for women with PROM and no overt signs of infection, there is no evidence of any substantial longterm eff ect in children at age 7 years associated with the prescription of either erythromycin or co-amoxiclav. Although there were some apparent diff erences in a small number of secondary outcomes-eg, improvement in respiratory function with erythromycin and increased bowel disorders with co-amoxiclav-and despite the fact that these changes are biologically plausible and consistent with the short-term outcomes from the ORACLE I trial, they should be viewed with caution because of the large number of comparisons made. The size of the ORACLE population made direct face-toface assessment of outcome impracticable and thus we elected to obtain proxy information about the children from parents. This was possible using well validated and standardised parent report tools, namely the Health Utilities Index and the Strengths and Diffi culties Questionnaire, to assess health-related quality-of-life and behaviour, respectively. Although these are validated ways of collecting outcome data, using a questionnaire may mean more subtle diff erences between the treatment groups were missed. Health conditions could only be assessed through parental report, which also leads to the potential for under reporting and some inaccuracy in ascertainment; in general, this might reduce the power to detect any associations present, since almost all the women in the study remained blind to their allocated treatment.
To achieve high follow-up rates using postal questionnaires is diffi cult. We used evidence-based strategies to optimise the response 25 and developed our own randomised evidence during the conduct of the study to maximise response. 11 We maintained regular contact with the women originally entered into the trial, allowing us to achieve a high follow-up rate of 75%. Nonetheless, disadvantaged groups are over-represented in the non-responders but there was no evidence of diff erential non-response between the randomised groups. Although this response rate is lower than that assumed in the initial power calculation, the increased prevalence of functional impairment resulted in power being increased overall.
Cognitive impairment is the most common disability associated with preterm birth. 26,27 School performance is aff ected by a range of disabilities, of which general cognitive function is the most important, although additional defi cits of executive function 28 and attention may be contributory in the preterm child. 2 Although we were unable to achieve direct cognitive assessment, we were able to use anonymised results from national curriculum tests, which allowed us to examine the scholastic attainment for virtually the whole population of surviving children in England, which we have used as a proxy for cognitive outcome. While the results showed no treatment diff erences in the level of performance reached with either antibiotic, this group of children as a whole have educational attainment below national norms, in keeping with the excess of preterm children with poor academic attainment. 2 Level scores are arguably rather weak measures of attainment; the power of these comparisons would have been increased by use of the raw scores for individual tests, but these are not available centrally.  In women with PROM, positive amniotic fl uid cultures are found in 32% of cases at the time of presentation 29 and in as many as 75% during subsequent labour. 30 Antibiotics might therefore be expected to improve clinical outcomes in this situation. There is also increasing evidence that, in addition to preterm birth, intrauterine infection is an independent antecedent of other disability, particularly cerebral palsy 31 and chronic lung disease. 32 Prescription of antibiotics for PROM could prevent neurological and respiratory disability by prolonging pregnancy or by reducing infl ammation. It is thus disappointing that the fi ndings of decreased neonatal morbidity after receipt of erythromycin in ORACLE I have not translated into long-term benefi t detectable within the measures we used. The reasons for this are not clear but might be linked to the length of antibiotic exposure, which in this group of women was fairly short, since about 60% gave birth within a week. 8 There is also some evidence that antibiotic administration to women with PROM may neither eradicate nor prevent intra-amniotic infection. In a recent study of 18 women with intra-amniotic infection following PROM, only three showed no evidence of infection or infl ammation after combination antibiotic treatment for 7-14 days and nine of 28 women with no evidence of intra-amniotic infl ammation at admission developed infl ammatory changes despite therapy; fi ve of these nine women developed positive amniotic fl uid cultures. 33 This evidence suggests that other strategies are necessary to prevent the progress of infection and infl ammatory changes in utero in the presence of ruptured membranes.
In the accompanying paper reporting the results of the ORACLE Children Study II 34 in the presence of spontaneous preterm labour with intact membranes, we have observed an increase in the risk of any functional impairment with the use of erythromycin (with or without co-amoxiclav) and, particularly, increases in the numbers of children with cerebral palsy associated with the use of both erythromycin and co-amoxiclav. These results would suggest further caution should be used when considering the routine treatment of women with antibiotics where there is uncertainty about the diagnosis of PROM. It is clear that we need to better understand the role of infection in women with PROM and the role and wider eff ects of antibiotics in the perinatal period. However, it is important to clarify that the women in ORACLE I with PROM were entered into the trial when there was clinical uncertainty about the need to prescribe antibiotics-ie, they showed no overt clinical signs of maternal or fetal infection. It is critical that women with evidence of clinical infection are treated with antibiotics, since clinical chorioamnionitis remains an important cause of maternal, fetal, and neonatal death.
The results of this study should not lead to fewer women with overt signs of maternal or fetal infection receiving treatment.