School academic performance of children hospitalised with a chronic condition.

OBJECTIVE: To examine academic outcomes among children hospitalised with a chronic health condition. DESIGN: Population-level birth cohort. SETTING: New South Wales, Australia. PARTICIPANTS: 397 169 children born 2000-2006 followed up to 2014. INTERVENTION/EXPOSURE: Hospitalisations with a chronic condition. MAIN OUTCOME MEASURES: Academic underperformance was identified as 'below the national minimum standard' (BNMS) in five literacy/numeracy domains using the national assessment (National Assessment Program-Literacy and Numeracy) data. Multivariable logistic regression assessed the adjusted ORs (aORs) of children performing BNMS in each domain at each grade (grades 3, 5 and 7, respectively). RESULTS: Of children hospitalised with a chronic condition prior to National Assessment Program-Literacy and Numeracy (NAPLAN) (16%-18%), 9%-12% missed ≥1 test, with a maximum of 37% of those hospitalised ≥7 times, compared with 4%-5% of children not hospitalised. Excluding children who missed a NAPLAN test, more children hospitalised with a chronic condition performed BNMS across all domains and grades, compared with children not hospitalised (eg, for BNMS in reading at grade 3: n=2588, aOR 1.35 (95% CI 1.28 to 1.42); for BNMS in numeracy at grade 3: n=2619, aOR 1.51 (95% CI 1.43 to 1.59)). Increasing frequency and bed-days of hospitalisation were associated with 2-3 fold increased odds of performing BNMS across all domains and grades. Children hospitalised with mental health/behavioural conditions had the highest odds of performing BNMS across all domains at each grade. CONCLUSIONS: Children hospitalised with a chronic condition underperform academically across literacy/numeracy domains at each school grade. Health and educational supports are needed to improve these children's academic outcomes.

The accuracy of hospital discharge data in recording major congenital anomalies in Australia.

BACKGROUND: There has been increasing use of hospital discharge data to identify congenital anomalies, with limited information about the accuracy of these data. OBJECTIVES: To evaluate the accuracy of hospital discharge data in ascertaining major congenital anomalies in infants. METHODS: All liveborn infants with major congenital anomalies born between 2004 and 2009 in New South Wales, Australia were included. They were separated into two study groups: (a) infants identified from the Register of Congenital Conditions with a corresponding record in linked hospital discharge data; and (b) infants with a recorded congenital anomaly in hospital data, but without a register record. For the first group, we assessed agreement (concordant diagnoses) and the proportion of anomalies with discrepant diagnoses in each dataset. For the second group, we determined the number of anomalies recorded only in hospital data and applied specific conditions restricting to those recorded in the birth admission, excluding nonspecific diagnoses, or those with relevant surgical procedures to minimize potential false positives or over-reporting. RESULTS: The first study group included 9,346 infants with an average 84% agreement in the ascertainment of major anomalies between hospital and registry data, and >93% agreement for cardiac, abdominal wall, and gastrointestinal anomalies. Discrepant diagnoses occurred on average in 20% of cases from hospital data and 17% from registry data, and were slightly reduced with the use of diagnoses recorded only in tertiary pediatric hospitals. The second group included 25,893 infants where anomalies were only recorded in hospital data, most commonly skin and unspecified anomalies. Excluding unspecified cases, those only diagnosed at the birth admission and restricting to surgical procedures reduced over-reporting by up to 96%. CONCLUSIONS: Hospital discharge data provide an acceptable means to ascertain congenital anomalies, but with variable accuracy for different anomalies. Application of specific conditions and limited to surgical procedures improves the utility of using hospital discharge data to ascertain congenital anomalies.

Survival of infants born with esophageal atresia among 24 international birth defects surveillance programs.

BACKGROUND: Esophageal atresia (EA) affects around 2.3-2.6 per 10,000 births world-wide. Infants born with this condition require surgical correction soon after birth. Most survival studies of infants with EA are locally or regionally based. We aimed to describe survival across multiple world regions. METHODS: We included infants diagnosed with EA between 1980 and 2015 from 24 birth defects surveillance programs that are members of the International Clearinghouse for Birth Defects Surveillance and Research. We calculated survival as the proportion of liveborn infants alive at 1 month, 1- and 5-years, among all infants with EA, those with isolated EA, those with EA and additional anomalies or EA and a chromosomal anomaly or genetic syndrome. We also investigated trends in survival over the decades, 1980s-2010s. RESULTS: We included 6,466 liveborn infants with EA. Survival was 89.4% (95% CI 88.1-90.5) at 1-month, 84.5% (95% CI 83.0-85.9) at 1-year and 82.7% (95% CI 81.2-84.2) at 5-years. One-month survival for infants with isolated EA (97.1%) was higher than for infants with additional anomalies (89.7%) or infants with chromosomal or genetic syndrome diagnoses (57.3%) with little change at 1- and 5-years. Survival at 1 month improved from the 1980s to the 2010s, by 6.5% for infants with isolated EA and by 21.5% for infants with EA and additional anomalies. CONCLUSIONS: Almost all infants with isolated EA survived to 5 years. Mortality was higher for infants with EA and an additional anomaly, including chromosomal or genetic syndromes. Survival improved from the 1980s, particularly for those with additional anomalies.

Five-year survival of infants with major congenital anomalies: a registry based study.

AIM: To determine survival of infants with major congenital anomalies (CA) and assess the effect of co-existing anomalies and gestational age. METHODS: All liveborn infants with major CA born in New South Wales (NSW), Australia, 2004-2009 were identified from the NSW Register of Congenital Conditions. Deaths were identified via record linkage to death registrations and five-year survival was estimated using Kaplan-Meier methods. RESULTS: There were 8521 liveborn infants with CA of whom 617 (7.2%) died within the first five years of life. Half of deaths occurred in the first week of life. The overall five-year survival rate was 92.8% (95%CI: 92.2-93.3) and 83.2% (95%CI: 79.0-87.4) for syndromes, 83.4% (95%CI: 80.9-85.9) for multiple, 85.1% (95%CI: 82.6-87.5) for chromosomal, 95.3% (95%CI: 94.8-95.8) for isolated and 96.2% (95%CI: 94.3-98.1) for non-Q chapter anomalies. Five-year survival for chromosomal, syndromes and sub-groups was higher for isolated compared with multiple anomalies ranging from 77.5% to 98.9% and 68.6% to 89.5%, respectively. Survival was lower for preterm (79.4%; 95%CI: 77.5-81.4) than for term infants (95.8%; 95%CI: 95.3-96.3). CONCLUSION: Nine in ten infants with major CA survive up to five years, although there is variability in survival across CA groups. Survival of infants with major congenital anomalies has improved in recent years.

The burden of hospitalized sports-related injuries in children: an Australian population-based study, 2005-2013.

BACKGROUND: There is concern about recent increase and severity of sports-related injuries in children. Despite the benefits of sports participation, injuries may carry long-term health consequences. We aimed to evaluate the prevalence, characteristics and types of hospitalized sports-related injuries in children. METHODS: Population-based study of all acute sports-related injuries requiring hospitalization in children 5 to 15 years of age in New South Wales (NSW), Australia, 2005-2013. Health information was obtained from the NSW Admitted Patient Data Collection, a census of all hospital admissions from public and private hospitals. Children with a recorded ICD10-AM injury code (S00-T79) and sport-related activity code (U50-U70) were included. Prevalence and trend in injuries by age group, sporting code, body region affected and type of injury were assessed. RESULTS: There was a total of 20,034 hospitalizations for sports-related injuries (2.7% of all hospitalizations in children aged 5-15 years), involving 21,346 recorded injuries in 19,576 children. The overall population hospitalization period prevalence was 227 per 100,000 children aged 5-15 years in 2005-2013, remaining stable over time (RR 0.99; 95% CI 0.98-1.00). Football codes such as rugby league/union and soccer combined represented nearly two thirds of the total (60%). The most common body regions affected were the forearm (31%) head (15%) and hand injuries (13%). Fractures accounted for 65% of injuries followed by dislocations (10%) and traumatic brain injury (10%). Compared to other age groups, children aged 5-8 years had double the proportion of shoulder (15% vs. 7%) while 13-15 year olds had higher proportion of lower-leg (14% vs. 8%) and knee (6% vs.2%) injuries. One in seven injuries sustained while playing rugby league/union, baseball and hockey were traumatic brain injuries. A total of 444 (2.2%) of children had more than one hospitalization for sports-related injuries. CONCLUSION: On average, six children were hospitalized every day for sports-related injuries in the last decade with trends remaining stable. The most common sports involved were football codes, one in three injuries involved the forearm and two thirds were fractures. These findings can be used to inform health policy and sporting governing bodies to target preventive interventions and promote safe sports participation in children.

Gestation at birth, mode of birth, infant feeding and childhood hospitalization with infection.

INTRODUCTION: Infections are a leading cause of mortality and morbidity in preschool children. We aimed to assess the impact of the co-occurrence of cesarean section, early birth and formula feeding on hospitalization with infection in early childhood. MATERIAL AND METHODS: Population-based retrospective record-linkage cohort study of 488 603 singleton livebirths ≥32 weeks' gestational age in New South Wales, Australia, 2007-2012. Multivariable Cox-regression was used to estimate independent and combined adjusted associations of gestational age, mode of birth (vaginal or cesarean section by labor onset) and formula feeding with time to first and repeat hospitalization with infection for children less than five years of age. RESULTS: In all, 95 346 (19.5%) children were hospitalized with infection, and of these 24.8% (23 615) more than once. Median age at first and repeat hospitalization was 1.1 and 1.7 years, respectively. Earlier gestation, modes of birth other than spontaneous vaginal, and formula feeding were independently associated with an increased risk of first and repeat hospitalization with infection. At 32-36 weeks' gestation, co-occurrence of perinatal factors (cf. spontaneous vaginal birth at 39+ weeks without formula feeding) was associated with a 2-fold and 1.5-fold increased risk of first and repeat hospitalization, respectively. For births at 37-38 weeks, the increased risk was 1.5-fold and 1.25-fold for first and repeat hospitalization, respectively. CONCLUSIONS: Cesarean section, labor induction, birth at <39 weeks and formula feeding increase the risk of infection-related hospitalization in childhood, which increases further when these factors co-occur. Reducing early planned birth and supporting breastfeeding are potentially cost-effective approaches to reducing the risk of hospitalization.

Acid suppressants for managing gastro-oesophageal reflux and gastro-oesophageal reflux disease in infants: a national survey.

OBJECTIVES: To evaluate the diagnosis and management of reflux and gastro-oesophageal reflux disease (GORD) in infants aged <1 year presenting to general practitioners (GPs). DESIGN, SETTING AND PARTICIPANTS: A nationally representative, prospective, cross-sectional survey of GP activity in Australia, 2006-2016 (Bettering the Evaluation And Care of Health Study). Annually, a random sample of around 1000 GPs recorded details for 100 consecutive visits with consenting, unidentified patients. OUTCOME MEASURES: Diagnoses of reflux and GORD and their management including prescribing of acid-suppressant medicines (proton pump inhibitors (PPIs) and histamine receptor antagonists (H2RAs)) and counselling, advice or education. RESULTS: Of all infants' visits, 512 (2.7%) included a diagnosis of reflux (n=413, 2.2%) or GORD (n=99, 0.5%). From 2006 to 2016, diagnostic rates decreased for reflux and increased for GORD. Prescribing of acid suppressants occurred in 43.6% visits for reflux and 48.5% visits for GORD, similar to rates of counselling, advice or education (reflux: 38.5%, GORD: 43.4% of visits). Prescribing of PPIs increased (statistically significant only for visits for reflux), while prescribing of H2RAs decreased. CONCLUSIONS: Overprescribing of acid suppressants to infants may be occurring. In infants, acid-suppressant medicines are no better than placebo and may have significant negative side effects; however, guidelines are inconsistent. Clear, concise and consistent guidance is needed. GPs and parents need to understand what is normal and limitations of medical therapy. We need a greater understanding of the influences on GP prescribing practices, of parents' knowledge and attitudes and of the pressures on parents of infants with these conditions.

Accidental pharmacological poisonings in young children: population-based study in three settings.

INTRODUCTION: Pharmacological poisonings in young children are avoidable. Previous studies report calls to poisons centres, presentations to emergency departments (ED) or hospital admissions. There are limited data assessing concurrent management of poisonings across all three settings. We aimed to describe accidental pharmacological poisonings in young children across our Poisons Information Centre (PIC), EDs and hospitals. METHODS: A population-based study in New South Wales, Australia, of PIC calls, ED presentations and hospital admissions for accidental pharmacological poisoning in children aged <5 years, 2007-2013. We examined trends, medicines responsible and subsequent management. Medicines were coded using ICD10-AM diagnosis codes (T36-50). RESULTS: Over 2007-2013, pharmacological poisonings accounted for 67,816 PIC calls, 7739 ED presentations and 2082 admissions. Rates (per 10,000 children) of PIC calls declined from 220 to 178; ED presentations were stable (∼22-24), with a decrease in emergency cases offset by an increase in semi- or non-urgent presentations; hospital admissions declined (8-5). Most PIC calls related to "non-opioid analgesics" (25%), and "topical agents" (18%). Nearly every day, one child aged <5 years was admitted to hospital for poisoning. "Benzodiazepines", "other and unspecified antidepressants", "uncategorised antihypertensives", and "4-aminophenol derivatives" accounted for over one-third of all admissions. Most PIC calls (90%) were advised to stay home, 6% referred to hospital. One-quarter of ED presentations resulted in admission. CONCLUSIONS: Poisonings reported to PIC and hospitals declined, however, non-urgent ED presentations increased. Strategies to reduce therapeutic errors and access to medicines, and education campaigns to improve Poisons Centre call rates to prevent unnecessary ED presentations are needed.

Long-term survival for infants born with orofacial clefts in Western Australia.

BACKGROUND: Only two population-based studies have reported survival beyond 15 years for individuals with orofacial clefts (OFC), and only for individuals with isolated OFC. Compared with the general population, long-term survival was similar for individuals with cleft lip only, and lower for individuals with cleft palate only. Results for those born with isolated cleft lip and cleft palate were inconsistent. METHODS: Using linked population-based health data, including a congenital anomaly register with active surveillance and diagnoses up to 6 years, we compared survival at 1, 5, and 20 years for infants born 1980 to 2010 with, and without OFC. RESULTS: Of the 8112 live born infants in the cohort, 186 died before 20 years; most (81%) died during infancy. Compared with infants without OFC, infants born with all types of isolated OFC ± additional minor anomalies had similar infant survival (around 99%), but we found lower survival for infants with all cleft types and an additional major anomaly (66-84%). From 1 to 5 years, only infants with cleft palate only and an additional major anomaly had lower survival (97%) compared with children without OFC (99.9%). From 5 to 20 years, children with all cleft types, with or without additional major anomalies had similar survival to children without OFC (98-100%). CONCLUSION: Parents with a child diagnosed with an OFC ± additional minor anomalies only can be reassured that the OFC does not influence survival rates in infancy, or long-term. Infant survival was lower only for children with OFC and additional major anomalies.

Maternal alcohol consumption during pregnancy and the risk of orofacial clefts in infants: a systematic review and meta-analysis.

BACKGROUND: The teratogenic effects of maternal alcohol consumption during pregnancy include anomalies of craniofacial structures derived from the cranial neural crest cells. The presence of specific craniofacial anomalies contributes to the diagnosis of fetal alcohol spectrum disorders. Cleft lip and palate [orofacial clefts (OFCs)], also derived from the cranial neural crest cells, are common congenital anomalies, but their relationship with prenatal alcohol consumption is unknown. METHODS: To evaluate the association between maternal consumption of alcohol during pregnancy and the occurrence of OFCs in infants, we conducted a systematic review and meta-analyses of published studies. We examined the associations between any alcohol consumption, binge level drinking, and heavy and moderate levels of consumption vs. no or low levels of consumption. RESULTS: After screening 737 publications, we identified 33 studies (23 case-control and 10 cohort studies). There was considerable heterogeneity in individual study design, quality measures and study results. Findings from random effects meta-analyses suggest no relationship between prenatal alcohol consumption and the occurrence of OFCs {pooled odds ratios for any alcohol intake and binge level drinking respectively: cleft lip with or without cleft palate 1.00 [95% confidence interval (CI) 0.86, 1.16] from 18,349 participants in 13 studies, 1.04 [95% CI 0.87, 1.24] [8763 individuals, 4 studies]; cleft palate only 1.05 [95% CI 0.92, 1.21] [21,459 individuals, 17 studies], 0.94 [95% CI 0.74, 1.21] [7730 participants, 4 studies]}. CONCLUSIONS: While we found no association between alcohol consumption during pregnancy and OFCs in infants, the influence of study design, particularly in relation to alcohol exposure measurement and OFC ascertainment cannot be ignored.

Descriptive epidemiology of cleft lip and cleft palate in Western Australia.

BACKGROUND: The reported birth prevalence of orofacial clefts (OFCs) varies considerably. This study describes the epidemiology of OFCs in an Australian population. METHODS: We studied infants diagnosed with cleft lip, with or without cleft palate (CL±P), and cleft palate only (CPO) since 1980 and reported to the population based Western Australian Register of Developmental Anomalies. We calculated prevalence rates by sex, Aboriginal status, geographic location, and socio-economic status. Associations between clefts and folate availability, pregnancy characteristics, pregnancy outcomes, other congenital anomalies, and age at diagnosis were also investigated. RESULTS: From 1980 to 2009, 917 infants with CL±P (12.05 per 10,000) and from 1980 to 2004, 621 infants with CPO (10.12 per 10,000) were registered. Prevalence rates for CL±P and CPO were 1.9 and 1.3 times higher, respectively, for Aboriginal Australians. Additional anomalies were reported for 31% of infants with CL±P and for 61% with CPO; chromosomal anomalies and other specific diagnoses accounted for 46% and 66%, respectively, of those with CL±P and CPO with additional anomalies. Almost all (99.7%) children with CL±P were diagnosed before 1 year of age, but 12% of CPO diagnoses were made after 1 year of age; 94% of these diagnoses were of submucous clefts and bifid uvula. CONCLUSIONS: These data provide a picture of the prevalence of OFCs in WA since 1980, and provide a useful reference for OFC data in Australia and internationally. The quality and completeness of the WARDA data are high, reflected in high prevalence rates, and proportions of clefts occurring with other anomalies.

Antibiotics for ureaplasma in the vagina in pregnancy.

BACKGROUND: Preterm birth is a significant perinatal problem contributing to perinatal morbidity and mortality. Heavy vaginal ureaplasma colonisation is suspected of playing a role in preterm birth and preterm rupture of the membranes. Antibiotics are used to treat infections and have been used to treat pregnant women with preterm prelabour rupture of the membranes, resulting in some short-term improvements. However, the benefit of using antibiotics in early pregnancy to treat heavy vaginal colonisation is unclear. OBJECTIVES: To assess whether antibiotic treatment of pregnant women with heavy vaginal ureaplasma colonisation reduces the incidence of preterm birth and other adverse pregnancy outcomes. SEARCH STRATEGY: We searched the Cochrane Pregnancy and Childbirth Group's Trials Register (31 May 2011). SELECTION CRITERIA: Randomised controlled trials comparing any antibiotic regimen with placebo or no treatment in pregnant women with ureaplasma detected in the vagina. DATA COLLECTION AND ANALYSIS: Three review authors independently assessed eligibility and trial quality and extracted data. MAIN RESULTS: We included one trial, involving 1071 women. Of these, 644 women between 22 weeks and 32 weeks' gestation were randomly assigned to one of three groups of antibiotic treatment (n = 174 erythromycin estolate, n = 224 erythromycin stearate, and n = 246 clindamycin hydrochloride) or a placebo (n = 427). Preterm birth data was not reported in this trial. Incidence of low birthweight less than 2500 grams was only evaluated for erythromycin (combined, n = 398) compared to placebo (n = 427) and there was no statistically significant difference between the two groups (risk ratio (RR) 0.70, 95% confidence interval (CI) 0.46 to 1.07). There were no statistically significant differences in side effects sufficient to stop treatment between either group (RR 1.25, 95% CI 0.85 to 1.85). AUTHORS' CONCLUSIONS: There is insufficient evidence to assess whether pregnant women who have vaginal colonisation with ureaplasma should be treated with antibiotics to prevent preterm birth.Preterm birth is a significant perinatal problem. Upper genital tract infections, including ureaplasmas, are suspected of playing a role in preterm birth and preterm rupture of the membranes. Antibiotics are used to treat women with preterm prelabour rupture of the membranes; this may result in prolongation of pregnancy and lowers the risks of maternal and neonatal infection. However, antibiotics may be beneficial earlier in pregnancy to eradicate potentially causative agents.

Trends in adverse maternal outcomes during childbirth: a population-based study of severe maternal morbidity.

BACKGROUND: Maternal mortality is too rare in high income countries to be used as a marker of the quality of maternity care. Consequently severe maternal morbidity has been suggested as a better indicator. Using the maternal morbidity outcome indicator (MMOI) developed and validated for use in routinely collected population health data, we aimed to determine trends in severe adverse maternal outcomes during the birth admission and in particular to examine the contribution of postpartum haemorrhage (PPH). METHODS: We applied the MMOI to the linked birth-hospital discharge records for all women who gave birth in New South Wales, Australia from 1999 to 2004 and determined rates of severe adverse maternal outcomes. We used frequency distributions and contingency table analyses to examine the association between adverse outcomes and maternal, pregnancy and birth characteristics, among all women and among only those with PPH. Using logistic regression, we modelled the effects of these characteristics on adverse maternal outcomes. The impact of adverse outcomes on duration of hospital admission was also examined. RESULTS: Of 500,603 women with linked birth and hospital records, 6242 (12.5 per 1,000) suffered an adverse outcome, including 22 who died. The rate of adverse maternal outcomes increased from 11.5 in 1999 to 13.8 per 1000 deliveries in 2004, an annual increase of 3.8% (95%CI 2.3-5.3%). This increase occurred almost entirely among women with a PPH. Changes in pregnancy and birth factors during the study period did not account for increases in adverse outcomes either overall, or among the subgroup of women with PPH. Among women with severe adverse outcomes there was a 12% decrease in hospital days over the study period, whereas women with no severe adverse outcome occupied 23% fewer hospital days in 2004 than in 1999. CONCLUSION: Severe adverse maternal outcomes associated with childbirth have increased in Australia and the increase was entirely among women who experienced a PPH. Reducing or stabilising PPH rates would halt the increase in adverse maternal outcomes.

Monitoring the quality of maternity care: how well are labour and delivery events reported in population health data?

Administrative or population health data sets (PHDS), such as birth and hospital discharge data, are used increasingly to evaluate maternity care. Use of PHDS requires reliable identification of diagnoses and procedures. The aim of this study was to determine the accuracy and reliability of the reporting of diagnoses and procedures related to childbirth in both individual and linked, birth and ICD10-coded hospital discharge data. Data from a population-based validation study of 1200 women provided the 'gold standard' for labour and delivery events and were compared with the hospital discharge and birth databases. Reporting characteristics (sensitivity, specificity, positive and negative predictive values) were determined for: induction, augmentation and obstruction of labour, modes of delivery (including failed instrumental delivery), episiotomy, perineal tears and repairs, and manual removal of the placenta. Differences in reporting by mode of delivery were also examined. Of the 1184 records available for review, 25% had labour induced, 25% had labour augmented and, of those who laboured, 17% had obstructed labour reported. Fourteen per cent had an elective/planned caesarean section (CS) including 2% that went into labour prior to the planned date, and 11% had an emergency, unplanned CS including 2% who had no labour. With the exception of augmentation and obstruction of labour, failed instrumental delivery and manual removal, there were high levels of accuracy for reporting of diagnoses and procedures during labour and delivery. There were no significant differences in reporting by mode of delivery. The findings suggest that PHDS-reported induction of labour, mode of delivery, and 3rd and 4th degree tears and repairs can be reliably used to evaluate maternity care. Consistency in reporting in birth and hospital discharge data from different countries and over time suggests the findings are likely to be generalisable to high-income countries.

How accurate is the reporting of obstetric haemorrhage in hospital discharge data? A validation study.

BACKGROUND: Routinely collected datasets are frequently used for population-based research but their accuracy needs to be assured. AIM: This study aims to assess the accuracy of hospital discharge data in identifying obstetric haemorrhage diagnoses and procedures, and estimate their population incidence. METHODS: The medical records of 1200 randomly selected women were reviewed and compared with obstetric haemorrhage diagnoses and procedures in the hospital discharge data. Sensitivity, specificity, and positive and negative predictive values were calculated using the medical records as the 'gold standard'. Estimates of population incidence were calculated and weighted by the sampling probabilities. RESULTS: Estimated population incidence for any antepartum haemorrhage was 1.8 per 100, and post partum haemorrhage was 7.2 per 100 women. Obstetric haemorrhage diagnosis and procedure codes tended to be underreported, with sensitivities ranging from 28.3% to 100%. All codes had specificities of 98.9% or greater. The identification of obstetric haemorrhage differed between levels of severity. CONCLUSION: The results indicate that population health datasets can be a reliable information source; however, these datasets could be improved with more complete documentation in medical records.

The accuracy of reporting of the hypertensive disorders of pregnancy in population health data.

OBJECTIVE: To assess the accuracy of hypertensive disorders of pregnancy reporting in birth and hospital discharge data compared with data abstracted from medical records. METHODS: Data from a validation study of 1200 women provided the 'gold standard' for hypertension status. The validation data were linked to both hospital discharge and birth databases. Hypertension could be reported in one, both, or neither database. RESULTS: Of the 1184 records available for review, 8.3% of women had pregnancy-related hypertension and 1.3% had chronic hypertension. Reporting sensitivities ranged from 23% to 99% and specificities from 96% to 100%. Using broad rather than specific categories of hypertension and more than one source to identify hypertension improved case ascertainment. Women with severe preeclampsia or adverse outcomes were more likely to have their pregnancy-related hypertension reported. When the hypertension reporting was discordant on the birth and hospital discharge data, the hospital data were more accurate. CONCLUSIONS: Pregnancy-related hypertension is reported with a reasonable level of accuracy, but chronic hypertension is markedly under-ascertained, even when cases were identified from more than one source. Milder forms of hypertension are more likely to go unreported. Studies utilizing population health data may overestimate the proportion of more severe forms of disease and any risk these conditions contribute to other outcomes.

Measuring maternal morbidity in routinely collected health data: development and validation of a maternal morbidity outcome indicator.

BACKGROUND: As maternal deaths become rare in many countries, severe maternal morbidity has been suggested as a better indicator of quality of care. OBJECTIVE: To develop and validate an indicator for measuring major maternal morbidity in routinely collected population health datasets (PHDS). METHODS: First, diagnoses and procedures that might indicate major maternal morbidity were compiled and used to sample possible cases in PHDS; second, a validation study of indicated cases was undertaken by review of birth admission medical records using a nested case-control study approach with 400 possible cases and 800 controls; finally "true" morbidity from the validation study was used to define a maternal morbidity outcome indicator (MMOI) with a high positive predictive value (PPV). Sensitivity, specificity, PPV, negative predictive value (NPV), and exact 95% confidence intervals (95% CI) were weighted by the sampling probabilities. RESULTS: There were 1184 records available for review. Of 393 possible cases only 188 were confirmed as suffering major morbidity (weighted PPV 47.3%, sensitivity 72.9%) and of the 791 initial noncases, 787 were confirmed as noncases (weighted NPV 99.5%, specificity 98.5%). Revision of the initial indicator with exclusion of noncontributing International Classification of Disease (ICD) codes provided a MMOI with population-weighted rate of 1.5%, PPV 94.6% (95% CI: 72.3-99.9), sensitivity 78.4% (95% CI: 55.2-93.1), specificity 99.9% (95% CI: 99.5-99.9), and 99.5% agreement with "true" morbidity (kappa 0.86). CONCLUSIONS: PHDS can be used reliably to identify women who suffer a major adverse outcome during the birth admission and have potential for monitoring the quality of obstetric care in a uniform and cost-effective way.

The accuracy of population health data for monitoring trends and outcomes among women with diabetes in pregnancy.

AIM: To assess the accuracy of routinely collected population birth and hospital datasets in identifying maternal pregestational diabetes mellitus (PDM) and gestational diabetes mellitus (GDM). METHODS: Information on maternal diabetes status was obtained from the medical records of a random sample of 1200 women and compared with routinely collected, population-based birth and hospital data. PDM and GDM are reported in both databases. Sensitivity, specificity, positive predictive value (PPV), negative predictive value and the kappa statistic were determined. RESULTS: Medical records were available for 1184 of the 1200 women sampled. 0.3% of women were classified with PDM and 4.8% with GDM. 'True' PDM was under-reported and misclassified in the birth data, but all cases were reported in the hospital data. GDM was also more completely and more accurately reported in the hospital data than in the birth data. Diabetes requiring insulin was more likely to be reported than non-insulin dependent diabetes. CONCLUSIONS: Hospital data were more sensitive and accurate (higher PPVs) than birth data and these measures were not improved by ascertaining diabetes from either of the two datasets. More severe forms of diabetes were more likely to be reported than less severe.

The prevalence of maternal medical conditions during pregnancy and a validation of their reporting in hospital discharge data.

Population health datasets are a valuable resource for studying maternal and obstetric health outcomes. However, their validity has not been thoroughly examined. We compared medical records from a random selection of New South Wales (NSW) women who gave birth in a NSW hospital in 2002 with coded hospital discharge records. We estimated the population prevalence of maternal medical conditions during pregnancy and found a tendency towards underreporting although specificities were high, indicating that false positives were uncommon.

Postpartum haemorrhage occurrence and recurrence: a population-based study.

OBJECTIVE: To determine the risk of occurrence and recurrence of postpartum haemorrhage (excessive bleeding after childbirth) among women having at least two consecutive pregnancies. DESIGN AND SETTING: Population-based study using longitudinally linked hospital discharge and birth records from New South Wales for the period 1 January 1994 to 31 December 2002. PARTICIPANTS: All 125,295 women having at least a first and second pregnancy resulting in a singleton birth at > 400 g or > or = 20 weeks' gestation in the study period. MAIN OUTCOME MEASURES: Risk of occurrence of postpartum haemorrhage (PPH) in any pregnancy, and of recurrence of PPH in subsequent (second and third) pregnancies. RESULTS: 5.8% of women (7327/125,295) had a PPH in their first pregnancy, and 4.5% (5318/117,968) had a first PPH in their second pregnancy. Among the 23,095 women who had three pregnancies in the study period, 4.4% (908/20,839) had a first PPH in their third pregnancy. The risk of recurrence in a second consecutive pregnancy was 14.8% (1082/7327), and in a third consecutive pregnancy (after two previous PPHs) was 21.7% (43/198); even with an intervening pregnancy with no PPH (ie, PPH in the first and third pregnancies only), the risk for the third pregnancy was 10.2% (111/1085). CONCLUSIONS: These consistently elevated risks of recurrence highlight the need for women with a history of PPH to have active management of the third stage of labour and to give birth in a hospital that has onsite blood cross-match facilities.