Risk of self-harm in patients with eating disorders: English population-based national record-linkage study, 1999-2021.

OBJECTIVE: Anorexia nervosa (AN) and bulimia nervosa (BN) are eating disorders associated with high rates of self-harm (SH). This is the first national study in England to quantify this association in a hospital population. METHOD: A retrospective cohort study using a linked national dataset of Hospital Episode Statistics for 1999-2021. The exposure cohort included individuals aged <35 years admitted to hospital with a diagnosis of AN or BN. The reference cohort included hospital controls. We calculated the rate ratio (RR) of SH in each cohort. The individuals in the two cohorts were matched on multiple socio-demographic indicators. The main outcome was a subsequent hospitalization or death record with an SH diagnosis. RESULTS: We identified 15,004 females and 1411 males with AN, and 6055 females and 741 males with BN. The RR with 95% confidence intervals (95%CI) for a subsequent admission with intentional self-harm after admission with AN was 4.9 (95%CI 4.7-5.1) in females and 4.8 (95%CI 3.9-5.8) in males. For BN it was 9.0 (95%CI 8.4-9.6) in females and 9.8 (95%CI 7.7-12.2) in males. There were strong associations between AN and BN and other SH. DISCUSSION: Women and men admitted to English hospitals with AN or BN have a very high risk of a subsequent admission with SH. For some SH behaviors, such as alcohol intoxication, the RR was >10-fold elevated. The magnitude of risk was higher for BN than for AN. Clinicians should be aware of the scale of risk increase. Providing those at risk with appropriate support is required. PUBLIC SIGNIFICANCE: This study is the first national study in an English hospital population that confirms and quantifies the association between eating disorders and self-harm. We have found that both women and men admitted to hospital with anorexia nervosa or bulimia nervosa are at an increased risk of subsequent admission with self-harm. It is important that clinicians are aware of this increased risk to support those at highest risk of self-harm.

Age-specific risk of breast cancer in women with neurofibromatosis type 1.

BACKGROUND: Young women with neurofibromatosis type 1 (NF1) are reported to have a higher risk of breast cancer than others, and this might have implications for screening programmes. Our aim was to calculate this risk. METHODS: An all-England linked data set of hospital admissions and deaths was analysed to determine age-specific rates of breast cancer in women with NF1 and controls. RESULTS: The age-specific excess risk of breast cancer, comparing the NF1 cohort with the control cohort, was elevated 6.5-fold (95% confidence interval 2.6-13.5) in women aged 30-39 years. There was a 4.4 (2.5-7.0) times higher risk among women aged 40-49. CONCLUSIONS: Women with NF1 develop breast cancer at younger ages than the general population.

Fall in population-based mortality from coronary heart disease negated in people with diabetes mellitus: data from England.

AIMS: Mortality rates for coronary heart disease (CHD) have declined substantially in developed countries. Diabetes mellitus is an important risk factor for CHD; its prevalence is increasing. We aimed to compare trends in population-based mortality for CHD without diabetes on the death certificate with trends for CHD with diabetes on the same certificate. METHODS: Analysis of an all-England dataset with multiple-cause coded mortality records from 1995 to 2010. Analysis of a similar dataset for Oxford, because it has the longest run of multiple-cause coded mortality in England, from 1979. Age-specific and age-standardized mortality rates were calculated. RESULTS: In the all-England dataset, there were 1 772 760 deaths with CHD and no mention of diabetes; and 173 184 deaths with CHD and diabetes on the same death certificate. Of deaths with CHD without a mention of diabetes, rates per million men fell by more than half from 2843 (95% confidence interval: 2822-2862) in 1995 to 1379 (1366-1391) in 2010; and those in women halved from 1324 (1314-1336) to 628 (621-635). Of deaths with CHD and diabetes, rates per million men increased from 194 (188-200) to 215 (210-220); and those for women fell, but only very slightly, from 114 (111-118) to 101 (98-104). The longer run in Oxford, from 1979, showed that rates in men and women without diabetes had fallen by two-thirds; but that rates for CHD with diabetes had not fallen. CONCLUSIONS: The fall in mortality from CHD without diabetes has been spectacular. There has been no comparable fall in mortality from CHD with diabetes.

Childbirth in women with intellectual disability: characteristics of their pregnancies and outcomes in an archived epidemiological dataset.

BACKGROUND: Historically, many women with intellectual disability (ID) did not have children. However, a few decades ago it became more common for women with ID to give birth. Our aim was to compare the completed pregnancies of mothers with ID with other completed pregnancies in a large, defined population. METHODS: We used the archived dataset of the Oxford record linkage study from 1970 to 1989, which included a specialized dataset of all maternities in a population of 850,000 people. RESULTS: There were 245,007 births, of which 217 were to mothers with a record of ID. Mothers with a record of ID were more likely than others to give birth at a young age, were much more likely to be unmarried (42% compared with 9%) and to have smoked in the pregnancy (54% compared with 23%). Comparing babies born of mothers with ID and others, percentages with very low birth weight (<2000 g) were similar at 1.9 and 1.8%. There was no excess of babies who were low birth weight for gestational age. Babies born of mothers with ID were much less likely than others to be breastfed (respectively, 33 and 70%). There were no significant associations with presentation at delivery, Caesarean section or forceps delivery. The distribution of Apgar scores at 1 min after birth was similar in babies of mothers with ID and other babies. The sex ratio of the babies was normal. Combining stillbirths and infant deaths, rates per 1000 were 27.9 for babies of mothers with ID and 13.4 in other babies (borderline significant at P = 0.07). CONCLUSIONS: Comparing pregnancies and outcomes of mothers with and without ID, there are some important differences, and some important similarities. It is possible, however, that there may be differences over time and between different places in the care and outcomes of pregnancies in mothers with ID. In the modern era, with electronic health records, it should be possible for maternity services, both locally and on a whole population basis, routinely to monitor the pregnancies, and their outcomes, of mothers with ID.

Maternal and perinatal factors associated with subsequent meningococcal, Haemophilus or enteroviral meningitis in children: database study.

We used a database of 248 659 births, with follow-up to subsequent disease, in the Oxford record linkage archive (1979-1999) to study the influence of family, maternal, and perinatal factors on subsequent hospital admission for meningococcal, Haemophilus, and enteroviral meningitis in the children. In this summary, we report key findings that were significant in multivariate analysis. Meningococcal meningitis was significantly associated with maternal smoking [odds ratio (OR) 2·1, 95% confidence interval (CI) 1·2-3·7]. Haemophilus meningitis was associated with having older siblings (e.g. second child compared to first-born, OR 3·3, 95% CI 2·0-5·6). Enteroviral meningitis was associated with low birth weight (OR 2·2, 95% CI 1·3-3·6) and male sex (OR 1·7, 95% CI 1·2-2·3). The mothers of six of the 312 children with enteroviral meningitis had previously had enteroviral meningitis themselves. We concluded that several maternal characteristics influence the risk of these types of meningitis.

Risk of benign tumours of nervous system, and of malignant neoplasms, in people with neurofibromatosis: population-based record-linkage study.

BACKGROUND: The neurofibromatoses (NF) are genetic disorders. Increased risks of some cancers in people with NF are well recognised, but there is no comprehensive enumeration of the risks across the whole range of site-specific cancers. Our aim was to provide this. METHODS: A linked data set of hospital admissions and deaths in England was used to compare rates of tumours in an NF cohort with rates in a comparison cohort, with results expressed as rate ratios (RR). RESULTS: The RR for all cancers combined, in people with both types of NF combined, was 4.3 (95% confidence interval (CI): 4.0-4.6), based on 769 cases of cancer in 8003 people with NF. Considering only people with presumed NF1 (as defined in the main article), the RR for all cancers excluding nervous system malignancies remained elevated (2.7, 95% CI: 2.4-2.9); and risks were significantly high for cancer of the oesophagus (3.3), stomach (2.8), colon (2.0), liver (3.8), lung (3.0), bone (19.6), thyroid (4.9), malignant melanoma (3.6), non-Hodgkin's lymphoma (3.3), chronic myeloid leukaemia (6.7), female breast (2.3) and ovary (3.7). CONCLUSION: Neurofibromatosis was associated with an increased risk of many individual cancers. The relationships between NF and cancers may hold clues to mechanisms of carcinogenesis more generally.

Risk of pneumonia and pneumococcal disease in people hospitalized with diabetes mellitus: English record-linkage studies.

BACKGROUND: The risk of invasive pneumococcal disease is higher in people with diabetes mellitus than those without. People with diabetes should be considered for routine pneumococcal immunization. This policy has been in place in England for more than a decade. We aimed to estimate, at the population level, the current scale of excess risk of pneumococcal disease in patients with diabetes, and whether the risks have decreased in recent years with the introduction of a pneumococcal vaccine. METHODS: We used two data sets of linked hospital admission and death records-the Oxford Record Linkage Study (1963-1998) and all-England linked hospital episode statistics (1999-2011). As a measure of relative risk, we calculated the rate ratio of pneumococcal disease in cohorts of people hospitalized with diabetes compared with cohorts without a record of diabetes. RESULTS: The risk of pneumococcal disease in patients hospitalized with diabetes mellitus has declined a little, but it is still high. The all-ages rate ratio in England declined from 1.92 (95% CI 1.89-1.94) in 1999-2002 to 1.68 (95% CI 1.65-1.71) in 2007-2011. In people aged under 60 years, rate ratios were higher and their decline was more substantial: rate ratios declined from 3.37 (95% CI 3.28-3.46) in 1999-2002 to 2.33 (95% CI 2.21-2.45) in 2007-2011. CONCLUSIONS: Patients admitted to hospital with diabetes mellitus remain at increased risk of pneumococcal infection despite a national immunization policy. Possible explanations for the elevated risk include low vaccine uptake or low effectiveness of available vaccine. Clinicians should be aware of the risk of pneumococcal infection in people with diabetes.

Avoidable mortality in people with schizophrenia or bipolar disorder in England.

OBJECTIVE: To quantify the extent of 'avoidable mortality' in those with schizophrenia or bipolar disorder and to quantify the effect a reduction in these might have on the mortality gap associated with severe mental illness. METHOD: A cohort was studied of people aged <75 years, discharged from inpatient care with schizophrenia or bipolar disorder in 2006-2007, and followed up for 365 days. Standardised mortality ratios (SMRs) were calculated followed by hypothetical SMRs, estimating the residual mortality gap if 'avoidable' causes and suicide in the cohorts had occurred at the same level as those in the general population. RESULTS: Avoidable deaths comprised 60.2% and 59.2% of all deaths in the schizophrenia and bipolar disorder cohorts respectively. All-cause SMRs were 4.23 (95% CI 3.85-4.60) and 2.60 (2.21-3.00) respectively. After discounting the excess attributable to avoidable causes and suicide, the SMRs fell to 2.38 (2.09-2.66) and 1.66 (1.35-1.98) respectively. CONCLUSION: Bringing mortality from avoidable causes and suicide down to general population levels would reduce the overall mortality excess in severe mental illness substantially, by about 50%, but would not eliminate it. Other underlying factors beyond those conventionally considered as 'avoidable' need further research.

Central nervous system abscesses and empyemas in England: epidemiological trends over five decades.

OBJECTIVES: To report on population-based epidemiological trends in central nervous system (CNS) abscesses and empyemas in England over five decades. METHODS: Trend analyses of age-sex-specific hospital admission and death rates using routinely collected English national hospital discharge records, mortality records, and annual population denominators from 1968 to 2019. RESULTS: Hospital admission rates for CNS abscesses and empyemas were stable in England until the late 1980s. In the last two decades of the study period (1999-2019), first-time admissions increased from 1.24 per 100,000 population in 1999 (95% confidence interval [CI] 1.14-1.35) to 2.86 in 2019 (95% CI 2.72-3.01). Admission rates were highest among infants and older adults, and were higher for males than females. There were small but significant increases in annual mortality rates for CNS abscesses and empyemas over the last two decades of the study period after accounting for population ageing, but mortality remained low at around 0.1-0.2 per 100,000 population. Mortality increased with advancing age; deaths in childhood were extremely rare. Case fatality rates where a relevant diagnosis was recorded as either the underlying or contributing cause were 4.3% and 9.7% respectively. CONCLUSIONS: The increase in CNS abscesses and empyemas in England might reflect improved case ascertainment, but the likelihood of a true rise in incidence should be considered.

Cancer of the body of the uterus: trends in mortality and incidence in England, 1985-2008.

OBJECTIVE: To report on trends in mortality and incidence for uterine cancer in England, 1985-2008. DESIGN: Database analysis. SETTING: England. POPULATION: Data sets of English national mortality and cancer registration statistics. METHODS: Analysis of data from death certificates with a mention of malignant neoplasm of corpus uteri or malignant neoplasm of uterus without specification of part, and analysis of incidence data from cancer registries in England. MAIN OUTCOME MEASURES: Age-adjusted and age-specific trends in rates. RESULTS: Mortality rates in the UK declined from 41.3 per million women in 1985 until about 1999 (30.0 per million), and then increased to 35.9 per million in 2008. Incidence rates for those aged 55 years and over showed little or no change from 1985 to the early 1990s, but then increased: for example, for women aged 60-64 years, the rates per million women in 1985, 1993, 2001 and 2008 were, respectively, 448, 461, 647 and 728. Rates for younger women showed a decline, followed by an increase: for example, for women aged 50-54 years the corresponding rates were 290, 254, 267 and 294. Over the period 2001-2008, the age-adjusted average annual percentage increase was 2.55% per annum (95% CI 2.13-2.98%). The increase in incidence rates, observed from the 1990s, preceded the increase in mortality by about 10 years. CONCLUSIONS: After many years of declining rates, uterine cancer has become more common in the first decade of this century. One possible contributor to this increase may be an association with the increase in body mass index in women, but other causes are possible and are discussed.

Death rates for asthma in English populations 1979-2007: comparison of underlying cause and all certified causes.

OBJECTIVE: To report on trends in mortality for asthma using all certified causes of death mentioned on death certificates (conventionally termed 'mentions'), not just the underlying cause. STUDY DESIGN: Retrospective analysis using death certificate information and population data. METHOD: Analysis of mortality records in the Oxford region (mentions available from 1979 to 2007) and English national data (mentions available from 1995 to 2007). The data were considered in periods defined by different national rules for selecting underlying cause of death (1979-1983, 1984-1992, 1993-2000, 2001-2007), and were also analysed as single calendar years. RESULTS: In Oxford, underlying cause mortality rates per million population in the four periods were 25, 32, 22 and 15, respectively. Rates for mentions were 44, 47, 41 and 29, respectively. Rule changes exaggerated the increase in underlying cause mortality in 1984-1992 (when 67% of asthma deaths were coded as underlying cause). Conversely, the decrease in underlying cause mortality for asthma by 2001-2007 is less than it seems (because just under 50% of asthma deaths in 2001-2007 were coded as underlying cause). Comparisons of trends in asthma and chronic obstructive airways disease (COPD) for individuals aged ≥ 55 years showed a decrease for both asthma and COPD in men; in women, a decrease in asthma and an increase in COPD was seen from the early 1990s. CONCLUSIONS: Approximately half of all deaths certified for asthma are missed when asthma mortality is analysed using underlying cause alone. The long-term decreasing trend in asthma mortality is real, and is not attributable to a trend in transferring certification from underlying to contributing cause. Nonetheless, caution is needed when comparing asthma deaths using underlying cause alone across periods that include changes to rules for the selection of underlying cause.

Cancer in patients admitted to hospital with diabetes mellitus aged 30 years and over: record linkage studies.

AIMS/HYPOTHESIS: The aim of this study was to determine the risk of cancer in people admitted to hospital for diabetes mellitus when aged 30 or older. METHODS: This study involved the analysis of two statistical datasets of linked hospital and mortality data, in an area in southern England, between 1963 and 1998 (the Oxford Record Linkage Study, ORLS1) and between 1999 and 2008 (ORLS2). Rates of cancer in the diabetes cohorts were compared with rates of cancer in reference cohorts and expressed as rate ratios. RESULTS: The rate ratio for all cancer in people admitted to hospital with diabetes was 1.01 (95% CI 0.95-1.06, based on 15,898 people with diabetes) for the years 1963-1998; and 1.09 (1.00-1.19, based on 7,771 people with diabetes) in the years 1999-2008. In both datasets, there were significantly high rate ratios for cancers of the liver (ORLS1 and ORLS2, respectively, 2.0 [95% CI 1.4-2.9]; 2.5 [95% CI 1.3-4.3]), pancreas (2.2 [95% CI 1.8-2.7]; 3.5 [95% CI 2.5-4.8]) and uterus (1.5 [95% CI 1.0-2.2]; 2.6 [95% CI 1.4-4.5]). There were significantly low rate ratios for cancer of the prostate (0.6 [95% CI 0.5-0.7]; 0.7 [95% CI 0.5-0.9]) and non-melanoma skin cancer (0.6 [95% CI 0.5-0.8]; 0.8 [95% CI 0.6-0.96]). CONCLUSIONS/INTERPRETATION: Diabetes mellitus was associated with an elevated risk of some site-specific cancers and a reduction of risk of others. Considering the risk in diabetes of all cancers combined, the elevation of risk, if any, is likely to be small and numerically less important than other known complications of diabetes.

Birthweight and the risk of childhood-onset type 1 diabetes: a meta-analysis of observational studies using individual patient data.

AIMS/HYPOTHESIS: We investigated whether children who are heavier at birth have an increased risk of type 1 diabetes. METHODS: Relevant studies published before February 2009 were identified from literature searches using MEDLINE, Web of Science and EMBASE. Authors of all studies containing relevant data were contacted and asked to provide individual patient data or conduct pre-specified analyses. Risk estimates of type 1 diabetes by category of birthweight were calculated for each study, before and after adjustment for potential confounders.Meta-analysis techniques were then used to derive combined ORs and investigate heterogeneity between studies. RESULTS: Data were available for 29 predominantly European studies (five cohort, 24 case-control studies), including 12,807 cases of type 1 diabetes. Overall, studies consistently demonstrated that children with birthweight from 3.5 to 4 kg had an increased risk of diabetes of 6% (OR 1.06 [95% CI 1.01-1.11]; p=0.02) and children with birthweight over 4 kg had an increased risk of 10% (OR 1.10 [95% CI 1.04-1.19]; p=0.003), compared with children weighing 3.0 to 3.5 kg at birth. This corresponded to a linear increase in diabetes risk of 3% per 500 g increase in birthweight (OR 1.03 [95% CI 1.00-1.06]; p=0.03). Adjustments for potential confounders such as gestational age, maternal age, birth order, Caesarean section, breastfeeding and maternal diabetes had little effect on these findings. CONCLUSIONS/INTERPRETATION: Children who are heavier at birth have a significant and consistent, but relatively small increase in risk of type 1 diabetes.

Benign breast disease and subsequent breast cancer: English record linkage studies.

BACKGROUND: Benign breast disease (BBD) increases the risk of breast cancer, but details of the relationship would benefit from further study in the UK. METHODS: Analysis of linked statistical abstracts of hospital data, including a cohort of 20 976 women with BBD in an Oxford data set and 89 268 such women in an English national data set. RESULTS: Rate ratios (RRs) for breast cancer, comparing BBD and comparison cohorts in these two data sets, were 2.3 (95% CI: 2.2-2.5) and 3.2 (3.0-3.3), respectively. RRs rose with increasing age at BBD diagnosis and remained elevated for at least 20 years after diagnosis. RRs were particularly high for a relatively small number of cancers occurring in the first few months after BBD diagnosis. CONCLUSIONS: Our findings accord well with those in other large studies, mostly done in the USA, in showing a sustained long-term cancer risk after BBD. They also demonstrate that known long-term risks of disease can be reliably identified from linked routine administrative hospital statistics. Most other studies omit cancers in the first few months after BBD. Such cases-presumably either misdiagnosed or miscoded-merit further study to determine whether in fact they include diagnoses of cancer that were initially missed.

Seasonal variation in hospital admission for road traffic injuries in England: analysis of hospital statistics.

OBJECTIVE: To report on seasonal variation in hospital admissions for traffic injuries to car and truck occupants, motorcyclists, cyclists, and pedestrians. METHODS: Descriptive study of hospital admission statistics in England, 1999-2004. MAIN OUTCOME MEASURES: national average monthly admissions for injury; severity of injury, defined as death in hospital or a hospital admission lasting four days or more. RESULTS: Admissions for car occupants were highest in the winter months but seasonal variation was not great (highest and lowest months: December, 16% above monthly average; June, 5% below). There was a summer peak and winter trough in admissions for adult cyclists (June, 34% above average; December, 27% below) and motorcyclists (August, 33% above average; January, 43% below). Admissions for child pedestrians were highest in late spring and lowest in mid-winter (May, 24% above average; December, 28% below). By contrast, admissions for adult pedestrians were higher in winter than summer (December, 33% above average; July, 17% below). From April to September, there were more admissions for pedestrians and cyclists in England (44 875 in the six years of the study) than for car occupants (34 582). For cyclists, proportionally more injuries in the winter months were severe. Severity of injuries to car occupants did not show seasonal variation. CONCLUSIONS: The public health "toll" of traffic injuries, measured as total numbers of people injured, varies substantially by season. Although it is important to reduce all injuries, the safety of pedestrians and cyclists, as unprotected road users, needs particular attention.

Immune-mediated and chronic inflammatory disease in people with sarcoidosis: disease associations in a large UK database.

BACKGROUND: Sarcoidosis is a multi-system disorder characterised by non-caseating granulomas. Coexistence of sarcoidosis with immune-mediated and chronic inflammatory diseases has been described in case series. However, the coexistence of two different diseases in individuals can occur by chance, even if each of the diseases is rare. AIM: To determine whether sarcoidosis necessitating hospital admission or day-case care coexists with a range of immune-mediated and chronic inflammatory diseases more commonly than expected by chance. DESIGN: Analysis of an epidemiological database of hospital admission and day-case statistics, spanning 30 years. RESULTS: 1510 patients with sarcoidosis were identified (mean age 44 years, median follow-up 19 years) who had been admitted to hospital or day-case care. Significant associations in the sarcoidosis cohort were identified with systemic lupus erythematosus (odds ratio (OR) 8.3; 95% CI 2.7 to 19.4), autoimmune chronic hepatitis (OR 6.7; 95% CI 1.8 to 17.1), multiple sclerosis (OR 3.3; 95% CI 1.7 to 5.6), coeliac disease (OR 3.1; 95% CI 1.01 to 7.3), thyrotoxicosis (OR 2.5; 95% CI 1.4 to 4.0), myxoedema (OR 2.2; 95% CI 1.2 to 3.7) and ulcerative colitis (OR 2.1; 95% CI 1.1 to 3.7). Weaker associations were found for diabetes mellitus with a first admission aged 30-49 years (OR 2.9; 95% CI 2.1 to 4.0) or age >50 (OR 1.7; 95% CI 1.2 to 2.3), but not for people age <30. No significant association with Crohn's disease (OR 1.52; 95% CI 0.61 to 3.14) or primary biliary cirrhosis (OR 3.75; 95% CI 0.77 to 11.0),was found. When all immune-mediated and chronic inflammatory diseases for which associations were sought were combined, the overall rate ratio associated with sarcoidosis was 2.2 (95% CI 1.9 to 2.6). CONCLUSION: This study adds epidemiological evidence to information from clinical reports that there is a connection between sarcoidosis and other immune-mediated and chronic inflammatory diseases.

Diagnostic outcomes following childhood non-specific abdominal pain: a record-linkage study.

AIMS: Non-specific abdominal pain (NSAP) is the most common diagnosis on discharge following admission for abdominal pain in childhood. Our aim was to determine the risk of subsequent hospital diagnosis of organic and functional gastroenterological conditions following a diagnosis of NSAP, and to assess the persistence of this risk. METHODS: An NSAP cohort of 268,623 children aged 0-16 years was constructed from linked English Hospital Episode Statistics from 1999 to 2011. The control cohort (1,684,923 children, 0-16 years old) comprised children hospitalised with unrelated conditions. Clinically relevant outcomes were selected and standardised rate ratios were calculated. RESULTS: From the NSAP cohort, 15,515 (5.8%) were later hospitalised with bowel pathology and 13,301 (5%) with a specific functional disorder. Notably, there was a 4.84 (95% CI 4.45 to 5.27) times greater risk of Crohn's disease following NSAP and a 4.23 (4.13 to 4.33) greater risk of acute appendicitis than in the control cohort. The risk of irritable bowel syndrome (IBS) was 7.22 (6.65 to 7.85) times greater following NSAP. The risks of inflammatory bowel disease (IBD), IBS and functional disorder (unspecified) were significantly increased in all age groups except <2-year-olds. The risk of underlying bowel pathology remained raised up to 10 years after first diagnosis with NSAP. CONCLUSIONS: Only a small proportion of those with NSAP go on to be hospitalised with underlying bowel pathology. However, their risk is increased even at 10 years after the first hospital admission with NSAP. Diagnostic strategies need to be assessed and refined and active surveillance employed for children with NSAP.