Studies with data from multiple sources
| Study, years of data collection and reference | Source of alternative data | Initial BPSU ascertainment | Proportion of cases with paediatric information after case tracing | Comments |
| Studies using multiple sources of ascertainment where a priori, the BPSU was not expected to ascertain all cases | ||||
| National Surveillance of HIV in Pregnancy and Childhood Data collected since 1986—years considered here 2012–2017 www.ucl.ac.uk/nshpc Unpublished data supplied by Peters H | The surveillance has established relationships with named contacts at paediatric and obstetric units across the UK and Ireland. Paediatricians may report through the BPSU or directly to the surveillance team | By the end of 2017, for years of birth 2012–2015, cases ascertained directly through BPSU were 1364/4490 (30.4%) | 90% (see comments for explanation of this proportion) | Cases are initially reported either through the BPSU (numbers given in initial ascertainment column), or directly from individual paediatric or obstetric units. Because of the long-established links between NSHPC and individual units and clinicians which encourage direct reporting, these figures are not a reliable guide to BPSU ascertainment levels. Clinical information including paediatric data is eventually available for around 90% of all cases. It is not possible to give precise data as this information comes from several sources, through a variety of routes, at different times |
| ADHD in transition (CATCh-Us) 2015–2016 Eke et al 5 | Child and Adolescent Psychiatry Surveillance System (CAPSS) | 202/315 (62.3%) There was no overlap in cases reported through both organisations (ie, no cases were reported by both paediatricians and psychiatrists) | Of the final 315 cases reported, 202 were reported through BPSU and 113 through CAPSS. There were no cases reported through both systems. This indicates that no useful information can be obtained about ascertainment as both groups are likely to have been mutually exclusive. An evaluation of the CAPSS ascertainment using an alternative service-based source of data suggested 18 of 76 (24%) eligible cases seen in the South London child psychiatric services were notified via CAPSS12 | |
| Visual Impairment and Blindness 2015–2016 Rahi4 | British Ophthalmic Surveillance Unit (BOSU) | 182/422 (43%) | The low ascertainment rate is due to ophthalmologists being the main secondary care doctor for the majority of these children. Of note, 139 (33%) cases were notified only through the BPSU illustrating the benefit of additional sources of ascertainment to the primary source (in this case BOSU) | |
| Studies comparing BPSU ascertainment with lab-based infectious disease databases | ||||
| Tuberculosis 2003–2005 Teo et al 13 | Enhanced Tuberculosis Surveillance scheme (ETS) covers England, Wales and Northern Ireland | 320/557 (57%) | This study was designed to assess the quality of ascertainment in the ETS rather than the BPSU | |
| Invasive Group B Streptococcal disease 2014–2015 Heath6 | Microbiology laboratory notifications to public health bodies in England, Scotland, Wales, Northern Ireland and Ireland | 49% (numbers not available) | 83% (657/856 paediatric and lab reports, 59/856 paediatric report only, 142/856 lab report only) | The increase in proportion ascertained from paediatricians shows value of backtracking and intensive case tracing |
| Intussception 2008–2009 Samad et al 11 | Hospital Episode Statistics | 190/227 (84%) | Case reports included those notified by paediatric surgeons who were added to BPSU reporting for this study. A capture–recapture analysis calculated the total estimated incidence as 233 cases, which would bring the BPSU ascertainment rate down to 82% | |
| Acute hepatitis 2014–2015 Ladhani8 | Laboratory reports to central public health organisations | 72/84 (86%) | 82/84 (98%) | The total number of cases includes those where no paediatric information was available but hospital admission was confirmed |
| Haemolytic Uraemic Syndrome Adak7 | Enhanced laboratory surveillance for verocytotoxin-producing strain of Escherichia coli
(VTEC) through public health bodies of England, Scotland, Wales, Northern Ireland and the Ireland | 297/365 (81%) | ||
| Congenital Syphilis 2010–2015 Simms et al 14 Simms15 | Public Health bodies of England, Scotland, Wales, Northern Ireland and Ireland Laboratory reports. GUMCAD (Clinical Activity Dataset of GUM clinics) Reporting by microbiologists directly to study | 13/15 (87%) | Although the final study identified 17 cases, the only published data on which ascertainment could be calculated were on the first 15 confirmed cases published in the BPSU Annual report 2013–2014 | |
| Studies comparing BPSU ascertainment with lab-based non-infectious disease databases | ||||
| Congenital Adrenal Hyperplasia 2007–2009 Khalid9 | Individual biochemistry laboratories which carries out diagnostic assays for CAH | 136/144 (94%) | This study involved intensive case tracing both ways—from laboratory reports to clinicians. The majority, but not all, laboratories carrying out relevant assays reported into the study. This study achieved 95% clinical questionnaire returns from all initial case reports which, along with case tracking from lab reports, suggests the ascertainment figure is reliable | |
| Elevated blood lead in children 2010–2012 Ruggles et al 3 | National Public Health Organisations (PHOs) Supraregional Assay Service Trace Element Laboratories | 32/46 (70%) Of 46 confirmed cases 32 reported through BPSU, 32 reported via laboratories of which 19 were also reported through BPSU, and 19 reported through PHOs of which 13 were also reported to BPSU | This study was run through Public Health England and involved the health protection teams in the public health organisations of each of the five countries. Theoretically each source should have been able to ascertain all cases, although there were some cases reported by only one source. While ascertainment in the BPSU was 70%, it was 65% from the laboratories and 41% through the PHOs | |
| Congenital Hypothyroidism 2011–2012 Knowles16 Knowles et al 10 Knowles et al 17 and additional data supplied by Knowles | Antenatal screening laboratories | 360/698 (52%) | 698/739 (94%) Includes 13 initially reported to BPSU with no clinical questionnaire and not traceable to lab reports, and 28 reliable lab reports where the clinician was not traceable | This study enhanced the initial ascertainment by intensive tracing of paediatric clinicians from lab reports, and the use of a mini-questionnaire to confirm cases status from paediatricians where a full questionnaire was not submitted |
ADHD, attention-deficit hyperactivity disorder; BPSU, British Paediatric Surveillance Unit; NSHPC, National Surveillance of HIV in Pregnancy and Childhood.