Skip to main content

Table 6 Outcomes and data fields assessed in pilot: records available and feasibility assessment

From: Assessing the impact of specialist home visiting upon maltreatment in England: a feasibility study of data linkage from a public health trial to routine health and social care data

Outcomes Data source: native field name Missing (n) Commentary of findings Impact
 Child in Need (CIN) status as of 31 March each year NPD > CIN:
Referral date
1 No data appear across one record—record to be excluded (this will apply to all fields below) Note: 34 records with dates prior to time point ranges (1997–2007). This is expected. None
  NPD > CIN:
Referral—no further action
42 No data collected in 2008/2009 time point (accounts for 38 records). Some blanks appear in 2009/2010 time point; however, the referral date on these records is prior to 1st April of that data collection year. *Assumption that time point cycle is Apr–Mar. Low: assumption missing data indicate further action was required and that the child was in need
  NPD > CIN:
Reason for closure
142 No pattern—further investigation required As above
Secondary a
 CIN categorisation NPD > CIN:
Category of Abuse
329 Only data from 2008/2009 accessed in pilot. For main phase data from 2012/2003 will be accessed and also ‘NPD > CIN: Latest category of abuse’ will be included which may improve data quality. Still to be determined
 Child looked after (CLA) status (1) NPD > CLA:
Category of need; Legal status; Placement; REC
0 All records returned are complete. None
 Child Protection registration (plan) (1) NPD > CIN:
Child Protection Plan (CPP) Indicator
195 Expected—all missing from 2010/2011 time point onwards. Data not collected during these years. Low: CPP flag can be determined from other fields
  NPD > CIN:
No. of previous Child Protection Plans
320 No pattern to missingness. Only 11 records have a value recorded, 9 of these are zero. As above
  NPD > CIN:
Child Protection Plan start date
320 Expected—not all children will have had a CPP. Only 11 records have a date recorded, these correspond with data captured in the ‘no. of previous CPPs’ above. As above
  NPD > CIN:
Child Protection Plan end date
327 Expected—only 4 records have an end date recorded. Corresponds with those records where a start date is recorded. Data check done—end dates are after the start date. As above
 Exploratory Markov chain modellingb NPD > CIN:
Date of initial child protection conference
327 Expected—not all children would have had a child protection conference. However further checks required to confirm validity of data. Medium: Low numbers may impact analysis
 Injuries and ingestions NHSD > A&E:
A&E diagnosis (diag n)
5981 45% missing (1650/6336 missing in BB trial—26% missing) Medium: All diag/treat/inv. fields to be used in combination to define inj/ingc
  NHSD > A&E:
A&E diagnosis2 char (diag2 n D)
3604 27% missing (1849/6336 missing in BB trial—29% missing) As above
  NHSD > A&E:
A&E investigation (invest n)
1728 13% missing (1396/6336 missing in BB trial—22% missing) As above
  NHSD > A&E:
A&E investigation2 char (invest n D)
1712 13% missing (1395/6336 missing in BB trial—22% missing) As above
  NHSD > A&E:
A&E treatment (treat n)
2349 18% missing (1411/6336 missing in BB trial—22% missing) As above
A&E treatment – 2 Char (treat2 n D)
2126 16% missing (1417/6336 missing in BB trial—22% missing) As above
  NHSD > A&E:
A&E diagnosisAnatomical Area (diaga n D)
9281 70% missing (4725/6336 missing in BB trial—74% missing)  
  1. (1) small numbers may be an issue—descriptive analysis will be used if necessary
  2. aAdditional fields were retrieved for secondary outcomes and assessed solely for presence (Special Educational Needs, Disability, Day care attendance, Early Years assessment, School attendance, Key stage one attainment)
  3. bTo explore probability of progression through each stage of child protection process
  4. cSame fields also contribute to assessment of subsequent pregnancies (via pregnancy-related A&E attendances)