Journal Club

Dr Karen Arnold

July 2015

Case Scenario• 18 month old boy• Multiple bruises• Knees, shins, right elbow, right hand, forehead• Boisterous boy with older siblings• “bruise easily”

• Patterns of bruising in preschool children- a longitudinal study

• Alison M Kemp et al

• Arch Dis Child 2015 100: 426-431

Aims

• Identify the prevalence and pattern of bruises in preschool children over time and to explore influential variables

Methods• Prospective longitudinal study• Children aged 0-6 years• South Wales• Recruited from well-baby clinics, hospital outpatient

clinics and mother & baby groups• Excluded motor disability, confirmed bleeding disorder or

suspected child abuse

• 2 phases in study• Phase 1 (Apr 2005 – Dec 2007)• Phase 2 (Apr 2008 – Aug 2011)• Both phases collected gender, age, developmental stage

and PMH• Phase 2 included ethnicity, family order and

socioeconomic status, parental report on cause of bruise

• Parents trained to recognise a bruise• Record number and location of bruises• Trained not to record cuts, abrasions, birthmarks• Bruises recorded on body map weekly for up to 12

weeks• Walking children had four weekly collections

• At each collection parents recorded the most advanced developmental milestone reached

• Classified as premobile (not rolling or rolling), early mobile (crawling or cruising) and walking

Validation• Parents recorded number, location and size of bruise• Random sample of 40 data collections from 40 different

children selected• Single research nurse verified lesions recorded by

independently repeating process• Parents not warned about visit

Analysis

• Longitudinal analysis performed using multilevel modelling with Poisson distribution and a log link function

Categorisation of sites where the locations of bruises were recorded at each data collection.

Alison M Kemp et al. Arch Dis Child 2015;100:426-431

Copyright © BMJ Publishing Group Ltd & Royal College of Paediatrics and Child Health. All rights reserved.

Results• 1002 parents approached• 380 gave informed consent• 328 children had at least one collection

(mean age 19 months, 54% female)• 2570 data collections• Average of 7.8 collections per child

• Ethnicity recorded in 280 children

269 White British

11 other ethnic groups

• Socioeconomic status

46% from least deprived area

15% two most deprived areas

• Validation process showed complete agreement between parents and research nurse for number and site of bruising

• For bruise size (within 5mm) agreement was 55% - size not included in analysis

Number and percentagewith 95% CI of collectionswith at least one bruise

Mean (SD) numberof bruises per collection

Mean (SD) numberof sites affectedper collection

Range, median and90th centile of numberof bruises per collection

PremobileAge range 0–11 months

68/10106.7%(95% CI 5.3 to 8.4)

0.09 (0.35) 0.08 (0.30)0–300

Early mobileAge range 4–18 months

218/47845.6%(95% CI 41.2 to 50.1)

0.80 (1.19) 0.59 (0.79)0–702

WalkingAge range 10–70 months

852/108278.8%(95% CI 76.2 to 81.1)

2.82 (2.77) 1.50 (1.23) 0–16

10–36 months 422/59471.0%(95% CI 67.3 to 74.5)

2.31 (2.59) 1.28 (1.17)0–1426

37–70 months 387/43090.0%(95% CI 86.8 to 92.4)

0–1637

OverallAge range 0–70 months

1138/257044.8%(95% CI 42.4 to 46.2)

1.37 (2.27) 0.77 (1.10)0–1604

Prevalence• Prevalence of bruising in first collection for each child

was:

– 5.3% premobile children– 55.8% early mobile children– 87.5% walking children

• 12 bruises recorded in 9/405 collections in children not rolling

2.2% ( 95% CI 1.2% - 4.2%)

• 75 bruises recorded in 59/605 collections in children rolling but not yet crawling

9.8% ( 95% CI 7.6% - 12.4%)

Premobile1010 collections

Early mobile478 collections

Walking1082 collections

Number of collections

Per centNumber of collections

Per centNumber of collections

Per cent

Left cheek 1 0.1 11 2.3 15 1.4

Right cheek 1 0.1 12 2.5 18 1.7

Left ear 0 0.0 1 0.2 0 0.0

Right ear 0 0.0 1 0.2 1 0.1

Head 12 1.2 32 6.7 34 3.1

Facial T 13 1.3 65 13.6 122 11.3

Eyes 0 0.0 8 1.7 12 1.1

Front trunk 2 0.2 0 0.0 43 4.0

Rear trunk 9 0.9 11 2.3 104 9.6

Neck 0 0.0 1 0.2 0 0.0

Buttocks 0 0.0 1 0.2 54 5.0

Genitalia 0 0.0 0 0.0 0 0

Upper arms 2 0.2 6 1.3 99 9.1

Hands 2 0.2 3 0.6 6 0.6

Lower arms 5 0.5 6 1.3 126 11.6

Front thighs 10 1.0 18 3.8 181 16.7

Back thighs 4 0.4 3 0.6 72 6.7

Feet 2 0.2 6 1.3 45 4.2

Below knees 14 1.4 99 20.7 693 64.0

All bruises 68 6.7 218 45.6 852 78.7

Distribution of percentage of 2570 collections from 328 children with at least one bruise by location and development stage.

Alison M Kemp et al. Arch Dis Child 2015;100:426-431

Copyright © BMJ Publishing Group Ltd & Royal College of Paediatrics and Child Health. All rights reserved.

• Some sites numerous bruises• Mean number bruises below knee 2.8• Mean number bruises to front of thigh 1.4

• Ears, neck, genitalia and hands were rarely bruised (<1% of collections)

• Majority premobile children had no bruise• 73% (97/133) never had a bruise in 1010 collections

• Independently mobile children• Wide variation between children• Considerable variation between collections

• Walking children• Majority of variation was between children• Suggests different children tend to sustain different amounts

of bruising

Factor Ratio of means 95% CIs

Development stage

 Premobile 0.06 0.04 to 0.09

 Early mobile (reference) 1

 Walking 2.28 1.73 to 2.99

 Gender (female) 0.84 0.64 to 1.09

 Any sibling 1.51 1.15 to 1.99

Season

 January–March (reference) 1

 April–June 0.99 0.82 to 1.20

 July–September 0.94 0.74 to 1.18

 October–December 0.82 0.63 to 1.06

Ethnicity

 White 1

 Ethnicity (not White) 0.90 0.46 to 1.79

 Not recorded 0.84 0.58 to 1.23

Deprivation

 Quintile 1 1

 Quintile 2 0.96 0.70 to 1.32

 Quintile 3 1.14 0.79 to 1.65

 Quintile 4 0.94 0.59 to 1.48

 Quintile 5 1.56 0.91 to 2.66

Conclusions• Strong relationship between presence, number and

location of bruises and motor development stage• Bruising in babies who were not yet rolling was

uncommon• Percentage of collections with at least one bruise and

the number of bruises at each collection increased with increased development stage

Comparisons and Contrasts

• Greater prevalence of bruising in this study for all developmental groups

• Comparative study with some children who have more than expected number of bruises

Safeguarding Considerations• Probability of abuse in this population low• Explanations, where available, compatible with bruise• If bruise pattern deemed unusual they were

independently reviewed by child protection team and child abuse was excluded

• Parents informed that referral to social services would be made if maltreatment suspected

Limitiations• Recruitment challenging as it required prolonged

participation• Over representation of less socially deprived population• Parents knowledge of cause of bruising

Analysis• Clearly focused issue?• Cohort recruitment?• Potential bias excluded?• Confounding factors?• Follow-up long/complete enough?• Results?• Can results be applied locally?• Implications?

Clearly focused issue?• Bruising in preschool children (0-70months)• Prevalence of bruising• Patterns of bruising

Cohort recruitment?• Wide cross section – well baby clinics, hospital OPDs,

mum & baby groups• 1002 approached, 328 in study• Under representation of lower socioeconomic groups• How was suspected child abuse identified and excluded?• Cohort of motivated parents agreeing to intensive

participation

Potential bias excluded?• Selection bias – motivated parents BUT prospective

study• Study compliance – some participants involved for 24

weeks• Subjective recording by parents• Inter – observer variability – reduced by training parents

Confounding factors?• Non-accidental injury• Suspected child abuse excluded• Unusual pattern of injury reviewed by child protection

team and abuse excluded

Was follow-up complete?• Over 12-24 weeks• Could potentially follow-up until 16 years looking for SC

involvement and look back at their history of bruising

Results• Bruising in non-rolling babies uncommon• Number of bruises increases with developmental stage• Ears, neck, genitalia, hands are rare sites for bruising in

any preschool child• 9% children have twice as many bruises as would be

expected

Case Scenario• 18 month old boy• Multiple bruises• Knees, shins, right elbow, right hand, forehead• Boisterous boy with older siblings• “bruise easily”