Is the lab useful to distinguish Septic Arthritis versus Transient Synovitis?

1
Is the lab useful to distinguish Septic Arthritis
versus Transient Synovitis?

• Laboratory rounds
• Jan 30, 2003
• Rob Hall PGY3

2
Case

• 3yo
• Refuses to weight bear on left hip
• No hx of trauma or fever
• URTI recently
• T 38.0
• Painful ROM right hip
• Kid looks well

• What is your differential?
• What tests would you order?
• How useful are your tests going to be?

3
Differential of Limp

• Trauma
• fracture, sprain, strain, contusion, abuse
• Infection
• Septic arthritis, osteomyelitis, cellulitis,
  fascitis, myositis
• Inflammation
• Toxic synovitis, JRA, HSP, serum sickness,
  rheumatic fever
• Other
• AVN, SCFE, tumor, leukemia, sickle cell

4
Are a wbc, ESR, or CRP useful for distinguishing
septic arthritis vs transient synovitis?

• Small retrospective studies with poor design
• Molteni 1978
• McCarthy1980
• Potentially meaningful studies
• Kunnamo 1987
• Del Beccaro 1992
• Kocher 1999

5
Kunnamo 1987American Journal of the Diseased
Child

• Prospective study of all kids referred to a
  tertiary pediatric referral center with arthritis
• N 278, only 18 with septic arthritis
• Gold standard was based on follow up if joint not
  tapped and no follow up data given

6
Kunnamo 1987American Journal of the Diseased
Child

• Variable Sensitivity Specificity
• CRP gt 20 94 92
• ESR gt 20 94 68
• WBC gt 12 59 95
• T gt 38.5 77
  95
• Tgt38.5 or 100 87
  CRP gt20

7
Kunnamo 1987American Journal of the Diseased
Child

• Small study
• Method problems
• Wide confidence intervals
• CRP gt 20 sensitivity 94 (95CI 72 99)
• Doesnt answer the question

8
Del Beccaro 1992Annals of Emergency Medicine

• Retrospective chart review
• Method problems
• N 132
• Septic arthritis 38
• Transient synovitis 94
• Gold standard problems
• Transient synovitis defined based on clinical
  course but 13 lost to follow up

9
Del Beccaro 1992Annals of Emergency Medicine

• Statistically significant differences
• Variable Septic Arthritis Transient
  synovitis
• Temp 38.1 37.2
• ESR 44 19
• WBC 13.2 11.2

10
Del Beccaro 1992Annals of Emergency Medicine
VARIABLE SENSITIVITY SPECIFICITY
ESR gt 20 79 72
T gt 37.5 66 68
T gt 38.0 45 85
WBC gt 15 26 84
ESR gt 20 and/or Temp gt 37.5 97 50
11
Del Beccaro 1992Annals of Emergency Medicine

• Wide confidence intervals
• Huge overlap between groups with each variable
• How can we use this information?
• ESR, CBC, temp not that helpful in isolation
• Combinations may be useful

12
Kocher 1999Journal of Bone and Joint Surgery

• Retrospective chart review of 282 patients
• Same metholodogical problems
• Found statistical differences b/w septic
  arthritis and transient synovitis for
• Hx of fever
• Ability to weight bear
• Joint effusion on Xray
• Temp
• ESR
• WBC

13
Kocher 1999Journal of Bone and Joint Surgery

• Multivariant analysis found four predictors of
  septic arthritis
• History of fever
• Non-weight bearing
• ESR gt 40
• WBC gt 12
• Developed an algorithm based on above 4 variables

14
Kocher 1999Journal of Bone and Joint Surgery
Predictors Transient Synovitis Group Septic Arthritis Group Probability of septic arthritis ()
0 22 0 0.2
1 54 1 3.0
2 19 14 40
3 5 53 93
4 0 30 99.6
15
Kocher 1999Journal of Bone and Joint Surgery

• This is NOT a validated prediction rule
• Again suggests that combination of information is
  useful

16
Approach to the Irritable Hip
17
Surprise! The lab alone is NOT the answer