Journal of Orthopaedic Surgery 2015;23(1):80-3
Preoperative antibiotics for septic arthritis in
children: delay in diagnosis
Simon BM MacLean, Christopher Timmis, Scott Evans, Dominik Lawniczak, Amit Nijran, Edward Bache
Birmingham Children’s Hospital, Birmingham, United Kingdom
ABSTRACT
Purpose. To review the records of 50 children who
underwent open joint washout for septic arthritis with
(n=25) or without (n=25) preoperative antibiotics.
Methods. Records of 50 children who underwent
open joint washout for presumed septic arthritis with
(n=25) or without (n=25) preoperative antibiotics
were reviewed. 17 boys and 8 girls aged 3 weeks to
16 years (median, 1.5 years) who were prescribed
preoperative antibiotics before joint washout were
compared with 12 boys and 13 girls aged one
month to 14 years (median, 2 years) who were not.
Following arthrotomy and washout, all patients were
commenced on high-dose intravenous antibiotics.
Patients were followed up for 6 to 18 months until
asymptomatic.
Results. Patients who were referred from places other
than our emergency department were twice as likely
to have been prescribed preoperative antibiotics
(p=0.0032). Patients prescribed preoperative
antibiotics had a longer median (range) time from
Address correspondence and reprint requests to: Mr Simon MacLean, Penthouse 5, Ludgate Lofts, 17 Ludgate Hill, Birmingham,
B3 1DW, United Kingdom. Email: simonmaclean81@gmail.com
symptom onset to joint washout (8 [2–23] vs. 4
[1–29] days, p=0.05) and a higher mean erythrocyte
sedimentation rate (93.1 vs. 54.3 mm/h, p=0.023)
at presentation. Nonetheless, the 2 groups were
comparable for weight bearing status, fever, and
positive culture, as well as the mean (range) duration
of antibiotic treatment (4.9 [4–7] vs. 4.7 [1–8] weeks,
p=0.586).
Conclusion. Preoperative antibiotics should be
avoided in the management of septic arthritis in
children. Their prescription delays diagnosis and
definitive surgery, and leads to additional washouts
and complications. A high index of suspicion and
expedite referral to a specialist paediatric orthopaedic
unit is needed if septic arthritis is suspected.
Key words: septic arthritis; children; antibiotics;
complications; arthrotomy
INTRODUCTION
Approximately 75% of septic joints occur in children
aged <5 years.1 Differentiating septic arthritis from
Vol. 23 No. 1, April 2015 Preoperative antibiotics for septic arthritis in children 81
other pathology in children is difficult.2,3 Algorithms
for the diagnosis of septic arthritis have been
reported.4 Synovial fluid analysis and culture are
imprecise and time-consuming.5 Laboratory results
may not be the gold standard for diagnosis. Delayed
treatment is associated with complications, additional
surgery, and long-term morbidity.6,7 Preoperative
antibiotics are generally not advised unless the child
is extremely unwell.7,8 Surgical drainage is suggested
for removal of inflammatory products that can cause
destruction of articular cartilage, even in the absence
of viable micro-organisms.9 Arthrotomy is preferred
for urgent decompression of infant hips,9 whereas
aspiration can be performed in superficial joints
and can be repeated if necessary.10,11 Aspiration can
also be used for septic arthritis of the hip in older
children with <4 days of symptoms.12 Diagnostic
arthrocentesis together with 2 weeks of exceptionally
high dose antibiotics (including 2 to 4 days of
intravenous antibiotics initially) is also advocated.13
These less-invasive surgical strategies require careful
and repeated review to gauge appropriate progress
and should never be used in infants aged <12 months
or those who are immunocompromised. Arthrotomy
is the treatment of choice for urgent decompression
of septic arthritis of the hip and other joints or when
other drainage methods have failed.7
In our hospital, the decision to perform joint
washout for septic arthritis is made by an experienced
clinician; the diagnosis is based on a combination of
clinical, biochemical, microbiological, and imaging
parameters. This study reviewed the records of 50
children who underwent open joint washout for septic
arthritis with (n=25) or without (n=25) preoperative
antibiotics.
MATERIALS AND METHODS
Records of 50 children who underwent open joint
washout from 2003 to 2010 for presumed septic
arthritis with (n=25) or without (n=25) preoperative
antibiotics (within 7 days of joint washout) were
reviewed. Patients with a history of penetrating injury
were excluded. 17 boys and 8 girls aged 3 weeks to
16 years (median, 1.5 years) who were prescribed
preoperative antibiotics before joint washout were
compared with 12 boys and 13 girls aged one month
to 14 years (median, 2 years) who were not. Of those
prescribed antibiotics, 10 received one antibiotic, 13
received 2 antibiotics, and 2 received 3 antibiotics. A
total of 14 different antibiotics were used; the most
common were flucloxacillin (n=12), cephalosporins
(n=9), and clindamycin (n=4).
All patients were provisionally diagnosed with
septic arthritis by a consultant within 24 hours
of admission. Radiologic investigation included
radiography (n=47), ultrasonography (n=15) to
confirm joint effusion, magnetic resonance imaging
(n=8) to confirm osteomyelitis in one and effusion
in 7 patients, and computed tomography (n=2). The
most common site involved in the respective patients
who had or had not received preoperative antibiotics
was the hip joint (10+13), followed by the knee joint
(6+8), shoulder (2+3), ankle (4+1), and elbow (3+0).
Following arthrotomy and washout, patients were
commenced on high-dose intravenous antibiotics
(penicillin, flucloxacillin, and benzylpenicillin).
Antibiotics were altered if necessary according to the
microbiology and allergy result. When inflammatory
markers approached normal, patients were switched
to oral antibiotics that continued for 4 to 8 weeks.
Patients were followed up for 6 to 18 months until
asymptomatic and were advised to return to our
hospital if any symptoms recurred.
The Chi-squared test was used for categorical
data, and the unpaired t test for continuous parametric
data. A p value of <0.05 was considered statistically
significant.
RESULTS
After a median of 5.5 (range, 2–9) years, 20 patients
had completed the last follow-up. There was no
significant difference between patients who were
or were not prescribed preoperative antibiotics in
terms of age, co-morbidities, weight bearing status,
fever, positive culture, C-reactive protein, or white
cell count at presentation (Table), as well as the mean
(range) duration of antibiotic treatment (4.9 [4–7] vs.
4.7 [1–8] weeks, p=0.586).
In the respective patient groups, most were
referred from other hospitals (15+8), followed by the
emergency department of our hospital (4+14), fracture
clinics (2+1), general practitioners (2+2), and in-
patient ward (1+0). Patients who were referred from
places other than our emergency department were
twice as likely to have been prescribed preoperative
antibiotics (p=0.0032).
Patients prescribed preoperative antibiotics had
a longer median (range) time from symptom onset
to joint washout (8 [2–23] vs. 4 [1–29] days, p=0.05)
and a higher mean erythrocyte sedimentation
rate (93.1 vs. 54.3 mm/h, p=0.023) at presentation.
Five patients prescribed preoperative antibiotics
had 8 additional washouts, whereas 4 patients not
prescribed preoperative antibiotics had 4 additional
82 SBM MacLean et al. Journal of Orthopaedic Surgery

Table
Diagnostic parameter of septic arthritis at admission in patients with or without preoperative antibiotics
Parameter Preoperative antibiotics p Value
Yes (n=25) No (n=25)
Mean erythrocyte sedimentation rate (mm/h) 93.1 54.3 0.023
Mean C-reactive protein (mg/l) 127.5 88.0 0.177
Mean white cell count (x109/l) 15.7 12.7 0.075
No. of patients able to weight bear 12 16 0.254
Mean temperature (ºC) 37.1 37.2 0.380
No. of patients with positive culture 13 12 0.777
Staphylococcus aureus 8 5 -
Coagulase-negative Staphlycoccus 2 1 -
Group B Streptococcus 1 3 -
Group A Streptococcus 1 2 -
Fusobacterium necrophorum 1 0 -
Escherichia coli 0 1 -

washouts. In the 6 patients prescribed preoperative
antibiotics, complications developed were proximal
femoral osteomyelitis (n=1), chondrolysis (n=2), and
hip dysplasia requiring (pelvic or femoral) osteotomy
(n=3). In the 3 patients not prescribed preoperative
antibiotics, complications developed were slight
hip stiffness (n=1), chondrolysis (n=1), and subtle
avascular necrosis of the femoral head (n=1).
DISCUSSION
The use of preoperative antibiotics in the presence
of septic arthritis delays diagnosis and thus may
increase joint morbidity. Children with septic arthritis
are more likely to be managed appropriately if their
initial presentation is to a specialist paediatric unit.
Patients who present to a peripheral unit or a primary
care unit are more likely to be commenced on
preoperative antibiotics. This results in a significant
delay in surgical treatment and a higher number
of additional washouts and complications. The
complication rate is increased as hyaline cartilage is
exposed to a higher bacterial load for a longer time.
Antibiotics reduce or mask the clinical parameters
of children who present with a septic joint and do
not effectively reduce bacterial load within the joint.
Antibiotic use also often results in negative cultures.
Nonetheless, the number of positive cultures in both
groups was similar. Preoperative antibiotics appear to
have little synovial penetration or effect. Arthrotomy
and open washout therefore remains the treatment of
choice to clear the microbial load from the joint.
In our study, non-specialist units usually prescribed
antibiotics when the diagnosis was uncertain (pyrexia
of unknown origin) or upper respiratory or urinary
tract infection was suspected. Accurate diagnosis is
difficult in children in the presence of vague clinical
signs and medical history. Prodromal symptoms and
systemic illness occur in both transient synovitis and
joint sepsis. Defining the exact time of symptom onset,
particularly in very young children, is difficult. No
single clinical or microbiological test can conclusively
indicate septic arthritis.
One limitation of our study was the small
number of patients with positive microbiological
cultures. Gram stain14 and synovial white cell count15
have low sensitivity to predict septic arthritis. Rapid
polymerase chain reaction–based diagnosis has 95%
sensitivity and 97% specificity,5 but is not routinely
performed in our unit. In addition, the exact time of
commencement of preoperative antibiotics was not
always certain. Thus, the effect of the duration of
antibiotic use on outcome could not be determined.
Nonetheless, the extent of hyaline cartilage damage
is proportional to the duration of its exposure to
the bacteria. Furthermore, postoperative functional
outcome was not assessed, although the wide age
distribution of children and level of function would
have made such a comparison futile.
CONCLUSION
Preoperative antibiotics should be avoided in the
management of septic arthritis in children. Their
prescription delays diagnosis and definitive surgery,
and leads to additional washouts and complications.
A high index of suspicion and expedite referral to
a specialist paediatric orthopaedic unit is needed if
septic arthritis is suspected.
DISCLOSURE
No conflicts of interest were declared by the authors.
Vol. 23 No. 1, April 2015 Preoperative antibiotics for septic arthritis in children 83

REFERENCES

1. Staheli LT. Infection. In: Staheli LT, editor. Fundamentals of pediatric orthopaedics, 4th ed. Lippincott Williams & Wilkins;
2008:99–115.
2. Caird MS, Flynn JM, Leung YL, Millman JE, D’Italia JG, Dormans JP. Factors distinguishing septic arthritis from transient
synovitis of the hip in children. A prospective study. J Bone Joint Surg Am 2006;88:1251–7.
3. Luhmann SJ, Jones A, Schootman M, Gordon JE, Schoenecker PL, Luhmann JD. Differentiation between septic arthritis and
transient synovitis of the hip in children with clinical prediction algorithms. J Bone Joint Surg Am 2004;86:956–62.
4. Kocher MS, Zurakowski D, Kasser JR. Differentiating between septic arthritis and transient synovitis of the hip in children:
an evidence-based clinical prediction algorithm. J Bone Joint Surg Am 1999;81:1662–70.
5. Yang S, Ramachandran P, Hardick A, Hsieh YH, Quianzon C, Kuroki M, et al. Rapid PCR-based diagnosis of septic arthritis
by early Gram-type classification and pathogen identification. J Clin Microbiol 2008;46:1386–90.
6. Matthews PC, Dean BJ, Medagoda K, Gundle R, Atkins BL, Berendt AR, et al. Native hip joint septic arthritis in 20 adults:
delayed presentation beyond three weeks predicts need for excision arthroplasty. J Infect 2008;57:185–90.
7. British Society for Children’s Orthopaedic Surgery. The management of acute bone and joint infection in childhood. A guide
to good practice. Available from http://bscos.org.uk/resources/Downloads/Blue-Book-Infection.pdf.
8. Nade S. Acute septic arthritis in infancy and childhood. J Bone Joint Surg Br 1983;65:234–41.
9. Shaw BA, Kasser JR. Acute septic arthritis in infancy and childhood. Clin Orthop Relat Res 1990;257:212–25.
10. Kocher MS, Mandiga R, Murphy JM, Goldmann D, Harper M, Sundel R, et al. A clinical practice guideline for treatment of
septic arthritis in children: efficacy in improving process of care and effect on outcome of septic arthritis of the hip. J Bone
Joint Surg Am 2003;85:994–9.
11. Wall EJ. Childhood osteomyelitis and septic arthritis. Curr Opin Pediatr 1998;10:73–6.
12. Wilson NI, Di Paola M. Acute septic arthritis in infancy and childhood. 10 years’ experience. J Bone Joint Surg Br
1986;68:584–7.
13. Paakkonen M, Peltola H. Management of a child with suspected acute septic arthritis. Arch Dis Child 2012;97:287–92.
14. Faraj AA, Omonbude OD, Godwin P. Gram staining in the diagnosis of acute septic arthritis. Acta Orthop Belg 2002;68:388–
91.
15. McGillicuddy DC, Shah KH, Friedberg RP, Nathanson LA, Edlow JA. How sensitive is the synovial fluid white blood cell
count in diagnosing septic arthritis? Am J Emerg Med 2007;25:749–52.