NEONATAL

OSTEOMYELITIS

C.

J.

M.

KNUDSEN,

E.

B.

HOFFMAN

From Thirty-four

the Red

Cross

War

Memorial

Childrens

Hospital,

Cape

Town common site involved. abnormalities, such as in 18 of the 19 hips surgically. Good

Swelling
metaphyseal involved. results

neonates with osteomyelitis were reviewed. The hip (19) was the most and pseudoparalysis were the most significant local signs. Radiographic rarefaction and/or joint subluxation were found on the initial radiographs with antibiotics and all infections of all sites and in 68% of hips. involvingjoints were

All patients were treated were achieved in 75%

drained

Bone andjoint sepsis in the first month oflife is rare. The diagnosis is difficult and often delayed as the clinical features differ significantly from infections occurring in older children. are high but permanent bone at risk. We describe of the disease compare our literature. In the post-antibiotic the survivors are and joint damage. era, survival frequently left The hip rates with

Table I. neonates

Predisposing

factors

in

22

Number Prematurity Skin/umbilical sepsis 7 7 6

is especially

the clinical and which facilitate results with

radiographic features early diagnosis and those reported in the

Delivered by Caesarian section (1 premature) Significantjaundice (2 premature) Pneumonia Meningitis

long-term

2 1

PATIENTS

AND

METHODS

Between 1977 and 1987, 34 neonates < 28 days old) were treated for osteomye!itis at the Red Cross War Memorial Childrens Hospital, an average of three per year. Infection was confirmed by a positive blood culture, by the finding of pus at surgery or by radiographic changes such as metaphyseal rarefaction or periosteal reaction. patients followed average The were follow-up Twenty-seven clinical analysed records and for radiographs and at least (range reviewed of all these all were The by 12 months. 1 to 12 years). personally retrospectively

FINDINGS Age at presentation. (range The average None age at presentation presented were present in the first was week

23 days
of life.

8 to 28). factors.

Predisposing

These

in 22 patients findings, two of 26 neonates Of slight pyrexia. alerted to the for the delay

(65%) and are listed in Table I. Clinical features. Based on the genera! groups were identified. Group I consisted who were not acutely ill at the time these, 24 were In 1 3 patients, severity of the in presentation who Group were apyrexia! and two had only the parents had not been condition, which which averaged accounted 10 days.

up as out-patients

of presentation.

was 43 months patients were

the authors. The other seven patients, at review, were assessed by reference records and radiographs.

not personally seen to their clinical

II consisted of the remaining eight acutely ill at initial presentation.

neonates Six were

FCS (SA) Orth, Orthopaedic Consultant E. B. Hoffman, FCS (SA) Orth, Orthopaedic Consultant Department of Orthopaedic Surgery, University of Cape Red Cross War Memorial Childrens Hospital, Rondebosch, ofSouth Africa. Correspondence

C. J. M. Knudsen,

Town and Republic at 7 Marne Africa.

septicaemic and two had meningitis. In this group, attention was so focused on the systemic illness that the subtle local abnormalities were frequently overlooked. There was an average delay of five days before the diagnosis was made. and Subtle, local abnormalities are listed in Table II. The were found in every case most frequent findings (Fig. 1) which affected

Avenue,

Newlands,

should be sent to Dr E. B. 7700, Cape Town, Republic and Joint

Hoffman of South Surgery

1990 British 0301-620X/90/5l47

Editorial Society ofBone $2.00 JBoneJointSurg[Br] 1990; 72-B:846-51.

were swelling and pseudoparalysis 95% of involved limbs.

846

THE

JOURNAL

OF BONE

AND

JOINT

SURGERY

NEONATAL

OSTEOMYELITIS

847

Proximal

humerus/shoulder

.8

Distal
.

humerus

.1
.2

Olecranon/elbow

Proximal

femur/hip

. 19

Distal

femur/knee

.6

.4

Proximal

tibia

.5

L
Neonate shoulder

Os calcis/subtalar

.1

Fig. with infection and pseudoparalysis

1 humerus! Fig. Sites of sepsis 2 in 34 neonates.

in the left proximal of the left arm.

Table II. 34 neonates Finding Pseudoparalysis Local Pain

Local

abnormalities

present

in

Number* 26 25 movement 20 12 was detected had either in

swelling on passive posture one and abnormality 95%

Abnormal #{149}at least every case,

of patients swelling

pseudoparalysis

and/orlocal

Table III. of infection

Radiographic at different

abnormalities sites Patients positive findings with

noted

at initial

presentation

Sites Proximal femur/hip

Findings 14 subluxed 12 metaphyseal 8 both ofthe 4 subluxed 4 metaphyseal 2 both ofthe 4 metaphyseal

Fig.

3 There between is metaphyseal the ischium and

18 of 19

rarefaction above

Infection of the left proximal femur/hip. rarefaction of the proximal femur and a gap the proximal femur of more than 5 mm.

Proximal shoulder

humerus/

6 of 8

rarefaction above rarefaction

Sites.

There

Distal Proximal

femur/knee tibia

4 of 6 5 of 5

4 metaphyseal rarefaction 3 periosteal reaction 2 both ofthe above 1 metaphyseal 1 metaphyseal 1 metaphyseal rarefaction rarefaction rarefaction

the most multiple

were 42 sites of infection in the 34 neonates, frequent being the hip (45%). In six cases (18%) sites were involved (Fig. 2).

Distal

humerus

1 of I 1 of 2 1 of 1

Olecranon/elbow Calcaneum/subtalar

Radiographic changes. Abnormalities such as subluxation, metaphyseal rarefaction or periosteal reaction were noted on the initial radiographs of 34 of the 42 sites of infection (8 1 (Table III). Of the 19 hips, 18 were

%)

subluxed

proximal

and/or femur

showed metaphyseal at presentation (Fig.

changes in 3). Subluxation

the of

VOL.

72-B, No. 5, SEPTEMBER

1990

848

C. J. M. KNUDSEN,

E. B. HOFFMAN

the

hip

was the

defined proximal

as

a gap femur

of and

greater ischium

than on

5 mm an AP

between

infection. There the staphylococcal

were no gram-negative infections, five (30%)

infections. Of were resistant

radiograph ofthe pelvis placed (Bertol, Macnicol

taken and

with the legs symmetrically Mitchell 1982).

Bacteriology. An organism was cultured from blood or pus in 25 patients (74%). Staphylococcus aureus, the most common, was grown in 18 patients (72%). fl-haemolytic streptococcus, the only other infective agent, was cultured in seven patients (28%). Of these, one patient had group A and six had group B JJ-haemolytic streptococca!

to c!oxaci!lin but sensitive to fucidic acid. These all occurred in premature neonates who had spent some time in hospital. Management. Al! patients were admitted to hospital and treated with intravenous cloxacil!in (200 mg/kg/day) for a minimum antibiotics therapy of 48 hours, were continued was later adjusted following blood culture. Oral for six weeks. according Initial to the antibiotic sensitivity

Fig. Figure 4a - Bilateral left. Pus was found group I). involvement at operation

4a with bilateral subluxation Both hips showing excellent

Fig.

4b femur Salvati on the 1978,

of the proximal femur/hip, in both hips. Figure 4b

and metaphyseal rarefaction of the proximal result at 46 months follow-up (Hallel and

::-T-

Radiographs 50 months after infection the left femoral neck (Hallel and Salvati

of the left hip reveal 1978, group II).

varus

of A destroyed (Hallel and left femoral Salvati 1978,

Fig.

6 34 months after infection

capital epiphysis, group III).

Table Hallel Number patients 13

IV. and of

Outcome in Salvati (1978)

19 hips,

using

the

classification

of

Group I

Findings Normal or some capital epiphysis loss of roundness head, neck

reports fucidic cus).

on the organism (penicillin acid for cloxacillin-resistant 100 mg/kg/day patients with for osteomyelitis

for the streptococcus; staphylococcus; sensitive not staphylococinvolving a

fluc!oxaci!lin Four
2 II Deformed, small in varus position Destroyed capital

joint responded no operation. surgical drainage

to early appropriate antibiotics and had The remaining 30 patients underwent open ; in no case was aspiration performed.

III

epiphysis

THE

JOURNAL

OF BONE

AND

JOINT

SURGERY

NEONATAL

OSTEOMYELITIS

849

Figure complete

7a

Four

months

after

osteomyelitis 58 months

of the later.

proximal

tibia

there

is anterior

bowing

of the

tibia.

Figure

7b

There

is

restitution

at follow-up

Table

V.

Outcome

at sites

other

than

the hip

Site Proximal humerus/shoulder

Number ofpatlents 6 2 6 5 1 I 1 I

Findings Normal epiphysis/shoulder epiphysis femur/knee tibia

Destroyed
Normal Normal

Distal Proximal

femur/knee tibia

Olecranon/Clbow

Normal olecranon/elbow Dislocated radial head Normal humerus talus

Distal

humerus

Os calcis/subtalarjoint

Flat topped

were

treated

by abduction

splintage

for

a further

three

months ; six of them relocated capital epiphyses ; the four that had destroyed capital epiphyses. Outcome. There were no deaths

and developed normal remained dislocated all in this series. The 19

affected hips were assessed using the classification of Hallel and Salvati (1978) and their long-term results are given in Table IV (see Figs 4 to 6). Thirteen hips (68%)
Figure 8a - Infection from thejoint. Figure 39 months later. of the right 8b - Minimal distal femur/knee metaphyseal
; pus changes

was drained at follow-up

were Table

assessed

The fate
V. Ofthe The sequelae.

as excellent. of infections at all other shoulder infections, 75% two distal shoulders femur

sites is given in resolved without

The hip approach. The

was

always were weeks at the

approached splinted

via

an

anterior

Luck

retained
tions of completely deformities and 8).

nearly
the

full movements

with destroyed epiphyses at follow-up. All infecand proximal tibia resolved

19 hips

in abduction

for an initial 10 (53%) were period. These

period of four still dislocated

following drainage; end of the four-week

with no leg length and normal or near

norma!

discrepancy, no epiphyses

angular (Figs

VOL.

72-B, No. 5, SEPTEMBER

1990

850

C. J. M. KNUDSEN,

E. B. HOFFMAN

Septic

arthritis

of the

knee

occurred

in four in none

of the of the

six cases of distal femoral infection but cases of infection of the proximal tibia.

spasm and subluxation pus

this combination could account of the hip in those cases where regard lateral clue that subluxation we suggest

for lateral there is no as such that an every

in the joint. We now

DISCUSSION
Greengard clinical form, form where where (1946) syndromes the child the child in the was the first to recognise osteomyelitis, two distinct in neonatal a benign

important

diagnostic

septicaemic neonate exclude occult hip practice involving because simultaneous to subject the lower the local sepsis

should have pelvic sepsis. In addition, all neonates limb to a routine signs in the hip

radiographs to it is our current

is not systemically is septicaemic and disease process. and by Dennison

ill, and a severe the osteomyelitis grouping

(1954).

with osteomye!itis pelvic radiograph, may be masked by

is an incident

A similar Potter
(1955)

was found by Hutter (1948) Thomson and Lewis (1950) and out that appearance satisfactory the

Pathology.
sites showed

pointed

at a distal site in the ipsilateral limb. With the exception of one knee, all infected radiographic signs of primary osteomyelitis rarefaction, periostea! changes subsequently. This of Ogden is the and Lister (1975), primary pathology

word benign only relates to the clinical of the child at presentation and is hardly a description of a condition which may lead to

at presentation (metaphysea! reaction) or developed these is similar confirming to the findings that osteomyelitis sepsis Ob!etz

permanent disability. Our findings support their opinion. In our patients who were systemically ill at presentation (group 2), all eight survived and five recovered without permanent were not suffered bone orjoint systemically permanent damage. Ofthe ill at presentation orjoint damage. there is as the 26 children who (group 1), five

and that joint Bacteriology.

developed secondarily. 1 960), Weissberg et a!

( 1974),
two

Hallel

and Salvati (1978) and Bergdah! a predominance of staphylococca! sites with bad results and in our series one elbow, staphylococca! streptococcal difference organism.

et al (1985) infections. (six hips,

all reported Of the 10 shoulders,

bone

The significance often delay in diagnosis attention ofthe doctor

of the two groups is that of the severely ill patients,

Diagnosis.
emphasised

The importance
by others

is not focused on the osteomyelitis. of early diagnosis has been 1 960 ; Lloyd-Roberts 1979).

one subta!ar joint), nine were due to infections and one was due to a Group A infection. Bergdahl et a! (1985) found no in the sequelae after infection by either

(Obletz

Isotope scanning is usually unreliable in the neonate (Ash and Gilday 1980; Bergdahl, Ekengren and Eriksson 1985), so radiographic (1952), (1974), pinching diagnosis depends abnormalities. on the subtle local and We found, as did Blanche

Management and outcome. The dramatic decline in the mortality from neonatal osteomyelitis during the postantibiotic era is a reflection of the efficacy of antibiotics in the control ofsepticaemia. cloxacillin in high doses initial therapy, except for been previously hospitalised. of a high incidence this selected group, We believe that intravenous (200 mg/kg/day) is the best the premature baby who has On the basis of our finding their bacteria treatment (1946), that surgical in

Obletz

(1960)

and Weissberg,
and pseudoparalysis limb) remained

Smith

and Smith

that swelling the involved the joint and

(confirmed by the most significant

signs. Unlike the infected in the neonate

older child with joint sepsis, in whom is held rigidly immobile, it is possible infant to obtain a reasonable range of

of cloxacillin-resistant we currently start did

with

fucidic

passive movement, despite sepsis. It is essential that this should not be misinterpreted. The white cell count was not of any assistance in the diagnosis, but the sedimentation rate was helpful (average 72 mm/hr Westergren). helpful in hips, 18 showed lateral subluxation and/or metaphyseal rarefaction on the initial radiographs. This is unlike the older child in whom radiographic changes at presentation are the exception rather than the rule. Hutter (1948) was the Radiographic features were extremely

We
Wilkinson

do

acid (30 mg/kg/day). not believe, as (1952) and Nicholson

Greengard (1960),

confirming

the

diagnosis.

Of the 19 infected

drainage of pus should be limited. Antibiotics cannot replace surgical drainage. We concur with Blanche (1952), Obletz (1960) and Lloyd-Roberts (1979) that the hip should be decompressed by adequate open arthrotomy. Ob!etz (1960) aspirated the hip as a diagnostic procedure, but this has been shown by Paterson (1970) to be unreliable and is not recommended. The outcome at sites other than the hip is shown in Table V. A favourable outcome of infections involving the tibia and femur/knee was reported by Trueta (1959) and by Lloyd-Roberts (1960). Angular deformities of the tibia! shaft (Fig. 7a) corrected with time (Fig. 7b) and no patient developed chronic osteitis. Involvement of the hip produces the most serious sequelae (Lloyd-Roberts 1960). Our series of 19 hips is the largest we could find in the literature. Previous authors failure have reported poor (Bergdahl et a! 1985)
THE JOURNAL

first to describe lateral displacement (1960) found this sign to be present Bergdahl et a! (1985) found it in most

of the hip, Obletz in all cases and cases of hip sepsis.

The early
explained

metaphyseal by the softer

rarefaction in the neonate may be bone of this age group, while the

lateral subluxation may be explained by distension of the capsule with pus. Lloyd-Roberts (1979) postulated that in the presence of a metaphyseal infection, the surrounding tissues, including the capsule and may hip flexors become develop lax and reflex

oedematous.

The adductors

results to 80%
OF BONE

ranging (Hallel
AND

from 50% and Salvati

JOINT SURGERY

NEONATAL

OSTEOMYELITIS

851

1978).

Our

patients, better. antibiotic We

with dosage,

68% our

excellent good

results, results surgical are

did due drain-

splintage of the three the

is, we believe, capital months epiphysis. after common

highly Further

suggestive splintage

of destruction for more than oflittle value. prematurity was ; all five having Kuo response infected et a! in hip

significantly to adequate

believe

immediate for

age and appropriate Obletz (1960) between (1979) hip and tion three emphasised

splintage. used splintage and 18 importance nine after

Of the 10 bad
a period varying the most months. Lloyd-Roberts of immobilising and were four some cloxacillin-resistant (1975) have shown premature babies. Conclusions. We

drainage is probably results in this series, associated factor staphylococcal an immature that the

weeks the position

infections. immune fate of the

in a stable radiographically

of abduction located

flexion.

Of the 19 hips in our series,

clinically stable weeks of abduc-

believe

splintage and all nine developed normally. Ten hips (53%) remained dislocated after four weeks of splintage and these were all treated by some form of abduction subsequently unstable been splint were destroyed. for a further became stable. those This in which finding (1978) who or moderate three The the months. Of those, six four that remained capita! epiphysis with that that dislocated head destruction by conservative infection. Failure of appropriate had of

is not sealed from the outset. Awareness of this and a knowledge ofthe clinical and radiographic

condition findings

at presentation make early diagnosis possible. Adequate antibiotic therapy and immediate open drainage is the best method to preserve the femora! capital epiphysis. Splintage of the frankly dislocated hip should be result continued for at least three months may be anticipated if the epiphysis because a good is not destroyed.

is in accord reported femoral

Hallel and Salvati hips with minimal have means a good within

chance of being reduced the first two months after hip after three months

to achieve

a stable

We would like to thank Mr Martin Singer for his help in preparing the script, Vera Barrow for the typing and Michael Wyeth for the illustrations. This study was supported by a grant from the Medical Research Council of South Africa. No benefits in any form have been received or will be received from a commercial party related directly or indirectly to the subject of this article.

REFERENCES Ash JM, Cilday osteomyelitis. Ekengren DL. The futility 1980; of bone 21 :417-20. scanning hematogenous features in neonatal osteoof neonatal 1982; 641952; 1955;

Lloyd-Roberts
1 5(Special

CC. Septic
Issue) :41-3.

arthritis In : Oblitz period.

in infancy. BE. Acute

Aust

Paediatr

J 1979;
arthritis 1960; period. Pediatrics of 42-

J Nuci Med

Bergdahl 5, myelitis.

J Pediatr

K, Eriksson M. Neonatal Orthop 1985 ; 5:564-8.

Bertol

P, Macnicol MF, Mitchell CP. Radiographic congenital dislocation of the hip. J Bone Joint B :I76-9.
in infants. osteitis

Nicholson JT. Discussion. the hip in the neonatal A:30.

suppurative

J Bone Joint
of the :23-30. hip

Surg [Am]
in neonatal

Surg [Br] Surg

Obletz BE. Acute suppurative arthritis Bone Joint Surg [Am] 1960; 42-A Ogden JA, Lister C. The pathology
1975; 55:474-8. DC. Acute suppurative 1970; Paterson

Blanche DW. Osteomyelitis 34-A :71-85.

J Bone Joint
in the newborn. in

[Am]
Lancet

of neonatal

osteomyelitis.
and

Dennison

WM.

Haematogenous hematogenous 1946; 30:135-45.

ii :474-6. Creengard J. Acute NorthAmerica Hallel Hutter Kuo osteomyelitis hip infancy. : end

Bone Joint Surg [Br] M C/in

result

arthritis in infancy 52-B :474-82. new-born. in new-born.

childhood.

J [Br]
1950;

Potter

CMC.
1954; 36-B

Osteomyelitis
:578-83.

in the

J Bone

Joint

Surg

T, Salvati EA. Septic study. C/in Orthop 1978;

arthritis of the 132:1 15-28. of osteomyelitis Orme joint

in infancy newborn

Thomson J, Lewis 25:273-9.

IC. Osteomyelitis types

Arch Dis Child

CC Jr. New concepts Pediatr 1948; 32 :522-9.

KN, Lloyd-Roberts and infantile bone 5 1-6. CC, and

in the

infant.

Trueta

J. The
clinical 671-80. and

three

vascular Smith

study.

of acute haematogenous J Bone Joint Surg

Smith 1974;

osteomyelitis : a [Br] 1959; 41-B: of neonatal

in infants.

IM, Soothill JF. Immunodeficiency infection. Arch Dis Child 1975; arthritis of infancy.

50:

Weissberg ED, osteomyelitis.

AL, Pediatrics

DH.

Clinical
DW. :84-5.

features

53 :505-10. Osteomyelitis

Lloyd-Roberts CC. Suppurative [Br] 1960; 42-B :706-20.

J Bone Joint

Surg

Wilkinson FR. Discussion. In : Blanche J Bone Joint Surg [Am] 1952; 34-A

VOL.

72-B,

No. 5, SEPTEMBER

1990