J. E. Hall, M.D., F.R.C.S.(C), and E. A. Silverstein, M.D., F.R.C.S.(C) The Hospital for Sick Children, Toronto
(Submitted April 24, 1962; accepted for publication June 20, 1962.)
ADDRESS: (J.E.H.) Medical Arts Building, Toronto, Ontario, Canada.
PEDIATRICS, June 1963
ACUTE
HEMATOGENOUS
OSTEOMYELITIS
1033
T
WENTY YEARS after the introduction of antibiotic therapy, it is obvious that the high hopes that were held for the complete eradication of osteomyelitis have not been fulfilled. The disease process has been modified considerably, and the earlychanges were all encouraging, but in the past few years less favorable trends have appeared.
To illustrate the first great change one can look at the statistics for osteomyelitis admissions to the Hospital for Sick Chil-dren, Toronto, from 1920 to 1930. They con-stituted 0.72% of all admissions, and the mortality rate was 20.7%. Between 1940 and 1950 the percentage of admissions fell to 0.19% and tile mortality rate to 3.6%.’
During tilis period there was a tendency to view osteomyelitis \vitil some com-placency, and sound physiological and sur-gical principles were overlooked. The
re-suit of this has been an increase in the rela-tive number of patients with osteomyelitis
requiring hospital admission and in the number going on to chronic
osteomyeli-3 Although some of the blame can be
placed on emergence of resistant strains of bacteria, some of it must fall on the meth-ods of treatment which came to be ac-cepted.
There are only two fundamental prin-ciples involved; one is selecting the proper antibiotic, and the other is seeing that it is able to get to the area in which it is re-quired. The first principle is obvious, and
tile development of newer forms of
peni-ciliin and broad spectrum antibiotics has so far enabled us to keep up with bacterial mutations in most cases. To emphasize the second principle it is necessary to consider the pathogenesis of acute hematogenous osteomyelitis.
When bacteria lodge in the cancellous area of the metaphysis of a long bone the usual process of acute inflammation begins. As the inflammatory exudate collects and the body attempts to wail it off, the phe-nomenon we recognize as swelling in the soft tissues results in greatly increased pres-sure in the rigid confines of bone. This in-creased pressure forces exudate down ha-versian canals and into every vascular chan-nd, rendering the bone in the area com-pletely avascular and resulting in its death. If the process is not arrested the exudate is forced out under the periosteum, which is lifted from the shaft of the bone, further reducing the local vascularity and resulting
in a subperiosteal abscess.
Visualizing this process, it is easy then to understand that unless the correct antibiotic in adequate dosage is supplied very early, before the involved area is sealed off from tile circulation, there can be very little chance of success with antibiotic therapy alone. It is also evident that to persist with conservative therapy when it is not control!-ing the disease, is futile, and some method of decompressing the involved area must be used. Equally futile is the system of changing antibiotics after it seems obvious that the first one has failed, since by then it will not be able to reach thc area in which
it is required.
HOSPITAL FOR SICK CHILDREN SERIES
TABLE I
CLASSIFICATION OF IATIENTS WIT!! DIAGNOSIS OF
OSTEOMYELITIS AT TILE lIOSPITAL FOR SIcK
CHILDREN, FROM .IANUARY, 1956, TO
JUNE, 1958
Cases
Diagnori8 (no.)
All cultures
Positive Negative
Cultures before antibiotics
Positive Negative
41
37
36
21
Chronic osteomyelitis I 15
Clinically diagnosed autc osteoniyelitis, no 33
x-ray proof
Acute hematogenous osteoiiiyelitis
Total 250
cases of osteomyeiitis of other bones were
found; of these, 115 were diagnosed as chronic osteomyelitis on admission.
Thirty-three patients had high fever,
tox-emia, pain in one limb, and marked tender-ness near the end of a long bone, and were diagnosed clinically as having acute hema-togenous osteomyelitis. Tiley were all seen within 36 hours of the start of their symp-toms, were given large doses of broad spec-trum antibiotics, and all responded dra-matically. Since no direct evidence, x-ray or otiler\vise, was obtainabh to substantiate the diagnosis, they were excluded from this series, but may represent osteomyelitis aborted by early effective treatment.
Two of the remaining 102 children died of fulminating septicemia, leaving 100 pa-tients with proven acute hernatogenous Os-teomyelitis. The age of these children ranged, without obvious pattern, from 2 weeks to 14 years. Nineteen were less than 2 years of age.
TABLE IL
RESULTS OF BLOOD CULTURES
Result \‘u,nber Per Cent
Blood Cultures (Table II)
Blood cultures were attempted in 78 pa-tients, \Vitil 41 positive resuits-an average of 53%. In tile 57 patients wilose cultures
were taken before any antibiotic had been given (in hospital or previous to admission)
tile percentage of positive cultures was 65%. When the antibiotic had been given before tile patient was sent to tile ilospital, there was less chailce of obtaining a positive
cul-ture.
Organisms and Sensitivities (Table III) Niost infections were caused by Staphy-lococcus pyogenes (59 of the 68 in which cultures were obtained). Only 28 of these
were sensitive to penicillin at tile level at winch it is teste(1 in our laboratory; 18 were classified as resistant and the other 10 as partially resistant.’
Hemolytic streptococcus was responsible for seven infections (all penicillin sensitive) and nonpyogenic staphylococcus was found
twice.
Since methycillin was not obtainable at that time, sensitivity studies to it were not carried out. The most effective combination was found to be erythromycin and chlor-amphenicol, and these drugs were used for most patients at that timi Because pro-longed therapy was sometimes required, in more recent cases novobiocin has been sub-stituted for chioramphenicol to prevent late hemopoietic complications.
Interval between Onset of Symptoms and
Beginning of Effective Treatment
The period of time whicil had elapsed from tile onset of symptoms to the start of
_____ effective antibiotic therapy was one of the
keys to the problem. The first 3 days were tile critical period, and as time
in-53 creased so did tile difficulty in treating the
47 infection and the likelihood of chronic os-teomyelitis.
The 100 patients were divided into
groups based on the above mentioned time
TABLE III
OIsaNIssts AND SENSITIVITY TO P:NIcILLIN
1035 Organism Positive Cultures (no.) Pyogenic Staphylococcus P(IIi(illiII sensitive Penicillin resistant Partially resistant llensolytic Streptococcus Penicillin sensitive Non-pyogenic Staphylococcus 59 28 10 18 7 2
days, 4 to 7 days, and over 7 days were used.
LESS THAN THREE DAYS (Table IV) : When
patients were seen in early stage of illness 87% were successfully managed. Nearly half required no aspiration or drainage since tiley responded quickly to conservative
tilerapy.
This group consisted of 54 patients, all treated with tilerapeutic closes of either penicillin or broad spectrum antibiotics and supportive therapy as indicated. Their progress was watched, and if at the end of 24 hours their temperature had fallen con-siderably, their pain was gone, and tileir tenderness was much less, the conservative therapy was continued. This occurred in 29
(
54%), of whom 27 were apparently cured(
up to 2 years); 2 went on to chronic os-teomyelitis.When the response was not dramatic after 24 hours, needle aspiration of the
sub-TABLE IV
DATA ON PATIENTS SEEN DURING TilE FIRST
THREE Divs OF ILLNESS
I
Therapy
Response
Success
----
Failure-
TotalAntibiotics alone
Antibiotics plus needle
aspiration
Antibiotics plus surgical
drainage Total 27 9 11 47 2 2 3 7 29 11 14 54 TABLE V
1)ATt ON I’ATIENTS SEEN FROM FOUJITII TO
SEVENTH l)&y OF ILI.NESS
Therapy
Response
---Success Failure Total
Antibiotics alone
Antibiotics plus nce(lle
iISI-ration
Antibiotics plus surgical (Irainage Total (1 6 7 19 ‘2 0 1 3 8 6 8 22
periosteal space was carried out, and tile patient was watched for approximately an-other 12 ilours. If the response was then favorable, no further intervention was con-sidered. Eleven patients who had needle aspiration showed a dramatic response within 12 ilours, and nine of these were apparently cured, another two going on to cilrOnic disease.
If after 36 hours and a needle aspiration, the local and general conditions were not markedly improved, open surgical drainage was performed. Fourteen patients were op-erated upon, with 11 successes and 3 fail-ures.
The cortex was not drilled in all cases of open drainage, especially when a large sub-periosteal abscess was encountered, since it was felt that the bone had probably already decompressed itself. Primary closure with-out drainage was used in almost all. In more recent cases continuous suction using the “Hemovac” system llas been considered a useful adjunct during the first 2 post-operative days, decreasing postoperative pain and swelling.
FOUR TO SEVEN DAYS (Table \7) : Patients
seen after 3 days were more likely to re-quire aspiration and/or surgical drainage,
but if this was done, a good measure of sue-cess could be expected. Almost the same proportion of apparently cured infections appeared in tilis group as in the one above, but a much larger proportion had needle
1036
eight cases did the initial symptoms subside on antibiotic therapy, and two of these pa-tients went on to develop chronic osteomye-litis. When patients in this category were admitted to tile orthopaedic service, aspira-tion was performed shortly after admission. In six patients this was all that was re-quired, and none of these six developed chronic disease (tip to 2 years). Six of the seven patients who were operated upon and drained had primary closure without drainage. None went on to chronic osteo-myelitis. The only residual infection in the operated group occurred in a tibia which was packed open for 10 days before see-ondary suture.
Moiu THAN SEVEN DAYS (Table VI) : This
group was composed of 24 patients, and the use of antibiotics alone was not very successful. Tile percentage of chronic in-fections was 54%, and in the 46% which were successfully managed, about two-thirds re-quired surgery.
The 12 patients who had antibiotics alone accounted for eight cases of chronic osteo-myelitis. Two more were found among the
four who had needle aspiration; and of the eight who had surgical drainage, three de-veloped chronic disease.
At this late stage in the disease, early
surgery provides the best chance for
sue-cess, although success is not very certain.
Age
The average age of children who had a successful outcome was 6.2 years. The
aver-TABLE VI
DATA ON PATIENTS SEEN 7 DAYS OR MORE AFrER ONSET OF SYMvrosIs
Therapy
Response
Success Failure Total
Antibiotics alone
Antibiotics plus needle aspi-ration
Antibiotics plus surgical
drainage
Total
4
2
5
11
8
2
3
13
12
4
8
24
age age of those in whom the treatment failed to prevent chronic osteomyelitis was
6.3 years. When the various age groups were
studied, treatment appeared no less likely to succeed at one age than another except that in the very young children, early diagnosis was less apt to be made. More of these chii-dren thus necessitated more radical treat-ment, with less certain results.
COMMENT
There have been many papers on this subject in the medical journals during the past few years, and nearly all have pointed out the lessons which are emphasized in this series.4”
The diagnosis must be made early and with certainty. Osteomyelitis must be sus-pected when a child complains of pain in an extremity, has a fever, is toxic, and has local tenderness near a joint. Less will be lost by waiting an additional 12 hours to be certain of the diagnosis than by giving in-adequate treatment. A single injection of penicillin will obscure the constitutional symptoms but will have no effect on the local disease, and this type of treatment is to be discouraged.
Since half of the eases in this series were due to organisms which were partially or completely resistant to penicillin, this does not seem to be the drug of choice in the treatment of acute hematogenous osteomye-litis (with the possible exception of the
newer forms of penicillin). If penicillin is used, while the result of the culture is awaited, the opportunity for aborting the disease will be gone, and it would seem more logical to start with a broad spectrum antibiotic, changing to penicillin, if mdi-cated, after the blood culture was re-turned.”
The patients who were treated on anti-biotics alone, and who failed to exhibit a good clinical response within the first 24 to
as-1037
piration (and in some cases operation)
in-tervened, then the percentage of failures was much less. There is, however, no real
proof
of the role of aspiration, since thenumbers are small and it must remain a clinical impression.
Local antibiotics were instilled at the time of aspiration in each case, so it is not possible to assess their role with accuracy, nor to separate the respective merits of the aspiration and the local antibiotic.
CONCLUSIONS
On the basis of this series, and of recent reports in the literature, the following would seem to be a reasonable regimen for the treatment of acute hematogenous osteo-myelitis.
Any child who has a fever, looks toxic, and has some local tenderness near the end of a long bone, must be suspected of hay-ing osteomyelitis. In the event of real doubt
as to diagnosis, wait about 12 hours in-stead of beginning inadequate treatment
which might then obscure the clinical pie-tore. Once the diagnosis is made, full treat-ment should begin with immobilization, in-travenous fluids as necessary, and full doses of broad spectrum antibiotics. A blood cul-ture should be taken before antibiotics are given.
The patient should then be watched closely during the next 24 hours for his clinical response. If at the end of that time his fever is not considerably lower and the local tenderness very much less, either as-piration or open drainage should be per-formed. If the clinical response is good and if the blood culture, when it is ob-tamed, indicates that the organism is sensi-tive to penicillin, then a change can be made to this drug. If the decision is to
as-pirate, and pus is obtained from the sub-periosteal space, local antibiotics such as bacitracin, 10,000 units, should be instilled and the clinical response watched for an-other 12 hours. If the fever remains ele-vated, or local tenderness is marked after
this
period, then open drainage should be undertaken.Decision as to whether or not to drill the cortex is an individual one. It was not felt
on the basis of this series that drilling, or the lack of it, had very much bearing on the eventual outcome, except in early cases where there was not much of a subperios-teal abscess. Drill holes were then made and pus was found under tension in the medullary cavity. When there was a large subperiosteal abscess and drilling was un-dertaken, there did not seem to be any pres-sure in the medullary cavity, indicating that it had already decompressed itself. Pri-mary closure should be undertaken with or without drainage. (Present preference is for closed suction because of decreased post-operative pain and swelling.)
If the clinical response has been good and no operation is necessary, antibiotics, should be continued for an arbitrary period of 3 weeks and then either discontinued or carried on, depending on the clinical find-ings at that time, and to some extent on the sedimentation rate and tile x-ray. If opera-tion has been done, antibiotic therapy should be carried on for 3 weeks from the time of the operation; decision to stop should be based upon the findings at that time.
SUMMARY
laid on tile fact that the diagnosis must be
made on clinical, not x-ray, findings. On tile basis of the recent literature and the
find-ings in this series, a rational scheme of
ther-apy for acute hematogenous osteomyelitis
has been suggested.
REFERENCES
1. Creen, M., Nyhan, W. L. Jr., and Fousek
M.D. : Acute hematogenous osteomyelitis.
PEDIATRICS, 17:368, 1956.
2. hung, \V., and McCavisk, D. F. : Acute hematogenous osteomyelitis : a report of 36 cases seen at Children’s Hospital 1950-58. Clin. Proc. Children’s Hosp., Wash-ington, D.C., 16:16:3, 1960.
:3. Winters, J. L., and Cahen, I. : Acute hema-togenous osteomvelitis : a review of sixty-six cases. J. Bone Joint Surg., 42A:691, 1960.
4. Shandling, B. : Acute haematogenous
osteo-myelitis : a review of 300 cases treated during 1952-1959. South African Med. J., 34:520, 1960.
5. Clarke, A. Ni. : Neonatal osteomyclitis: a dis-ease different from osteomvelitis of older children. Med. J. Aust., 1(8):237, 1958. 6. Kessel, A. W. L. : Acute osteomyclitis. Brit.
Med. J., 1:1352, 1956.
7. Harris, N. H. : Some problems in the diag-nosis and treatment of acute osteomyelitis.
J. Bone Joint Surg., 42B:535, 1960. 8. Trueta, J.: Acute haematogenous
osteomye-litis: its pathology and treatment. Bull. Hosp. Joint Dis., 14:5, 1953.
9. Trueta, J., and Morgan, J. D. : Late results in treatment of 100 cases of acute haematog-enoiss osteomvehtis. Brit. J. Surg., 41:449,
1954.
10. Buchman, J., and Fenton, R. L. : Role of surgical approach in treatment of acute