Secondary Logo

Journal Logo

Original Studies

Short- Versus Long-term Antimicrobial Treatment for Acute Hematogenous Osteomyelitis of Childhood

Prospective, Randomized Trial on 131 Culture-positive Cases

Peltola, Heikki MD*; Pääkkönen, Markus MD†‡; Kallio, Pentti MD, PhD*; Kallio, Markku J. T. MD* The Osteomyelitis-Septic Arthritis Study Group

Author Information
The Pediatric Infectious Disease Journal: December 2010 - Volume 29 - Issue 12 - p 1123-1128
doi: 10.1097/INF.0b013e3181f55a89


Acute hematogenous osteomyelitis of childhood (AHOM), previously a devastating disease,1–3 is traditionally treated with long courses of antimicrobials, commenced intravenously for at least 1 to 2 weeks. Bone drilling (trepanation) to drain subperiosteal pus and to obtain a sample for bacteriology is done before the institution of antimicrobial, and if adjacent joint is involved, capsulotomy and lavage are often performed.4–8 Usually, antimicrobial is discontinued only after the symptoms and signs have abated and the erythrocyte sedimentation rate (ESR) has normalized.4–8 The true need for all these treatments in everyday practice has never been examined in a sufficiently powered, prospective, randomized trial. Since no consensus on therapy exists, clinical pathways are based on results of case series, expert recommendations, and personal experience.9–16

Realizing the need for a prospective, randomized, equivalence trial,14 we conducted a prospective randomized trial on childhood AHOM in Finland. The aim was to examine the relevance of the current treatment policy by shortening the medication and tightening the indications for surgery. We learned that the treatment of this potentially serious condition can often be simplified.



We conducted a prospective, randomized, and controlled study of AHOM in Finland (first 50 cases caused by Staphylococcus aureus have been published earlier).17 Osteomyelitis formed a part of a comprehensive study on acute pediatric osteoarticular infections carried out in 7 referral hospitals (listed in Appendix) in 1983 to 2005; the data on septic arthritis are presented elsewhere.18 The study protocol was approved by the relevant ethical committees, and the child was included only if consent was obtained from the legal guardian. Only previously healthy children without major trauma, a skin ulcer, or pretreatment antimicrobials were included. The trial was designed, conducted, and analyzed independently of pharmaceutical companies.

When AHOM—fever, painful, and swollen limb without trauma; restriction of motion; often tender and warm area—was suspected in children aged between 3 months and 15 years, the clinician contacted a special ward of the Children's Hospital, Helsinki, 24 h/d. Each patient obtained a computer-generated number that randomized him or her to receive antimicrobial treatment for 20 or 30 days; the information was immediately recorded in the chart.


Since AHOM in industrialized countries is mostly caused by Gram-positive agents, clindamycin19,20 (40 mg/kg per day every 6 hours), or a first-generation cephalosporin (mentioned later in the text; 150 mg/kg per day every 6 hours) was used. This randomization was done by birthday (odd or even). Of cephalosporins, cephradine21–23 was our first choice, because it was the only first-generation agent available for parenteral and oral use. Later, withdrawal of cephradine from Scandinavia forced a change to intravenous cephalothin and oral cephalexin or cefadroxil (all administered as cephradine). The switch was not considered critical for the study, because these cephalosporins perform similarly.24 Since, in the first study years, Haemophilus influenzae type b (Hib) was a common cause of septic arthritis in young children,25 ampicillin or amoxicillin (both 200 mg/kg per day every 6 hours) was administered at the age of 0 to 4 years until the agent was isolated. The course was completed with only 1 antimicrobial).

Once vaccination had eliminated Hib, the ampicillin/amoxicillin was abandoned in 1997.26

Antimicrobial treatment was instituted intravenously for 2 to 4 days and completed orally with similar dosages. Serum antimicrobial concentrations were not assayed. In case of allergy, an alternative drug was used. Nonsteroidal anti-inflammatory drugs were given at the discretion of the physician (a pediatrician or a pediatric or orthopedic surgeon).

Identification of the Causative Agent, Role of Surgery, and Monitoring of Patients

Blood cultures were taken in all cases. Except diagnostic drilling, aiming mainly to obtain a representative sample for bacterial culture, surgery was kept as minimum. All further operations were performed by an experienced orthopedic or pediatric surgeon.

The present laboratory and radiographic investigations comprised plain radiograph on admission, and on day 10 and 19; basic blood analysis at presentation, and on day 5 and 10. As shown in Figure 1, the serum C-reactive protein level (CRP)27,28 and ESR4–8 were measured sequentially throughout illness. All CRP values exceeding 20 mg/L29 and the ESRs over 20 mm/h4 were deemed increased. Scintigraphy was performed in most cases, whereas computerized tomography and magnetic resonance imagining were performed only on demand. The data were recorded in special forms, and computerized and analyzed in Helsinki with use of Statview (SAS Institute). A researchers' meeting was held once a year.

C-reactive protein level (CRP), erythrocyte sedimentation rate (ESR), and blood leukocyte count (WBC) of the 67 cases in the short-term and 64 cases in the long-term treatment groups. Curves depicted with standard error of mean (SEM).

Discontinuation of Antimicrobial Therapy, and Control Visits

The antimicrobial treatment was discontinued when most (though not necessarily all) symptoms and signs of AHOM had subsided and the CRP value was <20 mg/L,29 no matter how high the ESR. If the CRP level remained elevated or reincreased notably, or clinical signs persisted, medication was continued until 2 normal CRP values were obtained. In cases of likely drug allergy, the agent was switched to the other agent recommended by the protocol.

Because osteoarticular infections have some tendency to reoccur10,30 and long-term sequelae may develop slowly,31,32 follow-up visits were scheduled at 2 weeks, and 3 and ≥12 months post hospitalization, with extra visits if needed. The liaison person executed all investigations, paying special attention to all potential sequelae, no matter how mild. Radiographs were performed, the CRP level, and ESRs were checked at each follow-up visit.

Outcome Measures, Sample Size, and Statistical Analysis

To avoid discussion on the accuracy of diagnosis,33 only culture positive cases were included. The primary end point was full recovery (having no symptoms or signs of AHOM at the end of the follow-up period, with no readministration of antimicrobial for an osteoarticular indication since the primary treatment). Secondary outcomes included all potential sequelae and the absence of disease after discontinuation of antimicrobial therapy.

The 95% confidence interval for the difference in success rates was calculated on the basis of normal approximation to the binomial distribution. This noninferiority test was based on the lower bound of the confidence intervals being within prespecified noninferiority margin of 15% and the upper bounds containing 0%. Assuming a 90% efficacy in both groups, a 80% power, and a 1-sided significance level of P = 0.025, 63 patients in each group were needed to test the null hypothesis (at least 15% difference in treatment results).



AHOM was diagnosed clinically in 183 cases of which 131 (72%) were culture-positive. Of these, 67 and 64 patients comprised the short- and long-treatment groups, respectively. The patient characteristics are summarized in Table, Supplemental Digital Content 1, Bone and blood grew bacteria in 35 cases, bone alone in 44 cases, and blood alone in 52 cases. Methicillin-susceptible S. aureus was the most common agent being responsible for 89% (n = 117) of cases, the remaining 11% (14) cases being caused by various agents. The extent of inflammation was comparable in the treatment groups as judged by slightly higher CRP values in the short-term but higher ESR in the long-term treatment group.

All age groups with a slight preponderance at school age were affected. The median age was 9.0 years in the short-term and 9.3 years in the long-term treatment group. Medical attention was sought within 3 days and 7 days in 47% (N = 61) and 84% (N = 110), respectively. Usually long bones of the lower extremity or pelvis were affected: the femur in 27%, tibia in 24%, pelvic girdle in 15%, calcaneum in 11%, and fibula in 10%.

An adjacent joint was involved in 13 and 11 cases in the short- and long-term treatment groups, respectively, the agents being S. aureus (N = 10 and 9), Hib (1 and 1), and Streptococcus pyogenes (2 and 1). Combined AHOM-septic arthritis induced greater changes in the laboratory indices than did plain AHOM; the initial CRP values and ESR's were 99 mg/L and 62 mm/h, and 81 mg/L and 44 mm/h, respectively. For ESR, the difference was significant (P = 0.0031).


The short- and long-term treatment groups comprised 67 and 64 children, respectively. In all, 75 (57%) children received clindamycin alone, combined with amoxicillin in 5 cases. Cephalosporin alone was used for 37 patients, combined with amoxicillin in 5 cases.Table 1 summarizes the 9 patients whose antimicrobials were modified, or medication was prolonged for more than a week. Likely allergy indicated a change to the alternative in 5 cases.

Patients for Whom Antimicrobial Treatment Was Modified or Prolonged for More Than a Week From Scheduled, in Order of Age

Antimicrobials were administered intravenously for 3.7 (median) and 4.1 days in the short- or long-term treatment groups, respectively. The overall range was 0 to 14 days, and 8 children were on intravenous medication for 7 days. The entire duration of antimicrobials in the short-term group was 20.0 days (no deviation in interquartile range, 90% range: 10–21 days), the shortest course (of cephalosporin) being only 9 days for S. aureus AHOM in the proximal femur of a 10-year-old boy. The long-term treatment group received antimicrobials for 30 days (no deviation in interquartile range, 90% range: 30–43 days). The extreme prolongations in antimicrobial treatment were 108 days in the short-term, and 91 days in the long-term treatment group (Table 1).


Corticotomy was performed in 62 cases, combined with arthrocentesis in 14 cases. Percutaneous bone aspiration was performed in 34 cases. When seeking arthritis, 8 children underwent joint aspiration; subperiosteal abscess was drained in 4 cases. The need for surgical intervention did not differ between the groups. Thirty-one (24%) children did not require surgery.


Most patients recovered rapidly without significant difference between the groups (Fig. 1). Importantly, after discontinuation of antibiotic therapy, no laboratory or clinical marker in the 20-day group deviated from the longer treatment group. Radiographic changes (periosteal reaction, osteolysis, or sclerosis) were, in all, first found with an increasing frequency, in 21% on day 1, 40% on day 10, and 41% on day 19. Thereafter, changes were seen in 35% on day 29. At 1 year, radiographic changes were observed in 15% of cases (16% of the short-term and 14% of the long-term treatment group). Initial bone-scan was positive in 74 (56%) cases.

Two weeks after hospitalization, 16 children in the short-term and 19 in the long-term treatment group had local pain, swelling, or tenderness. The presence or absence of symptoms or signs were not associated with the length of treatment, but with the time elapsed from the onset of disease to the institution of antimicrobial. If length of history (average) had been 7 or 5 days (P = 0.035), the differences between children with and without symptoms at 2 weeks post hospitalization were as follows: on admission, CRP value was 116 versus 76 mg/L (P = 0.006); ESR, 56 versus 45 mm/h (P = 0.046); normalization of CRP level in 14 versus 8 days (P = 0.0008); and normalization of ESR in 27 versus 17 days (P = 0.0005), respectively.

At 3 months, only 5 children had complaints. In the short-term treatment group, one 9-year-old boy had slight heel tenderness following calcaneal AHOM; the lesion healed in 6 months. In the long-term treatment group, there were 4 patients, all boys, aged 14, 13, 6, and 1 year(s). The first 3 complained of tenderness in the fibular, vertebral, or metatarsal region, respectively. The youngest, who developed AHOM at 1.6 years (No. 2 in Table 1), showed slight asymptomatic varus deformity following postvaricella tibial AHOM and pathologic fracture. Due to suspected sequestrum and chronic granulomatous disease (both later unconfirmed), antimicrobials were given for 108 days. The minor deformity was healing 2.3 years later.

Final follow-up was ≥12 months post hospitalization for all patients with the exception of 5 children (3 short-term and 2 long-term groups) who were lost to follow-up. However, all these patients had been found recovered at earlier stages, therefore the parents refused further checkups. Of the 126 attendees, following 2 minor sequelae were detected (Table 1): asymptomatic 8 degree varus deformity following tibia AHOM at age 1.6 years, and pain during exercise after ankle AHOM-arthritis at age 12 years. In this case, arthrosis of the tibiotalar joint was seen in radiograph. No patient required corrective osteotomy.


The main lesson learned from this study was that antimicrobials for 20 days, administered mainly orally, usually suffice in childhood AHOM without an obvious risk of reoccurrence or sequelae, provided the clinical response is good and the CRP concentrations normalize quickly. Our length of treatment was half that in a retrospective series of 146 cases from Canada.6 However, the antimicrobial should be chosen carefully, and large doses and the 4-times daily regimen with these time-dependent agents are probably necessary.

The initial intravenous period can be a few days only. Then, the course is completed orally with a well-absorbed agent, such as clindamycin or a first-generation cephalosporin. No serum assays are needed in our opinion. All this simplifies treatment, avoids complications,34 decreases costs, and allows discharge earlier.

Substantial regional differences prevail in the manifestations and severity of AHOM. However, a review of a number of reports from elsewhere does not support the view that our patients would not have been representative of those in industrialized countries. In the United States, the situation might be a little different since methicillin-resistant S. aureus and other earlier rare or unknown agents have become an issue.35,36 An important finding from this study was that prolonged history did not play a major role for the final outcome, but of course, treatment should be started without delay.

No major difference in clinical effectiveness between clindamycin and the cephalosporins was observed. Clindamycin penetrates slightly better into bone than β-lactams,37 but our intentional use of large doses for both agents (as recommended decades ago),4 probably concealed any special findings in this regard. Of note was the finding that two-thirds (87/131, 66%) of our patients grew bacteria, usually S. aureus, from blood. Thus, little question remained whether or not bacteremic staphylococcal infections can be treated (mostly) orally with first generation cephalosporins of clindamycin if large enough doses are used.

The failure to understand the pivotal difference between infection and inflammation38 is one reason why antimicrobials for AHOM (and many other bacterial infections) are given for long periods. Normalizing CRP values give good guidance when the antibacterial can be discontinued. To maximize the informative value of CRP measurements, sequential determinations are needed, and to best serve the patient, the results should arrive on the same day. One should be acquainted with the dynamics of this valuable yardstick: even when osteomyelitis heals uneventfully, the CRP level increases for 1 to 2 days, but then begins to decline and reaches 20 mg/L in 7 to 10 days.27,28,39 If the value continues to rise or remains high on the fourth day, the clinician should suspect a complication.28

Minimal surgery (percutaneous aspiration or drilling) aiming primarily at obtaining a representative sample for bacteriology, sufficed in most cases; 24% of our patients escaped all surgery and recovered uneventfully. However, surgery has a role in a few selected cases of AHOM, but in the every-day practice, its need is less than thought. Some new thinking is warranted in this important issue.

We realize a few limitations in our study. Collecting 131 culture-positive cases took many years, because AHOM in Finland is rare; the annual incidence at age 0 to 14 years is only 4.5 per 100,000.25 However, yearly researchers' meetings kept the protocol active and consistent. Not all children received solely the recommended agents nor was the treatment always exactly 20 or 30 days.

This said, we do not think we missed any recurrence of disease or sequela, although 5 (4%) patients did not return for the ≥1-year checkup. All had recovered in the earlier check-ups, and after the first months, recrudescences are very unlikely.40

We do not claim that our simplified treatment will heal 100% of childhood AHOM, but we believe that for most children who respond quickly and whose CRP values normalize within 10 days, large doses of clindamycin or a first-generation cephalosporin 4 times daily for about 20 days (intravenous for first 3–4 days) suffice. Extensive surgery is rarely needed. Further, randomized trials are warranted to confirm our results.


Members of the Osteomyelitis-Septic Arthritis (OM-SA) Study Group

Kari Aalto and Eeva Salo (Aurora Hospital, and Helsinki University Central Hospital, Hospital for Children and Adolescents, Helsinki), Juhani Merikanto (Helsinki University Central Hospital, Hospital for Children and Adolescents, Helsinki), Ilkka Anttolainen and Pentti Lautala (Päijät-Häme Central Hospital, Lahti), Marja Heikkinen (Kuopio University Hospital, Kuopio), Anita Hiippala and Niilo Kojo (Etelä-Saimaa Central Hospital, Lappeenranta), Ulla Kaski (Seinäjoki Central Hospital, Seinäjoki), and Pekka Ojajärvi (Jorvi Hospital, Espoo).


The authors thank Orion Pharma company, which supported the writing of this report through the University of Helsinki (Project Code 34490). The authors also thank Juho Ajanki for drawing the figure.


1. Lewis D. Acute osteomyelitis. JAMA. 1929;92:783–786.
2. Green WT, Shannon JG. Osteomyelitis of infants. A disease different from osteomyelitis of older children. Arch Surg. 1936;32:462–493.
3. Kenney WE. The prognosis in acute haematogenous osteomyelitis with and without chemotherapy. Surgery. 1944;16:477–484.
4. Tetzlaff TR, McCracken GH Jr, Nelson JD. Oral antibiotic therapy for skeletal infections of children. II. Therapy of osteomyelitis and suppurative arthritis. J Pediatr. 1978;92:485–490.
5. Kolyvas E, Ahronheim G, Marks MI, et al. Oral antibiotic therapy of skeletal infections in children. Pediatrics. 1980;65:867–871.
6. Karwowska A, Davies D, Jadavji T. Epidemiology and outcome of osteomyelitis in the era of sequential intravenous-oral therapy. Pediatr Infect Dis J. 1998;17:1021–1026.
7. Vinod MB, Matussek J, Curtis N, et al. Duration of antibiotics in children with osteomyelitis and septic arthritis. J Paediatr Child Health. 2002;38:363–367.
8. Gutierrez K. Bone and joint infections in children. Pediatr Clin North Am. 2005;52:779–794.
9. Mollam RA, Piggott J. Acute osteomyelitis in children. J Bone Joint Surg. 1977;59(-B):2–7.
10. Syrogiannopoulos GA, Nelson JD. Duration of antibiotic therapy for acute suppurative osteoarticular infections. Lancet. 1988;1:37–40.
11. Dahl LB, Hoyland AL, Dramsdahl H, et al. Acute osteomyelitis in children: a population based retrospective study 1965 to 1994. Scand J Infect Dis. 1998;30:573–579.
12. Christiansen P, Frederiksen B, Glazowski MJ, et al. Epidemiologic, bacteriologic, and long-term follow-up data of children with acute haematogenous osteomyelitis and septic arthritis: a ten-year review. J Pediatr Orthop. 1999;8:302–305.
13. Bonhoeffer J, Haeberie B, Schaad UB, et al. Diagnosis of acute haematogenous osteomyelitis and septic arthritis: 20 years experience at the University Children′s Hospital Basel. Swiss Med Wkly. 2001;131:575–581.
14. Le Saux N, Howard A, Barrowman NJ, et al. Shorter courses of parenteral antibiotic therapy do not appear to influence response rates for children with acute hematogenous osteomyelitis: a systematic review. BMC Infect Dis. 2002;2:16.Available at:–2334/2/16.
15. Lew DP, Waldvogel FA. Osteomyelitis. Lancet. 2004;364:369–379.
16. Malcius D, Trumpulyte G, Barauskas V, et al. Two decades of acute haematogenous osteomyelitis in children; are there any changes? Pediatr Surg Int. 2005;21:356–359.
17. Peltola H, Unkila-Kallio L, Kallio MJ; the Finnish Study Group. Simplified treatment of acute staphylococcal osteomyelitis of childhood. Pediatrics. 1997;99:846–850.
18. Peltola H, Pääkkönen M, Kallio P, et al. The OM-SA Study Group. Prospective, randomized trial of 10 days versus 30 days of antimicrobial treatment, including a short-term course of parenteral therapy, for childhood septic arthritis. Clin Infect Dis. 2009;48:1201–1210.
19. Nicholas P, Meyers BR, Lewy RN, et al. Concentration of clindamycin in human bone. Antimicrob Agents Chemother. 1975;8:220–221.
20. Kaplan SL, Mason EO, Feigin RD. Clindamycin versus nafcillin or methicillin in the treatment of Staphylococcus aureus osteomyelitis in children. South Med J. 1982;75:138–142.
21. Zaki A, Schreiber EC, Weliky I, et al. Clinical pharmacology of oral cephradine. J Clin Pharmacol. 1974;14:118–126.
22. Brosof AB, Spitzer TQ. Cephradine for the treatment of bone infections due to Staphylococcus aureus. Curr Ther Res. 1979;26:317–321.
23. Leigh DA. Determination of serum and bone concentrations of cephradine and cefuroxime by HPLC in patients undergoing hip and knee replacement surgery. J Antimicrob Chemother. 1989;23:877–883.
24. Lambert HP, O'Grady FW. Cephalosporins. In: Antibiotic and Chemotherapy. 6th ed. Edinburgh, United Kingdom: Churchill Livingstone; 1992:89–135.
25. Peltola H, Vahvanen V. A comparative study of osteomyelitis and purulent arthritis with special reference to aetiology and recovery. Infection. 1984;12:75–79.
26. Peltola H, Kallio MJ, Unkila-Kallio L. Reduced incidence of septic arthritis in children by Haemophilus influenzae type-b vaccination. Implications for treatment. J Bone Joint Surg. 1998;80(-B):471–473.
27. Peltola H, Räsänen JA. Quantitative C-reactive protein in relation to erythrocyte sedimentation rate, fever, and duration of antibiotic therapy in bacteraemic diseases of childhood. J Infect. 1982;5:257–267.
28. Roine I, Faingezicht I, Arguedas A, et al. Serial serum C-reactive protein to monitor recovery from acute hematogenous osteomyelitis in children. Pediatr Infect Dis J. 1995;14:40–44.
29. Peltola H, Jaakkola M. C-reactive protein in early detection of bacteremic versus viral infections in immunocompetent and compromised children. J Pediatr. 1988;113:641–646.
30. Uçkay I, Assal M, Legout L, et al. Recurrent osteomyelitis caused by infection with different bacterial strains without obvious source of reinfection. J Clin Microbiol. 2006;44:194–196.
31. Spindler T, Huenges R, Hoppe JE. Acute haematogenous osteomyelitis in childhood—correlation of clinical aspects, diagnostic parameters, therapy and prognosis. Klin Pädiatr. 1998;210:56–59.
32. Nunn T, Rollinson P. Haematogenous pyogenic bone and joint sepsis—reducing avoidable morbidity. S Afr Med J. 2007;97:456–460.
33. Schultz C, Holterhus PM, Seidel A, et al. Chronic recurrent multifocal osteomyelitis in children. Pediatr Infect Dis J. 1999;18:1008–1013.
34. Ruebner R, Keren R, Coffin S, et al. Complications of central venous catheters used for the treatment of acute hematogenous osteomyelitis. Pediatrics. 2006;117:1210–1215.
35. Martínez-Aguilar G, Hammerman WA, Mason EO Jr, et al. Clindamycin treatment of invasive infections caused by community-acquired, methicillin-resistant and methicillin-susceptible Staphylococcus aureus in children. J Pediatr Infect Dis. 2003;22:593–598.
36. Carrillo-Marques MA, Hulten KG, Hammerman W, et al. USA300 is the predominant genotype causing Staphylococcus aureus septic arthritis in children. Pediatr Infect Dis J. 2009;28:1076–1080.
37. Landesdorfer CB, Bulitta JB, Kinzig M, et al. Penetration of antibacterials into bone. Pharmacokinetic, pharmacokinetic and bioanalytical considerations. Clin Pharmacokinet. 2009;48:89–24.
38. Ginsburg I. The role of bacteriolysis in the pathophysiology of inflammation, infection and post-infectious sequelae. APMIS. 2002;110:753–770.
39. Pääkkönen M, Kallio MJ, Kallio PE, et al. Sensitivity of erythrocyte sedimentation rate and C-reactive protein in childhood bone and joint infections. Clin Orthop Relat Res. 2010;468:861–866.
40. Tice AD, Hoaglund PA, Shoultz DA. Risk factors and treatment outcomes in osteomyelitis. J Antimicrob Chemother. 2003;51:1261–1268.

osteomyelitis; osteoarticular infection; CRP; ESR; Staphylococcus aureus

Supplemental Digital Content

© 2010 by Lippincott Williams & Wilkins, Inc.