ARTICLE

 

Bacterial aetiology and antimicrobial susceptibility of osteoarticular infections at a tertiary-level paediatric unit in South Africa

 

 

A Horn^I; C Centner^II; M Laubscher^I; H Tootla^III

^IMMed (Orth), FC Orth (SA); Orthopaedic Research Unit, Division of Orthopaedic Surgery, University of Cape Town, South Africa
^IIMMed (Micro), FCPath; Division of Medical Microbiology, University of Cape Town, South Africa; National Health Laboratory Service, Groote Schuur Hospital, Cape Town, South Africa
^IIIMMed (Micro), FCPath; Division of Medical Microbiology, National Health Laboratory Service, Red Cross War Memorial Children's Hospital, Cape Town, South Africa

Correspondence

 

 

---

ABSTRACT

BACKGROUND. Osteoarticular infections (OAI) are very common in children living in low- and middle-income countries, yet the bacterial aetiology and antibiotic susceptibility of OAI in children are not well described.
OBJECTIVE. To determine bacterial aetiology and antibiotic susceptibility of OAI in paediatric patients in a regional hospital in Cape Town, South Africa.
METHODS. The study included all patients who underwent surgery for OAI over a 3-year period, and those with organisms identified from tissue, pus, fluid or blood. Duplicate cultures from the same patient were excluded if the organism and antibiotic susceptibility profiles were the same. Patients were categorised by age and class of infection (septic arthritis, acute osteomyelitis, fracture-related infection, postoperative sepsis and chronic osteomyelitis) and organisms were stratified accordingly.
RESULTS. We identified 132 organisms from 123 samples collected from 96 patients. Most cultured organisms were from children older than 3 years with acute haematogenous septic arthritis, osteomyelitis or both. Methicillin-sensitive Staphylococcus aureus (S. aureus) accounted for 56% (n=74/132) of organisms cultured. The Enterobacterales accounted for 17% (n=22/132) of organisms cultured, mostly in the fracture-related and postoperative infection groups. Of these, six each were extended-spectrum ß-lactamase producers and AmpC producers, respectively. There were no carbapenemase-producing Enterobacterales.
CONCLUSION. Methicillin-sensitive S. aureus is the most common infecting organism in paediatric OAI and anti-staphylococcal penicillin is the most appropriate empiric treatment for haematogenous OAI in our environment. In fracture-related or postoperative infections, Enterobacterales were more frequently cultured, and treatment should be guided by culture and susceptibility results.

---

 

 

Osteoarticular infections (OAIs) are a common cause of morbidity in paediatric patients. The incidence is reported to range from 1 in 800 to 1 in 5 000, with low- and middle-income countries (LMICs) reporting a higher incidence.^[1] In contrast to adults, these infections are predominantly haematogenous and only rarely related to previous surgery or fractures.

The management of these infections, including septic arthritis and osteomyelitis, usually consists of surgical drainage of pus followed by empirical antibiotic treatment based on the patient's age, severity of illness and the most likely causative bacterial organisms, considering local resistance patterns. Antibiotic treatment is subsequently adjusted to directed therapy when causative organisms and susceptibility results from intraoperative samples become available.

Two recent South African (SA) studies investigated the bacterial aetiology and susceptibility of isolates in chronic osteomyelitis.^[2,3] However, these studies included only adult patients, in whom the epidemiology of OAIs is different. In contrast, the bacterial aetiology and susceptibility of causative organisms for OAIs in children is not well described in SA, and local empirical antibiotic regimes are based on international data and anecdotal evidence. Findings from other sub-Saharan countries may also not apply to our context as SA has a more temperate climate than, for instance, Malawi^[4] or Tanzania,^[5] and our socioeconomic context is different.

This study aimed to identify the bacterial aetiology of OAIs in children at the Red Cross War Memorial Children's Hospital, Cape Town, South Africa, and to determine the antibiotic susceptibility of common infectious organisms to better inform the choice of empirical antibiotics for OAIs.

 

Methods

Following ethics approval from the University of Cape Town's Human Research Ethics Committee (ref. no. HREC 236/2021), we searched the prospectively maintained departmental database for all patients at our institution who underwent surgery for OAI between January 2018 and December 2020 (Fig. 1). Only patients operated at our paediatric hospitals were included, therefore excluding most patients older than 13 years as per local policy. The National Health Laboratory Service's (NHLS) database was searched to identify patients with positive culture results (blood, pus, and tissue fluid) during admission for OAI. Samples with no growth were excluded. Patients whose cultures yielded mixed growths, where no specific organism was identified, were also excluded. Duplicate isolates from the same patient were excluded if the cultures yielded growth of the same organism, with the same antibiotic susceptibility.

 

 

All samples were submitted to the NHLS microbiology laboratory at our institution for routine bacterial culture and susceptibility testing. Identification and susceptibility testing of organisms were performed using the VITEK 2 automated system (bioMérieux, France), biochemical or antigen-detection methods and disc or gradient diffusion antibiotic susceptibility testing methods where appropriate. Results of antibiotic susceptibility tests were interpreted using the Clinical Laboratory and Standards Institute (CLSI) guidelines for the relevant year. In this review, antibiotic susceptibility was classified as either susceptible or non-susceptible, with the non-susceptible category including both the 'intermediate' and 'resistant' categories as defined by the CLSI guidelines.

For the analysis, OAIs were divided into six categories: acute haematogenous septic arthritis, acute haematogenous osteomyelitis, acute haematogenous osteomyelitis and septic arthritis, chronic osteomyelitis, fracture-related infection and postoperative infection. Acute haematogenous infections were defined as infections that arose de novo with no predisposing injury or insult such as trauma or surgery. Fracture-related infection was defined as per the international consensus criteria.^[6] Postoperative infection was defined as infection arising at a previously sterile site following elective surgery. Patients were categorised into three age groups: those younger than 3 months, those between 3 months and 3 years and those who were 3 years and older. The age categorisation was based on published trends observed on the aetiology of OAIs based on age in children.^[7]

Descriptive analysis was performed using Stata version 14.2 (StataCorp., USA) and Microsoft Excel (Microsoft Corp., USA). Categorical variables were described using absolute values and percentages.

 

Results

A total of 123 samples collected from 96 patients were included in the analysis. Patient ages ranged from 1 month to 15 years (mean=6.5 years) old. A total of 132 organisms were isolated from these samples (Table 1). The distribution of organisms based on infection and age categories is summarised in Tables 2 and 3, respectively. Most organisms were cultured from patients older than 3 years (81%; n=107/132) and those presenting with acute haematogenous septic arthritis, osteomyelitis, or both (83%; n=109/132).

 

 

 

 

 

 

Bacterial aetiology and antibiotic susceptibility of commonly isolated or clinically relevant organisms

Staphylococcus aureus (S. aureus) accounted for 56% (n=74/132) of isolates cultured across the whole cohort. During the study period, no methicillin-resistant S. aureus was isolated. All S. aureus isolates were also susceptible to the following non-ß-lactam antibiotics: clindamycin, vancomycin, linezolid and ciprofloxacin. Forty percent of S. aureus isolates were resistant to trimethoprim-sulfamethoxazole.

Enterobacterales were the second-most frequent cause of osteoarticular infection in our cohort, particularly in fracture-related infections (48%) and postoperative infections (50%), while only accounting for 17% (n=22/132) of all isolates. Escherichia coli (n=5) and Enterobacter cloacae («=5) were the most frequently identified. Resistance to the ß-lactam antibiotics due to AmpC-producing ß-lactamases was identified in 6/22 (27%) isolates. Of these, all were susceptible to piperacillin-tazobactam, cefepime and trimethoprim-sulfamethoxazole, and five were susceptible to ciprofloxacin. Extended-spectrum ß-lactamase (ESBL) producing Enterobacterales accounted for 6/22 (27%) isolates. Of these isolates, only three were susceptible to ciprofloxacin and one to trimethoprim-sulfamethoxazole. Carbapenem-resistant Enterobacterales (CRE) were not isolated in this cohort.

Bacterial aetiology and antibiotic susceptibility of infrequently isolated but clinically relevant organisms

Only 7/132 (5%) ß-haemolytic streptococci were isolated. Of those, one was identified as Streptococcus agalactiae (Group B streptococcus) and, unsurprisingly, was isolated from an infant <3 months old. The rest were identified as Streptococcus pyoge«es (Group A streptococcus). Three of these ß-haemolytic streptococci were tested against and found to be susceptible to penicillin.

Streptococcus pneumoniae and Haemophilus influenzae, both of which have vaccines included in the childhood vaccination programme, were only isolated on one occasion each and both were tested against and found to be susceptible to penicillin and ampicillin, respectively.

Pseudomonas aeruginosa was isolated in 2% of isolates (n=3/132), two were from fracture-related infections and one from chronic osteomyelitis. All were susceptible to the anti-pseudomonal ß-lactam antibiotics piperacillin-tazobactam, ceftazadime, cefepime, imipenem and meropenem and two were susceptible to ciprofloxacin.

Mixed cultures

Infections where more than one organism was identified were present in 13 patients, of whom seven had haematogenous septic arthritis. Six patients had underlying conditions predisposing to infection or poor healing, such as spina bifida, open fractures, postoperative sepsis or a combination of the above.

 

Discussion

To our knowledge, this is the first SA study to investigate causative organisms and antibiotic susceptibility in paediatric osteoarticular infections.

Published data from LMICs are scarce and usually describe haematogenous infections only.^[4,8-11] We included fracture-related and postoperative infections, and despite relatively low numbers of these types of infection, this allowed a better understanding of the bacterial aetiology of these infrequent conditions.

In accordance with published data from high- and low-income countries, S. aureus was the most frequently isolated organism in our patients, regardless of age. Interestingly, no cases of methicillin-resistant S. aureus (MRSA) were identified. The incidence of MRSA osteoarticular infections in children has been reported to be as high as 40% in the USA and Australia,^[12-14] and rates of 29% and 17% have been reported in Japan and Thailand, respectively.^[15] India, in particular, has classified MRSA as an endemic infection, with studies reporting an incidence as high as 55%.^[8,16] In Africa, the reported incidence of MRSA is low, with many series also reporting no MRSA.^[4,17] However, incidences of 9% and 15% have been reported in Tanzania and Tunisia, respectively.^[5,11]

Kingella kingae (K. kingae) is considered the most frequent cause of osteoarticular infections in children between 3 months and 3 years of age.^[18] Kingella is a fastidious, slow-growing Gram-negative bacillus that is difficult to culture using standard techniques. Positive cultures are more often obtained when inoculating samples into blood culture bottles. However, the gold standard for diagnosis is molecular methods, including PCR using specific gene targets from the K. kingae RTX toxin locus.^[19] PCR testing for K. Kingae is not performed at our institution, which may explain why we had no cases of K. kingae in our cohort.

Enterobacterales, which include Gram-negative enteric bacteria such as E. coli, E. cloacae, Salmonella species and Klebsiella pneumoniae, are known to be frequent causes of bone and joint infections in neonates and young children.^[8,9] In our cohort, Enterobacterales accounted for 17% of overall infections, but in patients older than 3 months. These children in our cohort tended to have infections secondary to fractures and trauma or prior surgery. However, the small number of cases precludes the identification of any definitive associations. Salmonella species was responsible for only three cases of osteoarticular infections in our cohort. This is in stark contrast to the published series by Lavy et al.^[4] from Malawi, where Salmonella species accounted for more than 50% of culture-positive septic arthritis cases in their larger cohort of 204 children.

Antibiotic susceptibility for the Enterobacterales varied. There were no CRE isolated in our group, but six isolates were ESBL producers and another six were AmpC ß-lactamase producers, conferring resistance to widely used and available ß-lactam antibiotics.

There are several limitations to our study. The retrospective study design and our method of patient identification may have been inadequate to identify all cases of OAI during the study period. The subgroups 'fracture-related infection' and 'postoperative sepsis' included only a handful of patients, precluding any definitive antibiotic recommendations. As we do not have access to PCR testing for K. kingae, we were unable to provide any information regarding this common pathogen.

 

Conclusion

Methicillin-sensitive S. aureus was the most frequently isolated organism causing haematogenous osteoarticular infections in our population. This was followed by a wide spectrum of Enterobacterales causing mostly fracture-related or postoperative infections. Empiric treatment of haematogenous OAI's consisting of an anti-staphylococcal penicillin such as cloxacillin is still recommended. For postoperative or fracture-related infections, the addition of a ß-lactam with broad Gram-negative activity against AmpC-producing and ESBL Enterobacterales and Pseudomonas aeruginosa is recommended, due to the high incidence of Gramnegative organisms seen in these groups. Identification of the causative organisms in OAI remains essential and once established, antibiotic regimens should be adjusted accordingly.

Declaration. None.

Acknowledgements. None.

Author contributions. AH: Conceptualisation, data capture, write-up. CC: Data analysis. ML: Protocol development HT: Data analysis and oversight.

Funding. None.

Conflicts of interest. None.

 

References

1. Horn A, Wever S, Hoffman E. Complications following acute severe haematogenous osteomyelitis of the long bones in children. SA Orthop J 2019;18(3):23-29. https://doi.org/10.17159/2309-8309/2019/v18n3a1        [ Links ]

2. Ferreira N, Reddy K, Venter R, Centner C, Laubscher M. Antibiogram profiles and efficacy of antibiotic regimens of bacterial isolates from chronic osteomyelitis of the appendicular skeleton: A developing-world perspective. S Afr Med J 2021;111(7):642-648. https://doi.org/10.7196/samj.2021.v111i7.15516        [ Links ]

3. Mthethwa PG, Marais L. The microbiology of chronic osteomyelitis in a developing world setting. SA Orthop J 2017;16(2):39-45. https://doi.org/10.17159/2309-8309/2017/v16n2a4        [ Links ]

4. Lavy C, Thyoka M, Pitani A. Clinical features and microbiology in 204 cases of septic arthritis in Malawian children. Bone Joint J 2005;87(11):1545-1548. https://doi.org/10.1302/0301-620x.87b11.16735        [ Links ]

5. Ali AM, Maya E, Lakhoo K. Challenges in managing paediatric osteomyelitis in the developing world: analysis of cases presenting to a tertiary referral centre in Tanzania. Afr J Paediatr Surg 2014;11(4):308. https://doi.org/10.4103/0189-6725.143136        [ Links ]

6. Metsemakers W-J, Morgenstern M, McNally M, et al. Fracture-related infection: a consensus on the definition from an international expert group. Injury 2018;49(3):505-510. https://doi.org/10.1016/j.injury.2017.08.040        [ Links ]

7. Lorrot M, Gillet Y, Le Guen CG, Launay E, Cohen R, Grimprel E. Antibiotic therapy of bone and joint infections in children: proposals of the French Pediatric Infectious Disease Group. Arch Pediatr 2017;24(12):S36-S41. https://doi.org/10.1016/s0929-693x(17)30517-1        [ Links ]

8. Agarwal A, Aggarwal AN. Bone and joint infections in children: septic arthritis. Indian J Pediatrics 2016;83(8):825-833. https://doi.org/10.1007/s12098-015-1816-1        [ Links ]

9. Agarwal A, Aggarwal AN. Bone and joint infections in children: acute hematogenous osteomyelitis. Indian J Pediatr 2016;83(8):817-824. https://doi.org/10.1007/s12098-015-1806-3        [ Links ]

10. Lavy CB. Septic arthritis in Western and sub-Saharan African children: a review. Int Orthop 2007;31(2):137-144. https://doi.org/10.1007/s00264-006-0169-9        [ Links ]

11. Trifa M, Bouchoucha S, Smaoui H, et al. Microbiological profile of haematogenous osteoarticular infections in children. Orthop Traumatol Surg Res 2011;97(2):186-190. https://doi.org/10.1016/j.otsr.2010.10.005        [ Links ]

12. Gafur OA, Copley LA, Hollmig ST, Browne RH, Thornton LA, Crawford SE. The impact of the current epidemiology of pediatric musculoskeletal infection on evaluation and treatment guidelines. J Pediatr Orthop 2008;28(7):777-785. https://doi.org/10.1097/bpo.0b013e318186eb4b        [ Links ]

13. Saavedra-Lozano J, Mejías A, Ahmad N, et al. Changing trends in acute osteomyelitis in children: impact of methicillin-resistant Staphylococcus aureus infections. J Pediatr Orthop 2008;28(5):569-575. https://doi.org/10.1016/s0084-3954(09)79439-4        [ Links ]

14. Brischetto A, Leung G, Marshall CS, Bowen AC. A retrospective case series of children with bone and joint infection from Northern Australia. Medicine 2016;95(8):e2885. https://doi.org/10.1097/md.0000000000002885        [ Links ]

15. Hunter S, Chan H, Baker JF. Global epidemiology of childhood bone and joint infection: a systematic review. Infection 2022;50(2):329-341. https://doi.org/10.1007/s15010-021-01741-3        [ Links ]

16. Sodavarapu P, Sudesh P, Gopinathan NR, Jayashree M, Kumar P, Rangasamy K. Characteristics of musculoskeletal involvement in pediatric patients with disseminated sepsis in a tertiary care center. Indian J Orthop 2021;56(2):345-352. https://doi.org/10.1007/s43465-021-00488-1        [ Links ]

17. Visser HF, Visser A, Goller K, Goller R, Nel J, Snyckers CH. Paediatric septic arthritis in a tertiary setting: A retrospective analysis. SA Orthop J 2010;9(2):92-96.         [ Links ]

18. Ilharreborde B, Bidet P, Lorrot M, et al. New real-time PCR-based method for Kingella kingae DNA detection: application to samples collected from 89 children with acute arthritis. J Clin Microbiol 2009;47(6):1837-1841. https://doi.org/10.1128/jcm.00144-09        [ Links ]

19. Ceroni D, Cherkaoui A, Ferey S, Kaelin A, Schrenzel J. Kingella kingae osteoarticular infections in young children: clinical features and contribution of a new specific real-time PCR assay to the diagnosis. J Pediatr Orthop 2010;30(3):301-304. https://doi.org/10.1097/bpo.0b013e3181d4732f        [ Links ]

 

 

Correspondence:
A Horn
anria.horn@uct.ac.za

Accepted 23 August 2023