a b s t r a c t

Introduction: Transient synovitis (TS) is a common and benign cause of hip pain in children, but must be distinguished from more serious entities such as Septic arthritis, osteomyelitis, and pyomyositis. Our objectives were to determine the risk of missed bacterial musculoskeletal infection and rates of diagnostic testing in children diagnosed with TS.

Methods: We performed a cohort study using the Pediatric Heath Information System of children 1-10 years diagnosed with TS in the ED. We determined rates of missed bacterial musculoskeletal infection (defined as a new diagnosis of septic arthritis, osteomyelitis, or pyomyositis within 14 days of initial ED visit). We described the initial diagnostic evaluation and ED management of children diagnosed with TS and variability between sites. Results: We analyzed 6419 encounters from 37 hospitals. 62 (1.0%, 95%CI: 0.7-1.2%) children were diagnosed with a missed bacterial musculoskeletal infection. Children with missed infection were younger than those with- out (median age 2.6 vs. 4.6 years, p < 0.01). Serum laboratory testing was performed in 76% of encounters with minimal variation across sites. There was significant variation in the rates of hip ultrasound by site (2 to 92%), which has increased in use over time (from 42% in 2016 to 62% in 2021).

Conclusion: In this large observational study, missed bacterial musculoskeletal infection in children diagnosed with TS was rare but more common in Younger children. The optimal combination of bloodwork and radiographic testing, especially ultrasound, to distinguish TS from more Serious disease remains unclear.

(C) 2022

1. Introduction

Transient (also referred to as toxic) synovitis of the hip (TS) is a com- mon and benign cause of hip pain and limp in children presenting to the emergency department (ED) [1-3]. While the causes of TS are not fully known, it often develops in conjunction with upper respiratory viral symptoms, and commonly recurs [4]. The differential diagnosis for acute hip pain and limp in the pediatric ED is broad, and includes septic arthritis, an acute surgical emergency [5]. Therefore, TS is often a diag- nosis of exclusion [6].

The gold standard method for ruling out septic arthritis is invasive hip arthrocentesis [7-9]. Clinical prediction algorithms, most notably the “Kocher criteria,” have been in use since the early 2000s to avoid this pro- cedure in children at low risk of septic arthritis [6,9-14]. The Kocher criteria rely on serum white blood cell count, erythrocyte sedimentation rate , weight bearing status, and fever as predictive markers, although some practitioners use c-reactive protein (CRP) in addition to or instead of ESR [10,15]. The use of newer inflammatory markers such as procalcitonin have been described for risk stratification, but their use in current practice is unknown [14,16,17]. The optimal imaging

* Corresponding author at: Division of Emergency Medicine, Cincinnati Children’s Hospital Medical Center, 3333 Burnet Avenue, ML 2008, Cincinnati, OH 45229, USA.

E-mail address: [email protected] (M.J. Lipshaw).

strategy for children presenting with hip pain or limp, prior to diagnosing TS is unclear. The necessity of hip radiographs is controversial [10,18-20], and while hip ultrasound is increasingly described in the literature, it is of questionable utility, as presence of an effusion does not distinguish between TS and septic arthritis [20-25]. Bacterial osteomyelitis and pyomyositis can present similarly to septic arthritis, and similar inflam- matory markers have been used to diagnose these conditions [26-28]. While several clinical prediction rules exist in the literature, there is a lack of evidence for how they perform in the ED setting [29]. In addition, national estimates of the risk of missed bacterial musculoskeletal infec- tion in actual practice in children diagnosed with TS are lacking.

Therefore, the objectives of this investigation were to determine the risk of return visit with missed bacterial musculoskeletal infection (septic arthritis, osteomyelitis or pyomyositis) in children diagnosed with TS in the ED; as well as to describe the current diagnostic evaluation of TS in the ED.

1. Methods
   1. Study design and setting

We conducted a retrospective cohort study of children diagnosed with TS who were evaluated in EDs that contribute data to the Pediatric

https://doi.org/10.1016/j.ajem.2022.08.022

0735-6757/(C) 2022

Health Information System (PHIS). PHIS is a database that contains inpatient, ED, ambulatory surgery and observation encounter-level administrative data from over 50 not-for-profit, tertiary care pediatric hospitals in the United States affiliated with the Children’s Hospital Association (Lenexa, KS) [30] . Data are de-identified and subjected to reliability and validity checks before inclusion in the database. We included hospitals with complete clinical and billing data for our study period. This study was deemed exempt from review by our local institutional review board.

• 1. Study population

We included children aged 1 to 10 years who were treated in the ED between January 1, 2016, and December 31, 2021 and had a discharge diagnosis code for transient synovitis of the hip (International Classifica- tion of Diseases Tenth Revision [ICD-10] code M67.35) in any position.

We excluded children with complex chronic conditions (as defined by Feudtner et al. [31]), as the diagnosis and management of TS may be different in these patients. We excluded children with Diagnosis codes for bacterial musculoskeletal infection (osteomyelitis [M86], septic arthritis [M00 or M01], Lyme arthritis [A69.23], and pyomyositis [M60]) at the index visit or a prior encounter. We also excluded children with a previous TS encounter in the ED within the prior 30 days.

• 1. Measurements

We extracted the following variables from the PHIS database for each encounter: ED arrival date, age, sex, race, ethnicity, insurance status, encounter diagnoses, underlying CCC, procedures (hip arthrocentesis), laboratory testing (CBC, ESR, CRP, PCT, blood culture, Lyme disease test- ing, synovial fluid cell count and culture), imaging of the hip, pelvis or lower extremity (radiograph, nonvascular ultrasound, magnetic reso- nance imaging [MRI], and computed tomography [CT]), patient disposi- tion from the ED (discharge vs. hospital admission), and return visit information. We defined children as having had an arthrocentesis performed if there was a procedure code for hip arthrocentesis, imaging guidance for arthrocentesis, or a laboratory charge for body fluid exami- nation.

• 1. Outcomes

Our primary outcome was missed bacterial musculoskeletal infec- tion. We defined this outcome as presence of an ICD-10 diagnosis code for septic arthritis (M00 or M01), osteomyelitis (M86), or pyomyositis (M60) at an ED or inpatient encounter within 14-days following the index visit for TS. Secondary outcomes included rates of imaging, laboratory testing, arthrocentesis, and hospital admission at index visit.

• 1. Analysis

We described the demographic characteristics of the cohort using frequencies with proportions for categorical variables and median and interquartile range (IQR) for age. We determined the frequency of TS visits with missed bacterial musculoskeletal diagnosis by calculating a proportion with a corresponding 95% confidence interval (CI). We described characteristics of children who met our primary outcome, detailing type of bacterial musculoskeletal infection (septic arthritis, osteomyelitis, pyomyositis, or multiple types), as well as breakdown by age. To compare Demographic and clinical factors between children with and without missed bacterial musculoskeletal diagnosis, we used ?² and Fisher exact tests as appropriate for categorical variables and the Kruskal-Wallis test for the age variable. We calculated the propor- tion of children with laboratory studies, imaging studies, and proce- dures performed at the index TS encounter and described variation in practice based on site, age of the patient, and trend over time. Statistical

analyses were conducted in R (version 4.0.2; R Foundation for Statistical Computing, Vienna, Austria).

1. Results
   1. Study population

Over the study period, we identified 6835 ED encounters with a dis- charge diagnosis of TS from 37 PHIS hospitals, of which 6419 remained after exclusions (Fig. 1). The median age of the cohort was 4.6 years (IQR: 3.1-6.2), and 67% were male. 19.4% of patients were admitted to the hospital from the ED. Additional characteristics of the cohort are described in Table 1.

• 1. Missed bacterial musculoskeletal infections

Of the 6419 children with a discharge diagnosis of TS included in the analysis, 62 (1.0%, 95% CI: 0.7-1.2%) were diagnosed with a bacterial mus- culoskeletal infection within 14 days. Of these, 35 had septic arthritis (56%), 33 had osteomyelitis (53%), and 21 had pyomyositis (34%). 22 of the children with missed infection had multiple Infection types (35%), in- cluding 5 children with all three. The highest individual site rate of missed infection was 6% and the lowest was 0%. The median time to return with missed bacterial musculoskeletal infection diagnosis was 4 days.

Children with a return visit with a bacterial musculoskeletal infec- tion diagnosed were younger than those without (median age 2.6 vs. 4.6 years, p < 0.01). The ages of children with missed bacterial muscu- loskeletal infection are shown in Fig. 2 in comparison to the entire co- hort. The rate of missed bacterial musculoskeletal infection was 2.4% (n = 37) in children <3 years-of-age and 0.5% (n = 25) in children 3 years-of-age or older (p <=0.01). Children with and without return visits were similar in all other demographic categories. There was no differ- ence in rates of laboratory testing, imaging, procedures, and hospital admission at the index TS visit between those who did and did not have a missed bacterial infection (Table 1).

• 1. Laboratory testing, radiologic studies, arthrocentesis, and disposition

Seventy-six percent (n = 4851) of encounters were associated with serum laboratory testing. The most commonly ordered laboratory test

Fig. 1. Enrollment flow diagram. TS, transient synovitis.

Table 1

Clinical and demographic characteristics of children diagnosed with transient synovitis in the ED.

	Overall	No missed infection	Missed bacterial musculoskeletal infection	p-value
n	6419	6357	62
Age in Years (median [IQR])	4.62 [3.07, 6.23]	4.64 [3.09, 6.24]	2.57 [1.56, 4.80]	<0.01
Male (%)	4304 (67.1)	4265 (67.1)	39 (62.9)	0.57
Race (%)				0.24
Black	669 (10.4)	662 (10.4)	7 (11.3)
White	4272 (66.6)	4231 (66.6)	41 (66.1)
Other	920 (14.3)	915 (14.4)	5 (8.1)
Not Reported	558 (8.7)	549 (8.6)	9 (14.5)
Hispanic or Latino (%)	1506 (23.5)	1495 (23.5)	11 (17.7)	0.36
Insurance (%) Public	2871 (44.7)	2846 (44.8)	25 (40.3)	0.56
Private	3203 (49.9)	3171 (49.9)	32 (51.6)
Other	345 (5.4)	340 (5.3)	5 (8.1)
Hospital Admission (%)	1245 (19.4)	1232 (19.4)	13 (21.0)	0.88
CBC (%)	4604 (71.7)	4555 (71.7)	49 (79.0)	0.25
ESR (%)	4511 (70.3)	4463 (70.2)	48 (77.4)	0.27
CRP (%)	4301 (67.0)	4255 (66.9)	46 (74.2)	0.28
Blood Culture (%)	2448 (38.1)	2417 (38.0)	31 (50.0)	0.07
Arthrocentesis (%)	310 (4.8)	307 (4.8)	3 (4.8)	1
Radiograph (%)	4535 (70.6)	4495 (70.7)	40 (64.5)	0.36
Ultrasound (%)	3343 (52.1)	3310 (52.1)	33 (53.2)	0.96
MRI (%)	37 (0.6)	36 (0.6)	1 (1.6)	0.81
CT (%)	12 (0.2)	12 (0.2)	0 (0.0)	1

IQR, interquartile range; CBC, complete blood count; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; MRI, magnetic resonance imaging; CT, computed tomography.

was the CBC (72% of encounters), followed by ESR (70%), and CRP (67%). Blood cultures were performed in 38% of encounters, Lyme disease test- ing in 9%, and procalcitonin was only performed in 3% of encounters. The median percentage of serum laboratory testing performed by indi- vidual hospital EDs was 80%, ranging from 47% to 93% by site. Hospital variation in individual laboratory tests is shown in Fig. 3.

Eighty-seven percent (n = 5563) of encounters were associated with Radiologic testing. The most commonly performed radiologic test of the hip, pelvis, or lower extremity was radiography (71% of

Fig. 2. A: age distribution of children with TS, B: age distribution of children with missed bacterial musculoskeletal infection.

encounters), ultrasound was less common (52% of encounters), and MRI and CT were rare (0.6% and 0.2% of encounters, respectively). The median percentage of radiographs performed by individual hospital EDs was 69%, ranging from 53% to 90%. The median percentage of hip ultrasound use by site was 53% and ranged from 2% to 92%. Most sites did not perform any CTs or MRIs (median = 0, maximum CT % = 1, maximum MRI % = 6). Site to site variation in the various radiologic studies is shown in Fig. 3.

Arthrocentesis was done in 5% (n = 310) of encounters, though individual sites varied (median site: 4%, range: 0-19%). 19% of all index TS encounters resulted in hospital admission, with consider- able variation by site (median 20%, range: 0-47%). hospital length of stay for children admitted to the hospital with TS was gen- erally short (median LOS: 1 day, IQR 1-2 days). Between hospital variation in arthrocentesis and admission is shown in Fig. 3.

Rates of laboratory testing, imaging, and procedures by age are shown in Table 2. Notably, serum laboratory testing (67% vs 79%, p <=0.01), ultrasound (46% vs 54%, p <=0.01), and hip arthrocentesis (3.4% vs 5.3%, p <=0.01), were less frequently performed in children

<3 years old.

Use of ultrasound increased every year in the study (from 42% in 2016 to 62% in 2021). Use of all other diagnostic modalities remained relatively stable over the study period (Supplementary Fig. S1).

1. Discussion

In this multicenter, observational study of children diagnosed with TS of the hip in the ED, we found that 1% of children presented again within two weeks with a missed bacterial musculoskeletal infection. These children were significantly younger than the overall cohort with TS. We found that use of laboratory tests commonly included in clinical prediction algorithms for septic arthritis (such as CBC, ESR, and CRP) was common. Most TS encounters included radiographs; ultrasound use was variable across sites and increased in use year-to-year.

This study is the first to quantify risk of missed bacterial musculo- skeletal infection in children diagnosed in the ED with TS. Our results are important because existing clinical Prediction tools used to distin- guish TS from septic arthritis and other serious infections were gener- ally not ED-based in their design, and thus their accuracy in the ED setting has been unclear [29]. The 1% risk of missed infection we

Fig. 3. Use of emergency department diagnostic testing, arthrocentesis, and hospital admission by PHIS hospital.

CBC, complete blood count; ESR, erythrocyte sedimentation rate; CRP, C-reactive protein; CK, creatine kinase; CT, computed tomography; MRI, magnetic resonance imaging.

Table 2

Diagnostics and management of transient synovitis by age.

Age (years)	N	Admission	Bloodwork	X-Ray	US	CT	MRI	Any imaging	Arthrocentesis
1	690	19.9%	66.5%	66.2%	41.9%	0.3%	0.4%	81.0%	2.9%
2	848	19.3%	68.2%	70.3%	48.9%	0.0%	0.5%	85.6%	3.8%
3	1039	18.0%	74.9%	69.9%	52.1%	0.1%	0.0%	86.3%	4.5%
4	1071	17.7%	79.1%	70.9%	54.8%	0.1%	0.3%	87.5%	5.8%
5	981	19.0%	80.4%	70.8%	54.9%	0.2%	0.4%	89.0%	5.6%
6	685	19.1%	81.2%	71.2%	54.5%	0.1%	0.4%	86.3%	5.1%
7	489	20.4%	80.4%	72.4%	53.2%	0.6%	1.2%	89.2%	5.3%
8	301	23.9%	84.4%	77.1%	59.1%	0.3%	1.3%	89.7%	5.0%
9	182	25.8%	79.1%	69.2%	52.7%	0.0%	2.7%	83.5%	4.4%
10	133	23.3%	82.7%	76.7%	48.9%	0.8%	3.8%	91.7%	7.5%

X-ray, radiograph; US, ultrasound, CT, computed tomography; MRI, magnetic resonance imaging.

found is consistent with the range of a 0.1-5.3% risk of septic arthritis in children with zero or one predictor described by Kocher et al. in their original prediction rule, suggesting that commonly used prediction rules may function adequately in the pediatric ED sample [10]. The chil- dren with missed musculoskeletal bacterial infections in our study were not different than children with TS in most demographic categories or in diagnostic evaluation at their index visit, except that they were signifi- cantly younger. This increased rate of missed bacterial infections in the younger age group contrasts with the lower observed rates of serum laboratory testing, imaging, and arthrocentesis in younger chil- dren. Both the higher risk of missed infection as well as the reduced rates of diagnostic testing in this age group may be related to the diffi- culties in performing a reliable history and physical exam in pre- verbal and newly ambulatory children. As age is not considered in the Kocher prediction rule, clinicians may wish to have lower thresholds for further investigatory studies in children <3 years of age, especially if they have other higher risk features.

Admission rates in children with TS were highly variable between sites. Multiple factors could influence the decision to admit including local protocols (ie, admission for imaging or specialist consultation if not available 24 h a day), patient presentation (older children unable to walk or be cared for at home if non-ambulatory), or provider prefer- ence to admit to observe for clinical deterioration. While the risk of missed diagnosis was similar in children admitted compared to children discharged home, admission for observation may still be a useful diag- nostic tool; children admitted for observation and diagnosed with a bac- terial musculoskeletal infection during their admission would not receive the diagnosis of TS, and not be represented in our cohort.

The high rates of CBC and ESR use across all sites suggest that many clinicians are using the Kocher criteria [10,11] to risk-stratify children diagnosed with TS. The use of CRP was slightly less common and more variable, which may reflect the lack of its formal adoption into the Ko- cher prediction rule. Although recent work has suggested that there is a high utility of blood culture in febrile limping children [32], the use of blood culture was less common and highly variable between sites. Variation between sites may relate to Blood draw protocols, as blood cultures are not generally drawn off of existing intravenous lines, and therefore may be performed with the initial risk-stratification labora- tory studies to prevent a second venipuncture.

Radiographs were commonly used across all sites throughout the study period, likely due to the importance of ruling out fracture in the limping child. In contrast, the use of ultrasound was highly variable across sites, and increased in use each year. The increase in ultrasound use is likely due to increased availability at particular sites, however some increase may be due to ultrasound’s role in new clinical decision rules. In 2021, Zoabi et al. published a decision support algorithm using point-of-care ultrasound to diagnose TS in the ED. [33] Rather than serum laboratory testing, this algorithm uses the presence of a hip effu- sion on ultrasound, in combination with low-risk clinical features (age 3-10, lack of fever, ability to bear weight, symptoms <7 days in duration, and lack of ill appearance) to rule in TS, with a missed infection in only one of 1461 children in their cohort. This strategy could decrease other- wise unnecessary IV placement and laboratory work up. While the use of ultrasound in the diagnosis of TS is increasing, its effect on outcomes is unclear. Therefore, the cost-effectiveness of ultrasound in the diagnosis of TS, especially if not used in place of other imaging modalities or labo- ratory testing requires further evaluation. To date, no clinical prediction algorithms have successfully incorporated ultrasound with serum labo- ratory testing to aid in the evaluation of the limping child.

Our study has several limitations. First, our study was composed of

children who were diagnosed with TS at their First visit, and therefore cannot be generalized to all children presenting the ED with hip pain or limp, some of whom would have been diagnosed with septic arthritis or other more serious illnesses at the initial visit. However, our findings do describe the diagnostic methods that clinicians use to arrive at the di- agnosis of TS and can accurately determine rates of missed diagnosis in

children diagnosed with TS. Along these lines, our data were collected retrospectively which may introduce bias, especially in terms of work- up performed at the initial visit; for example, if imaging or blood- work led to the diagnosis of septic arthritis, the TS diagnosis would not have been assigned and the encounter would not have been in- cluded in our cohort. Our data were extracted from the administrative PHIS database, and therefore we do not have access to Physical exam findings or vital signs, and cannot assess whether historical patient fac- tors such as fever or inability to bear weight contributed to diagnostic work up. In addition, the PHIS database does not contain the results of laboratory tests or imaging, so we cannot evaluate risk of missed bacte- rial infections based on particular Laboratory test results or imaging findings. As our cohort was comprised of children evaluated at tertiary children’s hospitals, some children may have had laboratory studies, im- aging, or procedures performed at other hospitals prior to transfer, which we cannot account for. Our study relied on administrative diag- nostic codes, which can have variable accuracy [35]. The PHIS data can only identify procedures that were billed for, and some providers may not bill for all procedures, particularly point-of-care ultrasound, so rates of imaging and testing may be higher than reported [36]. We relied on follow-up visits in the same institution to determine missed diagno- ses. It is possible some children had revisits at other institutions, al- though this is likely a Rare occurrence [37,38]. Finally, the hospitals in our study are tertiary-care pediatric hospitals, and the practice patterns and patient characteristics may not generalize to community EDs.

In conclusion, we found a one-percent risk of missed bacterial musculoskeletal infection in children diagnosed with TS in the ED. Laboratory tests used in the Kocher criteria were most often used to diagnose TS, but ultrasound has increased in frequency. The util- ity and cost-effectiveness of ultrasound, compared to the Kocher criteria, requires further investigation.

Supplementary data to this article can be found online at https://doi. org/10.1016/j.ajem.2022.08.022.

Disclaimer

The views expressed in this article are those of the authors and not an official position of their institutions.

Financial disclosure

The authors have no financial disclosures to report.

Contributors’ statement

MJL and PSS conceptualized and designed the study, carried out the analyses, interpreted the results, and reviewed and revised the manu- script. MJL drafted the initial manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

CRediT authorship contribution statement Matthew J. Lipshaw: Writing - original draft, Supervision, Method-

ology, Investigation, Formal analysis, Conceptualization. Patrick S.

Walsh: Writing - review & editing, Methodology, Investigation, Formal analysis, Data curation, Conceptualization.

Declaration of Competing Interest

There are no conflicts of interest for any of the investigators.

Acknowledgements

The authors would like to thank Dr. Benjamin Kerrey for his review of the manuscript.

References

1. Nouri A, Walmsley D, Pruszczynski B, Synder M. Transient synovitis of the hip: a comprehensive review. J Pediatr Orthop B. 2014;23(1):32-6. https://doi.org/10. 1097/BPB.0b013e328363b5a3.
2. Do TT. Transient synovitis as a cause of painful limps in children. Curr Opin Pediatr. 2000;12(1):48-51. https://doi.org/10.1097/00008480-200002000-00010.
3. Landin LA, Danielsson LG, Wattsgard C. Transient synovitis of the hip. Its incidence, epidemiology and relation to Perthes’ disease. J Bone Joint Surg Br. 1987;69(2): 238-42. https://doi.org/10.1302/0301-620x.69b2.3818754.
4. McCarthy JJ, Noonan KJ. Toxic synovitis. Skeletal Radiol. 2008;37(11):963-5. https:// doi.org/10.1007/s00256-008-0543-3.
5. Naranje S, Kelly DM, Sawyer JR. A systematic approach to the evaluation of a limping child. Am Fam Physician. 2015;92(10):908-16.
6. Luhmann SJ, Jones A, Schootman M, Gordon JE, Schoenecker PL, Luhmann JD. Differ- entiation between septic arthritis and transient synovitis of the hip in children with clinical prediction algorithms. J Bone Joint Surg Am. 2004;86(5):956-62. https://doi. org/10.2106/00004623-200405000-00011.
7. Obana KK, Murgai RR, Schur M, Broom AM, Hsu A, Kay RM, et al. Synovial fluid cell counts and its role in the diagnosis of paediatric septic arthritis. J Child Orthop. 2019; 13(4):417-22. https://doi.org/10.1302/1863-2548.13.190022.
8. Cohen E, Katz T, Rahamim E, Bulkowstein S, Weisel Y, Leibovitz R, et al. Septic arthri- tis in children: updated epidemiologic, microbiologic, clinical and therapeutic corre- lations. Pediatr Neonatol. 2020;61(3):325-30. https://doi.org/10.1016/j.pedneo. 2020.02.006.
9. Liberman B, Herman A, Schindler A, Sherr-Lurie N, Ganel A, Givon U. The value of hip aspiration in pediatric transient synovitis. J Pediatr Orthop. 2013;33(2):124-7. https://doi.org/10.1097/BPO.0b013e31827268b8.
10. Kocher MS, Zurakowski D, Kasser JR. Differentiating between septic arthritis and transient synovitis of the hip in children: an evidence-based clinical prediction algo- rithm. J Bone Joint Surg Am. 1999;81(12):1662-70. https://doi.org/10.2106/ 00004623-199912000-00002.
11. Kocher MS, Mandiga R, Zurakowski D, Barnewolt C, Kasser JR. Validation of a clinical prediction rule for the differentiation between septic arthritis and transient synovitis of the hip in children. J Bone Joint Surg Am. 2004;86(8):1629-35. https://doi.org/10. 2106/00004623-200408000-00005.
12. Ryan DD. Differentiating transient synovitis of the hip from more urgent conditions. Pediatr Ann. 2016;45(6):e209-13. https://doi.org/10.3928/00904481-20160427-01.
13. Hunter S, Kioa G, Baker JF. Predictive algorithms in the diagnosis and management of pediatric hip and periarticular infection. J Bone Joint Surg Am. 2022. https://doi.org/ 10.2106/jbjs.21.01040.
14. Nickel AJ, Bretscher BS, Truong WH, Laine JC, Kharbanda AB. Novel uses of traditional algorithms for septic arthritis. J Pediatr Orthop. 2022;42(2):e212-7. https://doi.org/ 10.1097/bpo.0000000000002024.
15. Singhal R, Perry DC, Khan FN, Cohen D, Stevenson HL, James LA, et al. The use of CRP within a clinical prediction algorithm for the differentiation of septic arthritis and transient synovitis in children. J Bone Joint Surg Br. 2011;93(11):1556-61. https:// doi.org/10.1302/0301-620x.93b11.26857.
16. Caird MS, Flynn JM, Leung YL, Millman JE, D’Italia JG, Dormans JP. Factors distin- guishing septic arthritis from transient synovitis of the hip in children. A prospective study. J Bone Joint Surg Am. 2006;88(6):1251-7. https://doi.org/10.2106/jbjs.E. 00216.
17. Gamalero L, Ferrara G, Giani T, Cimaz R. Acute arthritis in children: how to discern between septic and non-septic arthritis? Children (Basel). 2021;8(10). https://doi. org/10.3390/children8100912.
18. Baskett A, Hosking J, Aickin R. Hip radiography for the investigation of nontraumatic, Short duration hip pain presenting to a children’s emergency department. Pediatr Emerg Care. 2009;25(2):78-82. https://doi.org/10.1097/PEC.0b013e318196ea43.
19. Heylen CE, Docquier PL, Dumitriu D. Transient synovitis of the hip : is systematic ra- diological screening necessary for the detection of Perthes disease? Acta Orthop Belg. 2021;87(2):263-8.
20. Safdar NM, Rigsby CK, Iyer RS, Alazraki AL, Anupindi SA, Bardo DME, et al. ACR ap- propriateness criteria((R)) acutely limping child up to age 5. J Am Coll Radiol. 2018; 15(11s):S252-62. https://doi.org/10.1016/j.jacr.2018.09.030.
21. Dubois-Ferriere V, Belaieff W, Lascombes P, de Coulon G, Ceroni D. Transient synovi- tis of the hip: which investigations are truly useful? Swiss Med Wkly. 2015;145: w14176. https://doi.org/10.4414/smw.2015.14176.
22. Plumb J, Mallin M, Bolte RG. The role of ultrasound in the emergency department evaluation of the acutely painful pediatric hip. Pediatr Emerg Care. 2015;31(1): 54-8. quiz 9-61. https://doi.org/10.1097/pec.0000000000000332.
23. Takundwa P, Chen L, Malik RN. Evaluation of hip pain and Management of Toxic Sy- novitis in the ultrasound era. Pediatr Emerg Care. 2021;37(1):34-8. https://doi.org/ 10.1097/pec.0000000000002314.
24. Thom C, Ahmed A, Kongkatong M, Moak J. Point-of-care hip ultrasound leads to ex- pedited results in emergency department patients with suspected septic arthritis. J Am Coll Emerg Physicians Open. 2020;1(4):512-20. https://doi.org/10.1002/emp2. 12167.
25. Zamzam MM. The role of ultrasound in differentiating septic arthritis from transient synovitis of the hip in children. J Pediatr Orthop B. 2006;15(6):418-22. https://doi. org/10.1097/01.bpb.0000228388.32184.7f.
26. Kiran M, Mohamed S, Newton A, George H, Garg N, Bruce C. Pelvic pyomyositis in children: changing trends in occurrence and management. Int Orthop. 2018;42(5): 1143-7. https://doi.org/10.1007/s00264-017-3746-1.
27. Nguyen A, Kan JH, Bisset G, Rosenfeld S. Kocher criteria revisited in the era of MRI: how often does the Kocher criteria identify underlying osteomyelitis? J Pediatr Orthop. 2017;37(2):e114-9. https://doi.org/10.1097/bpo.0000000000000602.
28. Toner E, Khaled A, Nakhuda Y, Mohil R. The limping child, a rare differential: Pyomyositis of the iliacus muscle -a case report. J Orthop Case Rep. 2019;9(2): 21-5. https://doi.org/10.13107/jocr.2250-0685.1352.
29. Tu J, Gowdie P, Cassar J, Craig S. Test characteristics of history, examination and in- vestigations in the evaluation for septic arthritis in the child presenting with acute non-traumatic limp. A systematic review. BMJ Open. 2020;10(12):e038088. https://doi.org/10.1136/bmjopen-2020-038088.
30. Children’s Hospital Association. PHIS. www.childrenshospitals.org; 2022. [accessed November 29 2021].
31. Feudtner C, Feinstein JA, Zhong W, Hall M, Dai D. Pediatric complex chronic condi- tions classification system version 2: updated for ICD-10 and complex medical tech- nology dependence and transplantation. BMC Pediatr. 2014;14:199. https://doi.org/ 10.1186/1471-2431-14-199.
32. El Helou RE, Kimia A, Harper MB, Joglekar N, Landschaft A. The utility of a blood cul- ture in the evaluation of the febrile limping child. Denver: Pediatric Academic Societies Annual Meeting; 2022..
33. Zoabi M, Kvatinsky N, Shavit I. Evaluation of a Point-of-care ultrasonography decision-support algorithm for the diagnosis of transient synovitis in the pediatric emergency department. JAMA Netw Open. 2021;4(7):e2116915. https://doi.org/ 10.1001/jamanetworkopen.2021.16915.
34. Gorelick MH, Knight S, Alessandrini EA, Stanley RM, Chamberlain JM, Kuppermann N, et al. Lack of agreement in pediatric emergency department discharge diagnoses from clinical and administrative data sources. Acad Emerg Med. 2007;14(7):646-52. https://doi.org/10.1197/j.aem.2007.03.1357.
35. Ng C, Payne AS, Patel AK, Thomas-Mohtat R, Maxwell A, Abo A. Improving point-of- care ultrasound documentation and billing accuracy in a pediatric emergency de- partment. Pediatr Qual Saf. 2020;5(4):e315. https://doi.org/10.1097/pq9. 0000000000000315.
36. Alpern ER, Clark AE, Alessandrini EA, Gorelick MH, Kittick M, Stanley RM, et al. Re- current and high-frequency use of the emergency department by pediatric patients. Acad Emerg Med. 2014;21(4):365-73. https://doi.org/10.1111/acem.12347.
37. Supat B, Brennan JJ, Vilke GM, Ishimine P, Hsia RY, Castillo EM. Characterizing pedi- atric high frequency users of California emergency departments. Am J Emerg Med. 2019;37(9):1699-704. https://doi.org/10.1016/j.ajem.2018.12.015.