Fussy baby presenting to ED poses diagnostic challenge

May 2008

An 18-month-old girl presented to the emergency department with a complaint of “fussiness” and a refusal to walk or bear weight. She was in her normal state of good health until two days prior, when she developed a fever to 102° F, appeared restless and uncomfortable and had a significant decrease in her appetite.

The following day, her childcare provider noticed that she seemed to avoid movement of her right leg and cried easily with any physical activity. That evening, she refused to walk or bear weight on either leg. Her family saw no evidence of injury to either leg. She had no history of recent trauma or any preceding systemic illnesses.

Her past medical history was unremarkable, with no previous hospitalizations or surgical procedures. She took only cetirizine hydrochloride (Zyrtec, McNeil-PPC Inc.) for seasonal allergies, received no other chronic medications and had no known drug allergies. The family history was noncontributory. She lived at home with her parents and an older sibling, who were all healthy. The patient was fully vaccinated to date.

Normal exam

Upon arrival to the ED, she had an axillary temperature of 100.9° F, a heart rate of 148 beats per minute, a respiratory rate of 28 breaths per minute and oxygen saturation of 98% on room air. She was fussy and cried when examined but was easily consoled when held by her mother. Her pupils were round and reactive, her tympanic membranes were clear, she had moist mucous membranes, and her neck was supple and nontender to palpation. No cervical or systemic adenopathy was noted. Her lungs were clear to auscultation bilaterally. Her cardiac exam was significant only for mild tachycardia, but no murmur, rubs or gallops were noted. Her abdomen was soft, nontender, without any mass or organomegally.

General inspection of her back, hips, pelvic girdle and lower extremities was unremarkable, with no evidence of erythema, warmth, swelling or ecchymosis. In particular, no tenderness to palpation or on range of motion of all the major joints could be elicited. Despite this apparently normal exam, she would not stand or bear weight for the examiner. Later during sleep, her hips were reexamined, and no asymmetry or abnormalities could be detected. Range of motion was normal, and vigorous movement did not appear to disturb her sleep.

Plain films of her chest and abdomen, as well as anterior, posterior and frog leg views of her hips and pelvis were all normal. A urinalysis was within normal limits. As the radiographs and physical exam were reassuring, no additional evaluation was performed. She was sent home with a scheduled follow-up appointment the next morning in the general pediatric clinic.

Lumbar puncture

Overnight, she was febrile again to 102° F and was increasingly irritable. The following morning in the clinic on physical examination, she was found to have an axillary temperature of 100.7° F, heart rate of 160 beats per minute and respiratory rate of 40 breaths per minute. She still refused to bear weight on either leg, and she cried with both neck flexion and passive hip rotation. She maintained her trunk in a generally rigid position and preferred to keep her right lower extremity flexed, abducted and externally rotated.

Due to concerns regarding her unusual neck and back rigidity, a lumbar puncture was quickly performed, which revealed a cerebrospinal fluid (CSF) glucose of 90 mg/dL, protein of 18 mg/dL, two white blood cell counts (WBCs) and one red blood cell count/mm³. No organisms were seen on Gram’s stain of the CSF. A complete blood count reported a WBC of 21,100/mm³, with a differential of 54% segmented neutrophils, 15% bands, 28% lymphocytes and 3% monocytes. The hemoglobin was 11.8 g/dL, the hematocrit was 35%, and the platelet count was 429,000/mm³. A C-reactive protein (CRP) was 18.3 mg/dL, and erythrocyte sedimentation rate (ESR) was 67 mm/hour. A urinalysis, urine culture and blood culture were obtained. The urinalysis was again normal.

The child was given intravenous ceftriaxone at the time of the lumbar puncture and admitted to the pediatric ward. What further evaluation would you perform at this time? Would you continue empiric IV antibiotics?

The child was initially continued on IV ceftriaxone at a dose of 50 mg/kg every 12 hours during the first 48 hours of admission while results of blood, urine and spinal fluid cultures were pending. She defervesced within 12 hours of admission and remained afebrile afterward.

A bilateral hip ultrasound was performed within a few hours after admission and revealed normal joint spaces with only scant (normal) symmetrical amounts of synovial fluid. The results of serial physical exams during the next 48 hours indicated that her pain appeared to now be more localized to her sacral and lumbar spine. Magnetic resonance imaging of the lower thoracic and lumbar spine, coccyx and pelvis was performed on the third hospital day.

The results revealed a markedly inflamed right sacroiliac joint, with extensive local myositis extending both anterior and posterior to the joint. No drainable fluid collections were noted, and the adjacent bone marrow of the sacrum and ileum did not indicate regional osteomyelitis.

Antibiotic therapy was changed to IV clindamycin. The patient’s blood, CSF and urine cultures failed to grow an organism. On hospital day five, the CRP had dropped to 3.4 mg/dL, though the ESR remained 67 mm/hour. That same day, she took several steps in the playroom and rapidly improved through the remainder of her hospitalization.

The patient was discharged to home on hospital day six and received a two-week course of IV clindamycin, followed by six weeks of oral clindamycin. On follow-up at the end of this course of therapy, she was asymptomatic and had resumed her normal daily activities.

Answer

Pyogenic sacroiliitis is an uncommon infection in young children that accounts for only 1.5% of all suppurative arthritis.

Unfortunately, due to its often insidious onset and less specific initial physical findings, diagnosis is frequently delayed. The average interval from onset of symptoms to definitive diagnosis in one case series was 3.9 weeks. In the preantibiotic era, pyogenic sacroiliitis was associated with a high level of morbidity and even some mortality. Today, most morbidity can be reduced by making a timely diagnosis.

Epidemiology

Pyogenic sacroiliac infections are actually much more common in young adults, with a mean age at presentation of 20 years, but they have been reported in children of all ages. The infection preferentially affects boys more often than girls. Other associated risk factors include recent significant trauma, pregnancy and IV drug abuse, which may explain the peak in young adulthood. It has also been reported as a local extension of regional skin and soft tissue infections, pelvic osteomyelitis and urinary tract infection. However, despite all these risk factors, the majority of pyogenic sacroiliitis cases are not associated with a predisposing condition. It is believed that most infections are due to hematogenous seeding of the sacroiliac joint, a process which may be enhanced by increased joint laxity from pregnancy or prior traumatic inflammation.

As with other pediatric musculoskeletal infections, the most common organism recovered is Staphylococcus aureus, though Streptococcus pneumoniae, Escherichia coli, Salmonella and Brucella species, Pseudomonas aeruginosa and Mycobacteria infections have also been reported. Pseudomonas infections of the sacroiliac joint are most often found in young adults with IV drug abuse.

Clinical manifestations

The initial presenting symptoms of pyogenic sacroiliitis are vague and nonspecific. Older patients frequently complain of buttock pain and develop a limp. Signs and symptoms in the younger patient may include refusal to bear weight, fever, abdominal pain and poorly localized hip and thigh pain. The sacroiliac joint is in close proximity to the retroperitoneum and is crossed by the sacral, subgluteal and obturator nerves, which leads to the wide variety of locations to which pain may be referred. An interesting finding that may lead the astute clinician to a timely diagnosis is that young children with pyogenic sacroiliitis often prefer to keep their ipsilateral leg extended and externally rotated or flexed at the hip and abducted when in the supine position. The preferential position of the ipsilateral leg is believed to be an attempt to relieve the frequent associated spasm of the piriformis muscle, which also sits near the sacroiliac joint.

The most common specific physical exam finding in pyogenic sacroiliitis is tenderness to palpation localized over the posterior aspect of the sacroiliac joint. This is unfortunately often missed due to the reluctance of examiners to physically turn the patient over during the exam, as such movement often exacerbates an already painful condition.

Localization of inflammation to the sacroiliac joint can be accomplished with the FABERE maneuver. FABERE is an acronym for Flexion of the knee with ABduction and External Rotation of the hip of the affected side and Extension of the contralateral leg. This maneuver is performed by placing the patient supine and placing the ipsilateral lateral malleolus above the patella of the opposite knee (with that leg extended). Outward and backward pressure is placed on the flexed knee, which compresses the sacroiliac joint. If pain is elicited, it is a strong indication of joint inflammation. This maneuver will also elicit pain in the case of hip joint inflammation and is more suggestive of sacroiliac inflammation when passive range of motion of the hip can be performed without pain. Age of the patient is a factor in the ability to use this test, as young infants may not cooperate or be able to communicate the location of pain. Another maneuver that is more feasible in older patients is to elicit regional tenderness through digital rectal examination. Younger children, as the child in this case, may have general discomfort on palpation of the lower abdomen, though this again is often nonspecific.

Laboratory, imaging, studies

It is common to find at least moderate elevation of the ESR (>30 mm/hour) and marked elevation of CRP levels. In addition, a polymorphonuclear leukocytosis is often found. Blood cultures are positive in a greater percentage than seen with other forms of arthritis, with rates of recovery as much as 50% in some case series. Although the results of these tests may raise the index of suspicion for a musculoskeletal infection, they are not particularly specific for pyogenic sacroiliitis. Often, multiple anatomic imaging studies may be required to confirm the diagnosis. Plain films of the sacroiliac joint are often initially inconclusive. Radionuclide bone scans are useful in identifying the sacroiliac joint as the focus of a possible deep-seated infection, though usually need to be followed by a more specific anatomic study. As with most other musculoskeletal infections, the MRI has become one of the most valuable tools in confirming the presence of inflammatory joint changes. Aspiration of the sacroiliac joint in the young child is difficult and rarely performed if the diagnosis is clear and the patient is responding to empiric treatment.

Differential diagnosis

The correct identification of the sacroiliac joint as the focus of a pyogenic infection is difficult in all ages. Infants and young children are often initially suspected of having septic arthritis of the hip, intravertebral disc infection (“discitis”), intraabdominal or retroperitoneal infections or even meningitis, as with this patient. Older children may be suspected of having osteomyelitis of the pelvis or femur or possible neoplastic disease. Accurate diagnosis of pyogenic sacroiliitis requires a high level of clinical suspicion and repeated thorough physical examination, as findings often evolve slowly during several days. It is important to avoid becoming falsely reassured by normal plain films and to have a low threshold to obtain a radionuclide scan or an MRI. If ultrasound and/or hip aspiration are negative in young infants being evaluated for potential septic arthritis of the hip, clinicians should quickly consider further imaging, such as an MRI or radionuclide scan in this challenging population.

Specific treatment of pyogenic sacroiliitis follows the general principles of other musculoskeletal infections. Broad-spectrum IV antibiotics with good Staphylococcus coverage should be given empirically and then tailored if positive cultures allow for specific antibiotic sensitivity testing. In the current era, concern must be given for methicillin resistance, but many of the community-acquired methicillin-resistant isolates of S. aureus remain susceptible to clindamycin, which is particularly effective in penetrating the regional pyomyositis (and occasional abscess fluid) frequently found anterior and posterior to well-established pyogenic sacroiliitis. IV drug users should be treated initially with antibiotics effective against P. aeruginosa.

An appropriate assessment of therapeutic response includes careful serial physical examination, monitoring for resolution of fever and pain and interval ESR and CRP levels. A combined course of IV antibiotics, followed by oral antibiotics, should be given for a minimum of four to six weeks, with duration guided by rapidity of improvement and normalizations of inflammatory markers.

The majority of pediatric patients with pyogenic sacroiliitis respond quickly to conservative treatment, with their symptoms resolving in one to two weeks. Unfortunately, those with significant delay in diagnosis and treatment are more likely to have extensive local and retroperitoneal abscess formation, with well-established regional osteomyelitis of the sacrum and ileum. These patients may require surgical intervention. Such patients are more likely to develop permanently fused joints, with associated growth abnormalities in the young and late chronic arthritis-related complaints.

Timely diagnosis of pyogenic sacroiliitis is a challenge to the best clinicians, requiring thorough serial physical exams and appropriate imaging studies if early efforts fail to uncover the cause of such general complaints as refusal to bear weight and/or poorly localized buttock pain. If diagnosed and treated in a timely manner, the prognosis for full recovery is excellent.

For more information:

• Lance Martin, DO, CPT, USA, MC, is a Pediatric Resident at San Antonio Uniformed Services Health Education Consortium.
• Martin Ottolini, MD, Col, USAF, MC, is a Consultant to the USAF Surgeon General for Pediatric Infectious Diseases, Wilford Hall Medical Center, Lackland AFB, Texas.
• Aprin H, Turen C. Pyogenic sacroiliitis in children. Clinical Orthopaedics and Related Research. 1993;287:98-106.
• Attarian DE. Septic sacroiliitis: the overlooked diagnosis. Journal of the Southern Orthopaedic Association. 2001;10:57-60.
• Carlson SA, Jones JS. Pyogenic sacroiliitis. American Journal of Emergency Medicine. 1994;12:639-641.
• Doita M, Yoshiya S, Nabeshima Y, et al. Acute pyogenic sacroiliitis without predisposing conditions. Spine. 2003;18:E384-E389.
• Feldman LS. Salmonella septic sacroiliitis: case report and review. Pediatr Infect Dis J. 2006;25:187-189.
• Sandrasegaran K, Saifuddin A, Coral A, Butt WP. Magnetic resonance imaging of septic sacroiliitis. Skeletal Radiol. 1994;23:289-292.
• Schaad UB, McCracken GH, Nelson JD. Pyogenic arthritis of the sacroiliac joint in pediatric patients. Pediatrics. 1980;66:375-379.