Case Presentation: A 70-year-old man with a history of bladder cancer treated with intravesical Bacillus Calmette-Guerin (BCG) therapy presented with acute on chronic back pain. Six months prior, he presented with lower back pain, was diagnosed with culture negative bacterial osteomyelitis, and underwent an L3-L5 laminectomy. He was discharged with an empiric 6-week antibiotic course with brief improvement in back pain. He subsequently worsened over six months, developed difficulty with ambulation, and represented to the hospital. On admission, the patient’s vitals were stable. He was uncomfortable-appearing, had exquisite point tenderness in the lumbar spine, and did not have neurologic deficits. Labs were remarkable for elevated inflammatory markers. MRI demonstrated progression of prior osteomyelitis. He was started on empiric broad spectrum antibiotics and admitted to medicine.Patient underwent debridement of his lumbar spine and tissue samples were sent for culture. Due to his refractory disease course, alternative causes for osteomyelitis were considered. Given his history of bladder cancer with intravesical BCG treatment along with imaging findings, mycobacterial infection was suspected. The patient was empirically treated with rifampin, isoniazid, ethambutol, and pyridoxine. Culture results from surgical tissue samples returned positive for mycobacterium, confirming the diagnosis of tuberculosis vertebral osteomyelitis. The patient responded successfully to treatment. He had incremental improvement in his back pain and mobility on outpatient follow up.
Discussion: This case is unique for several reasons. While Pott’s disease is a very well-known pathology, mycobacterium osteomyelitis after intravesical BCG therapy is a rare and morbid complication that requires treatment with antituberculosis agents for resolution. Perhaps the most unique element of this case is the clinical reasoning process that was required to reach this diagnosis. The diagnostic process was susceptible to several cognitive biases that were actively navigated against using type 2 analytical thinking. This patient’s initial working diagnosis was refractory bacterial osteomyelitis, which would have led to a more prolonged course of intravenous antibiotics. To combat this diagnostic momentum, we prevented premature closure by expanding our differential diagnosis. Our active acknowledgement of the possibility of diagnostic anchoring led to our recognition that this patient had undergone appropriate management of bacterial osteomyelitis with prior debridement and an extended completed course of intravenous antibiotics without resolution of symptoms. Though this patient did not have obvious risk factors for a tuberculosis infection, we avoided using the representative heuristic and considered this rare complication after the intravesical BCG therapy. With active acknowledgment of these cognitive biases, we were able to avoid delays in making this rare diagnosis and start prompt treatment.
Conclusions: Vertebral mycobacterial infection is a rare and morbid complication that should be considered in patients with history of intravesical BCG therapy who develop osteomyelitis refractory to antibiotics, independently of other common risk factors of mycobacterium. The path to diagnosis of rare conditions is fraught with cognitive biases. An active understanding of clinical reasoning principles will allow hospitalists to prevent diagnostic errors when approaching similar clinical conundrums.