Case Presentation: A 53 year old male Missouri resident with a prior history of intravenous drug use, traumatic subdural hematoma, seizure disorder and a diagnosis of human immunodeficiency virus (HIV) / acquired immunodeficiency syndrome (AIDS) in 1990 presented to his local hospital in April with a 1 day history of fever and confusion. He was hypotensive with pancytopenia, acute kidney injury and metabolic acidosis. Lumbar puncture was unremarkable. He was treated for septic shock of unknown source with intravenous broad-spectrum antibiotics, pressors and a bicarbonate infusion. After 24 hours without improvement, he was transferred to the intensive care unit at our institution due to persistent shock and the development of respiratory failure. CD4 count returned at 11 raising concerns for noncompliance with his antiretroviral therapy so coverage for pneumocystis pneumonia was added. Infectious Disease (ID) consultation was obtained and although it was acknowledged that the low CD4 count might be due to pancytopenia, their working diagnosis was acquired immunodeficiency syndrome. A history of risk factors for tick exposure was elicited from the family so Ehrlichia testing was recommended. The day following admission, unbeknownst to the bedside teams, the patient’s pancytopenia triggered a pathologist review of the peripheral blood smear. Possible intracellular bacteria in a single neutrophil was documented. The next day, the patient’s condition worsened with seizures and decreased blood pressures. The ID service then recommended the addition of empiric intravenous doxycycline. Two days later the Ehrlichia test resulted and was positive. Despite initiation of doxycycline, the patient continued to decline and on the 24th day, he expired due to multi-organ system failure.
Discussion: Ehrlichiosis was first recognized as a human disease in the late 1980s and Missouri leads the nation in reported cases. The case fatality rate is around 1% overall but patients who are co-infected with HIV are at risk of more severe disease. In a series of co-infected patients, 6 of 21 patients did not survive. Delay in diagnosis was much more common in nonsurvivors with a median of 5.5 days in patients who died compared to 1 day in survivors. This patient presented with fever and pancytopenia which is typical for patients with ehrlichiosis. Although ehrlichiosis was considered in the differential on the day of transfer, the risk of diagnostic delay was not fully appreciated by the involved physicians. The primary team was focused on the diagnoses of septic shock and pneumocystis pneumonia. The pathologist did not alert the bedside clinicians to the possibility of an intracellular organism on the smear which can be seen in ehrlichiosis. The ID service recognized that the low CD4 count could be spurious due to pancytopenia but didn’t make the potential causal link between ehrlichiosis and lymphopenia. The patient’s past history of substance of abuse was documented repeatedly and the possibility of medication noncompliance was deemed likely raising the spector of attribution error. The patient’s inability to confirm or refute risk factors for opportunistic infections versus ehrlichiosis also contributed to the diagnostic error.
Conclusions: Coinfection with HIV and ehrlichiosis should be considered a “ticking time bomb” where the need for urgent treatment to prevent death justifies initiation of definitive therapy based on a presumptive clinical diagnosis. Failure to do so can lead to irreversible multi-organ system failure and death.