Patient X was a 34-year-old, African American woman with a medical history of mental retardation, hypothyroidism, hypertension, and appendectomy. Her hypothyroidism and hypertension were well controlled on her home medication regimen of levothyroxine, lisinopril, atenolol, and hydrochlorothiazide. She had no history of alcohol, illicit drug, or tobacco use. Her immunizations were up to date.
Patient X lived with her parents and was independent or needed minimal assistance with her activities of daily living. She spoke English, although she was difficult to understand at baseline because of her mental retardation. She appeared well nourished and cared for.
Four days before admission, patient X had dyspnea, cough, fever, and wheezing that led her mother to take her to another hospital’s emergency department. According to the treating physician’s summary note, she was given azithromycin for bronchitis. No chest radiography or nasal swab for influenza was done at that time. She was discharged to home. Two days later, patient X stopped the azithromycin because of diarrhea. Two days after that (day 6 of illness), her mother took patient X to a primary care physician when she experienced increasing shortness of breath and wheezing. The primary care physician noted that patient X was febrile, tachypneic, and tachycardic; therefore, she was sent directly to our emergency department for treatment.
Upon arrival, patient X’s respirations were shallow and labored and she was tachypneic. Her initial vital signs were a heart rate of 117/min, a blood pressure of 149/87 mm Hg, a respiratory rate of 28/min, an oxygen saturation of 76% on room air, and a body temperature of 39.2°C. She was immediately fitted with a nonrebreather face mask with a fraction of inspired oxygen (Fio2) of 100%. Her chest radiograph revealed mild to moderate bilateral infiltrates suggestive of an infectious process. Respiratory support was increased to an Fio2 of 100% via bilevel positive airway pressure, but she was able to maintain an oxygen saturation of only 87%. Results of arterial blood gas analysis at that time were a pH of 7.27, a Pco2 of 44 mm Hg, a Po2 of 43 mm Hg, and a bicarbonate concentration of 21 mmol/L with an anion gap of 24.
Patient X was admitted to the medical intensive care unit with a chief complaint of hypoxic respiratory failure. She was immediately sedated and intubated. A notable laboratory test result was a normal white blood cell count of 5400/μL, which, when coupled with the presence of infiltrates, is indicative of an atypical pneumonia.7 The hallmark of an atypical pneumonia is the presence of pulmonary infiltrates with extrapulmonary signs and symptoms; that is, evidence of involvement of organs such as the brain, heart, liver, kidneys, or gastrointestinal tract, often without an elevated white blood cell count.7 Examples include confusion, diarrhea, bradycardia, elevated levels of liver enzymes, or shock.7
Other important laboratory test results for patient X were hyperglycemia (serum glucose level, 165 mg/dL; to convert to millimoles per liter, multiply by 0.0555), elevated level of aspartate aminotransferase (125 U/L; to convert to microkatals per liter, multiply by 0.0167), and a hemoglobin level of 11.5 g/dL. Table 1 summarizes laboratory values for patient X.
Patient X’s list of problems at the time of admission included hypoxic respiratory failure, metabolic acidosis, hyperglycemia, liver dysfunction, and anemia. The immediate plan of care included respiratory support, broad-spectrum antibiotics (ceftriaxone, vancomycin, and azithromycin), oseltamivir phosphate (Tamiflu), testing for influenzas A and B, and correcting the metabolic acidosis. A goal was to rule out Legionella infection as well as viral infection such as H1N1 infection. A mechanical ventilation protocol for adult respiratory distress syndrome (ARDS) was initiated; the protocol included the use of low tidal volumes on an assist-control setting, increased positive end-expiratory pressure, and high Fio2.
Patient X was tested for influenza A and influenza B, but the results were negative. On hospital day 4, another sputum sample was sent to the Maryland State Department of Health and Mental Hygiene to undergo more sensitive testing with the CDC’s Human Influenza Virus Real-Time Polymerase Chain Reaction (RT-PCR) Detection and Characterization Panel. That sample was later ruled positive for “human swine-like influenza virus” or H1N1 virus.8 This flu panel consists of influenza A PCR, influenza B PCR, and subtyping for influenza A. Diagnosis of H1N1 infection requires that the sample must test positive for both influenza A virus RNA and for subtype SWH1 (human swine-like influenza virus) RNA.8
Patient X received mechanical ventilation for the first 8 days of hospitalization. She had multiple episodes of desaturation associated with ARDS; therefore, the ventilator settings were manipulated many times to maximize oxygenation. Bilateral breath sounds were described as either “rhonchi,” “coarse,” or “crackles.” Fluid balance was carefully managed by monitoring the central venous pressure, administering furosemide, and restricting fluid input to avoid further compromising her pulmonary function. Additionally, she occasionally experienced self-limiting bursts of ventricular tachycardia lasting approximately 10 beats per episode.
Antimicrobial therapies were adjusted as Legionella was excluded and H1N1 was verified, but the choice of agents was influenced by an effort to avoid any drugs associated with increased QT intervals on an electrocardiogram. Eventually the antimicrobial agents consisted of oseltamivir phosphate (for H1N1) and cefazolin (added because 1 tube of blood out of 2 full sets of pan cultures tested positive for Staphylococcus hominis).
The healthcare team was vigilant in monitoring and maintaining serum levels of magnesium at greater than 2 mEq/L and serum levels of potassium at greater than 4 mEq/L to prevent further episodes of cardiac arrhythmia. Additionally, the subclavian triple-lumen catheter was pulled back to avoid irritating the heart, but the occasional bursts of ventricular tachycardia continued. During this time, patient X remained sedated, except when sedation was temporarily lightened for examinations.
On hospital day 2, the physician team noted that patient X had become precipitously thrombocytopenic (platelet count, 58 × 103/μL), with an increase in the international normalized ratio, prothrombin time, and activated partial thromboplastin time. Heparin was temporarily discontinued while a heparin PF4 antibody test was completed; the result ruled out heparin-associated antibody syndrome (heparin-induced thrombocytopenia). Treatment with subcutaneous heparin was reinitiated for prophylaxis of deep venous thrombosis.
Her hemoglobin level was 10.0 g/dL and her hemat-ocrit was 33%, which was indicative of anemia. On day 3, patient X’s level of d-dimer was elevated at greater than 20 μg/mL and the platelet count was 52 × 103/μL. The team suspected DIC and gave her vitamin K with a plan to transfuse fresh frozen plasma and platelets if her platelet count decreased to less than 15 × 103/μL. The fresh frozen plasma and platelets were not transfused because her platelet count remained stable, but her hemoglobin level decreased to 8.9 g/dL and her hemat-ocrit decreased to 28%, and therefore packed red blood cells were transfused instead.
On the same day, patient X had some minor oozing of blood from her triple-lumen catheter. For the next 2 days, she showed clinical signs of low-grade DIC (minor oozing of blood and ecchymosis at venipuncture sites), but the results of her laboratory tests were relatively stable and she continued treatment with subcutaneous heparin. Low-grade or chronic DIC represents a compensated state where the liver and bone marrow are able to produce clotting factors and platelets fast enough to not be overwhelmed by the rate of consumption by the hyperactive coagulation cascades of DIC.4
By day 5, her prothrombin time and international normalized ratio remained prolonged but her platelet count had increased to 88 × 103/μL and the activated partial thromboplastin time had begun to improve. It appeared that patient X’s DIC was resolving.
On hospital day 6, the direct care nurse discovered 2 cyanotic toes on patient X’s right foot. On the same day, she reported that patient X was no longer following commands when sedation was lightened. An urgent unenhanced computed tomography scan of the head revealed areas of lucency in the right frontal (1.9 × 1.3 cm), right anteroparietal (5.5 × 3.3 cm), left occipital (3.0 × 3.1 cm), and high left parietal (3.7 × 1.6 cm) lobes without evidence of midline shift or hemorrhage (see Figure). The report described these areas as being consistent with possible masses. A neurology consultation was obtained immediately. Patient X was given a heparin infusion to prevent thromboemboli from forming.
Although stroke was suspected, the healthcare team was diligent in ruling out any other reasons for the computed tomography results and the patient’s corresponding decline. A lumbar puncture was performed in search of evidence of influenza encephalitis, aseptic meningitis, or other opportunistic infection; the results were negative. Magnetic resonance imaging confirmed that the same areas of abnormality on the computed tomography scan were multiple acute and subacute strokes throughout both hemispheres. The Figure shows the location and size of each of the infarcts on the computed tomograms.
Magnetic resonance angiography of the neck and brain showed no evidence of proximal vascular occlusion, but vascular abnormalities were present along the cortical surface at the sites of the infarcts. These abnormalities were indicative of subacute blood products from petechial hemorrhages. Carotid Doppler studies confirmed normal arteries bilaterally. Doppler imaging was used to test for normal and equal flow from the jugular veins; no signs of venous thrombus were apparent. Both transthoracic and transesophageal echocardiograms verified normal size of the left and right ventricles and normal systolic function with an ejection fraction greater than 55%. No evidence was found of a structural defect, shunt, thrombus, or vegetation that could result in cardioemboli.
Because patient X continued with somnolence and decreased responsiveness for several days, an electroencephalogram was obtained to rule out seizures. No indication of seizures was apparent on the electroencephalogram; however, independent areas of slowing that corresponded to the regions of infarct were evident on the imaging studies.
The overall clinical picture painted by the test and laboratory results strongly supported the idea that patient X had had multiple strokes due to DIC. Because H1N1 infection was the only condition this patient had at the time DIC occurred, the healthcare team felt strongly that H1N1 infection was the likely trigger. By day 9, no evidence of DIC was apparent on coagulation studies. Additionally, follow-up magnetic resonance imaging showed no new strokes, and no clinical signs of further bleeding or formation of thromboemboli were present.
Evaluation and treatment by physical, occupational, and speech and language therapists was initiated when patient X was medically able to tolerate these interventions. Patient X had spontaneous eye opening and was intermittently able to follow simple commands, but she was confused and at times her speech was incomprehensible. She had left-sided weakness and could barely move her left extremities against gravity. She had a weak gag reflex and delayed swallowing. Patient X’s dysphagia was severe enough to require a percutaneous endoscopic gastrostomy tube, which was placed on hospital day 12. Care management worked with the family for support and discharge planning. Patient X became stable enough to be transferred to a telemetry unit on hospital day 13, where she stayed until she was ready for discharge.
Patient X’s entire acute care hospital course lasted 25 days. Upon discharge, the family asked to take her home rather than have her sent to a neurological rehabilitation facility. She needed assistance with activities of daily living as well as additional therapy. Therefore, home health care nursing, physical therapy, and speech therapy were continued on an outpatient basis. The consensus of the team was that patient X did not need long-term anticoagulation because her coagulopathy was due to an infection that had resolved. However, daily aspirin 81 mg was prescribed as part of her home medications.
Approximately 1 month after her discharge from the hospital, she returned to have her percutaneous endoscopic gastrostomy tube removed. Although she was not back to baseline, patient X had improved; she was consistently following commands, ambulating with assistance, talking, and able to eat safely. Outpatient therapy continues.