“Is there anything bothering you? Anything at all?” I asked.
“You,” she said.
She was in there. Gears were turning. But she had no chief complaint.
She seemed to prefer her eyes closed. With each question, she replied in the same manner: by lifting her head, opening her eyes, looking directly in mine, answering “no,” and resuming her slumber.
The Physical. Having already worn out my welcome, I decided to set aside the history and proceed with the physical. Before me was a 56-year-old African-American woman slouched over in a seated position, responsive to voice and following commands, but slow.
Slightly overweight and tachypneic to 32, her breath sounds were clear and equal, her expiratory phase was not prolonged, and she demonstrated no increased work of breathing. Oxygen saturation on facemask with reservoir bag placed by EMS was 100%. Fingerstick glucose was normal. With an effaced left nasolabial fold, weak grip strength in her left hand, and a weak left foot to dorsiflexion, I told the nurse it looked like she had had a stroke. He nodded. My thoughts turned to a head CT and the thrombolytic window.
Fortunately, the family showed up before I could do much harm. Head CT showed chronic white matter changes and small lacunar infarcts. The sister came into the room claiming that our patient had a stroke four years earlier, and hobbled around the house with a paretic limb since that time. She was generally of good nature, alert, interactive, always on the go. This was a significant departure from baseline.
Her daughter added that everything was fine that morning. She heard her mother make her way into the bathroom, and then noticed it was a long while, and she had not come out. “She always takes a long time to go,” said the daughter. “Usually she strains a lot to get anything out.”
Upon entry, they found our patient on the toilet slouched over and unarousable. So they called 911.
The answer must be in the urine. With baseline cerebropaths, a UTI is a favorite suspect for all changes in mentation. We wait anxiously for the urine to declare itself guilty and relieve us of having to think much more. But unfortunately, the catheterized urine specimen we obtained was entirely normal.
Her ECG was sinus rhythm with nonspecific T-wave abnormalities. CBC and chemistry were normal. I was preparing to perform a lumbar puncture.
What am I missing? Some component of the history maybe. I asked more questions, an extensive review of systems. Nothing remarkable. I had to change my approach. The subjective was clearly getting me nowhere, but what about the objective? Rather than words from her or her family, could the clues to her diagnosis lie in the observable signs at bedside?
Tachypnea. Whether the alveolar lumen, bronchioles, or proximal cartilaginous airways, tachypnea leads us to focus on the respiratory tree and pulmonary compliance. Our patient, however, had no increased work of breathing. She was breathing fast and deep, but there was no evidence by physical exam that she had a problem with compliance. In the supine position, she demonstrated no orthopnea. Her portable chest x-ray was normal.
Tachypnea without any corroborating lung findings must be a metabolic acidosis, I declared. Even without an anion gap by chemistry profile, I felt inclined to pursue a metabolic cause to this abnormal vital sign. So I took her off the facemask in preparation for an ABG.
Hypoxemia. Immediately, her arterial saturation trended down and leveled at 82% on room air. “Wow, she's hypoxic,” the nurse said. “What do you make of that?” An overweight patient, who was lethargic and slumped over with her neck flexed, most likely had a component of hypoventilation, I thought. If I sat her up and had her take deep, full breaths, she should reconstitute to above 92%. So I attempted to show my nursing colleague what I meant, that our patient's arterial hypoxemia was simply the result of hypoventilation from depressed mentation. Stimulating her enough to comply, we had our patient maintain an upright posture, and take several deep breaths. We watched the monitor, and cheered on as her saturation began to climb: 82%, 84%, 86% … 86% … 86%. It rose no further. I was in disbelief. What's more, I was beginning to get frustrated with the course of things, history and physical. “The waveform looks good,” said our nurse. “It's probably real.”
A Twist. Up until this moment, our patient was a case of “sudden onset depressed mentation with secondary hypoventilation and hypoxemia.” In that vein, investigations were aimed at determining a cause of rapid change in level of consciousness.
Once we demonstrate that hypoxemia is not from hypoventilation, our approach must change. No longer a sequela of depressed mentation, hypoxemia is now a primary derangement. In fact, it may be the cause, rather than the effect, of altered mental status. The course of the case turns away from cerebral investigations focused on sudden onset lethargy and toward a physiologic deconstruction of arterial hypoxemia.
We realize we must pursue hypoxemia regardless of the results of other investigations that could explain her altered mentation: a positive urinalysis or even a positive CSF. Either can explain lethargy, but outside of associated aspiration or hypoventilation, neither will explain hypoxemia.
Physiology. All causes emerge from four principles: hypoventilation of the alveolar lumen, a V/Q mismatch in favor of perfusion over ventilation (the extreme version being an intrapulmonary or intracardiac shunt), or a problem with the alveolar-capillary membrane impairing diffusion. Whatever our patient or family claimed, whatever symptoms she had, whatever course of events occurred this morning, her hypoxemia had to be explained by one of these principles.
With meaningful chest rise, she wasn't simply partaking in shallow airflow limited to tracheal and bronchial anatomic dead space. She must be moving air into and out of her respiratory bronchioles and alveolar lumen. She was not solely hypoventilated. By reconstituting to 100% on face mask and reservoir bag, a large shunt was unlikely.
Impairment of diffusion across the alveolar membrane secondary to an inflammatory or infectious process unseen by x-ray and unheard by auscultation was possible, but her presentation seemed to be rapid onset this morning, rather than slowly progressive.
We are left with V/Q mismatch. On the ventilation arm, air entry into the alveolar lumen would have to be impaired by something other than generalized hypoventilation: possibly alveolar edema, blood, pus, or gastric aspirate.
On the side of perfusion, blood flow would need to be increased beyond the capability of certain alveoli to deliver oxygen. This is uncommon unless associated with vascular obstruction elsewhere in the lung. Small occlusions produce dead space, not hypoxemia. But large occlusions can increase flow past neighboring segments and overwhelm the saturating ability of the adjacent, possibly collapsed, alveoli. So we must consider pulmonary embolus.
The history subordinate to the physical. She had a paretic limb as a risk factor, but no chest pain, shortness of breath, weakness, or any symptom attributable to PE. Her chief complaint did not exist.
What she did have was tachypnea unrelated to a loss of pulmonary compliance and hypoxemia unexplained by hypoventilation, lung auscultation, or chest x-ray. These findings alone meant we must accept the idea of a PE, regardless of the HPI. Marriage of the history and physical is not always possible, but markedly abnormal objective findings enslave us. They demand an explanation and force us to set aside a history that does not fit. They own our next clinical move.
Her CT pulmonary angiogram was positive for diffuse thromboemboli. She must have harbored a deep venous thrombus in her paretic limb. Straining to move her bowels, the cessation of her valsalva maneuver was characterized by a sudden drop in intrathoracic pressure. And the DVT siphoned into the pulmonary circuit.
Comments about this article? Write to EMN at[email protected].