One day, when I was on a very busy hospitalist service, the ER called me with an admission. "I've got a 35 year old woman, shoots meth, smokes, uncontrolled diabetic, comes in with shortness of breath," the harried ER attending reported. "X-ray looks like a bilateral pneumonia, and some parapneumonic effusions too, but not enough to tap. Oh, and she's pregnant, about 14 weeks by LMP. She's stable but the last bed we have is in ICU, so that's where she's going."
He told me he'd started some antibiotics for community-acquired pneumonia and written for an insulin sliding scale. It was mid-morning and I was rounding on patients to be discharged and patients who needed a lot of management and I couldn't go see my new patient right away. An ICU nurse called me. "I've got your new patient. She's got a temp of 102.7 and doesn't have anything for fever. Can I give her some Tylenol?"
I said yes and returned to my rounds with a feeling of unease. In residency I learned one cardinal rule of medicine: A pleural effusion in a febrile patient must be immediately sampled by thoracentesis, or removal of the accumulated fluid from the chest cavity using a percutaneous needle. Why? Because even though a simple pneumonia can result in the accumulation of some inflammatory fluid in the chest cavity, the fluid itself can become infected and cause a serious complication called empyema. In turn, an empyema can thicken and settle into a very difficult to remove honey-like consistency, which prevents the lung from expanding fully and will ultimately scar and damage the lung. An empyema is managed very differently from a simple parapneumonic effusion; it requires placement of a chest tube to drain the infected fluid out before it progresses further.
As I made rounds on other patients, my uneasiness grew more pronounced. The ER physician told me there wasn't enough pleural effusion to sample, and he'd seen more pneumonia than I had. I broke away from my other patients to take a look at the woman's X-ray. The film was expiratory and I really couldn't see any significant fluid in the chest cavity at all, so maybe the ER doctor was right. There was no lateral decubitus X-ray on the patient, and I thought about ordering one as I returned to my rounds. A lateral decubitus can often reveal a subtle pleural effusion, and helps determine whether the volume of fluid is sufficient for sampling. The following public domain photo demonstrates a large pleural effusion, indicated by arrow A:
Although I had many other patients to see, I went to check on the woman with pneumonia. As I walked into her room, I felt a hair on the back of my neck rising up. Not literally, of course. The hair--and it is only one--has become the symbol for the feeling I get when a patient is really sick. It sprouted during the second year of my residency and has been with me ever since.
The patient was a young woman who was sitting in bed, breathing shallowly and barely able to answer yes or no questions. Her face was greyish in color and she had dark circles under her eyes. I took her history: many years of IV meth use, tried to quit, wanted to quit, even managed to be clean during her pregnancies but relapsed, diagnosed with diabetes during her first pregnancy and it never went away. Didn't know she was pregnant this time, not sure of her LMP.
I examined her. She could only take shallow, mouse-like breaths and it looked as if she had to concentrate all her energy on mobilizing those breaths. I could barely hear any breath sounds at the bases of her lungs. I lost tactile fremitus in the middle of her left lung and heard egophony changes at the left base. When I palpated her belly, the top of her uterus rose well above her belly button, which suggested a gestational age over 20 weeks.
I ran back to Radiology and looked at the CT scan of the chest the ER had obtained. The radiologist was in the middle of a procedure so I had to look at the films myself. I'm not expert in reading CTs but even I could see the dense greyish effusions in both sides of the chest cavity. The image below shows several large pockets of fluid.
Image credit: Banga1 A. et al.: A study of empyema thoracis and role of intrapleural streptokinase in its management. BMC Infectious Diseases 2004, 4:19 
I was jittering mightily by now. I knew I had a sick pregnant woman with large pleural effusions which could well be empyemic and I needed to get the fluid sampled. I didn't want to take the time to get the whole thoracentesis tray out and fiddle with the big needles. I just wanted 60cc of the stuff to send to the lab.
I waylaid an ultrasound tech and he marked a large pocket of fluid in the left hand side of the chest for me. This was just for reassurance. I could have easily gone below the spot where I lost fremitus. Then I assembled the bits and pieces I use for a quick diagnostic thoracentesis, including a big syringe and a spinal needle.
I got the patient's consent and prepped the left side of her back. With a great deal more confidence than I felt, I inserted a needle into her pleural space and drew back a syringe full of fluid. Normally, pleural fluid is clear and straw-colored, like this. What came out of my patient's chest was milky tan/grey, the color of a bad latte.
It was pure pus.
There was no time to be horrified. I got on the phone with a general surgeon and an obstetrician. The former put in a big chest tube within an hour. The latter arranged for the patient to be transferred to a tertiary care center when her 22-week fetus died in utero. There she had a laparotomy to remove the products of conception--done surgically because she'd had two prior C-sections--and then she returned to our hospitalist service.
When I saw her next, I barely recognized her. She had good color in her face, and looked well. She was sad and remorseful about the loss of her pregnancy, and talked realistically about quitting meth and taking care of her diabetes. I worked with her on both of these issues. Then I discharged her back to her life.
The moral of this case: Whatever anyone tells you, check the facts for yourself. One doctor's small effusion is another's nighmare.