A Cautionary Tale from “The Tank”

Let this serve as a cautionary tale for our rising residents.

“EXI” or “The Tank” is our holding area for acutely intoxicated patients or those requiring direct, constant observation for de-compensated psychiatric disease, suicidal/homicidal thoughts/actions. The 6 beds housed there see a rapid turnover of patients, most of whom are simply intoxicated and need time to sober. The general “tank” patient meets several requirements for safe discharge, a fairly basic list including: ‘clinical sobriety’, insight into the reason for hospitalization, (1) tolerating oral intake, (2) ambulating with steady gait, and (3) clear communication.

Heed this warning: sick patients do end up in “the tank”.

I will present a recent case and discuss the circumstances under which a patient’s disposition was delayed.

It was an unusually busy Thursday afternoon; the room 9 buzzer seemed to go off every ten minutes for several hours straight. Presenting to triage was a 30’s yo female registered with a chief complaint of “visual changes, SOB that all started after using meth” (directly quoted from the patient’s registration). She was placed in a recliner in bay 24 and connected to a bedside monitor which includes cardiac monitoring, pulse oximetry, and a traditionally less accurate measure of respiratory rate.

On my initial interview the patient states she had injected intravenously a “standard” quantity of methamphetamine. This occurred approximately 1 hour prior to triage intake. Keeping with her triage complaint; the patient relays a history of relatively quick onset SOA, not associated with cough or chest pain, that began after using methamphetamine. She also noted some blurring of her vision. She denied a history of either of these complaints. She denied any medical problems. She states that she takes no medications at home.

The initial exam was fairly unimpressive. The patient was awake, she was talking (at times nonsensical and tangential), and I would document her as “anxious” but in no acute distress. She demonstrated some degree of psychomotor agitation but no hostility towards staff. She was tachycardic to 130 bpm on the bedside monitor, regular, with the appears of sinus tachycardia. She had strong palpable peripheral pulses, a brachial blood pressure was 179/89. Her lungs were clear to auscultation. She was mildly tachypneic to the mid 20’s. Her abdomen was soft and nontender. Her skin showed evidence of peripheral IV drug use with multiple track marks and sclerosed veins. Her mucus membranes were tacky and she requested a glass of water (1 of the 3 basic “sober” requirements checked).

Her initial plan of care was “water” and “time”, planned sober re-evaluation.

“ROOM 9”. Several unstable patients later, several procedures and I re-evaluate the patient. She is sleeping but rousable. Her mentation is stable but not improving. She remains tachycardic, but mildly improved on bedside monitor to the 120s bpm.

A peripheral line and IV fluid bolus were ordered. Initial palpation/landmark guided attempts at peripheral IV were unsuccessful. An ultrasound was placed at bedside for attempts at deeper peripheral venous access.

I was called to the patient’s bedside. She had attempted to ambulate to the bathroom (2 of the 3 basic “sober” requirements checked?) but had stumbled around wildly and needed assistance with ambulation. The clinical condition had deteriorated. She remained tachycardic. Her mental status was waning. Her breathing pattern was deep and labored (Kussmaul-type). A finger stick blood glucose returned at >600 mg/dL. Just like that, everything clicked. This patient wasn’t acutely intoxicated on methamphetamine; which can cause tachycardia, tachypnea, and anxiety; she was in DKA.

She was surprisingly still conscious. She did endorse a history of insulin dependent diabetes when asked directly about the condition. She was unsure of her last insulin use as she had been on an amphetamine binge, her best estimate was “a week ago”.

She had no peripheral venous access after several failed attempts. Phlebotomy was able to obtain blood from her. Screening labs were sent including a serum and urine tox.

She was taken to room 9 where an internal jugular central venous catheter was placed for aggressive volume resuscitation. The MICU service was consulted for admission. Potassium supplementation was started. IV insulin therapy was ordered. Venous blood gas showed a pH <6.7, pCO2 <22.0, undetectable bicarb, uncalculated base excess (extraordinarily negative I’m sure). Broad spectrum abx were initiated pending completion of her laboratory workup. WBC 50k. Urinary tract infection with ESBL E coli. AKI. NSTEMI. Transaminitis (diagnosed with hepatitis C).

Total time between triage arrival and ICU consultation… 4 hours.

We read frequently about hang-ups in diagnosis. In the post-encounter review of this patient I identified several hang-ups in my own medical decision making.

1) The first I will call “triage level” in lieu of some other established name. I urge caution when patients are placed in “the tank”, the hallways, or when the intake nurse/bedside nurse suggests a patient will be a “quick” or “easy” dispo (they rarely are). Also, the assigned triage level 1-5 is not infallible. Do not let the level 5 patient in the hallway put you at ease or lower your guard.

2) Anchoring bias. This patient threw the anchor when she presented with “visual changes, SOB that all started after using meth”. It was easy in the middle of a busy shift to anchor on the provided information. After all, a lot of this patient’s history and exam findings suggested acute methamphetamine intoxication.

3) Provider fatigue. This was an unusually busy weekday shift. Be sure to take time to breath. Don’t rush to pick up the next patient. Focus on the task at hand.

Luckily this patient suffered no long term consequence of her delayed diagnosis. After an brief ICU and inpatient stay the patient was successfully discharged to her home with her boyfriend.

Sick patients do end up in “The Tank”, in the hallways, and in First Care/Fast Tracks. Use caution, keep a broad differential, and re-evaluate frequently.