Wednesday, July 11, 2012

I have an admission for you: I don't know what to do with this patient

I will be the first to admit that emergency departments are overloaded and their docs work very hard.  But sometimes as a hospitalist getting slammed with his tenth admission in the last hour it seems like the ED has a handful of diagnoses, nay excuses, for getting the patient into the hospital rather than back on the street:


Chez Pane - If the patient has pain, pressure, or nonspecific discomfort below the ears and above the navel it is cardiac in nature.  Regardless of duration this requires admission and serial sets of enzymes.  This is actually French for "bring home the bread" as without these admissions the medical system would go bankrupt.

New Moan Ya - A serious infection requiring intravenous antibiotics that is evidenced by an abnormality detected on a chest x-ray by the ED physician regardless of the absence of fever, cough, sputum, or leukocytosis.  The abnormality, called an infiltrate, cannot be detected by anyone else.

CHF-U - Any patient with an elevated BNP has decompensated heart failure, regardless of signs, symptoms, or benign radiography.  These patients get a magical drug called lasiks which makes them urinate a lot so they are often also diagnosed with You Tee Eye.

Old - The patient is older than 65, takes medications, has a completely negative work-up, and wants to go home.  However they are old, and despite having taken care of themselves their entire lives and have no medical complaints they need to be admitted.

Sell You Light Us - If the skin is red and/or swollen it is infection and requires intravenous antibiotics.  Regardless of clinical history, bilaterally, absence of fever, or normal white blood cell count.  Even if they have a mouth, are not septic, and have not been trialled on oral antibiotics.

Sink Opie - Any hint of possible loss of consciousness or vertigo regardless of age, history. risk factors, and work-up has an insanely high risk of sudden cardiac death.

Stroke'n it - Any change in sensation, motor function, sensorium or headache is a cerebral vascular accident.  Regardless of physiological impossibility of the neurological  complaint and ready availability of an MRI.

You Tee Eye - When late for a golf date, any abnormal urine analysis despite absence of urinary symptoms, fever, or leukocytosis is concrete evidence of a urinary tract infection requiring intravenous antibiotics.

I kid. I kid.


Tuesday, June 26, 2012

Tucking In

In the era of shifts rather than calls, the art of "tucking in" a patient is being neglected.  In the past this generated pages to wake you up but now it generates unneeded business for your colleague who is covering for you.  When the "night float", "nocturnist" or "nocturnalist" is getting paged for a sleeping pill, acetaminophen, or a laxative you are getting a reputation of being lazy, short sighted, and a person who wastes clinical temporal resources.  While your therapeutic regimen may treat the patient, their symptoms do not spontaneously alleviate as soon as the medications hit the blood stream, there is a temporal lag.  In addition, your patient may develop new complaints only peripherally related to their presenting diagnosis.  Imagine, if you will, being a traveller trapped in a hotel room, that you forgot your luggage, and you have to call the front desk for permission to get up to use the lavatory.  This is the powerlessness experienced by patients, who may need trivialities to feel better but have to have a doctor's order to get them.  It is critical to see to the diagnosis and intervene but it is a necessity to see to your patient's comfort as well.  By nature, "tucking in" a patient uses adjunctive measures, medications that the benefits of comfort do not outweigh the risks of side effects, adverse reactions, or worsening the hemodynamic status of the patient.


Rate Controlled Unidirectional Gastrointestinal Motility
People don't like vomiting, nurses like it less than anyone else.  People also like a semblance of bowel regularity and consistency, too much and too little are both subject to complaint, a complaint that is geometrically proportional with age. Symptomatically treating nausea, vomiting, diarrhea, and/or constipation does not solve the problem, your diagnostic inquiry and therapeutic intervention must still proceed.  Where I trained ondansetron was the front line antiemetic of choice, however it can only be given twice in one day.  Therefore a breakthrough or back-up agent such as prochlorperazine, promethazine or metoclopramide is a good idea.  If their nausea and emesis is due to a systemic response to ischemia or infection, the management of the the cause will fix their symptoms, the time delay in improving cardiac perfusion and treating that urinary tract infection will be covered by anti-emetics.  If the nausea and vomiting are due to obstruction or ileus you will be better served with a nasogastric tube to low intermittent suction than just anti-emetics, alleviating the distention and pain which may or may not help the actual problem does help their symptoms, without hiding that massive amount of intestinal content that is just waiting to be ejected when their vomiting centers come back on-line.  Lastly, look the side effects of analgesia, opioids are notorious for causing nausea and vomiting, sometimes simply switching your analgesic regimen will help.
On the back end, people like a "normal" bowel routine.  Once the rate, composition, and volume changes they get concerned and want action.  Constipation must be excluded from ileus or obstruction.  In constipation you still have gas and no other symptoms aside from the abdominal discomfort.  Physicians' are the leading cause of in-hospital constipation, we make people take their iron supplement, PPI, and we fluid restrict them.  We address their pain needs without looking out for opiate-induced bowel dysfunction.  Once your sure it's constipation and not its malevolent cousins ileus or obstruction, you can always work on their bowels with bisacodyl, milk of magnesium (MOM), magnesium citrate, or enemas.  If they have kidney disease avoid the magnesium and phosphorus and try lactulose.


Feed Me
I just waxed poetic about this the other day.


Oh Dr. Sandman Bring Me a Dream...
Sleep is a valuable therapeutic tool.  People want to be "knocked out" a lá the late Michael Jackson to get that good nights' rest and get better.  However the layman does not understand that the medications we use as sleep aids can easily push one over into respiratory depression and failure.  On the flip side, a little sleep deprivation can move the recalcitrant patient toward discharge because they sleep better at home.  Hospitals are loud, obnoxious places.  You have a new, equally ill room mate, an open door to a lit corridor, you often times have things stuck to or in your body, you're probably tethered to some who's-is-what-is-it, you have apnea, telemetry, and bed alarms going off, and vital signs being checked at all hours of the night and labs drawn at other times.  IV pumps are loud and voices carry in the halls of healing.  Patients are not allowed to rest and do not feel in control.  This leads to an upset person when you preround at the crack of dawn.  Thus sleeping aids, which don't actually improve sleep but do make people forget that they woke, may be of benefit.  However only use them when you are sure that there is no other reason keeping the patient awake such as pain, anxiety, or delirium and that by giving them a respiratory depressant you will not precipitate respiratory failure.
First trim tethers such as IV fluids, Foley's, and NG-tubes.  Also get rid of alarms you don't need, like the telemetry and apnea alarm on your comfort care patient.  Eliminate unneeded lab draws and foster communication between nurses and patients so that evening vitals can be done before the patient decides to go to bed.  Consider non-chemical adjuncts to foster sleep such as no naps and increased activity during the day, turning the television off at bedtime, switching their phones off, or finding them some boring reading material.  Limit caffeine intake in the later part of the day.  Controlling pain can help with sleep, so make sure you have an adequate analgesic regimen in place.
Where I trained our formulary advocated zolpidem.  This is the pill that one of my nurses referred to as "the pill that turns my sweet 80+ year old patient into a psychotic nudist".  Benzodiazepines can also be used, either long or short acting, but benzo's are known to cause delirium, more so than zolpidem.  Diphenhydramine can also be used but is also deliriogenic.  Generally, if a patient is on benzodiazepines and not sedated I continue them to prevent withdrawal.  If the patient insists that diphenhydramine is the only thing they can take for sleep, give a small dose a try.  But neither of these medications are "goto" drugs for the sleeping aid naive.  Haloperidol is a good choice if their insomnia is due to hyperactive delirium.




Monday, June 25, 2012

The DIE-t Order

The start of a new academic medicine year reminds me of the initial shock of being a 'tern.  My first rotation was ambulatory medicine and I had recurrent waking nightmares about a patient presenting to the clinic with chest pain.  That initial day in ambulatory clinic the first patient simply wanted his medication explained and the second a refill.  Feeling saucy I was ready for the third until he said, "I have chest pain".  With that I ran from the room, forgetting further history or physical, and called my attending to staff, which sounded something like this: "The patient has chest pain.  In the thorax.  You know, in the chest area.  It is painful, in his chest." Awed by this clinical marvel, my attending sighed and took over, tempted to administer a benzodiazepine or a swift slap to the back of the head of the hyperactive intern.
My next rotation was the ICU, where experienced nurses carefully took us by the hand and shepherded the "intensivist interns" away from too much stupidity, tactfully paging me when they needed a critical intervention of acetaminophen or an antiemetic while deferring the more mundane ventilator adjustments and anxiolysis to the attending.
My third month was on the floors at the VA.  On-call.  With a suboptimally motivated or interested senior resident.  The resident had vanished before shift change leaving me to face a barrage of sign-offs and nursing shift change.  My first solo decision was...the diet order.  Never has a physician faced such consternation and mental gymnastics as I did when making this pivotal decision.  I thought back to my clinical nutrition rotation and pondered the need for enteral versus parenteral nutrition.  I weighed his medical problems, each one indicating a specific diet.  The nurse, patience fraying as the minutes ticked by finally threw me a clue with the infamous, "Well last time he was here he had a regular diet". Sold! One regular diet order coming up.
In medicine you need a system for making decisions, all of them are important but some are more important than others.  Some of them need to be made right now and some of them can be made later.  Americans are obsessed with food.  Obesity epidemic being corpulent evidence of the same.  Patients with intractable abdominal pain or who have just been told that a myocardial infarction has destroyed the pumping ability of their heart are more concerned with getting that next megaburger with ubër fries and a diet Coke then the next diagnostic and therapeutic step.  Been vomiting all day after eating? Family says feed her.  Had a stroke and with each meal starts hacking and turning cyanotic? Family says he can't get well if he doesn't eat.
This oral fixation drives nursing to page as soon as the diseased hit the floor, regarding the diet order.  Regardless of the vital signs, presenting complaint, operative or procedural needs the first thing I usually hear is "Can they eat?"  Despite a chronic history of "just say no" I still get the question.  Patients are usually blessedly monosyllabic while they eat, making it a perfect time to get a history or for nursing to get done.  However, food or withholding of the same is a therapeutic decision and should be made after  the physician has determined the problem and the solution for the same.  If there is no aspiration risk consider limited ice chips, patients are much happier if they don't have a dry mouth.  Unless they have been given a large dose of insulin or other anti-hyperglycemic medications, skipping one meal or even a whole days worth (as long as they are adequately intravenously hydrated) has never killed anyone.  If they did you would have a case report of hyperstarvation.
So how do you decide the diet order?
First decide if they can or cannot eat.  They cannot eat if they have a condition that will be worsened with intake or they will have a procedure the next day.  Typically they can eat up until midnight if they are having a procedure or operation.  Conditions that preclude intake are neurological insults (e.g. stroke, aspiration pneumonia) that have caused or increase the chance of aspiration, nausea ± emesis, abdominal pain of suspected gastrointestinal origin, blood coming out of the esophagus or rectum, a blood glucose > 400 mg/dL, and recent surgery or procedure until cleared by the surgeon/proceduralist.
Next choose their diet, basically look at their health problems and choose, e.g. hypertension = 2 g sodium, heart disease history = cardiac (trumps hypertension), diabetes = diabetic, both = cardiac diabetic, on dialysis = renal, hyperkalemia = restrict their potassium (which there is an abundance of in salt substitute, tomatoes, oranges, and bananas).  If they forgot their dentures, puree it.  If they have diet restrictions from the nursing home continue them.  If they have been losing weight, no liver disease, and cancer has been excluded or if they have low serum protein and albumin add some CIB shakes.  When in doubt about the calories ask a dietician to see them, when in doubt about the consistency ask the speech therapist to see them.  The longer you think they are going to live the stricter the diet needs to be.  The octogenarian with advanced cancer probably doesn't need to worry as much about her A1c as the newly diagnosed type I diabetic.
If they cannot eat you will need to come up with some form of nutrition other than the meager benefits of D5*.  As a rule of thumb, three days without food should be the temporal trigger for initiating some form of nutrition.  Always use the gut first if you can, consider a Dobhof for tube feeding if they cannot swallow with consideration for a PEG-tube (G-tube) if they are still deemed unable to swallow.  Be aware that neither tube decreases aspiration risk.  Sometimes the stomach needs to be bypassed and a GJ-tube can be used.  If the etiology of the NPO status precludes enteral feeding, i.e. a gastrointestinally-based intraabdominal etiology, consideration toward parenteral nutrition via a PICC or central line must be undertaken.  With either tube feeding (TF) or total parenteral nutrition (TPN) a dietician consultation is highly recommended.

*The amount of kcal (Cal) per liter of D5W = 5000 mg/dL x 3.4 kcal / g x 1 g / 1000 mg x 10 dL/1 L = 170 kcal/L.  Thus a patient receiving 150 mL/hr of D5 NS would receive 170 kcal/L x 150 mL/hr x 1 L/1000 mL x 24 hours = 612 kcal per day.  As the average caloric intake per day ~2000 kcal this is a starvation "diet" and catabolic.