Wednesday, January 15, 2014

Antibiotics in hospice

by: Alex Smith @AlexSmithMD

Case 1: You have a 94 year old woman with multiple medical problems in hospice who develops a fever (subjectively hot to the touch), shortness of breath, and a cough producing yellow sputum.  Her daughter asks if she can be treated with antibiotics "to make her feel better."  The patient is not well enough to make decisions, but in earlier conversations had stated a goal of remaining comfortable at home, while also hoping to live until her first great grandchild is born.  Should you treat with antibiotics?

Case 2: You have an 84 year old man in the hospital being treated for a fungal infection of the heart.  The condition is non-operable and he decides to focus on comfort. The plan is to discharge to hospice.  The infectious disease team recommends a 12 week course of IV antifungals, or at the every least oral anti-fungals.  The patient, who has already had one stroke from a fungal clot in his heart that shot up to his brain, is willing to take the medication.  Should you follow these recommendations?

Case 3: You have a 98 year old woman with advanced dementia on hospice in the nursing home.  She is unable to speak and is fed by hand.  She develops a fever, is urinating more frequently, grimacing and moaning.  Should you send a UA?  Should you treat with antibiotics?

Though loosly based on real world expereinces, these cases are all made up.

The role of antibiotics in hospice is far from clear-cut.  A recent paper by a group from the Oregan Health Sciences University led by Jennifer Albrecht, PhD, used a national dataset of hospice agencies to investigate use of antibiotics in hospice.  They found that 27% of patients received at least one antibiotic in the last week of life.  These numbers, they note, are lower than previous estimates.  That is still a lot of antibiotics!

The data from other studies seem to suggest that antibiotics are beneficial in terms of symptom relief only for urinary tract infections, as in Case 3.  The data are retrospective, and by no means definitive (see citations in Albrecht's paper). 

In my experience there are a lot of gray areas where the decision of whether or not to prescribe an antibiotic is far from clear.  How much is the antibiotic directed at symptom relief?  At prolonging life?  As the Jennifer Albrecht notes, the goals of patients in hospice are also often mixed, and many patients have some mixture of goals oriented toward quality of life and extending life, as in Case 1.  They also note that while preferences for CPR are commonly documented, preferences for antibiotic use are not. 

Patient, caregiver, and provider beliefs about the effectiveness of antibiotics vary widely.  If we struggle to constrain antibiotic prescribing for upper respiratory tract infections with healthy ambulatory patients in primary care, think of how much more charged these conversations can be at the end of life. 

We need more clarity in this area.  It's a rich area, involving symptom relief and communication, topics at the heart of hospice and palliative care.  Hopefully there are some researchers out there willing to take this on!

4 comments:

Bruce Scott said...

A fine topic for discussion.

I think I'd prefer to try to get a general sense of patient/family goals and then make recommendations within that context rather than ask a long serious of yes/no questions with gradations.

As far as the cases as presented, I think I'm probably treating in two of them. I'm giving a mealy-mouthed answer about the other.

In case 1) Are we pretty comfortable that we've got a respiratory infection here? Probably. Bronchitis vs pneumonia?Less clear. Lung exam might help, but isn't perfect. I'll probably treat here, although I might pick a bronchitis course rather than a pneumonia one. We've got PO choices. (Out of curiosity, how long until the grandchild is born? Let's make sure that we don't oversell the potential for how long she has to live.)

2) Side effect profile more potentially bothersome here. I think this one is easier if you've got a good relationship with the ID doc. The antifungal treatment isn't curative in intent here. So it must be an attempt at suppression. Have the ID doc give me a best estimate of how likely this is to actually help. Unless the patient has a port in place, I'm distinctly unthrilled by the IV antifungal plan. Does the ID guy think the PO meds have any likely effect on disease course? If the answer there is no, then you are looking at extra med, with side effect and med interaction potential (and cost) without any benefit. I'm inclined against treating on this one.

3) I'm treating. You've constructed the story such that I've got sufficient localizing symptoms. My nursing home nurses want me to get many more UAs than I want. I ask them for something to hang my hat on. In this case, you've given us fever, increased urinary frequency, and pain behaviors. I don't like my hospice patients to be in pain; I'll be either investigating or intervening or both. I might be sold on just treating with a best guess abx (based on facility antibiogram if you got one). I'm not faulting any hospice doc who wants to do a UA, though. My other question for this case is about whether or not there is anything else going on here. Constipation perhaps? She can't communicate her needs to us, so we need to be vigilant. I'll assume that if you were able to localize the pain, you would have said so. Also, I'll assume that someone has done a good skin survey regarding wounds as well.

Matthew Green said...

Very good post, thanks for sharing.

Ruchir said...

Truely relevant to practice in HPM and nursing homes as well. In " traditional practice of medicine"- antibiotics are seen as lifesavers. As HPM practitioners when we question its benefits- its views with skepticism by most practitioners, families and staff. Benefits and potential harm of antibiotics is a four different aspects; treatment of patient's symptoms, address family questions, address concerns of other healthcare practitioners and how to create more of this question.

Jim Mittelberger said...

Of note, in addition to the question as to what is the best care for the patient, the Medicare Hospice provider has to answer the questions as to whether the treatment is "curative" or "palliative," a regulatory question that is not relevant to patient or family except as to coverage.