QUESTIONS BEHIND MEDICINE
From Topics in Emergency Medicine, University of California, San Francisco, School of Medicine
| A CRITICAL EYE ON EVIDENCE-BASED MEDICINE —Jerome R. Hoffman, MD, Professor of Medicine, David Geffen
School of Medicine at the University of California, Los Angeles
|
| “New science”: advocates of evidence-based medicine (EBM) skeptical of clinical decision making, and maintain everything
should be based on literature; speaker believes that, while much to learn from EBM and its principles, need to maintain
healthy skepticism; in 1999, EBM hailed as great breakthrough of decade; however, nothing new about EBM, just
old set of principles dressed up in new language
|
| Clinical epidemiology: speaker’s specialty; deals with way of thinking about evidence, literature, and what we know
(similar to principles of EBM); need to define limits because EBM has become catchword, and notion prevails that clinicians
need to know and practice EBM; need to put EBM in perspective
|
| EBM’s founding principles: Gordon Guyatt, who coined phrase EBM, had done substantial work as clinical epidemiologist
long before he came up with term EBM; Guyatt said we get many things wrong due to our reliance on bad information,
eg, anecdotes, experience; Guyatt suggested EBM as explicit, judicious, and conscientious use of best evidence from research;
Guyatt also acknowledged EBM not perfect; in medical school, students learned procedures on basis of traditional
practice, using intuition, clinical experience, and pathophysiology, but Guyatt said these insufficient; EBM requires clinical
research-based evidence of efficacy before adopting practices and procedures; good principle but many problems with it
|
| Using EBM to solve clinical problems: student of EBM must first formulate question; many questions sound similar but
are different, eg, asking whether test for B-type natriuretic peptide (BNP) will aid diagnosis different from asking
whether BNP test will assist in evaluating patient presenting with acute dyspnea without obvious pulmonary edema but
with history of heart disease and chronic obstructive pulmonary disease (COPD), in whom prior probability intermediate
|
| EBM method: once question(s) formulated, EBM student conducts literature search, then assesses validity of results for
“good or bad” evidence (sometimes problematic); determine solution but recognize that patient has right to make decisions
and has values; view patient as individual, not population; method and numerous steps present obvious challenge in
environment such as emergency department (ED)
|
| Levels of evidence: EBM requires not just one set of evidence but hierarchy of evidence
|
 | N-of-1 randomized controlled trial: highest form of evidence; clinician places patient on drug, then switches patient to
placebo (or other treatment), then back to drug until patient expresses preference; patient serves as own control; this
avoids placebo effect and other confounders possible with larger number of patients; reflects interest in one patient
only, not whether drug good for population overall
|
| Meta-analysis: second highest type of evidence; systematic review of everything published on topic; if certain conditions
met, can mathematically combine evidence from studies; if conditions for inclusion in meta-analysis not met, can still
perform review of all studies on drug; meta-analysis considered almost as good as one-person trial
|
| Other forms of evidence: large randomized controlled trials; systematic reviews of observational studies; single observational
studies; disease-oriented evidence (based on measures of physiology, eg, blood pressure), which differs from patient-oriented
evidence (based on patient outcomes); unsystematic clinical observations at bottom of hierarchy
|
| Evaluation of quality of evidence: speaker’s area of expertise and important EBM principle; many ways studies can be
bad; allocation—how patients assigned to group; problems with selection bias common; blinding—especially important
with subjective outcomes; did trial coordinator have bias? outcomes—are they ones that matter or combination end points
such as “death, stroke, paronychia”; interpretation of data—biggest problem; “spin” common, ie, negative results presented
positively; EBM says clinicians must be critical thinkers about evidence; problem with discarding history of negative
studies for new positive study that should be interpreted only in light of existing studies; experts in EBM stress not
making interpretations based on point estimates, ie, interpretation should be within range of possibilities; sensitivity
analysis—if assumptions changed slightly, would results remain same? must take into account study dropout rate and
whether outcomes of dropouts would affect results; effect of missing data—common to have large amounts of missing
data; unknown how this affects results; number needed to treat (NNT)—must compare relative benefit, absolute benefit,
and NNT to achieve outcome, eg, save one life; with statin drugs, need to treat 250 people for 5 yr at high cost and many
side effects to save one life (absolute decrease in mortality 0.4% or relative decrease 20%); must look at NNT for benefit
and NNT for harm
|
| Limits of EBM: anecdote, experience, and expertise have value; notion that knowledge comes only from randomized clinical
trials disputed by life experience, eg, when crossing street, people know to keep their eyes open; problem with
finding time to perform EBM method; problem with gaining knowledge to do EBM; ability to critically evaluate clinical
evidence takes expertise difficult to acquire for most clinicians who may be experts in other fields; randomness itself
can go wrong; in medicine, randomness addressed with P values, but only small part of interpretation of data; most
critical appraisal addresses bias, not randomness, ie, ways in which nonrandom error involved in interpretation of results
|
| Hierarchy: important principle, but systematic review or meta-analysis usually asserts that only randomized controlled
trials count as evidence, negating whole idea of hierarchy and other kinds of evidence (eg, observational); way we look
at evidence and make inferences involves data and assumptions under which data interpreted; if EBM “science of evidence,”
must ask questions about—assumptions made prior to studies that results likely to be true; outcome scales
used; gold standards used; methodologic quality; how different types of evidence weighed; decision thresholds; meta-
analysis too simple to take all this into account; studies often confuse efficacy with effectiveness; efficacy occurs only
under ideal conditions; for simple therapies, eg, oral antibiotics, effectiveness may be close to efficacy, but not for
more complicated ones, eg, tissue plasminogen activator (tPA) for stroke; meta-analysis as method—not reliable or
high quality; P values do not indicate importance of result, only probability that it occurred by chance; subgroup analyses
almost invariably give wrong answer
|
| Conclusion: concepts of EBM important as tools but not mathematical science; we should learn from EBM but always be
skeptical of truth
|
| DEATH AND DYING —Steven Z. Pantilat, MD, Associate Professor of Clinical Medicine, Alan M. Kates and John M. Burnard
Endowed Chair in Palliative Care, and Director, Palliative Care Service, University of California, San Francisco, School
of Medicine
|
| Challenges in caring for dying: case—72-yr-old Chinese woman with widely metastatic ovarian cancer brought to ED
by ambulance for respiratory distress and altered mental status (patient unresponsive); O2 saturation 90% in high-flow
O2 ; heart rate 124 bpm, blood pressure (BP) 82/40 mm Hg; family said patient doing fine but found unresponsive in
morning; challenges for care include—unknown extent of disease and proximity to death; patient cannot communicate;
cultural issues; determining appropriate level of intervention; are interventions ultimately going to help patient or
add to her suffering? relief of suffering and cure of disease must be seen as twin obligations of medical profession
|
| Three needs of dying patients and their families: when asked, patients and families say they need 1) palliation of
symptoms (eg, pain, shortness of breath), 2) communication about illness and death, and 3) psychosocial support, expressed
as wanting to know physician and nurse “will be there for me”
|
| Symptoms: people facing end of life have many physical symptoms; data from Study to Understand Prognoses and Preferences
for Outcomes and Risks of Treatment (SUPPORT) show 40% of hospitalized patients, regardless of diagnosis, had
moderate-to-severe pain at least half of time in last 3 days of life (presumes analgesia available to these patients); this included
patients with heart failure and COPD (not usually considered painful conditions), either because of condition itself
(eg, edema in heart failure), comorbidity (eg, diabetic neuropathy), or interventions; patients facing end of life need to focus
on issues important to them (eg, saying good-byes, managing affairs), so treating pain and other symptoms to allow
them to do so becomes important; speaker cites comfort-care suites at his hospital for feel of home; suites allow patient’s
family to “move-in” to dying patient’s environment by providing amenities, eg, couch, crib, television, shelf for display
of spiritual objects, food
|
| Communication about illness and death: avoid common phrases such as “there’s nothing more we can do” (never true;
even if simply keeping patient clean and dry, always something that can be done); talk about positive actions being
planned; even withdrawing interventions should be communicated as active process to achieve particular goal
|
| Questions to ask: avoid asking “would you like us to do everything possible?” patient, family, and physician perceive
question differently; array of interventions frightening, but families do not think of this; far better to ask, “how were you
hoping we can help your mother?” having words to say to patients or their families just like knowing how to do procedures
and will help physician who has many patients and demands on time
|
| Withdrawing care: physicians never withdraw care; they are always providing care, even though life-sustaining interventions
actively taken away
|
| Talking to patients and their families: listen—first thing to do; patients often have much on their minds that they want
to tell clinicians; study showed that in primary care, patient able to talk average of 18 sec before being interrupted by physician;
use open-ended questions and wait (listen for 2 min); ask open-ended questions—eg, “how are you hoping we
can help you?” “how have things been going for you at home?” “what is bothering you the most?” more efficient way of
obtaining patient’s story than asking closed-ended questions; make empathic statements—that respond to family emotional
issues; eg, in study, 33 medical residents agreed to record code status discussions with patients; found that
residents missed opportunities for empathic connection; try to listen for opportunities to make people feel heard and help
establish trust and rapport quickly with patients
|
| Elicit values and goals of care: physicians’ language about what we do to patients, whereas patients’ language about what
they do in their lives; physician should say things such as “it must be really scary to think about being so sick” or ask questions
like, “when you think about the future, what worries you the most?” that remain sensitive to patient’s culture and enable clinician
to speak language of patient; define goals of care; ask patient “when you think about the future, what do you hope for?”
patient may articulate goal as one of not dying alone or of not experiencing pain
|
| Cultural issues: desire for information varies with culture; do not assume cultures monolithic; simply knowing someone
is from given culture does not inform physician on how that particular patient will respond; as much variation within culture
as between cultures; ask “is there anything I need to know about your culture concerning dying?” use third person, as
in, “this is what I know about people who have heart failure”
|
| Informed refusal: for some patients, cultural issues are such that they do not want information; we value information in
Western culture, but not all cultures do; physician should be willing to allow patients to waive their rights of decision making
but needs to hear this from patient or family; say, “some patients want to know details and some want to have the clinician
talk to someone else”; use professional translator, not family translator
|
| Use of resources: get help in talking to patients; in ED, patient not known to staff and often presents in extremis; seek assistance
from primary care physician (PCP); for patients enrolled in hospice, care nurse available 24 hr/day, 7 day/wk
(call hospice); culture where tradition to dress dying patient in new clothes before death for journey to next world demonstrates
importance of clinicians’ spending time with patient and family and asking them what they want
|
| Dying patient: patients face loss of all important relationships in life; caregivers also need to be aware of effect of looking
after dying patients on themselves
|
| Psychosocial support: conveys to patient “we will be there for you whatever happens”; difficult in ED; system must provide
nonabandonment for patient; ensure good care after leaving ED or being admitted to hospital; tell patient “we will
make sure you get care, even after you go home”; if best option for patients to return to hospital, tell them they can
|
| Ask about hope: dying patients have hope; asking about hope encourages hope; can ask patient what they hope for and receive
plenty of insight and information; in case of patient who keeps coming back to ED, sending back home not what patient
needs (patient actually afraid of dying at home alone)
|
| Growth and development: patients have tasks to accomplish before death, eg, saying good-bye or achieving closure in
relationship; speaker cites 5 important things to say before death, “forgive me, I forgive you, thank you, I love you, good-
bye;” even in ED, families may want to say something to patient who cannot hear because “power lies in telling”
|
| Caregivers and death: death has cumulative effect on health care providers; need to think about physician’s personal response;
speaker cites incident of denial of death during his residency (colleague talked about going out to eat right after
patient died); need to think about ways to acknowledge somebody has just died; develop rituals for when patient dies (eg,
moment of silence, saying prayer or poem); talk in medical rounds or case conferences about patients who died; data
show benefits gained from talking about mistakes include better integrated practice and less chance of burnout because
death dealt with and integrated into psyche
|
Educational Objectives
| The goal of this program is to educate the listener about critical issues in evidence-based medicine (EBM) and death
and dying. After hearing and assimilating this program, the clinician will be better able to:
|
| 1. Name some principles of EBM.
|
| 2. Discuss EBM’s hierarchy of evidence.
|
| 3. Identify some limitations of EBM methodology.
|
| 4. List 3 primary needs of dying patients and their families.
|
| 5. Implement better communication in caring for the dying.
|
Suggested Reading
Auerbach AD et al: End-of-life care in a voluntary hospitalist model: effects on communication, processes of care,
and patient symptoms. Am J Med 116:669, 2004; Barrett B et al: Placebo, meaning, and health. Perspect Biol Med
49:178, 2006; Beaton R et al: Exposure to duty-related incident stressors in urban firefighters and paramedics. J
Trauma Stress 11:821, 1998; Cowen AR: Death, dying and grieving. Dealing with personal and work-related tragedies.
Emerg Med Serv 27:110, 109, 1998; Edlich RF et al: On death and dying in the emergency department. J
Emerg Med 10:225, 1992; Hatala R et al: Beyond journal clubs. Moving toward an integrated evidence-based medicine
curriculum. J Gen Intern Med 21:538, 2006; Jurkovich GJ et al: Giving bad news: the family perspective. J
Trauma 48:865, 2000; Kalil AC: Evidence-based medicine and skepticism. Am J Med 119:e13; author reply e15,
2006; Kannry J et al: Using an evidence-based approach for system selection at a large academic medical center:
lessons learned in selecting an ambulatory EMR at Mount Sinai Hospital. J Healthc Inf Manag 20:84, 2006; Kho A
et al: Use of handheld computers in medical education. A systematic review. J Gen Intern Med 21:531, 2006; Marrow
J: Telling relatives that a family member has died suddenly. Postgrad Med J 72:413, 1996; Okoromah CA et
al: Evidence-based medicine curriculum: impact on medical students. Med Educ 40:465, 2006; Pantilat SZ: Care of
dying patients: beyond symptom management. West J Med 171:253, 1999; Porta M: Five warrants for medical decision
making: some considerations and a proposal to better integrate evidence-based medicine into everyday practice.
Commentary on Tonelli (2006), Integrating evidence into clinical practice: an alternative to evidence-based approaches.
Journal of Evaluation in Clinical Practice 12, 248-256. J Eval Clin Pract 12:265, 2006; Redley B et al:
Staff attitudes towards family presence during resuscitation. Accid Emerg Nurs 4:145, 1996; Tye CC: Sudden bereavement
in accident and emergency: the development and evaluation of a short educational course for qualified
nurses. Accid Emerg Nurs 4:88, 1996; van der Wal ME et al: Circumstances of dying in hospitalized children. Eur J
Pediatr 158:560, 1999; Vanbrabant P et al: What do we know about patients dying in the emergency department?
Resuscitation 60:163, 2004.
Faculty Disclosure
In adherence to ACCME guidelines, the Audio-Digest Foundation requests all lecturers to disclose any significant financial
relationship with the manufacturer or provider of any commercial product or service discussed. For this issue,
the faculty reported nothing to disclose.
Drs. Hoffman and Pantilat spoke at Topics in Emergency Medicine, presented October 24-27, 2005, in San Francisco
by the University of California, San Francisco, School of Medicine. The Audio-Digest Foundation thanks the speakers
and the sponsor for their cooperation in the production of this program.
|
