Evidence-based medicine

From Wikipedia, the free encyclopedia

Evidence-based medicine (EBM) aims to apply evidence gained from the scientific method to certain parts of medical practice. It seeks to assess the quality of evidence [1 relevant to the risks and !enefits of treatments (including lack of treatment). "ccording to the Centre for EvidenceBased Medicine# $Evidence%!ased medicine is the conscientious# e&plicit and 'udicious use of current !est evidence in making decisions a!out the care of individual patients.$[( EBM recogni)es that many aspects of medical care depend on individual factors such as quality and value%of%life 'udgments# *hich are only partially su!'ect to scientific methods. EBM# ho*ever# seeks to clarify those parts of medical practice that are in principle su!'ect to scientific methods and to apply these methods to ensure the !est prediction of outcomes in medical treatment# even as de!ate a!out *hich outcomes are desira!le continues. +racticing evidence%!ased medicine requires clinical e&pertise# !ut also e&pertise in retrieving# interpreting# and applying the results of scientific studies and in communicating the risks and !enefits of different courses of action to patients.

Contents

1 ,vervie* ( -lassification o (.1 Evidence%!ased guidelines o (.( Evidence%!ased individual decision making . /istory 0 1ualification of evidence o 0.1 -ategories of recommendations 2 3tatistical measures in evidence%!ased medicine 4 1uality of clinical trial pu!lications o 4.1 5imitations of availa!le evidence 6 Effectiveness 7 -riticism of evidence%!ased medicine 8 Evidence%Based +ractice 19 3ee also 11 :eferences 1( E&ternal links

Overview
;sing techniques from science# engineering# and statistics# such as meta%analysis of medical literature# risk%!enefit analysis# and randomi)ed controlled trials# EBM aims for the ideal that

healthcare professionals should make $conscientious# e&plicit# and 'udicious use of current !est evidence$ in their everyday practice. <enerally# there are three distinct# !ut interdependent# areas of EBM. =he first is to treat individual patients *ith acute or chronic pathologies !y treatments supported in the most scientifically valid medical literature. =hus# medical practitioners *ould select treatment options for specific cases !ased on the !est research for each patient they treat. =he second area is the systematic revie* of medical literature to evaluate the !est studies on specific topics. =his process can !e very human%centered# as in a 'ournal clu!# or highly technical# using computer programs and information techniques such as data mining. Increased use of information technology turns large volumes of information into practical guides. >inally# evidence%!ased medicine can !e understood as a medical $movement$ in *hich advocates *ork to populari)e the method and usefulness of the practice in the pu!lic# patient communities# educational institutions# and continuing education of practicing professionals. Evidence%!ased medicine has demoted ex cathedra statements of the $medical e&pert$ to the least valid form of evidence. "ll $e&perts$ are no* e&pected to reference their pronouncements to scientific studies.

Classification
=*o types of evidence%!ased medicine have !een proposed. [.

Evidence-based g idelines
Evidence%!ased guidelines (EB<) is the practice of evidence%!ased medicine at the organi)ational or institutional level. =his includes the production of guidelines# policy# and regulations. =his approach has also !een called evidence !ased healthcare. [0

Evidence-based individ al decision making

Evidence%!ased individual decision (EBI?) making is evidence%!ased medicine as practiced !y the individual health care provider. =here is concern that current evidence%!ased medicine focuses e&cessively on EBI?. [.

!istory
"lthough testing medical interventions for efficacy has e&isted since the time of "vicenna@s The Canon of Medicine in the 11th century#[2 [4 it *as only in the (9th century that this effort evolved to impact almost all fields of health care and policy. +rofessor "rchie -ochrane# a 3cottish epidemiologist# through his !ook Effectiveness and Efficiency: Random Reflections on Health Services (186() and su!sequent advocacy# caused increasing acceptance of the concepts !ehind evidence%!ased practice. -ochrane@s *ork *as honoured through the naming of centres of evidence%!ased medical research A Cochrane Centres A and an international organi)ation# the -ochrane -olla!oration. =he e&plicit methodologies used to determine $!est evidence$ *ere

largely esta!lished !y the McMaster ;niversity research group led !y ?avid 3ackett and <ordon <uyatt. =he term $evidence !ased$ *as first used in 1889 !y ?avid Eddy. [6 [. =he term $evidence%!ased medicine$ first appeared in the medical literature in 188( in a paper !y <uyatt et al.[7

" alification of evidence

Evidence%!ased medicine categori)es different types of clinical evidence and ranks them according to the strength of their freedom from the various !iases that !eset medical research. >or e&ample# the strongest evidence for therapeutic interventions is provided !y systematic revie* of randomi)ed# dou!le%!lind# place!o%controlled trials involving a homogeneous patient population and medical condition. In contrast# patient testimonials# case reports# and even e&pert opinion have little value as proof !ecause of the place!o effect# the !iases inherent in o!servation and reporting of cases# difficulties in ascertaining *ho is an e&pert# and more. 3ystems to stratify evidence !y quality have !een developed# such as this one !y the ;.3. +reventive 3ervices =ask >orce for ranking evidence a!out the effectiveness of treatments or screeningB

5evel IB Evidence o!tained from at least one properly designed randomi)ed controlled trial. 5evel II%1B Evidence o!tained from *ell%designed controlled trials *ithout randomi)ation. 5evel II%(B Evidence o!tained from *ell%designed cohort or case%control analytic studies# prefera!ly from more than one center or research group. 5evel II%.B Evidence o!tained from multiple time series *ith or *ithout the intervention. ?ramatic results in uncontrolled trials might also !e regarded as this type of evidence. 5evel IIIB ,pinions of respected authorities# !ased on clinical e&perience# descriptive studies# or reports of e&pert committees.

=he ;C Dational /ealth 3ervice uses a similar system *ith categories la!eled "# B# -# and ?. =he a!ove 5evels are only appropriate for treatment or interventionsE different types of research are required for assessing diagnostic accuracy or natural history and prognosis# and hence different $levels$ are required. >or e&ample# the ,&ford -entre for Evidence%!ased Medicine suggests levels of evidence (5,E) according to the study designs and critical appraisal of prevention# diagnosis# prognosis# therapy# and harm studiesB[8

5evel "B consistent :andomised -ontrolled -linical =rial# -ohort 3tudy# "ll or Done# -linical ?ecision :ule validated in different populations. 5evel BB consistent :etrospective -ohort# E&ploratory -ohort# Ecological 3tudy# ,utcomes :esearch# -ase%-ontrol 3tudyE or e&trapolations from level " studies. 5evel -B -ase%series 3tudy or e&trapolations from level B studies 5evel ?B E&pert opinion *ithout e&plicit critical appraisal# or !ased on physiology# !ench research or first principles

" ne*er system is !y the <rade Forking <roup and takes in account more dimensions than 'ust the quality of medical evidence.[19 $E&trapolations$ are *here data is used in a situation *hich has potentially clinically important differences than the original study situation. =hus# the quality of evidence to support a clinical decision is a com!ination of the quality of research data and the clinical @directness@ of the data.[11 ?espite the differences !et*een systems# the purposes are the sameB to guide users of clinical research information a!out *hich studies are likely to !e most valid. /o*ever# the individual studies still require careful critical appraisal.

Categories of recommendations
In guidelines and other pu!lications# recommendation for a clinical service is classified !y the !alance of risk versus !enefit of the service and the level of evidence on *hich this information is !ased. =he ;.3. +reventive 3ervices =ask >orce usesB[1(

5evel "B <ood scientific evidence suggests that the !enefits of the clinical service su!stantially out*eigh the potential risks. -linicians should discuss the service *ith eligi!le patients. 5evel BB "t least fair scientific evidence suggests that the !enefits of the clinical service out*eigh the potential risks. -linicians should discuss the service *ith eligi!le patients. 5evel -B "t least fair scientific evidence suggests that there are !enefits provided !y the clinical service# !ut the !alance !et*een !enefits and risks are too close for making general recommendations. -linicians need not offer it unless there are individual considerations. 5evel ?B "t least fair scientific evidence suggests that the risks of the clinical service out*eigh potential !enefits. -linicians should not routinely offer the service to asymptomatic patients. 5evel IB 3cientific evidence is lacking# of poor quality# or conflicting# such that the risk versus !enefit !alance cannot !e assessed. -linicians should help patients understand the uncertainty surrounding the clinical service.

=his is a distinct and conscious improvement on older fashions in recommendation and the interpretation of recommendations *here it *as less clear *hich parts of a guideline *ere most firmly esta!lished.

#tatistical meas res in evidence-based medicine

Evidence%!ased medicine attempts to e&press clinical !enefits of tests and treatments using mathematical methods. =ools used !y practitioners of evidence%!ased medicine includeB

5ikelihood ratios. =he pretest pro!a!ility of a particular diagnosis# multiplied !y the likelihood ratio# determines the posttest pro!a!ility. =his reflects Bayes@ theorem. =he differences in likelihood ratio !et*een clinical tests can !e used to prioriti)e clinical tests according to their usefulness in a given clinical situation.

=he area under the receiver operator characteristic curve (";-%:,-) reflects the relationship !et*een sensitivity and specificity for a given test. /igh%quality tests *ill have an ";-%:,- approaching 1# and high%quality pu!lications a!out clinical tests *ill provide information a!out the ";-%:,-. -utoff values for positive and negative tests can influence specificity and sensitivity# !ut they do not affect ";-%:,-. Dum!er needed to treat or Dum!er needed to harm are *ays of e&pressing the effectiveness and safety of an intervention in a *ay that is clinically meaningful. In general# DD= is al*ays computed *ith respect to t*o treatments " and B# *ith " typically a drug and B a place!o (in our e&ample a!ove# " is a 2%year treatment *ith the hypothetical drug# and B is no treatment). " defined endpoint has to !e specified (in our e&ampleB the appearance of colon cancer in the 2 year period). If the pro!a!ilities p" and pB of this endpoint under treatments " and B# respectively# are kno*n# then the DD= is computed as 1G(pB%p"). =he DD= for !reast mammography is 1G(72# so (72 mammograms need to !e performed to diagnose one !reast cancer. "s another e&ample# an DD= of 0 means if 0 patients are treated# only one *ould respond.

"n DD= of 1 is the most effective and means each patient treated responds# e.g.# in comparing anti!iotics *ith place!o in the eradication of Helicobacter pylori. "n DD= of ( or . indicates that a treatment is quite effective (*ith one patient in ( or . responding to the treatment). "n DD= of (9 to 09 can still !e considered clinically effective.[1.

" ality of clinical trial p blications

Evidence%!ased medicine attempts to o!'ectively evaluate the quality of clinical research !y critically assessing techniques reported !y researchers in their pu!lications.

=rial design considerations. /igh%quality studies have clearly%defined eligi!ility criteria# and have minimal missing data. <enerali)a!ility considerations. 3tudies may only !e applica!le to narro*ly%defined patient populations# and may not !e generali)a!le to clinical practice. >ollo*up. 3ufficient time for defined outcomes to occur can influence the study outcomes and the statistical po*er of a study to detect differences !et*een a treatment and control arm. +o*er. =his is the num!er of patients enrolled in the trial. " comple& mathematical calculation can determine if the num!er of patients is sufficient to detect a difference !et*een treatment arms. " negative study may reflect a lack of !enefit# or simply a lack of sufficient quantities of patients to detect a difference.

$imitations of available evidence

It is recogni)ed that not all evidence is made accessi!le# that this can limit the effectiveness of any approach# and that effort to reduce various pu!lication and retrieval !iases is required.

>ailure to pu!lish negative trials is the most o!vious gap# and moves to register all trials at the outset# and then to pursue their results# are under*ay. -hanges in pu!lication methods# particularly related to the Fe!# should reduce the difficulty of o!taining pu!lication for a paper on a trial that concludes it did not prove anything ne*# including its starting hypothesis. =reatment effectiveness reported from clinical studies may !e higher than that achieved in later routine clinical practice due to the closer patient monitoring during trials that leads to much higher compliance rates.[10

Effectiveness
=here are mi&ed reports a!out *hether evidence%!ased medicine is effective. ;sing the classification scheme a!ove %%dividing evidence%!ased medicine into evidence%!ased guidelines (EB<) and evidence%!ased individual decision (EBI?)%% may e&plain the conflict. It is difficult to find evidence that EBI? improves health care#[12 *hereas there is gro*ing evidence of improvements in the efficacy of health care *hen evidence%!ased medicine is practiced at the organi)ational level.[14 ,ne of the virtues of healthcare accreditation is that it offers an opportunity to assess the overall functioning of a hospital or healthcare organi)ation against the !est of the currently%availa!le evidence and to assist the hospital or healthcare organi)ation to move to*ards a more effective application of evidence%!ased medical.

Criticism of evidence-based medicine

%he ne trality of this section is disp ted& +lease see the discussion on the talk page. -ritics of EBM say lack of evidence and lack of !enefit are not the same# and that the more data are pooled and aggregated# the more difficult it is to compare the patients in the studies *ith the patient in front of the doctor A that is# EBM applies to populations# not necessarily to individuals. In The limits of evidence-based medicine#[16 =onelli argues that $the kno*ledge gained from clinical research does not directly ans*er the primary clinical question of *hat is !est for the patient at hand.$ =onelli suggests that proponents of evidence%!ased medicine discount the value of clinical e&perience. /o*ever# many proponents of EBM argue that the !est practice of EBM does not discount clinicians@ o*n e&perience. >or instance ?avid 3ackett *rites that $the practice of evidence !ased medicine means integrating individual clinical e&pertise *ith the !est availa!le e&ternal clinical evidence from systematic research$.[(

"lthough evidence%!ased medicine is !ecoming regarded as the $gold standard$ for clinical practice and treatment guidelines# there are a num!er of reasons *hy most current medical and surgical practices do not have a strong literature !ase supporting them. In some cases# such as in open%heart surgery# conducting randomi)ed# place!o%controlled trials *ould !e unethical# although o!servational studies may address these pro!lems to some degree.

-ertain groups have !een historically under%researched (racial minorities and people *ith many co%mor!id diseases)# and thus the literature is sparse in areas that do not allo* for generali)ing.[17 =he types of trials considered $gold standard$ (i.e. randomi)ed dou!le%!lind place!o% controlled trials) may !e e&pensive# so that funding sources play a role in *hat gets investigated. >or e&ample# pu!lic authorities may tend to fund preventive medicine studies to improve pu!lic health as a *hole# *hile pharmaceutical companies fund studies intended to demonstrate the efficacy and safety of particular drugs. =he studies that are pu!lished in medical 'ournals may not !e representative of all the studies that are completed on a given topic (pu!lished and unpu!lished) or may !e misleading due to conflicts of interest (i.e. pu!lication !ias).[18 =hus the array of evidence availa!le on particular therapies may not !e *ell%represented in the literature. " (990 statement !y the International -ommittee of Medical Hournal Editors that they *ill refuse to pu!lish clinical trial results if the trial *as not recorded pu!licly at its outset# may help *ith this# although this has to%date still not !een actioned. =he quality of studies performed varies# making it difficult to generali)e a!out the results.

"n additional pro!lem is that large randomi)ed controlled trials are useful for e&amining discrete interventions for carefully defined medical conditions. =he more comple& the patient population (e.g. severity of condition# co%mor!id conditions# etc) in the study# the more difficult it is to assess the treatment effect (i.e.# treatment mean % control group mean)# relative to the random variation (*ithin group variation of !oth the treatment and control groups). Because of this# a num!er of studies o!tain non%significant results# either !ecause there is insufficient po*er to sho* a difference# or !ecause the groups are not *ell%enough $controlled$. Ironically# the fe*er restrictions there are on *ho can participate in a study (i.e.# the greater the generali)a!ility of the results to the type of patient !eing seen in a real *orld setting) the less a!le the study to detect real differences !et*een groups for a given sample si)e. >urthermore# evidence%!ased guidelines do not remove the pro!lem of e&trapolation to different populations or longer timeframes. Even if several top%quality studies are availa!le# questions al*ays remain a!out ho* far# and to *hich populations# their results are $generali)a!le$. >urthermore# skepticism a!out results may al*ays !e e&tended to areas not e&plicitly coveredB for e&ample a drug may influence a $secondary endpoint$ such as test result (!lood pressure# glucose# or cholesterol levels) *ithout having the po*er to sho* that it decreases overall mortality or mor!idity in a population. In managed healthcare systems# evidence%!ased guidelines have !een used as a !asis for denying insurance coverage for some treatments *hich are held !y the physicians involved to !e effective# !ut of *hich randomi)ed controlled trials have not yet !een pu!lished. In some cases# these denials *ere !ased upon questions of induction and efficacy as discussed a!ove. >or e&ample# if an older generic statin drug has !een sho*n to reduce mortality# is this enough evidence for use of a much more e&pensive ne*er statin drug *hich lo*ers cholesterol more effectively# !ut for *hich mortality reductions have not had time enough to !e sho*nI (1). If a ne*# costly therapy that *orks on tumor !lood vessels causes t*o kinds of cancer to go into remission# is it 'ustified as an e&pense in a third kind of cancer# !efore this has specifically !een provenI ((). Caiser +ermanente did not change its methods of evaluating *hether or not ne*

therapies *ere too $e&perimental$ to !e covered# until it *as successfully sued t*iceB once for delaying IJ> treatments for t*o years after the courts determined that scientific evidence of efficacy and safety had reached the $reasona!le$ stage# and in another case *here Caiser refused to pay for liver transplantation in infants *hen it had already !een sho*n to !e effective in adults# on the !asis that use in infants *as still $e&perimental.$ (.). /ere again the pro!lem of induction plays a key role in arguments.

Evidence-Based 'ractice
In his 1884 inaugural speech[(9 as +resident of the :oyal 3tatistical 3ociety# "drian 3mith held out evidence%!ased medicine as an e&emplar for all pu!lic policy. /e proposed that Evidence% Based +ractice should !e esta!lished for education# prisons and policing policy and all areas of government.

#ee also

"dverse drug reaction "dverse effect (medicine) -linical trials *ith surprising outcomes -onsensus (medical) Epidemiology Evidence%!ased design Evidence%!ased management Evidence%!ased pharmacy in developing countries Evidence !ased practice <uideline (medical) /istory of medicine /ospital accreditation Medical algorithm Medical research Medicine Doce!o +lace!o (origins of technical term) 1uality control 3ystematic revie*

(eferences
1. Elstein "3 ((990). $,n the origins and development of evidence%!ased medicine and medical decision making$. Inflamm. Res. ! S"ppl #: S$%&-'. doi:$(.$(()*s((($$-((&(! )-#. +MI, $ !!%()&. (. 3ackett ?5# :osen!erg FM# <ray H"# /aynes :B# :ichardson F3 (1884). $Evidence !ased medicineB *hat it is and *hat it isn@t$. -M. !$# /)(#!0: )$-#. +MI, % '#&.

.. Eddy ?M ((992). $Evidence%!ased medicineB a unified approach$. Health affairs /+ro1ect Hope0 #& /$0: '-$). doi:$(.$!))*hlthaff.#&.$.'. +MI, $ 2&)#$$. 0. httpsBGGe&change.utmem.eduGe&ch*e!G!inGredir.aspI ;:5KhttpBGG***.ama)on.comGEvidence%Based%/ealthcare%H%Muir%<rayGdpG900.926(10 2. ?. -raig Brater and Falter H. ?aly ((999)# $-linical pharmacology in the Middle "gesB +rinciples that presage the (1st century$# Clinical +harmacolo3y 4 Therape"tics 46 (2)# p. 006%029 [008 . 4. Falter H. ?aly and ?. -raig Brater ((999)# $Medieval contri!utions to the search for truth in clinical medicine$# +erspectives in -iolo3y and Medicine 0. (0)# p. 2.9L209 [2.4 # Hohns /opkins ;niversity +ress. 6. Eddy ?M (1889). $+ractice policiesB *here do they come fromI$. .5M5 #2! /'0: $#2 6 $#2'6 $#)# passim. +MI, #!(&#&!. 7. <uyatt <# -airns H# -hurchill ?# et al. [MEvidence%Based Medicine Forking <roupN $Evidence%!ased medicine. " ne* approach to teaching the practice of medicine.$ .5M5 188(E(47B(0(9%2. +MI? 1090791 8. ,&ford -entre for Evidence%!ased Medicine 5evels of Evidence and <rades of :ecommendation 19. <:"?E *orking group. :etrieved on (996%98%(0. 11. "tkins ?# Best ?# Briss +"# et al ((990). $<rading quality of evidence and strength of recommendations$. -M. !#% /)& &0: $&'(. doi:$(.$$!2*bm1.!#%.)& &.$&'(. +MI, $ #( #' . 1(. =ask >orce :atings. :etrieved on (996%98%(0. 1.. Mc1uay# /enry H.E Moore# :. "ndre* (1886%92%91). Dum!ers Deeded to =reat. Bandolier. :etrieved on (994%94%(6. 10. $+atient -ompliance *ith statins$ -andolier Revie7 #((& 12. -oomarasamy "# Chan C3 ((990). $Fhat is the evidence that postgraduate teaching in evidence !ased medicine changes anythingI " systematic revie*$. -M. !#' /)&)!0: $($). doi:$(.$$!2*bm1.!#'.)&)!.$($). +MI, $ $&!&%. 14. Oealy ?M# "u!le =E# 3tone :"# et al ((992). $Effect of increasing the intensity of implementing pneumonia guidelinesB a randomi)ed# controlled trial$. 5nn. Intern. Med. $&! /$#0: %%$-'&. +MI, $2!2 &2'. 16. =onelli# M: ((991). $=he limits of evidence%!ased medicine$. Respiratory Care &2 /$#0: $&! 8$&&(. 17. :ogers# F" ((990). Evidence !ased medicine and 'usticeB a frame*ork for looking at the impact of EBM upon vulnera!le or disadvantaged groups. H Med Ethics. :etrieved on (996%96%1(. 18. >riedman# 53E :ichter# E? ((990). :elationship !et*een conflicts of interest and research results. D-BI +u!Med. :etrieved on (994%94%(6. (9. 3mith# ".>.M. (1884). $Mad co*s and ecstasyB chance and choice in an evidence%!ased society$. .o"rnal of the Royal Statistical 5ssociation6 Series 5 $ ': !2)-%!.

E)ternal links

Cochrane&org % @=he -ochrane -olla!orationB =he relia!le source for evidence in healthcare@ (systematic revie*s of the effects of health care interventions)# -ochrane 5i!rary Ma'or source of rigorous EBM evaluations.

*!("&gov % @;.3. +reventive 3ervices =ask >orce (;3+3=>)@# "gency for /ealth -are :esearch and 1uality. Ma'or source of EBM evaluations @Fhat Is Evidence%Based MedicineI@ % "merican -ollege of -ardiology CM*+&ca % @Evidence%!ased medicineB a commentary on common criticisms@# ?r. 3haron E. 3traus# ?r. >inlay ". Mc"lister# Canadian Medical 5ssociation .o"rnal# Jol 14.# Do 6# pp 7.6 % 701 (,cto!er .# (999) M+*&com&a % @Evidence%!ased medicineB useful tools for decision making@# Honathan -. -raig# 5es M. Ir*ig# Martin :. 3tockler# Medical .o"rnal of 5"stralia# vol 160# p (07% (2. ((991) ,#'-B&com % @Evidence%!iased medicineB Intention%to%treat analysis less conservativeI@. The Internet .o"rnal of Epidemiolo3y. 0(1). (996 .'/oteBook&co& k % @Evidence%!ased medicine (EBM)@# <eneral +ractice Dote!ook >ree content +(0&o)&ac& k % @BandolierB Evidence%!ased thinking a!out health care@# -andolier /1o"rnal0 >ree revie*s online #!EF&ac& k % @Detting the EvidenceB " 3c/":: Introduction to Evidence Based +ractice on the Internet@ (resource directory)# ;niversity of 3heffield E&tensive !i!liographies and links to online articles =:I+ ?ata!ase % @=:I+ ?ata!ase % EBM search engine@ (resource directory)# =:I+ Cno*ledge 3ervice. >ree. BM+&com % @Evidence !ased medicineB *hat it is and *hat it isn@tB It@s a!out integrating individual clinical e&pertise and the !est e&ternal evidence@# (editorial) -ritish Medical .o"rnal# vol .1(# p 61%6( (Hanuary 1.# 1884) BM+&com % @Evidence !ased medicineB 3ocratic dissent@# (Education and de!ate) -ritish Medical .o"rnal# vol .19# p 11(4%11(6 ("pril (8# 1882) CEBM&net % ,&ford -entre for Evidence%Based Medicine (;C) 3ome free content BM+&BM+1o rnals&com % @+arachute use to prevent death and ma'or trauma related to gravitational challengeB systematic revie* of randomised controlled trials@# <ordon - 3 3mith# Hill + +ell# -ritish Medical .o"rnal# Jol .(6# pp 1028%1041 ((9 ?ecem!er (99.) (-lassic argument that situations still e&ist *here :-=s are unnecessary.) EBOnCall&org % @Evidence compendia@ (evidence%!ased summaries of .7 on%call medical conditions)# Evidence%Based ,n%-all (EB,-) >ree Evidence%!ased medicine at the ,pen ?irectory +ro'ect

"ll te&t is availa!le under the terms of the <D; >ree ?ocumentation 5icense. (3ee Copyrights for details.) FikipediaP is a registered trademark of the Fikimedia >oundation# Inc.# a ;.3. registered 291(c)(.) ta&%deducti!le nonprofit charity.