La lecture critique de la littérature médicale dans le - CClin Sud-Est

Transcription

La lecture critique de la
littérature médicale
dans le groupe de lecture
DIALIN
L. Ayzac, CCLIN Sud Est
VI° journée annuelle du Réseau DIALIN – 22 septembre 2009
Les critères de validité des
publications médicales
• Où est publié l’article?
– Notion de comité de lecture
– Notion d’index de citations : Impact factor
• Qui sont les auteurs?
– Autres publications
– Notoriété
• La lecture critique
1
Objectifs
• Où se situe cette étape ?
– La recherche bibliographique a été réalisée et de
qualité
– Les articles ont été obtenus
• Quels sont les objectifs ?
– Lire les documents
– Juger de la qualité de leur contenu
– « Séparer le bon grain de l’ivraie »
– Sélection + évaluation
C’est le concept de lecture critique.
Juger la valeur des publications
• Quelle est la crédibilité de la publication ?
((validité interne))
– En fonction du type d’étude, identifier les différentes
étapes du protocole de l’étude et évaluer le niveau de
crédibilité de chaque étape
– Les conclusions sont-elles valides ?
• Quelle est l’applicabilité des informations
contenues dans la publication ? (validité
externe)
– Les conclusions sont-elles applicables à la pratique
médicale du praticien ou au projet de recherche du
réseau DIALIN?
2
Une méthode (1)
Deux livres
• 1. Salmi LR: Lecture critique et rédaction
médicale
édi l scientifique.
i tifi
C
Commentt lilire,
rédiger et publier une étude clinique ou
épidémiologique. Paris, Elsevier, 1998.
• 2. Landrivon G, Delahaye F: La recherche
clinique De l'idée
l idée à la publication. Paris,
Masson, 1995.
Une méthode (2)
Utiliser une grille de lecture
3
Une méthode (3)
8 étapes du protocole d’étude Ö 8 critères d’évaluation
• Pour chaque critère :
– Est-il possible de trouver l’information sur le
critère ?
– L’abord du critère est-il correcte ?
– Si l’abord est incorrect, cela menace-t-il la
validité de l’étude
l étude ?
Une méthode (4)
Situation du problème
Î Objectifs Î Hypothèse
• Il y a 4 grandes préoccupations possibles :
• L’histoire naturelle d’une maladie, son évolution
et son pronostic.
• L’utilisation,
L’ tilisation la performance et l’intérêt d’
d’un
n test
diagnostic.
•L’impact d’une intervention thérapeutique, de
dépistage, de prévention ou d’éducation
(utile/inutile/dangereux)
•La détermination d’une étiologie ou d’une
causalité
• Replacer l’article dans un de ces 4 cadres Ö une
idée claire sur les objectifs
• Y a-t-il une hypothèse à vérifier par rapport à
l’objectif ?
4
Une méthode (5)
•
Il existe une hiérarchie parmi les 5 modèles possibles de plan
d’études dans le niveau de preuve et dans le coût.
Plan de l’étude
Niveau de preuve
0
Coût
Une méthode (6)
• L’étude rapporte un cas intéressant et inhabituel ou une série de
cas
Plan de l’étude
Niveau de preuve
Rapport de
cas
0
Coût
5
Une méthode (7)
• L’étude transversale est la description de fréquence d’une
maladie, de ses facteurs de risque dans une population déterminée
à un temps déterminé.
Plan de l’étude
Niveau de preuve
Etude
transversale
Rapport de
cas
0
Coût
Une méthode (8)
Plan de l’étude
• L’étude cas-témoin est une étude d’observation, rétrospective,
comparant les caractéristiques de patients atteints d’une maladie
(les cas) avec celles d’individus indemnes de la maladie (les
témoins).
Niveau de preuve
Étude cas
témoin
Etude
transversale
Rapport de
cas
0
Coût
6
Une méthode (9)
Plan de l’étude
• L’étude de cohorte est une étude d’observation, prospective,
suivant un groupe d’individus exposés à des facteurs de risque
d’une maladie pendant une période de temps donnée. Le taux
d’incidence de la maladie peut être comparé à celui d’un groupe
témoin non exposé.
Niveau de preuve
Étude de
cohorte
Étude cas
témoin
Etude
transversale
Rapport de
cas
0
Coût
Une méthode (10)
Plan de l’étude
• L’essai contrôlé est une étude expérimentale, comparant le
résultat d’une intervention sur un groupe de sujets recevant
l’intervention au résultat dans un groupe semblable qui ne reçoit pas
l’intervention.
Niveau de preuve
Essai
randomisé
Étude de
cohorte
Étude cas
témoin
Etude
transversale
Rapport de
cas
0
Coût
7
Une méthode (11)
Plan de l’étude
• Le plan d’étude retenu est-il approprié à
la question posée ?
Une méthode (12)
Plan de l’étude
• Quel est le facteur étudié ?
• Exposition
• ou Intervention
Facteur étudié
• ou Test diagnostique
• Sont-ils bien décrits ?
• Comment sont ils mesurés ?
• Même méthode de mesure chez tout les sujets ?
• Même méthode de mesure dans tous les groupes ?
• Méthode de mesure à l’aveugle ?
• Pour les test diagnostiques : comparaison
indépendante avec l’étalon ?
8
Une méthode (13)
• Quel est le critère de jugement (facteur résultant) ?
Plan de l’étude
• Comment est il mesuré ?
Facteur étudié
• Même méthode de mesure chez tout les sujets ?
Critère de jugement
• Même méthode de mesure dans tous les groupes ?
• Méthode de mesure à l’aveugle ?
• Tous
T
les
l critères
itè
de
d jugement
j
t pertinents
ti
t sont-ils
t il
évalués ?
Une méthode (14)
Plan de l’étude
Facteur étudié
(facteur résultant)
Population étudiée / échantillon
• La sélection est-elle correcte ?
• Y a t il randomisation ?
• Les groupes diffèrent-ils
diffèrent ils par des
caractéristiques autres que le facteurs
étudiés ?
•Quel est la proportion de sujets atteignant
la fin du suivi ?
g
q
: y a-t-il un
•Pour les tests diagnostiques
large éventail de patients ?
9
Une méthode (15)
Plan de l’étude
Facteur étudié
Biais et facteurs de confusion
• Un biais est une erreur systématique qui contribue
à produire des estimation s systématiquement plus
élevées ou plus basses que la valeur réelle des
paramètres à estimer
• Biais dus à la sélection
• Biais dus à la mesure…
• Un facteur de confusion est un facteur qui modifie
les effets du facteur étudié sur le critère de jugement
du fait de son lien à la fois avec le facteur étudié et
avec le facteur de jugement.
jugement
• Sont-ils tous envisagés ?
• Sont-ils tous pris en compte ?
Une méthode (16)
Plan de l’étude
Facteur étudié
• Résultats statistiquement significatif
• Tests statistiques, place du p
• Intervalles de confiance
• Direction de l’association
• Force de l’association
• Taille de l’échantillon suffisant ?
• Cliniquement intéressant ?
Résultats
• Puissance du test / taille de l’échantillon ?
10
Une méthode (17)
• Y a t il toutes les réponses aux questions
posées ?
Plan de l’étude
• L’hypothèse est-t-elle vérifiée ?
Facteur étudié
• L’objectif est-il atteint ?
• Les résultats sont-ils acceptables
appliqués
li é à la
l population
l ti source d
de lla
publication ? (Validité)
Résultats
Synthèse
• Les résultats sont-ils acceptables pour
votre propre pratique ? (Applicabilité)
Un exercice corrigé (1)
11
Plan de l’étude
Facteur étudié
Le facteur étudié (l
(l’exposition
exposition ou ll’intervention
intervention qui est supposé avoir des conséquences sur
un problème de santé, une maladie ou un état clinique) est la présence de bactériémie à
Staphylococcus aureus (SAB) chez des patients hémodialysés chroniques.
Elle est mesurée dans ses conséquences en terme d’hospitalisation, de pronostique –
dissémination de l’infection , décès -et de coûts.
Le critère de jugement est la nature du site d’accès (cathéter, fistule native, fistule greffe).
12
Les facteurs de confusion et les biais
Les données sont collectées au moment de l’hospitalisation du patient et de sa saisie dans un
dossier informatisé. Qu’en est il si le patient n’est pas hospitalisé ou hospitalisé dans un autre
établissement? (Biais de sélection).
La mortalité est observée à la sortie du patient et 12 semaines après la survenue de la
bactériémie (pour éliminer un biais d’observation).
Pour l’étude de la mortalité et des coûts d’autres facteurs que le site d’accès rentre en ligne de
compte : ce sont les facteurs de confusions. Ils sont recueillis systématiquement : facteurs
liés au patient (gravité – APACHE II) ou à la prise en charge (antibiotiques, corticothérapie…)
Résultats
Les analyses statistiques (régression logistique multiple) ont consisté en une estimation du
risque relatif (Odds Ratio OR) et de son intervalle de confiance. Les intervalles contiennent 1.
Il n’y a donc pas d’association significative entre type de site d’accès et mortalité/coûts
13
Résultats
Synthèse
14
Au total
Plan de l’étude
Facteur étudié
Résultats
Synthèse
15
Fiche de lecture critique d’une étude clinique ou épidémiologique DIALIN (V3 08/03/2007)
(d’après SALMI L.R. Lecture critique et rédaction médicale scientifique. Comment lire, rédiger et publier une étude clinique ou épidémiologique. Paris :Elsevier ; 1995)
Référence : N° .....................
Auteurs : ..............................................................................................................................................................
Titre : ....................................................................................................................................................................
Revue : ................................................................................................................................................................
Année : ............................
Volume (n°) : .............. ( ....... ) Pages : .....................................................
Résumé :
Objectif et justification :
Schéma d’étude ; groupes comparés :
Population étudiée et nombre de sujets :
Conduite de l’étude :
Principaux résultats :
Cocher la case correspondant au respect du critère : 0 = oui, N = non, I = incomplet ; NA = ne s’applique pas,
NSP = ne sait pas. Une réponse cochée N à un critère en italique = étude inacceptable.
Critères
Schéma d’étude
Formulation claire de l’objectif
Hypothèse faite à priori
Schéma adapté à l’objectif
Procédures de sélection
Critères d’inclusion décrits et adéquats
Critères d’exclusion décrits et adéquats
Indication du nombre de refus avant l’étude
Procédures identiques dans tous les groupes
Taille d’échantillon adaptée
Règles éthiques respectées
Conduite de l’étude
Retraits indiqués, expliqués et raisonnables
Retraits équilibrés entre les groupes
Mesures principales fiables et valides
Analyse des résultats
Méthodes adaptées à la question et au schéma
Prise en compte des variables importantes
Résultats vérifiables des données brutes
Prise en compte des comparaisons multiples
Conclusion :
Qualité
Exceptionnelle
Très bonne
Plutôt bonne
Faible mais acceptable
Inacceptable
Pas qualifié pour juger
O
I
N
NA
NSP
Discussion, forces et faiblesses :
Cocher
Faits avérés ou réfutés :
Décision
Mettre un seul X devant un des deux mots clefs
INC
A inclure dans la biblio
EXC
A ne pas inclure dans la biblio
Justifications de l’inclusion
Justifications de l’exclusion
Mettre au moins un X devant un des mots clefs
IAV
Infection de l’accès vasculaire
HOS
Hors sujet
BAC
Bactériémie
MAQ
Mauvaise qualité
HCV
Hépatite virale C
AUT
Autre
HIV
SIDA
En clair :
VIR
Autres viroses
TIC
Taux d’incidence
TPR
Taux de prévalence
FAC
Facteurs de risque (de confusion)
REC
Recommandations
SUR
Surveillance
STA
Standardisation des résultats
INO
Innovation
EXC
Exceptionnel, publication princeps
Autres mots clefs réservés à la gestion des références bibliographiques
SEL
En cours de sélection
NOL
Non lu (non retenu par la sélection)
LEC
En cours de lecture (retenu par la sélection)
Lecteur : ...................................................
Date : ......... / ........./ .........
Commentaires
Relationship between Clinical Outcomes and Vascular
Access Type among Hemodialysis Patients with
Staphylococcus aureus Bacteremia
Jula K. Inrig,* Shelby D. Reed,† Lynda A. Szczech,* John J. Engemann,‡ Joelle Y. Friedman,†
G. Ralph Corey,‡ Kevin A. Schulman,§ L. Barth Reller,‡ and Vance G. Fowler, Jr.‡
*Department of Medicine, Division of Nephrology, ‡Department of Medicine, Division of Infectious Diseases, and
§
Department of Medicine, Duke University Medical Center, and †Center for Clinical and Genetic Economics, Duke
Clinical Research Institute, Durham, North Carolina
The association between hemodialysis vascular access type, costs, and outcome of Staphylococcus aureus bacteremia (SAB)
among patients with ESRD remains incompletely characterized. This study was undertaken to compare resource utilization,
costs, and clinical outcomes among SAB-infected patients with ESRD by hemodialysis access type. Adjusted comparisons of
costs and outcomes were based on multivariable linear regression and multivariable logistic regression models, respectively.
A total of 143 hospitalized hemodialysis-dependent patients had SAB at Duke University Medical Center between July 1996
and August 2001. A total of 111 (77.6%) patients were hospitalized as a result of suspected bacteremia; 32 (22.4%) were
hospitalized for other reasons. Of the 111 patients, 59.5% (n ⴝ 66) had catheters as their primary access type, 36% (n ⴝ 40) had
arteriovenous (AV) grafts, and 4.5% (n ⴝ 5) had AV fistulas. Patients with fistulas were excluded from analyses because of
small numbers. Patients with catheters were more likely to be white, had shorter dialysis vintage, and had higher Acute
Physiology and Chronic Health Evaluation II scores compared with patients with grafts. Unadjusted 12-wk mortality did not
significantly differ between patients with catheters compared with patients with grafts (22.7 versus 10.0%; P ⴝ 0.098); neither
did 12-wk costs differ by access type ($22,944 ⴞ 18,278 versus $23,969 ⴞ 13,731, catheter versus graft; P > 0.05). In adjusted
analyses, there was no difference in 12-wk mortality (odds ratio 1.63; 95% confidence interval 0.29 to 9.02; catheter versus graft)
or 12-wk costs (means ratio 0.84; 95% confidence interval 0.60 to 1.17; catheter versus graft) among SAB-infected patients with
ESRD on the basis of hemodialysis access type. Twelve-week mortality and costs that are associated with an episode of SAB
are high in hemodialysis patients, regardless of vascular access type. Efforts should focus on the prevention of SAB in this
high-risk group.
Clin J Am Soc Nephrol 1: 518 –524, 2006. doi: 10.2215/CJN.01301005
I
nfections are the second leading cause of mortality in
dialysis patients, accounting for nearly one fourth of all
deaths in patients with ESRD (1). Staphylococcus aureus is a
leading cause of bacteremia among hemodialysis-dependent
patients (1–11), and rates may be increasing (1). Complications
of S. aureus bacteremia (SAB) include sepsis, endocarditis, osteomyelitis, septic arthritis, and metastatic abscesses (12–14).
Vascular access is a major risk factor for bacteremia (15,16)
and infection-related hospitalization and mortality (17,18)
among hemodialysis patients. Evidence suggests that tunneledcuffed catheters might be associated with a higher frequency of
infectious complications than other forms of dialysis vascular
access. For example, the frequency of hospitalizations for infectious complications among patients with ESRD has increased
(1) at the same time that the proportion of patients who have
Received October 12, 2005. Accepted February 22, 2006.
Published online ahead of print. Publication date available at www.cjasn.org.
Address correspondence to: Dr. Jula K. Inrig, Duke University Medical Center,
North Pavilion, 2400 Pratt Street, Box 3646, Durham, NC 27705. Phone: 919-6687516; Fax: 919-668-7128; E-mail: [email protected]
Copyright © 2006 by the American Society of Nephrology
ESRD and have catheters as their primary vascular access has
climbed (19 –21). However, the association between vascular
access type and clinical outcomes among patients with ESRD
and SAB still is unclear.
Despite its clinical significance, the economic impact of SAB
among patients with ESRD remains incompletely characterized. Previous studies provided insight into the economic burden of S. aureus (22,23). Engemann et al. (24) found that the
mean cost of an episode of SAB among a cohort of patients with
ESRD was $24,034 per episode, and patients with complicated
SAB episodes incurred even higher costs. Similarly, Reed et al.
(25) found community-dwelling patients who had ESRD and
methicillin-resistant S. aureus (MRSA) bacteremia incurred
higher costs compared with patients with methicillin-sensitive
S. aureus (MSSA) bacteremia. Despite these observations, however, the potential relationship between intravascular access
type, costs, and clinical outcomes among patients with SAB is
undefined. Herein, we sought to analyze differences in resource
utilization, costs, and clinical outcomes by access type among
patients who have ESRD and are admitted to the hospital for
suspected SAB.
ISSN: 1555-9041/103-0518
Clin J Am Soc Nephrol 1: 518 –524, 2006
Materials and Methods
Study Design
A prospective cohort of hemodialysis patients with SAB was used to
describe and compare resource utilization, costs, and clinical outcomes
among patients with various types of hemodialysis access. Characteristics of these patients have been reported previously (24,25).
Patients and Study Setting
All adult patients who had ESRD, were undergoing hemodialysis,
and were admitted to Duke University Medical Center with SAB between July 1, 1996, and August 31, 2001, were identified prospectively
and included in this study. Daily reports were received from the
microbiology laboratory regarding all patients with at least one positive
blood culture for S. aureus, and all patients subsequently were evaluated and consented for enrollment within 36 h of identification of S.
aureus. Exclusion criteria were SAB in nonhospitalized patients, age
⬍18 yr, polymicrobial infection, neutropenia (absolute neutrophil
count ⬍1.0 ⫻ 109/L), or death before evaluation. For maintaining
independence of observations, the analysis was limited to the first
episode of SAB for each patient but was inclusive of recurrent infections
that occurred during the 12 wk after the date of the first positive blood
culture. The institutional review board of Duke University Medical
Center approved this study.
Data Collection
Clinical data were collected at the time of a patient’s hospitalization
and entered into an electronic database (Access; Microsoft Corp., Redmond, WA). The data that were collected included patient demographic characteristics, comorbid conditions, antibiotic therapy, complications of infection, duration of ESRD and hemodialysis, type of
vascular access, history of kidney transplant, and suspected cause of
the patient’s kidney disease. Acute Physiology and Chronic Health
Evaluation II (APACHE II) scores were calculated on the date of initial
positive blood culture (26). Outcomes (including death or metastatic
complications) were assessed during the initial hospitalization and 12
wk after the date of the initial positive blood culture. Data regarding
resource utilization (e.g., medical tests, procedures, length of hospitalization) were obtained by medical record abstraction. Cost data information for all patients was retrieved from the Duke University Medical
Center hospital accounting system (Transition Systems, Inc., Boca
Raton, FL).
Definitions
Patients were categorized into groups on the basis of type of intravascular access used for hemodialysis at the time of admission: (1)
Tunneled, cuffed, long-term catheter access; (2) synthetic arteriovenous
(AV) graft; or (3) native AV fistula. Patients with catheter access were
categorized further into two subgroups for sensitivity analyses: (1)
Patients with catheters as their sole active access type and (2) patients
with alternative active access types (maturing or active AV grafts or AV
fistulas). Corticosteroid use was defined as either chronic low-dose or
acute high-dose therapy used within the previous 30 d. Prevalent
congestive heart failure at baseline was ascertained through self-report
with medical record verification. Self-report was used to classify injection drug use. Sepsis was defined by standard criteria (27). Central
nervous system involvement was defined by a positive culture for S.
aureus obtained from cerebrospinal fluid. Infective endocarditis was
defined according to Duke criteria (28). Metastatic infections were
defined by the presence of an infection remote from the determined
primary focus of infection (e.g., vertebral osteomyelitis, distant abscess).
Methicillin susceptibility was determined from the oxacillin test using
Staphylococcus aureus Bacteremia and Outcomes
519
criteria of the National Committee for Clinical Laboratory Standards
(29).
Cost Data
The cost analysis was performed from the perspective of the health
system and includes all direct medical costs that are associated with
treatment of SAB and its sequelae during a 12-wk period. Because
patients were included in the analysis only when their hospitalization
was due to SAB, the entire cost of the admission was attributed to the
infection. Cost data were available for all patients who were admitted
after July 1, 1998. Hospital costs were obtained from the hospital cost
accounting system and were adjusted to 2001 US dollars using the
Consumer Price Index for Medical Care (30). Physician fees for inpatient and outpatient services were assigned using the 2001 reimbursement rates from the Centers for Medicare and Medicaid Services for
providers in North Carolina. Total inpatient costs for the initial hospitalization included inpatient costs and physician fees. Outpatient costs
and rehospitalization costs that occurred within 12 wk after the date of
the initial positive blood culture were included only for procedures and
admissions related to S. aureus infection. To be consistent with preferred methods (31), costs results were reported as means.
Statistical Analyses
Because of small numbers, patients with native AV fistulas as their
sole access were excluded from further analyses. Descriptive statistics
are presented as counts and percentages for discrete variables. Continuous variables are reported as means unless noted otherwise. In unadjusted analyses, continuous variables were compared using t tests when
the data approximated a normal distribution and Wilcoxon rank sum
tests when the data were non-normally distributed. Categorical variables were compared using Pearson ␹2 tests. Unadjusted comparisons
of costs were based on nonparametric bootstrapping because the data
did not approximate either a normal or log-normal distribution and
because of nonequal variances (32,33). A bias-corrected method was
applied to obtain the 95% confidence interval (CI) of the difference
between cost estimates. An unadjusted comparison of clinical outcomes
(mortality at discharge and at 12 wk) by access type was performed
using Pearson ␹2 tests.
Adjusted comparisons of costs were based on a multivariable generalized linear model (GLM) specified with a ␥ distribution and a log link
to account for heteroscedasticity and skewness associated with the cost
data (34 –36). Previous investigators have shown through a series of
simulated data sets that GLM are less biased than ordinary least
squares (OLS) regression on log-transformed cost data (34). GLM are
more resilient to various data structures that are often represented with
cost data, and the GLM approach typically is more conservative than
OLS because the standard errors tend to be greater than those estimated
by OLS (34). Cost analyses were performed using SAS PROC GLM.
Adjusted comparisons of clinical outcomes were performed using a
multivariable logistic regression model. Variables that were entered
into the regression model included demographic characteristics, comorbidities, APACHE II scores, infection-related factors (admitted
from home, nursing home, or other hospital), and infection with methicillin-sensitive S. aureus or MRSA. All analyses were performed
using SAS statistical software, version 8.2 (SAS Institute, Inc., Cary,
NC).
Results
Baseline Characteristics
From July 1996 to August 2001, 143 hemodialysis patients
who were admitted to Duke University Medical Center had
520
Clinical Journal of the American Society of Nephrology
positive blood cultures for S. aureus. Of these, 111 (77.6%)
patients were admitted for suspected bacteremia and 32 (22.4%)
were admitted for other reasons and therefore were excluded
from further analyses. Of the 111 patients, 59.5% (n ⫽ 66) had
a catheter as their sole access type for hemodialysis upon
admission, 36.0% (n ⫽ 40) had a synthetic AV graft, and 4.5%
(n ⫽ 5) had a native AV fistula.
Patient Characteristics According to Intravascular
Access Type
Patients with catheters as their sole access type were more
likely to be white and have a shorter dialysis vintage compared
with patients with AV grafts (Table 1). Patients with catheters
also had higher total APACHE II scores on diagnosis with SAB
compared with patients with AV grafts. There were no significant differences between groups with regard to gender, age,
admission from a nursing home, underlying cause of renal
disease, other comorbidities, or infection with MRSA.
Resource Utilization and Outcomes
In unadjusted analyses, resource utilization during the initial
hospitalization was similar between patients with catheters and
patients with AV grafts, with the exception of the number of
procedures that were performed related to a foreign device (4.1
versus 2.3 for catheter versus graft; P ⫽ 0.001; Table 2). During
the 3-mo follow-up period, 13.8% of patients with catheters
were re-hospitalized compared with 23.7% of patients with
grafts (P ⫽ 0.21) with a similar number of mean inpatient days
during the 12-wk period (12.6 versus 12.4 d for catheter versus
graft; P ⫽ 0.33).
Metastatic infections occurred in 22 (33.3%) patients with
catheters and in 12 (30.0%) patients with AV grafts (P ⫽ 0.72).
Metastatic complications included endocarditis (n ⫽ 17), septic
emboli (n ⫽ 8), abscess (n ⫽ 6), septic arthritis (n ⫽ 5), meningitis (n ⫽ 3), nonvertebral osteomyelitis (n ⫽ 3), vertebral
osteomyelitis (n ⫽ 3), septic thrombophlebitis (n ⫽ 2), cerebro-
Table 1. Baseline patient characteristics by access type among ESRD patients hospitalized with SABa
Characteristic
Tunneled, Cuffed Catheter
(n ⫽ 66)
AV Graft
(n ⫽ 40)
Pb
Male
Race
white
black
other
Age (mean ⫾ SD)
Duration of hemodialysis (mean ⫾ SD)
Admitted from
home
nursing facility
another hospital
Cause of renal failure
hypertension
diabetes
glomerulonephritis
cystic kidney disease
other
unknown
Comorbidities
cancer
HIV
corticosteroids within 30 d
diabetes
injection drug use
previous infective endocarditis
none
Total APACHE II score (mean ⫾ SD)
Infection with MRSA
32 (48.5%)
17 (42.5%)
0.55
18 (27.3%)
46 (69.7%)
2 (3.0%)
55.3 ⫾ 15.0
1.87 ⫾ 2.24c
2 (5.0%)
38 (95.0%)
0 (0%)
57.7 ⫾ 15.5
4.26 ⫾ 4.01c
0.02
54 (81.2%)
10 (15.2%)
1 (3.6%)
35 (87.2%)
3 (7.5%)
2 (5.0%)
0.46
34 (51.5%)
35 (53.0%)
3 (4.6%)
1 (1.5%)
8 (12.1%)
0 (0%)
25 (62.5%)
16 (40.0%)
4 (10.0%)
0 (0%)
3 (7.5%)
2 (5.1%)
0.27
0.19
0.27
0.43
0.45
0.29
3 (4.6%)
1 (1.5%)
6 (9.1%)
39 (59.1%)
5 (7.6%)
3 (4.5%)
14 (21.2%)
19.8 ⫾ 5.4
24 (36.4%)
1 (2.5%)
0 (0%)
2 (5.0%)
19 (47.5%)
4 (10.0%)
0 (0%)
13 (32.5%)
17.5 ⫾ 4.4
11 (37.5%)
0.59
0.43
0.44
0.24
0.66
0.17
0.20
0.043
0.35
0.42
⬍0.001
a
APACHE II, Acute Physiology and Chronic Health Evaluation II; AV, arteriovenous; MRSA, methicillin-resistant
Staphylococcus aureus; SAB, S. aureus bacteremia.
b
P values for continuous variables are based on t tests or Wilcoxon rank sum tests; P values for categorical variables are
based on ␹2 tests.
c
Missing data for three patients in the catheter group and three patients in the graft group.
521
Table 2. Unadjusted resource utilization, outcomes, and costs among patients with ESRD and SAB during their
initial hospitalization and at 3-monthsa
Tunneled, Cuffed
Catheter (n ⫽ 66)
AV Graft
(n ⫽ 40)
Difference with
95% CI or P
Resource utilization
initial hospitalization
inpatient days (mean ⫾ SD)
10.6 ⫾ 10.1
10.3 ⫾ 6.4
0.79
procedures related to (mean ⫾ SD)
foreign device
4.1 ⫾ 3.1
2.3 ⫾ 2.5
0.001
nonforeign device
0.3 ⫾ 1.1
0.1 ⫾ 0.3
0.75
12-wk period
rehospitalizationb
8 (13.8%)
9 (23.7%)
0.21
inpatient days (mean ⫾ SD)
12.6 ⫾ 11.1
12.4 ⫾ 6.8
0.33
Outcomes
metastatic infection
22 (33.3%)
12 (30.0%)
0.72
died
8 (12.1%)
2 (5.0%)
0.22
12-wk period
died
15 (22.7%)
4 (10.0%)
0.098
Costs
(n ⫽ 39)
(n ⫽ 23)
mean ⫾ SD
$16,679 ⫾ 12,063
$17,509 ⫾ 11,267
⫺$830 (⫺$7,161 to $5,117)
median (⫾IQR)
$13,776 (⫾82961 to 18,698) $13,663 (⫾10,868 to 23,607)
12-wk period
mean ⫾ SD
$22,944 ⫾ 18,278
$23,969 ⫾ 13,731
⫺$1,024 (⫺$8,459 to $6,792)
median (⫾IQR)
$17,392 (⫾11,945 to 29,425) $18,318 (⫾13,663 to 31,660)
a
CI, confidence interval; IQR, interquartile range.
Patients who died during initial hospitalization were excluded.
b
vascular event (n ⫽ 2), epidural abscess (n ⫽ 1), and other (n ⫽
5). More than one metastatic complication could be present in
the same patient.
Inpatient mortality during the initial hospitalization was
12.1% for patients with catheters and 5.0% for patients with AV
grafts (P ⫽ 0.22). Unadjusted 3-mo mortality did not differ
significantly on the basis of hemodialysis access type (22.7
versus 10.0% for catheter versus graft; P ⫽ 0.098). In the 15
patients who had catheters and died, cause of death was related
to SAB in eight and unrelated in seven. All four deaths in
patients with AV grafts were due to SAB or its complications.
Cost Analysis
Patients with catheters had similar mean costs during the
initial hospitalization for SAB ($16,679 ⫾ $12,063) compared
with patients with AV grafts ($17,509 ⫾ $11,267; difference
⫺$830; 95% CI ⫺$7,161 to $5,117). Mean total 3-mo costs that
were associated with SAB also were similar between patients
with catheters ($22,944 ⫾ $18,278) and patients with AV grafts
($23,969 ⫾ $13,731; difference ⫺$1,024; 95% CI ⫺$8,459 to
$6,792).
Multivariable Analysis
After adjustment for patient characteristics, the odds of death
at 3 mo did not differ significantly by access type (Table 3). In
contrast, an increasing APACHE II score (odds ratio [OR] 1.19;
95% CI 1.02 to 1.39) and bacteremia caused by MRSA (OR 10.33;
95% CI 2.15 to 49.66) were associated with increased mortality
at 12 wk. Twelve-week costs did not differ significantly between patients with catheters and patients with AV grafts
(means ratio 0.84; 95% CI 0.60 to 1.17) after adjustment for
patient characteristics.
Subgroup Analyses of Patients with Catheters
Among the 66 patients with catheters, 20 had alternative
active access (maturing or active AV grafts or fistulas), 18 had
failed AV grafts or fistulas, and the remaining 28 had no
alternative active or inactive access type. Inpatient and 12-wk
mortality did not differ between patients without alternative
active access (n ⫽ 46) compared with patients with maturing
AV fistulas or grafts (n ⫽ 20; inpatient mortality 13 versus 10%
[P ⫽ 0.73]; 12-wk mortality 26 versus 15% [P ⫽ 0.32]; catheter
only versus alternative active access, respectively). Among patients who had a catheter but no alternative active access, the
mean costs for the initial hospitalization were $18,162
(⫾$11,687) compared with $12,903 (⫾$12,740) among patients
with maturing AV grafts or fistulas (difference $5259; 95% CI
⫺$5056 to $12,627). At 12 wk, mean costs for patients who had
a catheter but no alternative access were $25,738 (⫾$19,415)
compared with $15,831 (⫾$13,198) for patients who had a catheter and a maturing AV graft or fistula (difference $9907; 95%
CI ⫺$944 to $20,183).
After adjustment for patient characteristics, there was no
difference in 12-wk mortality between patients who had a
catheter but no alternative access and those with a maturing AV
522
Table 3. Associations between clinical and demographic variables and mortality and costs among patients with
ESRD and SABa
12-Wk Mortality
12-Wk Costs
Variable
Odds Ratio
Catheter onlyb
Female gender
White race
Diabetes
Hypertension
History of kidney transplant
CNS involvement
Heart failure
Sepsis
APACHE II
MRSA
Admit from nursing facility/other hospital
a
1.63
0.24
0.90
1.72
3.02
0.29
3.32
2.11
3.93
1.19
10.33
0.69
95% CI
0.29
0.05
0.12
0.34
0.53
0.01
0.48
0.16
0.47
1.02
2.15
0.10
to
to
to
to
to
to
to
to
to
to
to
to
9.02
1.12
6.90
8.80
17.36
6.32
22.89
27.85
32.76
1.39
49.66
4.65
Means Ratio
0.84
0.89
1.19
1.12
0.86
1.05
1.39
1.54
1.14
0.97
1.09
1.45
95% CI
0.60
0.65
0.77
0.81
0.61
0.61
0.79
0.80
0.67
0.94
0.77
0.87
to
to
to
to
to
to
to
to
to
to
to
to
1.17
1.23
1.86
1.54
1.21
1.21
2.46
2.98
1.93
1.01
1.55
2.44
CNS, central nervous system.
Graft only served as the reference group.
b
graft or fistula (OR 1.71; 95% CI 0.25 to 11.75). In adjusted
analyses, 12-wk costs also did not differ significantly on the
basis of the absence or presence of alternative active access
(means ratio 1.52; 95% CI 0.98 to 2.35).
Discussion
SAB is a serious, common complication of intravascular access in hemodialysis-dependent patients (1). Our investigation
shows that morbidity, mortality, and costs that are associated
with an episode of SAB among hemodialysis-dependent patients are pronounced but do not vary significantly by different
forms of intravascular access.
The clinical impact of an episode of SAB in hemodialysis
patients is considerable and is highlighted by our cohort. Metastatic complications occurred in 34% of patients, including
infective endocarditis, osteomyelitis, and abscesses. The high
rate of metastatic complications among our patients is consistent with previous reports (10,12) and highlights the virulence
of S. aureus, particularly in an immunocompromised host with
an intravascular foreign device. It is interesting that the rate of
metastatic infections did not differ among patients on the basis
of their intravascular access type. This observation suggests
that the risk for metastatic complications that are associated
with SAB in hemodialysis patients is high regardless of the
potential source of infection. Overall mortality also was high in
this investigation but did not vary significantly among patients
with different vascular access types, even after adjustment for
differences in acute and chronic comorbidities.
Although vascular access type did not affect length of hospitalization, the number of procedures that were performed
during the initial hospitalization was highest among patients
with catheters. This reflects the standard practice at our institution of prompt removal of a hemodialysis catheter in any
patient with documented SAB whenever clinically possible.
Patients who had AV grafts and developed SAB also required
extensive medical therapy, because half of these patients required surgical removal or revision of their vascular access, and
approximately one quarter required re-hospitalization during
the follow-up period (37).
The findings from our cost analysis of hemodialysis-dependent patients with SAB are unique and show that overall costs
are pronounced (approximately $23,000 during 12 wk). Surprisingly, costs did not differ statistically between patients with
catheters and patients with AV grafts, despite the high surgical
costs associated with an AV graft revision/removal and replacement (38). As the total costs for treating the 111 hemodialysis patients who had SAB and were reported in this study
was $2,440,987, access-related SAB represents a significant economic burden to the health care system. It should be highlighted that these costs are in addition to the already substantial
health care costs associated with maintenance hemodialysis
and initial dialysis access placement (38 – 40).
There are a number of potential reasons that no significant
associations were noted between a patient’s hemodialysis access type and his or her ultimate clinical outcome. First, among
patients with AV grafts, complications and costs that are associated with an AV graft revision/resection can be significant
(37,38,41). In contrast, patients with catheters required a large
number of procedures (e.g., catheter removal and replacement,
prosthetic device removal, surgery for infective endocarditis),
which likely escalated the mortality rate and cost in this cohort.
The lack of our ability to detect a difference in costs among
patients who had a catheter and who did and did not have
alternative active access types may be due to the lack of power
within this subgroup analysis. Alternatively, the other active
access type may not have been readily usable at the time of
admission and therefore was unable to confer a cost savings; in
addition, if the maturing AV graft or fistula had been placed
recently, then management of postsurgical infectious complications may have increased costs in this subgroup.
Our study has a number of limitations. First, our data were
derived from an observational design with a modest sample
size, thereby limiting our ability to detect statistically significant differences between groups. Second, groups that were
based on access type had distinct differences in comorbidities;
however, we tried to identify and adjust for all potential confounders between groups in our statistical analysis. Third, although we did not capture costs for patients who were admitted to outside hospitals, follow-up was complete on all patients
at 12 wk, thus making the risk for missing an outside hospital
admission small. Fourth, we did not include in our cost analysis
outpatients who were treated for bacteremia, although this is
less relevant as it is standard practice for patients with documented SAB to be treated as inpatients and therefore this
should not have affected our results. Finally, referral bias may
overestimate the cost and outcome of SAB if only sicker patients were admitted to our institution. This risk is minimized
by the fact that the majority of patients in this study are cared
for by university-affiliated hemodialysis units that use our
institution for all hospital admissions.
Despite these limitations, our cohort highlights the significant morbidity, mortality, and costs that are associated with an
episode of SAB among hemodialysis patients. Furthermore, our
investigation suggests that a patient’s hemodialysis access type
is not significantly associated with differential costs or clinical
outcomes during an episode of SAB, even if the vascular access
is readily removable. Given the significant patient, physician,
and cost burden associated with an episode of SAB in patients
with ESRD, effective interventions to try to prevent S. aureus
infections in this high-risk group are warranted.
Acknowledgments
This research was supported in part by a research agreement between Duke University and Nabi Biopharmaceuticals. J.K.I. was supported by a clinical research fellowship endowment from Earl Dalbey
and by National Institutes of Health grant K12 RR-017630. V.G.F. was
supported by National Institutes of Health grant R01-AI059111.
This study was presented in abstract form at the European Renal
Association-European Dialysis and Transplant Association Annual
Congress, Lisbon, Portugal, May 15 to 18, 2004.
References
1. US Renal Data Systems: USRDS 2003 Annual Data Report:
Atlas of End-Stage Renal Disease in the United States, Bethesda, National Institutes of Health, National Institute of
Diabetes and Digestive and Kidney Diseases, 2003
2. Krishnasami Z, Carlton D, Bimbo L, Taylor ME, Balkovetz
DF, Barker J, Allon M: Management of hemodialysis catheter-related bacteremia with an adjunctive antibiotic lock
solution. Kidney Int 61: 1136 –1142, 2002
3. Mokrzycki MH, Schoppel B, Von Gersdorff G, Rush H,
Zdunek MP, Feingold R: Tunneled-cuffed catheter associated infections in hemodialysis patients who are seropositive for the human immunodeficiency virus. J Am Soc
Nephrol 11: 2122–2127, 2000
4. Boelaert JR, Van Landuyt HW, Gordts BZ, De Baere YA,
Messer SA, Herwaldt LA: Nasal and cutaneous carriage of
Staphylococcus aureus in hemodialysis patients: The effect of
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
523
nasal mupirocin. Infect Control Hosp Epidemiol 17: 809 – 811,
1996
Chow JW, Yu VL: Staphylococcus aureus nasal carriage in
hemodialysis patients: Its role in infection and approaches
to prophylaxis. Arch Intern Med 149: 1258 –1262, 1989
Saad TF: Bacteremia associated with tunneled, cuffed hemodialysis catheters. Am J Kidney Dis 34: 1114 –1124, 1999
Dryden MS, Samson A, Ludlam HA, Wing AJ, Phillips I:
Infective complications associated with the use of the
Quinton ‘Permcath’ for long-term central vascular access
in haemodialysis. J Hosp Infect 19: 257–262, 1991
Beathard GA: Management of bacteremia associated with
tunneled-cuffed hemodialysis catheters. J Am Soc Nephrol
10: 1045–1049, 1999
Tanriover B, Carlton D, Saddekni S, Hamrick K, Oser R,
Westfall AO, Allon M: Bacteremia associated with tunneled dialysis catheters: Comparison of two treatment
strategies. Kidney Int 57: 2151–2155, 2000
Poole CV, Carlton D, Bimbo L, Allon M: Treatment of
catheter-related bacteremia with an antibiotic lock protocol: Effect of bacterial pathogen. Nephrol Dial Transplant 19:
1227–1244, 2004
Marr KA, Sexton DJ, Conlon PJ, Corey RG, Schwab SJ,
Kirkland KB: Catheter-related bacteremia and outcome of
attempted salvage in patients undergoing hemodialysis.
Ann Intern Med 127: 275–280, 1997
Marr KA, Kong L, Fowler VG, Gopal A, Sexton DJ, Conlon
PJ, Corey GR: Incidence and outcome of Staphylococcus
aureus bacteremia in hemodialysis patients. Kidney Int 54:
1684 –1689, 1998
Schwab SJ, Beathard GA: The hemodialysis catheter conundrum: Hate living with them, but can’t live without
them. Kidney Int 56: 1–17, 1999
Sexton DJ: Vascular access infections in patients undergoing dialysis with special emphasis on the treatment of
Staphylococcus aureus. Infect Dis Clin North Am 15: 731–742,
2001
Hoen B, Paul-Dauphin A, Hestin D, Kessler M: EPIBACDIAL:
A multicenter prospective study of risk factors for bacteremia in chronic hemodialysis patients. J Am Soc Nephrol 9:
869 – 876, 1998
Powe NR, Jaar B, Furth SL, Hermann J, Briggs W: Septicemia in dialysis patients: Incidence, risk factors, and prognosis. Kidney Int 55: 1081–1090, 1999
Pastan S, Soucie M, McClellan WM: Vascular access and
increased risk of death among hemodialysis patients. Kidney Int 62: 620 – 626, 2002
Allon M, Depner TA, Radeva M, Bailey J, Beddhu S, Butterly D, Coyne DW, Gassman JJ, Kaufman AM, Kaysen
GA, Lewis JA, Schwab SJ; HEMO Study Group: Impact of
dialysis dose and membrane on infection-related hospitalization and death: Results of the HEMO study. J Am Soc
Nephrol 14: 1863–1870, 2003
Tokars JI, Miller ER, Alter MJ, Arduino MJ: National surveillance of dialysis-associated diseases in the United
States, 2000. Semin Dial 15: 162–171, 2002
Pisoni RL, Young EW, Dykstra DM, Greenwood RN, Hecking E, Gillespie B, Wolfe RA, Goodkin DA, Held PJ: Vascular access use in Europe and in the United States: Results
from the DOPPS. Kidney Int 61: 305–316, 2002
National Kidney Foundation: K/DOQI: Clinical practice
guidelines for chronic kidney disease: Evaluation, classifi-
524
22.
23.
24.
25.
26.
27.
28.
29.
30.
cation, and stratification. Available: http://www.kidney.
org. Accessed March 1, 2005
Rubin RJ, Harrington CA, Poon A, Dietrich K, Greene JA,
Moiduddin A: The economic impact of Staphylococcus aureus infection in New York City hospitals. Emerg Infect Dis
5: 9 –17, 1999
Abramson MA, Sexton DJ: Nosocomial methicillin-resistant and methicillin-susceptible Staphylococcus aureus primary bacteremia: At what costs? Infect Control Hosp Epidemiol 20: 408 – 411, 1999
Engemann JJ, Friedman JY, Reed SD, Griffiths RI, Szczech
LA, Kaye KS, Stryjewski ME, Reller B, Schulman KA, Corey RG, Fowler VF: Clinical outcomes and costs due to
Staphylococcus aureus bacteremia among patients on
chronic hemodialysis. Infect Control Hosp Epidemiol 36: 534 –
539, 2005
Reed SD, Friedman JY, Engemann JJ, Griffiths RI, Anstrom
KJ, Kaye KS, Stryjewski ME, Szczech LA, Reller B, Corey
RG, Schulman KA, Fowler VF: Costs and outcomes among
hemodialysis-dependent patients with methicillin-resistant or methicillin-susceptible Staphylococcus aureus bacteremia. Infect Control Hosp Epidemiol 26: 175–183, 2005
Knaus WA, Draper EA, Wagner DP, Zimmerman JE:
APACHE II: A severity of disease classification system.
Crit Care Med 13: 818 – 829, 1985
American College of Chest Physicians/Society of Critical
Care Med Consensus Conference: Definitions for sepsis
and organ failure and guidelines for the use of innovative
therapies in sepsis. Crit Care Med 20: 864 – 874, 1992
Durack DT, Lukes AS, Bright DK: New criteria for diagnosis of infective endocarditis: Utilization of specific echocardiographic findings. Duke Endocarditis Service. Am J
Med 96: 200 –209, 1994
National Committee for Clinical Laboratory Standards:
Performance Standards for Antimicrobial Susceptibility Testing:
Twelfth Informational Supplement, Wayne, PA, National
Committee for Clinical Laboratory Standards, 2002 [Approved standard M100 –S112]
Bureau of Labor Statistics Data: US Department of Labor
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
Statistics web site. Available at http://www.bls.gov/data.
Accessed October 15, 2005
Thompson SG, Barber JA: How should cost data in pragmatic randomised trials be analyzed? BMJ 320: 1197–1200,
2000
Briggs A, Gray A: The distribution of health care costs and
their statistical analysis for economic evaluation. J Health
Serv Res Policy 3: 233–245, 1998
Zhou XH, Melfi CA, Hui SL: Methods for comparison of
cost data. Ann Intern Med 127: 752–756, 1997
Manning WG, Mullahy J: Estimating log models: To transform or not to transform? J Health Econ 20: 461– 494, 2001
Blough DK, Ramsey SD: Using generalized linear models
to assess health care costs. Health Services and Outcomes
Research Methodology 1: 185–202, 2000
Manning WG, Basu A, Mullahy J: Generalized modeling
approaches to risk adjustment of skewed outcomes data.
J Health Econ 24: 465– 488, 2005
Minga TE, Flanagan KH, Allon M: Clinical consequences of
infected arteriovenous grafts in hemodialysis patients.
Am J Kidney Dis 38: 975–978, 2001
Lee H, Manns B, Taub K, Ghali WA, Dean S, Johnson D,
Donaldson C: Cost analysis of ongoing care of patients
with end-stage renal disease: The impact of dialysis modality and dialysis access. Am J Kidney Dis 40: 611– 622,
2002
Manns B, Tonelli M, Yilmaz S, Lee H, Laupland K, Klarenbach S, Radkevich V, Murphy B: Establishment and maintenance of vascular access in incident hemodialysis patients: A prospective cost analysis. J Am Soc Nephrol 16:
201–209, 2005
United States Renal Data System: The economic cost of
ESRD, vascular access procedures, and Medicare spending
for alternative modalities of treatment. USRDS. Am J Kidney Dis 30[Suppl 1]: S160 –S177, 1997
Tabbara MR, O’Hara PJ, Hertzer NR, Krajewski LP, Beven
EG: Surgical management of infected PTFE hemodialysis
grafts: Analysis of a 15-year experience. Ann Vasc Surg 9:
378 –384, 1995
Fiche de lecture critique d’une étude clinique ou épidémiologique DIALIN (V3 08/03/2007)
(d’après SALMI L.R. Lecture critique et rédaction médicale scientifique. Comment lire, rédiger et publier une étude clinique ou épidémiologique. Paris :Elsevier ; 1995)
Référence : N° BLA2588
Auteurs : Inrig JK, Reed SD, Szczech LA, Engemann JJ, Friedman JY, Corey GR, Schulman
KA, Reller LB, Fowler VG, Jr. ....................................................................................................................
Titre : Relationship between clinical outcomes and vascular access type among
hemodialysis patients with Staphylococcus aureus bacteremia. ..............................................
Revue : Clin J Am Soc Nephrol. ...............................................................................................................
Année : 2006 ..................
Volume (n°) : 1 (3)) ..Pages : 518-524 .......................................................
Résumé :
Objectif et justification : Comparer l’utilisation des ressources, les coûts et les résultats cliniques chez les
patients hémodialysés atteints d’une bactériémie à Staphylococcus aureus (SAB) selon le type d’accès
vasculaire.
Schéma d’étude ; groupes comparés : Etude de cohorte prospective avec 3 groupes : cathéters, fistule
greffe, FAV.
Population étudiée et nombre de sujets : 143 patients hémodialysés hospitalisés entre juillet 1996 et août
2001 dans un hôpital universitaire de Caroline du Nord avec SAB : 111 (77.6%) pour bactériémie suspectée, 40
(36%) pour une autre raison.
Conduite de l’étude : Sur les 111 patients, 66 (59.5%) cathéters, 40 (36%) fistule greffe et 5 (4.5%) FAV (ce
dernier groupe trop peu nombreux étant exclus des analyses
Principaux résultats :
Sur les analyses ajustées :
Pas de différence significative de mortalité à 12 semaines (OR cathéter/greffe 1.63 [0.29 9.03]95%
Pas de différence significative de coûts à 12 semaines (OR cathéter/greffe 0.84 [0.60 1.17]95%
Cocher la case correspondant au respect du critère : 0 = oui, N = non, I = incomplet ; NA = ne s’applique pas,
NSP = ne sait pas. Une réponse cochée N à un critère en italique = étude inacceptable.
Critères
Schéma d’étude
Formulation claire de l’objectif
Hypothèse faite à priori
Schéma adapté à l’objectif
Procédures de sélection
Critères d’inclusion décrits et adéquats
Critères d’exclusion décrits et adéquats
Indication du nombre de refus avant l’étude
Procédures identiques dans tous les groupes
Taille d’échantillon adaptée
Règles éthiques respectées
Conduite de l’étude
Retraits indiqués, expliqués et raisonnables
Retraits équilibrés entre les groupes
Mesures principales fiables et valides
Analyse des résultats
Méthodes adaptées à la question et au schéma
Prise en compte des variables importantes
Résultats vérifiables des données brutes
Prise en compte des comparaisons multiples
Conclusion :
Qualité
Exceptionnelle
Très bonne
Plutôt bonne
Faible mais acceptable
Inacceptable
Pas qualifié pour juger
O
I
N
NA
NSP
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
Discussion, forces et faiblesses : Echantillon faible
Cocher
Faits avérés ou réfutés :
X
Pas de différence significative de mortalité ajustée ou de coûts
ajustés selon le type d’accès vasculaire
Décision
Mettre un seul X devant un des deux mots clefs
X
INC
A inclure dans la biblio
EXC
A ne pas inclure dans la biblio
Justifications de l’inclusion
Justifications de l’exclusion
IAV
Infection de l’accès vasculaire
HOS
Hors sujet
X
BAC
Bactériémie
MAQ
Mauvaise qualité
HCV
Hépatite virale C
AUT
Autre
HIV
SIDA
En clair :
VIR
Autres viroses
TIC
Taux d’incidence
TPR
Taux de prévalence
X
FAC
Facteurs de risque (de confusion)
REC
Recommandations
SUR
Surveillance
STA
Standardisation des résultats
INO
Innovation
EXC
Exceptionnel, publication princeps
Autres mots clefs réservés à la gestion des références bibliographiques
SEL
En cours de sélection
NOL
Non lu (non retenu par la sélection)
LEC
En cours de lecture (retenu par la sélection)
Lecteur : LA
Commentaires
Date : 8/1/2009

La lecture critique de la littérature médicale dans le - CClin Sud-Est

Transcription

Documents pareils

www.icc-informatique.fr PLAN DE COURS ACCESS INITIATION 2

LOGICIEL MICROSOFT ACCESS

1ère rencontre du Scientific Advisory Board (SAB)

Formation Access perfectionnement

9 règles d`or pour mettre en place des tableaux de bord

MODULE ACCESS : AIDE A LA CREATION DE VOTRE

Access - Conception d`une base de données

Athéna TSINGARIDA, Quìmporte le flacon pourvu quòn ait lìvresse