Assessing Severity of Patients With Community-Acquired Pneumonia

Please download to get full document.

View again

of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Information Report
Category:

Documents

Published:

Views: 2 | Pages: 12

Extension: PDF | Download: 0

Share
Related documents
Description
itd
Tags
Transcript
  Assessing Severity of Patients withCommunity-Acquired Pneumonia  Jose Manuel Pereira, M.D. 1  Jose Artur Paiva, M.D., Ph.D. 1  Jordi Rello, M.D., Ph.D. 2 1 EmergencyandIntensiveCareDepartment,CentroHospitalarS.JoãoEPE, Faculdade de Medicina da Universidade do Porto, Porto,Portugal. 2 Critical Care Department, Vall d'Hebron University Hospital, Institutde Recerca Vall d'Hebron, CIBERes, Universitat Autonoma Barcelona,Barcelona, Spain.Semin Respir Crit Care Med 2012;33:272 – 283. Address for correspondence and reprint requests  José ManuelPereira, M.D., Emergency and Intensive Care Department, CentroHospitalar S. João EPE, Al Prof Hernâni Monteiro, 4200-319 Porto,Portugal (e-mail: jmcrpereira@yahoo.com). Community-acquiredpneumonia(CAP),describedin1892asthe  “ Captain of the men of Death ”  by Sir William Osler, 1 remains a common and serious worldwide health problemdespite all the advances in therapy with the emergence of potent and broad-spectrum antibiotics.TheoverallmortalityofCAPvaries, 2 – 7 reachingalmost50%in intensive care unit (ICU) patients requiring vasopressorsupport. 8 To mitigate negative outcomes, it is essential toidentify patients with severe CAP as soon as possible.Current international guidelines 9,10 consider severity of illnessassessmentavitalcomponentofpatient management,affecting not only site of care and diagnostic workup but alsoempiricalantibiotic therapyand adjuvant treatment.Howev-er, accurately assessing severity in CAP can be a challenge tophysicians.Thereisincreasingevidencethatcliniciansmaybothover-and underestimate the severity of CAP, particularly whenrelying on clinical judgment only, and therefore may oftendecide to inappropriately hospitalize or discharge patientsrequiring a different care pathway. Despite their widespreadusein clinicalpractice,traditional markerssuchasseverityof disease estimation by the patient, fever, or white blood cellcounts do not reliably assess disease severity and mortalityrisk. 11 The accuracy of several tools, mainly clinical scores andbiomarkers, to predict severe CAP has been extensivelystudied. This review discusses the available instruments toassess CAP severity, their role in the clinical practice, theiradvantages, and their limitations. Pneumonia-Speci 󿬁 c Severity Scores Clinical judgment has often been proved inadequate to thetaskofassessingseverity. 12 – 15 In1997,Fineetal 16 introduced Keywords ►  severity  ►  community-acquiredpneumonia ►  scores ►  biomarkers ►  bacterial load ►  DNAemia  Abstract  Despite all advances in its management, community-acquired pneumonia (CAP) is still animportantcauseofmorbidityandmortalityrequiringagreatconsumptionofhealth,social,and economic resources. An early and adequate severity assessment is of paramountimportancetoprovideoptimizedcaretothesepatients.Inthelast2decades,thisissuehasbeen the subject of extensive research. Based on 30 day mortality, several prediction ruleshave been proposed to aid clinicians in deciding on the appropriate site of care. In spite of beingwellvalidated,theirsensitivityandspeci 󿬁 cityvary,whichlimitstheir widespreaduse.Theutilityofbiomarkerstoovercomethisproblemhasbeeninvestigated.At thismoment,their full clinical value remains undetermined, and no single biomarker is consistently idealfor assessing CAP severity. Biomarkers should be seen as a complement rather thansuperseding clinical judgment or validated clinical scores. The search for a gold standard isnotover,andnew tools,likebacterialDNAload,areinthepipeline.Untilthen,CAPseverity assessment should be based in three key points: a pneumonia-speci 󿬁 c score, biomarkers,and clinical judgment. Issue Theme  Global Trends inCommunity-Acquired Pneumonia; GuestEditors, Marcos I. Restrepo, M.D., M.Sc.,F.C.C.P. and Antonio Anzueto, M.D.Copyright © 2012 by Thieme MedicalPublishers, Inc., 333 Seventh Avenue,New York, NY 10001, USA.Tel: +1(212) 584-4662. DOI  http://dx.doi.org/10.1055/s-0032-1315639. ISSN  1069-3424. 272    D  o  w  n   l  o  a   d  e   d   b  y  :   W  o  r   l   d   H  e  a   l   t   h   O  r  g  a  n   i  z  a   t   i  o  n   (   W   H   O   ) .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .  the best known of the prediction models, the PneumoniaSeverityIndex(PSI),followingastudyinover40,000patients.Thisscore,whichis based on 20 demographic, comorbid,andclinical variables was developed primarily to identify thosepatients who can safely be treated as outpatients and hasbeen validated in several, large, independent studies. 2,17 – 20 PSI strati 󿬁 espatients into 󿬁 ve riskclassesbased on the riskof death within 30 days: three with low riskof 30 day mortality(class I  ¼  0.1 to 0.4%; class II  ¼  0.6 to 0.7%; class III  ¼  0.9 to2.8%), a fourth with an increased risk (4 to 10%), and a  󿬁 fthwith a high risk (27%). 16 It performs consistently well as apredictor of mortality in CAP with aROC (area under thereceiver operating characteristic curve) values ranging from0.70 to 0.89 2,17,20,21 ( ►  Table 1 ). PSI is also a fair predictor of intensive care unit (ICU) admission (aROC ranging from 0.56to 0.85), 21 and it is recommended by the current AmericanThoracic Society/Infectious Diseases Societyof America (ATS/IDSA) guidelines to decide site of care. 9 According to the riskclass, patients can be treated as outpatients (class I or II) oradmitted to the hospital/ICU (class IV or V). Class III patientscan be managed as an outpatient or brief inpatient. Still, theoriginal score needs to be modi 󿬁 ed when it is used todetermine whether hospital admission is necessary. Hypoxia(arterial saturation of less than 90% or an arterial oxygenpressure lower than 60 mm Hg) as a complication of pneu-monia should be added as a sole indicator for admission forpatientsinriskclassesItoIIIasanadded “ marginofsafety. ” 16 Besides its complexity, which limits its implementation inclinical practice,thisscorehas otherlimitations ( ►  Table 2 ).Itoverestimates severity in older patients with comorbiditiesand, on the other hand, it may underestimate severity inyounghealthypatientswithsevererespiratoryfailure,aswasrecently demonstrated during the 2009 in 󿬂 uenza A (H1N1)pandemic. 22,23 It should also be stressed that this scale doesnotincludeimportant risk factorssuchaschronicobstructivepulmonary disease (COPD), diabetes, or other medical orpsychological conditions that may justify hospital admission.The fact that it relies on laboratory data limits its use bygeneral practitioners.The British Thoracic Society (BTS) proposes use of CURB-65 (confusion, urea  > 7 mmol/L; respiratory rate   30/min;blood pressure — systolic  < 90 mm Hg or diastolic   60 mmHg; age   65 years) rule, which was introduced in 2003 andhas been validated in over 12,000 patients from severaldifferentcountries. 10 Recently,ithasbeenshownthatsystolicblood pressure < 90 mm Hg alone is effective for risk predic-tion, providing a simpli 󿬁 cation of the srcinal score becausediastolicmeasurementsaddnoadditionalinformation. 24 Thisseverity scale is a 5-point scoring system with three riskcategories: 0 to 1 (low risk of mortality; class 0  ¼  0.7%; class1  ¼  3.2%), 2 (intermediate risk of mortality is 13%) and   3(high risk of mortality; class 3  ¼  17%; class 4  ¼  41.5%; class5  ¼  57%). 7 Studies assessing CURB-65 have shown it to be agood tool for prediction of mortality with an aROC rangingfrom 0.73 to 0.87. 21 CURB-65 can also be useful in determin-ing which patients may safely be treated at home (CURB-65class 0 or 1) and can  󿬂 ag certain hospitalized patients foradmission tothe ICU if theircondition deteriorates (CURB-65class   3). Class 2 patients can be treated as supervisedoutpatient or be admitted for a short hospital stay. As withPSI, data from different studies indicate a lesser performanceof this scorefor predicting ICU admission (aROC 0.60 to 0.78) 21 ( ►  Table 1 ).Two major differences between CURB-65 and PSI are thatthe former is easier to calculate, which favors its implemen-tation, and it does not directly address underlying disease.One limitation of this scoring system is that it mayunderestimate risk in elderly patients with comorbidities.Like PSI, another disadvantage of CURB-65 is its reliance onlaboratory investigations for calculation, which limits its useoutside the hospital setting ( ►  Table 2 ). This led to thedevelopment of a simpli 󿬁 ed version omitting the bloodurea nitrogen testing: the CRB-65. This modi 󿬁 ed BTS scorehas demonstrated its equivalence in risk strati 󿬁 cation com-paredwithbothPSIandCURB-65andcanbeusedbyprimarycare physicians to determine if severity is high enough towarrant hospital admission (CRB-65   2). 25 – 27 Most studieshave assessed CRB in the hospital setting, and only one has  Table 1  Discriminatory Power for 30-Day Mortality and Complications (ICU Admission or Need for Mechanical Ventilation and/orInotropic Support) of Different Pneumonia Speci 󿬁 c Clinical Scores Severity Score AUC for 30-Day Mortality AUC for CAP Complications PSI 0.70 – 0.89 0.58 – 0.85CURB-65 0.73 – 0.87 0.60 – 0.78CRB-65 0.69 – 0.78 0.57 – 0.77IDSA/ATS 2007 0.63 – 0.67 0.85 – 0.88SMART-COP Not assessed 0.83 – 0.87SCAP Not assessed 0.75 – 0.83PIRO-CAP 0.88 Not assessed AUC: area under the curve; CAP, community-acquired pneumonia; CRB-65, confusion, respiratory rate, blood pressure, 65 years of age; CURB-65,confusion, urea, respiratory rate, blood pressure, 65 years of age; IDSA/ATS, Infectious Diseases Society of America/AmericanThoracic Society; PIRO-CAP, predisposition, insult, response, organ dysfunction – community-acquired pneumonia; PSI, Pneumonia Severity Index; SCAP, severe community-acquired pneumonia; SMART-COP, systolic blood pressure, multilobar chest radiography involvement, albumin, respiratory rate, tachycardia,confusion, oxygenation, arterial pH. Seminars in Respiratory and Critical Care Medicine Vol. 33 No. 3/2012 Assessing Severity of Patients with Community-Acquired Pneumonia  Pereira et al.  273    D  o  w  n   l  o  a   d  e   d   b  y  :   W  o  r   l   d   H  e  a   l   t   h   O  r  g  a  n   i  z  a   t   i  o  n   (   W   H   O   ) .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .  examined its performance in a community setting 28 ; there-fore, further validation of this score in this setting ismandatory.Inarecentmeta-analysis, 29 theperformanceofPSIandthethree main iterations of CURB in predicting mortality fromCAP were evaluated. PSI was the most sensitive (0.90, 95% CI0.87 to 0.92) and less speci 󿬁 c (0.53, 95% CI 0.46 to 0.59) testwith a low false-negative rate, thus giving clinicians greatercon 󿬁 dence in identifying patients who have nonsevere CAPand may not need hospital admission. Conversely, CURB-65,CRB-65, and CURB scales were more speci 󿬁 c (speci 󿬁 cityranging from 0.77 to 0.92) and had higher positive predictivevalues than the PSI, which means a greater proportion of patientsinthehigherriskcategorieswerecorrectlyclassi 󿬁 ed.ThepoorersensitivityoftheCURB-65scalesmeansthatsomepatients may be incorrectly diagnosed and managed as non-severe even when they are actually at higher risk of death.According to these results, PSI may be preferred in settingswhere pneumonia mortality is relatively high due to its highsensitivity. Conversely, in settings with limited resourcesandwhere mortality is relatively low, the lower sensitivity of theCRB-65 is not a disadvantage, and its ease of use and higherspeci 󿬁 city may help clinicians to focus on those requiringmore clinical attention. 29 Up to 50% of deaths from CAP are unrelated to initialseverity. 30,31 Therefore,30daymortalitymaynotbetheidealmeasure to identify patients with severe pneumonia thatrequire the most intensive treatment, and other end pointssuchasICUadmissionshouldbesought.Infact,recentstudieshave identi 󿬁 ed delayed ICU admission for CAP patients as arisk factor for short-term mortality. 32 A scoring system thatcould accuratelyidentify patients needingICUadmission andtherefore allow earlier intensive therapy is potentiallydesirable.To address this issue, the IDSA/ATS recently reviewed riskfactors and developed major and minor criteria to identifypatients who require direct ICU admission. 9 These majorcriteria include need for invasive mechanical ventilation orvasopresssor support. For patients who do not meeteither of these two criteria, minor criteria have been purposed basedon CURB-65 and the original ATS guidelines 33 with newadditions. According to these guidelines, patients who ful 󿬁 llone of the major criteria or at least three of these minorcriteria should be admitted to an ICU.These criteria were subsequently validated by Brownet al. 34 Using as reference intensive care therapy in the ICUto de 󿬁 ne severe CAP, they observed that the 2007 IDSA/ATScriteria performed signi 󿬁 cantly better than CURB-65,SMART-COP (systolic blood pressure, multilobar chest radi-ography involvement, albumin, respiratory rate, tachycardia,confusion,oxygenation,arterialpH),andEspañaRulewithanaROCof0.88(95%CI0.85to0.90).Interestinglythisde 󿬁 nitionof severe pneumonia when compared with others proposedwas associated with similar 30 day mortality but clearlyindicated a group of patients with longer hospital length of stay.In a meta-analysis, 21 the predictive accuracy of 2001 ATScriteria 33 and 2007 IDSA/ATS criteria 9 for severe CAP wereanalyzed.Althoughthepooledsensitivityof2001ATScriteriadecreased from 66.7% to 61.2% according to 2007 IDSA/ATS  Table 2  Advantages and Limitations of Pneumonia-Speci 󿬁 c Scores Severity Score Advantages Limitations PSI Well validatedImproves outcomeGood performance inlow mortality risk patientsComplex to calculateOveremphasis of age and comorbiditiesExcludes risk factors such as COPD and diabetesPerforms less well for need for ICU/ventilatory or vasopressor supportLimited use outside hospital settingCURB-65 Well validatedSimple to calculateUnderestimates severity in young patientsDoes not take into account comorbiditiesPerforms less well for need for ICU/ventilatory or vasopressor supportLimited use outside hospital settingCRB-65 Suitable for community setting As for CURB-65IDSA/ATS 2007 Good performance forpredicting ICU admissionMinor criteria may be usefulto identify high-risk patientsNeed for ICU is not the most accurate measureof severity due to intercenter variability SMART-COP Good accuracy for predictionof need for ventilatory orvasopressor supportComplex to calculateMay underestimate severity in youngand previously   󿬁 t patientsPIRO-CAP Risk strati 󿬁 cation in high-risk patients Not widely validated CAP, community-acquiredpneumonia;COPD,chronicobstructive pulmonary disease;CRB-65,confusion,respiratoryrate,bloodpressure,65yearsof age; CURB-65, confusion, urea, respiratory rate, blood pressure, 65 years of age; ICU, intensive care unit; IDSA/ATS, Infectious Diseases Society of America/AmericanThoracicSociety;PIRO-CAP,predisposition,insult,response,organdysfunction – community-acquiredpneumonia;PSI,PneumoniaSeverity Index; SCAP, severe community-acquired pneumonia; SMART-COP, systolic blood pressure, multilobar chest radiography involvement,albumin, respiratory rate, tachycardia, confusion, oxygenation, arterial pH. Seminars in Respiratory and Critical Care Medicine Vol. 33 No. 3/2012 Assessing Severity of Patients with Community-Acquired Pneumonia  Pereira et al. 274    D  o  w  n   l  o  a   d  e   d   b  y  :   W  o  r   l   d   H  e  a   l   t   h   O  r  g  a  n   i  z  a   t   i  o  n   (   W   H   O   ) .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .  criteria, the pooled speci 󿬁 city increased from 84.6% to 88.6%.However, both criteria had higher sensitivity than equivalentcutoffsofthePSIandCURB-65/CRB-65scores.Theaccuracyof modi 󿬁 edATScriteriatopredict30daymortalityevaluatedbythe aROC is poor (0.63 to 0.67) 2,18 ( ►  Table 1 ).Data from three studies 35 – 37 using only the minor criteriashowed a low sensitivity (55.7%) but a very high speci 󿬁 city(91.7%) with a negative likelihood ratio of 0.51 (0.38 to 0.67).Probably, minor criteria are more helpful to identify a groupof patients athighriskofcomplication andmortality withouta major indication for ICU admission. This was validated in arecently published manuscript by Chalmers et al, 38 showingthat minor criteria correctly identify a signi 󿬁 cant proportionof patients as high risk. They compared favorably withalternative scoring systems with aROC 0.85 (0.82 to 0.88)for prediction of mechanical ventilation or vasopressor sup-port, 0.82 (0.82 to 0.88) for prediction of ICU admission and0.78 (0.74 to 0.82) for prediction of 30 day mortality.Toimprovetheabilitytopredictwhichpatientswillrequireintensive respiratory or vasopressor support, a new tool wasdevelopedbyCharlesetal:SMART-COP. 39 Thisseverityscoreisbased on the following features: systolic blood pressure lowerthan 90 mm Hg (2 points), multilobar chest radiographyinvolvement (1 point), low albumin level (1 point), highrespiratory rate (1 point), tachycardia (1 point), confusion (1point), poor oxygenation (2 points), and low arterial pH (2points).Ascore  3identi 󿬁 ed92%ofthepatientswhoreceivedrespiratory or vasopressor support, including 84% of patientswhodidnotneedimmediateICUadmission.Thereisasmoothrelationship between increasing score and need for intensiverespiratory or vasopressor support, and its accuracy (aROC0.87) issigni 󿬁 cantly higher thanPSI(aROC 0.69)andCURB-65(aROC 0.67) ( ►  Table 1 ). This higher accuracy was validated in 󿬁 ve different cohorts with consistent results. A simpler index(SMARTCOP)wasalso derivedfor useinprimarysettingswithpredictive accuracy similar to the srcinal score (aROC 0.80).Using this tool a score of   2 identi 󿬁 ed 90% of the patients atpresentation who received intensive care. An important con-tribution was that respiratory rate had a different threshold(  25 breaths/min) in young patients. As we previously men-tioned, there are concerns that existing pneumonia severityscores may underestimate CAP severity in younger and previ-ously 󿬁 tpatients.Unfortunately,SMART-COPisnottheperfectsolution because it failed to identify 15% of patients whorequired mechanical ventilation and/or inotropic support ina population younger than 50 years 40 ( ►  Table 2 ).TheSevereCommunity-AcquiredPneumonia(SCAP)scoreorEspañarulewasintroducedin2006. 41 Thescoreconsistsof eight variables that are separated into two major (pH  < 7.30and systolic blood pressure  < 90 mm Hg) and six minorcriteria (altered mental status, respiratory rate  > 30breaths/minute, uremia, oxygen arterial pressure  < 54 mmHg or PaO 2 :FiO 2  ratio   250 mm Hg, age   80 years, andmultilobar/bilateral lung involvement). At a cutoff of    10,this prediction rule showed a high sensitivity (92.1%) andgood speci 󿬁 city (73.8%) for identi 󿬁 cation of severe CAP (in-hospital mortality, mechanical ventilation, and/or septicshock), which proved superior to modi 󿬁 ed ATS criteria orCURB-65 and similar to PSI. The presence of one majorcriterion or at least two minor criteria allows physicians toidentify with high accuracy (aROC 0.92) patients at risk of complicatedCAPthatwillbene 󿬁 tfromadditionalmonitoringand more aggressive treatment. In a subsequent study, con-ductedintwolargecohortsofpatientshospitalizedwithCAP,theSCAPscorewasslightlymoreaccuratethanPSIandCURB-65 in predicting adverse outcomes such as ICU admission,needfor mechanicalventilation,severesepsis,and treatmentfailure and performed best for all of them. 42 The discrimina-tory power was good for all outcomes, except for treatmentfailure, where it was poor (aROC 0.61).Other proposed scores include A-DROP, 43 SOAR, 6 andCORB. 44 The A-DROP is a scoring system proposed in 2005by the Japanese Respiratory Society to assess CAP severity. 43 Itisamodi 󿬁 edversionofCURB-65basedon 󿬁 veclinicalfeatures:age (male   70 years; female   75 years), dehydration (bloodurea nitrogen   210 mg/L), respiration failure (SaO 2    90% orPaO 2    60 mm Hg), orientation disturbance (confusion), andbloodpressure(systolicbloodpressure  90 mmHg).Themaindifference with CURB-65 is the evaluation of the respiratorycondition. This prediction rule seems to be an accurate andclinically useful tool to assess CAP severity. According to thisscale, patients with a score of 0 should be managed as out-patients, those with a score of 1 or 2 as outpatients orinpatients, those with a score of 3 as inpatients, and thosewithascoreof4or5inanICU.Itisagoodpredictorofmortalitywith a predictive accuracy similar to CURB-65 and PSI. 45,46 SOAR is a new rule derived from BTS severity assessmentcriteria which comprises systolic blood pressure ( < 90 mmHg), oxygenation (PaO 2 :FiO 2  <  250), age (  65 years), andrespiratory rate (  30/min). 6 One point is given for thepresence of each item, and severe pneumonia is de 󿬁 ned bythe presence of a score   2. It is as sensitive and speci 󿬁 c ascurrent BTS recommended rules. The main potential advan-tage of SOAR in older people is the exclusion of urea andconfusion from the severity index because they are commonin this group of patients and can be confounded by multiplefactors in this population. Nevertheless, it does not seem toimprove identi 󿬁 cation of death from CAP within 6 weekscompared with the BTS rules.In 2007, a new, simple, clinical prediction tool, derivedfrom CRB-65, was proposed by Buising et al: the CORBscore. 44 Severe CAP, de 󿬁 ned by the combined outcome of death and/or requirementfor ventilatoror inotropicsupport,could be predicted by two or more of the following: acuteconfusion,oxygensaturation  90%,respiratoryrate  30/min,and either systolic blood pressure  < 90 mm Hg or diastolicblood pressure   60 mm Hg. Its advantages are that it issimple, uses predictive variables, does not require invasivetesting,andremovesbiasregardingpatientage.Itsaccuracyisnot perfect but is similar to that of CURB-65 and PSI.However, none of these scores strati 󿬁 es patients withhigh-severity CAP. Based in the PIRO concept (predisposition,insult,response,organdysfunction),anewpredictionruleforassessment of severity in ICU patients with CAP has beenproposed. 47 It is based on eight easily available variables, allwith known impact in CAP mortality: comorbidities (chronic Seminars in Respiratory and Critical Care Medicine Vol. 33 No. 3/2012 Assessing Severity of Patients with Community-Acquired Pneumonia  Pereira et al.  275    D  o  w  n   l  o  a   d  e   d   b  y  :   W  o  r   l   d   H  e  a   l   t   h   O  r  g  a  n   i  z  a   t   i  o  n   (   W   H   O   ) .   C  o  p  y  r   i  g   h   t  e   d  m  a   t  e  r   i  a   l .
Recommended
View more...
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks