ƅē�ĽǎƸƽǝÎÛƝŃǩǻșǻȓǎǚǝƝŃāǏĕċƢ�NjǃǏȒǴǫZqǏÞş

učŌ�1,2) �Ƨ�Ë3) Ý.Vk4) 7č��5) ƓƂū�6) æÔ�Ƭƫ

1) Ɗŏ�čĶNĒƖ ƚ�ðĔģ

2) � ĻejsCsƂŖċs,œŖċs

3) ��jsjsƖ ·�ƚ�ðĔCs

4) ��jsCsƂ��ĒƖ Þ×Ƃ

5) ľȉȒǣȖǿCģjsĒƖ Ő8Ƃ

6) � ĻejsCsƂŇ�ĞĦ±žǶȖǸș ċĄIJŞƂƎ

IncidenceAndriskfactorsofphlebitisandcoMplicatiOnsduetopeRipheralVENoUS catheterincriticallyillpatients

–AMOR-VENUSstudy-

ÏĞĦǏęė

• ICUNJ¯*ƿǞDžÎÛƝŃǩǻșǻȓǎǚǝƝŃāǏĕċƢ�NjǃǏȒǴǫȄǢǫǸșǏŨ×

•ƝŃǩǻșǻȓīĊǎƸƽǝu+�Ť!NjICU~�/~�Ő8�ǏÊČ�Ť!

ŁÇ

•ÎÛƝŃǩǻșǻȓǐ*Ɩ�ĽǎƸƵljÈǘƢĸǎ�ČƿǞljƵǝCĔǼȁǥǴǏ�LjNJƴǝ

ƸƸǚǃ70%Ĥ�Int JAntimicrob Agents2009;34(Supple):S38-42

JAMA1995;274:639-44

ICUǎƸƽǝÎÛƝŃǩǻșǻȓďĹǏ:NǐȞ

Ǔǔ100%??

ICUǎƸƽǝ��ƝŃǩǻșǻȓǏ¯*:Nȝ43-80%

InfectControlHosp Epidemiol 2004;25:266

ÎÛƝŃǩǻșǻȓǐu+ƹȞ

N�ē Ƣ�

m¸Ĉ 43-59%

ƝŃā 16.1-22.7%

þ0ȘŸ÷ǏÿǞ 14.2-23.9%

�¶¨G 6.9-17.5%

�Ö 0.44%

JInfus Nurs 2015;38:189

ƾǞǕNJǏÎÛƝŃǩǻșǻȓN�ēǎƒǁǝReview

ICU/ƅē�ĽǎƸƽǝÎÛƝŃǩǻșǻȓǏN�ēƢ�ǐȞ

IntensiveCareMed.2016Nov;42(11):1753-1765.

Variable Peripheral catheter(n=873)Extravasation andSuperficialPhlebitis 45(5.2%)

Accidental removal 55(6.3%)

Dysfunction 167(19.1%)

Catheters’ globalcomplicationsoring

None 620(71.0%)

Minor 238(27.3%)

Moderate 13(14.9%)

Severe 2(0.2%)

VerySevere 0(0%)

24.7/1000catheter-days

ÎÛƝŃǩǻșǻȓvs.��ƝŃǩǻșǻȓ

ICU�ĽǎƸƽǝ2ÌǏƝŃǩǻșǻȓďĹǏƀ©ǎƒǁǝRCT

Crit CareMed.2013Sep;41(9):2108-15

�Ľ1��DžǜǏƅjǍN�ēºǠíŶ

ÎÛƝŃǩǻșǻȓȝ1.04�

ƅjǍN�ēǏvĻȝCTCAE-v3.0Ǡ°Č

��ƝŃǩǻșǻȓȝ0.64�P=0.02

vs.

ÎÛƝŃǩǻșǻȓƹǛǏÃ]8Ǐ¦�ǐÅƹ�ƹ

A systematic review of extravasation and local tissue injury fromadministration of vasopressors through peripheral intravenous cathetersand central venous catheters☆,☆☆

Osama M. Loubani, MD, FRCPC a,⁎, Robert S. Green, MD, FRCPC a,b

a Departments of Critical Care Medicine and Emergency Medicine, Dalhousie University, Room 377, Bethune Building, 1276 South Park St, Halifax, Nova Scotia B3H 2Y9, Canadab Trauma Nova Scotia, 1276 South Park St, Centennial Building Room 1-026B, Halifax, Nova Scotia B3H 2Y9, Canada

a b s t r a c ta r t i c l e i n f o

Keywords:Vasopressor agentsIntravenous administrationPeripheral venous catheterizationCentral venous catheterizationAdverse effectsExtravasation of diagnostic and therapeutic materials

Purpose: The aim of this study was to collect and describe all published reports of local tissue injury or extrava-sation from vasopressor administration via either peripheral intravenous (IV) or central venous catheter.Methods: A systematic search of Medline, Embase, and Cochrane databases was performed from inceptionthrough January 2014 for reports of adults who received vasopressor intravenously via peripheral IV or centralvenous catheter for a therapeutic purpose. We included primary studies or case reports of vasopressor administra-tion that resulted in local tissue injury or extravasation of vasopressor solution.Results: Eighty-five articles with 270 patientsmet all inclusion criteria. A total of 325 separate local tissue injury andextravasation events were identified, with 318 events resulting from peripheral vasopressor administration and 7events resulting from central administration. There were 204 local tissue injury events from peripheral administra-tion of vasopressors,with an average duration of infusion of 55.9 hours (±68.1),median timeof 24hours, and rangeof 0.08 to 528 hours. In most of these events (174/204, 85.3%), the infusion site was located distal to the antecubitalor popliteal fossae.Conclusions: Published data on tissue injury or extravasation from vasopressor administration via peripheral IVs arederived mainly from case reports. Further study is warranted to clarify the safety of vasopressor administration viaperipheral IVs.

© 2015 Elsevier Inc. All rights reserved.

1. Introduction

1.1. Background

Vasopressor medications are commonly administered in theemergency department and intensive care unit to treat hemodynamicinstability in critically ill patients. Administration of vasopressors viacatheters located in large central veins has become the preferred routedue to concernsabout adverse events resulting fromperipheral intravenous(IV) use [1,2]—especially local tissue ischemia secondary to the vasocon-strictive properties possessed by this class of medications [2].

Despite these concerns, using a peripheral IV to administervasopressor may allow the medication to reach the patient sooner andreduce the time required to achieve hemodynamic stability and its con-comitant clinical benefits. Although peripheral IV access is readily

obtained in most patients, peripheral vasopressor administration isoften avoided to minimize the risk of potential local tissue ischemia.

1.2. Importance

In some critically ill patients, the requirement for a central venouscatheter (CVC) may delay administration of vasopressors whilethe catheter is placed (usually by a physician). This delay may haveunintended negative consequences because patients must remain in ahemodynamically unstable condition while the CVC is inserted. Inaddition, CVC insertion during emergency circumstances may increasethe risk of adverse events compared with CVC insertion for an electiveprocedure [3].

1.3. Goals of this investigation

The evidence cited for avoiding peripheral administration of vaso-pressors is a sparse collection of case studies and expert opinion [4].We sought to describe the literature for the current practice of avoidingperipheral administration of vasopressors due to concerns of local tissueischemia. We performed a systematic review to describe published re-ports on local tissue injury or extravasation during the administrationof vasopressor medications using a peripheral IV or a CVC, and the

Journal of Critical Care 30 (2015) 653.e9–653.e17

☆ Financial support: Dr Green is supported by a Clinician Scientist Award from theFaculty of Medicine at Dalhousie University in Halifax, Nova Scotia, Canada.☆☆ Conflicts of interest: The authors declare they have no competing interests.

⁎ Corresponding author at: Room 377, Bethune Building, 1276 South Park St, Halifax,Nova Scotia, Canada B3H 2Y9. Tel.: +1 902 425 3790.

E-mail address: Oloubani@dal.ca (O.M. Loubani).

http://dx.doi.org/10.1016/j.jcrc.2015.01.0140883-9441/© 2015 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Journal of Critical Care

j ourna l homepage: www. jcc journa l .o rg

the epidermis; (h) ulcer—lesion on surface of skin or a mucosal surfaceproduced by sloughing of inflammatory necrotic tissue; (i) amputationof limb or digit—removal of limb or other appendage or outgrowth ofthe body; (j) distal catheter location—IV catheters placed distal to theantecubital fossa of the upper extremity or the popliteal fossa of thelower extremity; (k) no long-term sequelae—patient returned to previ-ous level of function without deficits; (l) minor disability from pressorevent—patient returned to previous level of function, but with minordeficits; (m) major disability—patient unable to return to previouslevel of function because of severe deficits from event; (n) mortalitycontributed from pressor event—patient died, and pressor event wasfelt to be major contributor causing death; and (o) mortality unrelatedto pressor event—patient died, and pressor event was felt not to be im-portant contributor in causing death.

3. Results

3.1. Characteristics of study subjects

Our search identified a total of 86 371 references, of which 85 [5–89]met final inclusion criteria (Fig. 1). The κ statistic for agreement betweenreviewers for the final inclusion of papers based on full references was0.773. Of the 85 included references, 80 [5–21,23–35,37–64,66,68–88]presented individual data and were made up of case studies and case se-ries, while 5 [22,36,65,67,89] presented aggregate data and included 1[67] randomized controlled trial and 4 [22,36,65,89] case series. None ofthese studies directly compared the administration of vasopressors viacentral or peripheral catheters with rates of tissue ischemia or extravasa-tion. From the 85 included studies, a total of 270 patients with 325 sepa-rate events of local tissue injury or vasopressor extravasation arising fromIV administration of vasopressorswere identified. Overall, individual datawere presented on 195 patients, and aggregate data were presented on75 patients. Of the 85 studies included for review, 29 were written in 14languages other than English. Demographic and study characteristics ofincluded papers are presented in Table 1.

4. Main results

4.1. Studies on peripheral administration of vasopressors

Of the 325 separate events of local tissue injury or vasopressorextravasation associated with administration of vasopressors,318 events [6–17,19–27,29–66,68–75,77–89] resulted from periph-eral IV administration. Of these 318 events, there were 204[6–14,16,17,19–21,23–27,29–60,62–64,66,69–75,78,79,81–89] local

tissue injury events and 114 events [6,9,11,15,17,19,21,22,25,29,49,50,52,56,61,62,65,68,70,72–74,77,80,83–85,88,89] of extravasation of va-sopressor solution. Table 2 presents data on complications that resultedfrom the administration of vasopressor via peripheral IV. The occur-rence of extravasation was recorded independently of other complica-tions and may or may not have been related to other tissue injury.

Of the 204 local tissue injury events, there were 179 skin necrosisevents, 5 tissue necrosis events, and 20 gangrene events. Norepineph-rine (80.4%), dopamine (9.3%), and vasopressin (6.9%) were most com-monly administered in instances of local tissue complications. Thelocation of the peripheral IV through which vasopressor was infusedwas given in 194 of 204 events. In 174 (85.3%) events, the peripheralcatheter through which vasopressor was administered was located ina site distal to the antecubital or popliteal fossae. The average durationof vasopressor infusion before local tissue injury occurred was 55.9hours (±68.1), with a median of 24 hours and a range of 0.08 to 528hours. Fig. 2 provides a graphic representation of the local tissue injuryevents occurring as a function of the duration of peripheral vasopressorinfusion after which local tissue injury was noted. The occurrence oflong-term sequelae was reported in 182 of the 204 local tissue injuryevents. In many of the local tissue injury events, no long-term sequelae(77 [37.7%]) or minor disability (36 [17.6%]) was reported. However,major disability was reported in 9 (4.4%) of these events, and in 4(2.0%), the complications arising from use of peripheral vasopressorswere felt to be a major contributor to mortality.

Of the 114 events of extravasation of vasopressor solution, most(75.4%) did not result in any tissue injury. Norepinephrine (64.9%)and dopamine (22.8%) were most commonly administered in instancesof extravasation. The location of the peripheral IV through which vaso-pressor was infusedwas provided in 52 of the 114 extravasation events.In most of the reported events (39/52 [75.0%]), the peripheral catheterwas located distal to the antecubital or popliteal fossae. The averageduration of vasopressor infusion before extravasation occurred was35.2 hours (±51.0), with a median of 22 hours and a range of 0.25 to240 hours. The occurrence of long-term sequelae was reported in 112of the 114 extravasation events. Most of the patients who experiencedextravasation events had no long-term sequelae (90/112 [80.4%]).In 3 (2.7%) of 112 of these events, the patient suffered major disability,and in 1 (0.9%) of 112, the event was felt to be a major contributorto mortality.

4.2. Studies on central administration of vasopressors

Information regarding the occurrence of tissue injury or extravasa-tion resulting from vasopressor administration via CVC is presented in

Fig. 2. Duration of infusion of peripherally and centrally administered vasopressors, in hours, for events where local tissue injury occurred.

653.e13O.M. Loubani, R.S. Green / Journal of Critical Care 30 (2015) 653.e9–653.e17

JCrit Care2015;30:653

}Ū»ć85�Ǐƶdž2000��ƕǐ14%Ĥ�

74%13%

7%

1-3%

ƅē�ĽǎƸƵljÎÛƝŃǩǻșǻȓN�ēǠ�ƻŰƾǁȒǴǫȄǢǫǸșǐȞ

Age Typeofmedication infused (TPN,drug, bloodtransfusion, etc)

Severityofillness Replacement interval

Presence ofcertaincomorbidconditions

Infusion filtrationapparatus

Indwellingcatheter locationandduration

Infection

Increased durationofcatheterisation Intubation

Thenumberofcathetersused Intermittent infusion

Materialandlengthofthecatheter Weight

ȘRevEscEnferm USP2016;50:2:261-269ȘRevEscEnferm USP2016;50:22

ȘCrit CareMed2009;37:702-712ȘNursManage2005;36:52-54ȘInfus Nurs 2006;29:S1-S92ȘNurs Crit Care2008;3:10-15

ƝŃāǏȒǴǫȄǢǫǸș

ÎÛƝŃǩǻșǻȓǎƒǁǝđs�`ǐ�ǍƵ

Țąǎƅē�ĽǎƸƵljț

ICUNJǐȒǴǫȄǢǫǸșƺĐǍǝKł�ƺƴǝ

vasoactive agents,sedatives, antibiotics,parenteral nutrition,etc

ĉĆǎƸƵlj

ÎÛƝŃǩǻșǻȓǏu+īĊǐȞ

ĭ\NJǏÎÛƝŃīĊǏĉĆǐȞ

Ÿ÷ěũ�E� (INS:InfusionNurseSociety)

Ÿ÷~Ǝěũ� (CRNI:CertifiedRegisteredNurseInfusion)

Ÿ÷ĔòǏ¾ƉǙ¤ơǏĩv

CĔâXǏŤ!ƱijĴ¹ĿǠ«�

ąvƂƎǘǀƼǐ¯*[ƛ Ǐ

ǩǻșƷǝ¯*Ǡxų

Documento de Consenso sobre la prevención, diagnóstico y tratamiento de las infecciones por catéter venoso periférico en adultos

RESUMEN

El uso de catéteres vasculares es una práctica muy utiliza-da en los hospitales. El uso de catéteres venosos periféricos de corta duración se ha asociado con un elevado riesgo de bacte-riemia nosocomial, lo que comporta una no despreciable mor-bilidad y mortalidad. La etiología de estas infecciones suele ser frecuentemente por Staphylococcus aureus, lo que explica su gravedad. En este documento de consenso, elaborado por un panel de expertos de La Sociedad Española de Infecciones Car-diovasculares con la colaboración de expertos de la Sociedad Española de Medicina Interna, La Sociedad Española de Qui-mioterapia y la Sociedad Española de Cirugía Torácica y Car-diovascular, pretende establecer unes normes para un mejor uso de los catéteres venosos periféricos de corta duración. El Documento revisa las indicaciones para su inserción, manten-imiento, registro, diagnóstico y tratamiento de las infecciones derivadas y las indicaciones para su retirada; haciendo énfasis en la formación continuada del personal sanitario para lograr una mayor calidad asistencial. Seguir las recomendaciones del consenso permitirá utilizar de una manera más homogénea los catéteres venosos periféricos minimizando el riesgo de infec-ción y sus complicaciones.

Palabras Clave: Infección de catéter; catéter venoso periférico; bacter-iemia, prevención de la infección de catéter; diagnóstico de la infección de catéter; tratamiento de la infección de catéter, infección nosocomial

ABSTRACT

The use of endovascular catheters is a routine practice in secondary and tertiary care level hospitals. Short peripheral catheters have been found to be associated with the risk of nosocomial bacteremia resulting in morbidity and mortality. Staphyloccus aureus is mostly associated with peripheral cath-eter insertion. This Consensus Document has been elaborated by a panel of experts of the Spanish Society of Cardiovascular Infections in cooperation with experts from the Spanish So-ciety of Internal Medicine, Spanish Society of Chemotherapy and Spanish Society of Thoracic-Cardiovascular Surgery and aims at define and establish the norm for management of short duration peripheral vascular catheters. The document addresses the indications for insertion, catheter maintenance and registry, diagnosis and treatment of infection, indications for removal and stresses on continuous education as a driver for quality. Implementation of this norm will allow uniformity in usage thus minimizing the risk of infection and its compli-cations.

Key words: catheter infection, peripheral venous catheter, bacteremia, prevention of catheter infection, diagnosis of catheter infection, treatment of catheter infection, nosocomial infection

2016 Expert consensus document on prevention, diagnosis and treatment of short-term peripheral venous catheter-related infections in adult

1Servicio de Medicina Interna. Consorcio Sanitario de Mataró, Mataró, Barcelona.2Servicio de Microbiología Clínica y Enfermedades Infecciosas, Hospital General Universitario Gregorio Marañón, Madrid3Servicio de Medicina Interna. Hospital Universitario MontePrincipe, Madrid4Servicio de Enfermedades Infecciosas. Hospital Universitario Virgen del Rocío, Sevilla.5Servicio de Enfermedades Infecciosas Hospital Universitario Marqués de Valdecilla, Universidad de Cantabria. Santander.6Servicio de Enfermedades Infecciosas y Microbiología Clínica. Hospital Universitario Virgen Macarena. Sevilla7Servicio de Enfermedades Infecciosas. Hospital Universitario Donosti. San Sebastián.8Servicio de Cirugía Cardiovascular. Hospital Universitario Marqués de Valdecilla, Santander.9Servicio de Enfermedades Infecciosas, Complejo Hospitalario Universitario A Coruña, A Coruña.10Servicio de Enfermedades Infecciosas, Hospital Clinic-IDIBAPS. Universidad de Barcelona, Barcelona.11Unidad de Enfermedades Infecciosas. Hospital Universitario de Cruces, Bilbao. Universidad del País Vasco12Servicio de Cirugía Cardiaca. Hospital Universitario de A Coruña. A Coruña13Cardiothoracic and Vascular Surgery. Heart & Vascular Institute, Cleveland Clinic Abu Dhabi, Abu Dhabi, United Arab Emirates

Representing the a) Spanish Society of Cardiovascular Infections (SEICAV), b) Spanish Society of Internal Medicine (SEMI), c) Spanish Society of Chemotherapy (SEQ) and d) Spanish Society of Thoracic and Cardivacular Surgery (SECTCV)

Josep A. Capdevila1,a,b

María Guembe2

José Barberán3,b,c

Arístides de Alarcón4,a

Emilio Bouza2,a

M. Carmen Fariñas5,a

Juan Gálvez6,a

Miguel Ángel Goenaga8,a

Francisco Gutiérrez8,a

Martha Kestler2,a

Pedro Llinares9,a

José M. Miró10,a

Miguel Montejo11,a

Patricia Muñoz2,a

Marta Rodríguez-Creixems2,a

Dolores Sousa9,a

José Cuenca12,d

Carlos-A. Mestres10,13,a,d

on behalf the SEICAV, SEMI, SEQ and SECTCV Societies

Correspondence:Josep A. Capdevila Internal MedicineConsorcio Hospitalario de MataróMataró (Barcelona)E-mail: jcapdevila@csdm.cat

*Este Documento también se ha publicado en: Cirugía Cardiovascular:http://dx.doi.org/10.1016/j.circv.2016.06.001

Consensus document

230Rev Esp Quimioter 2016;29(4): 230-238

Documento de Consenso sobre la prevención, diagnóstico y tratamiento de las infecciones por catéter venoso periférico en adultos

RESUMEN

El uso de catéteres vasculares es una práctica muy utiliza-da en los hospitales. El uso de catéteres venosos periféricos de corta duración se ha asociado con un elevado riesgo de bacte-riemia nosocomial, lo que comporta una no despreciable mor-bilidad y mortalidad. La etiología de estas infecciones suele ser frecuentemente por Staphylococcus aureus, lo que explica su gravedad. En este documento de consenso, elaborado por un panel de expertos de La Sociedad Española de Infecciones Car-diovasculares con la colaboración de expertos de la Sociedad Española de Medicina Interna, La Sociedad Española de Qui-mioterapia y la Sociedad Española de Cirugía Torácica y Car-diovascular, pretende establecer unes normes para un mejor uso de los catéteres venosos periféricos de corta duración. El Documento revisa las indicaciones para su inserción, manten-imiento, registro, diagnóstico y tratamiento de las infecciones derivadas y las indicaciones para su retirada; haciendo énfasis en la formación continuada del personal sanitario para lograr una mayor calidad asistencial. Seguir las recomendaciones del consenso permitirá utilizar de una manera más homogénea los catéteres venosos periféricos minimizando el riesgo de infec-ción y sus complicaciones.

Palabras Clave: Infección de catéter; catéter venoso periférico; bacter-iemia, prevención de la infección de catéter; diagnóstico de la infección de catéter; tratamiento de la infección de catéter, infección nosocomial

ABSTRACT

The use of endovascular catheters is a routine practice in secondary and tertiary care level hospitals. Short peripheral catheters have been found to be associated with the risk of nosocomial bacteremia resulting in morbidity and mortality. Staphyloccus aureus is mostly associated with peripheral cath-eter insertion. This Consensus Document has been elaborated by a panel of experts of the Spanish Society of Cardiovascular Infections in cooperation with experts from the Spanish So-ciety of Internal Medicine, Spanish Society of Chemotherapy and Spanish Society of Thoracic-Cardiovascular Surgery and aims at define and establish the norm for management of short duration peripheral vascular catheters. The document addresses the indications for insertion, catheter maintenance and registry, diagnosis and treatment of infection, indications for removal and stresses on continuous education as a driver for quality. Implementation of this norm will allow uniformity in usage thus minimizing the risk of infection and its compli-cations.

Key words: catheter infection, peripheral venous catheter, bacteremia, prevention of catheter infection, diagnosis of catheter infection, treatment of catheter infection, nosocomial infection

2016 Expert consensus document on prevention, diagnosis and treatment of short-term peripheral venous catheter-related infections in adult

1Servicio de Medicina Interna. Consorcio Sanitario de Mataró, Mataró, Barcelona.2Servicio de Microbiología Clínica y Enfermedades Infecciosas, Hospital General Universitario Gregorio Marañón, Madrid3Servicio de Medicina Interna. Hospital Universitario MontePrincipe, Madrid4Servicio de Enfermedades Infecciosas. Hospital Universitario Virgen del Rocío, Sevilla.5Servicio de Enfermedades Infecciosas Hospital Universitario Marqués de Valdecilla, Universidad de Cantabria. Santander.6Servicio de Enfermedades Infecciosas y Microbiología Clínica. Hospital Universitario Virgen Macarena. Sevilla7Servicio de Enfermedades Infecciosas. Hospital Universitario Donosti. San Sebastián.8Servicio de Cirugía Cardiovascular. Hospital Universitario Marqués de Valdecilla, Santander.9Servicio de Enfermedades Infecciosas, Complejo Hospitalario Universitario A Coruña, A Coruña.10Servicio de Enfermedades Infecciosas, Hospital Clinic-IDIBAPS. Universidad de Barcelona, Barcelona.11Unidad de Enfermedades Infecciosas. Hospital Universitario de Cruces, Bilbao. Universidad del País Vasco12Servicio de Cirugía Cardiaca. Hospital Universitario de A Coruña. A Coruña13Cardiothoracic and Vascular Surgery. Heart & Vascular Institute, Cleveland Clinic Abu Dhabi, Abu Dhabi, United Arab Emirates

Representing the a) Spanish Society of Cardiovascular Infections (SEICAV), b) Spanish Society of Internal Medicine (SEMI), c) Spanish Society of Chemotherapy (SEQ) and d) Spanish Society of Thoracic and Cardivacular Surgery (SECTCV)

Josep A. Capdevila1,a,b

María Guembe2

José Barberán3,b,c

Arístides de Alarcón4,a

Emilio Bouza2,a

M. Carmen Fariñas5,a

Juan Gálvez6,a

Miguel Ángel Goenaga8,a

Francisco Gutiérrez8,a

Martha Kestler2,a

Pedro Llinares9,a

José M. Miró10,a

Miguel Montejo11,a

Patricia Muñoz2,a

Marta Rodríguez-Creixems2,a

Dolores Sousa9,a

José Cuenca12,d

Carlos-A. Mestres10,13,a,d

on behalf the SEICAV, SEMI, SEQ and SECTCV Societies

Correspondence:Josep A. Capdevila Internal MedicineConsorcio Hospitalario de MataróMataró (Barcelona)E-mail: jcapdevila@csdm.cat

*Este Documento también se ha publicado en: Cirugía Cardiovascular:http://dx.doi.org/10.1016/j.circv.2016.06.001

Consensus document

230Rev Esp Quimioter 2016;29(4): 230-238

hanced asepsis is not required if the endovenous segment of the PVC is not manipulated9 (III-B). As it is the case when inserting central venous lines, the use of additional protection measures like facemask is not recommended. However, this is a topic for consideration and analysis if in a given institution higher than expected rates of PVC-related bacteremia are observed.

Sterile gauze dressing or semi permeable transparent sterile dressing to cover the insertion site will be used23,24 (II-A). Ster-ile gauze dressing will be inspected and replaced every other day and transparent dressing should not stay in place over 7 days9. If there is humidity, sweating or blood it is more appropriate to use non-occlusive gauze dressing24,25 (III-B). Revision or replacement of dressing must be performed with single-use clean gloves9.

PVCs placed on urgent basis or without considering min-imal hygiene rules must be removed and replaced before 48 hours to avoid the risk of infection17,26,27 (II-A).

The use of techniques facilitating identification of veins as laser or ultrasound28,29 in patients with poor venous flow are also recommended for insertion. However, these techniques do not reduce the risk of infection. A meta-analysis on this topic showed that its routine use is not justified30 (I-A).

4. ChecklistThe adhesion to recommendations in the form of check-

list is associated to better results in prevention of post-inser-tion complications after insertion of central venous lines and PVCs10,31 (I-A). This is reflected in table 2.

2016 Expert consensus document on prevention, diagnosis and treatment of short-term peripheral venous catheter-related infections in adults

J. A. Capdevila, et al.

Rev Esp Quimioter 2016;29(4): 230-238 232

DEFINITIONS

Table 1 describes the levels of evidence and the strength of recommendations according to the criteria of the Infectious Disease Society of America (ISDA)15.

PVC is a catheter shorter than 7.62 cm (3 inches).Sepsis is a systemic inflammatory response syndrome sec-

ondary to an infection16. The term phlebitis is used if one of the following criteria was fulfilled: swelling and erythema > 4 mm, tenderness, palpable venous cord, pain or fever with local symptoms. Isolated swelling is not defined as phlebitis.

INSERTION

1. When?PVC will be inserted when the duration of a given endo-

venous therapy is expected to be shorter than 6 days and the PVC will not be used for major procedures as hemodialysis, plas-mapheresis, chemotherapy, parenteral nutrition, monitoring or administration of fluid large volumes. When any of these cir-cumstances is to be expected, it is preferable to insert a single-, double- or triple-lumen central venous line (peripherally insert-ed or not) as the risk of chemical phlebitis, the need for high-speed volume infusion or frequent manipulations do not sup-port a short catheter (I-A)17,18. An isolated transfusion does not need a central venous line insertion. Before placing any venous line, even peripheral, it is mandatory the evaluation of the actual need. Venous lines are often placed as routine; this meant to be an act reflecting the provision of care. It is also frequently shown that to treat the patient a “prophylactic” line was not mandato-ry. A study showed that up to 35% of peripheral venous lines place in the emergency department are unnecessary19.

2. Where?A PVC can be inserted in every accessible vein. However,

upper extremity veins are preferable for patient comfort and lesser risk of contamination. Some studies reported a higher risk of phlebitis after lines were placed at the cubital crease, thus becoming preferable avoiding this site in benefit of arm, forearm or dorsal aspect of the hand/wrist20,21 (II-A).

Furthermore, other patient-related factors like accessibil-ity to the venous system or comfort after insertion have to be taken into account. It does not make much sense to insert a PVC onto a central vein (III-A).

3. How?The insertion of PVC must be performed under maximal

aseptic techniques. It is not necessary to prep a surgical field as it is the when inserting a central venous line. The skin must be disinfected with 2% alcoholic chlorhexidine solution or, if not available, with a 70% iodine or alcohol solution9,22,23 (I-A).

The insertion site should not be touched after disinfection. The catheter must be handled from its proximal end when in-serted. The caregiver inserting the PVC must previously perform hand hygiene with water and soap and/or wash hands with alcohol solution. Single-use clean gloves must be used. An en-

Table 2 Checklist for an appropriate manipulation of peripheral catheters. If these are not fulfilled, the prompt removal of the catheter is advised (Evidence A)

-Insertion

-Correct hand hygiene

-Field disinfection

-Use single-use clean gloves

-Do not touch the insertion site

-Do not touch the endovenous segment of the catheter

-Sterile dressing (gauze or transparent)

-Manipulation

-Daily assessment of the need for the PVC

-Daily inspection of the insertion site

-Daily assessment of the function of the catheter

-Adequate replacement of infusion sets

-Catheter and events registry

-Fluid extravasation

-Presence of blood

-Inflammatory signs

-Dressing status

der aseptic conditions. A cotton swab should be used to take samples from purulent exudate if present. As PVCs should be of short duration and of easy replacement it is not justified to keep a catheter in situ while awaiting results from Microbiol-ogy if infection is suspected (III-B). We then believe that con-servative diagnostic techniques for diagnosis of infection are not applicable60,61 (III-A). Gram stain of a PVC segment may quickly draw the attention on the possibility of infection62.

TREATMENT

In the treatment of PVC infection, the first step is re-moval of the PVC as it has been mentioned above. Once the PVC is removed and blood samples taken for culture, the need for empirical antibiotic treatment will be related to the clinical condition of the patient (including fever and el-evation of biomarkers). Treatment should be directed to PVC bacteremia. Isolated positive tips cultures don´t need antibi-otic treatment.

If empirical antibiotic treatment is initiated, Gram-positive cocci (including methicillin-resistant S. aureus) and Gram-neg-ative bacilli (including P. aeruginosa) must be addressed ac-

cording to individual patient risk factors and the institutional flora. Other possible etiologies, albeit infrequent, have to be considered in special subsets of patients as those previously treated with antibiotics, with multiple comorbidities, immune depressed or hospitalized for long periods of time63. S. aureus has become an increasingly impactful etiologic pathogen for bacteremia as it has been shown in several studies3,4,64-66. For bacteremia related to central venous catheters, the etiology is well diversified.

A reasonable empirical regimen is a combination of dap-tomycin and a ß-lactam active against P. aeruginosa. In pa-tients with ß-lactam allergies, aztreonam, an aminoglycoside or a quinolone could be an alternative. In any case, treatment should follow sensitivity patterns at 24-72 hours after cultures are taken67,68 (I-A).

The duration of antibiotic treatment will be related to the isolated pathogen. S. epidermidis can be treated with removal of PVC if no other inert material that can be colonized and/or infected exists; duration of treatment should not be longer than 7 days. If no antibiotic treatment is given, the patient must be symptom-free and cultures must be negative upon removal of PVC.

2016 Expert consensus document on prevention, diagnosis and treatment of short-term peripheral venous catheter-related infections in adults

J. A. Capdevila, et al.

Rev Esp Quimioter 2016;29(4): 230-238 234

Table 3 Summary of recommendations and degree of evidence (see table 1).

Always assess the need of inserting a catheter. If necessary, a central venous line should be preferred over a PVC if duration of intravenous treatment longer than 6 days or blood transfusion, parenteral nutrition or chemotherapy I-A17,18

If possible, PVC should not be placed in the lower extremities or at the elbow crease due to higher risk of phlebitis II-A20,21

Insertion of PVC must be performed with the maximum hygiene with no need for a surgical field. There are no preferences as to which disinfectant solution to use I-A9,22,23

An sterile dressing must be used to cover the insertion site (gauze dressing or transparent semi permeate) II-A24,25

Adherence to pre-insertion checklist improves prevention of complication outcomes I-A10,31

The need for PVC should be assessed on daily basis. If it is not necessary, it is advisable to remove the PVC II-A18,32,33

The insertion site must be inspected daily. If abnormalities, malfunction or discomfort at the subcutaneous site, PVC should be removed III-A17,34-35

No antiseptic cream/gel should be used at the insertion site III-C36

Closed connectors to access the PVC can be used; its external surface must always be decontaminated II-A37

Infusion sets can be utilized up to 96 hours, exception made of blood transfusion or lipid emulsions I-A38,39

It is mandatory that the nursing files a daily record of the PVC III-A40

It is not advisable to remove PVC on a routine basis. PVC should be replaced when clinically indicated I-A18,41-51

It is advisable not to keep a PVC in place for over 5 days III-B

Unused catheters must be removed II-A10,53,55

When there is suspicion of PVC inserted under suboptimal conditions, it must be removed III-A56,57

If there is suspicion of infection, it is not indicated to use diagnostic technique leaving the PVC in place III-A

If there is a suspicion of catheter-related infection, the tip of the PVC must be submitted for Microbiology. Removed PVC non suspected to be infected not need Microbiology III-A

Empiric antibiotic treatment of PVC-related bacteremia has to be deescalated according to microbiology results I-A67,68

Continuous education in insertion and maintenance guidelines is an appropriate way to reduce complications I-A74-87

ICUǎƸƽǝŐ8ǏǽȑȅȓǐȞ

© 2014 Jiang et al. This work is published by Dove Medical Press Limited, and licensed under Creative Commons Attribution – Non Commercial (unported, v3.0) License. The full terms of the License are available at http://creativecommons.org/licenses/by-nc/3.0/. Non-commercial uses of the work are permitted without any further

permission from Dove Medical Press Limited, provided the work is properly attributed. Permissions beyond the scope of the License are administered by Dove Medical Press Limited. Information on how to request permission may be found at: http://www.dovepress.com/permissions.php

Therapeutics and Clinical Risk Management 2014:10 861–866

Therapeutics and Clinical Risk Management Dovepress

submit your manuscript | www.dovepress.com

Dovepress 861

S H O RT R E P O RT

open access to scientific and medical research

Open Access Full Text Article

http://dx.doi.org/10.2147/TCRM.S69585

Implementation of pharmacists’ interventions and assessment of medication errors in an intensive care unit of a Chinese tertiary hospital

Sai-Ping Jiang1,*Jian Chen2,*Xing-Guo Zhang1

Xiao-Yang Lu1

Qing-Wei Zhao1

1Department of Pharmacy, 2Intensive Care Unit, the First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, People’s Republic of China

*These authors contributed equally to this work

Background: Pharmacist interventions and medication errors potentially differ between the People’s Republic of China and other countries. This study aimed to report interventions administered by clinical pharmacists and analyze medication errors in an intensive care unit (ICU) in a tertiary hospital in People’s Republic of China.Method: A prospective, noncomparative, 6-month observational study was conducted in a gen-eral ICU of a tertiary hospital in the People’s Republic of China. Clinical pharmacists performed interventions to prevent or resolve medication errors during daily rounds and documented all of these interventions and medication errors. Such interventions and medication errors were categorized and then analyzed.Results: During the 6-month observation period, a total of 489 pharmacist interventions were reported. Approximately 407 (83.2%) pharmacist interventions were accepted by ICU physicians. The incidence rate of medication errors was 124.7 per 1,000 patient-days. Improper drug frequency or dosing (n 152, 37.3%), drug omission (n 83, 20.4%), and potential or actual occurrence of adverse drug reaction (n 54, 13.3%) were the three most commonly committed medication errors. Approximately 339 (83.4%) medication errors did not pose any risks to the patients. Antimicrobi-als (n 171, 35.0%) were the most frequent type of medication associated with errors.Conclusion: Medication errors during prescription frequently occurred in an ICU of a tertiary hospital in the People’s Republic of China. Pharmacist interventions were also efficient in preventing medication errors.Keywords: pharmacist, medication error, preva lence rate, type, severity, intensive care unit

IntroductionMedication errors (MEs) are among the most common preventable causes of adverse drug events.1,2 In 21 hospitals in the Netherlands, a retrospective study showed that 15% of adverse events were related to medications, and 21.2% of these events were considered preventable.3 In the Middle East, ME rates varied from 7.1% to 90.5% of prescription errors and from 9.4% to 80% of administration errors.4

Patients admitted in intensive care units (ICUs) are at a high risk of MEs because of critical illness, complex ICU environment, multiple medications, and frequent changes in medication therapy.5–8 Camiré et al9 reviewed data from 205 ICUs in 29 countries and found that the prevalence of MEs in ICUs was approximately 10.5 per 100 patient-days. Such a frequency of MEs was similar between prescription (54%) and administration (46%) phases. However, ME definitions and detection methods vary greatly in different reports.5

MEs may cause inefficient disease management for patients, additional workload for pharmacy personnel, increased cost for pharmacies and patients, frustration among

Correspondence: Qing-Wei ZhaoDepartment of Pharmacy, the First Affiliated Hospital, College of Medicine,

Zhejiang University, 79 Qingchun Road, Hangzhou, 310003, People’s Republic of ChinaTel 86 571 8723 3411Fax 86 571 8723 3411Email qingweizhao66@yeah.net

Journal name: Therapeutics and Clinical Risk ManagementArticle Designation: SHORT REPORTYear: 2014Volume: 10Running head verso: Jiang et alRunning head recto: Interventions and errors in an intensive care unitDOI: http://dx.doi.org/10.2147/TCRM.S69585

© 2014 Jiang et al. This work is published by Dove Medical Press Limited, and licensed under Creative Commons Attribution – Non Commercial (unported, v3.0) License. The full terms of the License are available at http://creativecommons.org/licenses/by-nc/3.0/. Non-commercial uses of the work are permitted without any further

permission from Dove Medical Press Limited, provided the work is properly attributed. Permissions beyond the scope of the License are administered by Dove Medical Press Limited. Information on how to request permission may be found at: http://www.dovepress.com/permissions.php

Therapeutics and Clinical Risk Management 2014:10 861–866

Therapeutics and Clinical Risk Management Dovepress

submit your manuscript | www.dovepress.com

Dovepress 861

S H O RT R E P O RT

open access to scientific and medical research

Open Access Full Text Article

http://dx.doi.org/10.2147/TCRM.S69585

Implementation of pharmacists’ interventions and assessment of medication errors in an intensive care unit of a Chinese tertiary hospital

Sai-Ping Jiang1,*Jian Chen2,*Xing-Guo Zhang1

Xiao-Yang Lu1

Qing-Wei Zhao1

1Department of Pharmacy, 2Intensive Care Unit, the First Affiliated Hospital, College of Medicine, Zhejiang University, Hangzhou, People’s Republic of China

*These authors contributed equally to this work

Background: Pharmacist interventions and medication errors potentially differ between the People’s Republic of China and other countries. This study aimed to report interventions administered by clinical pharmacists and analyze medication errors in an intensive care unit (ICU) in a tertiary hospital in People’s Republic of China.Method: A prospective, noncomparative, 6-month observational study was conducted in a gen-eral ICU of a tertiary hospital in the People’s Republic of China. Clinical pharmacists performed interventions to prevent or resolve medication errors during daily rounds and documented all of these interventions and medication errors. Such interventions and medication errors were categorized and then analyzed.Results: During the 6-month observation period, a total of 489 pharmacist interventions were reported. Approximately 407 (83.2%) pharmacist interventions were accepted by ICU physicians. The incidence rate of medication errors was 124.7 per 1,000 patient-days. Improper drug frequency or dosing (n 152, 37.3%), drug omission (n 83, 20.4%), and potential or actual occurrence of adverse drug reaction (n 54, 13.3%) were the three most commonly committed medication errors. Approximately 339 (83.4%) medication errors did not pose any risks to the patients. Antimicrobi-als (n 171, 35.0%) were the most frequent type of medication associated with errors.Conclusion: Medication errors during prescription frequently occurred in an ICU of a tertiary hospital in the People’s Republic of China. Pharmacist interventions were also efficient in preventing medication errors.Keywords: pharmacist, medication error, preva lence rate, type, severity, intensive care unit

IntroductionMedication errors (MEs) are among the most common preventable causes of adverse drug events.1,2 In 21 hospitals in the Netherlands, a retrospective study showed that 15% of adverse events were related to medications, and 21.2% of these events were considered preventable.3 In the Middle East, ME rates varied from 7.1% to 90.5% of prescription errors and from 9.4% to 80% of administration errors.4

Patients admitted in intensive care units (ICUs) are at a high risk of MEs because of critical illness, complex ICU environment, multiple medications, and frequent changes in medication therapy.5–8 Camiré et al9 reviewed data from 205 ICUs in 29 countries and found that the prevalence of MEs in ICUs was approximately 10.5 per 100 patient-days. Such a frequency of MEs was similar between prescription (54%) and administration (46%) phases. However, ME definitions and detection methods vary greatly in different reports.5

MEs may cause inefficient disease management for patients, additional workload for pharmacy personnel, increased cost for pharmacies and patients, frustration among

Correspondence: Qing-Wei ZhaoDepartment of Pharmacy, the First Affiliated Hospital, College of Medicine,

Zhejiang University, 79 Qingchun Road, Hangzhou, 310003, People’s Republic of ChinaTel 86 571 8723 3411Fax 86 571 8723 3411Email qingweizhao66@yeah.net

Journal name: Therapeutics and Clinical Risk ManagementArticle Designation: SHORT REPORTYear: 2014Volume: 10Running head verso: Jiang et alRunning head recto: Interventions and errors in an intensive care unitDOI: http://dx.doi.org/10.2147/TCRM.S69585

Therapeutics and Clinical Risk Management 2014:10submit your manuscript | www.dovepress.com

Dovepress

Dovepress

864

Jiang et al

of MEs (n 171, 35.0%) followed by cardiovascular drugs (n 76, 15.5%) and enteral/parenteral nutrition drugs (n 62, 12.7%). Considering specific drugs, we noted piperacil-lin/sulbactam (n 48, 9.8%), enteral nutrition drugs (n 76, 15.5%), and meropenem or imipenem (n 36, 7.4%) as drugs most commonly associated with MEs.

DiscussionCultures, laws, politics, clinical settings, and health care systems differ between the People’s Republic of China and other countries, such as the US or the UK. Moreover, the participation of pharmacists in ICU teams in the People’s Republic of China has been recently established. However, reports on the analysis of pharmacist interventions and the measurement of MEs in ICUs in the People’s Republic of China remain limited. To our knowledge, this study is the first to investigate pharmacist interventions and assess MEs in an ICU of a hospital in the People’s Republic of China. This

study provided insights into the value of pharmacy services in the People’s Republic of China’s health care system.

A high number of recommendations accepted by the physi-cians indicated that pharmacists in ICUs were well accepted by physicians. However, acceptance rate was greatly dependent on the type of recommendations. Extremely important recommen-dations (drug contraindication) or urgent need for physicians (consulting regarding drug information) could reach a high acceptance rate (100%); by contrast, recommendations requir-ing specific clinical skills from pharmacists (drug selection) presented a low acceptance rate (51.2%). Despite remarkable improvements in pharmacy services in hospitals in the People’s Republic of China, clinical skills of clinical pharmacists are essential for pharmaceutical care in a patient-centered pharmacy setting. Therefore, this area should be further improved.

In this study, the incidence rate of MEs was approxi-mately 124.7 per 1,000 monitored patient-days. This finding is close to that described by Camiré et al in which 29 countries

Table 2 Type and number of pharmacist’s recommendations in an intensive care unit of a Chinese tertiary hospital

Category of pharmacist recommendations n (%) recommendations n (%) recommendations accepted

Change in drug frequency or dosing: improper drug frequency or dosing 178 (36.4) 149 (83.7)Start new drugs: lack of use of needed drugs 81 (16.6) 72 (88.9)Change or stop drugs: potential ADR or occurrence of ADR 69 (14.1) 54 (78.3)Provision of drug information: lack of drug-related information 49 (10.0) 49 (100)Change or stop drugs: improper drug selection or indication 41 (8.4) 21 (51.2)Change or stop drugs: improper duration of drug treatment 27 (5.5) 22 (81.5)Change or stop drugs: drug–drug interactions 18 (3.7) 16 (88.9)Stop drugs: contraindication 11 (2.3) 11 (100)Others: wrong solvent, duplication, etc. 15 (3.1) 13 (86.7)Total 489 (100) 407 (83.2)

Abbreviation: ADR, adverse drug reaction.

Type

of m

edic

atio

n er

rors

Percentage of medication errors (%)

Others

Contraindication

Drug–drug interactions

Improper duration of drug treatment

Improper drug selection orindication

Potential or actual occurrence of ADR

Drugs ommision

Improper drug frequency or dosing

0.0% 10.0% 20.0% 30.0% 40.0%

3.2%

2.7%

4.7%

7.9%

10.6%

13.3%

20.4%

37.3%

Figure 1 Type of medication errors encountered in an intensive care unit of a Chinese tertiary hospital.Note: “Others” refer to medication errors, including wrong solvent, duplicated drug use, and so on.Abbreviation: ADR, adverse drug reaction.

CriticalCarepharmacistǐÊČƹȞ

Effect of critical care pharmacist's intervention on medication errors: Asystematic review and meta-analysis of observational studies

Tiansheng Wang, PharmD a, Neal Benedict, PharmD b, Keith M. Olsen, PharmD c, Rong Luan, MS a,d,Xi Zhu, MD d, Ningning Zhou, MS a, Huilin Tang, MS d, Yingying Yan, PhD candidate d,Yao Peng, MS a, Luwen Shi, MHPE a,⁎a Department of Pharmacy Administration and Clinical Pharmacy, School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing, Chinab Department of Pharmacy and Therapeutics, University of Pittsburgh School of Pharmacy, University of Pittsburgh Medical Center, Pittsburgh, PA, USAc Department of Pharmacy Practice, College of Pharmacy, University of Nebraska Medical Center, Omaha, NE, USAd Department of Pharmacy, Peking University Third Hospital, Beijing, China

a b s t r a c ta r t i c l e i n f o

Keywords:Pharmaceutical carePharmacist interventionMeta-analysisIntensive care unitCritical careMedication error

Pharmacists are integral members of the multidisciplinary team for critically ill patients. Multiple nonrandomizedcontrolled studies have evaluated the outcomes of pharmacist interventions in the intensive care unit (ICU). Thissystematic review focuses on controlled clinical trials evaluating the effect of pharmacist intervention on medica-tion errors (MEs) in ICU settings. Two independent reviewers searchedMedline, Embase, and Cochrane databases.The inclusion criteriawere nonrandomized controlled studies that evaluated the effect of pharmacist services vs nointervention on ME rates in ICU settings. Four studies were included in the meta-analysis. Results suggestthat pharmacist intervention has no significant contribution to reducing general MEs, although pharmacistintervention may significantly reduce preventable adverse drug events and prescribing errors. This meta-analysis highlights the need for high-quality studies to examine the effect of the critical care pharmacist.

© 2015 Elsevier Inc. All rights reserved.

1. Background

Over the past 3 decades, the role of the critical care pharmacist hasexpanded in the intensive care unit (ICU) from traditional dispensingresponsibilities to active members of multidisciplinary teams [1].Pharmacists play an important daily role in ICU patient care by assistingphysicians and other health care providers with pharmacotherapy deci-sionmaking andmonitoring, ultimately improving medication safety [2].

Medication errors (MEs) and adverse drug events (ADEs) occur athigh rates in the ICU [3] due to the (1) severity of disease of this popula-tion, (2) multiple medications administered, (3) high-risk nature of themedications, and (4) high frequency of changes in pharmacotherapyregimens [1,3-6]. Pharmacists are considered to be effective in reducingMEs andADEs by reducing the causes of these events throughpharmaco-therapy decision making and monitoring support [2,4]. Numerousstudies have assessed the effect of ICU pharmacists' intervention inreducing MEs [6-15]. However, these data are limited by the ability tocollect data without impeding care; thus, few of these studies arerandomized controlled trials [1] or are cohort studies that have an ICU

without a pharmacist as a control site [10,15]. Conversely, most ofthese studies are preintervention/postintervention designs [6-14],which are more realistic to conduct in the ICU setting.

It is, therefore, challenging to quantitatively evaluate the effectof ICU pharmacist interventions on MEs. Pooling data from availablestudies to perform a meta-analysis will help to better determine thetrue effect and magnitude of the interventions. Based on our literaturereview, no such meta-analysis exists. Therefore, we aimed to conducta systematic review and meta-analysis of the literature to assess theeffect of the ICU pharmacist intervention on MEs.

2. Methods

2.1. Scope

A systematic review of published works following the methodsspecified in the Cochrane Handbook for Reviews on Interventions [16]and the Meta-analysis of Observational Studies in Epidemiology guide-lines was conducted (Supplementary material 1) [17].

2.2. Searching

Studieswere identified by electronically searching theMEDLINE andEMBASE databases and the Cochrane Database until August 2014. Thefollowing keywordswere used in combination for the followingmedicalsubject headings and text words: “critical care,” “intensive care unit,”

Journal of Critical Care 30 (2015) 1101–1106

⁎ Corresponding author.E-mail addresses: tiansheng@bjmu.edu.cn (T. Wang), benedictnj@upmc.edu

(N. Benedict), kolsen@unmc.edu (K.M. Olsen), luanrong1989@163.com (R. Luan),xizhuccm@163.com (X. Zhu), keaidehaizimen@yeah.net (N. Zhou), hltang1985@163.com(H. Tang), yanyingying89@163.com (Y. Yan), pkulennoxyao@gmail.com (Y. Peng),shilu@bjmu.edu.cn (L. Shi).

http://dx.doi.org/10.1016/j.jcrc.2015.06.0180883-9441/© 2015 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Journal of Critical Care

j ou rna l homepage: www. jcc journa l .o rg

Effect of critical care pharmacist's intervention on medication errors: Asystematic review and meta-analysis of observational studies

Tiansheng Wang, PharmD a, Neal Benedict, PharmD b, Keith M. Olsen, PharmD c, Rong Luan, MS a,d,Xi Zhu, MD d, Ningning Zhou, MS a, Huilin Tang, MS d, Yingying Yan, PhD candidate d,Yao Peng, MS a, Luwen Shi, MHPE a,⁎a Department of Pharmacy Administration and Clinical Pharmacy, School of Pharmaceutical Sciences, Peking University Health Science Center, Beijing, Chinab Department of Pharmacy and Therapeutics, University of Pittsburgh School of Pharmacy, University of Pittsburgh Medical Center, Pittsburgh, PA, USAc Department of Pharmacy Practice, College of Pharmacy, University of Nebraska Medical Center, Omaha, NE, USAd Department of Pharmacy, Peking University Third Hospital, Beijing, China

a b s t r a c ta r t i c l e i n f o

Keywords:Pharmaceutical carePharmacist interventionMeta-analysisIntensive care unitCritical careMedication error

Pharmacists are integral members of the multidisciplinary team for critically ill patients. Multiple nonrandomizedcontrolled studies have evaluated the outcomes of pharmacist interventions in the intensive care unit (ICU). Thissystematic review focuses on controlled clinical trials evaluating the effect of pharmacist intervention on medica-tion errors (MEs) in ICU settings. Two independent reviewers searchedMedline, Embase, and Cochrane databases.The inclusion criteriawere nonrandomized controlled studies that evaluated the effect of pharmacist services vs nointervention on ME rates in ICU settings. Four studies were included in the meta-analysis. Results suggestthat pharmacist intervention has no significant contribution to reducing general MEs, although pharmacistintervention may significantly reduce preventable adverse drug events and prescribing errors. This meta-analysis highlights the need for high-quality studies to examine the effect of the critical care pharmacist.

© 2015 Elsevier Inc. All rights reserved.

1. Background

Over the past 3 decades, the role of the critical care pharmacist hasexpanded in the intensive care unit (ICU) from traditional dispensingresponsibilities to active members of multidisciplinary teams [1].Pharmacists play an important daily role in ICU patient care by assistingphysicians and other health care providers with pharmacotherapy deci-sionmaking andmonitoring, ultimately improving medication safety [2].

Medication errors (MEs) and adverse drug events (ADEs) occur athigh rates in the ICU [3] due to the (1) severity of disease of this popula-tion, (2) multiple medications administered, (3) high-risk nature of themedications, and (4) high frequency of changes in pharmacotherapyregimens [1,3-6]. Pharmacists are considered to be effective in reducingMEs andADEs by reducing the causes of these events throughpharmaco-therapy decision making and monitoring support [2,4]. Numerousstudies have assessed the effect of ICU pharmacists' intervention inreducing MEs [6-15]. However, these data are limited by the ability tocollect data without impeding care; thus, few of these studies arerandomized controlled trials [1] or are cohort studies that have an ICU

without a pharmacist as a control site [10,15]. Conversely, most ofthese studies are preintervention/postintervention designs [6-14],which are more realistic to conduct in the ICU setting.

It is, therefore, challenging to quantitatively evaluate the effectof ICU pharmacist interventions on MEs. Pooling data from availablestudies to perform a meta-analysis will help to better determine thetrue effect and magnitude of the interventions. Based on our literaturereview, no such meta-analysis exists. Therefore, we aimed to conducta systematic review and meta-analysis of the literature to assess theeffect of the ICU pharmacist intervention on MEs.

2. Methods

2.1. Scope

A systematic review of published works following the methodsspecified in the Cochrane Handbook for Reviews on Interventions [16]and the Meta-analysis of Observational Studies in Epidemiology guide-lines was conducted (Supplementary material 1) [17].

2.2. Searching

Studieswere identified by electronically searching theMEDLINE andEMBASE databases and the Cochrane Database until August 2014. Thefollowing keywordswere used in combination for the followingmedicalsubject headings and text words: “critical care,” “intensive care unit,”

Journal of Critical Care 30 (2015) 1101–1106

⁎ Corresponding author.E-mail addresses: tiansheng@bjmu.edu.cn (T. Wang), benedictnj@upmc.edu

(N. Benedict), kolsen@unmc.edu (K.M. Olsen), luanrong1989@163.com (R. Luan),xizhuccm@163.com (X. Zhu), keaidehaizimen@yeah.net (N. Zhou), hltang1985@163.com(H. Tang), yanyingying89@163.com (Y. Yan), pkulennoxyao@gmail.com (Y. Peng),shilu@bjmu.edu.cn (L. Shi).

http://dx.doi.org/10.1016/j.jcrc.2015.06.0180883-9441/© 2015 Elsevier Inc. All rights reserved.

Contents lists available at ScienceDirect

Journal of Critical Care

j ou rna l homepage: www. jcc journa l .o rg

of ICU patients. These quantitative results from meta-analysis areconsistent with the study by Campino et al [8], which showed a moresignificant effect of pharmacist intervention on potentially severe errorsthan moderate errors.

Lee et al [13] observed an increased risk of MEs. The reasons for thisdiscrepancy might include differing ICU populations, varying pharma-cist engagement, inadequate control of confounders such as severityof illness, pharmacist training, interprofessional team make-up, andlack of classification for MEs (ie, “lack of monitoring” was counted asan ME) [13]. Exclusion of this study from post hoc analysis generateda significant estimated effects with OR of 0.25 (0.22-0.28) and insigni-ficant heterogeneity (I2 = 0).

This meta-analysis highlights the need for additional evidence re-garding the impact of a critical care pharmacist on patient safety out-comes. Specifically, the identification and analysis of patient safetyoutcomes, such as ADE and ME rates, are needed from pharmacistswho are members of ICU care teams with activities that included inter-disciplinary rounds. As it stands, “best practice” reports on patient safetylists “clinical pharmacist consultation services” as a medium strength ofevidence regarding their impact and effectiveness [27]. It is possible thatthis is the highest strength of evidence that clinical pharmacy serviceswill receive due to feasibility and ethical concerns of better qualitytrial designs (ie, randomized controlled trials).

4.2. Quality of evidence

The evidence included in this analysis is at high risk for bias. Most in-cluded studies were preintervention/postintervention observationalstudies, and important confounding factors might not have been mea-sured with sufficient precision or might not have been measured atall. In addition, no statistical techniques such as propensity score or

Cox models have been used to adjust for biases. Only in 3 studies[9,11,12] were blinded reviewers used for outcome measurement.

Because of a relatively small number of studies includedand the inherent inability of before-and-after studies to eliminateunmeasured/residual confounding, the results of our meta-analysisshould be interpreted with caution. Future studies with better quality,for example, adjusting for potential confounders and having a controlunit, are needed to more accurately assess the impact of pharmacistson patient clinical outcomes.

4.3. Strengths and limitations

This meta-analysis systematically identified relevant observationalstudies to analyze and synthesize data to evaluate the impact of criticalpharmacist interventions on MEs. To our knowledge, this is the firstsystematic review that quantitatively evaluated this impact on differenttypes of MEs and ADEs. Nevertheless, observational studies are prone tobiases, such as prescription, selection, and misclassification, which willaffect the internal validity of this systematic review and meta-analysis.In addition, this analysis is limited by the source of included studies.For example, gray literature or unpublished data were not consulted.

5. Conclusion

The present systematic review and meta-analysis did not demon-strate a significant beneficial effect of the intervention on general MEs.However, the pooled analysis supported the role of pharmacists in re-ducing preventable ADEs and prescribing errors. Future high-qualitystudies that examine the effect of a critical care pharmacist arewarranted.Specifically, carefully designed and conducted observational studies thatclearly define the study population and total number of medication

Fig. 3. Subgroup analysis of MEs and preventable ADEs between ICU patients with pharmacist intervention or control, using random-effects model.

Table 4Outcomes and results of studies not reporting monitored patient-days

Study (year) Study design Patients Studyperiod (d)

Monitoredpatient-days

Type of ME Total no. of prescriptionduring study period

ME

Alagha et al (2011) [7] HCS B 139 B 150 NR Prescribing error (potentially severe, moderate, minor) B 1417 B 1107A 101 A 150 NR A 1096 A 391

Campino et al (2009) [8] HCS N/A B 150 NR Prescription error and transcription error B 4182 B 857N/A A 210 NR A 1512 A 47

Jiang et al (2012) [9] HCS B 409 B 91 NR Prescribing error NR B 687A 416 A 91 NR NR A 191

Nguyen et al (2014) [10] CBA N/A B 7 NR Did not specify B 1204 B 236N/A A 7 NR A 688 A 407

1105T. Wang et al. / Journal of Critical Care 30 (2015) 1101–1106

MedicationerrorƺYƁNJƻǝŐĄƑĚ��ČĕċºǠù�NJƻǝ

ƾǞǕNJǎǟƹLJljƵǝƾNj

•ÎÛƝŃǩǻșǻȓǎǚǝN�ēǏƢ�ǐ`RǎǚǜǕdžǕdžNJƴǝƲ

•ƅē�ĽǎƸƽǝÎÛƝŃǩǻșǻȓN�ēǏđsǐǓNjǡnj`RƿǞljƵǍƵƲ

•ÎÛƝŃǩǻșǻȓīĊǒǏƒ�ǐƩƵNjǐŝƷǂƱu+īĊǏÊăǐ�ÄNJƴǝ

ÏĞĦǏęė

• ICUNJ¯*ƿǞDžÎÛƝŃǩǻșǻȓǎǚǝƝŃāǏĕċƢ�NjǃǏȒǴǫȄǢǫǸșǏŨ×

•ƝŃǩǻșǻȓīĊǎƸƽǝu+�Ť!NjICU~�/~�Ő8�ǏÊČ�Ť!

¾ò•ĞĦǼDZǥȖȝi¿š7PƻǮȈșǽĞĦ

•}Ū�Ľȝƚ�ðĔyǎ*yǀDž18ç��Ǐ�Ľ

•ƚ�ðĔyǎ*y�ƹǛŹ0�ǕNJǎ¯*ƿǞljƵǝÎÛƝŃǩǻșǻȓǎƒǟǝ�`ǠJƚ

• ǣǦǽǩȋȝƝŃāĕċƢ�ƱȒǴǫZqǏª0

•ĞĦÌƑȝ3ȗÉƑȚ2018�1É1ÁƹǛ3É31ÁǠvț

}Ū�Ľ

ſÚ_ü1.+ICU*y�Ľ

Ɨh_ü1. ĞĦH<¬QǠğǀDžaN2. «�CƺĞĦH<[ƛNj3¼ǀDžaN3. 18çÍû

}ŪǩǻșǻȓÎÛƝŃŲƱ��ƝŃŲƱÎÛ¯*^��ƝŃŲƱŻÓǩǻșǻȓƱŀ?ŃǩǻșǻȓƱ?Ńǩǻșǻȓ

�*ĞĦNJǐǍƵDžǗǎƱĉêƘNJǐǨȆǽǣǦǽǏǖNJŊƵNjļƷǝ

ǼșǸJƚƠę

• ¿š�`ȚM¿šǏ'ĊpT�¥ŧ�ǎREDCapǒ*;ț

• �Ľ_Ï�`Ț�ĽĖƋÆǎREDCapǒ*;ț

• ǩǻșǻȓƒŽ�`ȚDay1ƹǛŹ0ǕNJƱǩǻșǻȓëǎŠƋȝÁƳREDCapǒ*;ƲŐ8�`ǏǖǩȓǻǏǮȃșǠ�l¿šǒź�ț

• �Ľŵ��`Țé ÆƱICUŹ0ÆƱĒƖŹƖÆȝREDCapǒ*;ț

ǨȖȑǥȖǼșǸȇșǴ�¡ȝREDCapǠ�Čv

¿š�`

1.ĒƖĒ�º 6.ICUĒÜ~�/~�~�Ő8�ǏÊă

2.ICUĒ�º 7.ICUǎƸƽǝŐ8¦�áü@}ĩǏÊă

3.ICU~�CǏÊă 8.ƝŃŲǎƒǀlj~ƎėǎǽȔșȀȖǬƿǞDžěũ� (IVnurse)ǏÊă

4.ICUěũ�5 9.ǩǻșǻȓǏvÌ�³ǏÊă

5.ƝŃāŢ¼_üǏÊă

10.ǩǻșǻȓƒŽŒô�ÖēǏŢ¼_üǏÊă

�Ľ_Ï�`

1.�ƪ 9.*y��ȚvȘĵ�ț

2.�4 10.ICU*yĊĎǩǻǯȒș

3.Ŵƍ/�ƅ 11.ICU*yÆǎǁNJǎ¯*ƿǞljƵǝǩǻșǻȓȚƝŃŲȘ?Ń]ȑǥȖțIǑ¯*a£Ș¯*Ľ

4.À�è (charlson,ǃǏ�) 12.ǩǻșǻȓ¯*Ƃ�

5.ƅē�ǴǮǣȚAPACHE2ȘSAPS2ȘSOFAț 13.ǩǻșǻȓǏÐŮȘǰǥǵ

6.ĒƖ*Ɩ�ÉÁ 14.ǾȔǺDzȖǬ¾ò

7.ICU*y�ÉÁ 15.�ŦǩǻșǻȓëǎICU*yÆǎ¦�ƿǞljƵǝŐ8�`ȚĥƥƱĀ�Ʊż�ț

8.ICU*yıŲȚ¤ŕyȘĒÜȘ·�hÑȘHCUț

ǩǻșǻȓƒŽ�`-ǩǻșǻȓ¯*Æ�`-

1.¯*ƿǞljƵǝŒī/ǩǻșǻȓǏĥƥIǑ¯*a£Ș¯*Ľ

6.ǩǻșǻȓ¯*ÆǏöìŐ

2.ǩǻșǻȓ¯*Ƃ� 7.ǩǻșǻȓ¯*ÆǏűƟóǏ�ČǏÊă

3.ǩǻșǻȓ¯*ÁÆ 8.ǩǻșǻȓħ6ÆǏħ6Yº4.ǩǻșǻȓǏÐŮȘǰǥǵ 9.ƞýō¤ŗ�ČÊăƱýō¤ŗǏ

Êă

5.ǾȔǺDzȖǬ¾ò

ǩǻșǻȓƒŽ�`-ÁƳǏŤ!Ơę/ǩǻșǻȓ¨GÆ�`-

1.ǃǞDŽǞǏǩǻșǻȓƹǛ¦�ƿǞDžŐ8ȚŘ8Ʊİ¡ƱÈjż�țNj¦�ÆƑ

7.ƝŃāǏÊăȚpain,redness,erythema,swelling,venous cord,steakformation,feverǠŠŷțNjǃǏÁÆ

2.ǩǻșǻȓ6*ƂǏ£ŚȚ4ÆƑëț 8. MajorcomplicationǏÊăȚƝŃāƱǩǻșǻȓƏdƱǩǻșǻȓƒŽ�ÖēƱōŒēƱ?Ńħ6ƱŒńƱŒīhÿ0Ʊġı´)ƱŅ´)ƱĘņféƱǮȖȂșǽȌȖǽē&ĺțNjǃǏÁÆ

3.ǾȔǺDzȖǬ¾ò 9.IVnurseǏƒ�ǏÊă

4.�Ŧǩǻșǻȓ¯*ÌƑ�Ǐ�ÖēǏÊăNjǃǏĥƥ

10. ICU~�/~�Ő8�ƒ�ǏÊăNjƒ�/{ȚƄNǙȓșǽǎƒǁǝǹǧǺǫâàǏÊăț

5.ǩǻșǻȓƒŽ�ÖēǏÊăNjǃǏÁÆ 11.ǩǻșǻȓ¨GǏÊăƱÁÆƱĊĎ

�Ľŵ��`

1.ICUŹ0�ÉÁ 5.ICUé ǏÊă2.ICUŹ0Æŵ� 6.Ɩ/é ǏÊă3.ĒƖŹƖ�ÉÁ 7.é ÁÆ4.ĒƖŹƖÆŵ�

ś|¾òȘǴǭdzȏșȓ

Ʈǩǻșǻȓ6*ƂǏś|Ƒƙěũ�ȝ4ÆƑëǎƝŃāǏŢ¼_üƠęǠś|

Ư¦�ƿǞljƵǝŐ8ǏŠƋ (ǩǻșǻȓë)ǃǞDŽǞǏǩǻșǻȓƹǛ¦�ƿǞDžŐ8�`ȚƜqǩȓǻǠǪȎȆǹȎǀlj�l¿šǎźǝƲ�l¿šNJİ¡Ʊ¦�ÌƑƱ¦�ż�ǠJƚǁǝƲț

ưś|ÌƑ¯*ƿǞDžǩǻșǻȓƺ¨GƿǞǝǕNJƱǘǀƼǐ}Ū�ĽƺICUǠŹ0ǁǝǕNJ

Ť!Ơę�řŤ!ƠęÎÛƝŃǩǻșǻȓǎƸƽǝƝŃāǏĕċ:NIǑÎÛƝŃāǏȒǴǫȄǢǫǸșǏª0

9äŤ!Ơę1. ÎÛƝŃŲǎƸƽǝmajorcomplicationǏĕċ:N2. ICUǎƸƽǝMĥƝŃŲǏ¯*Ćñ3. ÎÛƝŃǩǻșǻȓƒŽŒô�ÖēǏÊă4. MĥŐ8Ǐ¦�ȓșǽǏn��5. ƄNĢ�ǏŐ8Ǐ�ČĆñ6. ICU~�/~�Ő8�Ǐ>Õ7. ƝŃāǏGrade4ǎ1ƽljŜÓǁǝ (Sub-groupŜÓ)

ƝŃāǏvĻ

MajorcomplicationǏvĻ

ƝŃā Œīhÿ0

ǩǻșǻȓƏd ġı´)

ǩǻșǻȓƒŽ�Öē Ņ´)

ōŒē Ęņfé

?Ńħ6 ǮȖȂșǽȌȖǽē&ĺ

Œń

ęáǰȖȆȓǰǥǵ

5500ÏǏÎÛƝŃǩǻșǻȓ

Ù­1.ÎÛƝŃǩǻșǻȓƝŃāĕċƢ�Ǐ#ƣBƑǏÞţƝŃāĕċƢ�Ǡ5%Ʊ95%#ƣBƑƭ1%Nj�v→1825Ï

2.ÎÛƝŃǩǻșǻȓȒǴǫZqǏÞşȕdzǴǻǤǺǫY�ȍǼȓǎƸƵljƝŃāǙN�ēǏ-gƇǠ25%NjǁǝDžǗǎ�řǍǰȖȆȓºƱƝŃāǏĕċƢ�Ǡ5%Ʊã´ǼșǸǏ:NǠ10%Nj�v→5500Ï (į2750�)

ęá¿šº

15¿š

Ù­1ÁǏ½*Ɩ�Ľºȝ2Oś|ÌƑȝ90ÁƑ1�ĽƴDžǜǏ¯*Ϻȝ2Ï→1¿šƴDžǜ360Ï (�Ľ 180�)15¿šȝ5400Ï

ŜÓ¾ò

�řŤ!Ơę1.ƝŃāĕċ:Nȝ:N (%)ƸǚǑ95%#ƣBƑ2.ƝŃāĕċºȝ 1000ǩǻșǻȓ¯*ÁƴDžǜǏƝŃāĕċºIǑ95%#ƣBƑ→ repeatedmeasurementǎǍǝKł�ƺƩƼƱǃǏaNǎǐ

marginalCoxregression analysisȚÆƑǠļ�ǁǝaNțǘǀƼǐmixedeffectmodelȚÆƑǠļ�ǀǍƵaNțNJŜÓǠŔƶ

3.ȒǴǫȄǢǫǸșª0ȝ marginalCoxregression analysis�ĽNJƘ�@ǀƱ-gƇNjǀlj��ǏZqǠÞşǁǝ

�ƪ ǩǻșǻȓǏǰǥǵ ěũ�5 Ő8�ǏÊă

�4 ǾȔǺDzȖǬ¾ò űƟó�ČǏÊă Ő8 (8ĥƥ)ȝTWA

BMI öìŐ ¯*ÌƑ IVnurseǏÊă

SAPS2 ýō¤ŗǏÊă *yıŲ

¯*Ƃ� ħ6Yº IVnurseǏÊă

Ő8ȝǩǻǮȑȊȖƱǩȒǦȋƱȀǩȓdzȃȖƱȅǾǦĮƱPNƱ§ōŐƱȄǧȀǽǥȖ

Time-WeightedAverage(TWA)ǩȒǦȋĀ� (mEq/L)Ʊż� (ml/hr)Ʊ¦�ÆƑ (hr)ƹǛƱĀ�*ż�/1000(mEq/hr)ǏȂȑȌșǸǠ�¡ƾǏȂȑȌșǸǎ}ǀljTWAǠŞĪǁǝ

ǩǻǩǻǮȑȊȖǩǻǮȑȊȖIndex(μg/kg/min)Ǡ�¡ƾǏȂȑȌșǸǎ}ǀljTWAǠŞĪǁǝ

ȀǩȓdzȃȖĀ� (mg/ml)Ʊż� (ml/hr)Ʊ¦�ÆƑ (hr)ƹǛƱĀ�*ż� (mEq/hr)ǏȂȑȌșǸǠ�¡ƾǏȂȑȌșǸǎ}ǀljTWAǠŞĪǁǝ

ȅǾǦĮĀ� (g/ml)Ʊż� (ml/hr)Ʊ¦�ÆƑ (hr)ƹǛƱĀ�*ż� (g/hr)ǏȂȑȌșǸǠ�¡ƾǏȂȑȌșǸǎ}ǀljTWAǠŞĪǁǝ

ƝŃØƦõŻ]íƱż� (ml/hr)Ʊ¦�ÆƑ (hr)ƹǛƱõŻ]í*ż� (ml/hr)ǏȂȑȌșǸǠ�¡ƾǏȂȑȌșǸǎ}ǀljTWAǠŞĪǁǝ

ĞĦŭƈ

•ÁÏsŕ®ʼn�ģsĞĦŬ=¡�ßŋ¤(B)

H<¿š

•čĶNĒƖ•æŎƆůDrĒƖ•��js•úFƌ$ĒƖ•őč"(Ŗċjs•b�ĨĒƖ•ňðCjƿƵDžǕCĔǶȖǸș•ÁÏůDrĠSå�CĔǶȖǸș

Þş�

•�L�ĨĒƖ•��js•wc�ĨĒƖ•ïķĜĨ�ƂĒƖ•ƩĝjsCsƂƔ�ĒƖ•ġ¢�Ĩ�îCĔǶȖǸș•øċ�tƃzĒƖ•ÒAjsĒƖ• JA��ĶNĒƖ•Ɲ�ĜĨĶNĒƖ

ĞĦǴǭdzȏșȓ

• 2017�5É9Áȝresearchmeeting• 2017�5É�ÂȝčIRB²0• 2017�6É�ȝƚ�ðĔģƱICUěũ�¹Ŀ• 2017�7É1Á-14ÁȝȂǥȕǺǽĞĦȟč• 2017�8ÉȝUƤĂª0ȜµW• 2017�9É6ÁȝÁÏƚ�ðĔCs�CTGNJȆȔǷȖǻșDzȐȖ

• 2017�10ÉǚǜREDCap*;DzǴǻȋàĬ• 2017�10É24Áȝkick-offmeetingȟ·�Cs�• 2017�12ÉȝM¿šǒǏ¹Ŀ• 2018�1É1Á-3É31ÁȝÏťƨƐo