INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY OCTOBER 2 0 1 4 VOL 3 5 NO S3

O R I G I N A L A R T I C L E

Racial and Ethnic Disparities in Healthcare-Associated Infections in the United States 2009-2011

Anila Bakullari BS1 Mark L Metersky MD12 Yun Wang PhD3 Noel Eldridge MS4 Sheila Eckenrode RN1

Michelle M Pandolfi MSW MBA1 Lisa Jaser PharmD15 Deron Galusha MS16 Ernest Moy MD MPH4

BACKGROUND Little is known about racial and ethnic disparities in the occurrence of healthcare-associated infections (HAIs) in hosshypitalized patients

OBJECTIVE To determine whether racialethnic disparities exist in the rate of occurrence of HAIs captured in the Medicare Patient Safety Monitoring System (MPSMS)

METHODS Chart-abstracted MPSMS data from randomly selected all-payer hospital discharges of adult patients (18 years old or above) between January 1 2009 and December 31 2011 for 3 common medical conditions acute cardiovascular disease (composed of acute myocardial infarction and heart failure) pneumonia and major surgery for 6 HAI measures (hospital-acquired antibiotic-associated Clostridium difficile central line-associated bloodstream infections postoperative pneumonia catheter-associated urinary tract infections hospital-acquired methicillin-resistant Staphylococcus aureus and ventilator-associated pneumonia)

RESULTS The study sample included 79019 patients who had valid racialethnic information divided into 6 racialethnic groupsmdashwhite non-Hispanic (n = 62533) black non-Hispanic (n = 9693) Hispanic (n = 4681) Asian (n = 1225) Native HawaiianPacific Islander (n = 94) and other (n = 793)mdashwho were at risk for at least 1 HAI The occurrence rate for HAIs was 11 for non-Hispanic white patients 13 for non-Hispanic black patients 15 for Hispanic patients 18 for Asian patients 17 for Native HawaiianPacific Islander patients and 070 for other patients Compared with white patients the agegendercomorbidity-adjusted odds ratios of occurrence of HAIs were 11 (95 confidence interval [CI] 099-123) 13 (95 CI 115-153) 14 (95 CI 107-175) and 07 (95 CI 040-112) for black Hispanic Asian and a combined group of Native HawaiianPacific Islander and other patients respectively

CONCLUSIONS Among patients hospitalized with acute cardiovascular disease pneumonia and major surgery Asian and Hispanic patients had significantly higher rates of HAIs than white non-Hispanic patients

Infect Control Hosp Epidemiol 201435(S3)S10-S16

The US Census Bureau predicts that in 30 years no individual ethnic disparities in patient safety specifically concerning racial or ethnic group will represent a majority of the country healthcare-associated infections (HAIs) In fact others have as a whole1 Racial and ethnic minority groups are growing previously noted the gap in knowledge regarding racial and at a faster rate than the total US population resulting in a ethnic disparities in HAIs6 Prior studies have been based on diverse population in nearly every state Between 2000 and administrative data78 have assessed only a single HAI8 or 2010 the Hispanic population alone accounted for more than have focused on patient safety events in the Medicare pop-half of the total US population growth2 Similarly the Asian ulation9 resulting in a knowledge gap for the younger all-population grew at a faster rate than all other ethnic groups payer patient population

during this time period3 Due to the changing demographics In this study we measured racial and ethnic disparities in the frequency with which racial and ethnic minorities interact the rates of 6 HAIs included in the Medicare Patient Safety with the US healthcare system is rising Monitoring System (MPSMS) a national chart abstraction-

The existence of racial and ethnic disparities in outcomes based longitudinal database of 21 healthcare-associated ad-and process of care has been extensively documented4 with verse event measures jointly developed by federal agencies one study estimating excess healthcare costs of $24 billion and private healthcare organizations Despite the name of the between 2006 and 2009 for black and Hispanic men alone5 system during the study period charts were taken from an However there have been fewer investigations into racial and all-payer sample not only from patients covered by Medicare

Affiliations 1 Qualidigm Wethersfield Connecticut 2 Division of Pulmonary and Critical Care Medicine University of Connecticut School of Medicine Farmington Connecticut 3 Department of Biostatistics Harvard School of Public Health Boston Massachusetts 4 Agency for Healthcare Research and Quality US Department of Health and Human Services Rockville Maryland 5 Griffin Hospital Pharmacy Derby Connecticut 6 Section of General Internal Medicine Department of Internal Medicine Yale University School of Medicine New Haven Connecticut

Received April 23 2014 accepted May 1 2014 electronically published September 15 2014 copy 2014 by The Society for Healthcare Epidemiology of America All rights reserved 0899-823X201435S3-0003S1500 DOI 101086677827

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use

DISPARITIES IN HEALTHCARE-ASSOCIATED INFECTIONS S l l

These measures are considered relevant because they capture common and costly adverse medical events which are thought to be frequently preventable with optimum care10

in hospitalized patients

METHODS

Study Sample

Our study sample was drawn from the MPSMS1011 and was restricted to 6 HAI measures hospital-acquired antibiotic-associated Clostridium difficile central line-associated bloodstream infections (CLABSIs) postoperative pneumoshynia catheter-associated urinary tract infections (CAUTIs) hospital-acquired sterile-site methicillin-resistant Staphyloshycoccus aureus (MRSA) and ventilator-associated pneumonia (VAP) Due to the small number of cases the hospital-acquired vancomycin-resistant Enterococcus measure was exshycluded from this analysis Because both race and ethnicity information became available starting in 2009 we restricted our sample to patients discharged between January 1 2009 and December 31 2011 with 3 common medical condishytions acute cardiovascular disease (composed of acute myocardial infarction and heart failure) pneumonia and major surgery as defined by the Centers for Medicare amp Medicaid Services Surgical Care Improvement Project (SCIP)12 Patients with missing race or ethnicity information or patients younger than 18 years were excluded from the study sample

Patient Characteristics

Patient characteristics included demographics (age gender race and ethnicity) and comorbidities Race and ethnicity information was represented by 2 variables in the MPSMS data race (white black Asian American IndianAlaska Nashytive and other) and ethnicity (Hispanic non-Hispanic) Abshystractors determined a patients race and ethnicity by searchshying for appropriate documentation in the medical record face sheet If the face sheet did not contain information about raceethnicity or had conflicting documentation abstractors searched emergency department records history and physishycals nursing admission assessments physician assessments and electrocardiograms A predetermined inclusion list for each race and ethnic group vetted by the Agency for Healthshycare Research and Quality (AHRQ) Disparities Team inshyformed abstractors on how to categorize patients into the 6 groups The other group is composed of patients who could not be categorized into one of the 5 race and ethnicity catshyegories or had mixed race documented in the medical recshyord or face sheet Patient comorbidities were also identified from the medical records including history of heart failure obesity coronary artery disease renal disease cerebrovascular disease chronic obstructive pulmonary disease cancer diashybetes and smoking Using the race and ethnicity information we classified each patient into 1 of 6 mutually exclusive catshyegories white non-Hispanic black non-Hispanic Hispanic Asian Native HawaiianPacific Islander and other

Outcomes

Our outcomes were HAI-related adverse event rates for which patients were at risk during hospitalizations For exshyample for MRSA to be considered hospital acquired the patient must not have been documented as a carrier of MRSA or have a positive culture and sensitivity for MRSA on admission and must have had a positive culture and sensitivity for MRSA obtained from a sterile site more than 2 days after arrival to the hospital Algorithms for all these events are available online (httpwwwqualidigmorg indexphpcurrent-initiativesmpsms)

Our primary interests were 2 composite outcomes of the 6 included infection measures used in a previous study (1) the rate of occurrence of HAIs for which patients were at risk and (2) the rate of patients with 1 or more HAIs The unit of analysis was at the individual HAI level for the first outcome and the patient level for the second outcome Not all patients were at risk for all HAIs For example only pashytients who underwent bladder catheterization were at risk for CAUTI Furthermore a patient might have had more than 1 HAI

Statistical Analysis

We fitted mixed-effects models with a logit link function to evaluate the relationship between raceethnicity and the rate of occurrence of adverse events and the rate of patients with 1 or more adverse events Models were fitted with state-speshycific random intercepts to account for within-state and beshytween-state variations and were adjusted for patient charshyacteristics (age gender and comorbidities) Due to small sample size Native HawaiianPacific Islanders were combined with other for these analyses All models included 4 dummy variables black non-Hispanic Hispanic Asian and a comshybined group of Native HawaiianPacific Islander and other patients with the reference group being white non-Hispanic Analyses were conducted using SAS (ver 92) We obtained institutional review board approval through the IRB Solutions Human Investigation Committee which waived the requireshyment for informed consent

RESULTS

The final study sample included 79019 patients who had valid raceethnic information and were discharged from January 1 2009 to December 31 2011 including 33348 with acute cardiovascular disease 23368 with pneumonia and 22303 with major surgery as defined by SCIP12

Table 1 shows the patient demographic data and the rates of common comorbidities identified through chart abstracshytion Black patients were younger (average age 613 years SD 164) and had a higher percentage of females (569) than all other racial and ethnic groups White patients were older with an average age of 694 years (SD 159) Asian patients had the lowest rate of obesity (107) and the lowest percentage of females (491) Overall the most frequently occurring comorbidity was congestive heart failurepulmo-

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use

S l 2 INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY OCTOBER 2 0 1 4 VOL 3 5 NO S3

TABLE 1 Patient Characteristics by Race and Ethnicity in Patients with Healthcare-Associated Infections

Characteristic

Demographics

Age

Mean (SD) years

Age group

18-64 years

65-74 years

75-84 years

85 years and over

Gender

Female

Condition

Acute cardiovascular disease

Pneumonia

Surgical care

Comorbidities

Cancer

Diabetes

Obesity

Cerebrovascular disease

CHFpulmonary edema

COPD

Smoking

Renal disease

Aggregate

79019 (1000)

680 (163)

40891 (518)

13540 (171)

13942 (176)

10646 (135)

42708 (541)

33348 (422)

23368 (296)

22303 (282)

15324 (194)

28138 (356)

18503 (234)

13070 (165)

33006 (412)

24694 (313)

18930 (240)

20795 (263)

White

non-Hispanic

62533 (791)

694 (159)

30621 (490)

10883 (174)

11723 (188)

9306 (149)

33684 (539)

25700 (411)

18750 (300)

18083 (289)

12989 (208)

20730 (332)

14195 (227)

10424 (167)

25608 (410)

20647 (330)

14643 (234)

15685 (251)

Black

non-Hispanic

9693 (123)

613 (164)

6387 (659)

1467 (108)

1147 (118)

692 (71)

5517 (569)

4641 (479)

2659 (274)

2393 (247)

1450 (150)

4195 (433)

2729 (282)

1693 (175)

4674 (482)

2621 (270)

2944 (304)

3173 (327)

Hispanic

4681 (57)

635 (171)

2728 (583)

685 (146)

685 (146)

414 (88)

2419 (517)

2082 (445)

1289 (275)

1310 (280)

577 (123)

2280 (487)

1210 (259)

655 (140)

1877 (401)

997 (213)

927 (198)

1284 (274)

Asian

1225 (16)

684 (164)

574 (469)

210 (171)

280 (229)

161 (131)

602 (491)

595 (486)

341 (278)

289 (236)

192 (157)

552 (451)

131 (107)

204 (167)

521 (425)

223 (182)

177 (145)

412 (336)

Native Hawaiian

Pacific Islander

94 (01)

625 (165)

57 (606)

18 (192)

12 (128)

7 (75)

53 (564)

55 (585)

19 (202)

20 (213)

16 (170)

47 (500)

36 (383)

13 (138)

56 (596)

22 (234)

15 (160)

46 (489)

Other

793 (10)

622 (179)

524 (661)

108 (136)

95 (120)

66 (83)

433 (546)

275 (347)

310 (391)

208 (262)

100 (126)

334 (421)

202 (255)

81 (102)

270 (341)

184 (232)

224 (283)

195 (246)

NOTE Data are no () unless otherwise indicated CHF congestive heart failure COPD chronic obstructive pulmonary disease SD standard deviation

nary edema (412) followed by diabetes (356) and chronic obstructive pulmonary disease (313)

Table 2 shows the observed HAI rates among the 6 racial and ethnic categories The highest aggregate infection rate was for VAP (110) and the lowest aggregate infection rate was for sterile-site MRSA (01) These rates pertain only to the patient population at risk for the event which varies for each type of HAI for example almost all patients were at risk for MRSA and only a relatively small fraction were at risk for VAP Among those at risk the rate of VAP was the highest of the 6 infection measures for all racialethnic groups with the exception of Native HawaiianPacific Islander and other for which CAUTI rates were the highest The rate of at-risk patients who developed at least 1 HAI during the hospital stay was highest for Asian and Hispanic patients at 38 and 33 respectively This was followed by black white and other patients who had rates of 26 26 and 18 respectively Similarly the occurrence rate of HAI was 11 for non-Hispanic white patients 13 for non-Hispanic black patients 15 for Hispanic patients 18 for Asian patients 17 for Native HawaiianPacific Islander patients and 07 for other patients

The above pattern of observed rates does not change subshystantially after adjustment for patient age gender and coshymorbidities Compared with white non-Hispanic patients the adjusted odds ratios of occurrence of HAIs were 11 (95 confidence interval [CI] 099-123) for black patients 13

(95 CI 115-153) for Hispanic patients 14 (95 CI 107-175) for Asian patients and 07 (95 CI 040-112) for Native HawaiianPacific Islander and other patients comshybined (Figure 1A) The adjusted odds ratios for at least 1 HAI followed the same pattern as the adjusted odds ratios of occurrence rates of HAIs (Figure IB)

D I S C U S S I O N

We studied more than 79000 patients admitted to the hospital with acute cardiovascular disease pneumonia or major surgery as defined by SCIPmdashwhich represent approximately 26 of all Medicare hospitalizationsmdashto define risk-adjusted racial and ethnic disparities in HAIs For these conditions compared with non-Hispanic white patients Hispanic and Asian patients had higher rates of HAIs which remained statistically signifshyicant after adjustment for baseline patient characteristics and comorbidities Black patients had slightly higher rates of HAI than non-Hispanic whites but the difference was not statisshytically significant However review of the unadjusted HAI rates (Table 2) demonstrates that the higher overall HAI rate in Hispanics was in large part due to higher rates of CAUTI and VAP in these patients Asian patients had higher rates of C difficile CLABSI and CAUTI than white patients Also of inshyterest is that approximately half of the MPSMS-measured HAIs were CAUTIs The MPSMS definition of CAUTI (http wwwqualidigmorgindexphpcurrent-initiativesmpsms)

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use

TABLE 2 Impact of Race and Ethnic Characteristics on Specific Healthcare-Associated Infection Measures

White Aggregate non-Hispanic

Measure (n = 79019) (n = 62533)

Hospital-acquired antibiotic-associated Clostridium difficile 29159900 (05) 22447980 (05)

CLABSIs 765800 (13) 554449 (12) Catheter-associated UTIs 122136027 (34) 92428950 (32) Hospital-acquired sterile-site MRSA 5176833 (01) 3276833 (01) Postoperative pneumonia 51822803 (23) 41518456 (23) Ventilator-associated pneumonia 1711552(110) 1181109(106) Aggregate 2328202915 (12) 1768161707 (11)

Black Native Hawaiian non-Hispanic Hispanic Asian Pacific Islander Other in = 9693) in = 4681) (n = 1225) (n = 94) (n = 793)

376885 (05) 183476 (05) 10874 (11) 064 (00) 2621 (03) 13785 (17) 4382 (11) 3130 (23) 010 (00) 144 (23)

1664111 (40) 922024 (46) 28561 (50) 342 (71) 8339 (24) 119450 (01) 54557 (01) 31194 (03) 093 (00) 0776 (00) 602453(25) 321351(24) 7310(23) 122(46) 3211(14)

25255 (98) 24130 (185) 439 (103) 03 (00) 016 (00) 1223939(13) 17511920(15) 553108(18) 4234(17) 142007(07)

NOTE CLABSI central line-associated bloodstream infection MRSA methicillin-resistant Staphylococcus aureus UTI urinary tract infection

Dow

nloaded from httpsw

ww

cambridgeorgcore 04 N

ov 2020 at 010136 subject to the Cambridge Core term

s of use

S14 INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY OCTOBER 2 0 1 4 VOL 3 5 N O S3

A

Ct

nic (Ret) -

Other -

Hispanic -

Hispanic -

Asian -

lt-Decrease 1

067

1 1

1

i1dego

I I I

I

I lt

I I I i 110

I I I I I

I I

133

i

lncrease-gt i i i i

025 05 075 1 125 15 175 2 25

Adjusted odds ratio (odds of occurrence of adverse events)

B White-non-Hispanic (Ref)

Q_

gro

icity

et

hn

and

^ace

Other

Hispanic

Black-non-Hispanic

Asian

lt-Decrease i

063

i i

I

hoo i

i i

i i i i 134

I i I 109 I I I 1 138

lncrease-gt I i i i i i i

025 05 075 1 125 15 175 2 25

Adjusted odds ratio (odds of at least one adverse event)

FIGURE l Adjusted odds ratios for the occurrence of healthcare-associated infections (A) and at least 1 healthcare-associated infection (B) by race and ethnicity

depends primarily on the treating physician diagnosing and treating a UTI after an inpatient has received a urinary cathshyeter and is more sensitive than the widely used National Healthcare Safety Network definition13 Conversely the MPSMS definition of MRSA is very specific and includes only infections in normally sterile sites such as blood or spinal fluid

While there is extensive literature on racial and ethnic disshyparities in medical treatment and outcomes4 the literature on racial and ethnic disparities in patient safety is sparser and

there are limited data on racial and ethnic disparities in HAIs Studies using the claims-based AHRQ patient safety indicashytors14 have demonstrated higher rates of postoperative sepsis in blacks and Hispanics78 and in Asians7 Freedberg et al15 found a higher risk of recurrent C difficile infection in black patients while Murphy et al16 found that race and ethnicity predicted a risk of postdischarge C difficile but not hospital-onset C difficile Klevens et al17 found an almost doubled risk of hosshypital-acquired MRSA in blacks compared with whites In a multicenter study of more than 17000 patients Dick et al18

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use

DISPARITIES IN HEALTHCARE-ASSOCIATED INFECTIONS S l5

found that both black and white patients had similar rates of CLABSI and VAP Race and ethnicity had not previously been thought to be a significant risk factor for CAUTI

Several factors contribute to disparities in healthcare proshycesses and outcomes including language barriers unconshyscious bias and the more frequent use of lower-quality healthshycare facilities by minorities4 There is much evidence that the direct and indirect effects of economic and educational disshyparities play an important role4 For this reason our findings of a greater risk of HAIs in Hispanic and Asian patients compared with black patients were somewhat surprising Nashytionally both Asians and Hispanics have significantly higher mean household incomes and educational attainment than blacks with Asians having a higher mean household income and educational attainment than non-Hispanic whites2021

Therefore economic and educational disparities are unlikely to be the major explanation for our findings However both Hispanics and Asians have much lower rates of English proshyficiency than blacks and non-Hispanic whites22 Our finding of higher risk-adjusted national HAI rates in Hispanic and Asian patients suggests that language barriers may play an important role in the occurrence of HAIs Poor communishycation between healthcare providers and patients could result in increased HAI rates either directly or indirectly23 For exshyample ineffective communication could result in patients manipulating indwelling devices in a manner that could inshycrease the risk of infection or result in prolonged hospital stays for unrelated reasons which could also increase length of exposure to risk for HAIs Limited English proficiency is known to be a risk factor for adverse events in hospitalized patients24 but we found no studies that specifically examine the role of language barriers in HAIs Although the use of a professional interpreter is recommended whenever commushynicating with patients with limited English proficiency this recommendation is frequently not adhered to25 Our findings therefore provide a potential opportunity for improving HAI rates that has previously received little attention

Our study has some limitations that warrant discussion Although we reliably abstract race and ethnicity information from a patients chart the accuracy of this documentation is unknown For example a nonresponsive patient would be unable to provide information about their race or ethnicity thereby leaving that determination up to the admitting clishynician and introducing an opportunity for error We used a composite rate of HAIs encompassing 6 different infections As there are many factors that contribute to the development of HAIs and these differ among the specific infections our use of a composite rate may have obscured important racial and ethnic differences in relation to specific infection types The total number of several individual HAI types was small so comparisons based on the 6 types of individual HAI rates must be made with caution Finally our data set includes only hospital admissions for acute cardiovascular disease pneumonia and major surgery as defined by SCIP so the results are not representative of all-cause admissions

Our study has strengths worth highlighting First it was

based on a very large patient sample derived from randomly selected hospital admissions across the country and thus provides a more complete picture than similar studies based on a limited number of hospitals or a limited geographic region Furthermore unlike other data sets such as the Nashytional Healthcare Safety Network3 our data were obtained from chart abstraction thereby avoiding the potential bias associated with self-reported data

In summary we have shown that in a large national sample of patients hospitalized for acute myocardial infarction heart failure pneumonia and conditions requiring surgery Hisshypanic and Asian patients had higher risk-adjusted rates of HAIs than whites There was a nonsignificant trend toward an increased risk among black patients The primary cause for the differences seen was elevated CAUTI rates The higher HAI rates in Asian and Hispanic patients suggest the possishybility that language barriers between patients and healthcare providers could contribute to the occurrence of HAIs either directly or indirectly Improved availability and more frequent use of professional translation services might be a new opshyportunity for intervention to lower HAI rates Our finding of higher CAUTI rates in Hispanic and Asian patients in particular warrants further investigation as to the specific causes

A C K N O W L E D G M E N T S

We thank all the previous and current Medicare Patient Safety Monitoring System team members for their contributions to this work

Financial support This work was supported by the Agency for Healthcare Research and Quality US Department of Health and Human Services Rock-ville Maryland (contract HHSA290201200003C) Qualidigm was the conshytractor The content does not necessarily represent the official views or polshyicies of the Department of Health and Human Services nor does mention of trade names commercial products or organizations imply endorsement by the US government The authors assume full responsibility for the accuracy and completeness of the ideas

Potential conflicts of interest MLM reports having worked on various quality improvement and patient safety projects with Qualidigm the Centers for Medicare amp Medicaid Services and the Agency for Healthcare Research and Quality His employer has received remuneration for this work The views expressed in this article represent the authors and not their respective institutional affiliations All other authors report no conflicts of interest relevant to this article All authors submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and the conflicts that the editors consider relevant to this article are disclosed here

Address correspondence to Anila Bakullari BS Qualidigm 1290 Silas Deane Highway Suite 4A Wethersfield CT 06109 (abakullariqualidigmorg)

R E F E R E N C E S

1 US Census Bureau Newsroom US Census Bureau Projections Show a Slower Growing Older More Diverse Nation a Half Censhytury from Now December 12 2012 httpwwwcensusgov newsroomreleasesarchivespopulationcb 12-243html Acshycessed January 15 2014

2 Ennis SR Rios-Vargas M Albert NG 2010 Census Briefs The Hispanic Population 2010 May 2012 httpwwwcensusgov prodcen2010briefsc2010br-04pdf Accessed October 24 2013

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use

S l 6 INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY OCTOBER 2 0 1 4 VOL 3 5 NO S3

3 Hoeffel EM Rastogi S Ouk Kim M Shahid H 2010 Census Briefs The Asian Population 2010 March 2012 httpwww censusgovprodcen2010briefsc2010br-llpdf Accessed Ocshytober 24 2013

4 Institute of Medicine Unequal Treatment What Healthcare Proshyviders Need to Know about Racial and Ethnic Disparities in Healthcare Washington DC National Academy Press 2002

5 Thorpe RJ Richard P Bowie JV LaVeist TA Gaskin DJ Ecoshynomic burden of mens health disparities in the United States Int J Mens Health 201312195-212

6 Flores G Ngui E Racialethnic disparities and patient safety Pediatr Clin N A 2006531197-1215

7 Coffey RM Andrews RM Moy E Racial ethnic and socioecoshynomic disparities in estimates of AHRQ patient safety indicators Med Care 200543148-157

8 Romano PS Geppert JJ Davies S Miller MR Elixhauser A McDonald KM A national profile of patient safety in US hosshypitals Health Aff (Millwood) 200322154-166

9 Metersky ML Hunt DR Kliman R et al Racial disparities in the frequency of patient safety events results from the National Medicare Patient Safety Monitoring System Med Care 201149 504-510

10 Hunt DR Verzier N Abend SL et al Fundamentals of Medicare Patient Safety Surveillance Intent Relevance and Transparency AHRQs compendium of patient safety research advances in patient safety from research to implementation httpwww ahrqgovdownloadspubadvancesvol2Huntpdf Accessed December 15 2013

11 Wang Y Eldridge N Metersky ML et al National trends in patient safety for four common conditions 2005-2011 N Engl J Med 2014370341-351

12 Stulberg JJ Delaney CP Neuhauser DV Aron DC Fu P Ko-roukian SM Adherence to Surgical Care Improvement Project measures and the association with postoperative infections JAMA 20103032479-2485

13 Centers for Disease Control and Prevention National Healthshycare Safety Network (NHSN) September 24 2013 http wwwcdcgovnhsn Accessed February 7 2014

14 Agency for Healthcare Research and Quality AHRQuality Indicators Patient Safety Indicators Overview httpwww-

qualityindicatorsahrqgovModulespsi_resourcesaspx Acshycessed February 14 2014

15 Freedberg DE Salmasian H Friedman C Ahrams JA Proton pump inhibitors and risk for recurrent Clostridium difficile inshyfection among inpatients Am J Gastroenterol 2013 1081794-1801

16 Murphy CR Avery TR Dubberke ER Huang SS Frequent hosshypital readmissions for Clostridium difficile infection and the imshypact on estimates of hospital-associated C difficileburden Infect Control Hosp Epidemiol 20123320-28

17 Klevens RM Morrison MA Nadle J et al Invasive methicillin-resistant Staphylococcus aureus infections in the United States JAMA 20072981763-1771

18 Dick A Liu H Zwanziger J et al Long-term survival and healthshycare utilization outcomes attributable to sepsis and pneumonia BMC Health Serv Res 201212432

19 Chenoweth C Saint S Preventing catheter-associated urinary tract infections in the intensive care unit Crit Care Clin 2013 2919-32

20 US Census Bureau Educational Attainment Educational Atshytainment in the United States 2013mdashDetailed Tables January 1 2013 httpwwwcensusgovhhessocdemoeducationdata cps2013tableshtml Accessed January 15 2014

21 US Census Bureau The 2012 Statistical Abstract Income Expenditures Poverty and Wealth June 27 2012 http wwwcensusgovcompendiastatabcatsincome_expenditures _poverty_wealthhtml Accessed February 7 2014

22 Ryan C American Community Survey Reports Language Use in the United States 2011 August 2013 httpswwwcensusgov prod2013pubsacs-22pdf Accessed January 15 2014

23 Lion KC Rafton SA Shafii J et al Association between language serious adverse events and length of stay among hospitalized children Hosp Pediatr 20133219-225

24 Divi C Koss RK Schmaltz SP Loeb JM Language proficiency and adverse events in US hospitals a pilot study Int J Qual Health Care 200719(2)60-67

25 DeCamp LR Kuo DZ Flores G OConnor K Minkovitz CS Changes in language services use by US pediatricians Pediatrics 2013132(2)e396-e406

Downloaded from httpswwwcambridgeorgcore 04 Nov 2020 at 010136 subject to the Cambridge Core terms of use