JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

Journal of Infection and Public Health (2014) xxx, xxx—xxx1

The impact of the InternationalNosocomial Infection Control Consortium(INICC) multicenter, multidimensionalhand hygiene approach in two cities ofIndia

2

3

4

5

6

a b

c,∗ d a

b d a b

d a b

d b d

b d d

d

7

8

9

10

11

12

13

a Fortis Hospitals, Bangalore, India14

b Tata Memorial Hospital, Mumbai, India15

c International Nosocomial Infection Control Consortium (INICC), Buenos Aires,1

16

17

d Breach Candy Hospital Trust, Mumbai, India18

Received 30 May 2014; received in revised form 15 July 2014; accepted 24 August 201419

KEYWORDSCare;Developing countries;Hand hygiene;Hand washing;Healthcare workers;India;

Summary The fundamental tool for preventing and controlling healthcare-acquired infections is hand hygiene (HH). Nonetheless, adherence to HH guidelinesis often low. Our goal was to assess the effect of the International Nosocomial Infec-tion Control Consortium (INICC) Multidimensional Hand Hygiene Approach (IMHHA)in three intensive care units of three INICC member hospitals in two cities of Indiaand to analyze the predictors of compliance with HH. From August 2004 to July2011, we carried out an observational, prospective, interventional study to eval-uate the implementation of the IMHHA, which included the following elements:

21

22

23

24

25

26

27

∗ Corresponding author at: International Nosocomial Infection Control Consortium (INICC), Corrientes Ave. # 4580, Floor 12, Apt D,Buenos Aires 1195, Argentina. Tel.: +54 11 4861 5826; fax: +54 11 4861 5826.

E-mail addresses:1 URL: www.INICC.org.

http://dx.doi.org/10.1016/j.jiph.2014.08.0041876-0341/© 2014 Published by Elsevier Limited on behalf of King Saud Bin Abdulaziz University for Health Sciences. All rightsreserved.

Murali Chakravarthy , Sheila Nainan Myatra ,Victor D. Rosenthal , F.E. Udwadia , B.N. Gokul ,J.V. Divatia , Aruna Poojary , R. Sukanya , Rohini Kelkar ,Geeta Koppikar , Leema Pushparaj , Sanjay Biswas ,Lata Bhandarkar , Sandhya Raut , Shital Jadhav ,Sulochana Sampat , Neeraj Chavan , Shweta Bahirune ,Shilpa Durgad

Argentina

victor rosenthal@inicc.org, V.D. Rosenthal).

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

2 M. Chakravarthy et al.

(1) administrative support, (2) supplies availability, (3) education and training, (4)reminders in the workplace, (5) process surveillance and (6) performance feedback.The practices of health care workers were monitored during randomly selected 30-min

from 36.9% to 82% (95% CI 79.3—84.5; P = 0.0001). Multivariate analysis indicated thatcertain variables were significantly associated with poor HH adherence: nurses vs.physicians (70.5% vs. 74%; 95% CI 0.62—0.96; P = 0.018), ancillary staff vs. physicians(43.6% vs. 74.0%; 95% CI 0.48—0.72; P < 0.001), ancillary staff vs. nurses (43.6% vs.70.5%; 95% CI 0.51—0.75; P < 0.001) and private vs. academic hospitals (74.2% vs.66.3%; 95% CI 0.83—0.97; P <0.001). It is worth noticing that in India, the HH compli-ance of physicians is higher than in nurses. Adherence to HH was significantly increasedby implementing the IMHHA. Programs targeted at improving HH are warranted toidentify predictors of poor compliance.

28

29

30

31

32

33

34

35

36

37

38

39

40

Infection control;Intensive care units;International NosocomialInfection ControlConsortium;Multidimensionalapproach

© 2014 Published by Elsevier Limitedon behalf of King Saud Bin Abdulaziz University for Health Sciences. All rights reserved.

41

42

43

Healthcare-associated infection (HAIs) pose seri-45

ous threats to patient safety, including morbidity46

and mortality [1]. Within the mainstream litera-47

ture, the majority of studies on HAIs have been48

carried in high-income countries [2], whereas this49

problem has not been studied thoroughly in limited-50

resource countries. To counteract this, since 2002,51

the International Nosocomial Infection Control Con-52

sortium (INICC) has addressed the burden of HAIs53

by applying standardized definitions and meth-54

ods to measure and analyze HAI rates worldwide55

[3—7].56

Over a century ago, it was demonstrated that57

appropriate hand hygiene (HH) before patient con-58

tact was an essential intervention to prevent the59

transmission of cross-infections by health care60

workers (HCWs) [8]. It has been widely shown that61

an increase in the adherence to HH was related to62

the reduction of bacterial resistance and lower HAI63

rates [9—11].64

Successful interventions to improve HH have65

been analyzed in studies both from developed66

and developing countries [10,12—14]. Investigators67

have assessed the effectiveness of interventions to68

improve HH since the 1980s [15—17]. Since 1993,69

Rosenthal et al. have implemented multimodal pro-70

grams in Argentina combining administrative sup-71

port, supplies availability, education and training,72

process surveillance and performance feedback,73

which produced sustained increases in HH compli-74

ance [14], and associated reductions in the rates of75

HAI [10].76

The HH guidelines were published by the US Cen-77

ters for Disease Control and Prevention (CDC) in78

(WHO) in 2005 as part of the ‘‘Clean Care is Safer80

Care’’ campaign [19], and in 2009, by presenting a 81

compilation of previously published data, and a new 82

formulation for alcohol-based hand rub products, 83

among other recommendations [11]. 84

This is the first multisite study conducted in 85

India with the aim of determining the baseline rate 86

of adherence to HH by HCWs before patient con- 87

tact, analyzing risk factors for poor adherence and 88

assessing the impact of an INICC Multidimensional 89

HH Approach (IMHHA) in 3 hospitals from 2 cities. 90

The IMHHA includes the following components: 91

(1) administrative support, (2) supplies availabil- 92

ity, (3) education and training, (4) reminders in 93

the workplace, (5) process surveillance and (6) 94

performance feedback. 95

96

Study design 97

From August 2004 through July 2011, we carried 98

out an observational, prospective, interventional, 99

before-and-after multisite cohort study, which was 100

divided into two periods: a baseline period and 101

a follow-up period. The baseline period included 102

opportunities registered at each hospital during 103

their first 3 months of participation, and the follow- 104

up period included opportunities documented after 105

the fourth month of participation. Each hospital 106

started to participate in the study at different 107

times, and therefore, they have different lengths of 108

follow-up (from 4 to 36 months); but for all ICUs the 109

length of the baseline period is exactly the same (3 110

months). For comparing the rate of HH adherence, 111

the ICUs were aligned independently of the date at 112

which they started their participation in the study 113

over the 7-year period. 114

periods. We observed 3612 opportunities for HH. Overall adherence to HH increased

Materials and methods

44 Introduction

79 2002 [18], and by the World Health Organization

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

Impact of the INICC multicenter, multidimensional hand hygiene approach 3

Study setting115

This study was conducted in 3 ICUs of 3 INICC mem-116

ber hospitals from 2 cities in India, which were117

incorporated into the study over the 7-year study118

period. Each hospital has an infection control team119

(ICT) with at least one infection control practitioner120

(ICP) and one physician. The ICT member in charge121

of process surveillance at each hospital has a mini-122

mum of two years of experience in infection control123

practices and surveillance of HAI rates. Professional124

categories of HCWs included physicians, nurses,125

and ancillary staff (paramedical technicians, nurse126

aides, patient care technicians, laboratory team127

members, radiology team members, physiothera-128

pists, patient lift teams, and other paramedical129

personnel.)130

Background on INICC131

The INICC is an international, non-profit, open,132

multi-center HAI surveillance network with133

methodology based on the U.S. CDC/National134

Healthcare Safety Network (NHSN) [20]. The135

INICC is the first research network established136

internationally to measure, control and prevent137

HAIs in hospitals worldwide through the analysis138

of standardized data, which are collected on a139

voluntary basis by its member hospitals. Gaining140

new members since its international inception in141

2002, the INICC is now comprised by nearly 1000142

hospitals in 300 cities of 60 countries in Latin143

America, Asia, Africa, Middle East, and Europe,144

and has become the only source of aggregate stan-145

dardized international data on the epidemiology of146

HAIs worldwide [7].147

The IMHHA was implemented at the participat-150

ing ICUs when they began participation in the151

INICC program. The IMHHA includes 6 elements:152

(1) administrative support; (2) supplies availabil-153

ity; (3) education and training; (4) reminders in154

the workplace; (5) process surveillance and (6) per-155

formance feedback. For the purposes of analysis,156

we present the elements of the IMHHA individually.157

Nevertheless, it is worth noting that for an effective158

implementation of the approach, the 6 components159

need to interact simultaneously.160

Administrative support161

Hospital administrators agreed and committed162

to supporting the development of the IMHHA,163

attended infection control meetings on a monthly164

basis to discuss study findings, and allocated sup- 165

plies of HH products. 166

Supplies availability 167

Supplies were made available by placing alcohol- 168

based hand rub bottles in nursing stations, at the 169

ICUs’ and individual patient room entrances, and 170

near the site of patient care (at bedside tables 171

and/or on the feet of patient beds). Soap, paper 172

towels and sink water were supplied at the ICUs’ 173

entrances, nursing stations, and common areas in 174

the ICUs. 175

Education and training 176

At each ICU, the investigators of the ICT pro- 177

vided 30-min education sessions on HH practices 178

to HCWs in each work shift, at the beginning of 179

the study period and periodically (every month, 180

every 2 months, and every 6 months, in the respec- 181

tive 3 ICUs) during the follow-up period. Sessions 182

included the provision of information about the cor- 183

rect opportunities and techniques for HH. 184

Reminders in the workplace 185

Posters reminding employees of HH techniques and 186

opportunities were displayed around the hospi- 187

tal settings (i.e., hospital entrance, corridors, ICT 188

office and entrances, nursing stations, and beside 189

each alcohol-based hand rub bottle). Reminders 190

included simple instructions on HH practice in 191

accordance with the contents of the education and 192

training sessions. 193

Process surveillance 194

Process surveillance of HH practices consisted of 195

recording the potential HH opportunities, and the 196

number of HH episodes observed, both with water 197

and soap, or with alcohol-based hand rub prod- 198

ucts. HCWs’ HH practices were directly monitored 199

by a member of the ICT, who had received train- 200

ing sessions from a reporting manual, and who was 201

not an observed HCW [3,11]. To improve the inter- 202

reliability of the data, observers used standardized 203

monitoring processes, following a protocol and 204

completing standardized HH surveillance forms that 205

contained a uniform questionnaire for monitoring 206

HH practices [3]. The ICT member conducted unob- 207

trusive observations (that is, without interference 208

from the observer) at specific time periods selected 209

at random, distributed three times a week, dur- 210

ing 30 min each time and during all work shifts 211

(morning, afternoon and evening). HCWs were not 212

aware of the schedule of the monitoring period 213

by the ITC. The monitoring included HH compli- 214

ance before patient contact, and before an aseptic 215

task, because we started the study in August 2004, 216

148 The INICC Multidimensional Hand Hygiene149 Approach (IMHHA)

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

4 M. Chakravarthy et al.

and the ‘‘Five moments for HH’’ proposed by the217

WHO was not published until 2009. The potential218

confounders of HH included the type of ICU, the219

professional category, sex, the work shift, and the220

type of contact.221

Performance feedback222

On a monthly basis, the INICC Headquarters in223

Buenos Aires prepared and sent each ICU a final224

month-by-month report on compliance with HH.225

These reports displayed charts and contained a run-226

ning tally of HH compliance by HCWs in the ICUs,227

and compliance comparing several variables, such228

as sex, HCWs’ professional status, ICU type, con-229

tact type, and work shift. The results of the reports230

were reviewed every month at ICT meetings and231

the charts were posted in the participating ICUs to232

provide performance feedback to the HCWs work-233

ing in them [3]. Performance feedback started in234

the third month of participation in this approach235

[3].236

Training for process surveillance237

The ICT investigators were self-trained with a238

procedure manual designed by the INICC, which239

specified how to conduct HH process surveillance240

and how to fill in the forms [3]. ICT members had241

continuous telephone, email, and webinar access242

to a support team at the INICC Headquarters.243

Data collection and processing244

Every month, the ICT members from each partici-245

pating ICU completed INICC process surveillance HH246

forms and sent them to the INICC Headquarters in247

Buenos Aires, where the data were uploaded into248

a database and analyzed. Next, the members of249

the ICT at each participating ICU received a report250

on the HH compliance, showing compliance rates251

stratified by month, sex, HCWs’ professional sta-252

tus, the ICU, the ICU work shift, and the type of253

contact [3].254

Statistical methods255

Univariate analysis of variables associated with256

poor hand hygiene, and the impact of the INICC257

multidimensional hand hygiene approach258

The aggregated independent variables (type of hos-259

pital, sex of the HCWs, profession of the HCWs, the260

type of ICU, and the type of contact) of all of the261

observed HH opportunities and HH compliance dur-262

ing the study, and the comparison of HH compliance263

during the baseline period and during the follow-264

up period were compared using the Fisher’s exact265

test for dichotomous variables and the unmatched 266

Student’s t-test for continuous variables. 95% confi- 267

dence intervals (CI) were calculated using Stata 11 268

(StataCorp LP, 4905 Lakeway Drive, College Station, 269

TX, US). Relative risk (RR) ratios were calculated for 270

comparisons of the analyzed variables associated 271

with HH using EPI InfoTM V6 (Centers for Disease 272

Control and Prevention, 1600 Clifton Road Atlanta, 273

GA., US). P-values <0.05 by two-tailed tests were 274

considered significant. 275

Multivariate analysis of the variables associated 276

with poor hand hygiene 277

The aggregated described independent variables 278

of all observed HH opportunities and HH compli- 279

ance during the study were compared using logistic 280

regression for the dichotomous and continuous vari- 281

ables. Odds ratios (ORs) and their corresponding 282

95% CI were calculated for comparisons of the 283

analyzed variables associated with HH using PASW 284

Statistics 18. P-values <0.05 by two-tailed tests 285

were considered significant. 286

Multivariate analysis of the impact of the INICC 287

multidimensional hand hygiene approach 288

HH opportunities and HH compliance during base- 289

line and during follow-up were analyzed for changes 290

in HH compliance rates following an ICU joining the 291

INICC. We looked at the follow-up periods strati- 292

fied by three-month periods over the first year, and 293

yearly following the second year of participation. 294

We present the results of a logistic regression model 295

to describe the changes in HH compliance in INICC 296

participating ICUs over time since the beginning 297

of the HH surveillance. Odds ratios are presented, 298

comparing each time period since the start of the 299

surveillance with the baseline of 3 months. This is 300

we were able to adjust for the effect of each ICU 302

on HH compliance as a categorical variable in the 303

analysis. Because of the different length of follow- 304

up at each ICU (from 9 months to 3 years), for each 305

time period only ICUs with follow-up in that time 306

period were included in the baseline period used 307

for calculating the OR of HH compliance for that 308

period. 309

From August 2004 to July 2011 (7 years), we 311

recorded a total 3612 opportunities for HH before 312

patient contact, and before aseptic task. Char- 313

acteristics of participating hospitals are shown in 314

Table 1. 315

a large data set, with 3612 observations, and so 301

Results 310

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

Impact of the INICC multicenter, multidimensional hand hygiene approach 5

Table 1 The characteristics of the participating hospitals (from August 2004 to July 2011).

Hospital type ICU type Observations of opportunitiesfor HH in ICUs, n

Hospital 1 — academic teaching Medical surgical 2060Hospital 2 — private Surgical 310Hospital 3 — private Medical surgical 1242All hospitals 3612

ICU, intensive care unit; HH, hand hygiene.

Table 2 The distribution of hand hygiene compliance by the type of intensive care unit.

ICUs (n) Baseline period (HHcompliance/HHobservations)

Intervention period (HHcompliance/HHobservations)

RR (95% CI) P value

Medical 1 64.9% (72/111) 90.9% (1023/1131) 1.40 (1.1—1.8) 0.0039Medical surgical 1 40.1% (87/217) 69.4% (1279/1843) 1.73 (1.4—2.2) 0.001Surgical 1 13.5% (22/163) 19.7% (29/147) 1.46 (0.81—2.7) 0.1811All 3 36.9% (181/491) 74.8% (2336/3121) 2.0 (1.7—2.4) 0.0001

ICU, intensive care unit; HH, hand hygiene; CI, confidence interval.

Predictors of poor hand hygiene compliance316

We observed 2060 opportunities in academic317

hospitals and 1552 in private hospitals; 1084 oppor-318

tunities in males, and 2528 in females; 2154 in319

nurses, 1194 in physicians, and 264 in ancillary320

staff; 2127 were prior to non-invasive patient con-321

tacts, and 822 were prior to invasive procedures;322

1440 during the morning, 1139 during the after-323

noon, and 1033 during the night shift.324

Table 2 shows HH compliance distribution among325

the different ICU types in the baseline and inter-326

vention period.327

In Table 3, we present the HH compliance rates328

for the whole study period comparing each variable329

(sex, HCW professional status, type of procedure,330

type of ICU, and work shift). Their associations with331

poor HH were analyzed with univariate and multi-332

variate statistical methods as also shown in Table 3.333

Components of the INICC multidimensional334

hand hygiene approach335

During the follow-up period, the 6 components of336

the IMHHA were applied simultaneously in each337

ICU. All the aspects of the IMHHA were followed338

by all 3 ICUs, with the exception of the posting of339

reminders, which were only posted in 2 ICUs, at the340

ICU entrance and in common ICU areas. All ICUs341

counted on administrative support and available342

supplies for HH and alcohol-based hand rub prod-343

ucts. Process surveillance was conducted at the 3344

ICUs. All HCWs working at the 3 ICUs were provided345

with performance feedback and attended training346

sessions at regular intervals: 1 ICU on a monthly 347

basis: 1 ICU every 2 months, and 1 ICU every 6 348

months. 349

The impact of the INICC multidimensional 350

hand hygiene approach on hand hygiene 351

compliance 352

In Table 4, we present the results of a regression 353

model to describe the changes in HH compliance in 354

the INICC participating ICUs over the whole study 355

period. The baseline period of the INICC ICUs was 356

3 months, and their average follow-up period was 357

17.2 months (range 4—52). 358

359

Over the last decade, the INICC has constantly 360

struggled to reduce the burden of HAIs in Latin 361

America, Asia, Africa, Middle East, and Europe, 362

achieving successful results. Through the INICC 363

multimodal programs, compliance with infection 364

control measures and tools has been increased sub- 365

stantially, thereby reducing the HAI rates and their 366

adverse effects, such as mortality, as shown in many 367

scientific publications [10]. Since 2002, in adult 368

ICUs in 15 countries, the INICC has reduced the 369

rate of central line-associated bloodstream infec- 370

tion by 54% [21], of catheter-associated urinary 371

tract infection by 37% [22], of ventilator-associated 372

pneumonia by 56% [23], and of mortality by 58% 373

[21]. In pediatric ICUs from 5 countries, the INICC 374

Discussion

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

6 M. Chakravarthy et al.

Tabl

e3

Han

dhy

gien

eco

mpl

ianc

eby

type

ofva

riab

leus

ing

univ

aria

tean

alys

is,

logi

stic

regr

essi

on,

and

mul

tiva

riat

ean

alys

is.

Vari

able

%(#

HH

/#op

port

unit

ies)

Com

pari

son

RR95

%CI

Pva

lue

Adju

sted

OR

95%

CIP

valu

eTy

peof

hosp

ital

(bas

elin

e:pr

ivat

e)Pr

ivat

e74

.2%

(115

1/15

52)

1Ac

adem

ic66

.3%

(136

6/20

60)

Acvs

.Pr

0.89

0.83

—0.

970.

0052

0.49

0.42

—0.

600.

001

Sex

(bas

elin

efe

mal

e)Fe

mal

e71

.2%

(180

1/25

28)

Fvs

.M

0.93

0.85

—1.

010.

0858

1M

ale

66.1

%(7

16/1

084)

0.86

0.70

—1.

070.

186

Type

ofpr

ofes

sion

al(b

asel

ine:

phys

icia

ns)

Phys

icia

ns74

.0%

(883

/119

4)Ph

vs.

AS0.

590.

48—

0.72

0.00

011

Nur

ses

70.5

%(1

519/

2154

)N

svs

.Ph

0.95

0.88

—1.

040.

2616

0.77

0.62

—0.

960.

018

Anci

llary

staf

f43

.6%

(115

/264

)N

svs

.AS

0.62

0.51

—0.

750.

0001

0.23

0.16

—0.

310.

001

Type

ofco

ntac

t(b

asel

ine:

inva

sive

)In

vasi

ve82

.8%

(681

/822

)N

Ivs.

I0.

950.

87—

1.04

0.29

351

Non

-inv

asiv

e79

.0%

(168

0/21

27)

1.02

0.82

—1.

30.

850

Wor

ksh

ift

(bas

elin

e:ni

ght)

Afte

rnoo

n71

.6%

(815

/113

9)M

vs.

A0.

940.

85—

1.03

0.18

921.

0M

orni

ng67

.2%

(968

/144

0)M

vs.

N0.

950.

86—

1.05

0.25

740.

960.

79—

1.20

0.72

3N

ight

71.1

%(7

34/1

033)

Avs

.N

0.99

0.90

—1.

100.

8910

0.92

0.74

—1.

10.

447

ICU

,in

tens

ive

care

unit

;O

R,od

dsra

tio;

CI,

confi

denc

ein

terv

al;

AS,

anci

llary

staf

f;F,

fem

ale;

M,

mal

e;N

i,no

n-in

vasi

ve;

I,in

vasi

ve;

M,

mor

ning

wor

ksh

ift;

A,af

tern

oon

wor

ksh

ift;

N,

nigh

tw

ork

shif

t;N

S,nu

rsin

gst

aff;

Ph,

phys

icia

ns;

AS,

anci

llary

staf

f

has reduced the rate of central line-associated 375

bloodstream infection by 52% [24], of catheter- 376

associated urinary tract infection by 57% [25]of ventilator-associated pneumonia by 31% [26]and of mortality by 31% [24]. In neonatal ICUs in 379

10 countries, the INICC has reduced the rate of 380

ventilator-associated pneumonia by 33% [27]. 381

The baseline percentage of adherence to HH by 382

HCWs at our ICUs (37%) was within the wide and 383

variable range of percentages of compliance with 384

HH reported in previous studies, which vary from 9% 385

to 75% [11]. This is the first study that has showed 386

an increase in HH compliance in India as a result of 387

the implementation of the IMHHA. 388

However, there are some limitations in our study 389

that need to be addressed before describing and 390

explaining our findings. We did not measure the 391

opportunities as specified in 2009 by the WHO in, 392

‘‘My five moments for HH,’’ because the INICC 393

started the IMHHA in 1998 in Argentina and in 2002 394

internationally; that is, several years before these 395

WHO recommendations were published. However, 396

since 2009, the INICC has included the WHO’s ‘‘My 397

five moments for HH’’ in its process surveillance 398

forms and manuals [3,10,11,14]. It should be noted 399

also that, due to our limited budget, we did not 400

include more details about the HH techniques 401

[28]. In addition, as we applied an observational, 402

before—after method, the evidence may have less 403

strength and accuracy than other study designs. 404

Directly observing adherence typically involves a 405

Hawthorne effect, and represents only a sample 406

of all opportunities and we cannot overtly assure 407

inter-observer reliability. Finally, an analysis of 408

the compliance by the intensity of education 409

would be very useful for future infection control 410

interventions. 411

The evaluation of the impact of the IMHHA in 412

the ICU populations from 2 cities of India showed 413

that the 6 measures of the IMHHA were followed 414

by important improvements in HH practices. The 415

results of the multivariate analysis showed that 416

there was higher compliance in private hospitals 417

than in public and academic ones. The relationship 418

between the type of hospital and HH compliance 419

has not been assessed in the literature [29]. Never- 420

theless, there is published evidence that in limited- 421

resource countries, such as India, ICUs in the pri- 422

vate sector can influence the outcome of such pro- 423

grams due to the wider availability of resources and 424

greater administrative support, in contrast to pub- 425

lic ICUs that experience overcrowding and under- 426

staffing, which have been shown to hinder HCWs’ 427

efforts to perform HH [30,31]. Adherence to HH was 428

not statistically significantly different by the HCWs’ 429

sex, which is in contrast with the findings of Guinan 430

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

Impact of the INICC multicenter, multidimensional hand hygiene approach 7

Table 4 Hand hygiene improvement by year of participation from July 2004 to July 2011.

Years since joining INICC HHobservations

Number ofICUs Included

HH % (95% CI) AdjustedOR (95% CI)

P value

First 3 months (baseline) 491 3 36.9% (32.5—41.3) 1.0Months 4—6 437 3 44.4% (39.7—49.2) 1.04 (0.77—1.4) 0.803Months 7—9 413 3 60.3% (55.4—65.1) 1.91 (1.4—2.6) 0.001Months 10—12 381 2 65.1% (60.1—69.9) 1.78 (1.3—2.5) 0.0012nd year 1211 2 89.8% (88.1—91.5) 9.9 (7.3—13.3) 0.0013rd year 679 2 82.0% (79.3—84.5) 7.3 (5.3—9.9) 0.001

INICC, International Nosocomial Infection Control Consortium; HH, hand hygiene; ICU, intensive care units; CI, confidence interval;OR, odds ratio.

et al., were higher adherence to HH was found in431

females not related to health care [32]; as well as in432

relation to the type of contact and work shift, which433

also contrasts with the previous findings, such as in434

the study by Lipsett et al., which showed that lower435

HH compliance was found in low-risk situations436

[33]. In regard to the type of professional, compli-437

ance was lower among ancillary staff, which is con-438

sistent with the findings of Rosenthal et al. in 2005439

in which compliance was lower among ancillary440

staff compared to nurses [10]. The most surpris-441

ing outcome found by this research is that in India,442

the HH compliance of physicians is higher than443

nurses: nurses vs. physicians (70.5% vs. 74%; 95% CI444

0.62—0.96; P = 0.018), which is in contrast with pre-445

vious findings in the mainstream literature [34,35].446

This can be explained by the strongest commitment447

to preventing and controlling HAIs in India by doc-448

tors, and the lack of emphasis on hand hygiene449

importance in the nursing curricula [36]. We found450

greater compliance among physicians, which is the 451

opposite of the findings from other studies [37—39]. 452

We think that the impact of the IMHHA is directly 453

related to its components. Regarding administra- 454

tive support, there is published evidence that 455

higher HH adherence was associated to the sup- 456

port of administrative authorities, as shown by 457

Rosenthal et al. [14] The IMHHA included supplies 458

availability. In 2000, Bischoff et al. [40] showed that 459

easily accessible dispensers of alcohol-based hand 460

rub products revealed that the more dispensers 461

per bed, the higher adherence to HH. The IMHHA 462

also included education and training sessions, which 463

were other basic independent interventions iden- 464

tified to foster adequate HH performance. As 465

described by Dubbert et al., the regularity of 466

educational sessions improved HH compliance by 467

97% over four weeks [41]. Likewise, but within 468

the context of limited-resource countries, Rosen- 469

thal et al. showed that educating HCWs increased 470

36.9%

44.4%

60.3%65.1%

89.8%

82.0%

0.0%

10.0%

20.0%

30.0%

40.0%

50.0%

60.0%

70.0%

80.0%

90.0%

100.0%

1-3 months 4-6 months 7-9 months 10-12 months 2nd year 3rd year

Figure 1 Hand hygiene improvement by year of participation.

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

8 M. Chakravarthy et al.

adherence to HH and that compliance percentages471

could be further improved if performance feed-472

back was also provided to HCWs [14]. The IMHHA473

also included reminders at the workplace, which474

has been highlighted as an important tool to raise475

HCWs’ awareness of the relation between cor-476

rect HH performance and the reduction of HAIs477

[42]. The ICT team provided performance feed-478

back to the HCWs in each ICU on a monthly basis.479

This is the most motivating aspect of the IMHHA480

for HCWs. Knowing the outcome of their efforts481

reflected by the measurement of their practices482

and HAI incidence can be a rewarding or conscious-483

raising factor to ensure the IMHHA’s effectiveness484

[43]. From 1998 in Argentina, and 2002 internation-485

ally, the INICC has introduced process surveillance486

and performance feedback to measure and improve487

quality in healthcare by monitoring and providing488

constant feedback to HCWs not only of outcome489

data — that is, the HAI rates — but also by inform-490

ing the HCWs about the results of the process491

surveillance—rates of HH compliance and other492

basic but highly effective, evidence-based infection493

control practices — and we show that combining494

education with surveillance feedback can substan-495

tially reduce the risk of HAIs in ICUs [3—7,10,14]496

(Fig. 1).497

Conclusions498

The primary goal of the INICC is to promote499

infection control practices by providing free500

resourceful tools to address the burden posed by501

HAIs effectively, thereby leading to greater and502

steady adherence to infection control programs and503

guidelines. As demonstrated, the IMHHA improved504

HH compliance in India, and as shown in other INICC505

publications, thus contributing to the reduction of506

HAIs and the consequences attributable to them507

[10,44,45].508

Funding509

The funding for the activities carried out at the510

INICC headquarters were provided by the cor-511

responding author, Victor D. Rosenthal, and the512

Foundation to Fight against Nosocomial Infections.513

Competing interests514

None declared.515

Ethical approval 516

Every hospital’s Institutional Review Board agreed 517

to the study protocol, and patient confidentiality 518

was protected by codifying the recorded informa- 519

tion, and making it only identifiable to the infection 520

control team. 521

Acknowledgments 522

The authors thank the many health care profession- 523

als at each member hospital who assisted with the 524

conduct of surveillance in their hospital, including 525

the surveillance nurses, clinical microbiology lab- 526

oratory personnel, and the physicians and nurses 527

providing care for the patients during the study; 528

without their cooperation and generous assistance 529

this INICC would not be possible; Mariano Vilar and 530

Débora López Burgardt, who work at the INICC 531

headquarters in Buenos Aires, for their hard work 532

and commitment to achieving the INICC goals; the 533

INICC Country Coordinators and Secretaries (Altaf 534

Ahmed, Carlos A. Álvarez-Moreno, Anucha Apisarn- 535

thanarak, Luis E. Cuéllar, Bijie Hu, Namita Jaggi, 536

Hakan Leblebicioglu, Montri Luxsuwong, Eduardo A. 537

Medeiros, Yatin Mehta, Ziad Memish, and Lul Raka); 538

and the INICC Advisory Board (Carla J. Alvarado, 539

Nicholas Graves, William R. Jarvis, Patricia Lynch, 540

Dennis Maki, Gerald McDonnell, Toshihiro Mitsuda, 541

Cat Murphy, Russell N. Olmsted, Didier Pittet, 542

William Rutala, Syed Sattar, and Wing Hong Seto), 543

who have so generously supported this unique inter- 544

national infection control network. 545

References 546

[1] Jarvis WR. Selected aspects of the socioeconomic impact 547

of nosocomial infections: morbidity, mortality, cost, and 548

prevention. Infect Control Hosp Epidemiol 1996;17(8): 549

552—7. 550

[2] Safdar N, Crnich CJ, Maki DG. Nosocomial infections in 551

the intensive care unit associated with invasive medical 552

devices. Curr Infect Dis Rep 2001;3(6):487—95. 553

[3] Rosenthal VD, Maki DG, Graves N. The International Noso- 554

comial Infection Control Consortium (INICC): goals and 555

objectives, description of surveillance methods, and oper- 556

ational activities. Am J Infect Control 2008;36(November 557

(9)):e1—12. 558

[4] Rosenthal VD, Maki DG, Salomao R, Moreno CA, Mehta Y, 559

Higuera F, et al. Device-associated nosocomial infections in 560

55 intensive care units of 8 developing countries. Ann Intern 561

Med 2006;145(October (8)):582—91. 562

[5] Rosenthal VD, Maki DG, Mehta A, Alvarez-Moreno C, 563

Leblebicioglu H, Higuera F, et al. International Nosoco- 564

mial Infection Control Consortium report, data summary 565

for 2002—2007, issued January 2008. Am J Infect Control 566

2008;36(November (9)):627—37. 567

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

Impact of the INICC multicenter, multidimensional hand hygiene approach 9

[6] Rosenthal VD, Maki DG, Jamulitrat S, Medeiros EA, Todi SK,568

Gomez DY, et al. International Nosocomial Infection Control569

Consortium (INICC) report, data summary for 2003—2008,570

issued June 2009. Am J Infect Control 2010;38(March571

(2)):95—104, e2.572

[7] Rosenthal VD, Bijie H, Maki DG, Mehta Y, Apisarnthanarak573

A, Medeiros EA, et al. International Nosocomial Infection574

Control Consortium (INICC) report, data summary of 36575

countries, for 2004—2009. Am J Infect Control 2012;40(June576

(5)):396—407.577

[8] Raju TN. Ignac Semmelweis and the etiology of fetal and578

neonatal sepsis. J Perinatol 1999;19(4):307—10.579

[9] Simmons B, Bryant J, Neiman K, Spencer L, Arheart K.580

The role of handwashing in prevention of endemic inten-581

sive care unit infections. Infect Control Hosp Epidemiol582

1990;11(November (11)):589—94.583

[10] Rosenthal VD, Guzman S, Safdar N. Reduction in nosoco-584

mial infection with improved hand hygiene in intensive care585

units of a tertiary care hospital in Argentina. Am J Infect586

Control 2005;33(September (7)):392—7.587

[11] Pittet D, Allegranzi B, Boyce J. The World Health Organiza-588

tion Guidelines on Hand Hygiene in Health Care and their589

consensus recommendations. Infect Control Hosp Epidemiol590

2009;30(July (7)):611—22.591

[12] Lam BC, Lee J, Lau YL. Hand hygiene practices in a neonatal592

intensive care unit: a multimodal intervention and impact593

on nosocomial infection. Pediatrics 2004;114(November594

(5)):e565—71.595

[13] Allegranzi B, Sax H, Bengaly L, Richet H, Minta DK, Chraiti596

MN, et al. Successful implementation of the World Health597

Organization hand hygiene improvement strategy in a refer-598

ral hospital in Mali, Africa. Infect Control Hospital Epidemiol599

2010;31(February (2)):133—41.600

[14] Rosenthal VD, McCormick RD, Guzman S, Villamayor C, Orel-601

lano PW. Effect of education and performance feedback602

on handwashing: the benefit of administrative support in603

Argentinean hospitals. Am J Infect Control 2003;31(April604

(2)):85—92.605

[15] Preston GA, Larson EL, Stamm WE. The effect of private606

isolation rooms on patient care practices, colonization and607

infection in an intensive care unit. Am J Med 1981;70(March608

(3)):641—5.609

[16] Conly JM, Hill S, Ross J, Lertzman J, Louie TJ. Handwashing610

practices in an intensive care unit: the effects of an educa-611

tional program and its relationship to infection rates. Am J612

Infect Control 1989;17(December (6)):330—9.613

[17] Mayer JA, Dubbert PM, Miller M, Burkett PA, Chapman SW.614

Increasing handwashing in an intensive care unit. Infect615

Control 1986;7(May (5)):259—62.616

[18] Boyce JM, Pittet D. Guideline for Hand Hygiene in617

Health-Care Settings. Recommendations of the Healthcare618

Infection Control Practices Advisory Committee and the619

HICPAC/SHEA/APIC/IDSA Hand Hygiene Task Force. Soci-620

ety for Healthcare Epidemiology of America/Association for621

Professionals in Infection Control/Infectious Diseases Soci-622

ety of America. MMWR Recommend Rep 2002;51(October623

(RR-16)):1—45, quiz CE1-4.624

[19] WHO launches global patient safety challenge; issues guide-625

lines on hand hygiene in health, care. Indian J Med Sci626

2005;59(October (10)):461—3.627

[20] Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance628

definition of health care-associated infection and criteria629

for specific types of infections in the acute care setting.630

Am J Infect Control 2008;36(June (5)):309—32.631

[21] Rosenthal VD, Maki DG, Rodrigues C, Alvarez-Moreno632

C, Leblebicioglu H, Sobreyra-Oropeza M, et al. Impact633

of International Nosocomial Infection Control Consortium634

(INICC) strategy on central line-associated bloodstream 635

infection rates in the intensive care units of 15 636

developing countries. Infect Control Hosp Epidemiol 637

2010;31(December (12)):1264—72. 638

[22] Rosenthal VD, Todi SK, Alvarez-Moreno C, Pawar M, Karlekar 639

A, Zeggwagh AA, et al. Impact of a multidimensional infec- 640

tion control strategy on catheter-associated urinary tract 641

infection rates in the adult intensive care units of 15 devel- 642

oping countries: findings of the International Nosocomial 643

Infection Control Consortium (INICC). Infection 2012. 644

[23] Rosenthal VD, Rodrigues C, Alvarez-Moreno C, Madani N, 645

Mitrev Z, Ye G, et al. Effectiveness of a multidimensional 646

approach for prevention of ventilator-associated pneu- 647

monia in adult intensive care units from 14 developing 648

countries of four continents: findings of the International 649

Nosocomial Infection Control Consortium. Crit Care Med 650

2012;40(December (12)):3121—8. 651

[24] Rosenthal VD, Ramachandran B, Villamil-Gomez W, Armas- 652

Ruiz A, Navoa-Ng JA, Matta-Cortes L, et al. Impact 653

of a multidimensional infection control strategy on 654

central line-associated bloodstream infection rates in pedi- 655

atric intensive care units of five developing countries: 656

findings of the International Nosocomial Infection Con- 657

trol Consortium (INICC). Infection 2012;40(August (4)): 658

415—23. 659

[25] Rosenthal VD, Ramachandran B, Duenas L, Alvarez-Moreno 660

C, Navoa-Ng JA, Armas-Ruiz A, et al. Findings of the Inter- 661

national Nosocomial Infection Control Consortium (INICC), 662

Part I: effectiveness of a multidimensional infection con- 663

trol approach on catheter-associated urinary tract infection 664

rates in pediatric intensive care units of 6 develop- 665

ing countries. Infect Control Hosp Epidemiol 2012;33(July 666

(7)):696—703. 667

[26] Rosenthal VD, Alvarez-Moreno C, Villamil-Gomez W, Singh 668

S, Ramachandran B, Navoa-Ng JA, et al. Effectiveness of a 669

multidimensional approach to reduce ventilator-associated 670

pneumonia in pediatric intensive care units of 5 develop- 671

ing countries: International Nosocomial Infection Control 672

Consortium findings. Am J Infect Control 2012;40(August 673

(6)):497—501. 674

[27] Rosenthal VD, Rodriguez-Calderon ME, Rodriguez-Ferrer M, 675

Singhal T, Pawar M, Sobreyra-Oropeza M, et al. Findings of 676

the International Nosocomial Infection Control Consortium 677

(INICC), Part II: impact of a multidimensional strategy to 678

reduce ventilator-associated pneumonia in neonatal inten- 679

sive care units in 10 developing countries. Infect Control 680

Hosp Epidemiol 2012;33(July (7)):704—10. 681

[28] Mehta A, Rosenthal VD, Mehta Y, Chakravarthy M, Todi SK, 682

Sen N, et al. Device-associated nosocomial infection rates 683

in intensive care units of seven Indian cities. Findings of 684

the International Nosocomial Infection Control Consortium 685

(INICC). J Hosp Infect 2007;67(October (2)):168—74. 686

[29] Haley RW, Bregman DA. The role of understaffing and over- 687

crowding in recurrent outbreaks of staphylococcal infection 688

in a neonatal special-care unit. J Infect Dis 1982;145(June 689

(6)):875—85. 690

[30] Al-Wazzan B, Salmeen Y, Al-Amiri E, Abul A, Bouhaimed 691

M, Al-Taiar A. Hand hygiene practices among nursing 692

staff in public secondary care hospitals in Kuwait: self- 693

report and direct observation. Med Princ Pract 2011;20(4): 694

326—31. 695

[31] De Wandel D, Maes L, Labeau S, Vereecken C, Blot S. Behav- 696

ioral determinants of hand hygiene compliance in intensive 697

care units. Am J Crit Care 2010;19(May (3)):230—9. 698

[32] Guinan ME, McGuckin-Guinan M, Sevareid A. Who washes 699

hands after using the bathroom? Am J Infect Control 700

1997;25(5):424—5. 701

JIPH 368 1—10Please cite this article in press as: Chakravarthy M, et al. The impact of the International Nosocomial InfectionControl Consortium (INICC) multicenter, multidimensional hand hygiene approach in two cities of India. J InfectPublic Health (2014), http://dx.doi.org/10.1016/j.jiph.2014.08.004

ARTICLE IN PRESSJIPH 368 1—10

10 M. Chakravarthy et al.

[33] Lipsett PA, Swoboda SM. Handwashing compliance702

depends on professional status. Surg Infect (Larchmt)703

2001;2(3):241—5.704

[34] Pittet D, Simon A, Hugonnet S, Pessoa-Silva CL, Sauvan V,705

Perneger TV. Hand hygiene among physicians: performance,706

beliefs, and perceptions. Ann Intern Med 2004;141(July707

(1)):1—8.708

[35] Han K, Dou FM, Zhang LJ, Zhu BP. Compliance on hand-709

hygiene among healthcare providers working at secondary710

and tertiary general hospitals in Chengdu. Zhonghua Liu711

Xing Bing Xue Za Zhi 2011;32(November (11)):1139—42.712

[36] Kelcikova S, Skodova Z, Straka S. Effectiveness of hand713

hygiene education in a basic nursing school curricula. Public714

Health Nurs 2012;29(March—April (2)):152—9.715

[37] Wurtz R, Moye G, Jovanovic B. Handwashing machines,716

handwashing compliance, and potential for cross-717

contamination. Am J Infect Control 1994;22(4):228—30.718

[38] Watanakunakorn C, Wang C, Hazy J. An observational719

study of hand washing and infection control practices720

by healthcare workers. Infect Control Hosp Epidemiol721

1998;19(11):858—60.722

[39] Avila-Aguero ML, Umana MA, Jimenez AL, Faingezicht I,723

Paris MM. Handwashing practices in a tertiary-care, pedi-724

atric hospital and the effect on an educational program.725

Clin Perform Qual Health Care 1998;6(2):70—2.726

[40] Bischoff WE, Reynolds TM, Sessler CN, Edmond MB, Wen-727

zel RP. Handwashing compliance by health care workers:

the impact of introducing an accessible, alcohol-based hand 728

antiseptic. Arch Intern Med 2000;160(7):1017—21. 729

[41] Dubbert PM, Dolce J, Richter W, Miller M, Chapman 730

SW. Increasing ICU staff handwashing: effects of educa- 731

tion and group feedback. Infect Control Hosp Epidemiol 732

1990;11(April (4)):191—3. 733

[42] Julian KG, Subramanian K, Brumbach A, Whitener CJ. 734

Attitudes of healthcare workers and patients toward indi- 735

vidualized hand hygiene reminders. Infect Control Hosp 736

Epidemiol 2008;29(August (8)):781—2. 737

[43] Alp E, Ozturk A, Guven M, Celik I, Doganay M, Voss A. Impor- 738

tance of structured training programs and good role models 739

in hand hygiene in developing countries. J Infect Public 740

Health 2011;4(June (2)):80—90. 741

[44] Jaggi N, Rodrigues C, Rosenthal VD, Todi SK, Shah S, Saini 742

N, et al. Impact of an international nosocomial infection 743

control consortium multidimensional approach on central 744

line-associated bloodstream infection rates in adult inten- 745

sive care units in eight cities in India. Int J Infect Dis 746

2013;17(December (12)):e1218—24. 747

[45] Mehta Y, Jaggi N, Rosenthal VD, Rodrigues C, Todi SK, Saini 748

N, et al. Effectiveness of a multidimensional approach 749

for prevention of ventilator-associated pneumonia in 21 750

adult intensive-care units from 10 cities in India: find- 751

ings of the International Nosocomial Infection Control 752

Consortium (INICC). Epidemiol Infect 2013;141(December 753

(12)):2483—91. 754

Available online at www.sciencedirect.com

ScienceDirect

755

756