Fagernes & lingaas (2011) Factors interfering with the microflora on hands. journal of advanced...

ORIGINAL RESEARCH

Factors interfering with the microflora on hands: a regression analysis of

samples from 465 healthcare workers

Mette Fagernes & Egil Lingaas

Accepted for publication 13 August 2010

Correspondence to M. Fagernes:

e-mail: [email protected]

Mette Fagernes RN

PhD Student

Department of Internal Medicine, Vestfold

Hospital Trust, Tønsberg, Norway and

Institute of Nursing and Health Sciences,

University of Oslo, Norway

Egil Lingaas MD PhD

Head

Department of Infection Prevention,

Rikshospitalet University Hospital, Oslo,

Norway

FAGERNES M. & LINGAAS E. (2011)FAGERNES M. & LINGAAS E. (2011) Factors interfering with the microflora on

hands: a regression analysis of samples from 465 healthcare workers. Journal of

Advanced Nursing 67(2), 297–307. doi: 10.1111/j.1365-2648.2010.05462.x

AbstractAims. This paper is a report of a study of the impact of finger rings, wrist watches,

nail polish, length of fingernails, hand lotion, gender and occupation on hand

microbiology of healthcare workers.

Background. The impact of the above mentioned variables on hand microbiology

of healthcare workers is not well defined. Large scale studies suitable for multi-

variate analysis are needed to elucidate their role.

Methods. Both hands of 465 Norwegian healthcare workers were sampled by the

glove juice method during two study periods (2004 and 2007), and examined for total

number of bacteria and presence of Staphylococcus aureus, Enterobacteriacea and

non-fermentative Gram-negative rods. Multiple regression analysis was performed.

Results. The use of a wrist watch was associated with an enhanced total bacterial

count on hands compared to hands without a watch [(B) 3Æ25 (95% CI: 1Æ73–6Æ07),

P < 0Æ001], while the use of one plain finger ring increased the carriage rate of

Enterobacteriaceae [odds ratio 2Æ71 (95% CI: 1Æ42–5Æ20), P = 0Æ003]. The carriage

rate of Staphylococcus aureus was enhanced with fingernails longer than 2 mm

[odds ratio 2Æ17 (95% CI: 1Æ29–3Æ66), P = 0Æ004] and after recent use of hand lotion

[odds ratio 22Æ52 (95% CI: 4Æ05–125Æ30), P < 0Æ001]. No effect of nail polish was

observed. We found an association between occupation and carriage rate of

S. aureus and Enterobacteriaceae.

Conclusions. Health care workers should remove finger rings and watches at work.

Fingernails should be shorter than 2 mm, nail polish may be used.

Keywords: hand contamination, hand hygiene, healthcare workers, hospital

infection.

Introduction

Health care associated infections (HAI) are major sources of

morbidity and mortality worldwide (World Health Organi-

zation 2009). Microorganisms causing HAI are assumed to

be frequently transferred by the hands of healthcare

workers (HCWs). Consequently, hand hygiene is regarded

as one of the most fundamental infection prevention

practices (Larson 1988, Rotter 2007, World Health Orga-

nization 2009).

� 2010 Blackwell Publishing Ltd 297

J A N JOURNAL OF ADVANCED NURSING

Background

Many variables may potentially interfere with the risk of

hand contamination and the effect of hand washing and hand

disinfection. However, we still lack definitive answers with

regard to the influence of many of these variables, such as nail

polish, artificial nails, length of fingernails, wearing of finger

rings and wrist watches, use of hand lotion, gender and

occupation (World Health Organization 2009). Few studies

are published on these topics, and the results are partly

contradictory. There is a lack of studies suitable for multi-

variate analysis, which is a drawback since many of these

variables may be highly correlated and require large scale

studies to allow independent analysis of each single factor.

The inconsistency of the results is reflected in differing

recommendations on these issues in current guidelines on

hand hygiene (Larson 1995, Boyce & Pittet 2002, Pratt et al.

2007, World Health Organization 2009). We therefore need

more data to substantiate the role of these variables in order

to optimize hand hygiene among HCWs.

The study

Aims

The aim of the present study was to determine the impact of

the above mentioned variables in a large cohort of HCWs and

in ordinary clinical settings by the use of multivariate

analysis.

Study design

We used a cross-sectional design, appropriate for collecting

empiric data without interfering with the normal behaviour

of the study participants.

Participants

A convenience sample of 465 HCWs directly involved in

patient care (i.e. physicians, nurses, assistants, phlebotomists,

physiotherapists and radiography personnel) from three Nor-

wegian acute care hospitals were recruited into the study, 265

in study period 1 (2004) and 200 in study period 2 (2007).

They were interrupted between ordinary clinical work activ-

ities at least 2 hours after starting their shift and asked to

participate in the study. The different units were visited at

random days. No notification was give in advance, and no

extra hand hygiene was allowed before the hand samples were

taken. HCWs were collected based on the aim to include

approximately the same numbers with and without rings.

People were excluded if they had skin irritation or eczema,

if they had taken antibiotics during the previous 2 weeks or

had performed surgical hand disinfection during the preced-

ing 24 hours.

The following personal and work related data were

recorded: gender and occupation, length of fingernails (dom-

inant hand, fourth finger), nail polish (none, intact, chipped),

artificial nails, finger rings, wrist watch and/or bracelet (only

the second study period), hospital, time of day (day,

afternoon or night shift), time since work started, time since

most recent hand washing, hand disinfection and use of hand

lotion, and time since most recent glove use if hand hygiene

had been omitted after removing gloves.

Data collection and microbial methods

The data were collected during two separate study periods in

2004 and 2007 (Fagernes et al. 2007, Fagernes & Lingaas

2009). Both hands were sampled with a modified version of

the glove juice method. Each subject inserted their hand into

a sterile bag (Stomacher � 400 Classic; Seward, Worthing,

UK) containing 100 ml of sterile tryptic soy broth with

neutralizer as previously described (Fagernes et al. 2007,

Fagernes & Lingaas 2009). The bag was occluded around the

wrist, and the hand was massaged in a standardized manner

by an investigator for approximately 1 minute. The sampling

fluid was collected in a sterile container and total bacterial

counts were measured according to the European Norm 1499

(European Committee for Standardization 1997). Staphylo-

coccus aureus, Enterobacteriacea and non-fermentative

Gram-negative rods (NFGNR) were identified to the species

level, but were not quantified. The theoretical sensitivity for

detection of S. aureus was 500 CFU per hand in study period

1, and 90 CFU in study period 2. For detection of Gram-

negative rods, the theoretical sensitivity was 1000 CFU per

hand in study period 1. In study period 2 a change was made

and the sensitivity was 500 CFU per hand for subject 1–58

and 10 CFU per hand for subject 59–200.

Ethical considerations

Participation was voluntary and the HCWs were given oral

and written information before consenting to participate. The

studies were approved by the institutional review board at all

hospitals. All data were treated anonymously.

Data analysis

The median of the average number of bacteria on both hands

of each HCW and the presence of S. aureus, Enterobacteri-

M. Fagernes and E. Lingaas

298 � 2010 Blackwell Publishing Ltd

aceae and NFGNR on one or both hands were used as

outcome variables.

To account for the positive skewness of the measured data,

log-transformed data of total bacterial counts were used in

the analysis.

Separate regression models were constructed for total

bacterial count (linear regression) and for each organism

category (logistic regression). For all models, risk factors with

a P value of <0Æ2 as identified by univariate regression

analyses were incorporated into the multivariable regression

model. The least significant variables were thereafter removed

one by one until all remaining variables had a P £ 0Æ125.

Since the two parts of the study were separated by a period of

approximately 3 years, study number was included in all

models. All explanatory variables were included as categor-

ical variables, with the exception of work hours.

Use of wrist watches was recorded in the second part of the

study only. A separate analysis was therefore performed for

this variable. The watch carrying hands were compared with

a randomly selected hand of each HCW without a watch,

adjusted for hand dominance.

The fit of the linear model was assessed by inspection of the

residuals, while the fit of the logistic model was assessed by

use of the Hosmer and Lemeshow goodness-of-fit test. All

analyses were performed using the SPSSSPSS 16.0 (SPSS Inc.,

Chicago, IL, USA) statistical software package. The level of

statistical significance was set to 5%.

Validity and reliability

Hand samples were collected by the ‘Glove juice method’.

The method is considered to be the most valid and reliable

method to describe the transient and permanent flora on

hands (Paulson 1993).

Results

Hand samples were collected from a total of 465 HCWs. The

distribution of registered variables is shown in Table 1.

Total bacterial count

The median bacterial count recovered from the hands of 465

HCWs was 2,075,000 (range 2250–60,500,000). Occupation

(P = 0Æ004), finger rings (P = 0Æ002), length of fingernails

(P = 0Æ048), nail polish (P = 0Æ057), time since hand

disinfection (P < 0Æ001) and study number (P < 0Æ001) were

incorporated into the multivariable regression model. As

shown in Table 2, finger rings, time since hand disinfection

Table 1 Study participants and study variables

Variable

First study

period

(n = 265)

Second study

period

(n = 200)

Total

(N = 465)

Hospital

Hospital 1 132 (49Æ8) 155 (77Æ5) 287 (61Æ7)

Hospital 2 133 (50Æ2) 0 (0) 133 (28Æ6)

Hospital 3 0 (0) 45 (22Æ5) 45 (9Æ7)

Gender

Female 243 (91Æ7) 180 (90Æ0) 423 (91Æ0)

Male 22 (8Æ3) 20 (10Æ0) 42 (9Æ0)

Occupation

Nurse 148 (55Æ8) 114 (57Æ0) 262 (56Æ3)

Nursing assistant 46 (17Æ4) 12 (6Æ0) 58 (12Æ5)

Phlebotomist 23 (8Æ7) 26 (13Æ0) 49 (10Æ5)

Radiography

personnel

13 (4Æ9) 17 (8Æ5) 30 (6Æ5)

Physician 11 (4Æ2) 15 (7Æ5) 26 (5Æ6)

Physiotherapist 8 (3Æ0) 10 (5Æ0) 18 (3Æ9)

Other 16 (6Æ0) 6 (3Æ0) 22 (4Æ7)

Shift

Day shift 236 (89Æ1) 200 (100) 436 (93Æ8)

Afternoon shift 6 (2Æ3) 0 (0) 6 (1Æ3)

Night shift 23 (8Æ7) 0 (0) 23 (4Æ9)

Hours at work before sampling

3 80 (30Æ2) 62 (31Æ0) 142 (30Æ5)

4 56 (21Æ1) 30 (15Æ0) 86 (18Æ5)

5 50 (18Æ9) 24 (12Æ0) 74 (15Æ9)

6 37 (14Æ0) 50 (25Æ0) 87 (18Æ7)

7 15 (5Æ7) 31 (15Æ5) 46 (9Æ9)

>8 27 (10Æ2) 3 (1Æ5) 30 (6Æ5)

Finger ring

None 113 (42Æ6) 100 (50) 213 (45Æ8)

One plain 121 (45Æ7) 71 (35Æ5) 192 (41Æ3)

One decorative 31 (11Æ7) 19 (9Æ5) 50 (10Æ8)

More than one 0 (0) 10 (5Æ0) 10 (2Æ2)

Wrist watch

No – 121 (60Æ5) 121 (26Æ0)

Yes – 79 (39Æ5) 79 (17Æ0)

Not registered 265 (100) – 265 (57Æ0)

Length of fingernails (mm)

<2 179 (67Æ5) 151 (75Æ5) 330 (71Æ0)

2–2Æ9 65 (24Æ5) 35 (17Æ5) 100 (21Æ5)

>3 20 (7Æ5) 9 (4Æ5) 29 (6Æ2)

Not registered 1 (0Æ04) 5 (2Æ5) 6 (1Æ3)

Nail polish

No polish 206 (77Æ7) 171 (85Æ5) 377 (81Æ1)

Intact polish 18 (6Æ8) 17 (8Æ5) 35 (7Æ5)

Chipped polish 41 (15Æ5) 10 (5Æ0) 51 (11Æ0)

Not registered – 2 (1Æ0) 2 (0Æ4)

Artificial nails

No 264 (99Æ6) 197 (98Æ5) 461 (99Æ1)

Yes 1 (0Æ4) 3 (1Æ5) 4 (0Æ9)

Minutes since hand washing

<5 45 (17Æ0) 25 (12Æ5) 70 (15Æ1)

5–10 65 (24Æ5) 24 (12Æ0) 89 (19Æ1)

JAN: ORIGINAL RESEARCH Variables having an impact on hand contamination


and study number were included in the final model, and were

found to have a significant impact on the total bacterial count.

A separate analysis of hands with a wrist watch demon-

strated significantly higher total bacterial counts than on

control hands [unadjusted effect estimate: 5Æ70 (95% CI:

3Æ04–10Æ68), P < 0Æ001, adjusted effect estimate: 3Æ25 (95%

CI: 1Æ73–6Æ07), P < 0Æ001]. Variables controlled for in the

final regression model were hand disinfection, nail length,

nail polish and hospital.

Gram negative rods

Enterobacteriaceae were found on one or both hands of 75

(16Æ1%) HCWs. Hospital (P < 0Æ001), occupation

(P = 0Æ023), gender (P = 0Æ160), finger rings (P = 0Æ027), time

since hand washing (P = 0Æ006) and study number

(P < 0Æ001) were incorporated to the multivariable regression

model. Gender and time since hand washing were taken out of

the model during the multivariate analysis. As shown in

Table 3, ring wearing was found to have a significant impact

on the recovery of Enterobacteriaceae. Significant differences

were also shown between the three hospitals, study number

and between nurses and radiography personnel.

Non-fermentative Gram-negative rod species were identi-

fied on one or both hands of 164 (35Æ3 %) of 465 HCWs.

Hospital (P < 0Æ001), occupation (P = 0Æ167), finger rings

(P = 0Æ049), time since hand disinfection (P = 0Æ127) and

study number (P < 0Æ001) were incorporated in the multi-

variable regression model. Occupation and finger rings were

taken out of the model during the multivariate analysis, and

only hospital, time since hand disinfection and study number

were found to influence the occurrence of NFGNR. Table 4

shows effect estimates for the variables in the final model.

A separate analysis for watches showed an unadjusted

effect on NFGNR carriage [unadjusted effect estimate: OR

2Æ21 (95% CI: 1Æ21–4Æ03), P = 0Æ010] which disappeared

after adjusting for finger rings and nail polish in the final

model [adjusted effect estimate: OR 1Æ34 (95% CI: 0Æ64–

2Æ81), P = 0Æ442].

Table 1 (Continued)

Variable

First study

period

(n = 265)

Second study

period

(n = 200)

Total

(N = 465)

11–20 62 (23Æ4) 37 (18Æ5) 99 (21Æ3)

>20 93 (35Æ1) 114 (57Æ0) 207 (44Æ5)

Minutes since hand disinfection

<5 13 (4Æ9) 18 (9Æ0) 31 (6Æ7)

5–10 22 (8Æ3) 23 (11Æ5) 45 (9Æ7)

11–20 18 (6Æ8) 29 (14Æ5) 47 (10Æ1)

>20 94 (35Æ5) 97 (48Æ5) 191 (41Æ1)

Not done 118 (44Æ5) 33 (16Æ5) 151 (32Æ5)

Minutes since application of hand lotion

<5 9 (3Æ4) 1 (0Æ5) 10 (2Æ2)

5–10 3 (1Æ1) 2 (1Æ0) 5 (1Æ1)

11–20 4 (1Æ5) 1 (0Æ5) 5 (1Æ1)

>20 54 (20Æ4) 42 (21Æ0) 96 (20Æ6)

Not done 195 (73Æ6) 154 (77Æ0) 349 (75Æ1)

Minutes since glove use – if hand hygiene had been omitted after

glove removal

<5 5 (1Æ9) 0 (0Æ0) 5 (1Æ1)

5–10 1 (0Æ4) 1 (0Æ5) 2 (0Æ4)

11–20 2 (0Æ8) 1 (0Æ5) 3 (0Æ7)

>20 4 (1Æ5) 3 (1Æ5) 7 (1Æ5)

Not used 253 (95Æ5) 194 (97Æ0) 447 (96Æ1)

Not registered – 1 (0Æ5) 1 (0Æ2)

Values are given as n (%).

Table 2 Multivariate regression analysis of variables with an impact on the total number of bacteria on the hands of healthcare workers

(N = 465)

Variable

Unadjusted effect

(95% CI) P value

Adjusted effect

(95% CI) P value

Finger ring – 0Æ002 – 0Æ003

No ring Reference group – Reference group –

One plain ring 1Æ72 (1Æ23–2Æ39) 0Æ001 1Æ40 (1Æ02–1Æ90) 0Æ035

One decorative ring 1Æ82 (1Æ08–3Æ07) 0Æ024 1Æ50 (0Æ92–2Æ43) 0Æ102

More than one ring 3Æ53 (1Æ20–10Æ32) 0Æ022 5Æ53 (2Æ00–15Æ27) 0Æ001

Minutes since hand disinfection – <0Æ001 – 0Æ027

Not performed Reference group – Reference group –

<5 0Æ23 (0Æ12–0Æ44) <0Æ001 0Æ41 (0Æ22–0Æ77) 0Æ005

5–10 0Æ50 (0Æ29–0Æ87) 0Æ013 0Æ77 (0Æ45–1Æ30) 0Æ329

11–20 0Æ32 (0Æ19–0Æ55) <0Æ001 0Æ56 (0Æ33–0Æ96) 0Æ035

>20 0Æ45 (0Æ31–0Æ64) <0Æ001 0Æ66 (0Æ47–0Æ94) 0Æ020

Study period – <0Æ001 – <0Æ001

Study 1 (2004) Reference group – Reference group –

Study 2 (2007) 0Æ29 (0Æ21–0Æ39) <0Æ001 0Æ32 (0Æ23–0Æ43) <0Æ001



Staphylococcus aureus

Staphylococcus aureus was detected on one or both hands of

120 (25Æ8 %) of 465 HCWs. Hospital (P = 0Æ047), occupa-

tion (P < 0Æ001), length of fingernails (P = 0Æ010), time since

application of hand lotion (P = 0Æ009) and study number

(P = 0Æ934) were incorporated in the multivariable regression

model. Only hospital was removed from the model during the

multivariate analysis. Table 5 describes effect estimates for

the variables in the final model.

Discussion

Study limitations

The study includes hand samples from both hands of 465

HCWs collected in different clinical settings at three Norwe-

gian acute care hospitals. A cross-sectional design was used.

To compensate for the lack of randomization, the different

units were visited at random days. The results are expected to

be generalizable across international hospital settings, but

Table 3 Multivariate logistic regression analysis of variables with an impact on the occurrence of Enterobacteriaceae on the hands of

healthcare workers (N = 465)

Variable OR (95% CI) P value

Adjusted OR

(95% CI) P value

Hospital – <0Æ001 – 0Æ003

Hospital 1 Reference group – Reference group –

Hospital 2 0Æ24 (0Æ10–0Æ57) 0Æ001 0Æ68 (0Æ24–1Æ94) 0Æ470

Hospital 3 4Æ36 (2Æ25–8Æ45) <0Æ001 3Æ47 (1Æ65–7Æ32) 0Æ001

Occupation – 0Æ023 – 0Æ076

Nurse Reference group – Reference group –

Nursing assistant 0Æ48 (0Æ16–1Æ41) 0Æ182 0Æ55 (0Æ17–1Æ78) 0Æ318

Phlebotomist 1Æ66 (0Æ76–3Æ63) 0Æ202 1Æ24 (0Æ52–2Æ94) 0Æ634

Radiography personnel 3Æ24 (1Æ40–7Æ50) 0Æ006 3Æ98 (1Æ59–10Æ01) 0Æ003

Physician 2Æ39 (0Æ94–6Æ10) 0Æ068 1Æ07 (0Æ36–3Æ16) 0Æ909

Physiotherapist 1Æ85 (0Æ58–5Æ95) 0Æ300 1Æ56 (0Æ44–5Æ55) 0Æ496

Other 1Æ91 (0Æ66–5Æ50) 0Æ232 2Æ20 (0Æ65–7Æ46) 0Æ207

Finger ring – 0Æ027 – 0Æ019

No ring Reference group – Reference group –

One plain ring 1Æ88 (1Æ08–3Æ28) 0Æ026 2Æ71 (1Æ42–5Æ20) 0Æ003

One decorative ring 1Æ97 (0Æ87–4Æ44) 0Æ102 2Æ25 (0Æ89–5Æ68) 0Æ086

More than one ring 5Æ25 (1Æ38–19Æ94) 0Æ015 2Æ93 (0Æ72–11Æ97) 0Æ133

Study period – <0Æ001 – 0Æ001


Study 2 (2007) 5Æ47 (3Æ10–9Æ70) <0Æ001 3Æ52 (1Æ68–7Æ39) 0Æ001

Table 4 Multivariate logistic regression analysis of variables with an impact on the occurrence of non-fermentative Gram-negative rods on the

hands of healthcare workers (N = 465)


Adjusted OR

(95% CI) P value

Hospital – <0Æ001 – <0Æ001

Hospital 1 Reference group – Reference group –

Hospital 2 0Æ55 (0Æ34–0Æ88) 0Æ013 0Æ66 (0Æ37–1Æ19) 0Æ165

Hospital 3 10Æ63 (4Æ58–24Æ69) <0Æ001 10Æ02 (4Æ10–24Æ48) <0Æ001

Minutes since hand

disinfection

– 0Æ127 – 0Æ021


<5 0Æ35 (0Æ13–0Æ95) 0Æ040 0Æ23 (0Æ08–0Æ70) 0Æ009

5–10 1Æ09 (0Æ55–2Æ17) 0Æ805 0Æ64 (0Æ29–1Æ43) 0Æ280

11–20 0Æ69 (0Æ33–1Æ41) 0Æ307 0Æ37 (0Æ16–0Æ86) 0Æ020

>20 1Æ16 (0Æ75–1Æ81) 0Æ507 0Æ89 (0Æ53–1Æ49) 0Æ657

Study period – <0Æ001 – 0Æ092


Study 2 (2007) 2Æ57 (1Æ74–3Æ80) <0Æ001 1Æ61 (0Æ93–2Æ78) 0Æ092



potential differences were not explored. It is not known to

which degree the results can be generalized to other contexts

where hygienic aspects of finger rings are of interest, as in

kindergartens, food industry etc.

Gender

No differences were found between genders, neither regarding

bacterial load nor prevalence of potential pathogens. We are

aware of only one previous study comparing the hand

microflora of male and female HCWs. Larson (1981) reported

significantly higher prevalence of Gram negative rods among

40 male HCWs compared to 63 females. In our study the

prevalence of Enterobacteriaceae was 15Æ4% and 23Æ8%

among 423 women and 42 men respectively, but the difference

was not statistically significant (P = 0Æ156).

Occupation

Using nurses as reference, nursing assistants had higher

carriage rate of S. aureus, whereas the prevalence was lower

among radiography personnel. The latter group, however,

more frequently carried Enterobacteriaceae. No differences

were found in total bacterial counts between nurses and other

occupational groups. We have identified four published

studies comparing hand microflora of different healthcare

professionals. Larson (1981) found significantly higher prev-

alence of Gram negative rods among 31 physicians (42%)

than among 54 nurses (9%). Conversely, Horn et al. (1988)

found significantly higher prevalence of Gram negative

bacteria on the hands of oncology and dermatology nurses

compared to physicians from the same units. Larson et al.

(1986) measured total bacterial counts repeatedly among 12

nurses and 4 physicians and did not find significant differ-

ences. Daschner (1985) reported significantly higher bacterial

numbers and higher prevalence of Gram negative rods and S.

aureus on the hands of physicians compared to other HCWs

(N = 328).

Length of fingernails

Multivariate analysis demonstrated a statistically significant

correlation between fingernails longer than 2 mm and

prevalence of S. aureus, but no association with carriage

of Gram negative rods or total bacterial numbers. We have

identified two published studies on the influence of the

length of natural nails. Rupp et al. (2008) examined 192

samples from the dominant hand of 69 nurses over a 2-year

period and found increased bacterial counts with nail length

above 2 mm, but no difference in the recovery of Gram

Table 5 Multivariate logistic regression analysis of variables with an impact on the occurrence of Staphylococcus aureus on the hands of

healthcare workers (N = 459)


Adjusted OR

(95% CI) P value

Occupation – 0Æ001 – 0Æ002

Nurse Reference group Reference group

Nursing assistant 2Æ46 (1Æ37–4Æ44) 0Æ003 2Æ60 (1Æ39–4Æ90) 0Æ003

Phlebotomist 0Æ59 (0Æ26–1Æ33) 0Æ202 0Æ60 (0Æ26–1Æ38) 0Æ230

Radiography personnel 0Æ22 (0Æ05–0Æ93) 0Æ040 0Æ18 (0Æ04–0Æ82) 0Æ027

Physician 0Æ91 (0Æ35–2Æ36) 0Æ845 1Æ16 (0Æ44–3Æ11) 0Æ762

Physiotherapist 0Æ61 (0Æ17–2Æ16) 0Æ440 0Æ63 (0Æ17–2Æ34) 0Æ492

Other 2Æ53 (1Æ04–6Æ12) 0Æ040 2Æ55 (0Æ98–6Æ66) 0Æ056

Length of fingernails (mm) – 0Æ010 – 0Æ014

>2 Reference group – Reference group –

2–2Æ99 2Æ02 (1Æ24–3Æ27) 0Æ005 2Æ17 (1Æ29–3Æ66) 0Æ004

<3 1Æ89 (0Æ84–4Æ24) 0Æ124 1Æ34 (0Æ55–3Æ29) 0Æ518

Minutes since use of

hand lotion

– 0Æ009 – 0Æ006


<5 13Æ67 (2Æ85–65Æ69) 0Æ001 22Æ52 (4Æ05–125Æ30) <0Æ001

5–10 2Æ28 (0Æ37–13Æ88) 0Æ372 2Æ28 (0Æ36–14Æ67) 0Æ384

11–20 0Æ85 (0Æ09–7Æ76) 0Æ889 0Æ60 (0Æ060–5Æ99) 0Æ665

>20 1Æ55 (0Æ94–2Æ56) 0Æ084 1Æ38 (0Æ80–2Æ37) 0Æ247

Study period – 0Æ934 – 0Æ072


Study 2 (2007) 1Æ02 (0Æ67–1Æ55) 0Æ934 1Æ54 (0Æ96–2Æ46) 0Æ072



negative enteric bacteria. By swabbing the front of the

fingernails on the dominant hand of 100 nurses, including

the cuticle area, Wynd et al. (1994) were not able to detect

an influence of nail length on total bacterial numbers.

Neither of these two studies reported the prevalence of

S. aureus.

Recommendations on length of fingernails vary in different

guidelines for hand hygiene. Some guidelines use the phrase

short nails (Pratt et al. 2007), whereas Centers for Disease

Control and Prevention (CDC) (Boyce & Pittet 2002) and the

World Health Organization (WHO) (World Health Organi-

zation 2009) recommend nail length less than 1/4-inch

(6Æ3 mm), and 5 mm respectively. Based on the findings of

the present study and the results of Rupp et al. (2008), we

recommend that the fingernails of HCWs should not be

longer than 2 mm.

Nail polish

No impact of nail polish was detected in this study. We are not

aware of any other study examining the impact of nail polish

on the microflora of the whole hand. Three studies, sampling

the nails only, did not show any influence of polish on

bacterial counts before hand hygiene (Baumgardner et al.

1993, Wynd et al. 1994, Edel et al. 1998). However, two

studies with 100 and 61 participants respectively demon-

strated higher counts on polished nails after surgical scrub

(Wynd et al. 1994, Edel et al. 1998). The third study did not

detect differences after regular hand washing in 26 partici-

pants with nail polish on one hand only (Baumgardner et al.

1993). Among the guidelines cited above, only the UK

guideline recommends that HCWs refrain from using nail

polish. Our results do not support this recommendation.

Artificial fingernails

Due to low numbers, wearing of artificial fingernails was not

incorporated in the regression analysis.

Wrist watches

The guidelines from CDC and WHO do not address the issue

of wrist watches, except before surgical hand antisepsis

(Boyce & Pittet 2002, World Health Organization 2009),

while the English guideline state that wrist jewellery should

be removed prior to patient contact (Pratt et al. 2007). Only

two studies have previously been published on this issue.

Field et al. (1996) found that skin below wrist watches

harbours more bacteria than control skin on the opposite

wrist. Jeans et al. (2010) investigated the impact of wrist

watches on the bacterial counts on the wrist and finger tips.

They found that watch wearers had higher counts of bacteria

on their wrist compared to HCWs without a wrist watch.

They did not find any impact of wrist watches on the

bacterial load on finger tips when the watch was kept in

place. When the HCW removed the watch prior to sampling,

the manipulation of the watch resulted in increased counts of

bacteria on the fingertips (Jeans et al. 2010).

In the present study, we recovered more than three times as

many bacteria from hands with watches compared to control

hands. We recommend that HCW abstain from the use of

wrist watches at work.

Finger rings

The overall analysis showed that HCWs with finger rings had

enhanced total number of bacteria on hands. However, we

suspect that this finding is due to the lack of adjustment for

watches in the aggregated 2004/2007 data. Watches were

shown to significantly increase bacterial numbers in the 2007

study, which also demonstrated a significant correlation

between the use of rings and watches; 64% of ring wearers

used a wrist watch compared to 15% of the HCW without

ring.

For Enterobacteriaceae a significant increase in prevalence

was revealed for one plain ring only. We assume that the

failure to detect a significant effect of a single decorative ring

and multiple rings is due to insufficient statistical power.

These two groups were much smaller than the groups with a

single plain ring and no ring. No influence of rings was

detected for S. aureus or NFGNR.

We have identified more than 20 studies on the influence of

finger rings published in scientific journals since 1968

(Lowbury et al. 1968, Nicholson-Pegg 1982, Dewan &

Fergus 1985, Hoffman et al. 1985, Jacobson et al. 1985,

Athar et al. 1989, Field et al. 1996, Nicolai et al. 1997,

Salisbury et al. 1997, Trick et al. 2003, Alp et al. 2006,

Kelsall et al. 2006, Waterman et al. 2006, Fagernes & Nord

2007, Fagernes et al. 2007, Wongworawat & Jones 2007,

Al-Allak et al. 2008, Rupp et al. 2008, Yildirim et al. 2008,

Alur et al. 2009, Fagernes & Lingaas 2009, Stein &

Pankovich-Wargula 2009, Hautemaniere et al. 2010). Most

of these studies conclude that there is an association between

ring wearing and an enhanced bacterial load on hands, and

an increased prevalence of Gram negative bacteria. Results

from studies of the association between rings and hand

contamination after hand hygiene are less consistent. In

particular, several studies failed to show differences after

surgical hand antisepsis (Jacobson et al. 1985, Waterman

et al. 2006, Wongworawat & Jones 2007).



The CDC states that no recommendation can be made

about wearing rings in healthcare settings, and that this is an

unresolved issue (Boyce & Pittet 2002). WHO recommends

the removal of rings or other jewellery during health care, but

accept the use of simple wedding band during routine care

based on strong religious or cultural influences (World Health

Organization 2009). UK guidelines issued in 2007 state that

all wrist an ideally hand jewellery should be removed before a

shift of clinical work begins (Pratt et al. 2007). Our results

support this recommendation.

Hand lotion

Guidelines on hand hygiene commonly recommend frequent

use of hand lotion to maintain the integrity of the skin

(Larson 1995, Boyce & Pittet 2002, World Health Organi-

zation 2009). We found a significant association between the

use of hand lotion within 5 minutes before sampling and

recovery of S. aureus. One possible explanation for this

finding may be that the hands pick up staphylococci more

efficiently immediately after application of hand lotion.

However, it may be due to better recovery or enhanced

dispersion of S. aureus during sampling and plating due to the

influence of surface-active ingredients, or simply to statistical

chance. In a paper by Jacobson et al. (1985), reporting

bacterial counts on the hands of 12 volunteers, the authors

note that they observed that hand lotion increased the

bacterial count. However, no data were presented, and no

follow-up has been published. Further studies are needed on

this issue.

Hand washing and hand disinfection

There is a plethora of published studies on the efficacy of

different methods and agents for hand decontamination.

However, the present study is not a study of the immediate

effect of hand hygiene. It is a cross-sectional study taking into

account the time since hand washing or hand disinfection,

and also the risk of recontamination during ordinary health-

care activities between performance of hand hygiene and

sampling. We found a significant reduction of total bacterial

load on hands among HCW who had previously performed

hand antisepsis with alcohol, but no effect of previous hand

washing even within 5 minutes before sampling. As the risk

of recontamination is probably independent of the method

used for previous hand hygiene, this difference is most

probably a result of a sustained effect of alcohol on the

permanent microflora. A somewhat complex correlation was

found between the total number of bacteria recovered and the

time since hand disinfection. Compared to HCWs who had

not disinfected their hands on the day of sampling, a

significant reduction of bacterial load was observed for all

5-minute intervals after disinfection, except for samples

collected between 5 and 10 minutes after hand disinfection.

Most probably this is due to chance, even though this group

is statistically similar to the other groups (subject numbers,

CFU range). A possible explanation might be that alcohol has

a biphasic effect on hand microflora with an initial reversible

bacteriostatic effect followed by a slower bactericidal effect.

Previous hand disinfection was also associated with a

reduced prevalence of NFGNR. This is probably due to a

What is already known about this topic

• Healthcare associated infections are universal and their

prevention has high priority in healthcare facilities

worldwide.

• Even though hand hygiene is widely accepted as a

cornerstone of infection prevention, we still lack

answers to several questions on how to optimize hand

hygiene.

• Due to lack of valid information, international and

national guidelines on hand hygiene (WHO, UK, US

and others) differ in their recommendations regarding

ring wearing, use of wrist watches, nail length and nail

polish.

What this paper adds

• Wrist watches and finger rings are associated with

increased bacterial numbers on the hands of healthcare

workers.

• Long finger nails (>2 mm) enhance the carriage rate of

Staphylococcus aureus.

• Nail polish has no impact on hand contamination, while

the use of hand lotion may increase the carriage rate of

Staphylococcus aureus.

Implications for practice and/or policy

• Healthcare workers should keep finger nails short

(<2 mm), and remove all finger rings (included plain

wedding rings) and wrist watches during clinical work.

• Several guidelines on hand hygiene should be re-written

with regard to length of fingernails and the use of wrist

watches, finger rings and nail polish.

• Educational and clinical leaders must give a priority to

implementation and compliance to the guidelines on

hand hygiene.



sustained effect on the permanent skin flora, which frequently

contains NFGNR (Lucet et al. 2002). In contrast, Entero-

bacteriaceae and S. aureus, which are more typical represen-

tatives of temporary bacteria, were not affected by previous

disinfection. This may be due to contamination of the hands

in the time interval between hand disinfection and sampling.

Differences between the two study periods

We found significantly lower bacterial load on hands in the

second study period. This can probably be explained by a

significant increase in the use of alcoholic hand disinfection

from the first to the second study period. Also, a significantly

higher prevalence of Gram negative bacteria was observed,

which may be a result of enhanced sensitivity of the detection

method. These differences are taken into consideration in the

regression analysis by including study period as an indepen-

dent variable.

Conclusion

Due to lack of valid information about variables with a

potential impact on hand microflora, current guidelines on

hand hygiene have differing recommendations on these issues

(Larson 1995, Boyce & Pittet 2002, Pratt et al. 2007, World

Health Organization. 2009). The present study includes hand

samples from both hands of 465 HCW, and is to our

knowledge the largest study on this subject. The results show

that neither wrist watches nor rings should be used by

healthcare workers at work, and that fingernails should not

be longer than 2 mm. Nail polish does not seem to influence

the microflora on hands, but hand lotion may be a risk factor

and needs to be further examined.

It is a responsibility both for the individual healthcare

worker and the healthcare institutions to ensure patient safety

based on evidence based practice. Several international,

national and institutional guidelines on hand hygiene should

be re-written with regard to length of fingernails and the use

of wrist watches, finger rings and nail polish. Implementation

of the guidelines should been given priority by both educa-

tional and clinical leaders.

Acknowledgements

We greatly appreciate statistical advice from Magne Thore-

sen. We also thank Tone Herring, Hilde Kaasa, Anne

Ottestad Syvertsen, Terje Lingaas and Trond Lingaas for

technical assistance and May-Solveig Fagermoen for valuable

comments in the introductory part of this study. We are

grateful to all HCW participating in the study.

Funding

The study was funded by research grants from Helse Sør

RHF, Norway, which is a public hospital trust.

Conflict of interest

No conflict of interest has been declared by the authors.

Author contributions

MF and EL were responsible for the study conception and

design. MF performed the data collection. MF performed the

data analysis. MF and EL were responsible for the drafting of

the manuscript. MF and EL made critical revisions to the

paper for important intellectual content. MF provided

statistical expertise. MF obtained funding. MF and EL

provided administrative, technical or material support. EL

supervised the study.

References

Al-Allak A., Sarasin S., Key S. & Morris-Stiff G. (2008) Wedding

rings are not a significant source of bacterial contamination

following surgical scrubbing. Annals of the Royal College of

Surgeons of England 90(2), 133–135.

Alp E., Haverkate D. & Voss A. (2006) Hand hygiene among labo-

ratory workers. Infection Control and Hospital Epidemiology

27(9), 978–980.

Alur A.A., Rane M.J., Scheetz J.P., Lorenz D.J. & Gettleman L.

(2009) Simulated microbe removal around finger rings using dif-

ferent hand sanitation methods. International Journal of Oral

Science 1(3), 136–142.

Athar M.A., Stafford L. & Wootliff J.S. (1989) Bacterial population

of ring and control fingers in health care workers. Infection

Control Canada 4(2), 8–12.

Baumgardner C.A., Maragos C.S., Walz J. & Larson E. (1993)

Effects of nail polish on microbial growth of fingernails. Dispelling

sacred cows. AORN Journal 58(1), 84–88.

Boyce J.M. & Pittet D. (2002) Guideline for Hand Hygiene in

Health-Care Settings: recommendations of the Healthcare Infec-

tion Control Practices Advisory Committee and the HICPAC/

SHEA/APIC/IDSA Hand Hygiene Task Force. Infection Control

and Hospital Epidemiology 23(12 Suppl.), 3–40.

Daschner F.D. (1985) The transmission of infections in hospitals by

staff carriers, methods of prevention and control. Infection Control

6(3), 97–99.

Dewan P.A. & Fergus M. (1985) Rings in operating theatres. New

Zealand Medical Journal 98(793), 1094.

Edel E., Houston S., Kennedy V. & LaRocco M. (1998) Impact of a

5-minute scrub on the microbial flora found on artificial, polished,

or natural fingernails of operating room personnel. Nursing

Research 47(1), 54–59.

European Committee for Standardization (1997) European Commit-

tee for Standardization. Chemical Disinfectants and Antiseptics –



Hygienic Handwash – Test Methods and Requirements. Phase 2/

Step 2. European Committee for Standardization, Brussels, Belgium.

Fagernes M. & Lingaas E. (2009) Impact of finger rings on trans-

mission of bacteria during hand contact. Infection Control and

Hospital Epidemiology 30(5), 427–432.

Fagernes M. & Nord R. (2007) A study of microbial load of different

types of finger rings worn by healthcare personnel. Vard i Norden

27(2), 21–24.

Fagernes M., Lingaas E. & Bjark P. (2007) Impact of a single plain

finger ring on the bacterial load on the hands of healthcare

workers. Infection Control and Hospital Epidemiology 28(10),

1191–1195.

Field E.A., McGowan P., Pearce P.K. & Martin M.V. (1996) Rings

and watches: should they be removed prior to operative dental

procedures? Journal of Dentistry 24(1–2), 65–69.

Hautemaniere A., Cunat L., Diguio N., Vernier N., Schall C., Daval

M., Ambrogi V., Tousseul S., Hunter P.R. & Hartemann P. (2010)

Factors determing poor practice in alcoholic gel hand rub tech-

nique in hospital workers. Journal of infection and Public Health

3(1), 25–34.

Hoffman P.N., Cooke E.M., McCarville M.R. & Emmerson A.M.

(1985) Micro-organisms isolated from skin under wedding rings

worn by hospital staff. British Medical Journal 290(6463), 206–207.

Horn W.A., Larson E.L., McGinley K.J. & Leyden J.J. (1988)

Microbial flora on the hands of health care personnel: differences

in composition and antibacterial resistance. Infection Control and

Hospital Epidemiology 9(5), 189–193.

Jacobson G., Thiele J.E., McCune J.H. & Farrell L.D. (1985)

Handwashing: ring-wearing and number of microorganisms.

Nursing Research 34(3), 186–188.

Jeans A.R., Moore J., Nicol C., Bates C. & Read R.C. (2010)

Wristwatch use and hospital-acquired infection. Journal of

Hospital Infection 74(1), 16–21.

Kelsall N., Griggs R. & Bowker K. (2006) Should finger rings be

removed prior to scrubbing for theatre? Journal of Hospital

Infection 62(4), 450–452.

Larson E.L. (1981) Persistent carriage of gram-negative bacteria on

hands. American Journal of Infection Control 9(4), 112–119.

Larson E. (1988) A causal link between handwashing and risk of

infection? Examination of the evidence. Infection Control 9(1),

28–36.

Larson E.L. (1995) APIC guideline for handwashing and hand

antisepsis in health care settings. American Journal of Infection

Control 23(4), 251–269.

Larson E., McGinley K.J., Grove G.L., Leyden J.J. & Talbot G.H.

(1986) Physiologic, microbiologic, and seasonal effects of hand-

washing on the skin of health care personnel. American Journal of

Infection Control 14(2), 51–59.

Lowbury E.J.L., Blowers R. & Cunliff A.C. (1968) Aseptic methods

in the operating suite. Lancet 1, 705–709.

Lucet J.C., Rigaud M.P., Mentre F., Kassis N., Deblangy C.,

Andremont A. & Bouvet E. (2002) Hand contamination before

and after different hand hygiene techniques: a randomized clinical

trial. Journal of Hospital Infection 50(4), 276–280.

Nicholson-Pegg A. (1982) The wearing of wedding rings in the

operating department. NATNEWS 19(4), 19–25.

Nicolai P., Aldam C.H. & Allen P.W. (1997) Increased awareness of

glove perforation in major joint replacement. A prospective,

randomised study of Regent Biogel Reveal gloves. Journal of Bone

and Joint Surgery. British Volume 79(3), 371–373.

Paulson D.S. (1993) Evaluation of three microorganism recovery

procedures used to determine handwash efficacy. Dairy, Food and

Environmental Sanitation 13(9), 520–523.

Pratt R.J., Pellowe C.M., Wilson J.A., Loveday H.P., Harper P.J.,

Jones S.R., McDougall C. & Wilcox M.H. (2007) epic2: National

evidence-based guidelines for preventing healthcare-associated

infections in NHS hospitals in England. Journal of Hospital

Infection 65(Suppl. 1), S1–S64.

Rotter M.L. (2007) ‘‘I wash my hands of it!?’’ – Trends in hand

hygiene over the past decades. GMS Krankenhaushygiene Inter-

disziplinar 2(1), Doc07.

Rupp M.E., Fitzgerald T., Puumala S., Anderson J.R., Craig R., Iwen

P.C., Jourdan D., Keuchel J., Marion N., Peterson D., Sholtz L. &

Smith V. (2008) Prospective, controlled, cross-over trial of alcohol-

based hand gel in critical care units. Infection Control and Hospital

Epidemiology 29(1), 8–15.

Salisbury D.M., Hutfilz P., Treen L.M., Bollin G.E. & Gautam S.

(1997) The effect of rings on microbial load of health care workers’

hands. American Journal of Infection Control 25(1), 24–27.

Stein D.T. & Pankovich-Wargula A.L. (2009) The dilemma of the

wedding band. Orthopedics 32(2), 86.

Trick W.E., Vernon M.O., Hayes R.A., Nathan C., Rice T.W.,

Peterson B.J., Segreti J., Welbel S.F., Solomon S.L. & Weinstein

R.A. (2003) Impact of ring wearing on hand contamination and

comparison of hand hygiene agents in a hospital. Clinical

Infectious Diseases 36(11), 1383–1390.

Waterman T.R., Smeak D.D., Kowalski J. & Hade E.M. (2006)

Comparison of bacterial counts in glove juice of surgeons wearing

smooth band rings versus those without rings. American Journal of

Infection Control 34(7), 421–425.

Wongworawat M.D. & Jones S.G. (2007) Influence of rings on the

efficacy of hand sanitization and residual bacterial contamina-

tion. Infection Control and Hospital Epidemiology 28(3), 351–

353.

World Health Organization (2009) Who Guidelines on Hand Hygiene

in Health Care. World Health Organization, Geneva, Switzerland.

Wynd C.A., Samstag D.E. & Lapp A.M. (1994) Bacterial carriage

on the fingernails of OR nurses. AORN Journal 60(5), 796, 799–

805.

Yildirim I., Ceyhan M., Cengiz A.B., Bagdat A., Barin C., Kutluk T.

& Gur D. (2008) A prospective comparative study of the rela-

tionship between different types of ring and microbial hand colo-

nization among pediatric intensive care unit nurses. International

Journal of Nursing Studies 45(11), 1572–1576.



The Journal of Advanced Nursing (JAN) is an international, peer-reviewed, scientific journal. JAN contributes to the advancement of

evidence-based nursing, midwifery and health care by disseminating high quality research and scholarship of contemporary relevance

and with potential to advance knowledge for practice, education, management or policy. JAN publishes research reviews, original

research reports and methodological and theoretical papers.

For further information, please visit JAN on the Wiley Online Library website: www.wileyonlinelibrary.com/journal/jan

Reasons to publish your work in JAN:

• High-impact forum: the world’s most cited nursing journal and with an Impact Factor of 1Æ518 – ranked 9th of 70 in the 2010

Thomson Reuters Journal Citation Report (Social Science – Nursing). JAN has been in the top ten every year for a decade.

• Most read nursing journal in the world: over 3 million articles downloaded online per year and accessible in over 7,000 libraries

worldwide (including over 4,000 in developing countries with free or low cost access).

• Fast and easy online submission: online submission at http://mc.manuscriptcentral.com/jan.

• Positive publishing experience: rapid double-blind peer review with constructive feedback.

• Early View: rapid online publication (with doi for referencing) for accepted articles in final form, and fully citable.

• Faster print publication than most competitor journals: as quickly as four months after acceptance, rarely longer than seven months.

• Online Open: the option to pay to make your article freely and openly accessible to non-subscribers upon publication on Wiley

Online Library, as well as the option to deposit the article in your own or your funding agency’s preferred archive (e.g. PubMed).



Fagernes & lingaas (2011) Factors interfering with the microflora on hands. journal of advanced...

Documents

Transcript of Fagernes & lingaas (2011) Factors interfering with the microflora on hands. journal of advanced...