Hand Washing and Physicians: How to Get Them Together  • Author(s): Charles Salemi , MD, MPH; M. Teresa Canola , BSN, CIC; Enid K. Eck , RN, MPHSource: Infection Control and Hospital Epidemiology, Vol. 23, No. 1 (January 2002), pp. 32-35Published by: The University of Chicago Press on behalf of The Society for Healthcare Epidemiologyof AmericaStable URL: http://www.jstor.org/stable/10.1086/501965 .

Accessed: 17/05/2014 00:28

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .http://www.jstor.org/page/info/about/policies/terms.jsp

.JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range ofcontent in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new formsof scholarship. For more information about JSTOR, please contact support@jstor.org.

.

The University of Chicago Press and The Society for Healthcare Epidemiology of America are collaboratingwith JSTOR to digitize, preserve and extend access to Infection Control and Hospital Epidemiology.

http://www.jstor.org

This content downloaded from 195.78.109.107 on Sat, 17 May 2014 00:28:37 AMAll use subject to JSTOR Terms and Conditions

32 INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY January 2002

HAND WASHING AND PHYSICIANS: HOW TO GET THEM

TOGETHER

Charles Salemi, MD, MPH; M. Teresa Canola, BSN, CIC; Enid K. Eck, RN, MPH

OBJECTIVE: To determine the motivating and behavioralfactors responsible for improving compliance with hand washingamong physicians.

DESIGN: Five unobtrusive, observational studies record-ing hand washing after direct patient contact, with study resultsreported to physicians.

SETTING: A 450-bed hospital in a health maintenanceorganization with an 18-bed medical–surgical intensive care unit(ICU) and a 12-bed cardiac care unit.

METHODS: An infectious disease physician met individual-ly with participants to report study results and obtain a commitmentto hand washing guidelines. Follow-up interviews were conducted toevaluate behavioral factors and educational programs. Hand wash-ing study results were presented to all staff physicians by live andvideotaped inservice presentations and electronic mail (e-mail)newsletters. The importance of influencing factors and the educa-tional effectiveness of the hand washing program were evaluated.

RESULTS: Five observational hand washing studies wereconducted in the ICU between April 1999 and September 2000.Rates of physician compliance with hand washing were 19%, 85%,76%, 74%, and 68%, respectively. There were 71 initial encounters

and 55 follow-up interviews with the same physicians. Physicianinterviews revealed that 73% remembered the initial encounter,70% remembered the hand washing inservice presentations, and18% remembered the e-mail newsletters. Personal commitmentand meeting with an infectious disease physician had the mostinfluence on hand washing behavior. Direct inservice presenta-tions (either live or videotaped) had more influence than did e-mailinformation. Rates of ventilator-associated pneumonia did not sig-nificantly change before and during the study periods. A decreasein the rate of central-line–related bloodstream infections from 3.2to 1.4 per 1,000 central-line days was found, but could not be sole-ly attributed to improved physician compliance with hand washing.

CONCLUSIONS: Physician compliance with hand wash-ing can improve. Personal encounters, direct meetings with aninfectious disease physician, and videotaped presentations had thegreatest impact on physician compliance with hand washing at ourmedical center, compared with newsletters sent via e-mail. Localdata on compliance with hand washing and physician involvementare factors to be considered for physician hand washing compli-ance programs in other medical centers (Infect Control HospEpidemiol 2002;23:32-35).

Hand washing is one of the cornerstone activities toprevent nosocomial infections (NIs). Many studies of handwashing have shown an average of less than 50% compliancein intensive care units (ICUs), with nurses having highercompliance rates than physicians.1-9 Since the initial study bySemmelweis, there have not been many studies showing thathand washing has decreased rates of NIs.10,11 There havebeen no long-term studies focusing primarily on physicians.Our observational and educational study focused mainly onphysicians, with a long-term follow-up of nearly 2 years.

METHODS

Five unobtrusive, observational studies were per-formed in an 18-bed medical–surgical ICU and a 12-bed car-diac care unit in April 1998, October 1998, May 1999,November 1999, and September 2000. These studies wereperformed by healthcare workers in the ICU, usually nurs-es on modified sick leave who were performing additionalquality improvement functions. These healthcare workerswere trained by infection control professionals. They useda data collection form that contained criteria consistentwith standard hand washing guidelines to define accept-

able hand washing activities.12 These criteria included handwashing with soap, water, and direct friction for a minimumof 5 seconds after observed direct patient contact; not con-sidering activities such as contact with inanimate objects,handling medications, or delivery of a food tray as directpatient care; and if a patient was on special precautions forvancomycin-resistant enterococci, then considering contactwith inanimate objects an indication for hand washing.

The hand washing observations were done primarilyduring the day and evening shifts on weekdays. The dura-tion of each study was usually 1 to 2 weeks and the maingoal was to record the same number of hand washingobservations (approximately 250) for each study ratherthan a specific duration of the study. There was no record-ing of individual physicians or physician departments, or ofwhich hand washing products were used. The units studiedhad a total of 6 sinks that were located immediately outsidethe patients’ rooms and at the nurses’ stations. There wereno changes in nursing staffing ratios or hand washing prod-ucts during each study period. The hand hygiene productsavailable were bland soap, antimicrobial soap, and handlotion. An alcohol gel product was not used.

The authors are from Kaiser Foundation Hospital, Fontana, California.Address reprint requests to Charles Salemi, MD, MPH, Kaiser Foundation Hospital, 9961 Sierra Avenue, Fontana, CA 92335.

ABSTRACT

This content downloaded from 195.78.109.107 on Sat, 17 May 2014 00:28:37 AMAll use subject to JSTOR Terms and Conditions

Vol. 23 No. 1 HAND WASHING AND PHYSICIANS 33

An infectious disease physician had a face-to-faceinteractive meeting with 71 staff physicians during theirrounds of the ICU in May 1998. They were informed of theresults of the first hand washing observational study con-ducted in April (19% compliance by physicians comparedwith 73% compliance by nurses). The importance of handwashing was emphasized and a personal commitment tohand washing was obtained. These 71 staff physicians com-prised most of the attending physicians using the ICU.They were mainly from the departments of internal medi-cine, family medicine, surgery, and orthopedics. Familymedicine resident physicians and surgical residents froman outside program were not included in this study.

As part of the educational feedback during the courseof the study, two e-mail newsletters were sent, one in May1998 and the other in November 1998, to provide the resultsof the hand washing studies and indications for hand wash-ing. A second interview of 55 of the initial 71 physicians wasconducted in September 1999. Of the 55 physicians, 40recalled the initial presentation and interview. Additionalinformation was obtained from these 40 physicians to deter-mine specifically what educational approaches they recalledand to assess possible motivators for changes in hand wash-ing behavior, the influence of educational approaches, andbehavioral factors that influenced the improved compliance.

Our medical center conducts annual inservice pre-sentations on blood-borne pathogens for physicians, physi-cian assistants, and nurse practitioners as mandated by theOccupational Safety and Health Administration (OSHA).During these sessions, speakers from the departments ofinfection control, employee health, and infectious diseasediscuss blood-borne and airborne pathogen transmission.Inservice sessions were presented in April 1999 and April2000. The total time of each inservice presentation wasapproximately 45 minutes. There was usually a 10- to 15-minute period for presenting the indications for hand wash-ing and the results of the hand washing studies in the ICU.The initial presentation was live and videotaped for laterOSHA inservice presentations. The taping was donebecause of the large number of physicians involved and theneed for at least four OSHA inservice presentations.Infection control personnel and infectious disease physi-cians were present at each of the videotaped presentationsto administer the required posttest and to answer ques-tions. If a physician was unable to attend a scheduled inser-vice, he or she was required to view the videotape andanswer a posttest questionnaire. Five hundred physicians

(94% of the active physician staff) participated in this OSHAinservice presentation.

RESULTS

The observational studies were performed duringApril 1998, October 1998, May 1999, November 1999, andSeptember 2000. The rates of physician compliance withhand washing were 19%, 85%, 76%, 74%, and 68%, respec-tively. Data on compliance with hand washing and thechronology of events for the physicians and other healthcare workers are shown in the table and the figure.

The second interview included 55 (77%) of the initialphysicians contacted. Of these, 40 (73%) recalled the initialpresentation and interview. These 40 physicians wereincluded in the interview data. Seventy percent of themremembered the OSHA inservice presentation, but only18% remembered the e-mail newsletter.

During the second interview, physicians answeredquestions to determine factors that influenced compliancewith hand washing. A Likert scale was used to evaluatetheir answers, with 1 representing the least influence and 5representing the most influence. The following four factors,ranked in order of importance, had the most influence oncompliance with hand washing: (1) knowledge that handwashing prevents NIs (average score, 4.4); (2) poor physi-cian rates compared with nurse rates (average score, 3.7);(3) personal commitment to hand washing (average score,3.3); and (4) knowledge that repeat studies would be con-ducted (average score, 3.0).

FIGURE. Compliance with hand washing in relation to interventions. 1 =One-on-one interviews with commitment to hand washing with 71 physicians,May 1998; 2 = e-mail newsletter of initial hand washing data, May 1998;3 = e-mail newsletter of follow-up hand washing data, November 1998;4 = mandatory physician educational inservice presentation (live and video-taped), April 1999; 5 = follow-up interviews with 55 of the physicians initiallyinterviewed, September 1999; 6 = mandatory physician educational inser-vice presentation (live and videotaped), April 2000.

TABLEOBSERVATIONS OF COMPLIANCE WITH HAND WASHING

Apr 1998 Oct 1998 May 1999 Nov 1999 Sept 2000

Physicians 19% (9/48) 85% (40/47) 76% (26/34) 74% (31/42) 68% (26/38)Nurses 73% (78/107) 90% (150/167) 71% (141/198) 84% (93/111) 58% (76/131)Respiratory therapists 43% (36/84) 80% (20/25) 77% (24/31) 50% (14/28) 38% (7/18)Total 50% (123/245) 88% (210/239) 73% (191/263) 76% (138/181) 58% (109/187)

This content downloaded from 195.78.109.107 on Sat, 17 May 2014 00:28:37 AMAll use subject to JSTOR Terms and Conditions

34 INFECTION CONTROL AND HOSPITAL EPIDEMIOLOGY January 2002

Regarding the source of hand washing informationpresented to physicians, the infectious disease physicianhad the most influence compared with the infection controlprofessional or the ICU nurse. The following healthcareworkers, ranked in order of importance, had the most influ-ence on compliance with hand washing: infectious diseasephysicians (average score, 3.9); infection control profes-sionals (average score, 2.9); and ICU nurses (averagescore, 2.7).

Throughout the study period, the infection control NIsurveillance in critical care included central-line–relatedbloodstream infections and ventilator-associated pneumoniarates based on Centers for Disease Control and Preventioncriteria comparing our hospital’s rates with the NationalNosocomial Infections Surveillance (NNIS) System data.The rates were tabulated by quarters using NNIS Systemdata for a medical–surgical ICU in a nonteaching hospital.13

No significant difference in the ventilator-associated pneu-monia rates was found during the study period. The ratesfor ventilator-associated pneumonia remained at the 25thpercentile of NNIS System criteria. However, there was animprovement in the rates of central-line–related blood-stream infections. The average rate of central-line–relatedbloodstream infections before intervention was 3.0 versus1.4 per 1,000 central-line days postintervention.

DISCUSSION

Many articles have reported efforts to improve handwashing practices through educational and behavioralstrategies.1-8,10,14-17 These articles mainly involved nurses,and no article specifically addressed physicians. In a recentarticle, Pittet acknowledged the difficulty of achievingincreased compliance with hand washing.14 The major chal-lenge is sustaining the modified hand washing behavioruntil it becomes a “routine” activity.16 Some of the reportedreasons for poor compliance are skin irritation from theproduct, patient priority needs, or inadequate hand wash-ing facilities (ie, sinks or stations).

Comprehensive approaches combining research, edu-cation, and administrative strategies are necessary toimprove hand washing behavior among healthcare work-ers.3,16 Infection control teams have a unique opportunity tocontribute to the necessary research component by compil-ing accurate data on hand washing and NIs to facilitate physi-cian education that promotes changes in behavior.

Although observation and feedback are the methodsmost frequently reported to promote changes in behavior,3,14

behavioral modification is a complex process. Health educa-tion models of theories of behavior change are extensivelydiscussed in the literature, with many examples of educationprograms, resources, and strategies. Comprehensive pro-grams that incorporate information, motivation, and rein-forcement are essential to sustain changes in behavior.

Administrative interventions must include enhanc-ing the organization’s or the institution’s support for theinfection control program, educating physicians and staff,and ensuring the availability of a user-friendly physicalplant and resources for hand hygiene products.3

Increased NI rates in the ICU have been associatedwith poor compliance with hand washing. However, the fullimpact on ICU deaths cannot be adequately studied becauseof the ethical limitations associated with a control group nothaving adequate hand washing practices.4 These study lim-itations may affect the ability to identify short-term causalrelationships between reductions in the rate of NIs inresponse to specific strategies that improve hand washing.A direct association cannot be made between the reductionin the rate of central-line–related bloodstream infections andimproved compliance with hand washing because of themultifactorial nature of central-line–related bloodstreaminfections.

Physicians play an important role in improving com-pliance with hand washing because they are reinforcingrole models for other physicians.18 They are unlikely toattend any voluntary meetings on this topic, so interven-tions that ensure physician interaction can be difficult.15 Inour study, face-to-face interactions between physiciansplayed a pivotal role in improving behavior. The initial inter-view of 71 physicians included most of the staff in thedepartments of internal medicine, family medicine,surgery, and orthopedics who used the ICU. Medical stu-dents and residents were not included, mainly because wecould not ensure their availability during this long-termstudy.

The strengths and weaknesses of this study are asso-ciated with the involvement of infectious disease physi-cians. Physicians can exert peer pressure regarding theimportance of hand washing. Conducting a total of 126 face-to-face meetings with physicians required a major timecommitment that may not be possible in other medical cen-ters. Although these 126 encounters involved only severalminutes, infectious disease physicians spent a considerableamount of time waiting for physicians in the ICU for initialand follow-up interactions and keeping records so theythey did not interview the same physicians twice.

One of the interesting findings was that nearly thesame percentage of physicians remembered the OSHAinservices and the initial physician interview, whereas only18% recalled either of the e-mail educational newslettersabout hand washing results and criteria. Of the four recentOSHA inservice presentations, only one was live. The otherthree were videotaped presentations with infection controland infectious disease staff present to answer any questions.Our conclusion is that, at our medical center, direct presen-tations (either live or videotaped) are a more effective edu-cational tool than e-mail.

There is also a potentially significant interviewer biaswhen physicians are answering questions regarding theimportance of hand washing to prevent NIs or the sourcesof information that influence their hand washing behavior.There is probably not as much bias with questions aboutremembering meetings, previous interviews, or e-mails.Most of the physicians knew that hand washing preventsNIs, thus a program to inform physicians that hand wash-ing prevents NIs would probably not be effective.

Our definition for hand washing after direct patient

This content downloaded from 195.78.109.107 on Sat, 17 May 2014 00:28:37 AMAll use subject to JSTOR Terms and Conditions

Vol. 23 No. 1 HAND WASHING AND PHYSICIANS 35

contact is consistent with that of most published studies,including any contact with inanimate objects from patientson special infection control precautions (ie, contact precau-tions). To increase physician acceptance of the hand wash-ing study data, no data on hand washing were includedunless they met the study definition for direct patient con-tact. Data on hand washing were also reported for nursingand respiratory therapy staff to facilitate viewing the dataon physician hand washing in context. The initial poor rateof compliance with hand washing among physicians of 19%compared with 73% among nurses was a strong motivatingfactor for physicians to improve. In the final observationalstudy in September 2000, the rate of hand washing amongphysicians was 68%, which was higher than that amongnurses (58%).

At the onset of this study, there was an initial markedimprovement in the hand washing rate from 19% to 85%, butthere was a a gradual decrease in the hand washing rate to68% in September 2000. This reduction was not unexpect-ed, because our hand washing program has not had anyfurther face-to-face physician interactions for more than ayear and the last OSHA inservice presentation was 4months earlier in April 2000. The nursing and respiratorytherapy departments have addressed their hand washingcompliance through quality improvement activities.

Administrative support in our studies was best illus-trated by allowing ample time for infectious disease physi-cians to interact with physicians and instituting mandatoryOSHA inservice presentations for physicians where handwashing data and guidelines were presented.

As part of a national health maintenance organizationmedical care program, our medical center served as a pilotsite for an organization-wide hand washing initiative. Withincorporation of the successful strategies identified in theoriginal intervention, there is an opportunity to build on theexperience of the study designers and integrate these find-ings into a broader hand washing program. We developedrecommendations to improve physician motivation toimplement local hand washing programs based on our find-ings. Administrative mandates for hand washing will not bea strong motivating factor. Recognized physician leadersare needed as physician champions. Local baseline and fol-low-up data on hand washing will be the strongest motivat-ing factor for physicians. Other pertinent factors for con-sideration are that physicians are mainly motivated by theirresponsibility to provide quality care to their patients andthat they have to be shown that their actions or inactionshave adverse effects.

The final phase of the expanded national hand wash-ing program will be an extensive education program that

incorporates a focused approach and face-to-face interactionamong physicians. Live and videotaped programs will beused to provide education about hand washing compliance.Educational materials will be reinforced through operatingpolicies and procedures, posters and signage, and rewardand recognition initiatives. We documented that most physi-cians acknowledge that hand washing prevents NIs, thatphysicians are highly motivated by their poor performanceas compared with that of nurses, and that a personal com-mitment can be an effective tool to improve compliance withhand washing. Infectious disease physicians are role modelsand their discussion of hand washing data is the most effec-tive source of information for physicians.

REFERENCES1. Pittet D, Mourouga P, Perneger TV, the Infection Control Program.

Compliance with handwashing in a teaching hospital. Ann Intern Med1999;130:126-130.

2. Bischoff WE, Reynolds TM, Sesser CN, Edmond MB, Wenzel RP.Handwashing compliance by health care workers: the impact of intro-ducing assessable alcohol-based hand antiseptics. Arch Intern Med2000;160:1017-1021.

3. Larson EL, Bryan JL, Adler CM, Blane C. A multifaceted approach tochanging handwashing behavior. Am J Infect Control 1997;25:3-10.

4. Nyström B. Impact of handwashing on mortality in intensive care: exam-ination of the evidence. Infect Control Hosp Epidemiol 1994;15:435-436.

5. Farr BM. Reasons for noncompliance with infection control guidelines.Infect Control Hosp Epidemiol 2000;21:411-416.

6. Albert RK, Condie F. Hand-washing patterns in medical intensive careunits. N Engl J Med 1981;304:1465-1466.

7. Dubbert PM, Dolce J, Richter W, Muller M, Chapman W. IncreasingICU staff handwashing: effects of education and group feedback. InfectControl Hosp Epidemiol 1990;11:191-193.

8. Goldmann D, Larson E. Hand washing and nosocomial infection. N EnglJ Med 1992;327:120-122.

9. Watanakunakorn C, Wang C, Hazy J. An observational study of handwashing and infection control practices by healthcare workers. InfectControl Hosp Epidemiol 1998;19:858-860.

10. Larson E. Skin hygiene and infection prevention: more of the same ordifferent approaches? Clin Infect Dis 1999;29:1287-1294.

11. Doebbeling BN. Comparative efficacy of alternative hand-washingagents in reducing nosocomial infections in intensive care units. N EnglJ Med 1992;327:88-93.

12. Larson EL, 1992, 1993, and 1994 APIC Guidelines Committee. APICguideline for handwashing and hand antisepsis in health care settings.Am J Infect Control 1995;23:251-269.

13. Richards MJ, Edwards JR, Culver DH, Gaynes RP, the NationalNosocomial Infections Surveillance System. Nosocomial infections incombined medical-surgical intensive care units in the United States.Infect Control Hosp Epidemiol 2000;21:510-515.

14. Pittet D. Improving compliance with hand hygiene in hospitals. InfectControl Hosp Epidemiol 2000;21:381-386.

15. Pittet D, Hugonnet S, Harbarth S, et al. Effectiveness of a hospital-wideprogramme to improve compliance with hand hygiene. Lancet2000;356:1307-1312.

16. Kretzer EK, Larson EL. Behavioral intervention to improve infectioncontrol practices. Am J Infect Control 1998;26:245-253.

17. Simmons B, Bryant J, Neiman K, Spencer L, Arheart K. The role ofhandwashing in prevention of endemic intensive care unit infections.Infect Control Hosp Epidemiol 1990;11:589-594.

18. Handwashing Liaison Group. Hand washing, a modest measure—withbig effects. BMJ 1999;318:686.

This content downloaded from 195.78.109.107 on Sat, 17 May 2014 00:28:37 AMAll use subject to JSTOR Terms and Conditions