J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from Compliance With Universal Precautions In A Medical Practice With A High Rate Of HIV Infection

Sandra W Freeman, R.N., B.S.N., and Christopher V. Chambers, M.D.

AbsIrYlet: Btlcllgnnlllll: Universal precautions have been recommended to limit occupational exposure to the human immunodeftdency virus (HIV) and other infectious agents, but whether these recommendadons have been incorporated into routine pradice has not been demonstrated. Metbotls: Using a one-group, before-after design, we assessed the knowledge and attitudes concerning universal precautions and the level of compliance with these recommendations. 1be health care professionals had various levels of training and worked in an ambulatory practice with a high rate of mv. A total of 195 procedures involving potential exposure to various body 80ids were observed. Renin: No improvement in compliance with recommended precautions was observed following a didactic educational program for either 1atex glove use (44 percent versus 49 percent, X2 < 1, P > 0.2) or appropriate use of hand washing (34 percent versus 47 percent, X2 = 3.38, P = 0.07). Faculty demonstrated the lowest levels of adherence to universal precautions. While knowledge of precautions was high, s1aff members at all levels overestimated their own compliance with these recommendations. em.elflSlmls: Although the number of observations limits the conclusions, the results suaest that the basic protective measures included in universal precautions are not being routinely applied in ambulatory mecIkaI practice. Furthermore, didactic educational programs might not be suftlclent to improve compliance. Finally, fawlty in 1raining programs should monitor their own compliance with universal precautions because of their responsibilities as role models for physidans in 1raining. (J Am Board Fam Prad 1992; 5:313-8.)

In addition to its devastating effects on patients, In response to. increasing concerns regarding the epidemic of hwnan immunodeficiency virus the infectivity of HIV and hepatitis B virus (HIV) infection and the acquired immunodefi­ (HBV), universal precautions have been pro­ ciency syndrome (AIDS) has posed new problems posed by the CDC for use in all health care cen­ http://www.jabfm.org/ for health care providers. Concerns about pos­ ters in the United States to supplement existing sible acquisition of HIV infection in the work­ infection control policies.3,4 According to these place are supported by well-docwnented case re­ guidelines, health care workers should treat blood ports of occupational exposure to body fluids with and a limited nwnber of other body fluids from all subsequent seroconversion in health care workers patients as potential sources of HIV and HBV who have had no other recognized risk factors. l infections.3,4

Based on a prospective study of health care work­ Previous investigatorss found poor compliance on 24 September 2021 by guest. Protected copyright. ers occupationally exposed to HIv, the Centers with these guidelines by health care workers hav­ for Disease Control (CDC) has estimated that the ing regular contact with potentially infected pa­ risk of acquiring HIV after a needle-stick expo­ tients, and their study was terminated so that an sure from a known carrier is approximately 0.4 educational program could be implemented to percent.2 The risk of HIV infection from other motivate employees to use safer behaviors. In­ types of occupational exposure is unknown, but service education has improved knowledge and probably less. changed attitudes regarding the risks of acquiring HIV infection in other settings,6,7 but meaning­ ful changes in behavior have been more difficult Submitted, revised, 27 January 1992. to establish. Continuing medical education pro­ From the Department of Family Medicine, Jefferson Medical grams have had limited success in changing phy­ College, Thomas Jefferson University, Philadelphia. Address re­ sician practice.8,9 print requests to Christopher V. Chambers, M.D., Department of Family Medicine, Thomas Jefferson University, 1015 Walnut The primary goal of this study was to assess the Street, Room 401, Philadelphia, PA 19107. knowledge and attitudes concerning universal

Universal Precautions 313 J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from precautions and the level of compliance with presence was unobtrusive to the staff, who were, these recommendations by health care profes­ by design, unaware that observations were being sionals at various levels of training in a family made. For each potential exposure, she recorded medicine residency program. The outpatient the training level of the person observed, the practice where this study was conducted is located body fluid for which potential contact was antici­ in a large metropolitan hospital in the mid­ pated, and the appropriate use of gloves and hand Atlantic region of the United States. Epidemio­ washing. Observations were made for 1 hour each logic data from the CDC have suggested that the day in the office laboratory and at nursing stations number of mY-infected patients is likely to be on a rotating schedule in an attempt to minimize higher in this region than in other geographic any systematic bias in the selection of staff areas of the United States.l° According to the observed. computerized practice database, more than 200 At the end of this observation period, a ques­ mY-infected patients have received ongoing care tionnaire was distributed to all the office staff at a in this office. AIDS also has been the most com­ regularly scheduled conference. Ten questions mon admitting diagnosis to the practice inpatient addressed attitudes and perceptions regarding service since 1987. Given the high visibility of staff behaviors, and ten items explored knowledge AIDS-related care in the office and the formal of universal precautions and isolation procedures and informal teaching relating to the clinical care in the family medicine setting. Response options of mY-infected patients, we hypothesized that consisted of 5-point scales that ranged from compliance with universal precautions would be strongly disagree to strongly agree. high and that physicians with the most training The intervention consisted of a didactic lecture and experience in caring for mY-infected pa­ given by the second author after collecting the tients would be the most compliant with these questionnaires at the staff meeting. The risks of recommendations. A second goal of this study was occupational exposure were reviewed using ma­ to measure any changes in compliance with uni­ terials provided by the CDC (Hospital Infections versal precautions following a didactic, educa­ Program, Health Care Worker Surveillance tional program designed to make clinicians more Project, August 15, 1983-April 20, 1989), and aware of these recommendations. the importance of universal precautions was reinforced. Methods Compliance with universal precautions was http://www.jabfm.org/ This study was conducted in a university-based measured in the same manner for the second family practice outpatient office, where 12 at­ 3-week period, after which the questionnaire was tending physicians and 18 residents provide on­ again distributed. going primary care to approximately 10,000 ac­ Data from the clinical observations and the tive patients. An average of 8 3rd-year medical questionnaires were analyzed, and the chi-square

students rotate through the office every 6 weeks statistic was used to compare preintervention and on 24 September 2021 by guest. Protected copyright. for a required clerkship. Eight registered nurses postintervention compliance. Because of small and 5 medical technicians perform venipuncture subgroup sizes, tests of statistical significance and assist with other procedures involving poten­ were not used in the subgroup analyses. Selected tial exposure to various patient body fluids. data from the self-administered questionnaires Data for this study were collected over two are summarized in Table 1, also without statistical consecutive 3-week periods using a one-group, reference, because the preintervention and before and after (preexperimental) design. The postintervention groups did not consist entirely 6-week study duration was chosen to allow inclu­ of the same individuals. sion of the same group of medical students in the observation periods both before and after the Results educational intervention. In the first 3 weeks, Ninety-seven potential exposures to various body single-blinded observations of behaviors related fluids were observed in the preintervention pe­ to universal precautions were recorded. The ob­ riod, and 98 exposures were observed during the server was a registered nurse who also served as post-intervention period. Of al1195 observations, the coordinator for clinical trials in the office. Her 163 (84 percent) involved the handling of open

314 JABFP May-June 1992 Vol. 5 No.3 J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from

Attitudes Mean Preintervention and Postintenention Values* Attending Resident Medical Medical Physician Physician Student Nurse Technician Pre Post Pre Post Pre Post Pre Post Pre Post (n -8)(n .. 5) (n-14) (n=l3) (n .. 10) (n .. 9) (n = 3) (n =3) (n =4) (n .. 4) I feel I understand universal 4.1 4.1 4.1 4.8 4.1 4.2 3.3 3.7 4.3 4.8 precautions All patients should be 3.0 3.2 4.1 4.9 4.7 4.6 5.0 4.7 4.0 5.0 considered potentially infected with mv The facilities make fOllowing 2.8 3.2 3.7 3.4 3.6 3.9 4.3 4.7 4.5 4.8 universal precautions easy I feel my peers fOllow universal 2.5 2.0 2.5 2.1 3.2 2.7 3.0 3.3 4.3 4.0 precautions at all times I feel I follow universal pre­ 3.1 2.8 3.0 3.1 3.5 3.8 4.0 2.7 4.0 4.3 cautions at all times

*Mean values on a scale from 1 .. strongly disagree to 5 • strongly agree. containers or wet specimens, including urine were included in the 34 personnel who completed (93 exposures), vaginal discharge (68), and penile the second questionnaire. In both instruments discharge (4) in the office laboratory. The remain­ factual knowledge of all staff regarding HIV ing 32 (16 percent) involved venipuncture and transmission and universal precautions was high needle disposal. (range, 90 to 100 percent correct). The majority Latex gloves were worn during 44 percent of of the staff (88 percent) agreed or strongly agreed the potential exposures in the preintervention pe­ that they understood universal precautions as out­ riod and 49 percent of the exposures after the lined by the CDC. (fable 1). Consistent with this educational program (X2 < 1, P > 0.2). Glove use understanding, most of the staff (81 percent) was observed during 12 (52 percent) of 23 veni­ agreed or strongly agreed that all patients should

punctures before and 5 (55 percent) of 9 veni­ be considered potentially infected with HIv. The http://www.jabfm.org/ punctures after the intervention. For each of the exceptions were the attending staff, who neither other body fluids, gloves were worn during less agreed nor disagreed with this statement. The than 50 percent of the potential exposures ob­ members of each professional group tended to served. The wearing of latex gloves was examined rate their own compliance with universal precau­ in relation to the professional level of the tions as higher than that of their peers. staff person observed (Figure 1). Attending Staff members were also asked to estimate the physicians who completed their residency train­ percentage of time that they wore gloves for han­ on 24 September 2021 by guest. Protected copyright. ing before 1981 had the lowest rate of glove use dling specific body fluids. Attending and resident and did not improve after the educational physicians estimated that they wore gloves when program. handling blood 61 to 80 percent of the time, Appropriate hand washing was observed fol­ whereas students, nurses, and technicians esti­ lowing 34 percent of the potential exposures mated their own use at between 81 and 99 per­ in the preintervention period and following 47 cent. Overall, attending physicians judged them­ percent after the educational program (X2 = 3.38, selves as least compliant of all staff groups, and P = 0.07). When blood specimens were handled, technicians estimated themselves as most compli­ hand washing was recorded for only 28 percent of ant regarding glove usage. the observations made. Rates of hand washing were similar for attending physicians, residents, Discussion students, and technicians. Although universal precautions are recom­ A total of 39 personnel completed the first mended for all medical settings to prevent trans­ questionnaire, and most of these same individuals mission of infectious agents, including mY,3,4

Universal Precautions 315 100 J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from _ Preintervention n=11 ~ Postintervention 80

....c Q) u ~ Q) a.

20 o Attending t Attending * Nurse Resident Student Technician Latex gloves worn

Figure 1. Percen1age of obsenations for which latex gloves were wom during potential body fluid exposure. *lnadequate number of obsenations for a postintervention measurement. tResidency completed before 1981. ~idency completed after 1981. there are few data that show whether these rec­ versal precautions can improve following an ommendations have been adopted into routine educational intervention,6 a didactic program practice. In this study in an ambulatory family might not be sufficient to change how routine practice setting, both professional and nonprofes­ clinical procedures are performed. Results of sional staff demonstrated poor compliance with research in related areas have indicated that http://www.jabfm.org/ glove use and hand-washing recommendations. more creative educational interventions might Given the large number of patients receiving be necessary to change these behaviors.8,9 We treatment for AIDS and other InV-related con­ are currendy evaluating the effect of providing ditions in this medical practice, the finding that feedback to physicians regarding their compli­ the basic protective measures of glove use and ance with universal precautions relative to their

hand washing were employed for less than 50 peers. on 24 September 2021 by guest. Protected copyright. percent of the potential body fluid exposures was A disturbing finding in this study was the in­ surprising and suggests that clinicians have not verse relation between years of formal education made these practices routine. and adherence to universal precautions. Modeling Authors of a previous study of adherence to by faculty is an important part of the medical universal precautions by medical staff in an emer­ training of medical students and residents. The gency department reported similar levels of com­ clinical skills and behaviors of attending physi­ pliance. 5 Their study was aborted prematurely so cians have a strong influence on residents' sub­ that an in-service training program for employees sequent practice behaviors. ll,12 In this study could be implemented to effect changes in com­ attending staff were less compliant with recom­ pliance. In our study, substantive changes in be­ mendations for latex glove use than either resi­ haviors were not observed after a traditional edu­ dents or students. Faculty who completed their cational program. There are limited published residency training prior to the recognition of data that address whether medical education pro­ AIDS (i.e., before 1981) demonstrated the lowest grams can alter physician behavior.8,9 Although levels of compliance with universal precautions. provider knowledge and attitudes regarding uni- Because of the recency of the InV epidemic,

316 JABFP May-June 1992 Vol. 5 No.3 J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from senior clinicians might not be appropriate role tially infectious to eliminate a clinical judgment models for medical students and residents, who prior to taking appropriate protective measures. are learning to incorporate universal precau­ Excluding the observations relating to urine tions (formally recommended in 1987) into their would reduce further the already small number of clinical routine. Quill 11 has suggested that a true exposures in this study. Even so, although the special effort should be made to educate or re­ sample size limits the conclusions, glove use and educate physicians who serve as preceptors and hand washing were observed to be low for all body role models in areas where they demonstrated fluids. Finally, the before-after study design did low compliance. Researchers13 involved in a not control for exposure to those influences, other large survey of primary care physicians in Eng­ than the planned educational intervention, that land suggested that physicians in general prac­ might have affected compliance with universal tice are reluctant to change their infection precautions. These limitations do not diminish control procedures. Physicians in training pro­ the finding that basic protective measures against grams, however, might need to monitor their HIV infection are not being routinely used by own behaviors more carefully because of their health care personnel in the ambulatory setting. responsibilities as role models for physicians in This study should be replicated with larger sam­ training. ple sizes to assess more adequately the interac­ All staff and trainees tended to overestimate tions of factors that underlie this poor compliance their compliance with universal precautions. An before programs to change behavior can be interesting observation was that medical stu­ designed. dents, residents, and attending staff all stated that their own compliance with universal pre­ cautions was greater than that of their peers. A study of emergency department personnel also References showed that staff significantly overestimated 1. Centers for Disease Control. Update: acquired immunodeficiency syndrome and human immuno­ their compliance with these recommend a­ deficiency virus infection among health care work­ tions. 14 We did not examine the reasons for ers. MMWR 1988; 37:229-39. noncompliance with universal precautions by 2. Idem. Public Health Service statement on manage­ the medical personnel in our office. Trauma ment of occupational exposure to human immuno­

center personnel have indicated that there deficiency virus, including considerations regarding http://www.jabfm.org/ zidovudine postexposure use. MMWR 1990; often is inadequate time for donning appropri­ 39(RR-l):1-14. ate protective clothing and that cumbersome 3. Idem. Recommendations fur prevention of HN gloves, masks, and goggles can reduce dexterity transmission in health-care settings. MMWR 1987; during invasive procedures. 14 These problems 36(SuppI2):IS-18S. would seem to be less applicable to the proce­ 4. Idem. Update: universal precautions for prevention dures routinely performed in an ambulatory of transmission of human immunodeficiency virus, hepatitis B virus, and other bloodbome pathogens in on 24 September 2021 by guest. Protected copyright. practice. There is a general concern that the health-care settings. MMWR 1988; 37:377-82, protection provided by latex gloves is im­ 387-8. perfect, particularly with respect to needle 5. BaraffLJ, Talan DA. Compliance with universal pre­ sticks.4,5 Physicians in primary care, as a result cautions in a university hospital emergency depart­ of their long-term relationships with pa­ ment. Ann Emerg Med 1989; 18:654-7. 6. Wertz DC, Sounson JR, Liebling L, Kessler L, tients, might individualize the risk assessment Heeren TC. Knowledge and attitudes of AIDS 14 for HIV transmission with each patient and health care providers before and after education pro­ act accordingly. grams. Public Health Rep 1987; 102:248-54. There were methodologic limitations inherent 7. O'Donnell L, O'Donnell CR Hospital workers and in the design of this study. Strictly speaking, uni­ AIDS: effect of in-service education on knowledge versal precautions as outlined the CDC do not and perceived risks and stresses. NY State J Med by 1987; 87:278-80. include urine as an infectious source unless it is 8. White CWO Albanese MA, Brown DD, Caplan RM. grossly contaminated with blood. Nonetheless, The effectiveness of continuing medical education many hospitals, including ours, have recom­ in changing the behavior of physicians caring for mended that all body fluids be considered poten- patients with acute myocardial infarction. A con-

Universal Precautions 317 J Am Board Fam Pract: first published as 10.3122/jabfm.5.3.313 on 1 May 1992. Downloaded from

trolled randomized trial. Ann Intern Med 1985; 12. Ficklin FL, Browne VL, Powell RC, Carter JE. Fac­ 102:686-92. ulty and house staff' members as role models. J Med 9. UoydJS, Abrahmason S. Effectiveness of continuing Educ 1988; 63:392-6. medical education. Eval Health Professions 1979; 13. Foy C, Gallagher M, Rhodes T, Setters J, Philips 2:251-80. P, Donaldson C, et al. HIV and measures to con­ 10. Centers for Disease Control. 1988 Update: AIDS trol infection in general practice. BMJ 1990; and HIV infection in the United States. MMWR 300:1048-9. 1989; 38(SuppI4). 14. Campbell S, Maki M, Henry K. Compliance with 11. Quill TE. Medical resident education. A cross­ universal precautions among emergency depart­ sectional study of the influence of the ambulatory ment personnel. Presented at the Sixth Interna­ preceptor as a role model. Arch Intern Med 1987; tional Conference on AIDS. San Francisco. June 147:971-3. 20-24,1990. http://www.jabfm.org/ on 24 September 2021 by guest. Protected copyright.

318 JABFP May-June 1992 Vol. 5 No.3