HIV postexposure prophylaxis practices by US ED practitioners

Brief Reports

HIV Postexposure Prophylaxis Practices by US ED Practitioners ROLAND C. MERCHANT, MD AND REZA KESHAVARZ, MD, MPH To determine how often US ED practitioners have prescribed HIV postexposure prophylaxis (HIV PEP) and to discern how willing they are to offer it to patients, the authors surveyed 600 ED practitioners attending a national conference. According to their self-report, 11% had taken HIV PEP themselves. Sixty-eight percent had prescribed HIV PEP at some time. Of these, 92% had treated needlestick-injured health care workers, 48% sexual assault survivors, and 49% nonhealth care needlestick-injured persons. ED practitioners were more willing to offer HIV PEP after exposures to HIV-infected or high-risk sources than unknown or lowrisk sources, as well as after sexual assault than consensual sex. Female practitioners, those who had themselves taken HIV PEP, resident physicians, and ED practitioners with fewer than 6 years of clinical practice were generally more apt to offer HIV PEP. Educational campaigns appear to be necessary to help ED practitioners determine when HIV PEP is appropriate. (Am J Emerg Med 2003;21:309-312. © 2003 Elsevier Inc. All rights reserved.)

HIV post-exposure prophylaxis (HIV PEP) is a form of secondary HIV prevention that may help decrease the transmission of HIV. National guidelines from the Centers for Disease Control and Prevention (CDC) help direct occupational HIV PEP for health care workers.1 Although there are no national guidelines for nonoccupational HIV PEP in the United States, some groups have issued recommendations on its provision after sexual assault and consensual sex exposures as well as needlestick injuries to nonhealth care persons.2-5 Given the need for immediate evaluation and treatment of blood or body fluid exposures, the ED is a likely venue where HIV PEP is used. Little is known, however, of the extent of its use in the ED. Also unknown is how willing ED practitioners are to prescribe HIV PEP after both occupational and nonoccupational exposures. From the Department of Emergency Medicine, Mount Sinai School of Medicine New York, NY. Received July 24, 2002; accepted July 24, 2002. Presented at the First Annual New York State Region Society for Academic Emergency Medicine Regional Meeting, Rochester, NY, March 2001, and the 2001 Scientific Assembly of the American College of Emergency Physicians, Chicago, IL, October 2001. Current affiliation (R.C.M.): Section of Emergency Medicine, Brown University School of Medicine, Providence, RI Address reprint requests to Roland C. Merchant, MD, Rhode Island Hospital, Department of Emergency Medicine 593 Eddy Street, Davol 141, Providence, RI 02903. E-mail: rmerchant@ lifespan.org. Key Words: HIV post-exposure prophylaxis, HIV prevention, HIV, prophylaxis, post-exposure prophylaxis. © 2003 Elsevier Inc. All rights reserved. 0735-6757/03/2104-0011$30.00/0 doi:10.1016/S0735-6757(02)42242-5

Through a survey of ED practitioners attending a national EM conference, we aimed to describe ED practitioners’ self-reported prescribing of HIV PEP, their use of it for themselves (self-use), and their willingness to offer HIV PEP in clinical scenarios that describe typical ED patient visits for needlestick injuries and sexual encounters. METHODS The authors created an anonymous, written, voluntary, structured, self-administered questionnaire concerning different forms of emergency prophylaxis for a convenience sample study of ED practitioners. The survey was pilot tested on a small cohort of practitioners at Mount Sinai Hospital. The 20-item, multiple-part, closed-response questionnaire included the following: (1) practitioner demographic questions; (2) 7 clinical scenarios describing different patients who present to the ED within 1 hour of a needlestick injury, sexual assault, or consensual sex; and (3) questions regarding practitioner HIV PEP prescribing history and HIV PEP self-use. The Institutional Review Board of Mount Sinai School of Medicine approved the study. For the clinical scenarios, the practitioners were asked to indicate if they would offer HIV PEP to different patients (health care worker or nonhealth care worker adult or child) after specific exposures (needlestick, sexual assault, or consensual sex) and varied HIV risk levels (unknown, HIVinfected, high HIV risk factors, defined as injecting drug use, male-male sex, prostitution, or low HIV risk factors). To mimic the information typically known on initial ED presentation, the questionnaire specified that no further information was available (ie, the source could not be tested or queried). The questionnaire also stated that each exposure was significant (eg, bloody needlestick or unprotected sexual intercourse), and each patient was an ideal HIV PEP candidate (not pregnant and not HIV infected). The survey was conducted at the American College of Emergency Physicians 2000 Scientific Assembly in Philadelphia, Pennsylvania. Over 4,000 people attended the meeting. All ED practitioners who could prescribe medications (physicians, nurse practitioners, and physician assistants) and who visited the survey booth in the exhibition hall were asked to complete a questionnaire. ED practitioners were offered the chance to enter a $100 gift certificate lottery if they completed a questionnaire. A researcher was available to the respondents to clarify questions. The data were abstracted from the questionnaires and entered into SPSS 10.0 (SPSS, Chicago, IL). Only surveys over 90% complete were used. The statistical analyses in 309

AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 21, Number 4 ■ July 2003

310

TABLE 1.

ED Practitioner Self-Reported HIV PEP Prescribing History

ED practitioners who have prescribed HIV PEP (n ⴝ 600) To any age patient To patients under age 18 Circumstances for prescribing HIV PEP To any age patient (n ⴝ 395) Sexual assault Needlestick or sharp injury Health care worker Nonhealth care person Consensual sex Other (eg, blood splash) To patients under age 18 (n ⴝ 77) Sexual assault Needlestick or sharp injury Health care worker Nonhealth care person Consensual sex ED practitioners who have taken HIV PEP (n ⴝ 600) Circumstances (n ⴝ 61) Sexual assault Health care setting needlestick/sharp injury ED Practitioners Reporting Most Often Prescribed HIV PEP Regimen for a Given Exposure*

Zidovudine Zidovudine Zidovudine Zidovudine

alone ⫹ lamivudine ⫹ lamivudine ⫹ indinavir ⫹ lamivudine ⫹ nelfinavir

No.

Percent

395 77

68% 19%

198

48%

377 201 31 10

92% 49% 8% 2%

52

69%

6 23 1 61

8% 31% 1% 11%

4 60

7% 98%

Sexual Assault

Needlestick Health care Worker

Needlestick nonhealth care Person

Consensual Sex

40/399 (10%) 95/399 (24%) 110/399 (28%) 49/399 (12%)

31/406 (8%) 119/406 (29%) 160/406 (39%) 64/406 (16%)

25/398 (6%) 88/398 (22%) 110/398 (28%) 52/398 (13%)

18/391 (5%) 50/391 (13%) 46/391 (12%) 25/391 (6%)

*The remaining percentages were “other” and “I have never prescribed for this kind of exposure.”

SPSS 10.0 included generation of point estimates of proportions and 95% confidence intervals of these. RESULTS Six hundred questionnaires were over 90% complete and were used in the analysis. Of the respondents, 71% were male and 29% were female. Thirty-four percent were attending physicians in a teaching hospital, 26% in a community hospital, 33% were residents, and 7% were nurse practitioners or physician assistants. Sixty-one percent of the physicians were board-certified or eligible in EM, 3% in internal medicine, and 3% in family practice. Ninety-five percent of the physicians were from the United States, 2% from Canada, and 3% from other countries. Forty-two percent were from Pennsylvania and bordering states. The average of reported practice parameters was 9 years of clinical practice, 54,000 patient visits annually, and 39 clinical hours per week. Table 1 summarizes practitioner self-reported HIV PEP prescribing history and self-use. Tables 2, 3, 4 provides the responses to the clinical scenarios.

workers. The majority prescribing occupational HIV PEP likely reflects the existence of CDC guidelines for these exposures. However, despite a lack of national guidelines for nonoccupational HIV PEP, approximately half of ED practitioners had dispensed HIV PEP to nonhealth care persons who experienced needlestick injuries or to survivors of sexual assault. Few practitioners indicated experience with HIV PEP after consensual sexual exposures. A smaller group of practitioners (19%) had given HIV PEP to adolescents and children, the preponderance being after sexual assault. Only 11% of ED practitioners had ever taken HIV PEP themselves. The majority of respondents indicated using three-drug regimens for all exposures (zidovudine, lamivudine, and indinavir or nelfinavir). This distinction is noteworthy because some groups and scientists advocate reserving three-drug regimens for exposures to known HIVinfected sources,2,4 whereas others recommend three-drug regimens for all exposures.3,6 Although monotherapy with zidovudine is no longer recommended, some practitioners indicated that this was the type of therapy they had prescribed most often.

DISCUSSION

Willingness to Offer HIV PEP in Clinical Scenarios

Self-reported History of Prescribing HIV PEP Nearly two-thirds of respondents had prescribed HIV PEP, primarily after needlestick injuries to health care

In the clinical scenarios, practitioners offered HIV PEP comparatively equally across several groups: needlestick injuries in health care workers vs. nonhealth care persons

MERCHANT AND KESHAVARZ ■ HIV POST-EXPOSURE PROPHYLAXIS

TABLE 2.

311

Needlestick Injuries* Percentage of ED Practitioners Offering HIV PEP

Resident Physicians (n ⫽ 193)

Practitioners ⬍ 6 Years Clinical Practice (n ⫽ 272)

Practitioners ⱖ 6 Years Clinical Practice (n ⫽ 328)

All Respondents (n ⫽ 600)

Exposure source’s HIV risk

HIV PEP Self-Users (n ⫽ 60)

HIV PEP Non-SelfUsers (n ⫽ 520)

Health Care Worker

Unknown HIV-infected High-risk Low-risk

92 (81-97) 100 (94-100) 98 (91-99) 69 (56-81)

77 (74-81) 98 (96-99) 95 (92-97) 38 (33-42)

84 (77-89) 99 (97-100) 98 (95-100) 48 (40-55)

77 (73-81) 97 (95-99) 94 (91-96) 39 (34-44)

88 (82-92) 99 (97-100) 97 (93-99) 53 (46-61)

85 (80-89) 99 (96-100) 96 (92-98) 50 (44-56)

74 (69-79) 97 (95-99) 95 (91-97) 34 (29-39)

79 (76-82) 98 (97-99) 95 (93-97) 41 (37-45)

Nonhealth Care Adult

Unknown HIV-infected High-risk Low-risk

90 (80-96) 100 (94-100) 98 (91-100) 63 (49-76)

77 (73-80) 97 (96-99) 95 (92-96) 37 (33-41)

81 (74-86) 98 (94-99) 97 (93-99) 45 (38-53)

77 (72-81) 97 (95-98) 94 (91-96) 38 (33-43)

83 (77-88) 99 (96-100) 97 (93-99) 50 (42-57)

81 (76-86) 99 (97-100) 96 (93-98) 48 (42-54)

75 (70-80) 96 (93-98) 94 (90-96) 33 (28-38)

78 (74-81) 97 (96-98) 95 (93-96) 40 (36-44)

5-year-old

Unknown HIV-infected High-risk Low-risk

87 (75-94) 100 (93-100) 98 (91-100) 63 (49-75)

74 (70-77) 95 (92-96) 92 (89-94) 35 (31-40)

78 (70-84) 95 (91-98) 93 (88-96) 41 (38-50)

74 (69-78) 95 (92-96) 92 (89-94) 37 (32-42)

83 (76-88) 99 (96-100) 98 (94-99) 53 (46-60)

81 (76-85) 98 (95-99) 95 (92-98) 49 (42-55)

69 (64-74) 92 (89-95) 90 (86-93) 30 (24-35)

75 (71-78) 95 (93-97) 92 (90-94) 38 (34-42)

Patient

Females (n ⫽ 172)

Males (n ⫽ 424)

*Summary of survey questions: “A (healthcare worker, adult, or 5-year-old child) is stuck by a bloody needle from (an unknown source, an HIV-positive source, a source with high HIV risk factors, or a source with low HIV risk factors). Indicate the kind of prophylaxis (HIV PEP) you would offer, if any.”

(both adults and children), sexual assault exposures vs. needlestick injuries, and adults vs. children for both needlestick injuries and sexual assault. For all exposures, respondents indicated nearly identical willingness to provide HIV PEP after exposures to HIV-infected and high HIV risk factor sources. Practitioners offered HIV PEP significantly less after exposures to unknown and low-risk HIV sources. ED practitioner HIV PEP offers after low-risk exposures were approximately half of those from unknown sources. These differences were true for all exposure types and populations. As can be seen by the results, practitioner responses varied in the clinical scenarios by the stated risk levels. These findings are interesting because the concurrent HIV PEP guidelines and recommendations primarily separate exposures by characteristics of the exposure itself (eg, presence of blood on the needle, use of condoms) and not the sources’ perceived HIV risk.1-6 TABLE 3.

The slightly lower offering of HIV PEP after exposures to unknown sources in the health care worker needlestick injury scenario could indicate an understanding that these injuries are regarded differently by the CDC than exposures to known HIV-infected sources (“consider” vs. “recommend” HIV PEP).1 However, the lower offering in the sexual assault scenarios after exposures to unknown sources might reflect a lack of knowledge of the available recommendations on HIV PEP after sexual assault. Far fewer respondents indicated willingness to prescribe HIV PEP after consensual sex than sexual assault. The following reasons are possible for the difference: (1) a perception that HIV transmission could be more likely after sexual assault; (2) a greater awareness of available recommendations on HIV PEP after sexual assault; (3) local ED policies that endorse or dictate offering HIV PEP to sexual assault survivors; (4) lesser awareness of the recommenda-

Sexual Assault*

Resident Physicians (n ⫽ 193)

Practitioners ⬍ 6 Years Clinical Practice (n ⫽ 272)

Practitioners ⱖ 6 Years Clinical Practice (n ⫽ 328)

All Respondents (n ⫽ 600)

Exposure source’s HIV risk

HIV PEP Self-Users (n ⫽ 60)

HIV PEP Non-SelfUsers (n ⫽ 520)

Adult

Unknown HIV-infected High-risk Low-risk

90 (80-96) 97 (89-100) 95 (86-99) 68 (55-80)

71 (67-75) 96 (94-98) 93 (91-95) 40 (36-44)

78 (72-84) 97 (93-99) 96 (92-99) 49 (41-57)

70 (66-75) 96 (94-98) 92 (89-95) 41 (37-46)

81 (75-87) 97 (94-99) 94 (90-97) 56 (48-63)

79 (74-84) 97 (94-99) 94 (91-97) 53 (47-59)

68 (62-73) 96 (93-98) 93 (89-95) 35 (30-41)

73 (69-76) 96 (95-98) 93 (91-95) 44 (40-48)

5-year-old

Unknown HIV-infected High-risk Low-risk

87 (75-94) 98 (91-100) 98 (88-100) 60 (46-72)

73 (69-77) 94 (92-96) 90 (87-93) 36 (32-40)

81 (74-86) 95 (91-98) 94 (89-97) 41 (33-49)

72 (67-76) 94 (91-96) 90 (86-92) 38 (33-43)

85 (80-90) 97 (94-99) 95 (91-97) 50 (42-57)

81 (76-86) 97 (94-98) 94 (90-96) 45 (39-51)

69 (64-74) 92 (89-95) 89 (85-92) 34 (28-39)

75 (71-78) 94 (92-96) 91 (89-93) 39 (35-42)

Patient

Females (n ⫽ 172)

Males (n ⫽ 424)

*Summary of survey questions: “An adult/a 5-year-old child suffers sexual assault with unprotected intercourse from (an unknown assailant, an HIV-positive assailant, an assailant with high HIV risk factors, or an assailant with low HIV risk factors). Indicate the kind of prophylaxis (HIV PEP) you would offer, if any.”

AMERICAN JOURNAL OF EMERGENCY MEDICINE ■ Volume 21, Number 4 ■ July 2003

312

TABLE 4.

Consensual Sex*

Resident Physicians (n ⫽ 193)

Practitioners ⬍ 6 Years Clinical Practice (n ⫽ 272)

Practitioners ⱖ 6 Years Clinical Practice (n ⫽ 328)

All Respondents (n ⫽ 600)

Exposure source’s HIV risk

HIV PEP Self-Users (n ⫽ 60)

HIV PEP Non-SelfUsers (n ⫽ 520)

Adult

Unknown HIV-infected High-risk Low-risk

54 (41-70) 97 (89-100) 93 (84-98) 36 (23-50)

44 (40-49) 88 (85-91) 83 (79-86) 20 (17-24)

55 (48-63) 91 (85-95) 89 (83-93) 27 (21-35)

41 (37-46) 88 (85-91) 82 (78-85) 20 (16-24)

50 (42-57) 90 (86-94) 86 (81-91) 29 (22-36)

50 (44-57) 91 (87-94) 86 (81-90) 26 (21-32)

41 (36-47) 87 (83-91) 82 (78-86) 19 (15-24)

45 (41-50) 89 (86-91) 84 (81-87) 22 (19-26)

15-year-old

Unknown HIV-infected High-risk Low-risk

53 (40-66) 93 (84-98) 88 (77-95) 36 (23-50)

46 (42-51) 87 (83-89) 82 (72-85) 19 (16-22)

56 (48-64) 88 (82-93) 84 (78-89) 26 (20-33)

44 (39-48) 87 (83-90) 82 (78-86) 19 (15-23)

52 (45-59) 91 (85-94) 87 (82-92) 27 (21-34)

51 (45-57) 90 (86-93) 85 (80-89) 25 (20-31)

44 (38-49) 84 (80-89) 80 (76-85) 17 (13-22)

47 (43-51) 87 (84-90) 83 (80-86) 21 (18-24)

Patient

Females (n ⫽ 172)

Males (n ⫽ 424)

*Summary of survey questions: “An adult/a 15-year-old engaged in unprotected intercourse (NOT an assault) with (an unknown person, an HIV-positive person, a person with high HIV risk factors, or a person with low HIV risk factors). Indicate the kind of prophylaxis (HIV PEP) you would offer, if any.”

tions for using HIV PEP after consensual sexual encounters; or (5) a concern that HIV PEP after consensual sex could encourage subsequent risk-taking sexual behavior. An additional worrisome, but hopefully not present, reason could be a perception that HIV exposures after consensual sex are voluntary, and therefore less deserving of HIV PEP. Demographic Factors vs. Willingness to Offer HIV PEP in the Clinical Scenarios Factors other than the risk of HIV seroconversion could have affected practitioner willingness to offer HIV PEP. Some of these factors include practitioner gender, HIV PEP self-use, training status and years in clinical practice, and type of sexual exposure. Females and HIV PEP self-users were generally more likely to offer HIV PEP. This difference could reflect a more intimate knowledge of HIV PEP by these groups or an intuitive, emotional, encultured, or educated belief on the risk of HIV transmission through these exposures. Resident physicians and those in practice for less than 6 years overwhelmingly offered HIV PEP more frequently than those in practice for longer periods. This distinction probably underscores the newness of this type of therapy and perhaps its greater use in academic settings. Our study had several limitations. First, the sample could not be completely representative of national practice. Only 15% of total conference attendees completed the survey; however, some attendees were medical students, nurses, administrators, exhibitors, and non-ED practitioners, and therefore were not included in the study. Our survey likely represents approximately 3% of ED practitioners in the United States.7 Although the respondents were drawn from a national sample, over one-third were from Pennsylvania and neighboring states, and a third were physician residents. Accordingly, our results could only reflect practitioners from this region and could be skewed by the large group of resident physician respondents. Second, this survey relied on self-report data and recall, which might not be accurate. Actual prescribing patterns would have been preferable. However, the clinical scenario results would not have been affected by such. Third, actual HIV PEP offering could be

different in clinical practice. Willingness to offer in clinical scenarios might not correlate with actual practice. In summary, we report that 68% of ED practitioners surveyed have prescribed HIV PEP, primarily after health care needlestick injuries, and 11% had self-used HIV PEP. ED practitioners are clearly willing to provide and have prescribed HIV PEP for nonoccupational exposures despite a lack of national guidelines. Female practitioners, resident physicians, practitioners with less than 6 years of clinical practice, and HIV PEP self-users indicated a greater overall willingness to offer HIV PEP. Although the risk of HIV seroconversion exists after consensual sex, the practitioners surveyed were less willing in the clinical scenarios to offer HIV PEP after those exposures than for sexual assault. Further educational endeavors that improve ED practitioner awareness of HIV PEP guidelines and recommendations appear to be indicated. ACKNOWLEDGMENT The authors gratefully acknowledge the assistance and support of the faculty and residents of the Department of Emergency Medicine of the Mount Sinai School of Medicine, and the survey participants of the American College of Emergency Physicians.

REFERENCES 1. Centers for Disease Control: Public Health Service guidelines for the management of health care worker exposures to HIV and recommendations for postexposure prophylaxis. MMWR Morb Mortal Wkly Rep 1998;47:1-33 2. Gerberding JL, Katz MH: Post-exposure prophylaxis for HIV. Adv Exp Med Biol 1999;458:213-222 3. AIDS Institute: HIV prophylaxis following sexual assault: Guidelines for adults and adolescents, New York, AIDS Institute, New York State Department of Health, 1998 4. Bamberger JD, Waldo CR, Gerberding JL, et al: Postexposure prophylaxis for human immunodeficiency virus (HIV) infection following sexual assault. Am J Med 1999;106:323-326 5. Mayer KH, Kwong J, Singal R, Boswell S: Non-occupational postexposure HIV prophylaxis: Clinical issues and public health questions. Med Health R 2000;83:210-213 6. Puro V: Post-exposure prophylaxis for HIV infection. Lancet 2000;355:1556-1557 7. Moorhead JC, Gallery ME, Manale T, et al: A study of the workforce in emergency medicine. Ann Emerg Med 1998;31:595607