HIV non-occupational postexposure prophylaxis in a Canadian province: treatment completion and follow-up testing

Abstract

Provincial guidelines for HIV non-occupational postexposure prophylaxis (NPEP) were implemented on January 2005 in Alberta, Canada. Human immunodeficiency virus (HIV) NPEP was provided free of charge following approval by a medical officer of health. Between 1 January 2005 and 30 June 2007, 174 individuals were prescribed NPEP; 135 (78%) were women with a median age of 24 years. Sexual assaults accounted for 68% of exposures. NPEP was completed in 49% of cases. Individuals who completed NPEP were less likely to have been exposed by sexual assault (P = 0.04) and more likely to have received HIV follow-up testing (P = 0.03).Individuals who received at least one HIV follow-up test were older (P = 0.03) and more likely to have been exposed percutaneously (P = 0.003). Those who received no follow-up testing were less likely to have filled an NPEP prescription (P = 0.0001). New strategies are required to improve follow-up of individuals receiving NPEP, especially younger persons or sexual assault survivors.

Keywords

HIV NPEP sexual assault anti-HIV agents

INTRODUCTION

Evidence for the efficacy of non-occupational postexposure prophylaxis (NPEP) against human immunodeficiency virus (HIV) is limited to observational studies,^1–8 as well as extrapolations from occupational and perinatal data. Challenges include poor adherence and low rates of follow up, especially in sexual assault survivors.^2–4,6,7 Despite the lack of definitive efficacy data, the use of NPEP is widespread. In the USA, national recommendations for HIV NPEP have been available since 2005.⁹ In Canada, however, no national guidelines exist and recommendations vary by province.

Alberta is a Canadian province with a population of approximately three million, and is divided into nine health regions. In January 2005, a provincially funded NPEP protocol was developed by a committee of experts in HIV and public health and implemented in Alberta. It sets out standardized requisition and reporting processes, and provides guidelines for HIV NPEP and follow-up testing.¹⁰ Guidelines are provided for risk stratification, depending on the type or route of exposure, source risk factors, and timing of the exposure. A 28-day course of combination antiretroviral therapy is recommended for exposures to potentially infected blood or bodily fluids of a known or suspected HIV-infected source when that exposure poses a substantial risk for HIV transmission, and when NPEP can be initiated within 72 hours of exposure. A three-drug regimen is recommended for exposures to a known HIV-positive source (typically zidovudine plus lamivudine plus lopinavir/ritonavir). A two-drug regimen may be used in cases with unknown source HIV status, based on the level of risk (typically zidovudine plus lamivudine).

The regional medical officer of health (RMOH) is consulted for cases in which HIV NPEP may be indicated. The RMOH approves release of and payment for NPEP on a case-by-case basis, and reports case information to a central public health office (Alberta Health and Wellness [AHW]). If NPEP is indicated, a starter-pack is provided for timely initiation of therapy. Individuals started on NPEP are referred to a specialist physician within one to five days. Individuals continued on NPEP subsequently fill a prescription for the remainder of the 28-day treatment at one of four designated pharmacies (two in Northern Alberta and two in Southern Alberta) at seven-day intervals.¹⁰

The objectives of this study were to characterize NPEP cases and to identify factors associated with completing NPEP and receiving follow-up testing. Findings from this study will be useful in identifying areas for improvement in the implementation of NPEP programmes from a public health perspective. Knowledge of patient and exposure characteristics will aid in identifying opportunities for risk reduction initiatives.

METHODS

All approved NPEP cases in Alberta with a documented date of exposure between 1 January 2005 and 30 June 2007 were included. Patient demographic data, type of exposure and source information (as documented by the RMOH) were extracted from the AHW electronic database by one of the investigators (KW). In addition, we accessed outpatient pharmacy databases to identify duration of NPEP dispensed; AHW case records were also cross-referenced for any invoices from non-designated pharmacies in order to capture all NPEP medications dispensed. Follow-up testing results for anti-HIV and anti-hepatitis C virus (HCV) antibodies were obtained by one of the investigators (SP) from the Alberta Provincial Laboratory for Public Health, which performs all HIV and HCV testing in the province. Data from AHW, pharmacy and laboratory databases were merged (by SP), using patient names and personal health numbers, to develop an analytical database. Subsequently, all patient-identifying information was removed and replaced with unique study identifiers to preserve patient confidentiality.

Based on duration of NPEP dispensed, we described cases as ‘complete’, ‘partial’ or ‘none’ (if no outpatient NPEP prescription was filled at all) – this information was used as a proxy for treatment adherence. In order to assess treatment completion status, we cross-referenced dispensary records (fill dates and quantities) with the date of exposure and date of NPEP initiation. In cases where a starter-pack was given, we considered treatment complete if the entire authorized quantity of NPEP was dispensed. Cases in which treatment completion status could not be determined using available pharmacy or AHW documentation were described as ‘indeterminate’. Treatment completion status for all cases was assigned by one study investigator (KW). Since the primary objective of the study was to identify significant correlates with treatment completion, we compared individuals in the ‘complete’ category with all other individuals.

Serological testing is recommended at baseline, six, 12 and 24 weeks postexposure for HIV antibody, and at baseline, 12 and 24 weeks for HCV antibody.¹⁰ We evaluated follow-up testing by measuring the number of tests received after the baseline test (defined as the test performed on, or most immediately following, the date of exposure), as well as the time (in hours) from exposure to testing. In order to identify factors associated with follow-up testing, we compared individuals who received at least one follow-up test with those who received none. Those patients who were known HIV or HCV antibody positive prior to the exposure or positive at baseline testing were excluded from further serology analyses for HIV and HCV, respectively.

Descriptive analyses were performed using Stata version 10.0 or Microsoft Excel 2003. If an individual had more than one exposure, we used only information from the first exposure in our analysis. Continuous variables were compared using the Mann-Whitney U test, and categorical variables were compared using the χ ² test or Fisher's exact test if cell sizes were less than 10; P values <0.05 were considered statistically significant.

This study was approved by the University of Alberta Health Research Ethics Board and the University of Calgary Conjoint Health Research Ethics Board.

RESULTS

Between 1 January 2005 and 30 June 2007, 174 individuals were prescribed NPEP.

One individual was prescribed NPEP twice, in both cases following a sexual assault.

The majority of NPEP recipients were women (78%) and young (median age 24 years, mean 27 years, standard deviation 12 years). Sexual assaults accounted for 118 (68%) of exposures. Thirty-six (21%) were exposed percutaneously, including 19 accidental needlestick injuries, five human bites and one needlestick through injection drug use (IDU). Twelve (7%) were consensual sexual exposures and five (3%) were mucosal membrane exposures (Table 1).

Table 1

Demographic characteristics of non-occupational postexposure prophylaxis (NPEP) cases (n = 174)

Characteristic	n (%)
Median age (range)	24 years (4–69)
Gender
Female	135 (78)
Male	39 (22)
Exposure type
Sexual assault	118 (68)
Percutaneous	36 (21)
Consensual sex	12 (7)
Mucosal	5 (3)
Other	1 (0.6)
Not documented	2 (1)
Time to NPEP initiation
Within 24 hours	96 (55)
24–48 hours	54 (31)
48–72 hours	12 (7)
> 72 hours	9 (5)
Unknown	3 (2)
Source risk factors
Unknown assailant/source	87 (50)
Illicit drug use	39 (22)
IDU	27 (16)
Known HIV positive	25 (14)
History of incarceration	17 (10)
MSM	9 (5)
Known HCV antibody positive	7 (4)

IDU = injection drug use; HIV = human immunodeficiency virus; MSM = men who have sex with men; HCV = hepatitis C virus

NPEP was initiated within 72 hours postexposure in 162 (93%) cases. Most individuals (82%) were prescribed two drugs (primarily zidovudine and lamivudine); 16% were prescribed three drugs (typically zidovudine, lamivudine, and either nelfinavir or lopinavir/ritonavir). In three cases, the regimen was modified during the course of therapy.

A total of 118 (68%) individuals filled a prescription for NPEP. Treatment was completed for 86 (49%) cases and partially completed for 23 (13%) cases. Nine (5%) cases had unclear documentation and were classified as indeterminate. Fifty-six (32%) individuals did not fill an outpatient NPEP prescription.

Compared with all other individuals, those who completed NPEP were less likely to have been exposed by sexual assault (P = 0.04) and more likely to have received HIV follow-up testing (P = 0.03) (Table 2).

Table 2

Factors associated with non-occupational postexposure prophylaxis (NPEP) treatment completion

Characteristic	Completed NPEP (n = 86), n (%)	All others (n = 88), n (%)	P value*
Median age (years)	24	24	0.69
Gender			0.09
Male	24 (28)	15 (17)
Female	62 (72)	73 (83)
Type of exposure			0.33
Sexual assault	52 (61)	66 (75)	0.04
Percutaneous	21 (24)	15 (17)
Consensual sex	8 (9)	4 (5)
Mucosal	3 (4)	2 (2)
Other or not documented	2 (2)	1 (1)
Source risk factors
Unknown assailant	43 (50)	44 (50)	1.00
Illicit drug use	21 (24)	18 (21)	0.53
Known IDU	13 (15)	14 (16)	0.89
Known HIV positive	15 (17)	10 (11)	0.25
History of incarceration	5 (6)	12 (14)	0.12
MSM	4 (5)	5 (6)	1.00
HCV antibody positive	4 (5)	3 (3)	0.72
NPEP regimen prescribed			0.21
Two-drug	69 (80)	74 (84)
Three-drug	17 (20)	10 (11)
Four-drug	0 (0)	1 (1)
Had at least one follow-up test
HIV	58 (67)	45 (51)	0.03
HCV	40/80 (50)^†	36/84 (43)^†	0.36

IDU = injection drug use; HIV = human immunodeficiency virus; MSM = men who have sex with men; HCV = hepatitis C virus

*P values were calculated using the Mann-Whitney U test for continuous variables and χ ² or Fisher's exact tests for categorical variables

^†Individuals who tested positive at baseline or prior to exposure were excluded, as no further testing is required if already positive

One hundred and sixty-eight (97%) individuals received HIV testing at baseline. None were known seropositive prior to exposure nor found seropositive on baseline testing. One hundred and three (59%) individuals had one follow-up test (median time 6 weeks postexposure), 58 (33%) had two tests (14 weeks), 23 (13%) had three tests (26 weeks) and three (2%) had four tests (27 weeks). No HIV seroconversions were documented among those tested.

One hundred and sixty (95%) individuals received HCV testing at baseline. Six were known seropositive prior to exposure and four were found to be seropositive on baseline testing. Seventy-six (46%) individuals had one follow-up test (median time 10 weeks postexposure), 36 (22%) had two tests (22 weeks) and 11 (7%) had three tests (25 weeks). One individual developed detectable HCV antibodies 13 weeks postexposure, following a percutaneous knife injury where extensive blood exchange occurred with a known HCV-positive source with a history of incarceration and IDU.

Compared with those who received no follow-up testing, individuals who received at least one follow-up test were older (P = 0.03 for HIV, P = 0.04 for HCV) and more likely to have been exposed percutaneously (P = 0.003 for HIV, P = 0.005 for HCV). Those who received no follow-up testing were less likely to have filled an NPEP prescription (P = 0.0001 for HIV, P = 0.008 for HCV) (Table 3).

Table 3

Factors associated with receiving follow-up testing

Characteristic	≥1 HIV follow-up test			≥1 HCV follow-up test
Characteristic	Yes (n = 103), n (%)	No (n = 71), n (%)	P value*	Yes (n = 76),^† n (%)	No (n = 88),^† n (%)	P value*
Median age (years)	26	23	0.03	26	23	0.04
Gender			0.69			0.16
Male	22 (21)	17 (24)		13 (17)	23 (26)
Female	81 (79)	54 (76)		63 (83)	65 (74)
Type of exposure			0.02			0.005
Sexual assault	65 (63)	53 (75)		49 (65)	65 (74)
Percutaneous	29 (28)	7 (10)		22 (29)	10 (11)
Consensual sex	5 (5)	7 (10)		1 (1)	9 (10)
Mucosal	2 (2)	3 (4)		2 (3)	3 (3)
Other, not documented	2 (2)	1 (1)		2 (3)	1 (1)
Source risk factors
Unknown assailant	52 (51)	35 (49)	0.88	40 (53)	43 (49)	0.63
Illicit drug use	24 (23)	15 (21)	0.74	19 (25)	16 (18)	0.29
Known IDU	15 (15)	12 (17)	0.68	12 (16)	12 (14)	0.70
Known HIV positive	16 (16)	9 (13)	0.67	8 (11)	14 (16)	0.36
History of incarceration	12 (12)	5 (7)	0.44	10 (13)	7 (8)	0.31
MSM	6 (6)	3 (4)	0.74	3 (4)	6 (7)	0.51
HCV antibody positive	6 (6)	1 (1)	0.24	5 (7)	1 (1)	0.10
NPEP regimen prescribed			0.90			0.65
Two-drug	84 (82)	59 (83)		62 (82)	75 (85)
Three-drug	17 (17)	10 (14)		12 (16)	12 (14)
Four-drug	1 (1)	0 (0)		1 (1)	0 (0)
Treatment completion			0.001			0.03
Complete	58 (56)	28 (39)		40 (53)	40 (46)
Partial	18 (18)	5 (7)		14 (18)	9 (10)
No pharmacy data	21 (20)	35 (49)		16 (21)	36 (41)
Indeterminate	6 (6)	3 (4)		6 (8)	3 (3)

IDU = injection drug use; HIV = human immunodeficiency virus; MSM = men who have sex with men; HCV = hepatitis C virus

*P values were calculated using the Mann-Whitney U test for continuous variables and χ² or Fisher's exact tests for categorical variables

^†Individuals who tested positive at baseline or prior to exposure were excluded, as no further testing is required if already positive

In the subgroup of 25 exposures with a known HIV-positive source (11 percutaneous, 10 consensual sex, one sexual assault, two mucosal and one IDU), 16 were prescribed three drugs (one was later switched to a two-drug regimen), eight were prescribed two drugs and one was prescribed four drugs. In the latter case, it was documented that the source was on antiretroviral medications. Fifteen (60%) completed treatment, three (12%) partially completed treatment, six (24%) did not fill an NPEP prescription and one (4%) had indeterminate treatment completion status. Sixteen (64%) individuals had at least one HIV follow-up test. Differences between this subgroup and the overall study population were not statistically significant.

Fifty per cent of cases had an unknown source (74 sexual assaults, 12 percutaneous exposures and one mucosal exposure), in which the source was either unidentified or unavailable for testing. Differences between this subgroup and the overall study population were not statistically significant (data not shown).

DISCUSSION

Since the implementation of the Alberta NPEP protocol, the majority of NPEP cases were sexual assaults in young women. Although there have been no documented HIV seroconversions, low rates of follow-up testing and treatment completion were observed.

Sexual transmission is the most common route of HIV transmission in Canada.¹¹ In this study, the vast majority (91%) of sexual exposures were non-consensual. Certain exposure characteristics, such as genital trauma or multiple assailants, may increase transmission risk following sexual assault. A study performed in an emergency department in the USA found similar results; out of 620 patients evaluated for NPEP following sexual exposure, 94% of adult and 95% of paediatric exposures were due to reported or suspected sexual assault.¹²

Adherence to NPEP has been a challenge identified by many studies, but tends to be especially poor among sexual assault survivors.⁶ In British Columbia, 41% of sexual assault survivors continued treatment after receiving a starter-pack; however, only 11% completed treatment.⁷ Sexual assault presents barriers to adherence and follow up not only because of emotional and psychological trauma, but also because the assailant's HIV status is often unknown (63% of sexual assault cases in this study). In another study, 38% attended at least one follow-up visit and 13% completed treatment.¹³ Comparably low rates were reported in other studies,^2,3,14 and similarly this trend was also observed in the present study.

In contrast, studies of consensual sexual exposures have generally observed better treatment completion rates, ranging from 68% to 89%.^1,5,8 This trend was observed in the present study, although it was not statistically significant. Percutaneous exposures also trended towards higher rates of treatment completion. Although there is little non-occupational data specific to percutaneous exposures, studies in health-care workers, 84% of which are exposed percutaneously, reported that 77% completed treatment (either finished a PEP course or discontinued PEP because the source subsequently tested HIV-negative).¹⁵ Extrapolation of this information to non-occupational settings is difficult; however, since source sero-status or source risk factors are often available in health-care-related exposures, and health-care workers may be more educated and motivated to adhere to recommendations.

Although not measured in this study, one reason for non-adherence may have been adverse effects; other publications have consistently identified adverse effects as the most significant cause for NPEP discontinuation or regimen modification.^1,2,6–8 The USA NPEP registry reported that half of discontinuations were due to adverse effects,⁹ while occupational data also report similar findings.¹⁵ Reported overall incidences of adverse effects have been as high as 93% for three-drug NPEP and 66% for two-drug NPEP, with gastrointestinal discomfort being the most common.² Three-drug regimens generally have poorer tolerability and higher discontinuation rates^2,15 than two-drug regimens.

Another possible reason for NPEP discontinuation after the initial starter-pack is that the source subsequently tested HIV negative.¹⁰ This was the cause for 48% of discontinuations in the USA occupational PEP registry.¹⁵ This is unlikely to be an explanation in this study, as 50% of cases had an unknown source, in which the source was either completely unknown or their sero-status could not be obtained – rates of treatment completion in this subgroup were similar to those of the overall study population. In one study of sexual and IDU exposures, only 16% of source partners were available for testing and 75% of them tested HIV positive.¹

Abandonment of follow up is another major challenge identified in the literature. In the present study, although 59% and 46% had at least one follow up test for HIV and HCV, respectively, only 13% and 22% completed all testing as recommended by the protocol guidelines.¹⁰ Again, NPEP literature reports particularly poor follow up among sexual assault survivors, a trend also observed in the present study. The low rates of follow up testing found in this study may be due to the relatively high proportion of sexual assaults, and low proportion of percutaneous exposures. The correlation of follow up testing with older age may be a reflection of the higher proportion of sexual assaults in the younger population. Consistent with the correlations found in the present study, Garcia et al. ² found that in sexual assault survivors, completion of NPEP and college level education were associated with achieving follow up. In addition, further follow up testing would be discontinued if the source tested negative, although this is an unlikely explanation since there were no statistical differences in the unknown source subgroup.

Although the use of combination regimens has been advocated, there is no clear consensus on the preferred regimen. For exposures from sources of unknown HIV status, the Alberta NPEP protocol recommends two-drug NPEP.¹⁰ Zidovudine is recommended as part of the initial regimen based on data from a case-control study in the occupational setting.¹⁶ All regimens in the present study included at least two drugs, and all but three regimens contained zidovudine. In two of the three exceptions, the choice of a non-standard regimen was based on source treatment or resistance history. Although three-drug NPEP is recommended for exposures with a known HIV-positive source,¹⁰ out of 25 such cases, eight were prescribed two drugs. Moreover, not all three-drug regimens were used in cases with a known HIV-positive source. Decisions whether or not to use a third drug are typically based on the perceived level of transmission risk.

This study has several limitations. First, data were collected retrospectively – the current reporting process in Alberta includes only information known at baseline, but not outcomes upon follow-up, such as modifications or discontinuations of NPEP, duration of NPEP taken, adverse effects or laboratory abnormalities, follow-up serology, and source serology. Further information about the NPEP recipient, such as HIV risk factors and socioeconomic factors, may have been helpful in understanding possible reasons for poor follow-up. Second, this study did not directly measure adherence, since reasons for discontinuation of NPEP or follow-up testing were not measured. Also, the definition of ‘completed treatment’ using pharmacy refills does not necessarily guarantee actual consumption of medications as per guidelines. Third, certain associations in the data may have failed to reach statistical significance due to small sample size.

The need for strategies to improve adherence and follow-up has been discussed extensively in the literature. Besides serological testing, follow-up is essential for assessment of adherence, monitoring and management of adverse effects, risk reduction counselling, psychological support as well as follow-up for other forms of prophylaxis (such as sexually transmitted infections or emergency contraception). In contrast to the low rates of follow-up and adherence found in many retrospective analyses (including the present study), several prospective studies employing proactive follow-up strategies have reported substantially better outcomes. In one prospective study, NPEP recipients received five risk-reduction counselling sessions and three one-on-one medication adherence counselling sessions; 75% received six-month follow up and 78% completed treatment.¹ An unpublished pilot study of sexual assault survivors in South Africa, which involved medication adherence counselling and a tracing system for follow up, found that 89% received follow-up at four weeks, and relatively few (13%) discontinued NPEP.¹⁷

A potential concern is that the availability of publicly funded NPEP may disinhibit high-risk behaviours if primary prevention is not reinforced. The observational data presented here do not support this occurrence in Alberta. Although publicly funded, NPEP was not originally intended for consensual sexual exposures, NPEP was approved for 12 such cases due to a perceived high risk of HIV transmission. While other studies have also refuted this concern around behavioural disinhibition,^5,18 increased NPEP accessibility must be accompanied by efforts to promote primary prevention, such as risk reduction counselling.¹⁹ Finally, since the ultimate goal of NPEP is to prevent HIV transmission and given the relatively poor rate of follow-up testing in this and other studies, additional strategies to improve follow-up testing must receive further attention.

CONCLUSION

In summary, following the establishment of a provincial NPEP protocol, we found that the majority of NPEP cases were sexual assaults in young women. Although there have been no documented HIV seroconversions, issues of incomplete follow-up testing and premature NPEP discontinuation were identified. Further research is required to explore reasons for discontinuation and tolerability of NPEP regimens, questions that may be better answered by a prospective study. We believe our study highlights the importance of incorporating proactive strategies when establishing NPEP programmes in order to improve treatment adherence and follow up.

References

Kahn

, Martin

, Roland

, Feasibility of postexposure prophylaxis (PEP) against human immunodeficiency virus infection after sexual or injection drug use exposure: the San Francisco PEP Study. J Infect Dis 2001;183:707–14

Garcia

, Figueiredo

, Moretti

, Resende

, Bedoni

, Papaiordanou

. Postexposure prophylaxis after sexual assaults: a prospective cohort study. Sex Transm Dis 2005;32:214–9

Linden

, Oldeg

, Mehta

, McCabe

, LaBelle

. HIV postexposure prophylaxis in sexual assault: current practice and patient adherence to treatment recommendations in a large urban teaching hospital. Acad Emerg Med 2005;12:640–46

Meel

. HIV/AIDS post-exposure prophylaxis (PEP) for victims of sexual assault in South Africa. Med Sci Law 2005;45:219–24

Schechter

, do Lago

, Mendelsohn

, Behavioral impact, acceptability, and HIV incidence among homosexual men with access to postexposure chemoprophylaxis for HIV. J Acquir Immune Defic Syndr 2004;35:519–25

Sonder

, Van den Hoek

, Regez

, Brinkman

, Coutinho

. Trends in HIV postexposure prophylaxis prescription and compliance after sexual exposure in Amsterdam, 2000–2004. Sex Transm Dis 2007;34:288–93

Wiebe

, Comay

, McGregor

, Ducceschi

. Offering HIV prophylaxis to people who have been sexually assaulted: 16 months' experience in a sexual assault service. Can Med Assoc J 2000;162:641–5

Winston

, McAllister

, Amin

, Cooper

, Carr

. The use of a triple nucleosidenucleotide regimen for nonoccupational HIV post-exposure prophylaxis. HIV Med 2005;6:191–7

Merchant

. Antiretroviral postexposure prophylaxis after sexual, injection-drug use, or other nonoccupational exposure to HIV in the United States. Ann Emerg Med 2005;46:82–4

10.

Alberta Health and Wellness. Alberta guidelines for post-exposure prophylaxis in non-occupational settings. HIV, Hepatitis B, Hepatitis C, and sexually transmitted infections (Alberta NPEP protocol), 2006. See http://www.health.alberta.ca/professionals/NPEPprotocol.pdf (last accessed 28 May 2008)

11.

Public Health Agency of Canada. Canadian guidelines on sexually transmitted infections. Management and treatment of specific infections. Human immunodeficiency virus infections, 2006. See http://www.phac-aspc.gc.ca/std-mts/sti_2006/pdf/05sti2006h_e.pdf (last accessed 28 May 2008)

12.

Merchant

, Becker

, Mayer

, Fuerch

, Schreck

. Emergency department blood and body fluid exposure evaluations and HIV postexposure prophylaxis usage. Acad Emerg Med 2003;10:1345–53

13.

Olshen

, Hsu

, Woods

, Harper

, Harnisch

, Samples

. Use of human immunodeficiency virus postexposure prophylaxis in adolescent sexual assault victims. Arch Pediatr Adolesc Med 2006;160:674–80

14.

Kerr

, Cottee

, Chowdhury

, Jawad

, Welch

. The Haven: a pilot referral centre in London for cases of serious sexual assault. BJOG 2003;110:267–71

15.

Wang

, Panlilio

, Doi

, White

, Stek

Jr , Saah

. Experience of healthcare workers taking postexposure prophylaxis after occupational HIV exposures: findings of the HIV Postexposure Prophylaxis Registry. Infect Control Hosp Epidemiol 2000;21:780–5

16.

Cardo

, Culver

, Ciesielski

, A case-control study of HIV seroconversion in health care workers after percutaneous exposure. New Engl J Med 1997;337:1485–90

17.

Roland

, Myer

, Chuunga

, A prospective study of post-exposure prophylaxis following sexual assault in South Africa. In: 12th Conference on Retroviruses and Opportunistic Infections, 22–25 February 2005, Boston, MA

18.

Martin

, Roland

, Neilands

, Use of postexposure prophylaxis against HIV infection following sexual exposure does not lead to increases in high-risk behavior. AIDS 2004;18:787–92

19.

Roland

, Neilands

, Krane

, A randomized trial of standard versus enhanced risk reduction counseling for individuals receiving post-exposure prophylaxis following sexual exposure to HIV. In: 13th Conference on Retroviruses and Opportunistic Infections, 5–8 February 2006, Denver, CO