The important function that HIV testing serves in effective HIV prevention and control strategies is well documented.1,2 In addition to these public health successes, advances in the availability of and access to highly active antiretroviral therapy have raised awareness regarding the importance of knowing one's serostatus.3-6 However, little is known about testing behaviors and the characteristics of individuals seeking this testing at a population level. While reports from population-based surveys and the evaluation of individuals attending specific testing sites (eg, sexually transmitted disease clinics) have contributed to our knowledge of testing behaviors,7-12 the inherent limitations posed by self-reporting on prior tests and focus on specific risk groups (eg, men who have sex with men or injecting drug users) make it difficult to apply these findings to the general population.
Given the centralized nature of laboratory testing and reporting of HIV infection in the province of Alberta, Canada, we had an opportunity to examine HIV testing behavior at the population level on all newly diagnosed HIV-positive cases using laboratory data to validate the time and number of prior tests for each case. Serologic screening for HIV began in 1985 and nonnominal reporting of HIV infection to provincial public health authorities became effective on May 1, 1998. All enzyme immunoassay anti-HIV screening (except third-party testing) and all confirmatory HIV testing in the province is conducted at the Provincial Laboratory for Public Health (PLPH). Laboratory specimens that are submitted for HIV testing to the PLPH are accompanied by a laboratory test requisition form, which includes patient demographics, clinical history, and risk exposures.
In this report, we characterize the magnitude and determinants of first-time and repeat testing for HIV infection among newly diagnosed cases in Alberta between 2000 and 2001. Given the roles of concurrent illnesses such as chronic hepatitis C (HCV) and hepatitis B (HBV) infections on the morbidity and mortality of HIV-infected individuals,13 a second objective was to examine the extent of co-infection with these viruses among our study population. To our knowledge, this is the first time that results from such a population-based study using laboratory data on HIV testing and co-infection with HCV or HBV have been reported in Canada.
Using the PLPH database, we extracted all specimens that were received between January 1, 2000 and December 31, 2001 at the PLPH and were screened positive for HIV infection. Through record linkage using a combination of unique identifiers (provincial health number, name, gender, and date of birth), we identified the number of individuals who were confirmed positive via Western blot for HIV infection during this period. To the extent possible, we excluded repeat positive testers (ie, individuals who had tested positive for HIV infection prior to January 2000 and individuals who repeated their test to confirm a positive diagnosis). We also removed from the analyses all individuals who were 2 years and younger, since at least an 18-month follow-up of laboratory testing is required before HIV infection can be confirmed. For the purposes of this report, we removed from the study population all individuals tested for HIV during the study period who resided outside of Alberta (n = 19).
To differentiate individuals who were confirmed positive for HIV infection at their first test (ie, first-time testers) from repeat negative testers who then became seropositive for HIV infection (ie, repeat testers) we determined whether the individuals who were newly diagnosed during 2000 and 2001 had previous negative HIV test results, and if so, when the test was performed at the PLPH. We used HIV tests performed between January 1, 1992 and December 31, 2001 to determine HIV testing history, since 1992 was the first complete year for which both positive and negative tests were computerized in the PLPH database. We counted any negative test reports that were within 1 month of each other as a single test since persons at risk for infection are counseled to retest a second time within this period. Using methods that have been previously described,11 we identified individuals who had recent infections as those with a confirmed positive HIV test within 1 year of a previous negative test report. To identify any geographic differences in the characteristics of first-time and repeat testers, we examined the city of residence of the study population and assigned individuals residing north of and including the city of Red Deer to northern Alberta.
We identified any significant differences between first-time and repeat testers using the χ2 test, and where appropriate, the Fisher exact test. The independent variables examined included age, gender, risk factors, area and population of residence, testing agency, and co-infection with HCV or HBV. We conducted logistical regression analyses to further define independent predictors of first-time vs. repeat testers for HIV infection. For all the variables, we conducted separate analyses including and excluding missing values to determine the effect (if any) of the missing information. Only results that were not affected by missing values are reported here. We did not conduct statistical analyses for variables with a value of <5 in any given cell.
Between January 2000 and December 2001, 140,199 specimens were screened for HIV infection at the PLPH. Record linkage identified 175 and 223 newly diagnosed cases of HIV infection in Alberta during 2000 and 2001, respectively, corresponding to a diagnostic rate of 6.08 per 100,000 population during 2000 and 7.64 per 100,000 population during 2001 (2000 and 2001 population figures obtained from Alberta Health and Wellness, http://www.health.gov.ab.ca/system/rhas/require/list.htm). Of the 398 cases, 278 (69.8%) were newly diagnosed with HIV at their first test (Fig. 1). During 2000, 73.1% were diagnosed at their first test for HIV infection. This proportion was 67.3% during 2001 (P = 0.81). Among repeat testers for HIV infection, the mean number of previous negative tests was 3.4 (range = 2-11 tests). The median interval between the last negative and first positive test was 648 days (range = 53-2678 days). Among repeat testers, 62 (51.7%) were diagnosed within 1 year of their last negative test.
The characteristics of first-time and repeat testers for HIV infection in our study population are shown in Table 1. While not statistically significant at the P < 0.05 level, compared with first-time testers, repeat testers were more likely to be female, to be younger in age when diagnosed with HIV infection, to reside in northern Alberta and in larger urban centers (population >200,000), to indicate injection drug use as their primary risk exposure, and to be tested for HIV infection at medical clinics or health centers (including walk-in clinics) or at remand/correctional institutions (Table 1). HCV status using the PLPH database was unknown for 98 individuals (24.6%) in our study population. Among those who were tested for HCV antibody at the PLPH, 161 (53.7%) were seropositive for HCV infection. Among repeat testers and first-time testers, 69 (64.5%) and 92 (47.7%), respectively, were positive for HCV antibody. Co-infection with HBV was unknown for 153 individuals (38.4%) in our study population. Among those who were tested for HBV surface antigen, core antigen, or e-antigen at the PLPH, 47 (19.1%) were positive for at least 1 of these 3 markers. Among repeat testers and first-time testers, 11 (13.6%) and 36 (22%), respectively, were one of these 3 markers.
In univariate analyses, compared with first-time testers, repeat testers were likely to be younger when diagnosed with HIV infection (P = 0.01, odds ratio [OR] = 1.03), more likely to reside in northern Alberta (P = 0.01, OR = 2.2), more likely to indicate injection drug use as their primary risk exposure to HIV infection (P = 0.02, OR = 2.8), and more likely to have been tested for infection at a remand or correctional facility (P = 0.05, OR = 4.8, Table 2). However, after adjusting for all other variables in the model, only residence in northern Alberta and injecting drug use were significantly associated with repeat testing. Compared with first-time testers, repeat testers were 1.8 times more likely to reside in northern Alberta (CI = 1.01-3.2) and 1.9 times more likely to indicate exposure to HIV infection through injecting drug use (CI = 1.1-3.2, Table 2).
Results from secondary analysis of laboratory data suggest that the majority of HIV cases (69.8%) who were newly diagnosed between 2000 and 2001 had no previous test for HIV infection since 1992 in Alberta. Where available, analysis of evolving band patterns in Western blots and clinical histories of first-time testers suggests that 62 (22.3%) may be seroconvertors or recent infections (data not shown). Among repeat testers, 62 (51.7%) were diagnosed within 1 year of their last negative test. These findings suggest that 275 individuals or approximately 69% of the study population may not have discovered their serostatus soon after infection.
Our finding confirms similar reports from other groups in Canada, Australia, Europe, and the United States, suggesting that an increased proportion of people may be postponing getting tested for HIV infection even if they suspect they have been exposed to the virus.14-20 Among repeat testers in our study population, HIV testing was sought infrequently, with a median interval of almost 2 years between the last negative and first positive test. This finding is cause for concern given reports that repeat testers may be at greater risk for HIV transmission.21,22 It has been proposed previously that positive test results among repeat negative testers represent a failure of the HIV prevention programs and missed opportunities for effective pre- and posttest counseling.23 Preliminary data examining testing behaviors among those who remain negative for HIV infection suggest that among residents of southern Alberta, out of 26,789 individuals who tested negative for HIV infection during 2001, 19,304 (72.1%) were first-time testers. These findings reinforce the need to make the most of each test-seeking event with proper counseling and other relevant support services.
The differences in the characteristics of repeat and first-time testers residing in northern vs. southern Alberta are interesting and likely reflect the different dynamics of the HIV epidemic in these 2 regions. Routine surveillance of reported HIV cases suggests that a disproportionate number of newly reported cases between 1998 and 2001 in Northern Alberta were among aboriginal persons, injecting drug users, and women.24 In contrast, during this period reported HIV cases in southern Alberta were comprised largely of caucasians and men who have sex with men. Currently, the PLPH database does not capture information on ethnicity so we were unable to conduct further analyses with respect to this variable. Clearly this and other factors need to be considered in formulating and evaluating effective HIV testing strategies. The impact of geographic, social, and economic factors on HIV test seeking behaviors among these individuals also merits further investigation.
Among those with a laboratory test result at the PLPH (n = 300), the fact that 53.7% had a confirmed laboratory test indicating co-infection with HCV is cause for concern. Among first-time testers with a documented test for HCV infection at the PLPH, 47.7% were co-infected with HCV. This proportion increased to 64.5% among repeat testers. Of note, 78 (20.6%) of our study population had no documented laboratory test for HCV antibody in the PLPH database. All screening (except third-party testing) and all confirmatory testing for HCV in the province are conducted at the PLPH. While it is possible that a minority of these individuals knew their HCV status from a previous test conducted out of province or in a private laboratory through a third-party agreement, and while we may not have successfully linked all HCV test reports to HIV case reports due to the availability of anonymous testing for HIV infections, these explanations seem unlikely to account for all the individuals with unknown HCV status in our study population. The proportion of individuals without a documented test result for HBV infection in the PLPH database was 24.6%, but this proportion may be explained by the fact that other private laboratories within the province also conduct testing for HBV infection. A more thorough investigation of HCV/HIV or HBV/HIV co-infections utilizing different data sources is warranted to determine the true prevalence of these co-infections in Alberta. However, our results clearly support the need for testing and counseling strategies to take into account additional risks associated with HCV infection, given the high prevalence of HCV/HIV co-infection in our study population.
Given some inherent limitations of this study, our findings must be interpreted with caution. The results reported here were based on information provided through HIV test requisitions that are completed by the health care providers prior to any knowledge of the HIV status of their patient. The data may therefore be prone to misclassification. However, whenever possible, the medical virologists at the PLPH confirm all information on the test requisition when communicating, by phone, the test results from newly diagnosed positive cases to the requesting physician. Clearly, enhanced data, particularly on risk behaviors including needle sharing, specific sexual practices, and HIV status of partner, would provide insight into reasons for testing and HIV acquisition. The PLPH is currently in the process of revising its test requisition form to capture information on ethnicity and enhanced information on risk exposure. Since provision of a provincial health number and nominal reporting of HIV cases to provincial public health authorities are not currently required in Alberta, we could not conduct a thorough linkage of data stored at the PLPH with HIV/AIDS cases reported to the provincial health office to complete missing information. Furthermore, we were unable to completely verify whether any of the cases newly diagnosed at the PLPH had tested positive previously in another province. The number of newly diagnosed cases is therefore an overestimate of the reported number of cases by Alberta Health to Health Canada.25 Similarly, we may have misclassified some repeat testers as first-time testers if individuals had been tested for HIV infection outside of Alberta or if they were screened at a private laboratory due to a third-party request. Of note, while the PLPH conducts screening and confirmatory testing of HIV infection for inmates in municipal and provincial facilities, the PLPH only conducts confirmatory HIV testing for those in federal institutions, so we may have underestimated the extent of repeat testing among this group. Last, although we attempted to eliminate duplicate patients from the data set prior to analysis, this process was incomplete. Information on at least 3 of these 4 unique identifiers was available in the PLPH database for 96.7% of the individuals tested for HIV during the study period. For the remaining individuals, the PLPH database contained information on at least 2 of the 4 unique identifiers. There are a few anonymous test sites in Alberta and it is possible that some of the people testing at these sites could not be identified. However, we believe that anonymous testing is requested infrequently. One of the 2 main sexually transmitted disease clinics in the province reported only 23 requests for anonymous HIV testing between May 1998, when HIV infection became notifiable, and May 2002.26
Despite these limitations, the use of a centralized, population-based laboratory system can serve a vital function in providing timely data on the epidemiology of HIV. Most newly diagnosed cases between 2000 and 2001 in Alberta tested positive at their first test for HIV infection. We need to investigate the complex and multifaceted factors that have contributed to HIV test-seeking behaviors among this group. For repeat testers who seroconvert and for those who are co-infected with HIV/HCV or HIV/HBV, each previous negative test result for these viral markers can be viewed as a missed opportunity for prevention. Different approaches may be required to address this issue.
The authors thank Rhonda Gordon and Barry Chamberlain for extracting the HIV, HCV, and HBV testing data from the PLPH database. We also thank Dr. Chris Archibald from Health Canada for helpful comments on the manuscript.
1. Valdiserri RO, Holtgrave DR, West GE. Promoting early HIV diagnosis and entry into care. AIDS.
2. Rogheram-Borus MJ, Newman PA, Etzel MA. Effective detection of HIV. J Acquir Immune Defic Syndr.
3. Rosenberg ES, Altfeld M, Poon SH, et al. Immune control of HIV-1 after early treatment of acute infection. Nature.
4. Ammassari A, Trotta MP, Murri R, et al. Correlates and predictors of adherence to highly active antiretroviral therapy: overview of published literature. J Acquir Immune Defic Syndr.
5. Brocklehurst P, Volmink J. Antiretrovirals for reducing the risk of mother-to-child transmission of HIV infection. Cochrane Database Syst Rev.
6. Valdiserri RO, Ogden LL, McCray E. Accomplishments in HIV prevention science: implications for stemming the epidemic. Nat Med.
7. Houston SM, Archibald CP, Strike C, et al. Factors associated with HIV testing among Canadians: results of a population-based survey. Int J STD AIDS.
8. MacKellar DA, Valleroy LA, Secura GM, et al. Repeat HIV testing, risk behaviors, and HIV seroconversion among young men who have sex with men: a call to monitor and improve the practice of prevention. J Acquir Immune Defic Syndr.
9. Weinstock H, Sweeney S, Satten GA, et al. HIV seroincidence and risk factors among patients repeatedly tested for HIV attending sexually transmitted disease clinics in the United States, 1991 to 1996. J Acquir Immune Defic Syndr Hum Retrovirol.
10. Phillips KA, Paul J, Kegeles S, et al. Predictors of repeat HIV testing among gay and bisexual men. AIDS.
11. McFarland W, Fischer-Ponce L, Katz MH. Repeat negative human immunodeficiency virus (HIV) testing in San Francisco: magnitude and characteristics. Am J Epidemiol.
12. Calzavara L, Burchell AN, Major C, et al. Increases in HIV incidence among men who have sex with men undergoing repeat diagnostic HIV testing in Ontario, Canada. AIDS.
13. Martinez EH. Hepatitis B and hepatitis C co-infection in patients with HIV. Rev Med Virol.
14. Girardi E, Sampaolesi A, Gentile M, et al. Increasing proportion of late diagnosis of HIV infection among patients with AIDS in Italy following introduction of combination antiretroviral therapy. J Acquir Immune Defic Syndr.
15. Poznansky MC, Coker R, Skinner C, et al. HIV positive patients first presenting with an AIDS defining illness: characteristics and survival. BMJ.
16. Gillieatt SJ, Mallal SA, French MAH, et al. Epidemiology of late presentation of HIV infection in Western Australia. Med J of Aust.
17. Wortley PM, Byers RH, Sweeney PA, et al. HIV testing patterns: where, why, and when were persons with AIDS tested for HIV? AIDS.
18. Nakashima AK, Campsmith ML, Wolfe MI, et al. Late versus early testing of HIV: 16 sites, United States, 2000-2003. MMWR.
19. Geduld J, Romaguera A, Esteve A, et al. Late diagnosis of HIV infection among reported AIDS cases in Canada and Catalonia, Spain. Paper presented at: XIV International AIDS Conference; July 2002; Barcelona, Spain. Proceedings from the XIV International AIDS Conference.
20. Njihia J, Remis RS, Swantee C, et al. Marked increase in first-time diagnoses in Ontario, 2002. Paper presented at: 12th Annual Conference of the Canadian Association for HIV/AIDS Research; April 10-13, 2003; Halifax, Nova Scotia.
21. Hankins C, Alary M, Parent R, et al. Continuing HIV transmission among injection drug users in Eastern Central Canada: the SurvUDI Study, 1995 to 2000. J Acquir Immune Defic Syndr.
22. Leaity S, Sherr L, Wells H, et al. Repeat HIV testing: high-risk behaviour or risk reduction strategy? AIDS.
23. Summers T, Spielberg F, Collins C, et al. Voluntary counseling, testing, and referral for HIV: new technologies, research findings create dynamic opportunities. J Acquir Immune Defic Syndr.
26. Jayaraman GC, Preiksaitis JK, Larke B. Mandatory reporting of HIV infection to public health authorities and opt out prenatal screening for HIV in Alberta: effect on testing rates. CMAJ.