HIV Surveillance among General Population Shyam P. Lohani, PhD Overview of HIV surveillance in General population About 33.2 million people living with HIV worldwide including an estimated 2.5 million new infections worldwide in 2007. Globally HIV incidence peaked in the late 1990s and stabilized despite increasing incidence in many regions of the world. Behavioral change to prevent infection such as increased utilization of condoms, delay in first sexual contact and fewer sexual partners played a key role in the decline in adult HIV infection rates in many areas. However, the number of people living with HIV in East Asia, Eastern Europe and Central Asia doubled between 2001 and 2007. In 2007, estimated four million (3.3- 5.1 million) HIV positive resides in South and South East Asia, second only to Sub Saharan Africa (WHO, 2007). HIV/AIDS is the greatest public health crisis in today’s time with leading mortality among women of reproductive age group in some high burden countries in Africa (WHO, 2010). Globally, 15.4 million adult women (15 years and older) are now living with HIV which is more than ever before. In 2008 alone estimated 430,000 children were newly infected with HIV. The most frequent source of HIV infection in infants and children is transmission from mother-to-child during pregnancy, labor and delivery, or breastfeeding (WHO, 2010). A report reveal prevalence of transmission of HIV from the mother to the child in developing countries is about 30%, the highest risk of infection reported at the time of labor (GTZ, 2003). The prevention of mother-to-child transmission (PMTCT) is a highly effective intervention and has great potential to improve child survival and health. Comprehensive programs for prevention of mother-to-child transmission of HIV (PMTCT) including ARV therapy and prophylaxis can significantly reduce the number of infants with HIV-infection and promote better health for their mothers and families in general (WHO, 2007). Global status of PMTCT services Mother-to-child transmission (MTCT) of HIV can occur during pregnancy, during delivery or through breastfeeding. The risk of MTCT is between 15 and 30% in non-breastfeeding children and between 25 and 45% in breastfeeding children. The risk factors for MTCT of HIV include high maternal viral load, low CD4 count, impaired cell mediated immunity, prolonged rupture of membrane, prematurity, low birth weight, bacterial chorioamnionitis, occurrence of STI, first twin, use of obstetric procedures such as forceps and electrodes, and subclinical mastitis (WHO 2001). TM PMTCT refers to comprehensive, family-focused clinical and supportive services provided along with other public health initiatives to prevent the transmission of HIV from a woman to her infant (WHO, 2007). Programs should be available to pregnant women that include HIV counseling during antenatal care, HIV testing, post counseling including counseling to couples; use of condoms; antiretroviral

PDF Editor

1

therapy as prophylaxis for MTCT prevention, infant feeding counseling and support, care and HIV/AIDS treatment services available in the country. Everyday about 17% of children under the age of 15 are infected with HIV (WHO, 2007). An estimated 430,000 newly born were infected with HIV in 2008, over 90% of them through mother-to-child transmission (MTCT) and out of which more than 95% occurs in low and middle income countries. About 70% of these infected children died in 2007 without treatment. The risk of MTCT ranges from 20% to 45% without appropriate program intervention (WHO, 2010). As a mode of transmission, MTCT accounts for more than 10% of all new HIV infections globally (WHO, 2007). Antiretroviral drugs reduce the viral load and can reduce the rates of mother-to-child transmission. Only about ten percentages of pregnant women in low and middle income countries were offered prophylactic antiretroviral therapy to prevent transmission to their newborns in 2005, a modest increase of coverage from 2003, which was just 7.6% (WHO, 2006). Though the coverage of services to prevent mother-to-child transmission of HIV (PMTCT) is expanding in low- and middle-income countries. Overall, 33 per cent of pregnant women living with HIV in these countries received antiretroviral regimens, including antiretroviral therapy to prevent transmission of the virus to their infants, in 2007 (UNICEF, 2008). PMTCT program in Nepal The total fertility rate (TFR) in Nepal in mid-1976 was estimated at 6.3 births per woman, contraceptive use among currently married women was low (3 percent) and the proportion marriage was high (MoH, 1977). Given these demographic parameters, no immediate change in the fertility rate in Nepal seemed probable. In recent years however, several researchers have noted that a fertility transition has been underway in Nepal. National Demographic and Health Survey (NDHS, 2006) showed an unprecedented decrease in the TFR of 3.1 births per woman in 2006 (MoH, New ERA and Macro International Inc, 2007) from 4.1 births per woman in 2001 (MoH, New ERA and ORC Macro, 2002).

In Nepal, hospital based PMTCT services were established by the government in three different sites in 2005 and increased in number to four in 2006. Nepal had formed PMTCT working group and national TM protocol and guidelines have been developed. Tribhuvan University Teaching Hospital and Maternity hospital are PMTCT sites in Kathmandu. With the establishment of total four PMTCT sites in the year 2006 the service has been increasingly being utilized by pregnant mothers at ANC as compared to when it was first started in 2005.

PDF Editor

2

Data Source for HIV surveillance in General Population HIV surveillance among ANC clinic attendees is used as a proxy for the general population, especially in the case of generalized epidemics, as it is a source of low cost data for the general population. The population of Nepal is about 28 million and national estimates indicate that approximately 70,000 adults and children are living with HIV in Nepal (NCASC, 2009). Thus, ANC clinical attendees are used as a proxy for remaining 27.93 million population of the country. About 0.9 million annual pregnancies occurs in Nepal and approximately 1800 pregnancies are estimated to occur in HIV positive women with HIV prevalence of 0.2% that accounts to an annual birth of about 450-800 HIV infected newborns (Shrestha, 2005). Although the number of mothers needing PMTCT service tends to be 1800, only 262 new HIV positive pregnant women has been identified since 2006 and only 124 women being provided with navirapine as a prophylactic drug. Not all pregnant women attending ANC undergo pretest counseling. Out of 1.33 million pregnant women who attended pretest counseling at different PMTCT sites in Nepal only 1.23 million pregnant women had undergone HIV testing since the year 2006, and moreover, there was a decreasing trend of new pretest counseling at ANC and labor every year (NCASC, 2009). Those figures reveal the problem of low utilization of PMTCT service by the pregnant women but the reason behind in lacking the uptake of the services is less identified in Nepal. Male involvement in ANC along with his wife tends to be a supportive initiation for utilizing the PMTCT services and coping with HIV positive status. In Nepal, a study conducted in one of the public maternity hospital indicate that the most prominent barriers to male involvement in ANC and maternal health included low levels of knowledge, social stigma, shyness/embarrassment and job responsibilities (Mullany, 2005). The reduction of 37-50% of mother to child transmission can be achieved using short course antiretrovirals and optimizing obstetrics care practices. In order to use such preventive measures, knowledge of pregnant women and her partner on HIV/AIDS, MTCT and ongoing PMTCT service seems to be of paramount important. Moreover, couples' VCT can provide life-saving benefits; reduce HIV transmission, prevention of HIV, reinfection, sexually transmitted infections, and unintended pregnancies (Allen, Karita, Chomba, et al., 2007; WHO 2008). Unsupportive partners’ attitudes are likely to create a barrier to women’s program participation. The fear of the partner’s reaction in case of positive sero-status disclosure is a major obstacle for women to participate in voluntary counseling and testing and further PMTCT program enrolment. Joint counseling and HIV-testing for couples seems to be a key to success here (Theurin, Mbezi, Luvanda et.al., 2009). TM Studies have shown that the utilization of PMTCT services by the pregnant women is influenced both by factors related to the health system such as accessibility of VCT services, and by individual factors such as fear of disclosure of HIV results, lack of male partner support, fear of domestic violence, abandonment and stigmatization. Participation of men in the antenatal care of their spouses and couple

PDF Editor

3

VCT increase the utilization of interventions to prevent HIV-1 transmission (Byamugisha , Tumwine, Semiyaga, et.al., 2010). The data from ANC surveillance is easy to record and present. However, data from ANC sentinel surveillance shows only of those women who utilize the service and usually from urban centers. All ANC sites are located in government health institutions in urban areas and thus a large proportion of the rural populations and those who do not access governmental facilities do not get recorded in the national surveillance system. Moreover, the women attending ANC clinics are for young and of reproductive age group hence sexually active may not use protective measures. For these reasons, data generated from ANC surveillance are biased and need to be interpreted and used with caution. ANC sentinel surveillance in a low level/concentrated epidemic The main objective of ANC sentinel surveillance are to know the extent of spread of HIV among the general population, to observe HIV trends, to estimate burden of HIV and identify geographical areas of high burden to allocate resource. In a concentrated epidemic, HIV is likely to be first seen among behaviors of populations with most at high risk populations and then spreads to their low-risk partners (general population). Therefore, in a low level/concentrated epidemic as in Nepal, surveillance in populations with high-risk behaviors is a critical first priority. Only relying on the ANC sentinel surveillance and extrapolating to general population group for detecting the emergence of HIV could be misleading and gives a false positive impression. Therefore, in low prevalence settings, the limited available resources should not be used to expand ANC surveillance; instead several homogenous districts should be combined and considered as a single epidemiological zone for the purpose of planning and estimating burden. Both sentinel surveillance and populationbased surveys have strengths and weaknesses, but taken together provide complementary information and can provide a clearer picture of both overall trends and geographical distribution of HIV in a country (Asiimwe-Okiror, 2005). For program planning purposes, not just ANC data but all available program data sources should be used (e.g. PMTCT, blood bank, STI, IBBS, ART, and any community based surveys, such as demographic and health survey) in order to extrapolate for general population. NCASC should analyze all available TM data to prepare a comprehensive report for general population. Also, there should be central guidelines on the use of those data sources and building local resources for triangulation of data.

PDF Editor

4

HIV trends using ANC sentinel surveillance data HIV surveillance has been carried out among women attending antenatal clinics in more than 115 countries worldwide (UNAIDS/WHO/CDC, 2003). This is usually based on anonymous, unlinked, cross sectional surveys of pregnant women attending antenatal clinics in the public health sector. Only firsttime attendees are included to minimize the chance of any woman being included more than once. UNAIDS/WHO estimates of national adult HIV prevalence and the demographic impact of HIV in countries with generalized epidemics have been based on prevalence data collected over time from pregnant women attending antenatal clinics (Walker, 2001; UNAIDS, 2006). Because of differences in prevalence between urban and rural areas, country-specific prevalence is often estimated separately for urban and rural areas, and then combined to obtain a national, weighted estimate of adult prevalence (Ghys, 2004). Historically, the prevalence curve for non-urban areas was adjusted downward by 20% because surveillance systems often did not cover rural areas well, and it was assumed that HIV prevalence was lower in those areas that were excluded from surveillance (Walker, 2001). Antenatal sentinel surveys are usually conducted annually or bi-annually around the same time of the year and involve anonymous, unlinked sampling of blood from pregnant women attending selected antenatal clinics in the public health sector. The main strength of antenatal clinic surveillance is that it provides ready and easy access to a cross-section of sexually active pregnant women from the general population, and it can be used to assess trends in the epidemic over time. In generalized epidemics, HIV prevalence among pregnant women has been considered a good approximation of prevalence among sexually active men and women aged 15–49 years (Grassly et. al., 2004). HIV prevalence data from 26 countries with generalized HIV epidemics were compared which showed that unadjusted estimates of adult HIV prevalence derived from antenatal clinic surveillance are almost always higher than those from national population based household surveys (Gouws et.al, 2008). Nepal has no national surveillance guidelines for survey; however a study suggested a sample size of 400 pregnant women should be consecutively recruited at each sentinel site for monitoring changes in HIV prevalence over a time period (WHO, 2008). The required sample size for monitoring changes in HIV prevalence over time depends on two variables: (i) the baseline HIV prevalence, and (ii) the magnitude of change in prevalence that the program wants to detect from one point of time to another. It is known that the smaller the rate of prevalence, the larger is the required sample size; the smaller the change in magnitude in prevalence that the surveillance wants to detect over time, the TM larger is the sample size to detect a statistically significant difference. To detect a 30% decrease in prevalence (from 1% to 0.7%) in areas with a baseline prevalence of 1%, at 95% confidence level and 80% power, a sample size of 14,000 is required; and to detect a 20% decrease (from 1% to 0.8%) with the same baseline prevalence, the required sample size would be 34,000. Thus, the sample size of 400 tested at ANC sites does not have enough power for monitoring trends even in high prevalence areas. However, data for all the districts can be pooled to get a sufficiently large sample size to monitor trends in country level. In

PDF Editor

5

pooling the data, appropriate adjustments should be made for any differences in socio-demographic characteristics of the populations (WHO, 2008). Prevalence 30% decrease (1-0.7%) 20% decrease (1-0.8%)

Sample size 14,000 34,000

The year-wise fluctuation in a country may be a reflection due to variations attributed to the small sample size rather than to real changes in prevalence of HIV. Thus, administrative decisions at the district level should not be based on changes in prevalence noted from one year to the next. The interpretation of trends requires careful analyses and understanding of data and should be done at central level by NCASC. For this purpose, national guidelines are needed to investigate unusual fluctuations in HIV prevalence in a country. PMTCT program data for surveillance As HIV testing is unlinked and anonymous (i.e. women are not aware that they were tested for HIV) in ANC sentinel surveillance, the data are not biased due to refusal of HIV testing unlike data from PMTCT programs. However, due to the unlinked nature of HIV testing, the HIV test results cannot be communicated to the participants, which deny them an opportunity to know their HIV status and receive HIV services, such as PMTCT, ART, and care and support services. With the rapid expansion and increased availability of PMTCT and care and treatment services, women have a right to know their HIV status. Thus, it is ethically difficult to justify unlinked anonymous testing unless all sites are also providing PMTCT services. PMTCT programs also collect HIV testing information from the same pregnant women as collected by sentinel surveillance sites. Therefore, PMTCT program data could be used for HIV surveillance and could replace sentinel surveillance data. However, the PMTCT program collects monthly aggregated data and reporting is usually incomplete and prone to errors, which could affect the use of PMTCT data for surveillance. Additionally, HIV prevalence estimates from PMTCT program data could be biased if some women refuse to take the HIV test. PMTCT program data offer several advantages if used for HIV surveillance such as it could provide greater coverage and representativeness as the program expands further, the number of women tested is much larger than that tested sentinel surveillance and larger sample sizes would improve the precision of HIV prevalence estimates and it may save costs incurred for sentinel surveys. TM As the PMTCT program expands and its quality improves and the data can be used for monitoring trends in HIV prevalence and for estimating prevalence at the district and national level. However, before discontinuing ANC sentinel surveillance, potential biases in the PMTCT must be evaluated.

PDF Editor

An operational study at selected sites in each district should be conducted to test the feasibility of using PMTCT data and to characterize any biases introduced by women refusing HIV testing. The utility of 6

PMTCT data for surveillance could be increased if regular reporting of PMTCT program data from all centers is ensured and their quality is systematically checked. Also, ANC sentinel surveillance data can be used for evaluation of the quality of PMTCT. Obtaining representative data for HIV estimates Data available from the ANC surveillance tend to overestimate HIV prevalence in the general population. The representativeness of ANC surveillance data is largely compromised by absence of the male population and the limited inclusion of rural populations. Potential sources of obtaining data for the male population are blood donors, para-military and military recruits and potential government employees undergoing general health check-up before recruitment and health checkups for students going abroad and job seekers in foreign countries. However, each of these data sources has some plausibility. Data source ANC data (HIV% amongst ANC attendees) Blood donors data

Strengths Early indication of prevalence among general population Provides opportunity to monitor trend in blood donors

Military recruits

Trend on young male population HIV trend in young population

Weaknesses ANC coverage may not be representative such as Urban vs rural Data is based on single HIV screening test and eligibility requirement for blood donation may result into bias Usually sensitive and may not be easily accessible Not representative

Young population

Not representative

Potential Government employee Students going abroad

Monitoring prevalence among blood donors was one of the early surveillance systems put in place. However, data from this source are particularly prone to bias. Indeed, over time blood transfusion services have changed their policies regarding eligibility for blood donation. Blood donors in recent years are therefore likely to be at lower risk of HIV infection compared with those in earlier years, resulting in bias. Screening of repeat blood donors can provide an opportunity to monitor trends in HIV incidence in this group. However, these trends may not be generalizable to the wider population, as their behavior is likely to differ from the general population given their knowledge of their sero-status (Ghys, 2006). Data from blood donors (surprisingly decreased from prevalence of 0.44 in 2002 to 0.09 in 2009 in Nepal TM although the blood collection is in increasing trend (NRCS, 2009) is based on a single HIV screening test and is also subject to selection bias. Prior to considering blood donor data for use in surveillance, it is important to characterize the magnitude and direction of the bias in using these data as a proxy for the general population. In some countries, potential military recruits serve as an important group to provide data on the young male population. However, these data are usually sensitive and may not be easily accessible and available.

PDF Editor

7

Blood collection, supply and TTIs prevalence – nationwide Indicators

2058/59 2001/02 2002

2059/60 2002/03 2003

2060/61 2003/04 2004

2061/62 2004/05 2005

2062/63 2005/06 2006

2063/64 2006/07 2007

2064/65 2007/08 2008

2065/66 2008/09 2009

Total Blood Collection 72,459 73,758 76,647 82,677 1,03,067 1,15,720 1,21,512 1,36,580 Total Blood Supply 89,414 1,08,643 1,17,125 1,24,142 1,36,630 1,45,162 1,54,107 1,78,652 HIV Prevalence % 0.44 0.44 0.37 0.24 0.40 0.18 0.09 0.09 HBsAg Prevalence % 0.87 1.24 0.88 0.85 0.90 0.37 0.56 0.51 HCV Prevalence % 0.53 0.57 0.42 0.49 0.40 0.54 0.54 0.36 VDRL Prevalence % 0.25 0.23 0.14 0.17 0.30 0.24 0.23 0.18 Note: We prepare blood components from whole blood and supply to the patients. Thus, total number of blood supply is more than total number of blood collection. Source: NRCS, 2009

Private laboratories are another option for obtaining a representative sample of the general population. However, this group is unlikely to truly represent the low-risk general population rather it is likely to be a mix of low-risk and high-risk populations. As Government scales-up PMTCT in the private sector, these PMTCT sites should be included for surveillance. Access to rural populations for surveillance will remain challenging unless HIV testing services are further expanded to rural areas thus capturing those ignored population. Obtaining a truly representative general population sample is difficult without a community based probability sampling approach. In order to make appropriate adjustments to surveillance data, operations research is needed to determine differences between populations who are captured in the surveillance systems and those who are not. For the purpose of estimating burden of HIV in general population, periodic population based surveys that use probability sampling would be needed to calculate appropriate calibration factors to adjust surveillance data. In addition to using population-based survey estimates of adult HIV prevalence to TM calibrate adult prevalence curves fitted to antenatal clinic data in countries where such populationbased surveys have been conducted, a comparison of these estimates can also be used to inform the adjustment factor needed to correct for potential surveillance bias in countries where national population-based surveys have not been conducted.

PDF Editor

8

Surveillance among general population Providing that the methodology used in national population-based surveys is sound and reliable, and potential bias of survey non-response is adjusted for, the population-based survey estimates can be used to adjust antenatal clinic-based prevalence in countries where such surveys have been conducted. The reliability of HIV prevalence estimates obtained from population-based household surveys depends on sound survey and sampling methodology, including taking representative samples of the adult population in relevant subgroups, ensuring high quality of data and specimen collection and employing sound laboratory methods for HIV testing while maintaining high ethical standards. A major challenge in population-based surveys is to minimize the level of non-response, either due to refusal to participate or due to absence of household members at the time of the survey. Mishra (2008) studied the impact of non-response on HIV prevalence in national population- based surveys in 14 countries and showed that non-response did not significantly bias HIV estimates.

Conclusions In areas with low general HIV prevalence, ANC sentinel surveillance only serves limited purpose; in these areas, surveillance among MARPs is most critical and useful. The current sample size of 400 per site/district is too small to conclusively monitor trends at the sub-national level. However, for the high prevalence regions/districts, data could be pooled to monitor trends at the national level. Year-wise changes in prevalence at the district level should not drive administrative and financial decisions. Although promising, PMTCT program data cannot immediately replace regular surveillance. The decision can be made only after appropriate evaluation of the possible biases and improving the quality of PMTCT program recording and reporting. The current surveillance system for the general population focuses on PMTCT data is not truly representative of the general population. For the purpose of HIV estimates, periodic population surveys would be needed to supplement facility-based data from ANC clinic attendees. Recommendations Followings are the recommendations for surveillance among general population:  

Carry out of multi-site operation research using individual level primary data to characterize bias TM among women who do not accept HIV test in the PMTCT program. Scale up PMTCT services to the rural areas of the country in order to make it more representatives to general population and strengthening the PMTCT program, and particularly, improve quality of recording and reporting. Use all available data sources for program planning purpose at the district level, such as PMTCT, ANC sentinel surveillance, IBBS, and blood donors’ data.

PDF Editor 

9



   

Explore the possibility of obtaining data on potential military and para-military recruits, from government pre-employment health check-ups and health checkups for students going abroad and job seekers in foreign countries should be explored to capture surveillance data representing young population. Conduct operations research to characterize the magnitude and direction of bias in using blood bank data as a proxy for the general population should be cautiously considered. Carry out periodic population-based surveys that can capture more representative data from the general population to calibrate ANC data if possible for estimations should be conducted. Strengthen human resource and institutional capacity in surveillance especially for data analyses and interpretation of the results for programming. Develop national guidelines and tools on data analyses and triangulation from multiple sources and preparation of report on HIV for each epidemic region of the country.

References Allen, S., Karita, E., Chomba, E., Roth, D.L., Telfair, J., Zulu, I et. al. (2007) Voluntary counseling and testing for HIV through influential networks in two African capital cities. BMC Public Health 2007, 7:349doi:10.1186/1471-2458-7-349 Byamugisha, R., Tumwine,J.K., Semiyaga,N.& Tylleskar,T. (2010). Determinants of male involvement in the prevention of mother-to-child transmission of HIV program in Eastern Uganda: a cross-sectional survey. Reproductive Health 2010, 7:12 doi:10.1186/1742-4755-7-12. GTZ (2003), Prevention of Mother to Child Transmission of HIV in Kenya, Tanzania and Uganda. Available at: gtz.de/de/dokumente/en-gtz-pmtct.pdf (Accessed on 10 August, 2010). Ministry of Health (MoH) (1977). Nepal Fertility Survey 1976. First Report. Nepal Family Planning and Maternal and Child Health Project, World Fertility Survey. Kathmandu. Ministry of Health (MoH) 2007. Nepal Demographic and Health Survey 2006. New ERA and Macro International Inc. Ministry of Health (MoH) (2002). Nepal Demographic and Health Survey 2001.Kathmandu, Nepal and Calverton, New ERA and ORC Macro, DHS+, Maryland, USA. Mullany, B.C (2005). Barriers to and attitudes towards promoting husbands’ involvement in maternal health in Katmandu, Nepal. Social Science & Medicine. Volume 62, Issue 11, June 2006, Pages 2798-2809

TM NCASC (2009). Prevention of mother to child transmission services in Nepal. Monthly Report on HIV/AIDS Statistics. Kathmandu: NCASC. Available at:www.ncasc.gov.np/pics/category/files/1259600213.pdf (Accessed on 10 August, 2010).

PDF Editor Phillips, J. F. (eds) Fertility Transition in South Asia. Oxford University Press, Oxford.

Shrestha, L. (2005). Prevention of Mother to Child Transmission of HIV in Nepal. 3rd International conference on HIV pathogenesis and treatment. NCASC, Kathmandu,Nepal. Abstract No.TuPe5.2P18. 10

[online] Available at:www.aegis.com/conferences/iashivpt/2005/TuPe5-2P18.html (Accessed on 1 August, 2010). Theuring,S., Mbezi,P., Luvanda,H., Harder, B.J., Kunz, A.& Harms, G.(2009). Male involvement in PMTCT Services in Mbeya Region, Tanzania. AIDS Behav (2009) 13:S92–S102. DOI 10.1007/s10461-009-9543-0 UNICEF. (2008). Children and AIDS. [online] Available at:www.uniteforchildren.org/.Children_and_AIDS_Fourth_Stocktaking_Report_EN_120209.pdf (Accessed on 1 August, 2010). World Health Organization (2001). HIV-AIDS in South Asia Region. World Health Organization. Regional Office for South-East Asia. New Delhi World Health Organization (2007). Guidance on global scale up of the prevention of mother to child transmission of HIV. Avenue Appia, Geneva, Switzerland. Available at: www.unicef.org/aids/files/PMTCT_enWEBNov26.pdf (Accessed on 17th June, 2010). World Health Organization (2007). Prevention of Mother to child transmission. Department of HIV/AIDS. World Health Organization (2008). Prevention Technical consultation to review HIV Surveillance. SEARO, WHO. World Health Organization (2010). HIV/AIDS program towards universal access by 2010. Avenue Appia, Geneva, Switzerland. World Health Organization (2010). PMTCT Strategic Vision 2010-2015. Avenue Appia, Geneva, Switzerland. Walker, N., Stanecki, K.A., Brown, T. et al (2001). Methods and procedures for estimating HIV/ AIDS and its impact: the UNAIDS/WHO estimates for the end of 2001. AIDS 2003;17:2215–25. UNAIDS. 2006 Report on the global AIDS epidemic. Geneva: UNAIDS. Ghys, P.D., Brown, T., Grassly, N.C., et al (2004). The UNAIDS Estimation and Projection Package: a software package to estimate and project national HIV epidemics. Sex Transm Infect; 80 (suppl 1):i5–9. UNAIDS/WHO/CDC (2003). Guidelines for conducting HIV sentinel sero-surveys among pregnant women and other groups. Geneva: WHO and UNAIDS. Nepal Red Cross Society (2009). Prevalence of HIV, HBsAG, HCV, VDRL. Kathmandu, Nepal.

TM Mishra, V., Barrere, B., Hong, R., et. al. (2008). Evaluation of bias in HIV sero-prevalence estimates from national household surveys. Sex Transm Infect.

PDF Editor

Asiimwe-Okiror, A., Knight, R & Gouws, E. (2005) Guidelines for measuring national HIV prevalence in population based surveys. Geneva: UNAIDS and WHO.

11

Grassly, N.C., Morgan, M., Walker, N., et al. (2004). Uncertainty in estimates of HIV/AIDS: the estimation and application of plausibility bounds. Sex Transm Infect; 80 (suppl 1): i31–8. Gouws, E., Mishra, V & Fowler, T.B. (2008). Comparison of adult HIV prevalence from national population-based surveys and antenatal clinic surveillance in countries with generalised epidemics: implications for calibrating surveillance data. Sex Transm Infect; 84 (Suppl I):i17–i23. doi:10.1136/sti.2008.030452. Ghys, P.D., E Kufa, E & George, M.V. (2006). Measuring trends in prevalence and incidence of HIV infection in countries with generalised epidemics. Sex Transm Infect;82 (Suppl I): i52–i56. doi: 10.1136/sti.2005.016428.

TM

PDF Editor

12