Transitioning from HIV sentinel surveillance to programme based surveillance

[Guest guest]

Transitioning from HIV sentinel surveillance to programme

based surveillance

Commentary

Indian J Med Res 132, September 2010, pp 245-247

Kumar et al1 have presented evidence in support of potential use of data

emerging from screening of pregnant women at prevention of parent-to-child

transmission of HIV centers (PPTCT) in place of the annual HIV sentinel

surveillance (HSS) data for determining HIV prevalence trends.

Ever since its introduction in 1998 in India2, annual HIV sentinel surveillance

(HSS) has remained the mainstay for monitoring trends of HIV epidemic in high

and low risk populations and also for HIV disease burden estimation in India.

The HSS programme was initially critiqued for its non-uniformity and lack of

adequate coverage3; however, its expansion over the last decade has

significantly improved the coverage.

In 2008, HSS was conducted at 660 sites among pregnant women (ANC sites) in the

country4. Although HSS has remained the most reliable and readily available HIV

data source in India, there have always been concerns about the survey

methodology.

HIV annual sentinel surveillance involves collection of blood samples from 400

consecutive women attending antenatal clinic facility at ANC survey site for the

first time in a time frame of 3 months5.

Justification for such an approach is to prevent selection bias.

However, there have always been concerns about implementing this approach,

especially at ANC sites with large attendance and which are operated by many

specialists6. As part of routine ANC care, blood samples are collected for

haemoglobin estimation and/ or syphilis testing and an aliquot is utilized for

HIV testing after removing the identifiers. In this approach there are almost no

instances of refusals and the data obtained are unlikely to suffer from refusal

bias.

The methodology of unlinked anonymous testing (UAT) practiced in India for ANC

surveillance does not require informed consent for HIV testing. However,

globally, there is a growing insistence on conduct of HIV surveillance with

appropriate informed consent and counselling.

It is possible that the Indian HSS programme will make the necessary amendments

to comply with this recommendation in the near future. It has been commented

that in absence of counselling due to UAT in HSS programme, possible opportunity

for intervention directed at PPTCT is missed. The limitations to

generalizability of HSS data are that the sample collection period is only for 3

months and not throughout the year and the sample size of 400 can hardly be

considered as representative of the district7.

Most of the HSS data are captured from the surveillance sites that are housed in

public hospitals or dispensaries and representation of the private sector is

very poor. The profile of clients who visit the public sector is completely

different than that of clients visiting the private sector.

Educated class with higher income more often opts for private sector and it is

quite possible that HIV prevalence among women attending private hospitals is

lower than those visiting public hospitals.

Thus, for the ANC surveillance data to be more generalizable, it is important to

extend the HSS network to achieve better geographic coverage, include more sites

in the private sector and also increase the sample size considerably. However,

all of these would significantly increase the operating cost of the HSS

programme.

With the documented evidence of prevention of mother-to-child transmission of

HIV8-10, more and more countries have been focusing on increasing coverage and

screening of pregnant women for HIV and providing appropriate anti-retroviral

medications to HIV infected pregnant women to protect their babies from HIV. The

PPTCT programme has been rolled out in India by the National AIDS Control

Organization and 6437 PPTCT centers are functional as on September 200911.

These centers carry out testing of pregnant women for HIV on an ongoing basis

throughout the year. The reach of PPTCT centers in terms of width and depth and

the number of samples collected is significantly higher when compared to the HSS

programme.

This strategy offers the test to all pregnant women reporting to ANC facility

and might have a selection bias if many pregnant women visiting PPTCT centers

refuse or opt out of HIV testing.

Additionally, while HIV testing algorithms in HSS and PPTCT are a little

different, the quality control and quality assurance mechanisms are identical in

HSS as well as PPTCT.

The number of PPTCT sites far exceeds those in HSS; and provide much greater

coverage in terms of number of women tested and are likely to be more

representative than HSS due to continuous geographic expansion.

Therefore, through the PPTCT programme an opportunity has now been created to

obtain year-round HIV prevalence data in pregnant women.

Public health experts and programme managers in India have started considering

the possibility of using PPTCT data and gradually waning off the HIV sentinel

surveillance. As the information will be collected as part of the ongoing PPTCT

programme, significant costs for conducting HSS are likely to be averted.

However, while transitioning from HSS data to PPTCT data we have to be cautious.

Many PPTCT centers are facing major challenges in data entry, analysis and

reporting due to lack of adequate infrastructure and human resources to

efficiently handle the emerging data8. Kumar et al have rejected a large number

of PPTCT sites from their analysis, probably for this very reason1.

There could be different challenges in managing the PPTCT data at the site,

district, state and national level, but with systematic approach to fault

finding and appropriate modifications in the programme; the PPTCT data would

become more and more reliable over the next few years. Timely data entry and

implementation of pre-defined quality assurance and quality control protocols

will improve the PPTCT programme data quality in future.

The analysis presented by Kumar et al shows that the HSS and PPTCT data

correlated well at the State level, but not at the district and site levels1. It

could be a function of number of samples, but the precise reasons for this

observation need to be investigated.

The reported proportion of women opting out of HIV testing in the PPTCT settings

was around 32 per cent for the years 2005-2007 and a significant variation was

observed across various States. Determining the precise reasons for opting out

is crucial to the proposed transition from HSS to PPTCT.

Another significant point is that the HSS programme is ongoing since 1998 and we

have data available for comparison. Data elements or variables collected over

the years in HSS will have to be included in data collection under the PPTCT

programme for retaining the ability to analyze trends over years.

The phase III of the National AIDS Control Programme has emphasized on districts

for programme implementation as well as data analysis.

Kumar et al1 have observed a lower correlation between HSS and PPTCT data at

district level probably due to increased opt-out rates in the PPTCT programme.

It is possible to estimate cut-off values for PPTCT opt-out rates which are

likely to yield acceptable correlation between HSS and PPTCT data.

It will be important to elevate the quality of the PPTCT programme to maintain a

minimum opt-out rate at each site.

It may not be prudent to switch over completely to PPTCT based HIV trend

analysis yet, it might be helpful to continue HSS till the PPTCT programme

becomes adequately mature and starts yielding robust data. Issues related to

acceptable opt-out rates, reasons for opt-out, differences in HIV test protocols

and overall improvement in data quality and timely reporting in PPTCT programs

should urgently be addressed.

Sanjay Mehendale

National AIDS Research Institute (ICMR)

G-73, MIDC, Bhosari, Post Box 1895

Pune 411 026, India

smehendale@...

References

1. Kumar R, Virdi NK, Lakshmi PVM, Garg R, Bhattacharya M, Khera A. Utility of

prevention of prevent-to-child transmission (PPTCT) programme data for HIV

surveillance in general population of India. Indian J Med Res 2010; 132 : 256-9.

2. HIV Fact Sheet Based on HIV Surveillance, Data in India 2003-2006. National

AIDS Control Organization, Ministry of Health and Family Welfare, November 2007.

Available at:

http://www.nacoonline.org/upload/NACO%20PDF/HIV_Fact_Sheets_2006.pdf, accessed

on May 28, 2010.

3. National Institute of Health and Family Welfare (NIHFW), GOI Orientation

workshop for specialists from regional coordinating centers. National annual

sentinel surveillance for HIV: Introductory document by NIHFW 1998 Jan 19-21,

New Delhi.

4. United Nations General Assembly Special Session (UNGASS). India Country

Progress Report (2010) New Delhi: Ministry of Health and Family Welfare; 2010

5. Operational Guidelines for HIV Sentinel Surveillance, 2008: National AIDS

Control Organization, Ministry of Health and Family Welfare, Government of

India, New Delhi 2008.

Available at:

http://www.nacoonline.org/upload/Publication/M & E%20Surveillance,%20Research/Oper\

atioal%20Guidelines %20for%20HIV%20Sentinel%20Surveillance%202008.pdf, accessed

on May 28, 2010.

6. Lal S. Surveillance of HIV/AIDS epidemic in India. Indian J Commun Med 2003;

28 : 3-9.

7. World Health Organization. Technical Consultation to Review HIV Surveillance

in India: HIV surveillance in the general population. New Delhi: WHO SEARO;

2008. p. 15-21.

8. Connor EM, Sperling RS, Gelber R, Kiselev P, G, O'Sullivan MJ, et al.

Reduction of maternal-infant transmission of human immunodeficiency virus type 1

with zidovudine treatment. Pediatric AIDS Clinical Trials Group Protocol 076

Study Group. N Engl J Med 1994; 331 : 1173-80.

9. The Joint United Nations Program on HIV/ AIDS (UNAIDS). Prevention of HIV

transmission from mother-to-child: Report of Meeting on planning for programme

implementation, Geneva, 23-24 March 1998. Available at:

http://data.unaids.org/Publications/IRC-pub03/meetingmarch98_en.pdf, accessed on

May 30, 2010.

10. Volmink J, Siegfried NL, van der Merwe L, Brocklehurst P. Antiretrovirals

for reducing the risk of mother-to-child transmission of HIV infection. Cochrane

Database Syst Rev 2007 Jan 24; (1):CD003510.

11. Children and AIDS: Fourth Stocktaking Report 2009. Available at:

http://www.unicef.org/india/media_5896.htm,