Large-scale STI services in Avahan improve utilization and treatment seeking behaviour amongst high-risk groups in India: an analysis of clinical records from six states

Background Avahan, the India AIDS Initiative, implemented a large HIV prevention programme across six high HIV prevalence states amongst high risk groups consisting of female sex workers, high risk men who have sex with men, transgenders and injecting drug users in India. Utilization of the clinical services, health seeking behaviour and trends in syndromic diagnosis of sexually transmitted infections amongst these populations were measured using the individual tracking data. Methods The Avahan clinical monitoring system included individual tracking data pertaining to clinical services amongst high risk groups. All clinic visits were recorded in the routine clinical monitoring system using unique identification numbers at the NGO-level. Visits by individual clinic attendees were tracked from January 2005 to December 2009. An analysis examining the limited variables over time, stratified by risk group, was performed. Results A total of 431,434 individuals including 331,533 female sex workers, 10,280 injecting drug users, 82,293 men who have sex with men, and 7,328 transgenders visited the clinics with a total of 2,700,192 visits. Individuals made an average of 6.2 visits to the clinics during the study period. The number of visits per person increased annually from 1.2 in 2005 to 8.3 in 2009. The proportion of attendees visiting clinics more than four times a year increased from 4% in 2005 to 26% in 2009 (p<0.001). The proportion of STI syndromes diagnosed amongst female sex workers decreased from 39% in 2005 to 11% in 2009 (p<0.001) while the proportion of STI syndromes diagnosed amongst high risk men who have sex with men decreased from 12% to 3 % (p<0.001). The proportion of attendees seeking regular STI check-ups increased from 12% to 48% (p<0.001). The proportion of high risk groups accessing clinics within two days of onset of STI-related symptoms and acceptability of speculum and proctoscope examination increased significantly during the programme implementation period. Conclusions The programme demonstrated that acceptable and accessible services with marginalised and often difficult–to-reach populations can be brought to a very large scale using standardized approaches. Utilization of these services can dramatically improve health seeking behaviour and reduce STI prevalence.


Background
In India, high risk groups (HRG) which include female sex workers (FSWs), men who have sex with men (MSM), injecting drug users (IDUs) and transgenders (TGs) have high HIV prevalence based on data from the national sentinel surveillance 2009 and other studies [1][2][3]. Concentrated epidemics amongst HRGs require an integrated service approach for HIV prevention efforts [4][5][6]. Avahan is a large scale HIV prevention programme in six states of India which have a population of about 300 million. The main components of the Avahan intervention for HRGs are peer-led outreach education, condom promotion and distribution, clinical services for managing sexually transmitted infections (STIs), community empowerment and structural interventions.
High burden of STIs amongst HRGs has been reported from most developing nations [7]. Improved primary care of symptomatic and asymptomatic STIs through programmes targeting vulnerable groups has been promoted globally to achieve scaling-up, high coverage and decrease in STI burden [8][9][10]. STI prevention amongst high risk groups is the cornerstone for prevention efforts in concentrated HIV epidemics [11,12].
Provision of quality STI services to HRGs is low in developing countries [13]. Developing acceptable, accessible STI services for HRGs in resource-constrained settings has been challenging due to prevalent stigma, discrimination, socio-cultural barriers and affordability issues. Studies have demonstrated that when STI services are designed to address these issues, utilization by HRGs improves [14]. To further improve service utilization and health seeking behaviour, standardized treatment guidelines, a robust STI service delivery system, regular supervision and tracking coverage are critical. However, until now interventions tracking coverage using unique identifiers or individual tracking systems amongst HRGs have been reported rarely and at small scale [15].
We report here an analysis using routine clinical monitoring data which is based on individual tracking of HRGs in Avahan. The analysis focuses on STI services provided to HRGs; assessing utilization, health seeking behaviour and STI syndrome trends between 2005 and 2009.

The programme
Avahan is an HIV prevention programme for FSWs, MSM, TGs and IDUs implemented in six states of India with historically the highest prevalence of HIV: Andhra Pradesh, Tamil Nadu, Maharashtra, Karnataka, Manipur and Nagaland [16]. The programme is implemented by seven lead implementing partners through a network of about 130 local non-governmental (NGOs) and community-based organizations (CBOs) from 2004 to date. While the drivers of the epidemic in the four southern states are mainly sexual, the epidemics in the two north-eastern states are driven by injecting drug use [10].
At the individual level, primary and secondary prevention of STIs is a key Avahan strategy. Avahan-supported clinics provide STI services to HRGs in the six intervention states .In addition, the clinical services in the two north eastern states also address injection related infections with less emphasis on STIs compared to the southern states. Avahan-supported STI services are standardized across the states in which an essential STI service package was defined [17]. The package consisted of syndromic management of symptomatic infections as per Indian national guidelines; presumptive treatment for gonorrhoea and chlamydia at the first visit which was repeated if the individual had not attended the clinic for any STI check ups in the previous six months; quarterly STI check ups and six-monthly syphilis screening. IDUs were provided services for symptomatic STIs and injected related infections.

Routine clinical monitoring data base
A component of the monitoring system was individual tracking data pertaining to registration and clinical services. At the NGO-level (NGOs provided service coverage at the district level), each individual was provided a unique identification number which was based on the project identification number, the district code and the number of the peer educator responsible for that individual. This unique identifier was used to maintain paperbased NGO clinic records for the individual over time. Information for each clinical visit was recorded in a standardized clinical encounter form, entered into a computerized database by the NGOs/CBOs and collated electronically by the lead implementing partners. Quality control of the database was maintained by the individual implementing units and state lead partners.

Individual STI tracking data
Individual STI data was collated from all lead implementing partners to create the pan-Avahan individual tracking STI database. Data variables from the database and their description are listed in Table 1.
Locally described typologies from across the states were merged to define uniform typologies across Avahan. The typologies mainly reflected the place of solicitation for FSWs and were classified as such. Street based FSWs were those who solicited clients on the streets; while home based FSWs solicited clients at home, similarly bar based FSWs were found mainly soliciting in bars; and brothel based FSWs soliciting points were organized brothels. Amongst MSM, the typologies were classified as per self-reported sexual identity generally based on roles in anal sex. "Kothis" were self-identified anal receptive MSM, "Panthis" were self-identified anal insertive partners and "Double-deckers" were self identified anal insertive as well as anal receptive partners.

Data sources include
(1) Avahan programme generated HRG size estimates: at the start of the programme in a district or sub-district, NGOs conducted a formal external mapping and size estimation exercise. Some state-level lead implementing partners updated these numbers on a regular basis (every 12 to 18 months) using programme data; others conducted periodic formal size estimation exercises. Size estimates were done separately for FSWs, MSM, IDUs and TGs and were available consistently from 2007 onwards [18,19].
(2) Data from clinical encounter forms at the level of individual clinics: registration details obtained at the first clinic visit included including age, sex and typology. At this time a unique clinic number was assigned. Clinical encounter forms labelled with the individual's clinic number were filled for each clinic visit.
The merged database of registrations and subsequent clinical visits were cleaned in consultation with the lead implementing agencies. During the cleaning process clinical encounter forms missing unique clinic numbers or dates of visits were deleted (for details, see Table 1). In the case of the variable, duration of symptoms prior to clinic visit, the format for recording was changed from a categorical to a continuous variable in 2007. Data was recoded to categorical variables to utilize all years of data.

Analytical approach
The data was cleaned and merged using MS-Access 2003™ and analyzed using statistical package STATA™ version 10. Four broad areas of enquiry formed the basis of the analysis of the paper. The sections below describe in details the analytical approaches and data sources: (1) Demography and proportion of HRGs reaching the clinics Age at first clinic visit, number of years into sex work, number of clients per week by typology were analyzed. The proportion of HRGs using clinical services was estimated using the number of individuals from clinic records and the estimated denominator from the programme. This analysis for the clinical coverage of HRGs was restricted to the period 2007 to 2009 as the previous years were still being utilized to increase the scale of services.

(2) Utilization of the clinics
The analysis included the number of visits per year by individuals. Trends in number of repeat visits by HRGs by typology over the programme implementation period were analyzed, adjusted for age. "STI visits", were defined as visits by HRGs who attended the clinics for relevant symptoms, regular-check ups or follow up visits within 14 days of a previous symptomatic visit.
(3) Improving health seeking behaviour and changing trends in STI syndromes Treatment seeking behaviour was analyzed by duration of symptoms prior to reporting to the clinics, trends in regular STI check ups and internal examination. STI syndromes diagnosed and trends were analyzed by adjusting confounders which included age and subpopulation groups. "Cohorts for the year", were defined as those HRGs coming to clinics for the first time in that year and followed up subsequently.

(4) Quality of clinical services provided
Prescription analysis of correct treatment given as per standard packs described in the Avahan Clinic Operating Guidelines and Standards (COGS) were analyzed.

Results
A total of 431,434 high risk individuals made 2.7 million visits to the targeted programme clinics in the six states. The HRGs consisted of 331,533 FSWs, 10,280 IDUs, 82,293 MSM and 7,328 TGs.
The age distribution of HRGs varied by typology at their first clinic visit, as shown in Table 2.Amongst FSWs, the bar-based were the youngest with a mean age of 24.7 years (SD 5.0); among the MSM the youngest were "Panthis" with a mean age of 27.8 years (SD 7.7). The reported number of commercial clients in the previous week was significantly higher amongst the highway-based FSWs who had an average of 12 clients per week; followed by the brothel-based (eight clients per week) and street-based (six clients per week). Amongst the MSM who reported selling sex, "double-deckers" had the highest number of partners with four clients per week. The mean number of years in sex work for FSWs and MSM were 3.1 (SD 4.4) and 1.8 (SD 3.6) respectively.
Individuals made an average of 6.  Table 3.
New sex workers (defined as new into the project as shown in Table 4) who accessed the Avahan clinics increased from 43, 394 in 2005 to 220,877 in 2008. However, retention within the cohorts by years of follow up showed a declining trend and ranged from 22% to 25% over the five year period of follow up.
The number of repeat visits was found to increase consistently amongst FSWs, MSM and IDUs from 2005 to 2009 as shown in Table 5. HRGs who visited the clinics more than four times per year increased over the years. An increasing proportion of HRGs attended the clinics for regular STI check-ups during the period. The proportion of clinic attendees undergoing internal examination (i.e. vaginal speculum or proctoscopy) increased from 10 % to 53 % amongst FSWs, from 1% to 54% amongst MSM and from 0.7% to 27 % amongst TGs from the year 2005 to 2009. Treatment seeking behaviour improved with an increasing proportion of HRGs coming to the clinics within two days of the onset of symptoms.
There was a declining trend in the proportion of all syndromes diagnosed amongst HRGs from 2005 to 2009, as shown in Table 6. There was a decline in STI syndromes occurred amongst FSWs, MSM, TGs and IDU (p=<0.001). Amongst MSM self identity of being anal receptors, penetrates or both did not match the STI syndromes diagnosed. While the STI regular check up visits increased over the quarters of follow-up, there  was a distinct decreasing trend in syndromes diagnosed, as shown in Fig -1.
Based on comparisons of the records of syndrome treatment packets prescribed and the recorded diagnosis, treatment was dispensed correctly 61% to 92% of times. On an average, correct treatment for FSWs with a diagnosis of lower abdominal pain was the poorest at 61%; FSWs with vaginal discharge were treated correctly over 90% of the time, as shown in Table 7.

Discussion
The analysis of five years of individual clinic tracking records of HRGs shows improved health seeking behaviour, declining trends in STI syndromes and increasing utilization of services provided by Avahan clinics across six states. More detailed analysis of the Avahan programme data and quality issues have been described in previous reports that showed a high level of infrastructure by Avahan and quality STI services [20]. These data indicate that quality STI services were brought to scale across the Avahan programme districts, resulting in reduced prevalence of STI syndromes amongst individuals attending the Avahan clinics.
Prevention of STI transmission to and from sex workers is critical to limiting the establishment and expansion of these epidemics at the population level. The role of HRGs and client groups in the epidemiology of a particular STI depends upon the frequency and nature of commercial sex transactions and the transmission dynamics of each STI. As STI treatment and prevention programmes improve in quality and expand in scope, the duration of infectiousness and perhaps the transmission efficacy of the targeted STI should decrease [21][22][23].
However, HRGs experience a high degree of social marginalization and discrimination in society especially from healthcare providers and therefore do not have adequate or equitable access to health services. Access to services can be a key motivator for many FSWs to interact with programme staff and to participate in programme activities [24].
Access to health services is determined by three factors, the health seeking behaviour of the population, the health care provider's attitudes and the healthcare delivery systems [25,26]. In the present analysis, the healthcare seeking behaviour was studied amongst marginalized groups who showed improving health seeking behaviour reflected by the increased number of clinic visits, the increasing proportions coming for STI check-ups and early treatment seeking behaviour trends over the programme implementation period. Trends in utilization, however, did vary as per the typology of FSWs, MSM and IDUs. Recent research reveals a more complex picture of STI epidemiology amongst HRG [27]. This complexity is based on HRG populations who actually are diverse sub-populations within themselves, each with distinct population characteristics. In the present analysis brothel based female sex workers were easily accessible compared to the street based or home based sex workers. Amongst MSM the "Kothis" were less accessible compared to other typologies. Health seeking behaviour thus may be related to the population characteristics within sex worker groups. Increasing proportions of internal examination during routine STI check-ups in FSWs and MSM indicate that over a period of time, both health care providers and clinic attendees reached a degree of comfort and accepted it as a norm. Thus, improving health seeking behaviour and acceptable service provision is possible on a large scale.
A report of a cohort of female sex workers in Pune provided with STI services followed over a nine-year period showed declining trends of genital ulcer disease prevalence while the vaginal discharge syndrome remained stable [28]. In our analysis, all STI syndromes diagnosed amongst those who attended clinics showed a declining trend, while regular STI check-up visits increased consistently The present data show a declining trend in syndromes diagnosed possibly because of the essential service package approach followed in Avahan clinics. Though in the absence of controls this cannot be a direct measure of causality the scale of such a declining trend in STI syndromes in all six states of India could not be possible without a large intervention amongst HRGs such as Avahan.
Providing quality STI services encourage STI patients to seek care at such facilities. Quality services are technically sound and based on evidence-informed standard guidelines [29]. Though perceived quality was not measured in the present analysis, high standards were maintained by using the STI service guidelines detailed in the COGS which were developed early in the Avahan programme. Experience in STI control programmes indicates that assessment and improvement of service quality is an essential part of programme management, leading to a more effective and efficient use of resources . Recent efforts to assess the quality of STI services have relied on review of patient records, simulated patients and observational methods as data sources, which are difficult to implement in resource-constrained scenarios [30]. According to the WHO protocol for STI case management, Prevention Indicator 6 (PI 6), now renamed as (HIV-prevention indicator) PI 11 measures different components of STI case management including history taking, examination and correct treatment given as per the clinical diagnosis [31]. In the present analysis, correctness of treatment was measured by comparing the treatment packets with the syndromes diagnosed. The analysis found that correct treatment, ranging from 61% to 95%, was given for all syndromes. In comparison, another study in Nairobi, Kenya showed that correct treatment for syndrome management was given 33% of times with a range of 9% to 63% [32]. In studies carried out in the rural health districts of KwaZulu-Natal, South Africa, and correct treatment improved when treatment packs were standardized [33]. Our analysis is one of the first to report high standards of correct treatment for STIs using coded packs implemented on a large scale in India. The present analysis showed that the population coming to the clinics is changing constantly, gauging by the changing clinic attendance patterns. In addition to HRGs being socially, culturally and economically marginalized, mobility itself presents barriers to health care access [34]. Changing patterns of new individuals coming in and drop outs amongst the clinic attendees over the years reveal the dynamic nature of the cohorts being followed. This perhaps reflects mobility amongst HRG population in India. There is a gap in our understanding of the dynamics of mobility in sex work and its impact on STI prevention [35]. Hence, in India structural interventions and STI prevention strategies need to address mobility so that impact of these strategies is not mitigated.

Limitations of the analysis
The analysis of the paper was based on clinical records and hence the findings cannot be generalized to the community at large Another key limitation of the analysis was that the individual tracking system was NGOspecific. Sex workers in India are a mobile population and often move, either for short term or longer term, from one solicitation point to another [36]. Accordingly, such individuals registered in one NGO clinic may have registered again in another NGO clinic(s) given the programme's extensive reach in 83 of 129 districts in six states. This multiple registration may explain the over 100% clinic utilization of certain typologies of HRGs. The denominators for the size estimates were calculated by the state lead implementing partners through various research agencies, hence may not be of similar quality across the states giving rise to under or over estimates [37]. Finally, this analysis examined the data across all Avahan sites. The authors recognize that there are potential geographical variations across the state and within districts both in utilization as well as with service delivery models (e.g., static clinics, mobile clinics, and  health camps) but such analyses were beyond the scope of this paper.

Conclusions
The analysis of routine MIS data from STI clinical services shows that utilization of health services by marginalized groups can be dramatically improved with efforts to make clinics acceptable and accessible through quality services; along with outreach efforts to promote health seeking behaviour. Utilization of quality services results in improved treatment seeking behaviour and decrease in prevalence of STI syndromes.