Primary study interventions
In Bangladesh, NSP were expanded significantly starting in 1998 with full implementation occurring in 2000. By 1999, the NSP served a population of 3,500 PWID, covered 89-93% of the PWID in Dhaka, and distributed up to three syringes and six needles per injector per week with an average needle and syringe exchange rate of 73% among program participants .
Azim 2008  and Azim 2009  recruited male PWID from NSP, detoxification centers, and clinics in Dhaka, Bangladesh to analyze prevalence and incidence of HIV and HCV infection. Azim 2008  examined 3,759 male PWID from 1999–2006 and documented an HIV prevalence increase from 1.4% (1999–2000) to 7% (2006). HCV infection prevalence decreased from 66.5% (1999–2000) to 56.4% (2006).
Azim 2009  analyzed PWID HIV incidence from 1998–2007. The HIV incidence among PWID remained stable at 1.16/100 PY (1998–1999) to 1.29/100 PY (2007) while HCV infection incidence among PWID decreased from 37.50/100 PY (1998–1999) to 11.58/100 PY (2007).
Azim 2009 reported decreases in several injection related behaviors; borrowing of needles/syringes decreased 10.6% while lending of needles/syringes decreased 29.8%. Additionally, long term PWID were more likely to utilize the NSP; those who had at least a 10 year history of injection drug use were nearly three times more likely to utilize the NSP compared to PWID who had been injecting for less than 10 years.
Porto Alegre Brazil
NSP programs in Porto Alegre were started in 1996 with cooperation from the public health system in conjunction with municipal and state authorities. In 1998, the NSP program distributed approximately 48,000 needles/syringes, increasing to over 150,000 in 2002 with an exchange rate of nearly 50%; the NSP serve a population of approximately 7,000 to 8,000 PWID.
Caiaffa 2003  recruited PWID from non-institutionalized locations in Porto Alegre in 1998 (when NSP was still in its scale up and expansion period), and again in 2000–2001 (when NSP had become fully established in the city). PWID in both samples were predominately young (average age: 28–31 years old), male (81-84%), and injected mainly cocaine (74-87%). Both samples had moderate to high levels of sharing in the last six months (36-59%), and 60% had ever visited the NSP at least once. HIV prevalence among PWID in Porto Alegre during NSP implementation increased from 48.5% (1998) to 64.3% (2000–2001). No explanation was given by authors related to the changes in HIV infection prevalence seen among locations without NSP in place.
Dagou & Luzhai China
In 1998 the Ministry of Health of China promoted social marketing of safe injection as part of their HIV/AIDS prevention strategy. The first NSP were started in Dagou and Luzhai with ramp up beginning in 2002. The number of NSP locations in China increased from 93 sites in March 2006 to 729 by December 2006.
Wu 2007  analyzed the impact of NSP on HIV prevalence in Dagou and Luzhai and compared results to two cities in China that did not have NSP programs in place, Yu’nan and Yongning. PWID were recruited into the study from several locations including detoxification centers and community outreach. In all four locations in which PWID were recruited (Dagou, Luzhai, Yu’nan, and Yongning), PWID were predominantly male and between 20 and 39 years of age.
Both locations that had NSP documented decreases in HIV and HCV infection prevalence. In Dagou, HIV prevalence decreased from 40% (2002) to 33.6% (2003) (p = 0.16), while HCV infection prevalence decreased from 98.7% (2002) to 88.5% (2003) (p < 0.01). In Luzhai, HIV prevalence decreased from 56.2% (2002) to 53.2% (2003) (p = 0.54), while HCV infection prevalence decreased from 88.7% (2002) to 84.5% (2003) (p = 0.22).
In the cities that did not have NSP in place, changes in HIV and HCV infection prevalence varied. In Yu’nan, HIV prevalence remained stable at 17.6% (2002–2003) (p = 0.99), while HCV prevalence decreased from 88.1% (2002) to 58.5% (2003) (p < 0.01). In Yongning HIV prevalence increased from 22.4% (2002) to 24.1% (2004) (p = 0.68) while HCV infection prevalence increased from 81.6% (2002) to 88.2% (2004) (p = 0.07). No explanation was given by authors related to the changes seen in HIV and HCV infection prevalence seen among locations without NSP in place.
Lang Son Province Vietnam & Ning Ming County China
The “Cross-Border” intervention took place in the regions of Lang Son Province, Vietnam and Ning Ming County in the Guangxi province in China. The intervention involved packaged harm reduction services, including a pharmacy-based voucher program for acquiring clean needles/syringes along with clean injecting equipment and condoms. On average, 7,000 to 10,000 needles/syringes were distributed per month at each location, serving a population of approximately 3,000 PWID in each region. Three studies reported on changes in HIV prevalence in these locations, Hammett 2006 , Des Jarlais 2007  and Hammett 2012 ; this review includes the more complete analysis utilizing a longer follow-up period from Hammett 2012.
Hammett 2012  documented changes in HIV prevalence and HIV incidence among new PWID over an eight year period in conjunction with NSP expansion in both locations beginning in 2002 of which six years of data was available for Ning Ming while seven years of data was available for Lang Son. 2125 PWID were included in the Ning Ming sample and 2677 PWID were included in the Lang Son sample.
In Ning Ming, HIV prevalence decreased from 17% to 14% after 12 months, and then stabilized at 11% after 72 months (p = 0.003). In Lang Son, HIV prevalence decreased to 43% after 12 months and decreased further to 23% after 84 months (p < 0.001). When examining only new PWID, defined as injectors that had injected for three years or less, HIV incidence in Ning Ming decreased from 12/100 PY to 9/100 PY after 12 months and stabilized at 11/100 PY after 72 months. Among new PWID in Lang Son the HIV incidence decreased from 22/100 PY to 16/100 PY after 12 months and decreased further to 3/100 PY after 84 months (p < 0.001).
NSP was implemented in Estonia in 2003 with the majority of services provided in the capital city of Tallinn, where nearly 75% of the PWID in Estonia are located. The NSP serve a population of approximately 10,000 PWID and have increased needle/syringe distribution greatly since implementation, from 18,000 needles/syringes distributed in 2003 to over 770,000 by 2009.
Uusküla 2011  recruited 1,027 PWID between 2005 and 2009, of which 168 were new injectors (defined as having injecting histories of three years or less). The sample was predominately male (80%), Russian (80%), and young (mean ages: 24 to 27 years old). Among new injectors, subjects were predominately male (74-82%), Russian ethnicity (78-89%), and mainly injected fentanyl (48-61%) or amphetamine (32-47%). High levels of receptive needle/syringe sharing were documented among new PWID, with rates ranging from 74-79%.
HIV prevalence in the entire sample decreased slightly from 54% (2005) to 50% (2009). However, when examining only new injectors, HIV prevalence decreased from 34.2% (2005) to 15.8% (2009) (p = 0.046); after controlling for age, gender, injection frequency and NSP use, the change in overall HIV prevalence among new injectors remained statistically significant (χ2 = 8.31, p = 0.016). Estimated HIV incidence among new injectors decreased from 20.9/100 PY (2005) to 9/100 PY (2009) (p = 0.026).
Several injection-related behaviors were measured in the study among the PWID sample; during the study period, receptive sharing of needles/syringes decreased 5% and the percentage of injectors that utilized NSP increased from 44% in 2005 to 76% in 2009.
Chiang Rai Thailand
The NSP in Chiang Rai Province Thailand originally began in 1992 in three of the nine Akha hill tribe villages in northern Thailand. Five thousand needle and syringe kits were provided by the government for vaccination and were subsequently distributed among 46 PWID in three villages from 1992 to May 1994. During the period of evaluation, needles were not allowed to be distributed for the purposes of needle exchange; however, there was an agreement made that allowed for this small village to receive needles and syringes in response to the elevated number of new infections among PWID in the hill tribe villages. In 1995 a grant from the Australian government allowed implementation of NSP in all nine villages. The NSP served approximately 132 PWID and allowed up to 12 needles/syringes to be acquired per month for each PWID.
Gray 1998  analyzed the impact of the NSP on HIV prevalence among PWID in Akha hill tribe villages in Chiang Rai Province Thailand from 1993–1996. All PWID in this location were included in the study; 46 were part of the 1993–1994 sample while 132 were part of the follow-up 1995–96 sample. The samples were typically male (85%) and injected primarily heroin.
Over the course of the study, the HIV prevalence among PWID decreased from 33% (1993) to 18% (1995–1996).
National surveillance report interventions
Islamic Republic of Iran
The first NSP in the Islamic Republic of Iran was established in 2003 . Programs slowly increased after 2003 as laws were reformed so PWID were not arrested if they were covered by prevention and care services including drop-in centers which offer NSP . In 2007 there were 4,665,512 needles/syringes distributed, with an average of 41 needles/syringes distributed per PWID per year; during the same time period NSP sites increased from 170 in 2008 to 637 by 2010 .
The Center for Disease Management at the Iranian Ministry of Health reported national surveillance data on annual newly reported cases of HIV among PWID undergoing testing at surveillance sites from 1986 to 2007. The number of newly reported cases among PWID continued to increase through the late 1990s and early 2000s with 2332 cases in 2003, and a peak of 3145 new cases in 2004. However, with the implementation and scale-up of NSP services, the number of newly reported cases among PWID began to drop to 2,293 cases in 2005, 1,658 cases in 2006, and 426 cases in 2007.
In early 1997 NSP was first introduced in Lithuania via mobile and illegal underground needle/syringe distribution. In 2006 NSP was legalized in Lithuania, and by 2009, there were 12 NSP sites in Lithuania serving approximately 3,200 PWID. The number of needles/syringes distributed increased starting in 2006 and by 2008 313,000 needle/syringes were distributed to PWID, decreasing to 188,000 in 2009. No explanation was given by the report as to the reason for the decrease in distribution in the last year of data analysis.
Astrauskiene 2010  reported national newly reported HIV cases among PWID, collected by the Ministry of Health of Lithuania. The number of newly reported HIV infections among PWID decreased from 85 cases in 2005 to 62 cases in 2006, 59 cases in 2007, and 42 cases in 2008. However, in 2009, there was an increase with 117 newly reported cases among PWID. The increase in cases in 2009 coincided with a significant decrease (40%) in needle/syringe distribution.
In 2005, Taiwan instituted NSP on a trial basis throughout the country. By 2006 NSP in Taiwan had ramped up to full-scale operation with a 900% increase in coverage of PWID during the first two years of operation. A total of 450,000 needles/syringes were distributed in the first year, expanding to nearly four million in 2007, serving approximately 60,000 PWID.
A 2010 Taiwan Center for Disease Control (CDC)  document reported national surveillance data of newly reported HIV cases among PWID. Surveillance data captured in this analysis include all HIV incident PWID cases registered with CDC Taiwan between 2006 and 2010. The number of newly reported HIV cases among PWID decreased from 1,693 cases in 2006 to 114 cases in 2010.
NSP programs in Vietnam were started as pilot projects in 1993, with major scale-up occurring in 2005 as a result of the HIV Prevention Project implemented by the World Bank. In 2006, HIV/AIDS law mandated harm reduction activities that included NSP, leading to expansion from 21 provinces in 2005 to over 60 provinces by 2009. During this same time period, the number of needles/syringes distributed increased from two million in 2006 to over 11 million in 2007, serving approximately 215,000 PWID.
Vietnam surveillance data among at risk groups collected newly reported HIV cases from 2005 through 2009, coinciding with the period of expanding NSP in the country . During this time period, the number of HIV cases among PWID decreased from 11,358 in 2005 to 7,947 cases in 2009.