Injection drug users (IDUs) are at high risk of infection with HIV, hepatitis C (HCV) and other blood-borne pathogens transmitted by sharing syringes and other injection equipment. Globally, almost 20% of the world's 15.9 million IDUs are infected with HIV [1] and in some studies HCV prevalence among IDUs is >90% [2].

Transmission of these pathogens can be prevented by eliminating syringe sharing among IDUs. This requires that sterile syringes be available at appropriate times and in sufficient quantities to supply a sterile syringe for each injection. In most settings, IDUs' avenues for obtaining sterile syringes are limited to syringe exchange programs (SEPs) and pharmacies. SEPs have proven effective in reducing syringe sharing [3], but the number of these programs - and the overall number of syringes they distribute - is not sufficient to provide IDUs with a sterile syringe for each injection. In Latin America, for example, only 5 of 20 countries are known to have implemented SEPs, which serve only approximately 2% of the region's IDUs [4].

Pharmacies can provide a more comprehensive and convenient source of syringes for IDUs, as they generally exceed SEPs in number of locations and hours of operation. In some areas, however, pharmacy-based syringe access is hampered by laws requiring a prescription for purchase. Allowing purchase without a prescription has been shown to increase the number of syringes sold and reduce sharing among IDUs. In the United States, for example, where syringe access laws vary by state, Connecticut and Minnesota saw substantial increases in the number of IDUs who reported pharmacy purchase of syringes and decreases in syringe sharing following repeal of syringe prescription laws [5, 6]. Similarly, in New York City, an Expanded Syringe Access Demonstration Program (ESAP) was associated with a significant increase in the proportion of IDUs who obtained syringes from pharmacies, and these IDUs were less likely than others to report syringe sharing [7].

Unfortunately, even where the "laws on the books" allow syringe purchase without a prescription, IDUs still encounter barriers to purchase. U.S.-based studies have obtained objective measures of IDUs' ability to purchase syringes using the "mystery shopper" method, in which study personnel enter pharmacies and attempt to purchase a sterile syringe according to a predetermined script. These studies have documented refusal rates of 31-59% in areas where syringe purchase without a prescription is legal [8–12].

In Mexico, where syringe purchase without a prescription is also legal, IDUs report being refused or overcharged for syringes at retail pharmacies and link these refusals directly with syringe sharing [13, 14]. In Tijuana, a northwestern Mexico border city adjacent to San Diego, California, 59% of IDUs report receptive syringe sharing in the past 6 months and HCV prevalence is 96% [15, 16]. HIV prevalence among Tijuana's male IDUs, female IDUs and female IDUs who engage in sex work is 4%, 10%, and 12%, respectively [17, 18] and as many as one in 125 persons aged 15-49 in the city are estimated to be HIV-positive [19]. We undertook this study to obtain an objective measure of barriers to pharmacy-based syringe purchase among IDUs in Tijuana and assess the need for pharmacy-based HIV prevention interventions.

Overall we identified 189 retail pharmacies in the three targeted colonias and completed purchase attempts at 164 of them. The other 25 pharmacies were excluded because either the pharmacy had closed since the validated list was completed (n = 17) or it was determined not to be an eligible retail pharmacy at the time of the purchase attempt (n = 8; e.g., sold botanicals only, pediatric medications only). Of the 164 purchase attempts, one was excluded because the mystery shopper varied from the predetermined script and one was excluded due to missing data, leaving a total of 162 eligible purchase attempts included in our analysis.

Mystery shopper syringe purchases

Only 46 (28.4%) of the 162 eligible mystery shopper purchase attempts were successful. The median price per syringe purchased was 7 pesos (IQR: 5-10). Table 1 compares the characteristics of successful and unsuccessful purchase attempts. There were no statistically significant differences between the pharmacies where syringes were successfully purchased and those where the purchase attempt was unsuccessful; however, female mystery shoppers were more likely to have a successful purchase outcome than male shoppers, with marginal significance (p = 0.058).

	Successful (%) N = 46	Unsuccessful (%) N = 116	P-value
Mystery shopper sex
Male	19 (41.3)	67 (58.7)	.058
Female	27 (58.7)	49 (42.2)
Median number of pharmacy staff who waited on mystery shopper (IQR)	1 (1-2)	1 (1-2)	.406
Sex of staff person 1
Male	23 (50.0)	54 (47.0)	.727
Female	23 (50.0)	61 (53.0)
Approximate age of staff person 1	31 (25-45)	35 (27-44)	.260
Sex of staff person 2
Male	7 (46.7)	11 (40.7)	.710
Female	8 (53.3)	16 (59.3)
Approximate age of staff person 2	40 (30-55)	31 (25-44)	.109
Median number of customers within 10 feet of shopper	1(0-2)	0 (0-1)	.154
Number of customers in store
< 5	43 (97.7)	113 (98.3)	.825
5-10	1 (2.3)	2 (1.7)
> 10	0	0

Figure 1 presents the reasons for unsuccessful purchase attempts. One-third (35.3%) of the pharmacies told the mystery shopper they did not sell syringes and another one-third (31.0%) said they had no syringes in stock. The remaining purchase attempts failed either because the pharmacy requested a prescription (20.7%), charged more than the maximum established price of 10 pesos (3.5%; price range 12-15 pesos), referred the mystery shopper to another pharmacy (2.5%), or for some other reason (7.0%; e.g., told that syringes were only sold in packages of 10, only sold 10 cc syringes).

In some cases, the mystery shoppers provided additional comments regarding their experiences during the syringe purchase attempt. In five cases, the shopper noted that although their purchase attempt was unsuccessful the person who waited on them treated them kindly, as if this experience were out of the ordinary. In two cases in which the purchase attempt was successful, the shopper noted that the seller told them to "take care" or "be careful." There were two cases in which the shopper specifically reported that the person who waited on them was angry or upset, and two others where they reported being actively observed (e.g., "[They] looked at me from head to toe, studied me [before they said] 'I can't sell it to you without a prescription.'"). In three other failed attempts the shopper reported being completely ignored by the pharmacy staff (e.g., "[They] continued watching television, didn't even look at me.").

In six cases, the mystery shopper reported lack of knowledge about the pharmacy policy regarding syringe sales or that pharmacy staff disagreed about whether to sell them the syringe. These interactions were described as follows:

• "He/she was going to sell it to me but a voice from behind the wall told him/her that I needed a prescription."

• "The guy was going to help me but they girl said they had run out...she was shaking her head at the guy as if to say 'No, don't sell.'"

• "One (the one that was selling) said yes, but then asked the boss who said they didn't have any."

• "The lady was going to sell to me, the man wasn't. 'We don't have syringes because we don't have a fridge for the insulin.'"

• "He/she had to call a supervisor by radio to ask if he/she could sell it to me or not. [The supervisor said]...not without a prescription."

• "One said that with a prescription and the other one said that there was no problem. In the end it was sold."

To our knowledge, this is the first published mystery shopper study of syringe access conducted in a developing country and the first anywhere to compare mystery shopper outcomes with a concurrent telephone survey. We documented a very low level of success in purchasing sterile syringes at retail pharmacies in Tijuana, Mexico, and a high level of discordance between stated pharmacy syringe sales policies and those experienced by the mystery shoppers.

The low mystery shopper success rate in this study corroborates IDU reports of substantial barriers to pharmacy-based syringe purchase in Tijuana. In qualitative studies, IDUs have linked these barriers directly to risky injection practices, including syringe sharing and scavenging through medical and household waste for used syringes [13, 14]. A quantitative study of IDUs in Proyecto El Cuete similarly demonstrated an independent association between experiencing barriers to pharmacy-based syringe purchase and receptive syringe sharing, syringe reuse, and a higher number of lifetime abscesses [14]. In light of the mounting evidence regarding restrictive syringe sales practices in Tijuana and their direct contribution to risky injection behaviors, structural interventions are needed to modify these sales practices.

This study provides preliminary insights into the reasons for restrictive syringe sales practices in Tijuana. First, we found high levels of discordance between stated pharmacy syringe sales practices and mystery shopper outcomes. The fact that practices experienced by the mystery shoppers were more restrictive than those stated in the telephone survey suggests pharmacies are less likely to sell syringes to suspected IDUs. Our mystery shoppers were study staffers who, although not current drug users, had a history of injection drug use and for the purposes of the study dressed in a manner consistent with IDUs in the area; it is thus reasonable to believe that they were suspected of injection drug use. Further, seven pharmacies in our telephone survey willingly stated that their decision to request a prescription for syringe purchase hinged on the appearance of the customer. IDUs in Tijuana have spoken at length with us about the perceived role of appearance in their attempts to purchase syringes, and we have demonstrated an independent association between homelessness - which influences the ability of IDUs to maintain the cleanliness of their person and clothing - and encountering barriers to syringe purchase [14]. These findings indicate that suspecting a person of injection drug use is a motivating factor for pharmacy staff in denying syringe purchase in Tijuana.

U.S. studies have identified a number of reasons why pharmacies deny syringes to suspected IDUs. These include business considerations including worries regarding store theft, the security of pharmacy staff and customers, and increased drug use and discarding of used syringes near the pharmacy [26–34]. Individual attitudes of pharmacy staff also play a role; these include negative attitudes toward drug use and drug users, concerns that distributing syringes increases drug use and the belief that selling syringes is not appropriate for pharmacists in their role as health care professionals [26, 28, 30, 31, 34]. Studies that incorporate interviews with pharmacy owners, pharmacists and clerks in Tijuana are needed to determine whether these factors influence syringe sales decisions and identify other factors amenable to intervention. These studies constitute the next phase of our research activities in Tijuana.

One of these other factors may be misunderstanding of the laws regarding pharmacy syringe sales in Mexico. Almost one-third of the pharmacies interviewed by telephone said they require a prescription for syringe purchase. It is possible that pharmacy management and staff do not have an accurate understanding of the laws governing syringe sales in Mexico. Alternatively, it is possible that, for reasons that remain unclear, they feel the need to obtain a prescription despite what the law allows. In qualitative interviews with IDUs, they posited that pharmacies fear retribution from police if they are caught selling syringes to IDUs [14]. None of the pharmacies in this study cited fear of police to the mystery shopper or telephone interviewer, but prior studies in Tijuana by our research team have found high rates of arrest for possession of sterile syringes among IDUs, even though possession of these syringes is allowed under Mexican law [35]. If fear of police is indeed a factor in pharmacy syringe sales practices then police behavior would need to be targeted as part of any pharmacy-based intervention.

Regardless of their reasons for requesting a prescription, it is clear that pharmacies' prescription policies are not applied consistently. Our study found discordance between pharmacy prescription policies reported in the telephone survey and policies encountered by the mystery shoppers at the same pharmacies. As expected, most of these discrepancies went in one direction, with pharmacies reporting not requiring a prescription but asking the mystery shopper for one anyway. However, we also identified pharmacies that sold syringes to the mystery shopper despite a stated policy of requiring a prescription. Further, our mystery shoppers reported encountering disagreements among pharmacy staff regarding whether or not to sell the shopper a syringe; these disagreements may explain some of the discrepancies we encountered in comparing our mystery shopper and telephone survey findings, as the outcome may have depended greatly upon which pharmacy staffer waited on the mystery shopper or answered the survey call. Interventions that seek to bring pharmacies into compliance with Mexican laws allowing over-the-counter syringe sales will need to ensure that these policies are understood and implemented consistently across the staff within each pharmacy.

Finally, the overall rate of successful purchases achieved in this study was lower than the success rates achieved by U.S.-based studies employing a similar methodology [8–12]. Although our findings cannot be extrapolated to other developing countries - or even to other regions in Mexico - they raise the possibility that access to sterile syringes through retail pharmacies in developing countries where syringes are legally available without a prescription may be more limited than in developed countries with the same policies. This possibility should be investigated by researchers in developing countries where IDUs constitute a substantial proportion of new HIV cases.

U.S.-based mystery shopper studies have documented significant differences in outcomes depending on whether a retail pharmacy is a chain or independent establishment. Because we were not able to determine the chain/independent status of several pharmacies while developing our validated pharmacy list, we did not include this variable in our study. We also identified only a very small number of pharmacies that attempted to overcharge the mystery shoppers for syringes, which was inconsistent with our prior qualitative findings that overcharging is common [13, 14]. This may be attributed to the fact that, for reasons of staffing and safety, we did not conduct any purchase attempts after 6:00 pm. Given that IDUs have reported higher likelihood of overcharging in the late evenings and early mornings, our findings may underestimate the frequency of overcharging. IDUs have also reported a higher likelihood of overcharging when they appear to be in opiate withdrawal, which our mystery shoppers clearly were not.

We documented a substantial difference between the "laws on the books" that govern syringe sales in Mexico and the actual sales practices of retail pharmacies in Tijuana. A clearer understanding of what motivates these sales practices is needed. Structural interventions that build upon this knowledge should be prioritized among public health efforts to expand syringe access and reduce transmission of HIV and other blood-borne pathogens among IDUs.