Water, sanitation, and hygiene access among people who inject drugs in Tijuana and San Diego in 2020–2021: a cross-sectional study

This was a cross-sectional study based on data from interviewer-administered surveys [conducted computer-assisted programmable interviews – CAPI] in the participants’ language of preference [English or Spanish] to PWID in the Tijuana–San Diego metropolitan area as part of the La Frontera binational cohort study. Data were collected by trained interviewers in 2020–2021 using street outreach and mobile vans in Tijuana and San Diego, as previously described in Strathdee et al.,2021 [26]. Eligibility criteria included individuals aged ≥18 or older, who reported injecting drugs within the last month in San Diego, Tijuana, or in both cities. Only participants who completed the WASH component questionnaire collected 1 week after the baseline survey formed part of this study. Participants’ informed consent was obtained, and monetary reimbursements were provided. This study received ethics approval from the institutional review boards at UCSD and UCLA in the United States (IRB# 800668), and Xochicalco University in Mexico (IRB # 191390).

Social and geographic independent variables of interest included city of residence and housing status. Housing status was classified as permanent housing, sheltered or unsheltered homelessness [27, 28], based on people’s reporting of the places where they have lived or slept, and the main place where they slept in the past 6 months. Permanent housing included people who reported sleeping in their parent’s, their own, their spouse’s/sex partner’s, family or friend’s house or apartment. Sheltered homelessness included people who reported sleeping in migrant worker’s camp, asylum seekers shelter, shelter/welfare residence, workplace, rented room (hotel, motel, or other rooming house), deportee shelter/camp, correctional institution (jail, prison, detention center), drug treatment center, medical care facility (i.e., hospital, hospice, or nursing home), or rented garage. Unsheltered homelessness included people who reported sleeping in their car, bus, truck, or other vehicle, abandoned building, on the streets, beach, parks, canal, woods, and shooting gallery.

Other covariates described were age, weight, income, gender identity (including man, woman, and transgender), and engagement in sex work. Sex work was defined as a source of income described as ‘prostitution or sex work’ in the past 6 month.

WASH access assessment included questions related to drinking water, sanitation, and hygiene (including handwashing, cleaning of wounds/abscesses, and showering/bathing) access in the past 6 months. First, drinking water access was assessed by identifying the main water source used, availability of drinking water sources, time spent collecting water, and the number of glasses of water and other beverages consumed in the last week. Second, sanitation access was determined by identifying the main sanitation facility used (i.e., open defecation, type of toilet/latrine), availability of the facility, whether the facility was shared with other households, the privacy (i.e., toilet with door that locks), the functionality, and experience of violence/harassment using a toilet. Hygiene access was measured with questions on availability of soap and water, water sources used for handwashing, availability of alcohol-based rub, and handwashing practices (after a bowel movement, before eating, and before preparing food). Additionally, we described water sources used for cleaning wounds and abscesses, for preparing drugs for injection, and showering/bathing practices in the last week (e.g., frequency and water source type). We included changes in water and soap availability since the COVID-19 pandemic (March 2020) and a measurement of cost (in USD) spent to access WASH services in the last month.

Basic WASH access variables were operationalized according to the WHO/UNICEF JMP definitions [6]. Basic drinking water access included the use of an improved water source in ≤30 minutes away roundtrip. Water sources definitions are described in Fig. 1. Basic sanitation access was defined as the use of improved facilities not shared with other households. Improved sanitation facilities were those which separate excreta from human contact, including flush/pour flush toilets connected to piped sewer systems, septic tanks, or pit latrines; pit latrines with slabs, and composting toilets. Basic hygiene access was defined based on the availability of a handwashing facility with soap and water. JMP definitions for water sources were used also for handwashing, bathing, cleaning wounds and abscesses, and preparing drugs for injections [6]. Bathing was defined as having > 4 bathes per week with improved water sources. Additionally, a set of complementary basic WASH variables were included to measure continuity of basic WASH services (i.e., always/24 hours service availability), and for basic hand hygiene we also measured the use of improved water sources and soap availability.

Key WASH outcomes were selected to describe social and geographic inequalities in WASH access and to assess WASH insecurity factors associated with them. These included 1) basic drinking water, 2) basic sanitation, 3) basic hygiene, 4) bathing, and improved water source for 5) cleaning wounds/abscesses and 6) preparing drugs for injection. WASH insecurity outcomes were measured as the lack of these six key WASH access indicators.

Of 612 participants in La Frontera, 586 (95.8%) completed the WASH survey and were included in analysis (Supplementary Table 1). All 26 participants who did not complete the WASH survey were from San Diego, and 25 of them were people experiencing homelessness. City of residence (p < 0.001) and housing status (p < 0.001) were significantly different between those who did and did not complete the WASH survey; there were no significant differences in survey completion based on gender (p = 0.555) and sex work status (p = 0.225).

Descriptive statistics were generated by estimating means and standard deviations (SD) along with medians and interquartile ranges (IQR) – according to their distribution – for the interval variables and frequencies and percentages for the categorical or binary variables. Number of glasses of water per day were described as a frequency and were used to estimate the daily water intake in milliliters per kilogram (ml/kg) of participant’s weight. Forty ml/kg were described as the recommended daily water intake [29]. Open defecation was described as both a binary variable and as a frequency (e.g., times a person openly defecated in the last week).

Inequality analysis included two components, the description of key WASH access outcomes by housing status and city of residence, and the assessment of those stratifiers as factors associated with key WASH insecurity outcomes. Inequalities accessing WASH were described as percentages of PWID who reported access to key WASH outcomes by city of residence and housing status. Inequalities assessment analyzed those social and geographic stratifiers as factors associated with outcomes of WASH insecurity (i.e., lack of WASH access). Measures of association between independent variables (stratifiers) and WASH insecurity outcomes were assessed using bivariate and multivariable analyses. We estimated crude (RR) and adjusted risk ratios (aRR) using a generalized linear model approach, leveraging a modified-Poisson distribution, and a log link, with a sandwich estimator to estimate robust standard errors [30]. Multivariable regression models included city of residence and housing status as independent variables and were adjusted for gender, and sex work as covariates. Two-way interactions were tested between independent variables to identify reference categories in each stratifier. Ninety-five percent confidence intervals (95% CI) were computed to identify significant results (alpha = 0.05). Data analyses were performed using R version 4.0.2.

Eighty-nine percent of the participants had access to basic drinking water, and 56% had basic drinking water access ‘always available’ (24-hours a day) (Table 2). Participants spent a median of 4 minutes (range from 0 min to 12 hours) traveling to their main drinking water source and returning from it. More than half (51%) of the participants reported drinking a beverage other than water when they felt thirsty, such as soda (35%) and sweetened beverages (8%). It was common that participants felt thirsty without having access to drinking water every day/multiple times per day (20%) or multiple times per week (25%). Participants drank a median of four glasses (IQR: 2–6) of water per day, and 96.9% had a daily water intake under the medical recommendation, with a median 16.4 ml/kg (IQR: 7.7–23.3) water intake per day. Drinking water availability did not change after the onset of the COVID-19 pandemic (March 2020) for 77% of the participants; however, for 18% of individuals, access became more limited. Participants spent a median of $5 USD (IQR: $2.5, $10), ranging from $0.3 to $200 USD, on drinking water per month. For the previous 6 months, the main drinking water source was packaged water (48%) and tap water (44%) (Table 3). Six percent of the participants reported no access to water sources, and 5% used unimproved water sources or surface water for drinking at least once in the last 6 months.

Only 28% of participants (32% in San Diego and 23% in Tijuana) reported basic sanitation access, and even fewer (25%) had basic sanitation access ‘always available’ (24-hours a day) (Table 2). Open defecation was reported by 38% of participants, with a median of five times (IQR: 2, 7) per week. Participants spent a median of 2 minutes (range from 0 min to 90 min) traveling to a sanitation facility and 20% spent more than 10 min. Toilet functionality and privacy was reported by 62 and 50% of the participants respectively. Experiencing violence or harassment using the toilet was reported by 13% of the participants (16% of women and 12% of men). Participants spent a median of $4 USD (IQR: $1.5, $50) on sanitation services per month.

Thirty-eight percent of the participants reported basic hand hygiene access, and 29% reported basic hand hygiene access with soap and improved water sources ‘always available’ (24-hours a day) (Table 2). For 68% of participants, soap availability did not change after the COVID-19 pandemic started, but it got worse for 24% of the sample population. More than half of the participants (61%) always or usually had alcohol-based rub available for hand hygiene practices. The main water source for handwashing was tap water (87%). Eight percent used surface water or unimproved water sources for handwashing at least once in the last 6 months. Always handwashing after bowel movements was reported by 41% of participants, before eating by 33%, and before preparing food by 36% of participants.

In terms of body hygiene practices, 46% of the participants reported bathing access (i.e., > 4 baths/week with improved water sources). Participants bathed a median of four times (IQR: 2, 7) per week and 94% used improved water sources (94%). Access to improved water sources for cleaning wounds or abscesses, and for preparing drugs for injection was greater than 90% (92 and 94% respectively). Tap water was the most common water source for bathing, cleaning wounds and abscesses, and for preparing drugs for injections (90, 79, and 81% respectively). Five percent reported no water sources for cleaning wounds and abscesses, and 4% used surface water or unimproved water sources for preparing drugs for injection. Participants spent a median of $10 USD (IQR: $5, $20), ranging from $0.50 to $250 USD on hygiene and $22.5 (IQR: $10.5, $50) USD, ranging from $2.3 to $405 USD on all WASH services per month.

San Diego (SD) residents had higher access to WASH services in comparison to Tijuana (TJ) residents (Fig. 2), especially in accessing basic hygiene (SD 55.8% vs TJ 30.0%) and sanitation (SD 43.1% vs TJ 20.0%). We observed a social gradient in WASH access by housing status across all key WASH access variables, except for access to basic drinking water. Participants living in permanent housing had higher access to key WASH services in comparison to people experiencing sheltered or unsheltered homelessness. These differences were notable between individuals living in permanent housing (PH) and those who experienced unsheltered homelessness (UH) in shower/bathing access (PH 79.2%, vs UH 30.9%), basic sanitation (PH 49.4% vs UH 12.7%), and basic hygiene (PH 57.1% vs UH 32.7%).

About PWID-specific water need, public and private tap water were the most common water sources for cleaning wounds/abscesses in both cities (SD 81.2%, TJ 88.4%) (Fig. 3). About the same proportion of participants from both cities reported no water sources for cleaning wounds/abscesses (SD 3.0%, TJ 3.3%). Surface and unimproved water sources for this use were reported only by participants residing in Tijuana (5%). San Diego residents used more packaged water (15.8%) to clean wounds/abscesses than those living in Tijuana (3.3%).

Participants living in permanent housing used almost exclusively private tap water to clean wounds/abscesses (93.5%). Among individuals experiencing sheltered (SH) and unsheltered homelessness, private (SH 70.0%, UH 55.6%) and public taps (SH 10.0%, UH 14.8%) were the most common water sources for this use. Individuals experiencing unsheltered homelessness used more packaged water (20.4%) to clean wounds/abscesses than who were experiencing sheltered homelessness (13.3%) and those living in permanent housing (3.9%). More than 3% of individuals experiencing homelessness used surface or unimproved water sources for cleaning wounds/abscesses (SH 3.3%, UH 3.7%). Also, more than 3 % of sheltered (3.3%) and unsheltered (5.6%) individuals experiencing homelessness reported no water sources for these uses.

In multivariable analysis, participants who resided in Tijuana had 1.68 (aRR 95%CI: 1.02, 2.76) times more basic drinking water insecurity, 1.45 (aRR 95%CI: 1.28, 1.64) times more basic hygiene insecurity, and 1.21 (aRR 95%CI: 1.06, 1.39) times more bathing insecurity in comparison to those who resided in San Diego (Table 4). No significant differences were found in insecurity accessing sanitation and improved water sources for cleaning wounds or for preparing drugs by city of residence.

Participants experiencing sheltered homelessness had 2.40 (aRR: 95%CI: 1.18, 4.90) times more insecurity for basic drinking water insecurity than those living in permanent housing. Participants experiencing unsheltered homelessness had 3.12 (aRR 95%CI: 1.55, 6.29) times more insecurity accessing improved water sources for cleaning wounds and abscesses and 2.58 (aRR: 95%CI: 1.36, 4.89) for preparing drugs for injection, 2.03 (aRR 95%CI: 1.07, 3.86) times more basic drinking water insecurity, 1.84 (aRR 95%CI: 1.52, 2.22) times more bathing insecurity, and 1.68 (aRR 95%CI: 1.48, 1.92) times more basic sanitation insecurity. No significant differences were found in hand hygiene insecurity by housing status.

In our study, PWID access to basic drinking water was lower than the national averages in both countries (Mexico and the US report > 99% access). This is suboptimal as the international target is universal basic access [6]. Likewise, low quantity of water intake per day [29], feeling thirsty without available drinking water sources, and drinking soda and sweetened beverages instead of water increase PWID risk of metabolic diseases. Drinking water insecurity (i.e., lack of access) and suboptimal water intake can lead to acute or chronic dehydration (i.e., 1–2% loss of body water), impaired cognitive function (e.g., short term attention and memory loss), and urolithiasis and other kidney diseases in the long term [31]. Unsheltered homelessness status and residence in Tijuana were associated with basic drinking water insecurity, which is likely related to infrastructure inaccessibility (e.g., lack of public drinking fountains). A recent mixed methods study among PWID and who were experiencing unsheltered homelessness residing in Tijuana found that 40% reported dehydration in the week prior to the survey, attributing it to their limited access to drinking water sources [16]. Similarly, a study among people experiencing homelessness – many who use substances – in San Diego found that increased distance from sources of safe drinking water may heighten risk of dehydration [23]. Our study complements previous work, highlighting the severe water insecurity and dehydration risk among this population of PWID.

According to international monitoring data, basic sanitation access in urban areas has been described as 100% for the US and 94% for Mexico, and both countries report that fewer than 1% of the population practice open defecation [6]. Despite residing in metropolitan areas that are each high-income relative to the country in which they are located, participants’ access to basic sanitation was very low (< 30%) and open defecation was a common practice (~ 40%). The Project for Sanitation Justice in San Diego reported in 2022 that 49% of census tracts don’t have public sanitation facilities, which are often closed or locked and almost none are open 24 hours [32]. Although no similar information is available for Tijuana, access is expected to be similar or more limited. We identified that participants who did have access to toilets reported that the toilets were often non-functional, without privacy (i.e., toilet with door that locks), and frequently experienced violence using them. Further, basic sanitation insecurity was more challenging among individuals experiencing unsheltered homelessness in both cities, increasing their risk of infectious diseases. Particularly, open defecation poses risks not only to PWID, but is a major public health risk in both cities due to fecal contamination in soil and surface water. For instance, unhoused individuals with insufficient access to sanitation services may be unable to avoid contributing to human fecal contamination of the environment where they live and the Tijuana River and San Diego River Watersheds pollution [22, 33]. River water contamination can further affect environmental health during storm events and when untreated water reaches ocean and estuary environments [34, 35]. Increases in 24-hour public restroom access, especially in neighborhoods with people experiencing homelessness could therefore have a large community health impact by reducing open defecation [36]. In the Tijuana River, fecal contamination is also contributed to by untreated sewage and lack of wastewater treatment capacity. Fecal contamination was high immediately upstream of the location where the community who lives inside the river canal resides. In the case of the Tijuana River, the lack of capacity in the wastewater treatment system contributes to untreated wastewater discharges, especially during rain events, which has been described as the most important source of human contamination [22, 37, 38]. Similarly, in the San Diego River, researchers found chemical markers indicating that high levels of fecal contamination during storm events are primarily from leaky sewers and not open defecation [39].

Official data in Mexico reports that 91% of individuals have access to basic hand hygiene, and although there are no official data for the US, nearly universal access is expected [6]. Yet, PWID participating in this study were far from achieving this national access to basic hand hygiene. Basic hygiene access insecurity could be limiting hygiene practices at key moments (i.e., after a bowel movement, before eating, and before preparing food). Additionally, although the use of unimproved and surface water sources was not common, they did represent the main water source for hand hygiene for some participants. A study in Zimbabwe reported that contaminated water for handwashing was associated with hand contamination with E. coli [40].

In this study we described bathing access based on the number of baths per week and the water sources used. However, the location or facilities this population used for bathing remain unclear from this study. Different studies among PWID and who experience unsheltered homelessness residing in Tijuana and San Diego have reported business establishments (~ 30%) such as private stores or gas stations [16, 23] were the most common places to bathe, which have sinks and do not have showers or baths. Considering that our study population could share this practice, we are unsure that bathing access was from a proper facility – e.g., different than a sink in a public restroom, and this issue should be further studied.

Although most PWID participants reported using improved water sources for drug injection preparation and for cleaning wounds and abscesses, the use of surface water and other unimproved water sources remained present (5–8%). This is an important finding because the lack of safely managed water sources for these behavior practices may exacerbate health risks, including the re-occurrence of abscesses among the population. In different contexts, abscesses – and other skin and soft tissue infections (SSTI) – are common (~ 50%) among PWID [13, 16, 23], which highlights the need to conduct more research on water sources associated with SSTI among PWID. Particularly among PWID and who experienced unsheltered homelessness residing in the Tijuana, SSTI were significantly associated with use or contact with contaminated surface water [16]. A study of people experiencing unsheltered homelessness who used drugs in rural areas of Central Appalachian Kentucky supported the notion that harm reduction services providing clean water for preparing drugs, cleaning skin, and handwashing facilities prior to injection would help prevent bacterial infections and abscesses among PWID and who experienced unsheltered homelessness [12]. Harm reduction programs can provide PWID with access to WASH services and education about hygiene practices around drug preparation [17], decreasing PWID’s risk of SSTI and water-related infectious diseases.

Despite Tijuana being one of the most developed cities in Mexico [41], it does not have the same WASH infrastructure and availability of free public services for residents that San Diego offers. This includes access to public drinking fountains, sanitation and hygiene facilities at public spaces, and mobile hygiene services.

In San Diego, after the Hepatitis A outbreak in 2017 and during the COVID-19 outbreak in 2020, the local government set up temporary handwashing stations and portable toilets for individuals experiencing unsheltered homelessness [23, 42]. However, study staff in San Diego observed that many toilets in public spaces (e.g., parks) were locked or unavailable outside of daylight hours. In contrast, in Tijuana during the COVID-19 outbreak in 2020, local businesses and nonprofit organizations provided handwashing stations; yet few public facilities were established, which may explain geographic differences on basic hand hygiene insecurity.

Further, differences between both cities in infrastructure and availability of programs for individuals experiencing homelessness may explain bathing inequalities. About half of the population experiencing homelessness residing in the San Diego River used restroom sinks at businesses, portable toilets or public restrooms to bathe, and the remainder used service providers, shelters, homelessness service facilities, or saved up money to stay in a hotel room once per month to shower [23]. On the other hand, among the PWID and who experienced unsheltered homelessness residing in the Tijuana River, few bathing facilities were available, and there was a higher reported levels of contact with unimproved water sources, such as surface water – including irrigation water (clean tap water in Mexico), or purchased packaged water [16]. Furthermore, relative to their income (participants in Tijuana had lower income than in San Diego), many PWID in this region spent a large quantity of money on WASH services every month, especially on hygiene supplies. Although there are differences between Tijuana and San Diego that require tailoring an approach to each context, we found few differences in term of WASH access suggesting that policy responses should be similar.

WASH access is integrally connected to shelter/housing. PWID and who experienced unsheltered homelessness had significantly higher experiences of WASH insecurity than those who lived in permanent housing. Lack of access to private tap water sources – linked with housing access – among participants experiencing unsheltered homelessness may be related to the higher use of packaged water. In many cases water bottles and sterile water were provided by harm reduction services and were the only water source for preparing drugs for injection and cleaning wounds for the sampled population. Intersectional vulnerabilities, like experiencing unsheltered homelessness, exacerbate WASH insecurity in distinct fashions for both men and women that prevent them from participating in daily activities [8, 43].

Sanitation and bathing facilities are services that require infrastructure that is less accessible outside a formal housing setting. In two studies of communities experiencing homelessness in Los Angeles, CA and Belo Horizonte, Brazil, during the night, public toilets and bathing facilities are usually closed and open defecation becomes a coping survival strategy [8, 44]. Many individuals in the Brazilian study reported urinating and defecating in the open near where they sleep [44], avoiding violence, but increasing environmental risk for themselves and others around them. Individuals experiencing homelessness are at higher risk of WASH insecurity, including people experiencing sheltered homelessness, which can be living in hotel rooms, garages, or cuarterias (building divided into small informal living spaces) where bathrooms are not always available. Furthermore, WASH insecurity perpetuates a cycle of poverty among populations experiencing unsheltered homelessness [8], and criminalization of homelessness in both cities also contribute to experiences of WASH insecurity [16, 23]. Particularly hygiene insecurity (bathing and handwashing) can exacerbate social exclusion, discrimination, police victimization, and loss of dignity and self-esteem.

Our findings lead us to propose two possible avenues to achieve WASH security among populations of PWID and who experienced unsheltered homelessness in our study population: 1) a housing first (HF) model where WASH is available within the household setting, and 2) mobile/public WASH facilities to meet all other WASH needs. HF principles include immediate access to affordable housing, and independent individualized and flexible support [45]. Mobile WASH facilities can help by bringing WASH services to PWID who are experiencing homelessness. In San Diego there are public and nonprofit mobile hygiene services available that provide shower services and hygiene supplies to populations experiencing homelessness, while in Tijuana these services are limited to a small number of nonprofit organizations. Furthermore, to access sanitation services, in Tijuana (as in many other locations in Mexico) there is a cost incurred for use of public bathrooms (~$0.25–$0.5 USD) and in San Diego, the San Diego City Council is considering charging for public restrooms as well, limiting the accessibility for anyone experiencing homelessness or who have scarce financial resources [46].

Current JMP definitions of basic WASH access have two implications for the population of PWID studied. First, the unit of measure is the ‘household’, which excludes individuals experiencing homelessness. Second, this definition does not include data on temporal availability of these facilities, therefore we do not know whether sanitation facilities are ‘always available.’ We report access using the JMP definitions and additionally described an extended version of them, including 24-hour availability of WASH services. Particularly, for basic hand hygiene access, we also incorporated availability of soap and the use of improved water sources, no guarantee among marginalized populations. The proposed extended definitions highlight the accessibility gaps hidden in basic WASH access definitions, which is particularly important among communities experiencing homelessness that use public facilities to solve daily needs. There are other scales to measure water insecurity [47], yet we defined WASH insecurity indicators in a simple way based on JMP definition for basic WASH access, highlighting the need to include sanitation and hygiene insecurity, which are usually left out.

In addition to the JMP water quality standards for drinking water, we extended the JMP water sources definitions for handwashing, bathing, and PWID-specific water needs. Moreover, WASH literature has typically focused on hand hygiene but not on bathing. More data should be collected on bathing, including frequency, water sources, facilities, soap/shampoo availability, and health outcomes.

Due to a restricted sample size, we could not analyze intersectional interactions of WASH insecurity between more than one social and geographic stratifier at time. There was also potential selection bias based on the significance testing for those included vs excluded, though excluded participants were < 0.05% of the parent study sample, and also because participants residing in San Diego who engaged in cross-border drug use were oversampled since the goal of the parent study was to study risks of infectious diseases and overdose among PWID in relation to cross-border mobility. This study was based on a self-reported questionnaire about WASH access, which is prone to recall and response biases if participants could not accurately remember details or if they tended to respond with what they believed was a socially desirable answer. Social desirability bias may have led some behaviors, like open defecation, to be under-estimated. The WASH questionnaire also included questions that can feel private, such as handwashing and open defecation practices. We did not include questions about the facilities used for bathing. Additionally, access to laundry services, oral hygiene, and water sources for rinsing syringes and handwashing before injection practice are important variables that could be collected in future research but were not explored in this study. Future efforts to implement mobile WASH services or other interventions should involve PWID to ensure that the services meet their needs.