High prevalence of abscesses and self-treatment among injection drug users in Tijuana, Mexico

Summary

Background

Soft tissue infections are common among injection drug users (IDUs), but information on correlates and treatment in this highly marginalized population is lacking.

Methods

Six hundred twenty-three community-recruited IDUs in Tijuana, Mexico, completed a detailed interview on abscess history and treatment. Univariate and multiple logistic regressions were used to identify factors independently associated with having an abscess in the prior 6 months.

Results

Overall, 46% had ever had an abscess and 20% had had an abscess in the past 6 months. Only 12% had sought medical care for their most recent abscess; 60% treated the abscess themselves. The most common self-treatment method was to apply heated (24%) or unheated (23%) Aloe vera leaf. Other methods included draining the wound with a syringe (19%) or knife (11%). Factors independently associated with recent abscess were having income from sex work (adjusted odds ratio (aOR) 4.56, 95% confidence interval (CI) 2.08–10.00), smoking methamphetamine (aOR 1.65, 95% CI 1.05–2.62), seeking someone to help with injection (aOR 2.06, 95% CI 1.18–3.61), and reporting that police affected where they used drugs (aOR 2.14, 95% CI 1.15–3.96).

Conclusions

Abscesses are common among IDUs in this setting, but appropriate treatment is rare. Interventions to reduce barriers to medical care in this population are needed. Research on the effectiveness of Aloe vera application in this setting is also needed, as are interventions to provide IDU sex workers, methamphetamine smokers, and those who assist with injection with the information and equipment necessary to reduce abscess risk.

Introduction

Soft tissue infections are common among injection drug users (IDUs). A prospective study of IDUs in Amsterdam reported an incidence of one abscess per 3 years of injection.¹ In San Francisco, 32% of street-recruited IDUs were found to have an abscess, cellulitis, or both, based on physical examination at the time of interview.² Community-based studies of IDUs in Vancouver³ and Baltimore⁴ found 6-month abscess prevalence of 22% and 11%, respectively. Among IDUs recruited through needle exchange programs and supervised injection facilities that provide information and materials to promote safer injection and prevent abscesses, prevalence of abscesses has been found to be lower, at <10% over 12–18 months.^5,6

Abscesses can also lead to serious infections, including endocarditis, osteomyelitis, and septicemia. Despite these potentially serious consequences, IDUs often delay medical care for abscesses. In San Francisco,² 77% of IDUs with a history of abscess had been treated by a doctor, but 48% reported at least one abscess for which they had not sought treatment; 27% had lanced their own abscess and 16% had used antibiotics acquired on the street. Delaying or avoiding care can result in complicated clinical presentations requiring more complex treatment and higher morbidity and mortality. In Seattle, the median time between symptom onset and seeking care among IDUs with abscess was found to be 5 days, with 35% of patients ultimately requiring hospitalization.⁷ Serious soft tissue infections were reported to be the most common reason for emergency department visits and the second most common reason for hospital admission in a cohort of IDUs in Vancouver.⁸ A hospital records review in San Francisco found that among patients admitted with a primary diagnosis of soft tissue infection, 70% had injected drugs in the past 12 months.⁹ There appear to have been no empirical studies focused on barriers to abscess care among IDUs, although at least one study has cited lack of financial resources and concern about negative or punitive interactions with healthcare providers as contributing to delays in care.¹⁰

Similarly, there have been relatively few studies examining risk factors for abscesses among IDUs. Identified risk factors include poor injection hygiene practices like licking the needle before injecting² and inconsistently washing hands before injecting,⁵ while skin cleaning before injection is protective.^4,11 Subcutaneous or intramuscular injection (‘skin popping’),^2,11 drawing blood into the syringe before injecting (‘booting’),¹¹ injecting heroin and cocaine together (‘speedballing’),^1,2,11 and high injection frequency^1,12 have also been associated with abscesses among IDUs. Further, HIV infection^1,3 and female gender^1,3,6,11,12 have been associated with abscess in multiple studies, as have contextual factors like engaging in sex work^1,3 and unstable housing.^3,5

Other studies have suggested that the type of drug used may contribute to abscess risk. A prior study in Tijuana, Mexico, found that IDUs who injected white or colored methamphetamine were more likely to report abscesses than those who injected clear methamphetamine.¹³ Black tar heroin injection has also been associated with soft tissue infections. Black tar is sometimes ‘cut’ with contaminated adulterants, including soil, which can promote inoculation of bacteria and subsequent soft tissue infection.¹⁴ The thick consistency of black tar heroin commonly results in a high degree of venous sclerosis, making intravascular injection difficult and increasing the likelihood that IDUs may resort to subcutaneous injection.

Tijuana, Baja California, Mexico is situated on the Mexico–USA border adjacent to San Diego, California, on a major drug trafficking route. Increased drug trafficking in recent years has created a local consumption market in Tijuana, where black tar heroin and methamphetamine predominate. We examined the prevalence and treatment of abscesses among IDUs in Tijuana and identified correlates of recent abscess to identify opportunities for prevention and treatment interventions.

Methods

Recruitment

Between April 2006 and April 2007, 1056 IDUs were recruited in Tijuana into a prospective study of behavioral and contextual factors associated with HIV, syphilis, and tuberculosis (TB) infections, which has been previously described.¹⁵ Eligibility criteria included being ≥18 years of age; having injected illicit drugs within the past month, as confirmed by inspection of injection stigmata (‘track marks’); ability to speak Spanish or English; being able to provide informed consent; and having no plans to permanently move out of the city in the next 18 months. Methods were approved by the Institutional Review Board of the University of California, San Diego, and the Ethics Board of the Tijuana General Hospital.

Respondent-driven sampling (RDS) was used to recruit participants.¹⁶ Briefly, a diverse group of ‘seeds’ (heterogeneous by age, gender, and neighborhood) was selected and given uniquely coded coupons to refer their peers to the study. Waves of recruitment continued as subjects returning with coupons were given coupons to recruit members of their social networks. Recruitment and interviews were conducted by indigenous outreach workers through the use of a modified recreational vehicle and a storefront office.

Data collection

At baseline and semi-annually thereafter, participants completed an interviewer-administered survey eliciting information on sociodemographic, behavioral, and contextual characteristics. At the first follow-up visit, a cross-sectional survey on abscess history and treatment was added to the study instrument, which included questions on lifetime and recent history of abscess (i.e., “Have you ever had an abscess?”, “When was the last time you had an abscess?”), followed by detailed questions regarding the characteristics of the most recent abscess. Among these characteristics were the symptoms accompanying the abscess, e.g., ‘swelling’, ‘fever’, ‘pus or fluid under the skin’, and ‘oozing of clear, white, or yellow fluid or pus from the skin’. Participants who had an abscess at the time of the interview were instructed to provide details of the most recent abscess prior to the current abscess, unless the current abscess was their first, and were offered treatment by on-site medical personnel or referred to a local clinic or hospital for care. Questions on modes of self-treatment were formulated based on field observations and piloting of the cross-sectional survey instrument.

Serological testing for HIV antibody was conducted using the Determine^® rapid HIV antibody test (Abbott Pharmaceuticals, Boston, MA, USA) and confirmatory testing with an HIV-1 immunoassay and immunofluorescence assay. Syphilis serology was conducted using the rapid plasma reagin (RPR) test (MacroVue, Becton Dickenson, Cockeysville, MD, USA); RPR-positive samples were confirmed using the Treponema pallidum particle agglutination assay (TP-PA; Fujirebio, Wilmington, DE, USA).

Statistical analysis

Descriptive statistics were used to characterize lifetime abscess history and details of the most recent abscess and its treatment. To identify factors associated with abscess in the past 6 months, we conducted a cross-sectional univariate analysis using Chi-square tests for categorical variables and t-tests and Wilcoxon rank-sum tests for normally and non-normally distributed continuous variables, respectively. Variables that achieved significance of ≤0.10 were entered into a multiple logistical regression model in a manual, stepwise fashion to identify baseline variables independently associated with reporting abscess in the 6 months prior (p < 0.05). Given previous findings that abscesses are more frequent among female than male IDUs we also tested for interactions between gender and other exposure variables.

As described previously,¹⁵ we explored the potential effects of RDS on our estimates using the RDS Analysis Tool (version 5.6.0, October 2006; Cornell University) and WinBUGS (version 1.4.1, 2004; Imperial College and Medical Research Council, UK). Odds ratios and 95% confidence intervals produced by the RDS analysis were compared to our multivariate logistic regression models. Since no significant differences between the RDS-adjusted and unadjusted models were identified, unadjusted models are presented.

Results

Overall, 653 participants (62%) completed the abscess questionnaire at the first semi-annual follow-up visit. Of these, we excluded 20 who did not report injection in the past 6 months and 10 for whom abscess history questions were missing. Of the remaining 623 IDUs, 82% were male and the median age was 37 years (interquartile range (IQR) 32–43 years). The median time since first injection was 15 years (IQR 9–23 years). Most (88%) injected daily during the past 6 months; the most common drugs injected were heroin alone (81%) or in combination with methamphetamine (51%), followed by methamphetamine alone (14%), and heroin and cocaine together (6%). In addition, 22% smoked methamphetamine during the past 6 months.

Abscess characteristics

Two hundred eighty-five IDUs (46%) reported ever having an abscess (median 3, IQR 2–5) and 127 (20%) reported an abscess in the past 6 months. IDUs who reported a history of abscess were asked to provide detailed information on the characteristics of their most recent abscess. A majority reported that their most recent abscess was on the arm (70%), followed by the leg (13%), neck (6%), hand (6%), or foot (3%).

Almost half (47%) reported pus or fluid under the skin and 20% reported oozing of pus or fluid from the skin. Almost all reported swelling (99%) and pain (94%), and a majority reported redness (78%). Less common symptoms were heat at the abscess site (23%), fever (21%), and vomiting (5%).

Abscess treatment

A majority (60%) of IDUs treated their most recent abscess on their own. For most (97%) this was the only source of treatment reported. The most common self-treatment was applying heated (24%) or unheated (23%) Aloe vera leaf to the wound; interestingly, only one participant reported both of these application methods. Other treatments included applying a rag soaked in salt water (21%), using fingers to press out the pus (20%), draining the wound with a syringe (19%), taking medication (17%) (most commonly penicillin, which along with other first-line antibiotics is available over-the-counter in Mexico), draining the wound with a knife (11%), and applying hydrogen peroxide (10%). Among those who drained the wound themselves, 11% did nothing to sterilize the area before incision. Forms of cleansing included water (40%), alcohol (28%), hydrogen peroxide (23%), Aloe vera (15%), and iodine (2%). Six percent applied an antibiotic to the affected area. Brief descriptions of this method suggest the antibiotic pills were crushed and the powder applied directly to the incision site. Overall, a majority (54%) reported using only one self-treatment method, 30% used two methods, and 15% used three or more methods.

Of those who self-treated their most recent abscess, 89% reported that the abscess healed after self-treatment. Thirty-one percent said the abscess healed in less than 1 week, 50% in 1–3 weeks, and 19% said it took >3 weeks to heal. Ninety-eight percent of those who reported invasive self-treatment (i.e., draining the abscess with a knife or syringe) reported healing after self-treatment compared to 86% of those who did not use invasive treatment (p = 0.022). Those who used multiple methods of self-treatment were not significantly more or less likely to report healing than those who used a single method, and location of the abscess on the body was not significantly associated with healing after self-treatment.

Only 12% of IDUs went to a clinic or hospital for treatment of their most recent abscess. Of these, 56% waited >5 days after the abscess developed to seek care. Almost half (47%) of those who sought treatment were treated at the hospital, with the remainder citing a clinic or private physician. The majority of those who sought treatment (76%) had the wound drained; 56% were given penicillin or another antibiotic and 24% received medication for pain.

Factors associated with abscess in the past 6 months

Table 1 presents selected results from the univariate analysis of factors associated with reporting an abscess in the past 6 months. Participants who reported an abscess were more likely to be female, to report that their primary income came from sex work, and that they had traded sex for money or drugs. They were also more likely to report smoking methamphetamine, seeking someone to help them to inject (‘hit doctor’), and injecting with a family member or spouse. Negative interactions with police, including being arrested for having track marks and being asked for money from police, were also significantly associated with recent abscess, as was reporting that police affected where they used drugs. Injection practices like licking the needle or skin before injecting were rare (2% and 1%, respectively).

Table 1.

Factors associated with abscess in the past 6 months among 623 IDUs in Tijuana, Mexico

Variables	Abscess % (n = 127)	No abscess % (n = 496)	p-Value
Demographics
Male	72	85	<0.01
Median age (IQR)	39 (33–43)	37 (32–43)	0.14
Homeless past 6 months	8	14	0.07
Primary source of income past year
Informal work/odd jobs	59	61	0.74
Legal job with pay	19	15	0.33
Prostitution/sex work	13	3	<0.01
HIV-positive	5	6	0.52
Active syphilis (≥1:8)	12	7	0.10
Drug use (past 6 months)
Injected heroin alone	83	80	0.35
Injected methamphetamine alone	15	14	0.72
Injected heroin and meth together	55	50	0.38
Smoked methamphetamine	31	19	<0.01
Form of heroin usually injecteda
Black tar	100	99	0.25
Color of meth usually injecteda
Clear (crystal)	79	73	0.26
White or other color	21	27
Injected daily	90	87	0.45
Injected drugs alone	56	56	0.99
Injected drugs with friends	46	45	0.79
Injected with family member/spouse	16	7	<0.01
Median number uses per syringe	5 (3–10)	5 (3–10)	0.59
Sought someone to help you inject	20	9	<0.01
Receptive syringe sharing	40	40	0.90
Receptive works sharing	47	43	0.43
Traded sex for money or drugs	14	4	<0.01
Locations injected drugs
Your home	69	63	0.16
Shooting gallery	42	47	0.33
Someone else’s home	13	12	0.88
Police and criminal justice (past 6 months)
Arrested for sterile syringes	11	9	0.50
Arrested for used syringes	9	12	0.44
Arrested for track marks	23	15	0.03
Police asked you for money	39	28	0.02
Police affected where you use drugs	15	7	<0.01
Incarcerated	57	50	0.14
Injected during incarceration	17	13	0.18
Shared syringes during incarceration	16	10	0.06

IDUs, injection drug users; IQR, interquartile range.

^a.

Table 2 presents the results of the multiple logistic regression analysis for factors independently associated with abscess in the past 6 months. Participants who reported an abscess were significantly more likely to report injecting with a family member or spouse, smoking methamphetamine, seeking someone to help them inject, trading sex for money and drugs, and reporting that police affected where they use drugs. Gender was not independently associated with abscess, nor did its addition to the model substantially change the estimates shown in Table 2; therefore, it was excluded from the final model. There were no statistically significant interactions with gender.

Table 2.

Factors associated with abscess in the past 6 months^a

Variable	aOR (95% CI)
Principle source of income was prostitution/sex work	4.56 (2.08–10.00)
Smoked methamphetamine	1.65 (1.05–2.62)
Sought someone to help you inject	2.06 (1.18–3.61)
Police affected where you use drugs	2.14 (1.15–3.96)

aOR, adjusted odds ratio; CI, confidence interval.

^aAll variables refer to exposures during the past 6 months.

Discussion

We found a high prevalence of recent abscess and self-treatment among IDUs in Tijuana, Mexico. Of particular concern are the high prevalence of self-drainage under unhygienic conditions and low prevalence of seeking medical care. This combination of reticence to seek care and improper self-treatment creates an environment in which abscesses that could be treated easily in a medical setting may progress to more severe and even life-threatening infections requiring complex and expensive care.

Appropriate interventions to expand abscess care and utilization need to be identified and implemented.

Approximately half of the IDUs treated their abscess with either heated or unheated Aloe vera leaf. We are unaware of any other studies that have reported the use of Aloe vera leaf in treating abscesses among IDUs. The efficacy of Aloe vera to treat skin abscesses, improve wound healing, and decrease inflammation is not well studied, nor is there consistency in the methods by which Aloe vera has been used or tested in the existing literature. The therapeutic use of Aloe vera remains an area of controversy, in part due to the differences in plant storage, treatment, and the use of different plant components alone or in combination to treat myriad diseases. Polysaccharides in the gel from the inner leaf may have anti-inflammatory and antibacterial properties.^17,18 Heating the gel prior to application may degrade these unstable polysaccharides and decrease their therapeutic potential;¹⁹ however, heat may also increase circulation in the infected skin, improving immune system defenses at the infected site, facilitating healing. Further study of the biochemistry and pharmacology of components of the Aloe plant (i.e., leaf exudates vs. inner gel) and whether heating changes the efficacy of the therapy is warranted.

IDUs whose primary income source was sex work were more than four times as likely to report an abscess. This finding is consistent with two prior studies,^1,3 one of which attributed this association to “the generally poor circumstances under which the addicted prostitutes live, work, and inject drugs”.¹ In our study, IDUs who reported sex work as their primary income source had marginally higher incomes than those who reported other primary sources of income and were not more likely to be homeless or to share syringes, suggesting that the conditions under which they inject may be no less hygienic than other IDUs (data not shown). Another hypothesis that warrants examination is that sex workers are more likely to be exposed to bacteria. Recent studies on methicillin-resistant Staphylococcus aureus (MRSA) have underscored skin-to-skin contact as a risk factor for skin pathogen transmission, with outbreaks seen among men who have sex with men, prisoners, wrestlers, football teams, and lap dancers.^20–28 We postulate that IDU sex workers in Tijuana are similarly at high risk due, in part, to frequent skin contact with clients. Therefore, sex workers may have an added occupational exposure risk beyond their own injection drug use that may account for their high prevalence of skin abscesses. Whether MRSA is a concern for IDUs in Tijuana is unclear, as we are unaware of any MRSA prevalence studies among IDUs in Mexico. Appropriate MRSA studies would help to inform effective and efficient treatment of abscesses in this population.

One of every five IDUs who reported an abscess sought someone to help them inject, which was independently associated with recent abscess. This is consistent with a prior study that found an association between requiring help to inject and cutaneous injection-related infection among IDUs recruited at a supervised injection facility in Vancouver, a city known for its large number of IDUs injecting powder heroin and cocaine.⁶ The authors of that study proposed that injection assistance may increase the risk of bacterial exposure when the assister injects first, introducing bacteria to the shared syringe, but this hypothesis is not well supported since their analysis controlled for syringe sharing. In our setting, we hypothesize that venous sclerosis, which is common among long-time injectors of black tar heroin, leads IDUs to seek help with injection. Venous sclerosis increases the likelihood of skin-popping, which has been shown in several studies to be associated with abscesses. Thus, it is possible that the association between abscess and seeking injection help is not causative. However, it may also be that the assister or ‘hit doctor’ is more colonized with bacteria than the average IDU and serves as an ‘infection hub’ for those whom he/she assists. In either case, our findings suggest that persons who assist others with injection are important targets for interventions that promote more hygienic injection practices, including skin and hand cleansing prior to injection.

We examined variables related to law enforcement, as prior studies in this population identified policing practices as correlates of syringe sharing²⁹ and HIV infection.¹⁵ Interestingly, we found an independent association between abscess occurrence and reports that police had affected where the IDU used drugs. Fear of police has been associated with a number of risky injection practices, including syringe sharing,³⁰ rushed injections,^31,32 and the use of shooting galleries,^30,33,34 that may contribute to abscess risk. Why this particular variable and not others more directly associated with injection practices (e.g., rushed injection) achieved significance in our model warrants further exploration.

Of interest, we identified an association between abscess and methamphetamine smoking, but not injecting. There are few studies investigating the differences in routes of methamphetamine administration in relation to skin infection risk. Methamphetamine is known to cause formication and skin picking, leading to abrasions that violate the protective barrier against invading pathogens.³⁵ Smoking methamphetamine, as opposed to injection of the drug, may involve the contamination and subsequent sharing of specific drug paraphernalia (e.g., pipes, straws) that transmit pathogens that colonize the nasopharynx more readily from person to person. An alternative explanation is that frequent methamphetamine smoking is associated with dental caries, cracked lips, and sores around the mouth and nose, which may promote colonization. Several studies suggest that crack smoking can promote hepatitis C virus transmission, and possibly HIV transmission, through these potential pathways.^36–38 Further study is needed on the difference in abscess risk based on the method of methamphetamine administration, as well as to identify the specific behaviors responsible for this finding.

Our study relied on self-reports of abscess rather than clinical diagnosis and therefore may have suffered from misreporting. However, a prior study that compared abscess self-report and physical examination among IDUs found a concordance of 89%.⁴ We did not provide a formal definition of ‘abscess’ to study participants, leaving them to define for themselves what constitutes and abscess – a definition that may have varied across study participants; however, we did collect information on symptoms to help ensure the validity of the self-report. Further, selection bias may have affected our results, as IDUs with abscesses that progressed to serious levels of severity, including endocarditis and/or death, would not have been captured in the study. This may help to explain the very high rate of healing we found after invasive self-treatment. Similarly, we were unable to capture information on serious complications that directly resulted from self-treatment methods. Our study omitted questions on specific disinfection practices (e.g., pre-injection skin cleansing and hand washing) that have previously been associated with abscess and may have confounded our findings. Due to the cross-sectional nature of our study we cannot establish temporality of exposures and outcomes and thus are limited to identifying correlates rather than predictors.

In conclusion, we found that abscesses are common among IDUs in Tijuana, but appropriate and timely medical treatment of these infections is rare. Interventions are needed to reduce barriers to the availability and utilization of medical care for abscesses and to provide IDU sex workers, ‘hit doctors’, and those who smoke methamphetamine with the information and equipment necessary to reduce abscess risk.