Shooter’s Abscess: Foot Infections in People Who Inject Drugs

Abstract

Musculoskeletal infections are common among people who inject drugs. Little is known about the prevalence and characteristics of skin and soft-tissue infections in this patient population, especially in the lower extremity. The aim of this study is to describe the clinical characteristics, bacterial pathogens, and clinical outcomes in adults with foot infections due to intravenous drug use. Nine patients for whom the podiatric service was consulted for lower-extremity abscess secondary to intravenous drug use were included in this study. Most infections (66.7%) were located on the dorsal foot. Two of the abscesses (22.2%) resulted in amputation, with one toe amputation and one ray amputation. Most infections grew methicillin-resistant Staphylococcus aureus (MRSA) (66.7%). Although our investigation had limitations, we conclude that it is likely people who inject drugs will present with infections due to MRSA and will require surgery to resolve the infection.

Musculoskeletal infections are common among people who inject drugs [1]. These patients are more prone to having multiple medical comorbidities, including human immunodeficiency virus (HIV) and hepatitis C. People who inject drugs require longer hospital stays to treat related infections, which can cause substantial resource utilization [2,3]. Complications due to skin and soft-tissue infections are common, including direct extension of subcutaneous abscess into vital areas or structures, necrotizing fasciitis, myositis, osteomyelitis, bacteremia, and sepsis [4].

Little is known about the prevalence and characteristics of skin and soft-tissue infections in people who inject drugs [5]. The paucity of data is often attributed to this populations’ aversion to medical treatment and delay of proper care [6]. Much of the current literature focuses on skin and soft-tissue infection in the upper extremity. There is little published data on the clinical presentation or treatment results of shooter’s abscess of the foot, where injections have been noted to occur 19% of the time [7]. The aim of this study is to describe the clinical characteristics, bacterial pathogens, and clinical outcomes in adults with foot infections due to intravenous drug use.

Methods

After approval by the University of Texas Southwestern Medical Center institutional review board, medical records of patients admitted to Parkland Hospital (Dallas, TX) treated between August 2019 and January 2022 were reviewed. Nine patients for whom the podiatric service was consulted for lower-extremity abscess secondary to intravenous drug use were identified. An abscess was defined as any cutaneous manifestation with erythema, induration, fluctuance, and purulent drainage. Intravenous drug use was identified via either a documented admission of intravenous drug use in the anatomic area of infection or clinical evidence of injection in the foot or ankle with positive drug screening results. Clinical evidence of injection was defined as localized erythema with injection marks about the anatomic area of interest. Patients were excluded if they denied intravenous drug use and had negative drug screening.

Data were collected with respect to patient demographics, medical comorbidities, anatomic locations of the infection, lab values, and microbiologic culture results. The outcomes evaluated were the type of surgical intervention required and need for lower-extremity amputation. Patient data were summarized using descriptive statistics. Median, mean, and standard deviation were used for continuous variables, and frequencies and percentages were used for categorical variables.

Results

The mean ± SD age of the study cohort was 38.2 ± 7.7 (range, 29–51) years and 55.6% of subjects were men. Demographics and clinical characteristics are presented in Table 1. There were no bilateral cases, and the majority of infections (66.7%) were located on the dorsal foot. Drug screening results were positive for heroin and methamphetamines in 44.5% of patients; heroin alone in 22.2% of patients; methamphetamines alone in 11.1% of patients; cocaine and marijuana in 11.1% of patients; and heroin, fentanyl, and marijuana in 11.1% of patients. The mean ± SD white blood cell count was 13.4 ± 6.2 10⁹ cells/L, C-reactive protein 5.9 ± 5 mg/L, and erythrocyte sedimentation rate 56.9 ± 23.0 mm/hr at admission.

Table 1. Patient Demographics, Comorbidities, and Clinical Characteristics

Table 1. Patient Demographics, Comorbidities, and Clinical Characteristics

All nine patients required surgery to treat their infections. Two of the abscesses (22.2%) resulted in amputation, with one toe amputation and one ray amputation. Of the nine patients, 44.5% were treated with composite graft application (Integra bilayer, Integra LifeSciences, Princeton, New Jersey), 11.1% underwent initial primary closure, 22.2% were packed open and eventually went on to delayed primary closure during the admission, 11.1% underwent negative pressure wound therapy application and eventual split thickness skin graft, and 11.1% healed by secondary intention.

The majority of infections grew methicillin-resistant Staphylococcus aureus (MRSA) (66.7%). Five patients had single pathogen infections with MRSA. One patient grew MRSA and Streptococcus species, one patient grew Streptococcus species alone, one grew methicillin-sensitive Staphylococcus aureus (MSSA), and one displayed growth of multiple organisms (including Proteus mirabilis, Klebsiella pneumonia, and Prevotella intermedia) (Table 2).

Table 2. Microbiological Cultures

Table 2. Microbiological Cultures

Discussion

The results of this cohort study indicate that MRSA is a common pathogen in this patient population. In other studies from our group, the incidence of MRSA is low at 9.8% [8]. Our results are similar to the published literature as demonstrated by Pong et al [9], who compared forearm and wrist infections in intravenous drug users and those who do not use drugs over a 10-year period. Among intravenous drug users, MRSA was the most likely bacterial pathogen [9]. Although MRSA was the most common isolate for our study, both Streptococcus species and MSSA were noted in three of nine specimens. This is similar to the published literature by Ho et al, who evaluated 130 intravenous buprenorphine users and found an incidence of 20% MSSA isolate from both tissue and blood cultures [10]. Cutaneous and subcutaneous abscesses in intravenous drug users have been found to be polymicrobial, which include anaerobic as well as aerobic gram-positive cocci in more than 50% of cases [4]. In contrast, our study reported an incidence of 88.9%.

Saldana and colleagues [11] noted that 40% of skin and soft-tissue infections are polymicrobial in people who inject drugs. We only noted that to be the case in one of nine specimens. They found that gram positive cocci were commonly isolated but according to more recent data, Staphylococcus aureus (including MRSA) is isolated less frequently among intravenous drug users due to an increase in commensal organisms of the oral cavity [11]. Summanen [12] showed that 67% (n = 86) of abscesses from intravenous drug users were of oral origin compared to 25% in nonintravenous drug users. Compared with noninjection drug users, cutaneous abscesses in injection drug users are less likely to involve Staphylococcus aureus, including MRSA, and more likely to involve Streptococci and anaerobes [13]. As our study was retrospective, we did not collect whether patients licked the needle or their skin prior to injection. In a 40-person interview series, Deutscher and Perlman [14] reported three subjects who perform licking as a ritualistic practice, and four as a means of cleaning the needle.

One limitation with the present study is small sample size. Given the small cohort, generalizations to this population are limited. There are wide variations in bacterial presence across the nation and world so generalizations regarding present species may be reduced as well. These differing pathogens may drive outcomes depending on what region people who inject drugs are treated. Data were also collected from a single institution with a limited number of patients treated by the podiatric surgery service. Thus, these results might not be representative of our entire institution or even other institutions throughout the country.

Parkland Health and Hospital is a safety net hospital that provides medical care to uninsured and under-insured patients, creating an underlying selection bias that may not represent the nation as a whole. The hospital serves an estimated 15,000 homeless patients [15]. It may be that the individuals in this paper were more likely than others to have housing instability – sleeping at more than one location in the past week or sleeping at a shelter, both of which increase the risk of nasal MRSA colonization by three times. Showering at a shared facility increases the odds by 13.7 times [16].

Although our sample was small, the podiatry team at Parkland Hospital has a specialized focus on lower-extremity infections and limb preservation, which could skew outcomes toward a positive result. Due to the retrospective nature of our study, we may not have included all people who inject drugs with our inclusion criteria, as we only included patients where our service was consulted, and with appropriate documentation. These limitations tend to underestimate the population of our investigation.

Although our investigation had limitations, we conclude that it is likely that people who inject drugs will present with infections due to MRSA and require antibiotic coverage for such. We hope that this data can be used by others in similar urban practice settings to improve care of this patient group. We recommend further longitudinal studies to track the clinical characteristics and trends of this difficult to treat patient population.