Microbiology of Musculoskeletal Infections in People Who Inject Drugs at a Rural Tertiary Care Center : Infectious Diseases in Clinical Practice

Secondary Logo

Journal Logo

Original Article

Microbiology of Musculoskeletal Infections in People Who Inject Drugs at a Rural Tertiary Care Center

London, Abby BS; Lin, Dan MPH; Schade, Meredith MD‡,§

Author Information
Infectious Diseases in Clinical Practice: November 2022 - Volume 30 - Issue 6 - e1198
doi: 10.1097/IPC.0000000000001198

Abstract

The mortality of the opioid epidemic has been well documented.1 Nonfatal infectious complications of injection drug use (IDU) garner less attention than deaths but have significant morbidity. Injection drug use tends to occur in younger individuals with limited medical comorbidities. Infections of the musculoskeletal (MSK) system associated with IDU include septic arthritis, vertebral osteomyelitis, skin, and soft tissue infections. The consequences of complex musculoskeletal infections (MSKIs) may include loss of productivity, permanent disability, and even death.

The microbiology of MSKI in people who inject drugs (PWID) is usually the commensal flora of the skin.2 Unfortunately, methicillin-resistant Staphylococcus aureus (MRSA) infections have risen in community-onset infections.3,4 In addition to commensal flora, organisms colonizing the drug or devices used for injection may also lead to infection.5 With the acceleration of the opioid epidemic, our investigation aimed to determine the number of MSKIs in PWID, describe the anatomic location and microbiology of the infections, and assess outcomes at 1 year at our rural academic medical center.

METHODS

Data Source and Study Design

We performed a single center, 11-year case series review of adults admitted to the Penn State Health Milton S. Hershey Medical Center (PSMSHMC) with MSKIs related to self-reported IDU. The PSMSHMC is a 628-bed academic medical center located in rural central Pennsylvania. The period studied was from January 1, 2008, to December 31, 2018. As no diagnosis codes exist for injection drug-use associated infections, codes for MSKIs and codes related to IDU were assembled. Codes used are located in the appendix, https://links.lww.com/IDCP/A50. Penn State Health i2b2 database was queried using the TrinetX application to identify persons potentially admitted with MSKIs related to IDU. All medical records found by the query were individually reviewed by the authors to determine whether the infection was related to IDU.

Patient Consent Statement

The Human Subjects Protection Office of Penn State College of Medicine determined the study did not require formal institutional review board review as it met criteria for exempt research.

Definitions

A case was defined as an MSKI due to IDU based on patient history of self-reported drug use and physical examination findings of MSKIs documented in the medical record. Patients were characterized as young (18–35 years), middle-aged (36–55 years), or older adults (51–65 years). A diagnosis of infection required a positive culture from an abscess or a sterile site such as blood, tissue, or synovial fluid. Date of infection was date of admission to the hospital. Infections of the axial skeleton included vertebral column, thoracic cage, pelvis, sacroiliac joints, sternum, or chest wall. Infections of the appendicular skeleton included hip, knee, leg, foot, shoulder, arm, elbow, and hand. Soft tissue infections were defined as infections of skin, subcutaneous tissue, fascia, or muscle without bone involvement. Demographics and 1-year outcomes were obtained from the electronic medical record.

Data Collected

For patient characteristics, categorical data (age, sex, race, comorbidities) were summarized as number (percentages) of subjects. Age at time of first admission for MSKIs was adopted when patients had more than one infection recorded. Anatomy and microbiology of MSKIs were assessed overall and by treatment (medical treatment alone vs medical and surgical treatment). One-year outcomes were assessed as: treatment completed, treated and relapsed with same infection, deceased, and lost to follow-up.

RESULTS

A total of 849 medical records were identified with the query, and each record was individually reviewed by 2 of the authors (A.L. and M.S.). Eighty-six infections were found in 82 patients who conformed to the study definition. Cases of MSKIs in PWID increased significantly over the study period, from 1 in 2008 to 17 in 2018 (Fig. 1).

F1
FIGURE 1:
Distribution of MSKI cases in PWID from 2008 to 2018 at the Penn State Hershey Medical Center.

Patient Characteristics

The majority of MSKIs in PWID were found in persons 50 years or younger (Table 1). The race and ethnicity of our patients closely approximated that of our region. Medicaid was the most common insurance coverage. Tobacco use (87.8%) and hepatitis C positivity (64.6%) were the most frequent comorbidities.

TABLE 1 - Patient Demographics and Characteristics
Characteristic Total Patients, n (%)
N = 82
Age
 18–35 y 38 (46.3)
 36–50 y 31 (37.8)
 51–65 y 13 (15.9)
Sex
 Female 33 (40.2)
 Male 49 (59.8)
Race/ethnicity
 White, non-Hispanic 63 (76.8)
 African American 5 (6.1)
 Hispanic 13 (15.9)
 Other 1 (1.2)
Insurance type
 Medicaid 49 (59.8)
 Medicare 4 (4.9)
 Medicare/Medicaid 8 (9.7)
 Commercial insurance 11 (13.4)
 Uninsured 10 (12.2)
Tobacco use disorder 72 (87.8)
HCV* 53 (64.6)
Hypertension 9 (11.0)
Asthma 7 (8.5)
Type 2 DM 6 (7.3)
Alcohol use disorder 4 (4.9)
HIV* 3 (3.7)
CKD 3 (3.7)
Hypothyroidism 2 (2.4)
COPD 2 (2.4)
Mental health disorder
 Depression or Anxiety 15 (18.3)
 Bipolar disorder 7 (8.5)
 Schizophrenia or schizoaffective disorder 3 (3.7)
 ADHD 1 (1.2)
 PTSD 1 (1.2)
*HCV Ab test missing = 10, HIV Ag/Ab test missing = 14.
ADHD indicates attention-deficit/hyperactivity disorder; CKD, chronic kidney disease; COPD, chronic obstructive pulmonary disease; HIV, human immunodeficiency virus; HCV, hepatitis C; PTSD, posttraumatic stress disorder; Type 2 DM, type 2 diabetes mellitus.

Microbiology

Gram-positive infections predominated (74.5%) with S. aureus the most frequent organism isolated (Table 2). Methicillin-susceptible S. aureus (MSSA) was slightly more common (36%) than MRSA (31.4%). In contrast, MSSA is isolated twice as often as MRSA in the general adult population at our center (Director of Microbiology, PSMSHMC, personal communication).

TABLE 2 - Anatomic and Microbiological Patterns of MSKI Episodes With Treatment Methods
Total episodes* Episodes With Medical Treatment Only Episodes With Medical and Surgical Treatment
N = 86 n = 32 n = 52
Anatomic infection site, n (%)
 Axial 53 (61.6) 29 (56.9) 22 (43.1)
 Appendicular 22 (25.6) 1 (4.5) 21 (95.5)
 Axial and appendicular 2 (2.3) 1 (50.0) 1 (50.0)
 Soft tissue 8 (9.3) 1 (12.5) 7 (87.5)
 Axial and soft tissue 1 (1.2) 0 1 (100.0)
Organism, n (%)
 MSSA 31 (36.0) 13 (43.3) 17 (56.7)
 MRSA 27 (31.4) 4 (14.8) 23 (85.2)
Pseudomonas aeruginosa 10 (11.6) 6 (60.0) 4 (40.0)
Serratia marcescens 4 (4.7) 1 (25.0) 3 (75.0)
 Viridans group streptococci 3 (3.5) 1 (33.3) 2 (66.7)
Eikenella corrodens 2 (2.4) 1 (50.0) 1 (50.0)
Enterobacter cloacae 2 (2.4) 2 (100.0) 0
Candida albicans 2 (2.4) 2 (100.0) 0
Candida parapsilosis 2 (2.4) 0 2 (100.0)
Haemophilus species 1 (1.1) 0 1 (100.0)
Enterococcus faecalis 1 (1.1) 0 1 (100.0)
Klebsiella oxytoca 1 (1.1) 0 1 (100.0)
Propionibacterium acnes 1 (1.1) 0 1 (100.0)
 MRSE 1 (1.1) 0 1 (100.0)
Prevotella 1 (1.1) 0 1 (100.0)
*Two patients with axial infection self-discharged prior to treatment.
MRSE indicates methicillin-resistant Staphylococcus epidermidis.

Pseudomonas aeruginosa was the most common gram-negative organism (11.6%). Musculoskeletal infections due to Candida species and streptococci were uncommon.

Anatomy

Infections frequently involved the axial skeleton (61.6%), with infections of the lumbar and thoracic spine the most common (Table 2). The sternoclavicular joint, often described as involved by infection in IDU, was involved in only 3 individuals. Infections of the lower extremity joints were more prevalent than the upper extremity. There were 12 infections involving hips, 5 of the knees, and only 2 each of the elbow and shoulder.

Management of Infections

Infections of the large joints included operative debridement and antimicrobial therapy. Infections of the spine were treated primarily with parenteral antimicrobials; surgery was reserved for failure of antibiotic management.

One-Year Outcomes

Sixty-two of the infections (72.1%) were deemed to be resolved at 1 year (Fig. 2). Sixteen patients (18.6%) were lost to follow-up, including 2 who self-discharged. Five patients (5.8%) had relapsed with the same infection, and 3 patients (3.5%) were documented deceased at 1 year. Cause of death for the 3 deceased persons were out of hospital cardiac arrest, septic shock in setting of decompensated liver failure from alcohol use disorder, and 1 had no etiology documented.

F2
FIGURE 2:
Distribution of 1-year outcomes among 86 episodes MSKIs in PWID at the Penn State Hershey Medical Center.

DISCUSSION

Because of the illicit nature of substance abuse, population-level prevalence of IDU is difficult to measure.6 What we observe, and are better able to characterize, are increasing numbers of infections related to IDU.7 The typical patient in our cohort was a young White male with a median age of 38.5 years, with few comorbidities. Others have previously reported male predominance for IDU. Chandrasekar and Narula8 found that 70% were male, had a median age of 32 years, and had a low burden of medical illness. Allison et al9 analyzed a group from 1998 to 2005 and also found a male predominance. DiGiorgio et al10 examined a later cohort, 2013–2018, and noted a male predominance of 76%. In contrast, we noted a higher percentage of females at 40%. This increase in IDU by women has been noted in other studies. In an analysis of IDU-related infective endocarditis patients, Leahey et al11 found that 46% of their cohort was female.

Anatomy

The anatomic location of MSKIs in PWID reflects hematogenous seeding or local extension from a contiguous source.2 The spine is involved because of its abundant vascularity, and in our study, a majority of infections involved the spine. Others have documented the frequency of spinal infections in IDU. Blecher et al12 described a 60% increase in the overall incidence of spinal infections from 2003 to 2017. In Louisiana, DiGiorgio et al10 identified 45 patients with spinal epidural abscess related to IDU from 2013 to 2018. They postulate that after opioid prescribing restrictions were put in place in Louisiana in the mid 2010s, a transition from oral to intravenous opioid use occurred. In response to one of the highest overdose rates in the nation, Pennsylvania passed legislation to implement a statewide prescription drug monitoring program in 2014, which went live in 2016.13 Our data show a sharp rise in infections from 2015 to 2018. State legislation and increasing physician awareness of the risks of opioids may have also contributed to changes in drug use in our cohort.

Along with infections of the spine, infections of the appendicular skeleton have increased. Using the Nationwide Inpatient Sample, Oh et al14 found the proportion of patients with IDU-related septic arthritis increased from 5% in 2000 to 11% in 2013. In a review of septic arthritis during the opioid epidemic, Ross et al15 noted the knee to be the most commonly infected joint. We found hips were infected more commonly than knees by a 2:1 margin. Groin injection is described in substance use once peripheral veins are no longer accessible.16 Soft tissue infections were a minority in our study, which did not capture persons evaluated in the emergency department and discharged. Monteiro et al17 reported that although PWID experience high rates of cutaneous abscesses and cellulitis, they often delay medical care and self-treat the infections. In a study examining the microbiology of cutaneous abscesses in both IDU and non-IDU, Jenkins et al18 found aerobic gram-positive organisms in 90%, consistent with commensal flora. They also noted that compared with non-IDU, abscesses found in PWID were more often polymicrobial, perhaps reflecting prior treatment or attempts at drainage.

Microbiology

The organisms responsible for infection in PWID may be introduced at the time of injection and with drug preparation. The patient's skin is the source for gram-positive bacteria. From the Centers for Disease Control Emerging Infections Program, persons who inject drugs were estimated to be 16 times more likely to develop invasive MRSA infections than others.3 Aerobic gram-positive organisms were the etiologic agents in 73% of the infections at our hospital with S. aureus responsible for the majority. However, we noted a higher percentage of MSSA infections by a slight margin.

The needle, syringe, drug, and drug diluent may also be a source for infection. Preparation of drug with water from toilets or puddles leads to exposure to environmental gram negatives.19

For example, Pseudomonas has been isolated from the interior of a syringe.20 In a report of 10 cases of P. aeruginosa infective endocarditis in PWID, the major source of P. aeruginosa was believed to be standing water, unboiled tap, or toilet water that contaminated paraphernalia used for injections.21 The isolation of Pseudomonas or Serratia from a sterile site, such as a disc space, should raise concern that IDU may be the cause.

In a retrospective study using a cohort of 800 US hospitals, McCarthy et al22 examined all bacterial infections in PWID during the period 2012–2017. On a much larger scale, they too found the same high percentage of S. aureus infections, a smaller number of gram-negative infections, and fewer infections due to viridans group streptococci.22

Outcomes

In our cohort, infections of the large joints were treated with arthrotomy and drainage as well as antibiotics. Patients with spinal infections received parenteral antibiotics and were less frequently taken for operative intervention. By chart review, we were able to document 72% of the patients had completed treatment and were without infection at 1-year follow-up. An alarming 28% had relapsed with the same infection, died, or were lost to follow-up. In reviewing the microbiology of relapsed infections, all 5 were due to S. aureus with 3 due to MRSA and 2 from MSSA. Four of the 5 patients with relapse of infection had undergone operative debridement as part of their initial management.

There are limitations to our study including its retrospective design and reliance on diagnosis codes to identify the cohort. During the period studied, there was a change in methodology from International Classification of Diseases, Ninth and Tenth Revisions, codes, which may have impacted identification of infections. Identification of IDU as the etiology of infection was based on patient self-reporting with toxicology screening infrequently performed to confirm drug use. The data derived are from a single large center in rural Pennsylvania and may not reflect what is occurring in other regions.

In conclusion, we describe an increasing number of complicated MSKIs related to IDU in the period studied. Staphylococcus aureus continues to predominate as the most frequent isolate, but environmental gram-negative organisms are also found and may be related to drug use practices. Spinal infections are more prevalent than infections of the joints, but all have potential for long-term morbidity. Infectious disease physicians and orthopedic surgeons are charged with confronting opioid use disorder as the driver and MSKIs that result in this at-risk population.

REFERENCES

1. Hedegaard H, Minino A, Warner M. Drug Overdose Deaths in the United States, 1999–2019. NCHS Data Brief no. 428. Hyattsville, MD: National Center for Health Statistics; 2020.
2. Gordon R, Lowy F. Bacterial infections in drug users. N Engl J Med. 2005;353(18):1945–1954.
3. Jackson K, Bohm M, Brooks J, et al. Invasive methicillin-resistant Staphylococcus aureus infections among people who inject drugs—six sites, 2005–2016. MMWR. 2018;67(22):625–628. doi:10.15585.
4. Parikh MP, Octaria R, Kainer MA. Methicillin-resistant Staphylococcus aureus bloodstream infections and injection drug use, Tennessee, USA, 2015–2017. Emerg Infect Dis. 2020;26(3). doi:10.3201/eid2603.191408.
5. Tuazon CU, Miller H, Shamsuddin D. Antimicrobial activity of street heroin. J Infect Dis. 1980;142(6):944–944. doi:10.1093/infdis/142.6.944.
6. Bradley H, Rosenthal EM, Barranco MA, et al. Use of population-based surveys for estimating the population size of persons who inject drugs in the United States. J Infect Dis. 2020;222(suppl 5):S218–S229. doi:10.1093/infdis/jiaa318.
7. Miller AC, Polgreen PM. Many opportunities to record, diagnose, or treat injection drug–related infections are missed: a population-based cohort study of inpatient and emergency department settings. Clin Infect Dis. 2019;68(7):1166–1175. doi:10.1093/cid/ciy632.
8. Chandrasekar PH, Narula AP. Bone and joint infections in intravenous drug abusers. Clin Infect Dis. 1986;8(6):904–911. doi:10.1093/clinids/8.6.904.
9. Allison DC, Holtom PD, Patzakis MJ, et al. Microbiology of bone and joint infections in injecting drug abusers. Clin Orthop. 2010;468(8):2107–2112. doi:10.1007/s11999-010-1271-2.
10. DiGiorgio AM, Stein R, Morrow KD, et al. The increasing frequency of intravenous drug abuse–associated spinal epidural abscesses: a case series. Neurosurg Focus. 2019;46(1):E4. doi:10.3171/2018.10.FOCUS18449.
11. Leahey PA, LaSalvia MT, Rosenthal ES, et al. High morbidity and mortality among patients with sentinel admission for injection drug use-related infective endocarditis. Open Forum Infect Dis. 2019;6(4). doi:10.1093/ofid/ofz089.
12. Blecher R, Yilmaz E, Ishak B, et al. Recent increase in the rate of spinal infections may be related to growing substance-use disorder in the state of Washington: wide population-based analysis of the Comprehensive Hospital Abstract Reporting System (CHARS) database. Spine. 2019;44(4):291–297. doi:10.1097/BRS.0000000000002819.
13. Bakewell BK, Miller C, Sherman M, et al. Opioid Prescribing Patterns by Drug Type: The Pennsylvania Experience. Durham, NC: Research Square; 2021. Available at: https://www.researchsquare.com/article/rs-731312/latest.pdf. Accessed May 10, 2022.
14. Oh DHW, Wurcel AG, Tybor DJ, et al. Increased mortality and reoperation rates after treatment for septic arthritis of the knee in people who inject drugs: nationwide inpatient sample, 2000–2013. Clin Orthop. 2018;476(8):1557–1565. doi:10.1097/01.blo.0000534682.68856.d8.
15. Ross JJ, Ard KL, Carlile N. Septic arthritis and the opioid epidemic: 1465 cases of culture-positive native joint septic arthritis from 1990–2018. Open Forum Infect Dis. 2020;7(3):ofaa089. doi:10.1093/ofid/ofaa089.
16. Mackenzie AR. High prevalence of iliofemoral venous thrombosis with severe groin infection among injecting drug users in North East Scotland: successful use of low molecular weight heparin with antibiotics. Postgrad Med J. 2000;76(899):561–565. doi:10.1136/pmj.76.899.561.
17. Monteiro J, Phillips KT, Herman DS, et al. Self-treatment of skin infections by people who inject drugs. Drug Alcohol Depend. 2020;206:107695. doi:10.1016/j.drugalcdep.2019.107695.
18. Jenkins TC, Knepper BC, Jason Moore S, et al. Microbiology and initial antibiotic therapy for injection drug users and non-injection drug users with cutaneous abscesses in the era of community-associated methicillin-resistant Staphylococcus aureus. Merchant R, ed. Acad Emerg Med. 2015;22(8):993–997. doi:10.1111/acem.12727.
19. Harris M, Scott J, Hope V, et al. Navigating environmental constraints to injection preparation: the use of saliva and other alternatives to sterile water among unstably housed PWID in London. Harm Reduct J. 2020;17(1):24. doi:10.1186/s12954-020-00369-0.
20. Rajashekaraiah KR, Rice TW, Kallick CA. Recovery of Pseudomonas aeruginosa from syringes of drug addicts with endocarditis. J Infect Dis. 1981;144(5):482–482. doi:10.1093/infdis/144.5.482.
21. Reyes MP, Ali A, Mendes RE, et al. Resurgence of Pseudomonas endocarditis in Detroit, 2006–2008. Medicine (Baltimore). 2009;88(5):294–301. doi:10.1097/MD.0b013e3181b8bedc.
22. McCarthy NL, Baggs J, See I, et al. Bacterial infections associated with substance use disorders, large cohort of United States hospitals, 2012–2017. Clin Infect Dis. 2020;71(7):e37–e44. doi:10.1093/cid/ciaa008.
Keywords:

injection drug use; septic joint; vertebral osteomyelitis

Supplemental Digital Content

Copyright © 2022 The Author(s). Published by Wolters Kluwer Health, Inc.