Under the skin: The relationship between subcutaneous injection and skin infections among people who inject drugs

Abstract

Objectives:

People who inject drugs (PWID) are at increased risk for numerous negative health outcomes. Subcutaneous injections (aka skin popping) can result in greater risk of skin and soft tissue infections (SSTIs), but less is known about PWID who choose this route of administration. This study compares subcutaneous injectors to intravenous injectors, characterizes those who inject subcutaneously, and examines whether subcutaneous injection is associated with SSTIs in the past year.

Methods:

A cohort of hospitalized PWID (n=252) were interviewed regarding injection-related behaviors, history of SSTI, and knowledge of subcutaneous injection risk. We examined differences between those who do and do not inject subcutaneously and used a negative binomial regression model to estimate adjusted odds associating subcutaneous injection and SSTI.

Results:

Participants averaged 38 years, with 58.3% male, 59.5% White, 20.6% Black, and 15.9% Latinx. PWID who performed subcutaneous injection were not demographically different from other PWID; however, the mean rate of past year SSTIs was higher for persons injecting subcutaneously than for those who did not (1.98 vs 0.96, p<.001). Persons injecting subcutaneously did not differ from those who injected intravenously in terms of their knowledge of subcutaneous injection risk (p=.112) and knowledge score was not associated with SSTIs (p=.457).

Conclusions:

PWID who perform subcutaneous injections are demographically similar to other PWID but had higher rates of past year SSTIs. Knowledge of subcutaneous injection risk was not associated with SSTI risk.

Introduction

When comparing routes of illicit drug administration, injection drug use (IDU) is associated with the most diverse and harmful effects.¹ IDU is associated with bacterial infections (e.g., blood sepsis, skin and soft tissue infection [SSTI], endocarditis), viral infections such as HIV and Hepatitis C, and self-reported poor physical and mental health.² Intravenous IDU is the most common injection route; however, people who inject drugs (PWID) also commonly inject subcutaneously (under the skin; aka skin popping), often due to an inability to obtain a vein. Poor vascular structure can result from years of injection into the same vein.³

A limited number of studies have shown that non-intravenous injection, such as subcutaneous or intramuscular injection, is a significant risk factor for SSTIs.^4–6 Some PWID report accidentally missing their vein (“missed hit”) and inadvertently injecting subcutaneously.⁷ Hope et al. reported that over half of their PWID cohort had a history of ever missing their vein when aiming to inject intravenously, sometimes with substantial frequency; those who had experienced a missed hit were two times more likely to have experienced an SSTI.⁸ Pirozzi et al. analyzed a small cohort of PWID (n=49) presenting with lower extremity SSTI secondary to subcutaneous injection, of which 85.7% underwent surgical debridement, 63.3% in a formal operating room setting.⁹ In a recent analysis, Wright et al. identified subcutaneous or intramuscular injection as a risk factor for SSTIs. Of those who performed subcutaneous injection, 77.3% (n=136) reported having an SSTI in the past year.¹⁰ Fink and colleagues assessed PWID in California to identify determinants of SSTI and prevalence of self-treatment. Findings suggested that women had greater odds of a recent SSTI compared to men and that Latinx participants had higher odds of abscess self-treatment.¹¹ Additional studies have found similar differences by sex, with past work suggesting possible reasons for such differences, including that women may be more likely to rely on a partner for injection, lack knowledge of proper injection technique, and have less accessible veins.^12–16 Overall, demographics of the study population in these reports are predominately white males, which limits generalizability to a more diverse population.^4,6,8–10

The association between subcutaneous injection and SSTIs may also be a function of drug type. In a cross-sectional study addressing two urban PWID populations in the United States, Summers et al. found that having an abscess was independently associated with using black tar heroin as opposed to powder heroin.¹⁷ The authors reported that of those injecting black tar heroin, 63% performed subcutaneous injection when having difficulty accessing a vein; comparatively, 36% of those injecting powder heroin (more common in the Northeastern United States) performed subcutaneous injection after having venous access difficulties.¹⁷ Black tar heroin, with its thick and gummy consistency, is more difficult to inject and can lead to venous scarring with chronic use.¹⁸ PWID using black tar heroin report significantly more abscesses per year versus those injecting powder heroin.¹⁷ Past work has also shown that those injecting a combination of heroin and cocaine (speedballs) report a greater number of abscesses as compared to methamphetamine and cocaine injectors.^5,6,14

SSTIs have significant clinical and economic implications.^4,19,20 Tookes et al. reported that treatment for SSTIs in PWID cost a safety-net hospital $11.4 million over a 12-month period.²¹ Ciccarone et al. found that 32% of PWID admitted to San Francisco General Hospital experienced a drug-related SSTI and that SSTIs were among the most common diagnoses among patients.²² SSTIs can lead to severe consequences, including surgery, amputation, sepsis or death. Though past studies have demonstrated an association between SSTIs and subcutaneous injection, the work is limited.

As subcutaneous injection can significantly increase the risk of SSTI in PWID, the primary goals of this study are to 1) characterize those who report subcutaneous injection, 2) compare those engaging in subcutaneous injection to those injecting intravenously, and 3) examine predictors (subcutaneous injection, intramuscular injection, years IDU, injection frequency, HIV positive status, SSTI knowledge, sex, race, and ethnicity) of SSTI. We hypothesize that 1) women will report more subcutaneous injections than men, 2) those who report subcutaneous injection will have a lengthier injection history and report a greater number of SSTIs, and 3) knowledge of SSTI risk associated with subcutaneous injection will be greater in those who inject intravenously.

Methods

Study Design

After obtaining Institutional Review Board approval, hospitalized PWID (n=252) were recruited for a randomized trial providing education on hygienic injection practices from an academic safety-net hospital in Boston.²³ Participants who were at least 18 years old and reported IDU a minimum of three times in the week preceding hospitalization met inclusion criteria. The current analysis focused on baseline data from the 60 to 90-minute in-person interview designed to assess injection practices over the last 3 months.

Measures

We assessed the following background characteristics: age, sex, race, ethnicity, and education.

SSTIs. Participants were asked to self-report the number of SSTIs experienced within the past year. SSTIs were defined as “abscesses (red, hardish infected lumps that contain pockets of pus), ulcers (open infected sores that looks like a crater), and cellulitis (a more widespread skin infection) that occur at the injection site.”

HIV/AIDS and HCV diagnosis. Participants were asked to self-report any history of HIV/AIDS or HCV.

Drug and injection behaviors. Participants were asked how frequently they injected drugs over the last 3 months. Age initiated IDU, as well as primary drug injected were reported. Participants were asked, “How many times did you inject under your skin (skin pop rather than into the vein).” Similarly, they were also asked “How many times did you inject into your muscle (muscling).” Responses were coded as yes/no if any subcutaneous or intramuscular injection was reported.

Knowledge of subcutaneous injection risk. We used a single “True/False” item to assess knowledge of subcutaneous injection risk: “You can get a skin infection more easily if you inject drugs under your skin (skin pop) rather than into a vein.”

Analytical Methods

We report descriptive statistics to summarize sample characteristics. We used t-tests and χ2-tests to examine differences between PWID reporting subcutaneous injection and those who did not, addressing hypotheses related to demographic characteristics, injection history, number of SSTIs, and knowledge of SSTI risk. A negative binomial regression model, adjusting for sex, race, ethnicity, intramuscular injection, diagnosis of HIV, years of injection use, frequency of injection, and response to the infection knowledge question, was used to estimate the adjusted association between subcutaneous injection and number of SSTIs reported in the last year.

Results

Demographic and drug use characteristics

Participants averaged 37.9 (± 10.7) years of age, 58.3% were male, 59.5% were White, 20.6% were Black, 19.8% identified other or mixed racial origins, and 15.9% were Latinx (Table 1). Mean years of education was 11.6 (± 2.25). Just over 90% said their primary injection drug was opioids. Most participants (96.8%) reported a lifetime diagnosis of Hepatitis C. Additionally, 32 (12.7%) reported a positive diagnosis of HIV/AIDS. Mean age of initiating IDU was 25.8 (± 9.6), the mean number of past year SSTIs was 1.54 (± 2.31), and 53.0% reported that subcutaneous IDU could increase the likelihood of skin infection on the knowledge item. Intramuscular injection was relatively uncommon (n=43, 17.15%) in this cohort.

Table 1:

Background Characteristics and Injection Drug Use Behaviors and Practices by Subcutaneous Injection (Skin Pop).

		SKIN POPPED?
	Sample (n = 252)	No (n = 108)	Yes (n = 144)	p = a
Age	37.9 (± 10.7)	38.6 (± 11.0)	37.4 (± 10.7)	.381
Sex (Male)	147 (58.3%)	66 (66.1%)	81 (56.2%)	.439
Race
White	150 (59.5%)	58 (53.7%)	92 (63.9%)
Black	52 (20.6%)	26 (24.1%)	26 (18.1%)	.259
Other	50 (19.8%)	24 (22.2%)	26 (18.1%)
Latinx (Yes)	40 (15.9%)	18 (16.7%)	22 (15.3%)	.765
Yrs Education	11.6 (± 2.25)	11.3 (± 2.63)	11.8 (± 2.25)	.145
HIV Positive (Yes)	33 (13.2%)	19 (17.6%)	14 (9.8%)	.070
Ever HCV (Yes)	242 (96.8%)	103 (96.3%)	138 (97.2%)	.683
Primary IDU Drug (Opioids)	227 (90.1%)	96 (88.9%)	131 (91.0%)	.584
Age Initiated IDU	25.8 (± 9.6)	25.7 (± 10.4)	26.0 (± 9.0)	.804
# SSTIs (Past Year)	1.54 (± 2.31)	0.96 (± 1.51)	1.98 (± 2.68)	<.001
Intramuscular Injection (Yes)	43 (17.1%)	10 (9.3%)	33 (23.0%)	.004
Knowledge of subcutaneous injection risk	133 (53.0%)	51 (47.25%)	82 (57.3%)	.112

^aP-value for t-test for differences in means or Pearson χ² – test of independence.

Subcutaneous and intravenous injection comparisons

Participants who injected subcutaneously were compared to those who injected intravenously. Findings from t-tests showed that participants who reported subcutaneous skin injection had significantly higher (1.98 vs 0.96) unadjusted rates of past year SSTIs (Table 1). They were also significantly more likely to report intramuscular injections (23.0% vs 9.3%). In the subcutaneous injection group, 57.3% (82 out of 144) answered the knowledge of subcutaneous injection risk question correctly, whereas 47.3% (51 out of 108) of non-subcutaneous injectors answered correctly (p=.112). Between group differences on the other background characteristics were not significant statistically at the .05 level (see Table 1).

Predictors of SSTIs

Adjusting for demographic characteristics, years of IDU, total injection frequency, HIV status, and intramuscular injection, the mean rate of past year SSTIs was almost 2 times higher (IRR = 1.96, 95%CI 1.38; 2.78, p < .001) for persons injecting subcutaneously than for those who did not (Table 2). Number of past year SSTIs was not associated significantly with years of IDU (IRR = 1.01, 95%CI 1.00; 1.03, p = .120), knowledge of subcutaneous injection risk (IRR = 0.94, 95%CI 0.79; 1.11, p = .435), HIV status (IRR = 0.68, 95%CI 0.38; 1.22, p = .199) or intramuscular injection (IRR = 1.22, 95%CI 0.86; 1.73, p = .273). Persons with greater frequency of IDU had significantly higher mean rates of SSTIs (IRR = 1.28, 95%CI 1.06; 1.55, p = .012).

Table 2:

Negative Binomial Regression Model Estimating the Adjusted Association of Skin Popping with the Number of Self-Reported Skin Infections in the Past Year (n = 251).

	IRR	(95% CI)	z (p = )
Gender (Male)	0.76	(0.52; 1.10)	−1.47 (.142)
Race (White)	0.86	(0.56; 1.31)	−0.72 (.470)
Ethnicity (Latinx)	1.01	(0.64; 1.60)	0.04 (.966)
Years IDU	1.01	(1.00; 1.03)	1.55 (.120)
Knowledge	0.94	(0.79; 1.11)	−0.78 (.435)
Log(Total Injection Frequency)	1.28	(1.06; 1.55)	2.52 (.012)
HIV+ (Yes)	0.68	(0.38; 1.22)	−1.29 (.199)
Muscled (Yes)	1.22	(0.86; 1.73)	1.10 (.273)
Skin Popping (Yes)	1.96	(1.38; 2.78)	3.77 (<.001)
Constant	0.46

Discussion

Drug injection is associated with a multitude of negative health outcomes.²⁴ This study, situated in a city where PWID predominantly use powder heroin as opposed to tar heroin, reaffirms the association between subcutaneous injection and SSTIs. The number of self-reported SSTIs was approximately two times higher in those who report injecting subcutaneously, even when controlling for injection frequency.

Increased risk of SSTI is at least partially a result of soft tissue damage secondary to injection of acidic drug solutions, venous collapse, and inadequate injection technique, with injection sites becoming vulnerable to infections when exposed to organisms during the injection process.⁸ This tissue damage can lead to necrosis and create an environment that is particularly susceptible to the growth of anaerobic organisms, leading to infection.^25,26

The majority of our cohort (57%) skin-popped and persons who perform subcutaneous injections were not demographically distinct. Contrary to our hypotheses, women did not report injecting subcutaneously more than men. Sex was also unrelated to number of SSTIs, which is inconsistent with past work.^12–16 Sex differences related to subcutaneous injection have been unexplored and we speculated that women would be more likely to skin pop due to having difficulty injecting intravenously.^12–16 However, this was not the case and may be related to extraneous variables we did not examine. In addition, a prior analysis found that unintentional subcutaneous injection was more common in PWID who had injected for longer and more frequently.⁸ We did not find differences in length of injection history between those who injected subcutaneously vs intravenously and cannot state definitively whether participants skin popped intentionally. Those who skin popped were more likely to inject intramuscularly, thus leading us to surmise that difficulty with venous access may play a role. Though our study serves as an initial look at characteristics associated with subcutaneous injection, many questions remain surrounding decision-making associated with choosing alternate forms of injection.

Limited past research has examined the association between knowledge of SSTIs or subcutaneous injection risk and injecting subcutaneously. We found no differences in knowledge between those engaging in subcutaneous vs. intravenous injection. In addition, knowledge was not associated with increased SSTI risk. In a qualitative study, Gilbert et al. reported that there was not a lack of knowledge about the relationship between subcutaneous and SSTI in their cohort of 12 PWID. Rather, subcutaneous injection seemed to be motivated by lack of or missed venous access.²⁷ These findings suggest that knowledge deficits may not contribute to choice to inject subcutaneously, but more likely, lack of injection options for PWID who have difficulty with venous access.

This study had limitations. The cohort consisted of hospitalized patients which may include people predisposed to higher risk of SSTI (approximately two-thirds had admission diagnoses of SSTI) and therefore our findings may not be generalizable to all PWID.²³ Additionally, the cohort of PWID only includes those who had access to and felt an urgent need to seek treatment at a hospital. As many SSTIs are self-treated by PWIDs outside of hospitals, population differences could account for any divergence between the results of this study and previously published reports. History of SSTIs were self-reported and therefore may be subject to inaccuracies based on the participant’s recall. We also assume that subcutaneous injection represents a problem with venous access rather than an intentional act. There are no qualitative reports suggesting that subcutaneous injection has advantages in terms of greater or faster euphoria. In fact, effects from intravenous injection are typically felt within 15 to 30 seconds, whereas effects from subcutaneous injection can take approximately 3 to 5 minutes after administration.^28,29 One important strength of this study is the diverse cohort - nearly half women and half non-white - as compared to most reports with predominantly white male participants.

If correct information about the SSTI risk of subcutaneous injection was universal, some PWID might choose not to inject subcutaneously. But for many drug injectors, the inability to find venous access may be irrevocable. Indeed, other work demonstrates that some PWIDs seek out other drug injectors who might be able to find venous access and help with injection.^13,30 Certainly, one way to reduce SSTIs (and other infectious complications) of drug use would be to change route of administration; intranasal use of heroin or cocaine, for instance would be a SSTI risk reduction avenue.^31,32 Additional access to harm reduction interventions, such as safe injection facilities, could allow for better training of safer injection techniques for PWID and others in their network, which could decrease disease risk and reduce the economic impact of drug injection.

Conclusions

PWID who perform subcutaneous injections are not demographically different from other PWID. Those who inject subcutaneously had higher rates of past year SSTIs when compared to intravenous injectors. Knowledge that one can get a SSTI more easily if they inject drugs under their skin rather than into a vein was low in both those who do and do not inject subcutaneously. Hygiene education surrounding skin and needle cleaning, provision of new needles, and assistance in locating veins for injection may reduce SSTIs in PWID and could be expanded among hospitalized patients and incorporated in syringe exchange protocols.²³