Skip to main content
Download PDF
Download ePub

Economic evaluation of the effect of needle and syringe programs on skin, soft tissue, and vascular infections in people who inject drugs: a microsimulation modelling approach

Harm Reduction Journal volume 21, Article number: 126 (2024) Cite this article

A Correction to this article was published on 09 September 2024

This article has been updated

Abstract

Background

Needle and syringe programs (NSP) are effective harm-reduction strategies against HIV and hepatitis C. Although skin, soft tissue, and vascular infections (SSTVI) are the most common morbidities in people who inject drugs (PWID), the extent to which NSP are clinically and cost-effective in relation to SSTVI in PWID remains unclear. The objective of this study was to model the clinical- and cost-effectiveness of NSP with respect to treatment of SSTVI in PWID.

Methods

We performed a model-based, economic evaluation comparing a scenario with NSP to a scenario without NSP. We developed a microsimulation model to generate two cohorts of 100,000 individuals corresponding to each NSP scenario and estimated quality-adjusted life-years (QALY) and cost (in 2022 Canadian dollars) over a 5-year time horizon (1.5% per annum for costs and outcomes). To assess the clinical effectiveness of NSP, we conducted survival analysis that accounted for the recurrent use of health care services for treating SSTVI and SSTVI mortality in the presence of competing risks.

Results

The incremental cost-effectiveness ratio associated with NSP was $70,278 per QALY, with incremental cost and QALY gains corresponding to $1207 and 0.017 QALY, respectively. Under the scenario with NSP, there were 788 fewer SSTVI deaths per 100,000 PWID, corresponding to 24% lower relative hazard of mortality from SSTVI (hazard ratio [HR] = 0.76; 95% confidence interval [CI] = 0.72–0.80). Health service utilization over the 5-year period remained lower under the scenario with NSP (outpatient: 66,511 vs. 86,879; emergency department: 9920 vs. 12,922; inpatient: 4282 vs. 5596). Relatedly, having NSP was associated with a modest reduction in the relative hazard of recurrent outpatient visits (HR = 0.96; 95% CI = 0.95–0.97) for purulent SSTVI as well as outpatient (HR = 0.88; 95% CI = 0.87–0.88) and emergency department visits (HR = 0.98; 95% CI = 0.97–0.99) for non-purulent SSTVI.

Conclusions

Both the individuals and the healthcare system benefit from NSP through lower risk of SSTVI mortality and prevention of recurrent outpatient and emergency department visits to treat SSTVI. The microsimulation framework provides insights into clinical and economic implications of NSP, which can serve as valuable evidence that can aid decision-making in expansion of NSP services.

Introduction

Injection drug use and associated high-risk injecting behaviours (e.g., needle or syringe sharing) are a major public health issue, as they increase the risk of overdose [1] and HIV and hepatitis C virus (HCV) infection [2,3,4]. However, harms associated with injection drug use extend beyond overdose and blood borne infections.

Among people who inject drugs (PWID), skin, soft tissue, and vascular infections (SSTVI) at drug injection sites are the leading cause of emergency department (ED) visits and hospitalizations, globally [5,6,7,8,9,10,11,12]. SSTVI are bacterial infections of the skin and subcutaneous soft tissues, which can lead to inflammatory response (e.g., pain and swelling) as well as formation of lesions and bullae [13]. Some cases of SSTVI may manifest as abscess or cellulitis, which can be treated with antibiotics or incision and drainage procedures [14]. However, without timely treatment, SSTVI can progress into necrotizing infections or sepsis [15]. Sharing and reusing of injection equipment, frequency of injections, and years of injection drug use are all known to increase the risk of SSTVI [16,17,18,19].

In Canada, a confluence of factors may be contributing to increasing SSTVI morbidity among PWID [20, 21]. In recent years, changes in consumption patterns owing to the introduction of more powerful but shorter acting synthetic opioids such as fentanyl in the unregulated market means the frequency of injection drug use may be on the rise [22,23,24]. Meanwhile, because PWID limit their use of health services owing to past experiences of mistreatment and/or stigma [25], treatment for SSTVI is often delayed [26], leading to more serious morbidity and costly hospitalizations [27].

One of the first harm reduction strategies employed was needle and syringe programs (NSP), where PWID are provided with sterile hypodermic needles and other equipment at low to no cost. In Montreal, Canada, NSP started as a grassroots community initiative under CACTUS-Montreal (Centre d’Action Communautaire auprès des Toxicomanes Utilisateurs de Seringues) in 1989 [28], with the aim of preventing HIV transmission [29]. An early study of Quebec’s NSP found that there was an 11% decline in the needle sharing rate (31% to 20%) within two years of beginning operation [30]. More recent studies on NSP reveal that these programs are associated with significant reductions in HIV and HCV transmission among PWID [31, 32]. In addition, NSP may help mitigate SSTVI, as they reduce needle sharing and reuse of injection equipment while contributing to injection cessation [33,34,35,36,37,38]. The reduction in risky injection behaviour and promotion of injection cessation are achieved through outreach programs that accompany NSP, which include education on safe and hygienic injection practices and needle disinfection [36].

Despite evidence of lower injection risk behaviour associated with NSP, Canadian modelling studies of harm reduction interventions have focused on supervised consumption sites and their role in prevention of HIV and HCV transmission [39,40,41,42]. A recent study found cost-effectiveness of NSP in relation to mitigating fatal overdose events [43], but the cost-effectiveness of NSP relative to SSTVI remains uncertain. Additionally, the impact of NSP on the public health burden of SSTVI and the health service utilization patterns for treating SSTVI among PWID is unclear. By addressing these knowledge gaps, we will be able to demonstrate the economic value of NSP as well as whether there are additional benefits of NSP beyond prevention of HIV and HCV. Therefore, in this study, we aim to model the cost-effectiveness of NSP versus not having NSP with a focus on the treatment of SSTVI among PWID.

Methods

Model type and health states

We constructed a 13-state transition microsimulation model to capture patterns of progression of SSTVI and the corresponding health service utilization in a population of PWID (Fig. 1). We chose this individual-based, state-transition model due to its ability to model each individual’s unique clinical pathways, incorporate individual’s history in the occurrence of future events, and handle a relatively large number of health states [44, 45]. To model the transmission of SSTVI, we chose a microsimulation framework as opposed to an agent-based framework, as the proportion of PWID who shared needles with other individuals was low (15%) [46]. The health states were based on disease progression, treatment guidelines, and health service utilization patterns for treating SSTVI (see Supplementary Materials for the list of health states) [14, 47]. We set the time horizon for this model as 5 years with the cycle length of 1 week. We chose this specific cycle length because clinical practice guidelines explicitly state that the duration of antibiotic treatment be 5–7 days [14, 47, 48], and the 7-day period is often used to identify separate episodes of SSTVI or identify complications from an existing episode [9]. Details of the model assumptions are described in Supplementary Materials.

Fig. 1
figure 1

Model schematic for progression of SSTVI in PWID. Keys: H = healthy; np = non-purulent; p = purulent; OP = outpatient; ED = emergency department; IP = inpatient; IPC = inpatient complications; SSTVI = skin, soft tissue, and vascular infections; Trt = treatment. Note: H and H* indicate the same health state with the same health state utilities. For ease of visualization, we created the H* state in the diagram

Full size image

Data sources

To derive the parameters for our model, we used six administrative health databases provided by the Government of Québec (see Supplementary Materials for more detail; Research file publication date: 2009–2019; Data use approval date: 20 September 2021).Footnote 1 Further information regarding these datasets can be found by visiting the Institut de la Statistique du Québec (ISQ) webpage at: https://statistique.quebec.ca/services-recherche/donnees/administratives [49]. All inferences, opinions, and conclusions drawn in this publication are those of the authors, and the Government of Québec is not responsible for the compilations or the interpretation of the results produced using the research files.Footnote 2

To identify values for model parameters that influenced the characteristics of injection practices among PWID, we used the results from surveillance of PWID conducted by Quebec Provincial Public Health Institute (INSPQ) [46]. Finally, we retrieved the health state utility weight associated with varying degrees of severity of SSTVI (e.g., outpatient versus inpatient) from the published literature [50,51,52,53].

Model structure

Figure 1 illustrates the model schematic of how PWID with SSTVI treat their infection. The model schematic and health states informing the model were adopted from the Infectious Disease Society of America’s (IDSA) “Practice Guidelines for the Diagnosis and Management of Skin and Soft Tissue Infections” [14] and the University of California San Francisco (UCSF) Infectious Disease Management Program’s “Guideline for the Management of Suspected Skin and Soft Tissue Infections in Adults” [47]. The model reflects how individuals transition through states of self-treatment or care within the healthcare system until the infection is resolved, they die, or reach the end of the time horizon.

Our model includes two types of SSTVI: (1) purulent and (2) non-purulent (International Classification of Diseases [ICD]-9/10 codes in Supplementary Materials). Purulent SSTVI include abscess, furuncle, carbuncle, and folliculitis. Non-purulent SSTVI include cellulitis, erysipelas, and necrotizing fasciitis. To account for deaths from other causes (e.g., overdose, injury, cancer), we created an absorbing state specifically for ‘other-cause mortality’ (OCM) and incorporated the probability of OCM when we computed the transition probabilities for each health state. We calculated the probability of OCM specific to PWID by calculating the mortality rate (i.e., number of cases per 1000 population) for each age group (e.g., 20–24, 25–29, etc.) using the administrative data.

Epidemiologic features of simulated cohort

We created two synthetic cohorts with 100,000 Quebec adults (age range: 18–65 years) who were either ‘Healthy’ or had SSTVI according to the observed prevalence of SSTVI in the province. We used one cohort to simulate a scenario with NSP, and the other cohort to simulate a scenario without NSP. We created our cohorts based on the distributions of demographic characteristics of PWID and injection risk behaviour characteristics from the INSPQ surveillance of the PWID [46] (Table 1).

Table 1 List of parameters, distributions, and sources
Full size table

During the data generation step, we based the proportion of PWID who share and reuse needles on the probabilities of needle sharing and reusing captured in INSPQ surveillance data [46]. We used the risk ratio estimates corresponding to the effectiveness of NSP on sharing and reusing needles [34] and multiplied the inverse of these risk ratios to derive the probabilities of sharing and reusing needles under the scenario without NSP. Consequently, the proportion of PWID who shared and reused needles was higher without NSP, leading to a higher probability of acquiring SSTVI in the cohort without NSP compared to the cohort with NSP.

To estimate the prevalence of SSTVI and transition probabilities, we used Quebec’s provincial administrative health databases (January 2009-December 2019) with a validated algorithm to identify PWID (sensitivity: 0.85, specificity: 0.80) [54]. We then searched the records of physician visits, ED visits, and hospital admissions in the administrative health databases between January 1, 2009 and March 31, 2019 to identify cases of SSTVI (see Supplementary Materials).

To derive the individual probability of SSTVI, we estimated the proportion of the general population (including individuals who do not use drugs) with SSTVI between January 2009 and March 2019 using the above administrative health databases. We set this value as the “base” probability of SSTVI if the PWID did not inject in the past month, did not share needles, and did not reuse injection equipment. We then included relative risks of SSTVI occurrence associated with injection risk behaviour (e.g., injection frequency, sharing injection equipment, or reusing needles and syringes) [16, 18, 19] and changes in injection risk behaviour associated with the implementation of NSP [34]. For PWID with behavioural risk factors, this “base” probability was multiplied by the “risk multiplier” (i.e., risk ratios or odds ratios from the literature; see Supplementary Materials for more detail).

Cost parameters

We assessed the direct healthcare costs from the public payer perspective (Quebec Ministry of Health and Social Services), which were captured through the provincial public system. We report all costs in 2022 Canadian dollars after adjusting for inflation using Statistics Canada’s Consumer Price Index for Quebec (Health and Personal Care Products) [55]. We used records of physician visits, ED visits, and hospital admissions to derive cost estimates at each stage of SSTVI treatment. Complications during episodes of hospitalization were based on the length of hospital stay and record of procedure codes. We assumed that hospitalization longer than 7 days with a record of surgical procedures constituted complication in inpatient settings, which we designated as the ‘inpatient complication’ health state. To calculate the cost of SSTVI in each treatment setting (e.g., outpatient, ED, and inpatient), we took the average cost of treating the case at each of the settings for each type of SSTVI, which we derived from the administrative data (details described in Supplementary Materials).

Due to the lack of data on the cost of NSP in Quebec, we adjusted the annual cost of NSP from Ontario to account for the difference in population between these two provinces. Since NSP are available to all PWID, we divided the total cost of NSP over the 5-year period by the number of individuals in the simulation and the number of weeks during the same period to derive the weekly average cost of NSP per PWID.

We applied the discount rate of 1.5% for costs and health outcomes (with 0% and 3% as sensitivity analyses) in accordance with the Canadian Agency for Drugs and Technologies in Health’s recommendations [56]. More specific assumptions and algorithms used to create the input parameters are enumerated in the accompanying Supplementary Materials.

Analysis

We conducted a cost-utility analysis of a needle and syringe program, in which we computed the incremental cost-effectiveness ratio (ICER) to derive the cost per quality adjusted life year (QALY). In addition, we calculated the specific healthcare costs for each treatment setting (i.e., location of care) and type of SSTVI to estimate the economic burden in each scenario.

Relatedly, we conducted a recurrent-event survival analysis on data from the microsimulation output (details in Supplementary Materials), in which we estimated the risk of recurrent outpatient visits, emergency department visits, and hospitalizations associated with NSP compared to not having an NSP. We used the marginal means model [57], an extension of the Cox proportional hazards model, to estimate the hazard ratios (HR) and 95% confidence intervals (CI) of the association between NSP and the risk of recurrent use of the healthcare system for treating SSTVI. We chose to run the marginal means model because multiple episodes of the outcome over the course of follow-up from the same individual could be correlated. In addition, the marginal approach provides a population-level estimate of the cumulative hazard, which could then be used to calculate the expected number of events that the individual experienced up to a given time [58]. Where the within-subject correlation is complex and unknown, the marginal means model can account for it in a flexible and unbiased manner (i.e., the robust standard error is valid even when the correlation dependence structure is mis-specified) [59].

To estimate the effect of NSP on the likelihood of SSTVI mortality, we ran competing risk survival analysis. We chose this analytic framework to account for the presence of competing events (e.g., other-cause mortality). We constructed the event-specific cumulative incidence function and ran the Fine-Gray model for competing risk regression [60], as it enables interpretation of hazard ratios similar to a Cox proportional hazards model.

Finally, we conducted a probabilistic sensitivity analysis (PSA) with 1,000 model runs on 10,000 individuals (for each treatment strategy). For parameters with uncertainty (e.g., costs, QALY, and risk ratios), we assigned a statistical distribution (e.g., beta, gamma, or lognormal) in which its mean value was considered the base case scenario. We then computed the quantile-based credible intervals at the 2.5th and 97.5th percentiles to reflect the range of values from the PSA iterations. Based on the PSA outputs, we constructed a PSA scatter plot and fitted the cost-effectiveness acceptability curves to determine the probability of cost-effectiveness of NSP. We derived the model parameters (e.g., transition probabilities and costs) from the administrative data using SAS 9.4. We then conducted model simulation and statistical analyses in R version 4.2.3.

Results

Tables 2 and 3 display the results from analyses of the public health burden of SSTVI. We estimated that NSP had a protective effect against SSTVI mortality, with 788 fewer deaths per 100,000 PWID (No NSP: 3,360 deaths from SSTVI vs. With NSP: 2,572 deaths from SSTVI in simulated cohort of 100,000). This corresponded to hazard ratio of 0.76 (95% CI = 0.72–0.80; Fig. 2). Upon examination of health service use, we estimated that health service utilization over the 5-year period remained lower under the scenario with NSP (outpatient: 66,511 vs. 86,879; ED: 9920 vs. 12,922; inpatient: 4282 vs. 5596). Relatedly, NSP was associated with modest reduction in the relative hazard of recurrent outpatient visits (HR = 0.96; 95% CI = 0.95–0.97) for purulent SSTVI as well as outpatient (HR = 0.88; 95% CI = 0.87–0.88) and ED visits (HR = 0.98; 95% CI = 0.97–0.99) for non-purulent SSTVI. However, the hazard of recurrent inpatient SSTVI (including complications) remained similar across the two cohorts.

Table 2 SSTVI mortality and health service use with and without NSP (95% CI)
Full size table
Table 3 Hazard ratios for SSTVI mortality and health service use with NSP
Full size table
Fig. 2
figure 2

Cumulative incidence function for SSTVI mortality with and without NSP. Note: The results in this figure were compiled using data from the © Government of Québec (Research file publication date: 2009–2019; Data use approval date: 20 September 2021)

Full size image

Table 4 summarizes the results of the cost-utility analysis, with discounted cost and QALY estimates. When 1.5% discount rate was assumed, the incremental cost under the NSP scenario compared to the no NSP scenario was $1,207 with 0.017 QALY gains, which correspond to an ICER of $70,278 per QALY. At 0% and 3% discount rates, we also observed similar increase in costs and QALYs with an ICER of $69,455 per QALY and $71,108 per QALY, respectively. When we assumed different scenarios for the annual cost of NSP per PWID ($300–$400 per person), the ICER ranged from $64,103 per QALY to $92,333 per QALY.

Table 4 Estimates of cost, QALY, and ICER with and without NSP
Full size table

Figures 3 and 4 characterize the results of the PSA. Figure 3 illustrates the cost-effectiveness plane, in which we plotted the incremental cost and incremental effectiveness for each of the 1,000 PSA runs. All 1,000 points lie in the quadrant corresponding to higher effectiveness and higher costs. Figure 4 depicts the cost-effectiveness acceptability curve corresponding to different willingness-to-pay thresholds ranging from $0 to $300,000 per QALY. If the decision maker is willing to pay $100,000 per QALY, the probability of cost-effectiveness associated with NSP is 65.1%. Similarly, if the decision maker is willing to pay $200,000 per QALY, then there is 90.9% probability that NSP remains cost-effective.

Fig. 3
figure 3

Cost-effectiveness plane of incremental costs and incremental QALY from probabilistic sensitivity analysis with 1,000 iterations of the simulation. Note: The results in this figure were compiled using data from the © Government of Québec (Research file publication date: 2009–2019; Data use approval date: 20 September 2021)

Full size image
Fig. 4
figure 4

Cost-effectiveness acceptability curve with the probability of cost-effectiveness of NSP by willingness-to-pay thresholds from probabilistic sensitivity analysis with 1,000 iterations of the simulation. Note: The results in this figure were compiled using data from the © Government of Québec (Research file publication date: 2009–2019; Data use approval date: 20 September 2021)

Full size image

Discussion

To our knowledge, this is the first study to evaluate cost-effectiveness of needle and syringe programs in relation to skin, soft tissue, and vascular infections for people who inject drugs. In our model, we assessed the epidemiologic impact of NSP on the burden of SSTVI in PWID as well as economic implications of SSTVI on the Quebec healthcare system compared to a counterfactual scenario had NSP not been implemented. We estimated that having NSP led to fewer SSTVI deaths over the five-year period compared to not having NSP, with lower relative hazard of SSTVI mortality. Relatedly, under the scenario with NSP, there were fewer contacts with the healthcare system as well as lower hazards of recurrent outpatient visits and ED visits to treat SSTVI.

Our results provide evidence that complements earlier Canadian studies on harm reduction strategies that demonstrated their cost-effectiveness [39,40,41,42, 61]. An earlier economic modelling study of NSP in Ontario concluded that NSP was associated with 359 new HIV infections averted ($106,928 per disability-adjusted life-year averted) [61]. Other studies have focused on the cost-effectiveness of supervised consumption sites, which also distribute needles and injection equipment. These latter studies have concluded cost-effectiveness based on prevention of HIV and HCV, which resulted from reduction in high-risk injection behaviours and increased uptake of safer injection practices [41, 42, 62]. Our findings remain significant because we observed that NSP reduce the epidemiological and economic burdens of SSTVI. In addition to existing evidence of effectiveness of NSP against blood-borne infections, our study highlights complementing benefits of NSP by demonstrating their clinical and cost-effectiveness in addressing bacterial infections.

The relatively high ICER observed should be interpreted with caution because the higher ICER could in large part be due to the lower rate of mortality from SSTVI under the ‘With NSP’ scenario compared to the ‘No NSP’ scenario. Under the former scenario, there were more individuals who were alive. This meant that there were a greater number of individuals who were at risk of health service use as well as those who incurred the NSP costs (i.e., fixed costs needed to operate the programs) at each cycle under the ‘With NSP’ scenario. In other words, by virtue of saving lives, having the NSP resulted in higher direct healthcare costs among those alive, which in turn increased the incremental costs for the healthcare system. In addition, the incremental QALY gains from having NSP remained very small, which then resulted in higher incremental cost per incremental QALY (i.e., ICER).

Despite the relatively high ICER, other important benefits of the NSP include lower recurrent health service use under the ‘With NSP’ scenario. The observed differences in health service utilization patterns could be due to how NSP operate in Quebec. Some of the NSP in the province (also known as ‘injection equipment access centres’) provide access to medical care and treatment (e.g., wound care to treat injection-related wounds) [63]. In addition, NSP in Quebec are often operated by community organizations unaffiliated with the healthcare system [64], which helps establish trust with the PWID population [65]. With high level of trust between clients and NSP staff [66], PWID may utilize NSP as an alternative to physician visits when seeking medical care for SSTVI, and NSP may bridge the gap between medical care and PWID.

In addition to the health benefits to individuals, the lower hazard of recurrent SSTVI associated with NSP has important implications on the Quebec health care system. The lower hazard ratios associated with NSP suggest that they may not only operate as a lower cost alternative to formal medical service but also as a preventive strategy against SSTVI for PWID. However, our results are to be interpreted with caution. The findings of recurrent-event analysis were from a simulated cohort (i.e., synthetic data). The model relied on simplified assumptions of the cohort of PWID, and the model assumptions may not have accurately reflected the complexities of the real-world patterns of SSTVI incidence and recurrence. While this could serve as preliminary evidence on the benefits of NSP against recurrent SSTVI, evidence from real-world data may be needed to strengthen the current findings and to inform decision making.

Other benefits to the healthcare system may include reduction in early departures from healthcare settings (i.e., patient-directed discharge). Owing to a small number of early departures from medical institutions observed in our simulation (i.e., rare outcome), we were only able to see a modest reduction in the number of patient-directed discharges. However, NSP mitigate needle sharing and reuse of injection equipment [34,35,36,37,38], which are factors that elevate the risk of early departures among PWID [67, 68]. Further, community-level treatment and care could reduce the risk of self-directed discharges from hospitals and emergency departments [69]. Many NSP in Quebec operate at the community level outside the healthcare system [64], and NSP coverage in the province is high at 82% [70]. With expanded operation, NSP may contribute to additional reduction in hospital re-admissions that arise from complications and worsened health outcomes following patient-initiated early departures [71].

Strengths

First, our model framework captured the real-world complexities surrounding individuals’ unique clinical pathways for the treatment of SSTVI. Second, we used Quebec administrative health data and extensive public health surveillance on PWID conducted by the INSPQ to inform our model parameters. The administrative databases allowed our model inputs to reflect the health service utilization patterns of PWID and current costs and burdens of SSTVI in Quebec’s healthcare system in each care setting (e.g., outpatient, ED, and inpatient). Relatedly, the INSPQ surveillance data captured the most up-to-date population characteristics of PWID, which are not readily available in traditional administrative databases. Third, in addition to the economic evaluation, we conducted survival analysis that accounts for recurrent events and competing risks, enabling the assessment of both cost-effectiveness and clinical effectiveness of NSP. Thus, our study provides a more comprehensive overview of the comparative effectiveness between the ‘NSP’ and the ‘No NSP’ scenarios.

Limitations

First, our model may not have adequately captured the variabilities in real-world clinical practice, as we assumed that the quality of care delivered, and health system performance remained the same over the course of follow-up for both cohorts. To mitigate this concern, we varied the risk ratios corresponding to injection risk behaviours as well as the cost and utility parameters in our PSA, where the intervention and the control cohorts had different values for these parameters at each iteration of the PSA. Second, due to the lack of availability of NSP costs in Quebec, we used the NSP cost estimates from Ontario. In addition, we were unable to divide NSP costs into fixed cost and variable cost components and how NSP affected each component due to the lack of available information beyond the total cost of operating the NSP. While we were unable to model the effect of NSP on specific cost components, we conducted deterministic sensitivity analyses to mitigate concerns surrounding the NSP costs per PWID in Quebec by re-running the model with five different estimates of NSP cost per PWID per year. Third, due to a lack of Quebec-specific information on the effectiveness of NSP against injection risk behaviours, we relied on the risk ratio estimates derived from epidemiological studies conducted in other parts of the world. This may raise concerns around the validity of model findings due to discrepancies between the sources of the estimates and the Quebec regional contexts. To mitigate these concerns, we calibrated our model probabilistically, which accounted for parameter uncertainties by generating posterior distributions that reflect the range of plausible values for the risk ratio estimates. Fourth, there was uncertainty around the extent to which multiple injection risk behaviours interact and elevate the risk of SSTVI. To account for this uncertainty, we conducted Bayesian model calibration using the sample importance resampling approach, which generated a posterior set of risk ratio estimates based on 659 unique sets of risk ratios (details of model calibration in Supplementary Materials).

Conclusion

Needle and syringe programs are effective in preventing and reducing transmission of SSTVI. Compared to not having NSP, having NSP is associated with reduced hazard of SSTVI mortality as well as lower relative hazard of recurrent contacts with the healthcare system for treating SSTVI. Expansion of NSP services may maximize their benefits and further reduce harms associated with SSTVI in PWID.

Availability of data and materials

The administrative data that support the findings of this study are available only through the CADRISQ site, secure research environment accessible only to accredited researchers in Quebec for research purposes. Therefore, restrictions apply to the availability of these data, and data requests must be made with permission from the Institut de la Statistique du Québec. The R code used in the current study is available in the GitHub repository (URL: https://github.com/jihoon-jay/Microsimulation).

Change history

Notes

  1. Une partie de la compilation est effectuée à partir de données provenant du © Gouvernement du Québec (année de la publication du Fichier de recherche: 2009–2019).

  2. Le Gouvernement du Québec n’est pas responsable des compilations ni de l’interprétation des résultats produits à l’aide du Fichier de recherche.

Abbreviations

CAD:

Canadian dollar

CADRISQ:

Centre d’accès aux données de recherche de l’Institut de la statistique du Québec

CCI:

Canadian Classification of Health Interventions

CI:

Confidence interval

CMG:

Case mix group

DAA:

Discharge against medical advice

DIN:

Drug Identification Number

ED:

Emergency department

HCV:

Hepatitis C virus

HR:

Hazard ratio

I&D:

Incision and drainage

ICD:

International Classification of Diseases

ICER:

Incremental cost-effectiveness ratio

ICU:

Intensive care unit

IDSA:

Infectious Disease Society of America

INC:

Incremental

INSPQ:

Institut national de la santé publique du Québec/Quebec Provincial Public Health Institute

IP:

Inpatient

IPC:

Inpatient complications

IRR:

Incidence rate ratio

ISQ:

Institut de la statistique du Québec

MCSE:

Monte Carlo standard error

NSP:

Needle and syringe programs

OCM:

Other cause mortality

OP:

Outpatient

OR:

Odds ratio

PSA:

Probabilistic sensitivity analysis

PWID:

People who inject drugs

QALY:

Quality adjusted life years

REF:

Reference

SSTVI:

Skin, soft tissue, and vascular infections

TMP-SMX:

Trimethoprim-sulfamethoxazole

UCSF:

University of California San Francisco

References

  1. Mathers BM, et al. Mortality among people who inject drugs: a systematic review and meta-analysis. Bull World Health Organ. 2013;91(2):102–23.

    Article  PubMed  PubMed Central  Google Scholar 

  2. Mathers BM, et al. Global epidemiology of injecting drug use and HIV among people who inject drugs: a systematic review. Lancet. 2008;372(9651):1733–45.

    Article  PubMed  Google Scholar 

  3. Trickey A, et al. The contribution of injection drug use to hepatitis C virus transmission globally, regionally, and at country level: a modelling study. Lancet Gastroenterol Hepatol. 2019;4(6):435–44.

    Article  PubMed  PubMed Central  Google Scholar 

  4. Degenhardt L, et al. Global prevalence of injecting drug use and sociodemographic characteristics and prevalence of HIV, HBV, and HCV in people who inject drugs: a multistage systematic review. Lancet Glob Health. 2017;5(12):e1192–207.

    Article  PubMed  PubMed Central  Google Scholar 

  5. Tookes H, et al. A cost analysis of hospitalizations for infections related to injection drug use at a County Safety-Net Hospital in Miami, Florida. PLoS ONE. 2015;10(6):e0129360.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Kendall CE, et al. A cohort study examining emergency department visits and hospital admissions among people who use drugs in Ottawa, Canada. Harm Reduct J. 2017;14(1):16.

    Article  PubMed  PubMed Central  Google Scholar 

  7. Marks M, et al. Needles and the damage done: reasons for admission and financial costs associated with injecting drug use in a Central London Teaching Hospital. J Infect. 2013;66(1):95–102.

    Article  PubMed  Google Scholar 

  8. Takahashi TA, et al. Predictors of hospitalization for injection drug users seeking care for soft tissue infections. J Gen Intern Med. 2007;22(3):382–8.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Lewer D, Harris M, Hope V. Opiate injection-associated skin, soft tissue, and vascular infections, England, UK, 1997–2016. Emerg Infect Dis. 2017;23(8):1400–3.

    Article  PubMed  PubMed Central  Google Scholar 

  10. Dwyer R, et al. Prevalences and correlates of non-viral injecting-related injuries and diseases in a convenience sample of Australian injecting drug users. Drug Alcohol Depend. 2009;100(1–2):9–16.

    Article  CAS  PubMed  Google Scholar 

  11. Topp L, et al. Prevalence and predictors of injecting-related injury and disease among clients of Australia’s needle and syringe programs. Aust N Z J Public Health. 2008;32(1):34–7.

    Article  PubMed  Google Scholar 

  12. Blome A, et al. Emergency department medication dispensing reduces return visits and admissions. Am J Emerg Med. 2020;38(11):2387–90.

    Article  PubMed  Google Scholar 

  13. Ki V, Rotstein C. Bacterial skin and soft tissue infections in adults: a review of their epidemiology, pathogenesis, diagnosis, treatment and site of care. Can J Infect Dis Med Microbiol. 2008;19(2):173–84.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Stevens DL, et al. Practice guidelines for the diagnosis and management of skin and soft tissue infections: 2014 update by the infectious diseases society of America. Clin Infect Dis. 2014;59(2):147–59.

    Article  PubMed  Google Scholar 

  15. Sartelli M, et al. 2018 WSES/SIS-E consensus conference: recommendations for the management of skin and soft-tissue infections. World J Emerg Surg. 2018;13:58.

    Article  PubMed  PubMed Central  Google Scholar 

  16. Doran J, et al. Factors associated with skin and soft tissue infections among people who inject drugs in the United Kingdom: a comparative examination of data from two surveys. Drug Alcohol Depend. 2020;213:108080.

    Article  PubMed  Google Scholar 

  17. Barocas JA, et al. Impact of medications for opioid use disorder among persons hospitalized for drug use-associated skin and soft tissue infections. Drug Alcohol Depend. 2020;215:108207.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Jawa R, et al. Association of skin infections with sharing of injection drug preparation equipment among people who inject drugs. Int J Drug Policy. 2021;94:103198.

    Article  PubMed  PubMed Central  Google Scholar 

  19. Wright T, et al. Prevalence and severity of abscesses and cellulitis, and their associations with other health outcomes, in a community-based study of people who inject drugs in London, UK. PLoS ONE. 2020;15(7):e0235350.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Mosseler K, et al. Epidemiology, microbiology, and clinical outcomes among patients with intravenous drug use-associated infective endocarditis in New Brunswick. CJC Open. 2020;2(5):379–85.

    Article  PubMed  PubMed Central  Google Scholar 

  21. Gomes T, et al. Trends in hospitalizations for serious infections among people with opioid use disorder in Ontario, Canada. J Addict Med. 2022;16(4):433–9.

    Article  PubMed  Google Scholar 

  22. Kenny KS, et al. Association of illicit fentanyl use with injection risk practices among people who inject drugs. AIDS Behav. 2023;27(6):1757–65.

    Article  PubMed  Google Scholar 

  23. Scheim AI, et al. The Ontario integrated supervised injection services cohort study of people who inject drugs in Toronto, Canada (OiSIS-Toronto): Cohort Profile. J Urban Health. 2021;98(4):538–50.

    Article  PubMed  PubMed Central  Google Scholar 

  24. Bouchard M, et al. Estimating the size of the fentanyl market in British Columbia. 2020.

  25. Biancarelli DL, et al. Strategies used by people who inject drugs to avoid stigma in healthcare settings. Drug Alcohol Depend. 2019;198:80–6.

    Article  PubMed  PubMed Central  Google Scholar 

  26. Ozga JE, et al. A community-based study of abscess self-treatment and barriers to medical care among people who inject drugs in the United States. Health Soc Care Community. 2022;30(5):1798–808.

    Article  PubMed  Google Scholar 

  27. Lloyd-Smith E, et al. Determinants of hospitalization for a cutaneous injection-related infection among injection drug users: a cohort study. BMC Public Health. 2010;10:327.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Hankins CA. Syringe exchange in Canada: good but not enough to stem the HIV tide. Subst Use Misuse. 1998;33(5):1129–46.

    Article  CAS  PubMed  Google Scholar 

  29. CACTUS Montreal. Our History. 2023 [cited 2023 October 2, 2023]; Available from: https://cactusmontreal.org/en/all-about-cactus/history/.

  30. Hankins CA, et al. Le programme d’échange de seringues de CACTUS-Montréal: Description du service et profil comportemental de la clientèle. Revue sexologique. 1993;1:57–75.

    Google Scholar 

  31. Aspinall EJ, et al. Are needle and syringe programmes associated with a reduction in HIV transmission among people who inject drugs: a systematic review and meta-analysis. Int J Epidemiol. 2014;43(1):235–48.

    Article  PubMed  Google Scholar 

  32. MacNeil J, Pauly B. Needle exchange as a safe haven in an unsafe world. Drug Alcohol Rev. 2011;30(1):26–32.

    Article  PubMed  Google Scholar 

  33. Werb D, et al. Patterns of injection drug use cessation during an expansion of syringe exchange services in a Canadian setting. Drug Alcohol Depend. 2013;132(3):535–40.

    Article  PubMed  PubMed Central  Google Scholar 

  34. Bartholomew TS, et al. Reduction in injection risk behaviors after implementation of a syringe services program, Miami, Florida. J Subst Abuse Treat. 2021;127:108344.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  35. Marotta PL, et al. Assessing the relationship between syringe exchange, pharmacy, and street sources of accessing syringes and injection drug use behavior in a pooled nationally representative sample of people who inject drugs in the United States from 2002 to 2019. Harm Reduct J. 2021;18(1):115.

    Article  PubMed  PubMed Central  Google Scholar 

  36. Fernandes RM, et al. Effectiveness of needle and syringe Programmes in people who inject drugs—an overview of systematic reviews. BMC Public Health. 2017;17(1):309.

    Article  PubMed  PubMed Central  Google Scholar 

  37. Sanchez DP, et al. Wounds and skin and soft tissue infections in people who inject drugs and the utility of syringe service programs in their management. Adv Wound Care (New Rochelle). 2021;10(10):571–82.

    Article  PubMed  Google Scholar 

  38. Thakarar K, et al. Injections and infections: understanding syringe service program utilization in a rural state. Harm Reduct J. 2021;18(1):74.

    Article  PubMed  PubMed Central  Google Scholar 

  39. Jozaghi E, Reid AA, Andresen MA. A cost-benefit/cost-effectiveness analysis of proposed supervised injection facilities in Montreal, Canada. Subst Abuse Treat Prev Policy. 2013;8:25.

    Article  PubMed  PubMed Central  Google Scholar 

  40. Enns EA, et al. Potential cost-effectiveness of supervised injection facilities in Toronto and Ottawa, Canada. Addiction. 2016;111(3):475–89.

    Article  PubMed  Google Scholar 

  41. Bayoumi AM, Zaric GS. The cost-effectiveness of Vancouver’s supervised injection facility. CMAJ. 2008;179(11):1143–51.

    Article  PubMed  PubMed Central  Google Scholar 

  42. Andresen MA, Boyd N. A cost-benefit and cost-effectiveness analysis of Vancouver’s supervised injection facility. Int J Drug Policy. 2010;21(1):70–6.

    Article  PubMed  Google Scholar 

  43. Adams JW, et al. Modeling the cost-effectiveness and impact on fatal overdose and initiation of buprenorphine-naloxone treatment at syringe service programs. Addiction. 2022;117(10):2635–48.

    Article  PubMed  PubMed Central  Google Scholar 

  44. Caro JJ, et al. Modeling good research practices–overview: a report of the ISPOR-SMDM Modeling Good Research Practices Task Force–1. Value Health. 2012;15(6):796–803.

    Article  PubMed  Google Scholar 

  45. Krijkamp EM, et al. Microsimulation modeling for health decision sciences using R: a tutorial. Med Decis Making. 2018;38(3):400–22.

    Article  PubMed  PubMed Central  Google Scholar 

  46. Leclerc P, et al. Surveillance des maladies infectieuses chez les personnes qui utilisent des drogues par injection. 2023 [cited 2023 December 28, 2023]; Available from: https://www.inspq.qc.ca/publications/3362.

  47. Mulliken JS, Doernberg SM. UCSF Medical Center Guideline for the Management of Suspected Skin and Soft Tissue Infections in Adults. 2019 [cited 2023 October 10, 2023]; Available from: https://idmp.ucsf.edu/sites/g/files/tkssra4251/f/UCSF%20SSTI%20Guideline%20FINAL_0.pdf.

  48. Barton M, et al. Guidelines for the prevention and management of community-associated methicillin-resistant Staphylococcus aureus: A perspective for Canadian health care practitioners. Can J Infect Dis Med Microbiol. 2006;17(Suppl C):4C-24C.

    Article  PubMed  PubMed Central  Google Scholar 

  49. Institut de la Statistique du Québec. Dictionnaire de données. [cited 2023 December 20, 2023]; Available from: https://statistique.quebec.ca/services-recherche/donnees/administratives.

  50. Chung KC, et al. A cost-utility analysis of amputation versus salvage for Gustilo type IIIB and IIIC open tibial fractures. Plast Reconstr Surg. 2009;124(6):1965–73.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Matza LS, et al. Health state utilities associated with post-surgical Staphylococcus aureus infections. Eur J Health Econ. 2019;20(6):819–27.

    Article  PubMed  PubMed Central  Google Scholar 

  52. Lee BY, et al. Universal methicillin-resistant Staphylococcus aureus (MRSA) surveillance for adults at hospital admission: an economic model and analysis. Infect Control Hosp Epidemiol. 2010;31(6):598–606.

    Article  PubMed  PubMed Central  Google Scholar 

  53. Sertkaya A, et al. Analytical Framework for Examining the Value of Antibacterial Products. 2014 [cited 2023 October 9, 2023]; Available from: https://aspe.hhs.gov/sites/default/files/migrated_legacy_files//44241/rpt_antibacterials.pdf.

  54. Janjua NZ, et al. Identifying injection drug use and estimating population size of people who inject drugs using healthcare administrative datasets. Int J Drug Policy. 2018;55:31–9.

    Article  PubMed  Google Scholar 

  55. Statistics Canada. Table 18-10-0005-01 Consumer Price Index, annual average, not seasonally adjusted. 2023 [cited 2023 October 10, 2023]; Available from: https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=1810000501.

  56. Canadian Agency for Drugs and Technologies in Health, CADTH methods and guidelines: guidelines for the economic evaluation of health technologies: Canada. 2017, Ottawa, Ontario: Canadian Agency for Drugs and Technologies in Health.

  57. Wei LJ, Lin DY, Weissfeld L. Regression analysis of multivariate incomplete failure time data by modeling marginal distributions. J Am Stat Assoc. 1989;84(408):1065–73.

    Article  Google Scholar 

  58. Therneau TM, Grambsch PM. Multiple events per subject. In: Modeling survival data: extending the Cox Model. Springer; 2000. p. 169–229.

    Chapter  Google Scholar 

  59. Amorim LD, Cai J. Modelling recurrent events: a tutorial for analysis in epidemiology. Int J Epidemiol. 2015;44(1):324–33.

    Article  PubMed  Google Scholar 

  60. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc. 1999;94(446):496–509.

    Article  Google Scholar 

  61. Optima Decision Science. Model-based cost-effectiveness and impact assessment of needle-syringe programs in Ontario, Canada from 2006–2015. 2017 [cited 2023 October 16, 2023]; Available from: http://optimamodel.com/pubs/Ontario%20NSP%20evaluation%20report.pdf.

  62. Kerr T, et al. Safer injection facility use and syringe sharing in injection drug users. Lancet. 2005;366(9482):316–8.

    Article  PubMed  Google Scholar 

  63. Gouvernement du Québec. Distribution of new injection equipment. 2019 [cited 2023 November 30, 2023]; Available from: https://www.quebec.ca/en/health/advice-and-prevention/alcohol-drugs-gambling/distribution-of-new-injection-equipment.

  64. A week after opening, Montreal's safe injections sites report success, in CBC News. 2017.

  65. Yeung MEM, et al. Opioid-related emergency department visits and deaths after a harm-reduction intervention: a retrospective observational cohort time series analysis. CMAJ Open. 2023;11(3):E537–45.

    Article  PubMed  PubMed Central  Google Scholar 

  66. Treloar C, et al. Trust and people who inject drugs: the perspectives of clients and staff of Needle Syringe Programs. Int J Drug Policy. 2016;27:138–45.

    Article  PubMed  Google Scholar 

  67. Ti L, et al. Factors associated with leaving hospital against medical advice among people who use illicit drugs in Vancouver, Canada. PLoS ONE. 2015;10(10):e0141594.

    Article  PubMed  PubMed Central  Google Scholar 

  68. McNeil R, et al. Hospitals as a “risk environment”: an ethno-epidemiological study of voluntary and involuntary discharge from hospital against medical advice among people who inject drugs. Soc Sci Med. 2014;105:59–66.

    Article  PubMed  PubMed Central  Google Scholar 

  69. Jafari S, et al. A community care model of intravenous antibiotic therapy for injection drug users with deep tissue infection for “Reduce Leaving Against Medical Advice.” Int J Ment Health Addict. 2015;13(1):49–58.

    Article  PubMed  Google Scholar 

  70. Laniece Delaunay C, et al. Public health interventions, priority populations, and the impact of COVID-19 disruptions on hepatitis C elimination among people who have injected drugs in Montreal (Canada): a Modeling Study. Int J Drug Policy. 2023;116:104026.

    Article  PubMed  PubMed Central  Google Scholar 

  71. Ti L, Ti L. Leaving the hospital against medical advice among people who use illicit drugs: a systematic review. Am J Public Health. 2015;105(12):e53–9.

    Article  PubMed  PubMed Central  Google Scholar 

  72. Hope V, et al. Frequency, factors and costs associated with injection site infections: findings from a national multi-site survey of injecting drug users in England. BMC Infect Dis. 2008;8:120.

    Article  PubMed  PubMed Central  Google Scholar 

  73. Nosyk B, et al. Cost-effectiveness of population-level expansion of highly active antiretroviral treatment for HIV in British Columbia, Canada: a modelling study. Lancet HIV. 2015;2(9):e393-400.

    Article  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

We would like to thank the Institut de la statistique du Québec for data linkage and hosting us at their secured research environment, Centre d’accès aux données de recherche de l’Institut de la statistique du Québec (CADRISQ). We would also like to thank Dr. Jonathon Campbell, PhD and Dr. Alton Russell, PhD for their methodological insights in the model development and analytic approaches, and Dr. Julie Bruneau, MD MSc and Dr. Ahmed Bayoumi, MD MSc for their clinical insights in the model development process.

Funding

This work was supported by the Canadian Institutes of Health Research (CIHR), Project Grant programme, grant number 436351 (principal investigator: DP).

Author information

Authors and Affiliations

  1. Department of Epidemiology, Biostatistics, and Occupational Health, McGill University, 2001 McGill College Avenue, Suite 1200, Montreal, QC, H3A 1G1, Canada

    Jihoon Lim, W. Alton Russell, Mariam El-Sheikh, David L. Buckeridge & Dimitra Panagiotoglou

Authors
  1. Jihoon Lim
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. W. Alton Russell
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Mariam El-Sheikh
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. David L. Buckeridge
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Dimitra Panagiotoglou
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

JL conceived the study, wrote the code to analyze the data, conducted statistical analysis, and drafted and revised the paper. ME-S wrote the code to analyze the data and reviewed and revised the draft paper. DLB reviewed and revised the draft paper. DP acquired the funds to enable this study, reviewed and revised the draft paper, and supervised this project. All authors read and approved the final manuscript.

Corresponding author

Correspondence to Dimitra Panagiotoglou.

Ethics declarations

Ethics approval and consent to participate

The McGill University Research Ethics Board (A08-E53-19B; first approval date: 2019/08/26) and Quebec Access to Information Commission (1026586-S; 2021/09/20) approved the protocol. We rounded all frequencies to the nearest multiple of 5 or 10, as required by the Institut de la Statistique du Québec. Due to the retrospective nature of this study, informed consent was waived.

Consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

Additional information

Publisher's Note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

The original online version of this article was revised: the author would like to include missing co-author W. Alton Russell in the author group and has been updated.

Supplementary Information

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated in a credit line to the data.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Lim, J., Russell, W.A., El-Sheikh, M. et al. Economic evaluation of the effect of needle and syringe programs on skin, soft tissue, and vascular infections in people who inject drugs: a microsimulation modelling approach. Harm Reduct J 21, 126 (2024). https://doi.org/10.1186/s12954-024-01037-3

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12954-024-01037-3

Keywords

Harm Reduction Journal

ISSN: 1477-7517