key: cord-0742029-4wmdqi7i
authors: Djuric, Olivera; Mancuso, Pamela; Zannini, Angela; Nicolaci, Antonio; Massari, Marco; Zerbini, Alessandro; Belloni, Lucia; Collini, Giorgia; Sampaolesi, Fabio; Celotti, Anna; Boni, Iulica; Giorgi Rossi, Paolo
title: Are Individuals with Substance Use Disorders at Higher Risk of SARS-CoV-2 Infection? Population-Based Registry Study in Northern Italy
date: 2021-05-05
journal: Eur Addict Res
DOI: 10.1159/000515101
sha: 4b26618cb2c38bf52e791a38cac991560ab4e379
doc_id: 742029
cord_uid: 4wmdqi7i

BACKGROUND AND AIM: This study assesses whether individuals with substance use disorder are at greater risk of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection than people in the general population. METHODS: A population-based study was conducted including 3,780 individuals, diagnosed with alcohol or other drug dependence and cared for by the addiction service (AS) in the province of Reggio Emilia. Standardised incidence ratios (SIRs) and relative 95% confidence intervals (CIs) of being tested and of being SARS-CoV-2 positive in the population of interest compared with those in the general population of Reggio Emilia were calculated. RESULTS: Both individuals with alcohol and those with other drug use disorders had a lower risk of being SARS-CoV-2 positive (SIR = 0.69; 95% CI 0.32–1.30, SIR = 0.56; 95% CI 0.24–1.10, respectively), despite higher rates of being tested than the general population (SIR = 1.48; 95% CI 1.14–1.89, SIR = 1.51; 95% CI 1.20–1.86, respectively). Among HIV-negative persons, 12.5% were positive to SARS-CoV-2, while none was positive among HIV-positive persons. HCV-infected AS clients had a higher risk of both being tested for SARS-CoV-2 (SIR = 1.99; 95% CI 1.26–2.98) and of resulting positive (SIR = 1.53; 95% CI 0.50–3.58). CONCLUSIONS: Individuals with alcohol and/or other drug use disorders are at higher risk of being tested for SARS-CoV-2 infection but at lower risk of resulting positive than the general population. Further research is warranted in order to support our findings and to address plausible factors underpinning such associations.

were calculated. Results: Both individuals with alcohol and those with other drug use disorders had a lower risk of being SARS-CoV-2 positive (SIR = 0.69; 95% CI 0.32-1.30, SIR = 0.56; 95% CI 0.24-1.10, respectively), despite higher rates of being tested than the general population (SIR = 1.48; 95% CI 1.14-1.89, SIR = 1.51; 95% CI 1.20-1.86, respectively). Among HIVnegative persons, 12.5% were positive to SARS-CoV-2, while none was positive among HIV-positive persons. HCV-infected AS clients had a higher risk of both being tested for SARS-CoV-2 (SIR = 1.99; 95% CI 1.26-2.98) and of resulting positive (SIR = 1.53; 95% CI 0.50-3.58). Conclusions: Individuals with alcohol and/or other drug use disorders are at higher risk of being tested for SARS-CoV-2 infection but at lower risk of resulting positive than the general population. Further research is warranted in order to support our findings and to address plausible factors underpinning such associations.

The outbreak of coronavirus disease 2019 (CO VID19), caused by the novel coronavirus severe acute respiratory syndrome coronavirus 2 (SARSCoV2), has rapidly developed into a pandemic, imposing health, eco nomic, and social burdens of unmeasurable proportions. People with substance use disorders (SUDs), including alcohol, are expected to be at higher risk of SARSCoV2 infection due to the presence of coexisting chronic health conditions, including chronic respiratory disease, immu nosuppression due to HIV infection, or other chronic dis eases, or indirectly as an effect of healthneglecting be haviours, poverty, poorquality housing, high population densities, homelessness, imprisonment, and/or poor mental health. In addition, smoking, which is highly prevalent in the SUD population [1] , may contribute to increasing risk of SARSCoV2 infection and may exac erbate COVID19 severity [2] .

As SUD is associated with numerous cardiorespiratory and metabolic conditions, it has been inferred that people with SUDs are also at increased risk of COVID19 [3] [4] [5] . However, despite rapidly accumulating evidence on CO VID19, the actual risk of the disease in this population has never been directly estimated. The aim of this study was to assess whether individuals with alcohol and/or other drug use disorders are more at risk of SARSCoV2 infection than people in the general population and whether a diagnosis of HIV or HCV further increases this risk.

This longitudinal registrybased cohort study was conducted on a total of 3,780 individuals, residents in the province of Reggio Emilia on December 31, 2019, who were diagnosed with alcohol or other drug use disorders and cared for by the addiction service (AS -in Italian, SerDP) between January 19, 1982 and February 20, 2020. This cohort was followed up to compare the COVID19 cu mulative incidence observed in SUD patients with that observed in the general population during the first phase of the epidemic, that is, from February 27, 2020 to April 12, 2020.

The International Classification of Diseases and Related Health Problems, 10th Revision (ICD10) was used to code SUDs. Pa tients with a diagnosis of mental or behavioural disorders due to the use of alcohol (ICD10 F10), opioids (ICD10 F11), cannabi noids (ICD10 F12), or cocaine (ICD10 F14) were selected from the AS database.

Drugs other than alcohol were classified as opioids, cocaine related, cannabinoids, or other. If a patient was concurrently diag nosed with two or more SUDs, the diagnosis made first was con sidered as the exposure. Alcohol and other drug dependences were classified as follows: if both alcohol and illicit drug abuse were present, the latter was considered, based on the fact that people who are dependent on illicit drugs are more likely to have an alco hol use disorder than people with alcoholism to have a drug use disorder (National Institute on Alcohol Abuse and Alcoholism -NIAAA, 2006).

The AS carries out prevention, treatment, and rehabilitation of use disorders, addiction to legal and illegal psychoactive substanc es, and gambling. It provides care to people with addiction either directly through the local health authority clinics or through pri vate clinics that have entered into an agreement with the National Health Service. Referral to the AS can be voluntary or mandatory, as a complementary penalty by order of the police and judicial au thorities.

The first COVID19 case in Reggio Emilia has been diagnosed on February 27, 2020. Up to April 12, 2020, 5,032 cases were diag nosed, corresponding to a cumulative incidence in the general population of 9.5/1000 inhabitants.

In the province of Reggio Emilia, a surveillance system with dedicated software that collects data on patients tested for SARS CoV2 and, if they tested positive, on the outcome of their disease. The surveillance is fed by several sources: the local health author ity public health service epidemiological investigations, contact tracing, and symptom surveillance of contacts isolated at home, laboratory reports, emergency room and hospital electronic re cords, and death certificates. The system was developed for the purpose of managing each individual case. During the first phase of the epidemic, only symptomatic patients had been tested, with very few exceptions.

The SARSCoV2 database, including all diagnoses from the be ginning of the epidemic up to April 12, 2020, was linked with the registry of SUDs extracted from the AS database. The database re ports the substance userelated diagnoses, the type of care received, the most recent HIV and HCV tests done, and their results. Both the registries are linked with the resident population register to deter mine whether patients are residents of the province of Reggio Emilia.

Age, sex, district of residence standardised incidence ratios (SIRs), and relative 95% confidence intervals (CIs) of being tested and of being SARSCoV2 positive in the population of interest compared with the general population of the province of Reggio Emilia were computed by dividing the number of observed cases in the population of interest by the expected number of cases. The expected number of cases was computed applying age, sex, and districtspecific rates from the general population of province of Reggio Emilia and to the SUD cohort age and sex groups. SIRs were also calculated for the following subcohorts: according to the first diagnosis (alcohol use disorder and drug use disorder), having at least a diagnosis of alcohol use disorder and use of opioids, hav ing a HCVpositive test in one's lifetime, having a negative HCV test, and never tested for HCV. Due to the different distribution of both the prevalence of AS clients and the spread of COVID19 within the province, SIRs were also adjusted for the district of res idence. All analyses were performed by using Stata version 13.0 (College Station, TX, USA).

Eur Addict Res DOI: 10.1159/000515101

Overall, 149 (3.9%) AS patients were tested for SARS CoV2 virus, of whom 17 (11.4%) resulted positive (Ta ble 1). In the group of individuals with alcohol depen dence as first diagnosis, 64 (4.4%) were tested, of whom 9 (14.1%) were positive, while among those with a drug de pendency as first diagnosis, 3.6% were tested, of whom 9.4% were positive.

Among those tested for HCV, there were twice as many individuals who were positive for SARSCoV2 among HCVpositive persons than among HCVnega tive persons [21.7% (5/23) versus 9.7% (10/103)] (Ta ble 2). In the group of HIVnegative persons, 12.5% (11/88) were positive for SARSCoV2 virus, while none of the HIVpositive persons tested positive (0/6).

Both individuals with alcohol and those with drug de pendence as first diagnosis had a lower risk of being SARSCoV2 positive (SIR = 0.69; 95% CI 0.32-1.30, SIR = 0.56; 95% CI 0.24-1.10, respectively), despite the higher rate of testing in these populations than the gen eral population (SIR = 1.48; 95% CI 1.14-1.89, SIR = 1.51; 95% CI 1.20-1.86, respectively). Only HCVinfected AS patients had a higher risk of both being tested for SARS CoV2 (SIR = 1.99; 95% CI 1.26-2.98) and of testing pos itive (SIR = 1.53; 95% CI 0.50-3.58).

Although compatible with a random fluctuation, we observed lower risk of contracting SARSCoV2 virus for both alcohol and for drugdependent individuals, over all as well as for those who were HIVpositive, despite a higher probability of being tested for SARSCoV2. To our knowledge, this is the first study to directly assess the risk of infection in this population, which makes a com parison with previous literature impossible. Some indi rect evidence, based on the distribution of cooccurring conditions in SUD patients (which represent a risk factor for infectious diseases in general), suggests that people who smoke crack cocaine may be more likely to have more severe COVID19 disease [4] . With the limitation of low precision and high risk of bias, the study by Saeei di et al. [6] reported a high rate (10.8%) of opium addic tion in 93 patients admitted to a COVID19 referral hos Tested for SARSCoV2  Positive for SARSCoV2   male  female  total  male  female  total  male  female  total   Overall  3,059  721  3,780  117  32  149  14  3  17  First diagnosis  Alcohol use disorder  1,057  393  1,450  50  14  64  9  0  9  Drug use disorder  2,002  328  2,330  67  18  85  5  3  8  Opioids  977  196  1,173  38  10  48  2  3  5  Cocaine  495  50  545  8  3  11  1  0  1  Cannabis  530  82  612  21  5 pital and 4 times higher the mortality associated with a diagnosis of opium addiction. A reliable interpretation of these results is even more challenging when considering the association between smoking and the risk of severe COVID19. Although peerreviewed studies that have evaluated the risk of SARSCoV2 infection among smokers are lacking, me taanalyses suggest that smokers have a higher risk of se vere COVID19 disease and death [7] . In a prospective study of 500,000 adults, analysis adjusted for important baseline characteristics showed that current smokers were slightly more likely than never smokers to test posi tive for COVID19 [8] .

Our results suggest that the risk of contracting the SARSCoV2 virus is not higher in HIVpositive AS pa tients than it is in the general population, consistently with what Del Amo and colleagues [9] reported in a co hort study including 60 Spanish HIV hospitals, Guo et al. [10] observed in HIV/AIDS patients in two districts in Wuhan, China, and Vizcarra observed in a large cohort study conducted in Madrid after inclusion of suspected COVID19 cases [11] . However, all this should be inter preted with caution since most of these authors did not control their analyses for comorbidities and other con founders.

The strength of our study is that, to our knowledge, it is the first observational study to directly assess the risk of SARSCoV2 infection in a marginalized population of individuals with an SUD. Our results contradict current beliefs and provide a new perspective on the risk factors for SARSCoV2 infection.

The main limitation of this study is the limited gener alisability of the results. While the increased testing in this population could be the result of more intensive medical followups, the protective effect we noted was the result of a mix of behavioural, human, social (social isolation), environmental, and biological factors that reduced expo sure to the virus and the probability of infection and/or of having symptoms. Behaviours and the social environ ment vary between countries and communities. Up to 90% of alcoholdependent people and more than 60% of highrisk opioid clients in Italy belong to the hardto reach population as they do not seek or have not been referred to rehabilitation or drug replacement treatment [12] .

Individuals with alcohol and/or other drug use disor ders are at higher risk of being tested for SARSCoV2 infection but at lower risk of resulting positive than the general population. Further research is warranted in or der to support our findings and to address plausible fac tors underpinning such associations.

The following are members of the Reggio Emilia COVID19 Working Group: Obs, observed number of cases; Exp, expected number of cases; SIR, standardised incidence rate; CI, confidence interval; AS, addiction service; SARSCoV2, severe acute respiratory syndrome coronavirus 2. * Adjusted for age, sex, and district of residence.

Tobacco smoking rates in a National cohort of people with sub stance use disorder receiving treatment

Risk factors of the sever ity of COVID19: a metaanalysis

Collision of the COVID19 and addiction epidemics

Mitigating and learning from the impact of COVID19 infec tion on addictive disorders

Are patients with alcohol use dis orders at increased risk for covid19 infec tion

Opium addiction and COV ID19: truth or false beliefs

World Health Organization. Smoking and COVID19. Scientific brief

Smoking and the risk of COVID19 infection in the UK bio bank prospective study

Incidence and sever ity of COVID19 in HIVpositive persons re ceiving antiretroviral therapy: a cohort study

A survey for COVID19 among HIV/AIDS patients in two districts of Wuhan, China

De scription of COVID19 in HIVinfected indi viduals: a singlecentre, prospective cohort

Alla mani A. Alcohol consumption, alcohol de pendence, and related mortality in Italy in 2004: effects of treatmentbased interventions on alcohol dependence

We thank Jacqueline M. Costa for English language editing.

The study was approved by the Area Vasta Emilia Nord Ethics Committee on 04/07/2020 (protocol number 2020/0045199). Pa tients' informed consent to participate in the study was obtained whenever possible, given the retrospective nature of the study.

The authors have no conflicts of interest to declare.

The study was conducted using exclusively institutional funds of the Azienda Unità Sanitaria localeIRCCS di Reggio Emilia. There was no external funding source for this study.