ORIGINAL RESEARCH published: 18 February 2021 doi: 10.3389/fphar.2021.625610 Long-Term Outcomes of Patients With Cocaine Use Disorder: A 18-years Addiction Cohort Study Arantza Sanvisens 1, Anna Hernández-Rubio 1, Paola Zuluaga 1, Daniel Fuster 1, Esther Papaseit 2, Sara Galan 1, Magí Farré 2 and Robert Muga 1* 1 Department of Internal Medicine, Hospital Universitari Germans Trias i Pujol-IGTP, Universitat Autónoma de Barcelona, Department of Medicine, Badalona, Spain, 2Department of Clinical Pharmacology, Hospital Universitari Germans Trias i Pujol- IGTP, Universitat Autònoma de Barcelona, Badalona, Spain Objective: Cocaine Use Disorder (CUD) has been associated with multiple complications and premature death. The purpose of the present study was to analyze the relationship Edited by: between baseline medical comorbidity and long-term medical outcomes Qi Wang, Southern Medical University, China (i.e., hospitalization, death) in a cohort of patients primarily admitted for detoxification. Reviewed by: In addition, we aimed to analyze cause-specific mortality. Yadong Guo, Central South University, China Methods: longitudinal study in CUD patients admitted for detoxification between 2001 Sun Junhong, and 2018. Substance use characteristics, laboratory parameters and medical comorbidity Shanxi Medical University, China by VACS Index were assessed at admission. Follow-up and health-related outcomes were Keiichi Kadoyama, Himeji Dokkyo University, Japan ascertained through visits and e-health records. Kaplan-Meier and Cox regression models *Correspondence: were used to analyze survival and predictors of hospitalization and death. Robert Muga rmuga.germanstrias@gencat.cat Results: 175 patients (77.7% men) were included. Age at admission was 35 years [IQR: 30–41 years], 59.4% of the patients being intranasal users, 33.5% injectors, and 7.1% Specialty section: smokers. Almost 23% of patients had concomitant alcohol use disorder, 39% were This article was submitted to Neuropharmacology, cannabis users and 9% opiate users. The median VACS Index score on admission was 10 a section of the journal points [IQR: 0–22]. After 12 years [IQR: 8.6–15 years] of follow-up there were 1,292 Frontiers in Pharmacology (80.7%) ED admissions and 308 (19.3%) hospitalizations. The incidence rate of ED Received: 03 November 2020 admission and hospitalization was 18.6 × 100 p-y (95% CI: 15.8–21.8 × 100 p-y). Accepted: 14 January 2021 Published: 18 February 2021 Mortality rate was 1.4 × 100 p-y (95% CI: 0.9–2.0 × 100 p-y) and, baseline Citation: comorbidity predicted hospitalization and mortality: those with VACS Index >40 were Sanvisens A, Hernández-Rubio A, 3.5 times (HR:3.52, 95% CI: 1.19–10.4) more likely to dye with respect to patients with Zuluaga P, Fuster D, Papaseit E, Galan S, Farré M and Muga R (2021) VACS < 20. Long-Term Outcomes of Patients With Cocaine Use Disorder: A 18-years Conclusion: addiction care warrants optimal stratification of medical comorbidity to Addiction Cohort Study. improve health outcomes and survival of CUD patients seeking treatment of the disorder. Front. Pharmacol. 12:625610. doi: 10.3389/fphar.2021.625610 Keywords: cocaine use disorder, VACS index, comorbidity, mortality, hospitalization Frontiers in Pharmacology | www.frontiersin.org 1 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder INTRODUCTION baseline medical comorbidity, use of health resources, and long- term health outcomes among those seeking treatment for CUD. Cocaine is the second most widely used illegal drug in Western Europe after cannabis. According to the European Monitoring Center for Drugs and Drug Addiction (EMCDDA), about four MATERIALS AND METHODS million people aged 15–64 have used cocaine in the last year, and the number of users has increased in recent years (European This was a longitudinal study of patients admitted to the Monitoring Center for Drugs and Drug Addiction (EMCDDA), addiction treatment unit of a tertiary hospital (Germans Trias 2019). i Pujol University Hospital) between January 2001 and May 2018. According to the EDADES population survey on alcohol and The unit admits patients diagnosed with SUD in an area in the drugs, 2% of the general population aged 15–64 in Spain uses cocaine north of Barcelona (Spain) with 400,000 inhabitants. There were (Observatorio Español de las Drogas y las Adicciones, 2019). 837 admissions for addiction treatment between January 2001 Moreover, 54% of people who have used cocaine in the last year and May 2018, of which 195 (23.3%) were due to CUD in 175 have used it in the last month (Observatorio Español de las Drogas y patients. In those who were admitted more than once, only the las Adicciones, 2019). In Catalonia (Spain), 24% of people who seek first admission was analyzed. treatment for substance use disorder (SUD) have a cocaine use The patients came from local primary care centers and from two disorder (CUD), and this percentage has increased in recent years municipal outpatient addiction clinics, one in Badalona (250,000 (Subdirecció General de Drogodependències, 2018). inhabitants) and the other in Santa Coloma de Gramenet (150,000 CUD has been associated with serious systemic complications, inhabitants), both located in the metropolitan area of Barcelona frequent use of healthcare resources (i.e., emergency department (Spain). The main admission criteria for CUD treatment were as (ED) admissions, hospitalization), and premature death follows: failure in outpatient treatment, concurrent dependency on (Degenhardt et al., 2011; Butler et al., 2017). In fact, cocaine use other addictive drugs or alcohol, serious concurrent medical can aggravate inflammatory diseases and alter immune functions problems, severe impairment of psychosocial functioning, lack of that favor the progression of cardiovascular, respiratory, or family and social support, and use of crack or freebase cocaine or infectious diseases (Taylor et al., 2016; Bachi et al., 2017). In intravenous cocaine abuse, among others. addition, it has been communicated that cocaine intoxication All patients received a diagnosis of CUD according to the can cause acute kidney injury, hepatotoxicity, and disseminated Diagnostic and Statistical Manual of Mental Disorders, fourth Ed intravascular coagulation (Vitcheva, 2012; Filho et al., 2019). On the (DSM-IV) and fifth Ed (DSM-5). Due to the transition from other hand, cocaine use has been associated with an increased risk of DSM-IV to DSM-5, not all participants were evaluated under the HIV infection and hepatitis C virus (HCV) infection even in the same DSM. absence of injecting drug use (Macías et al., 2008; Deiss et al., 2012). On admission, data on the use of cocaine and other substances In addition, patients with CUD show a high prevalence of (i.e., alcohol, cannabis, opiates) were collected, including age of onset, psychiatric comorbidities, such as mood disorders (12–62%), route of administration, and duration. DSM criteria were used to anxiety disorders (21–45%) and suicidal tendencies, among diagnose AUD. Cannabis and opiates use was ascertained through others (Vergara-Moragues et al., 2012; Warden et al., 2016). urinalysis at admission. For the purposes of this study, patients were Cocaine use has also been associated with traffic accidents and classified according to the route of cocaine administration as either violence (Giovanardi et al., 2005; Pavarin et al., 2011). intranasal or non-intranasal users (i.e., injectors, smokers). Polysubstance use is common among cocaine users, especially In all patients, blood samples were drawn for biochemical and alcohol consumption, but the concurrent use of marijuana and hematological parameters, and for serologic testing for human opiates is also prevalent, and has been associated with poor health immunodeficiency virus (HIV) infection and hepatitis C virus outcomes (Stinson et al., 2005; Timko et al., 2018). In the US, the (HCV) infection. Anthropometric data (i.e., height and weight) second wave of the National Epidemiologic Survey on Alcohol were obtained as well. and Related Conditions (NESARC) estimated that 79% of people Medical comorbidity on admission was analyzed using the VACS with CUD have a concomitant alcohol use disorder (AUD) Index (Veterans Aging Cohorts Study Index). VACS Index assigns a (Stinson et al., 2005). score based on age and on blood parameters, such as hemoglobin, A recent systematic review and meta-analysis on healthcare platelets, aspartate and alanine aminotransferase levels, creatinine, utilization demonstrated that in SUD patients, hospitalization HIV infection, CD4 lymphocytes, HIV RNA, and HCV infection. and ED admissions are 5 and 7 times more frequent, respectively, The VACS Index ranges from 0 to 164 points, with a higher score compared to the general population (Lewer et al., 2020). In indicating greater comorbidity. The VACS Index has been associated addition, the death rate of patients with CUD ranges from 0.5 with an increased risk of hospitalization and death in patients with to 6.2 × 100 person-years (p-y) and is considered to be 4 to and without HIV infection (Blackstock et al., 2013; Justice et al., 2013; 8 times higher than the death rate of the general population Tate et al., 2013). In HIV-negative patients, HIV RNA is considered (Arendt et al., 2011; Degenhardt et al., 2011). undetectable. Our hypothesis is that the chronicity of CUD is suggestive of the presence of multiple medical complications, which leads to Follow-Up, Comorbidity and Outcomes the excessive use of healthcare resources (i.e., ED visits and The patients were followed up until September 30, 2018 through hospitalization). We aimed to analyze the relationship between in-person visits and review of the ED visits and hospitalization Frontiers in Pharmacology | www.frontiersin.org 2 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder e-health records of the Catalan health department. The diagnoses TABLE 1 | Sociodemographic characteristics, alcohol and substance use, and clinical and blood parameters in 175 patients admitted for the treatment of made during the ED visits and hospitalization were coded CUD in metropolitan Barcelona, Spain. according to the 10th revision of the International Classification of Diseases (ICD-10). The diagnostic coding was N = 175 carried out by two members of the research team (AS and RM) Men, n (%) 136 (77.7) independently; coding discrepancies were resolved by consensus. Age, median [IQR] 35 [30–41] Mortality was analyzed by cross-referencing the data with the BMI, (kg/m2) (n  144), median [IQR] 19.5 [16.4–23.0] National Death Index as of September 30, 2018. The causes of Drug related parameters n (%) death were established by reviewing the medical history. Age at starting cocaine use, median [IQR] 22 [18–26] Route of cocaine administration (n  170) Injected 57 (33.5) Ethics Intranasal 101 (59.4) All patients gave written informed consent, and the study design Smoked 12 (7.1) was approved by the Ethics Committee of the Germans Trias i Duration of CUD (months) (n  164), median [IQR] 111 [36–180] AUD 40 (22.9) Pujol University Hospital (approval number PI-13-082). The Urine screening at admission methods were in compliance with the ethical standards for Opiates (n  158) 14 (8.9) medical research and the principles of good clinical practice in Cannabis (n  160) 63 (39.4) accordance with the World Medical Association’s Declaration of Antecedent of injection drug use (n  172) 82 (47.7) Helsinki. Hematology parameters Median [IQR] Platelet count (x109L) (n  174) 219 [184–270] Hemoglobin (g/dL) (n  174) 14.2 [12.9–15.3] Statistical Analysis Biochemistry Median [IQR] Descriptive statistics were expressed as median and interquartile Creatinine (mg (dL) (n  173) 0.9 [0.78–0.99] range (IQR) for quantitative variables and as absolute frequencies Glomerular fíltration rate (n  173) 97.6 [83.9–109] and percentages for qualitative variables. Aspartate aminotransferase (U/L) 24 [18–40] Alanine aminotransferase (U/L) (n  173) 25 [16–46] We used the chi-square test to detect significant differences in Comorbidity n (%) qualitative variables and t-Student test for differences in quantitative HIV Infection (n  174) 43 (24.7) variables. The Kruskal-Wallis equality-of-populations rank test and CD4+ T cell (x109L), median [IQR] 457 [214–730] Mann-Whitney U test were used to analyze differences in the Viral load (copies/mL), median [IQR] 1,000 [49–22,000] distribution of episodes during follow-up. HCV Infection (n  173) 82 (47.4) FIB-4 (n  173), median [IQR] 0.76 [0.54–1.28] Rates were calculated in p-y by dividing the number of observed Advanced liver fibrosis (>3.25) 8 (4.6) events during the study period by the sum of all individual follow- VACS index (n  173), median [IQR] 10 [0–22] up times. The survival estimates were analyzed using the Kaplan- <20 123 (71.1) Meier curves. Cox regression models were used to analyze the risk 20–39 38 (22.0) >40 factors of first hospitalization after discharge and mortality. The sex, 12 (6.9) variables related to substance use, and the VACS Index score were AUD, alcohol use disorder; BMI, body mass index; CUD, cocaine use disorder; HCV, included in the analysis. All covariates that were statistically hepatitis C virus; HIV, human immunodeficiency virus; IQR, interquartile range. significant in the univariate analysis were included in the multivariate analysis. Prior to implementing the statistical models, we checked the proportional hazard assumptions using The laboratory test results for hemoglobin, total cholesterol, tests and graphs based on the Schoenfeld residuals. and gamma-glutamyl transferase were 14.2 g/dl [IQR: p-values <0.05 were considered statistically significant. 12.9–15.3 g/dl], 170 mg/dl [IQR: 147–201 mg/dl], and 28 U/L Statistical analysis was performed using Stata software (version [IQR: 17–62 U/L], respectively. Moreover, 24.7% of the 11.1, College Station, Texas, United States). patients were HIV-positive and 47.4% were anti-HCV positive. The body mass index (BMI) was 19.5 kg/m2 [IQR: 16.4–23.0 kg/m2]. The median VACS Index score on RESULTS admission was 10 points [IQR: 0–22; range 0–69]. Table 1 shows the sociodemographic characteristics, alcohol and Patient Characteristics at Baseline substance use, and clinical and blood parameters. The study included 175 patients (77.7% men) aged 35 years [IQR: 30–41 years]. The age of onset of cocaine use was 22 years [IQR: 18–26 years], with 59.4% of the patients being intranasal users, 33.5% Follow-Up, ED Visits, Hospitalizations and injectors, and 7.1% smokers. Moreover, 22.9% of the patients had Comorbidity concomitant AUD. According to the screening for illegal drugs in Table 2 shows the socio-demographics, substance use characteristics, urine samples, 39.4% were cannabis users and 9% opiate users. and clinical and blood parameters of CUD patients that were Overall, 58.1% of the patients used alcohol or other substances in hospitalized during follow-up. Hospitalization was significantly addition to cocaine. The most common combinations of substances more frequent in women (p  0.049), in those with antecedent of were cocaine and cannabis (27.5%), cocaine and alcohol (14.4%), and injecting drug use (p  0.028), and in HIV-positive and HCV-positive cocaine, alcohol, and cannabis (7.5%). patients (p  0.001 and p  0.047, respectively). Frontiers in Pharmacology | www.frontiersin.org 3 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder TABLE 2 | Baseline characteristics of CUD patients according to hospitalization and death during follow-up. Hospitalization Death Yes No p value Yes No p value N = 97 N = 78 N = 27 N = 148 Men, n (%) 70 (72.2) 66 (84.6) 0.049 21 (77.8) 115 (77.7) 0.993 Age, median [IQR] 36 [31–41] 34.5 [30–40] 0.392 37 [32–42] 34.5 [30–40.5] 0.323 BMI, (kg/m2) (n  144), median [IQR] 18.5 [15.6–22.0] 21.0 [17.6–23.4] 0.030 19.4 [14.6–22.7] 19.6 [16.7–23.1] 0.334 Drug related parameters Age at starting cocaine use, median [IQR] 21 [17–26] 19.5 [18–25] 0.484 22 [18–24] 20 [17–26] 0.963 Route of cocaine administration (n  170) Injected 33 (35.5) 24 (31.2) 0.509 17 (63.0) 40 (28.0) 0.001 Intranasal 52 (55.9) 49 (63.6) 10 (37.0) 91 (63.6) Smoked 8 (8.6) 4 (5.2) 0 (0) 12 (8.4) Duration of CUD (months) (n  164), median [IQR] 120 [42–180] 96 [36–180] 0.968 96 [36–204] 114 [48–180] 0.604 AUD 25 (25.8) 15 (19.2) 0.306 4 (14.8) 36 (24.3) 0.279 Urine screening at admission Opiates (n  158) 8 (9.1) 6 (8.6) 0.909 3 (13.0) 11 (8.1) 0.445 Cannabis (n  160) 34 (37.8) 29 (41.4) 0.639 12 (50.0) 51 (37.5) 0.248 Antecedent of injection drug use (n  172) 52 (55.3) 30 (38.5) 0.028 21 (80.8) 61 (41.8) <0.001 Hematology parameters Platelet count (×109L) (n  174) 226 [189–279] 210 [181–262] 0.148 193 [139–233] 224 [191–272] 0.005 Hemoglobin (g/dL) (n  174) 14.2 [13–15.2] 14.1 [12.9–15.3] 0.861 14.7 [12.6–15.1] 14.1 [13.0–15.3] 0.973 Biochemistry Creatinine (mg (dL) (n  173) 0.89 [0.78–0.99] 0.9 [0.78–0.97] 0.876 0.9 [0.74–1.01] 0.89 [0.79–0.99] 0.955 Glomerular fíltrate rate (n  173) 96.5 [82.8–107.5] 98.3 [85.7–112.4] 0.248 97.7 [83.7–112.4] 97.6 [84.1–108.4] 0.997 Aspartate aminotransferase (U/L) 25 [19–39] 23 [17–41] 0.429 38 [22–73] 23 [18–36.5] 0.006 Alanine aminotransferase (U/L) (n  173) 25.5 [16–43.5] 22 [15–51] 0.877 42 [21–65] 23 [15–42] 0.109 Comorbidity HIV Infection (n  174) 33 (34.4) 10 (12.8) 0.001 14 (51.8) 29 (19.7) <0.001 CD4+ T cell (x109L), median [IQR] 486 [242–771] 382 [202–562] 0.314 364 [202–457] 562 [242–792] 0.097 Viral load (copies/mL), median [IQR] 2,700 [49–22,000] 337 [49–8,100] 0.325 4,050 [49–15,000] 570 [49–22,000] 0.725 HCV Infection (n  173) 52 (54.2) 30 (39.0) 0.047 21 (77.8) 61 (41.8) 0.001 FIB-4, median [IQR] (n  173) 0.76 [0.54–1.30] 0.78 [0.53–1.26] 0.732 1.5 [0.8–2.2] 0.7 [0.5–1.1] <0.001 Advanced liver fibrosis (>3.25) 3 (3.1) 5 (6.5) 0.294 4 (14.8) 4 (2.7) 0.006 VACS index, median [IQR] (n  173) 10 [0–27] 10 [0–15] 0.110 21 [11–33] 10 [0–19] 0.001 <20 64 (66.7) 59 (76.6) 0.349 13 (48.1) 110 (75.3) 0.005 20–39 24 (25.0) 14 (18.2) 9 (33.3) 29 (19.9) >40 8 (8.3) 4 (5.2) 5 (18.5) 7 (4.8) AUD, alcohol use disorder; BMI, body mass index; CUD, cocaine use disorder; HCV, hepatitis C virus; HIV, human immunodeficiency virus; IQR, interquartile range. The median follow-up time was 12.1 years [IQR: 8.6–15.1 years] In terms of ED admissions, 19% were related to trauma/ with a total time of 1,973.2 p-y. At the end of the study, there were injuries (i.e., fractures, contusions, wounds), 19% to non- 1,292 (80.7%) ED admissions and 308 (19.3%) hospitalizations. The specific/unclassified symptoms, 10% to substance use, and median number of ED admissions and hospitalizations per patient 8.5% to mental disorder. were 5 [IQR: 1–10] and 1 [IQR: 0–2], respectively. The vast majority Regarding hospitalization, almost 40% of the episodes were (85.1%) of the patients presented at least one episode of ED related to mental health, 11% to the liver/digestive system, and admission or hospitalization. Statistically significant differences 10.4% to respiratory conditions (i.e., pneumonia). Figure 4 shows were observed in the distribution of episodes (either ED the distribution of ED admission and hospitalization episodes admission or hospitalization) according to gender and HIV according to the main diagnosis. status; specifically, women (z  −2.704, p  0.007) and HIV- Table 3 shows the risk factors for hospitalization. Specifically, positive patients (z  −2.291, p  0.022) had a greater number being women (hazard ratio (HR): 1.61, 95% CI: 1.02–2.58), of episodes (Figure 1). The probability of having an ED admission presenting concomitant AUD (HR: 1.42, 95% CI: 1.18–1.71), and or hospitalization was 50% after 2.5 years (95% CI: 1.9–3.7 years) having VACS Index >40 (HR: 2.59, 95% CI: 1.13–5.94) were (Figure 2A). significantly associated with a higher probability of hospitalization. The incidence rate of ED admission or hospitalization was 18.6 × 100 p-y (95% CI: 15.8–21.8 × 100 p-y), which was significantly Mortality and Causes of Death higher in women (rate ratio (RR): 1.8, 95% CI: 1.22–2.60, Of the patients included in the study, 15.4% (n  27) died during p  0.002) and in patients with concomitant AUD (RR: 2.0, the follow-up and the death rate was 1.4 × 100 p-y (95% CI: 95% CI: 1.3–2.9, p < 0.001). Figure 3A shows the incidence of 0.9–2.0 × 100 p-y). Table 2 shows the differences between those episodes according to baseline medical comorbidity. who died and those who survived; all-cause mortality was Frontiers in Pharmacology | www.frontiersin.org 4 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder FIGURE 1 | ED admissions or hospitalizations by (A) sex and (B) HIV infection status in a cohort of 175 patients admitted for treatment of CUD in metropolitan Barcelona, Spain. FIGURE 2 | Kaplan-Meier estimates (95% confidence intervals) for (A) ED visits or hospitalization episodes of 175 CUD patients and for (B) survival after being admitted for detoxification in metropolitan Barcelona, Spain. FIGURE 3 | Comorbidity (VACS Index)-adjusted incidence rates (95% confidence intervals) of (A) ED admission or hospitalization episodes and (B) mortality in a cohort of 175 patients seeking treatment of CUD in metropolitan Barcelona, Spain. significantly associated with current or past injecting drug use Figure 2B shows the estimate (Kaplan-Meier) of survival after (p  0.001 and p < 0.001, respectively), HIV infection (p < 0.001), admission for CUD treatment. The multivariate analysis showed that HCV infection (p < 0.001), advanced liver fibrosis (p  0.006) and medical comorbidity was the only predictor of death; patients with higher scores in VACS Index (p  0.005). VACS >40 showed 3.5 times greater probability of death (HR 3.52, Frontiers in Pharmacology | www.frontiersin.org 5 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder FIGURE 4 | Distribution of 1,292 ED admissions and 308 hospitalizations according to ICD-10 diagnostic codes. TABLE 3 | Cox regression model for predictors of hospitalization and death in a cohort of 175 patients admitted for treatment of CUD in metropolitan Barcelona, Spain. Hospitalitzation Death Unadjusted Multivariate Unadjusted Multivariate HR (95% CI) HR (95% CI) HR (95% CI) HR (95% CI) Women 1.89 (1.21–2.96) 1.61 (1.01–2.58) 1.07 (0.43–2.65) Age at starting cocaine use: 5 years increase 1.15 (0.93–1.43) 0.93 (0.57–1.52) Non-intranasal cocaine 0.94 (0.62–1.41) 2.30 (1.05–5.03) 1.64 (0.69–3.92) Duration of CUD 0.99 (0.99–1.00) 1.00 (0.99–1.01) AUD 1.23 (1.05–1.44) 1.27 (1.08–1.48) 0.96 (0.67–1.38) Urine screening at admission Opiates 0.72 (0.35–1.50) 1.28 (0.38–4.32) Cannabis 0.84 (0.55–1.29) 1.49 (0.67–3.32) VACS index <20 1 1 1 1 20–39 1.30 (0.81–2.08) 1.24 (0.77–2.02) 2.39 (1.02–5.59) 1.86 (0.73–4.72) >40 2.63 (1.25–5.52) 2.64 (1.20–5.83) 4.35 (1.55–12.2) 3.52 (1.19–10.4) Statistically significant differences are in bold. AUD, alcohol use disorder; CUD, cocaine use disorder; CI, confidence interval; HR: hazard ratio. 95% CI: 1.19–10.4) compared to patients with VACS Index <20 hospitalization (Tate et al., 2013; Barakat et al., 2015; Hotton (Table 3). Figure 3B shows death rates according to VACS Index. et al., 2017). To the best of our knowledge, this is the first time that The cause of death was determined in 77.7% (21/27) of cases; VACS Index was analyzed in a cohort of HIV-positive and HIV- the main causes were drug-related in 40% (n  8; 5 overdoses, 3 negative patients with CUD. Our results support the use of unattended deaths in the context of current drug use), cancer in medical comorbidity rates in patients with SUD who start 19% (n  4), infectious diseases in 9.5% (n  2), cardiovascular treatment, although more studies are required to confirm these diseases in 9.5% (n  2), and liver cirrhosis in 9.5% (n  2). findings. In this study, VACS Index at baseline reflected moderate organ system damage, even though 47 and 25% of the patients had HCV DISCUSSION and HIV infection, respectively. Despite moderate comorbidity, those with a VACS Index score over 40 were up to 2.6 times more This longitudinal study in patients with CUD who were followed likely to require hospitalization than those with a VACS score for 12 years confirms the prognostic value of a comorbidity index under 20. On the other hand, women with CUD and AUD were at in predicting the risk of hospitalization and death in patients a higher risk of hospitalization. Concomitant AUD is frequent in seeking treatment for the disorder. VACS is a multiorgan system CUD and has been described as an indicator of poor health injury index validated in 2013 for HIV-positive patients, although outcomes (Timko et al., 2018). Specifically, cocaethylene, the it has been described as a reliable index for HIV-negative patients metabolite resulting from the concomitant use of alcohol and as well and as a predictor of health outcomes such as cocaine, has a known toxicity (Jones, 2019). This potent stimulant Frontiers in Pharmacology | www.frontiersin.org 6 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder is more toxic than cocaine itself and has a longer half-life. On the In terms of digestive/liver comorbidity, our findings are other hand, the increased risk of hospitalization for women with consistent with those observed in other studies (Pavarin et al., CUD requires an accurate evaluation of the continuum of care 2011). Liver decompensation was another frequent reason for and care coordination after discharge. clinical attention; however, a study in patients coinfected with It was interesting to confirm that medical comorbidity was the HIV and HCV was unable to demonstrate an association between only predictor of death in this cohort with a high prevalence of cocaine/crack use and evolution of liver fibrosis (Martel- polysubstance use. Some studies on CUD indicate that the risk of Laferrière et al., 2017). Therefore, it is likely that alcohol abuse death is higher in men, in those with a history of injected drugs, in in the patients could explain those findings. those with an early onset of use, in those who drink alcohol, or in Cardiovascular complications in this long-term followed-up those with psychiatric comorbidity (Arendt et al., 2011; de la cohort were less frequent than expected, despite the extensive Fuente et al., 2014). However, there are hardly any studies on the scientific literature on CUD and acute coronary syndrome (Lippi medical comorbidity of CUD other than HIV infection and HCV et al., 2010; Carrillo et al., 2011). However, the results are infection. VACS Index analyzes kidney and liver function in consistent with those reported in other cohorts with low addition to age, hemoglobin and HIV and HCV infections, thus frequency of coronary ischemic complications (Qureshi et al., reflecting the general health status. The recent 2019 version 2014). includes albumin, blood cell count, and BMI (Tate et al., This study has several limitations that should be mentioned. 2019), which may improve the prognostic value of VACS. First, socioeconomic and baseline psychiatric comorbidity It was interesting to note that BMI was below 19.5 kg/m2 in data were not available, which could have facilitated the 50% of the patients in this study and below 16.4 kg/m2 in 25%, interpretation of some findings during follow-up. Second, which indicates that being underweight might be associated with temporary changes in cocaine use (i.e., remission or CUD or an underlying disease. A population study in adults aged exacerbation of use) were not analyzed. In this sense, 18–45 years demonstrated that the BMI of cocaine users is lower retention in care is critical for achieving remission of CUD. than that of non-users, although those BMI values were much The high rate of ED admissions of the patients in this study and higher than those observed in this study (Qureshi et al., 2014). the diagnoses of episodes related to continued substance use On the other hand, death rates in this hospital-based cohort suggest low treatment retention. Third, patients from this were higher than that reported by another Spanish study of study were evaluated with different DSM versions; however, patients recruited in outpatient clinics (de la Fuente et al., criteria for admission were similar throughout the study 2014), although clearly lower than that reported by our group period and mainly related to the severity of CUD. Fourth, in past decades (Sanvisens et al., 2014). this study had a limited number of patients, which impairs the ED admissions for accidents/injuries and non-specific interpretation of some associations due to lack of statistical symptoms were the most frequent during follow-up, power (i.e., route of cocaine administration). In addition, this suggesting that they could be related to continued substance study was carried out in a single unit which limits the use or complications derived from such use. The first systematic generalization of the findings. review and meta-analysis on healthcare utilization in patients In contrast, the strength of this study among patients seeking with SUD was published in 2019 and shows that hospitalization treatment for CUD highlights the challenges in measuring and ED admissions are 5 and 7 times more frequent, respectively, medical comorbidity with an index that has proven to be in this group than in the general population (Lewer et al., 2020). useful in the context of SUD. Most studies on cocaine-related In Spain, 38.4% of drug-related ED admissions can be attributed morbidity are conducted in EDs with patients with acute to cocaine (Miró et al., 2019). Furthermore, a study reveals that intoxication (Arendt et al., 2011; Qureshi et al., 2014; Miró 18% of those admitted to an ED for cocaine use are readmitted in et al., 2019; Santurtún et al., 2020), which prevents an accurate the following year (Miro et al., 2010). clinical assessment of comorbidity. Finally, understanding the Mental health-related complications, accidents/injuries, risk factors for mortality allows us to target preventive respiratory/lung related conditions, and digestive/liver diseases interventions to increase retention in care among those were other diagnoses frequently observed during follow-up. seeking treatment for the disorder. About 40% of the episodes were related to SUD either due to an associated mental illness or to trauma/accidents associated with substance use. These results are consistent with those DATA AVAILABILITY STATEMENT presented in individuals who use illicit drugs (Aitken et al., 2013; Kendall et al., 2017). The pulmonary complications The datasets generated for this study are available on request to resulting from cocaine use could be due to several reasons, the corresponding author. although the route of administration is relevant (Mégarbane and Chevillard, 2013); cocaine’s respiratory toxicity can be immediate (i.e., acute lung injury or hypersensitivity reaction) ETHICS STATEMENT or delayed (i.e., chronic obstructive pulmonary disease, cancer) (Mégarbane and Chevillard, 2013). A recent study shows that The studies involving human participants were reviewed and cocaine users are at increased risk for pulmonary hypertension approved by the Ethics Committee of the Germans Trias i Pujol (Alzghoul et al., 2020). University Hospital (approval number PI-13-082). The patients/ Frontiers in Pharmacology | www.frontiersin.org 7 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder participants provided their written informed consent to FUNDING participate in this study. This work was funded by the Ministry of Science, Innovation and Universities, Carlos III Health Institute (ISCIII), Spain and AUTHOR CONTRIBUTIONS European Fund for Regional Development (FEDER), (grant numbers RD16/0017/0003, PI17/00174, PI20/00883); the AS and RM designed the study and wrote the first draft of the Ministry of Health, Social Services and Equality, National Plan manuscript. AHR and SG reviewed e-health records and on Drugs (PNSD), Spain (grant number 2018/020 and 2020/024); managed the databases. AS managed the literature searches the Agency for Management of University and Research Grants, and statistical analysis. RM, PZ and DF recruited the study Government of Catalonia (grant number 2017SGR316). DF is population and took care of patients. AS and RM reviewed the attached to the Research Intensification Program of the Carlos III literature and made contributions to the interpretation of data. Health Institute (grant number INT19/00026). PZ is attached to AS, EP, MF, and RM contributed to the discussion section. All the the Joan Rodés Program of the Carlos III Health Institute (grant authors revised and approved the final manuscript. number JR20/00016). acute kidney injury associated with cocaine use: report of three cases. BMC REFERENCES Nephrol. 20, 99. doi:10.1186/s12882-019-1279-0 Giovanardi, D., Castellana, C. N., Pisa, S., Poppi, B., Pinetti, D., Bertolini, A., et al. Aitken, C., Kerr, T., Hickman, M., Stoové, M., Higgs, P., and Dietze, P. (2013). A (2005). Prevalence of abuse of alcohol and other drugs among injured drivers cross-sectional study of emergency department visits by people who inject presenting to the emergency department of the University Hospital of Modena, drugs. Emerg. Med. J. 30, 421–422. doi:10.1136/emermed-2012-201170 Italy. Drug Alcohol Depend. 80, 135–138. doi:10.1016/j.drugalcdep.2005.04.010 Alzghoul, B. N., Abualsuod, A., Alqam, B., Innabi, A., Palagiri, D. R., Gheith, Z., Hotton, A. L., Weber, K. M., Hershow, R. C., Anastos, K., Bacchetti, P., Golub, E. T., et al. (2020). Cocaine use and pulmonary hypertension. Am. J. Cardiol. 125, et al. (2017). Prevalence and predictors of hospitalizations among HIV-infected 282–288. doi:10.1016/j.amjcard.2019.10.008 and at-risk HIV-uninfected women. J. Acquir. Immune Defic. Syndr. 75, Arendt, M., Munk-Jørgensen, P., Sher, L., and Jensen, S. O. (2011). Mortality e27–e35. doi:10.1097/QAI.0000000000001278 among individuals with cannabis, cocaine, amphetamine, MDMA, and opioid Jones, A. W. (2019). Forensic drug profile: cocaethylene. J. Anal. Toxicol. 43, use disorders: a nationwide follow-up study of Danish substance users in 155–160. doi:10.1093/JAT/BKZ007 treatment. Drug Alcohol Depend. 114, 134–139. doi:10.1016/j.drugalcdep.2010. Justice, A. C., Modur, S. P., Tate, J. P., Althoff, K. N., Jacobson, L. P., Gebo, K. A., 09.013 et al. (2013). Predictive accuracy of the veterans aging cohort study index for Bachi, K., Mani, V., Jeyachandran, D., Fayad, Z. A., Goldstein, R. Z., and Alia-Klein, mortality with HIV infection: a north american cross cohort analysis. J. Acquir. N. (2017). Vascular disease in cocaine addiction. Atherosclerosis 262, 154–162. Immune Defic. Syndr. 62, 149–163. doi:10.1097/QAI.0b013e31827df36c doi:10.1016/j.atherosclerosis.2017.03.019 Kendall, C. E., Boucher, L. M., Mark, A. E., Martin, A., Marshall, Z., Boyd, R., et al. Barakat, L. A., Juthani-Mehta, M., Allore, H., Trentalange, M., Tate, J., Rimland, D., (2017). Erratum to: a cohort study examining emergency department visits and et al. (2015). Comparing clinical outcomes in HIV-infected and uninfected hospital admissions among people who use drugs in Ottawa, Canada. Harm older men hospitalized with community-acquired pneumonia. HIV Med. 16, Reduct. J. 14, 42. doi:10.1186/s12954-017-0169-7 421–430. doi:10.1111/hiv.12244 Lewer, D., Freer, J., King, E., Larney, S., Degenhardt, L., Tweed, E. J., et al. Blackstock, O. J., Tate, J. P., Akgün, K. M., Crystal, S., Duggal, M., Edelman, E. J., (2020). Frequency of health-care utilization by adults who use illicit drugs: et al. (2013). Sex disparities in overall burden of disease among hiv-infected a systematic review and meta-analysis. Addiction 115, 1011–1023. doi:10. individuals in the veterans affairs healthcare system. J. Gen. Intern. Med. 28 1111/add.14892 Suppl 2, S577. doi:10.1007/s11606-013-2346-z Lippi, G., Plebani, M., and Cervellin, G. (2010). Cocaine in acute myocardial Butler, A. J., Rehm, J., and Fischer, B. (2017). Health outcomes associated with infarction. Adv. Clin. Chem. 51, 53–70. doi:10.1016/S0065-2423(10)51003-5 crack-cocaine use: systematic review and meta-analyses. Drug Alcohol Depend. Macías, J., Palacios, R. B., Claro, E., Vargas, J., Vergara, S., Mira, J. A., et al. (2008). 180, 401–416. doi:10.1016/j.drugalcdep.2017.08.036 High prevalence of hepatitis C virus infection among noninjecting drug users: Carrillo, X., Curós, A., Muga, R., Serra, J., Sanvisens, A., and Bayes-Genis, A. association with sharing the inhalation implements of crack. Liver Int. 28, (2011). Acute coronary syndrome and cocaine use: 8-year prevalence and 781–786. doi:10.1111/j.1478-3231.2008.01688.x inhospital outcomes. Eur. Heart J. 32, 1244–1250. doi:10.1016/j.ycar.2012.01. Martel-Laferrière, V., Nitulescu, R., Cox, J., Cooper, C., Tyndall, M., Rouleau, D., 08210.1093/eurheartj/ehq504 et al. (2017). Cocaine/crack use is not associated with fibrosis progression de la Fuente, L., Molist, G., Espelt, A., Barrio, G., Guitart, A., Bravo, M. J., et al. measured by AST-to-Platelet Ratio Index in HIV-HCV co-infected patients: a (2014). Mortality risk factors and excess mortality in a cohort of cocaine users cohort study. BMC Infect. Dis. 17, 80. doi:10.1186/s12879-017-2196-0 admitted to drug treatment in Spain. J. Subst. Abuse Treat. 46, 219–226. doi:10. Mégarbane, B., and Chevillard, L. (2013). The large spectrum of pulmonary 1016/j.jsat.2013.07.001 complications following illicit drug use: features and mechanisms. Chem. Degenhardt, L., Singleton, J., Calabria, B., McLaren, J., Kerr, T., Mehta, S., et al. Biol. Interact. 206, 444–451. doi:10.1016/j.cbi.2013.10.011 (2011). Mortality among cocaine users: a systematic review of cohort studies. Miró, Ò., Dargan, P. I., Wood, D. M., Dines, A. M., Yates, C., Heyerdahl, F., et al. Drug Alcohol Depend. 113, 88–95. doi:10.1016/j.drugalcdep.2010.07.026 (2019). Epidemiology, clinical features and management of patients presenting Deiss, R. G., Lozada, R. M., Burgos, J. L., Strathdee, S. A., Gallardo, M., Cuevas, J., to European emergency departments with acute cocaine toxicity: comparison et al. (2012). HIV prevalence and sexual risk behaviour among non-injection between powder cocaine and crack cocaine cases. Clin. Toxicol. 57, 718–726. drug users in Tijuana, Mexico. Global Publ. Health 7, 175–183. doi:10.1080/ doi:10.1080/15563650.2018.1549735 17441692.2010.549141 Miro, O., Galicia, M., Sanchez, M., and Nogué, S. (2010). Factores que determinan la European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) (2019). reconsulta a urgencias tras una atención urgente por consumo de cocaína. European monitoring Centre for drugs and drug addiction (2019), European Emergencias 22, 408–414, http://emergencias.portalsemes.org/numeros-anteriores/ drug report 2019: trends and developments. Brussels, Belgium: Publications volumen-22/numero-6/factores-que-determinan-la-reconsulta-a-urgencias-tras-una- Office of the European Union. atencion-urgente-por-consumo-de-cocaina/. Filho, J. C. C. L., Ogawa, M. Y., De Souza Andrade, T. H., De Andrade Cordeiro Observatorio Español de las Drogas y las Adicciones (2019). EDADES Gadelha, S., Fernandes, P. F. C. B. C., Queiroz, A. L., et al. (2019). Spectrum of 2017—informe web actualizado marzo 2019 OEDA. Available at: moz- Frontiers in Pharmacology | www.frontiersin.org 8 February 2021 | Volume 12 | Article 625610 Sanvisens et al. Burden of Cocaine Use Disorder extension://a8bef36a-8935-4d90-87ae-d41d2b63b348/enhanced-reader.html? Tate, J. P., Sterne, J. A. C., and Justice, A. C. (2019). Albumin, white blood cell openApp&pdfhttp%3A%2F%2Fwww.pnsd.mscbs.gob.es%2Fprofesionales% count, and body mass index improve discrimination of mortality in HIV- 2FsistemasInformacion%2FsistemaInformacion%2Fpdf%2FEDADES_2017_ positive individuals. AIDS 33, 903–912. doi:10.1097/QAD. Informe.pdf (Accessed March 30, 2020). 0000000000002140 Pavarin, R., Lugoboni, F., Mathewson, S., Ferrari, A. M., Guizzardi, G., and Taylor, A. L., Denniston, M. M., Klevens, R. M., McKnight-Eily, L. R., and Jiles, R. B. Quaglio, G. (2011). Cocaine-related medical and trauma problems: a (2016). Association of hepatitis C virus with alcohol use among U.S. Adults: nhanes consecutive series of 743 patients from a multicentre study in Italy. Eur. 2003–2010. Am. J. Prev. Med. 51, 206–215. doi:10.1016/j.amepre.2016.02.033 J. Emerg. Med. 18, 208–214. doi:10.1097/MEJ.0b013e3283440f25 Timko, C., Han, X., Woodhead, E., Shelley, A., and Cucciare, M. A. (2018). Qureshi, A., Chaudhry, S., and Suri, M. (2014). Cocaine use and the likelihood of Polysubstance use by stimulant users: health outcomes over 3 years. J. Stud. cardiovascular and all-cause mortality: data from the third national health and Alcohol Drugs 79, 799–807. doi:10.15288/jsad.2018.79.799 nutrition examination survey mortality follow-up study. J. Vasc. Interv. Neurol. Vergara-Moragues, E., González-Saiz, F., Lozano, O. M. O., Betanzos Espinosa, P., 7, 76–82. doi:10.1212/wnl.78.1_meetingabstracts.p07.031 Fernández Calderón, F., Bilbao-Acebos, I., et al. (2012). Psychiatric comorbidity in Santurtún, A., García Blanco, A., Fdez-Arroyabe, P., Santurtún, M., and cocaine users treated in therapeutic community: substance-induced versus independent Zarrabeitia, M. T. (2020). Cocaine in hospital admissions for diseases of disorders. Psychiatr. Res. 200, 734–741. doi:10.1016/j.psychres.2012.07.043 the circulatory system and as the underlying cause of death: analysis and Vitcheva, V. (2012). Cocaine toxicity and hepatic oxidative stress. Curr. Med. discussion. Cardiovasc. Toxicol. 20, 20–27. doi:10.1007/s12012-019- Chem. 19, 5677–5682. doi:10.2174/092986712803988929 09537-6 (Accessed October 15, 2020) Warden, D., Sanchez, K., Greer, T., Carmody, T., Walker, R., dela Cruz, A., et al. Sanvisens, A., Vallecillo, G., Bolao, F., Rivas, I., Fonseca, F., Fuster, D., et al. (2014). (2016). Demographic and clinical characteristics of current comorbid Temporal trends in the survival of drug and alcohol abusers according to the psychiatric disorders in a randomized clinical trial for adults with stimulant primary drug of admission to treatment in Spain. Drug Alcohol Depend. 136, use disorders. Psychiatr. Res. 246, 136–141. doi:10.1016/j.psychres.2016.09.007 115–120. doi:10.1016/j.drugalcdep.2013.12.022 Stinson, F. S., Grant, B. F., Dawson, D. A., Ruan, W. J., Huang, B., and Saha, T. Conflict of Interest: The authors declare that the research was conducted in the (2005). Comorbidity between DSM-IV alcohol and specific drug use disorders absence of any commercial or financial relationships that could be construed as a in the United States: results from the national epidemiologic survey on alcohol potential conflict of interest. and related conditions. Drug Alcohol Depend. 80, 105–116. doi:10.1016/j. drugalcdep.2005.03.009 Copyright © 2021 Sanvisens, Hernández-Rubio, Zuluaga, Fuster, Papaseit, Galan, Subdirecció General de Drogodependències (2018). Sistema d’Informació sobre Farré and Muga. This is an open-access article distributed under the terms of the Drogodependències de Catalunya. Available at: http://drogues.gencat.cat/ca/ Creative Commons Attribution License (CC BY). The use, distribution or professionals/epidemiologia/sid/ (Accessed October 15, 2020). reproduction in other forums is permitted, provided the original author(s) and Tate, J. P., Justice, A. C., Hughes, M. D., Bonnet, F., Reiss, P., Mocroft, A., et al. the copyright owner(s) are credited and that the original publication in this journal is (2013). An internationally generalizable risk index for mortality after one year of cited, in accordance with accepted academic practice. No use, distribution or antiretroviral therapy. AIDS 27, 563–572. doi:10.1097/QAD.0b013e32835b8c7f reproduction is permitted which does not comply with these terms. Frontiers in Pharmacology | www.frontiersin.org 9 February 2021 | Volume 12 | Article 625610