Journal of Substance Abuse Treatment 53 (2015) 71–77

Contents lists available at ScienceDirect

Journal of Substance Abuse Treatment

Factors Associated with Hospitalization for Blood-Borne Viral Infections Among Treatment-Seeking Illicit Drug Users Ifeoma N. Onyeka, M.B.B.S., M.Sc.P.H. a,⁎, Olubunmi Olubamwo, M.B.B.S., F.M.C.Path., M.P.H. a, Caryl M. Beynon, B.Sc., D.L.S.H.T.M., M.Sc., Ph.D. b, Kimmo Ronkainen, M.Sc. a, Jaana Föhr, M.D. c, Jari Tiihonen, M.D., Ph.D. d,e,f, Pekka Tuomola, M.D. c, Niko Tasa, D.B.A. c, Jussi Kauhanen, M.D., Ph.D., M.P.H. a a

Institute of Public Health and Clinical Nutrition, Faculty of Health Sciences, University of Eastern Finland, Kuopio, Finland Independent Research Consultant, Liverpool, United Kingdom c Helsinki Deaconess Institute, Helsinki, Finland d Department of Forensic Psychiatry, University of Eastern Finland, Niuvanniemi Hospital, Kuopio, Finland e National Institute for Health and Welfare, Helsinki, Finland f Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden b

a r t i c l e

i n f o

Article history: Received 19 May 2014 Received in revised form 25 December 2014 Accepted 6 January 2015 Keywords: Blood-borne pathogen HIV Hepatitis C Substance abuse Hospitalization Risk factor

a b s t r a c t Blood-borne viral infections (BBVIs) are important health consequences of illicit drug use. This study assessed predictors of inpatient hospital admissions for BBVIs in a cohort of 4817 clients seeking treatment for drug use in Finland. We examined clients’ data on hospital admissions registered in the Finnish National Hospital Discharge Register from 1997 to 2010 with diagnoses of BBVIs. Cox proportional hazards regression analyses were separately conducted for each of the three BBVI groups to test for association between baseline variables and hospitalizations. Findings were reported as adjusted hazard ratios (aHRs). Based upon primary discharge diagnoses, 81 clients were hospitalized for HIV, 116 for hepatitis C, and 45 for other types of hepatitis. Compared to those admitted for hepatitis C and other hepatitis, drug users with HIV had higher total number of hospital admissions (294 versus 141 and 50 respectively), higher crude hospitalization rate (7.1 versus 3.4.and 1.2 per 1000 person-years respectively), and higher total length of hospital stay (2857 days versus 279 and 308 respectively). Trends in hospitalization for all BBVI groups declined at the end of follow-up. HIV positive status at baseline (aHR: 6.58) and longer duration of drug use (aHR: 1.11) were independently associated with increased risk for HIV hospitalization. Female gender (aHR: 3.05) and intravenous use of primary drug (aHR: 2.78) were significantly associated with HCV hospitalization. Having hepatitis B negative status at baseline (aHR: 0.25) reduced the risk of other hepatitis hospitalizations. Illicit drug use coexists with blood-borne viral infections. To address this problem, clinicians treating infectious diseases need to also identify drug use in their patients and provide drug treatment information and/or referral. © 2015 Elsevier Inc. All rights reserved.

1. Introduction Blood-borne viral infections (BBVIs) such as HIV, hepatitis C (HCV), and other types of hepatitis are important health consequences of illicit drug use (EMCDDA, 2004). Intravenous drug use has been implicated in the majority of BBVIs in many parts of the world due to risky behaviors such as sharing of non-sterile injecting equipment (Aceijas & Rhodes, 2007; Alter, 2006; Garten et al., 2005; Grogan, Tiernan, Geogeghan, Smyth, & Keenan, 2005; Li et al., 2006; Loebstein et al., 2008; Maher, Chant, Jalaludin, & Sargent, 2004; Mathers et al., 2008; Nelson et al., 2011; Partanen, Vikatmaa, Tukiainen, Lepäntalo, & Vuola, 2009; Pereira et al., 2013; Xia, Luo, Bai, & Yu, 2008). BBVIs have also been reported among drug users ⁎ Corresponding author at: Institute of Public Health and Clinical Nutrition, Faculty of Health Sciences, University of Eastern Finland, Kuopio, P.O.Box 1627, FI-70211 Kuopio, Finland. Tel.: +358 403552905. E-mail address: ifeoma.onyeka@uef.fi (I.N. Onyeka). http://dx.doi.org/10.1016/j.jsat.2015.01.005 0740-5472/© 2015 Elsevier Inc. All rights reserved.

who share non-injecting drug paraphernalia such as straws, tubing, or pipes (Caiaffa et al., 2011; Tortu, McMahon, Pouget, & Hamid, 2004) and those who are homeless (Andıá et al., 2001). Engaging in some sexual behaviors (such as having unprotected sex and multiple sex partners) and skin penetration practices (such as tattooing and piercing) also increase the risk of exposure to BBVIs (Brodish et al., 2011; Fethers, Marks, Mindel, & Estcourt, 2000; Fry & Lintzeris, 2003; Liu, Grusky, Li, & Ma, 2006; Niccolai, Shcherbakova, Toussova, Kozlov, & Heimer, 2009; Turner et al., 2006). Globally, BBVIs among drug users is a concern to public health systems because they could be a bridge population for transmitting these infectious diseases to sero-negative drug users and to non-drug using members of the society through social and sexual interactions (Liu et al., 2006; Niccolai et al., 2009). Morbidity arising from BBVIs further impacts health systems through treatment and inpatient hospital admissions. For example, a study conducted in three HCV treatment centers in the Netherlands (Helsper et al., 2012) found that the costs per cured HCV patient including side effects were €28,500 and €15,400 for

72

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77

genotype 1/4 and genotype 2/3 respectively. This, in addition to their drug use, could have considerable public health and financial burden both to the society and drug-using individuals. Monovalent and combination vaccines against hepatitis A and B are currently available (Lugoboni, Quaglio, Civitelli, & Mezzelani, 2009) but there is none for HIV and HCV, so primary prevention and treatment are the only strategies. The use of hospital services for BBVIs in drug-using population is less studied. Studies have shown differential hospitalization by injectors, females, and persons who received highly active antiretroviral therapy (Barash, Hanson, Buskin, & Teshale, 2007; Bassetti, Hoffmann, Bucher, Fluckiger, & Battegay, 2002; Floris-Moore et al., 2003; Schoenbaum, Lo, & Floris-Moore, 2002; van Haastrecht et al., 1996). However, these studies were limited by focusing on HIV and on opiate users. Predictors of hospitalization might differ for BBVIs other than HIV and for persons who use other types of illicit drugs. To address this gap, we considered a wide range of BBVIs in a population using different types of drugs. This study aimed to assess patterns of hospitalization, and factors associated with increased risk of hospitalization for BBVIs among a cohort of Finnish drug users seeking treatment. 2. Methods 2.1. Study population The study population consisted of 4817 clients (3365 males and 1452 females) who sought treatment for drug use between 1997 and 2008 at Helsinki Deaconess Institute (HDI). The HDI is a large public utility foundation located in Helsinki that offers drug treatment services to the residents of Helsinki and other surrounding municipalities that comprise the Greater Helsinki area (about 1.3 million residents). Services are rendered free to the clients but their municipalities pay the service fees. The service where the clients sought treatment provides treatment for illicit drug users, but there were some occasional clients including minors with severe alcohol use problems, and polydrug users with alcohol and prescription medicines contributing most to the reasons for seeking treatment. This study population made up the epidemiological part of “huumehoito tietokanta” (HUUTI, translated as drug treatment database) consortium research project, and includes all consecutive drug users who sought treatment at HDI during 1997–2008. Here, we focused on a subset of clients with diagnoses of BBVIs during the follow-up period. The Ministry of Social Affairs and Health of Finland, and research ethics committees of HDI and North-Savo Hospital District approved the HUUTI project. Informed consent was not required because data released to the researchers were anonymized and the clients were not contacted. 2.2. Data collection Clinicians used a structured questionnaire to conduct interview at each client’s first visit in order to obtain their self-reported drug use history, and their social, and medical and psychiatric histories. Full details of the cohort have been described elsewhere (Onyeka et al., 2012; Onyeka et al., 2013). Clients’ data were linked to the Finnish National Hospital Discharge Register (FHDR) using personal identifiers. The FHDR has a total coverage of inpatient care provided at all hospitals and municipal health centres since 1969, and contains admission and discharge dates, discharge diagnoses, personal identity codes, hospital identifier codes and other information (Haikonen, Lunetta, Lillsunde, & Sund, 2013). The administrative health and social register system in Finland is reliable with good accuracy and completeness (Gissler & Haukka, 2004; Sund, 2012). The follow-up period was from the first day of the first visit to HDI until 31 December 2010. 2.3. Definitions Causes of inpatient hospital admissions were coded using the 10th version of the International Classification of Diseases (ICD-10); clients’

records contained the main/primary diagnosis and 1–3 additional/secondary diagnoses. By examining primary discharge diagnoses that fell within the disease-related chapters of the ICD-10 coding system (A00–R99), BBVIs were defined as follows: HIV (B20–B24), HCV (B17.1 and B18.2), and other hepatitis (B15.0, B15.9, B16.2, B16.9, B17.8, B18.1, B18.9, and B19.9). Apart from HCV, ICD-10 codes denoting other hepatitis subtypes were grouped as “other hepatitis” in order to increase statistical power. 2.4. Statistical analyses Statistical analyses were carried out using SPSS version 21 for windows. Clients’ demographic data and other baseline characteristics were summarized using frequencies, mean and standard deviation (SD). Baseline differences between each of the three BBVI groups versus the other clients were tested using χ 2 test or Fischer’s exact test for categorical variables and Mann–Whitney test for continuous variables. BBVIs were analyzed based upon primary (or main) discharge diagnoses and we restricted analyses to primary diagnoses for the first admission. The proportions of BBVI hospitalizations per year from 1997 to 2010 were calculated. Crude hospitalization rates (CHRs) were calculated by dividing the observed total number of hospitalizations for each BBVI group by the total person-years (PY) of follow-up for the cohort, expressed per 1000 PY. Cox proportional hazards regression analyses were performed separately to determine baseline variables that independently predicted the risk of being hospitalized for HIV infection, HCV, and other hepatitis. 2.5. Covariates for multivariate analyses We assessed the association between baseline variables and BBVI hospitalizations first in univariate models, and subsequently in multivariate models. To test for the assumptions of Cox model, log minus log and log survival functions were plotted for relevant covariates and those that violated the assumptions were excluded from the models. Apart from socio-demographic variables (age, and gender), we also considered clients’ living and health conditions, and drug use behaviors that might predispose them to BBVIs including: homelessness (defined as the presence or absence of postal code/address), BBVI status at baseline, duration of illicit drug use (defined as the difference between chronological age and age at initiation of illicit drug use), route of administration of primary drug, and past month frequency of primary drug use. For primary drugs, cannabis was chosen as a reference category because preliminary Kaplan–Meier analysis showed that cannabis users had significantly higher (or better) survival for hospitalization than the other illicit drugs (data not shown). Due to large amounts of missing data for BBVI status at baseline, these important variables were handled in a special way than the other covariates in the models by using the “missing data” as a reference category. Variables with P-value ≤ 0.05 in the univariate analyses were included in multivariate model. Results of the multivariate models were presented as adjusted hazard ratios (aHRs) and 95% confidence intervals (95% CIs). Predictor variables with P-value of ≤ 0.05 were considered significant. 3. Results A total of 4817 clients were followed up: 3365 males and 1452 females. The mean follow-up period was 8.6 years (SD = 3.3, range 0.01–13.9 years) and clients contributed a total of 41567.5 personyears. Using the national hospital discharge register, 76.7% (n = 3693) were hospitalized at least once, and a higher proportion of females (84.5%, n = 1227) experienced hospitalizations compared to males (73.3%, n = 2466). The overall mean number of hospital admission was 4.7 times (SD = 7.3), and the overall mean length of hospital stay was 67.3 days (SD = 198.5). The overall CHR was 540.2/1000 PY (95% CI: 533.1–547.3) and the overall standardized hospitalization ratio

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77

was 5.4 (95% CI: 5.3–5.5). Based upon primary discharge diagnosis, the main contributors to hospital admissions included psychosis (n = 622), schizophrenia (n = 604), depression (n = 497), cardiovascular diseases (n = 223), hepatitis C (n = 116), HIV (n = 81), and other types of hepatitis (n = 45). Detailed information about hospitalization in this cohort has been provided elsewhere (Onyeka et al., in press). 3.1. HIV infection At the end of the follow-up period, 81 clients had primary diagnoses of HIV infection. There were statistically significant differences in baseline characteristics between those hospitalized for HIV and the rest of the clients. As shown in Table 1, clients admitted for HIV were significantly older, homeless, more likely to have HIV positive status at baseline, and they had longer duration of illicit drug use relative to the other clients. They mainly used stimulants as primary drugs which they mostly administered intravenously compared to the other clients. Both groups were similar in terms of gender and past month frequency of drug use (Table 1). The total observed number of hospitalization with primary diagnosis of HIV was 294. Clients in this subgroup had the highest CHR (7.1/1000 PY, 95% CI = 6.3–7.9) and the longest lengths of stay (LOS) compared to the other two BBVIs, with total LOS of 2857 days (mean = 0.59, SD = 8.08). There was a decline in the proportions of clients with primary diagnoses of HIV. Following a peak in 2001 (3.0%), the proportion of clients eventually declined to 0.7% in 2010 (Fig. 1). The results of Cox regression analyses are also shown in Table 1. In the multivariate model, only HIV positive status at baseline (aHR: 6.58, 95% CI = 2.96–14.60) and longer duration of drug use (aHR: 1.11, 95% CI = 1.01–1.21) were independently associated with increased risk for HIV hospitalization. 3.2. Hepatitis C infection As shown in Table 2, 116 clients were hospitalized with primary discharge diagnoses of HCV. At baseline, these clients were significantly

73

more likely to be older, females, had longer duration of drug use, mainly used stimulants as primary drug and mostly injected the primary drug compared to the other clients (Table 2). Although a higher proportion of clients with HCV hospitalizations reported positive tests for HCV at baseline compared to the other clients, this difference was only weakly significant at P = 0.06. A total of 141 hospital admissions with primary diagnoses of HCV were observed in this study cohort, with CHR of 3.4/1000 PY (95% CI = 2.9–4.0). The total lengths of hospital stay were 279 days (mean = 0.06, SD = 1.00). The trend in HCV hospitalizations first peaked at 2.7% in 2001, achieved a second peak in 2004 (3.0%), and then declined to 0.5% in 2010 (Fig. 1). The full multivariate Cox model showed that female gender (aHR: 3.05, 95% CI = 1.71–5.46) and intravenous use of primary drug (aHR: 2.78, 95% CI = 1.00–7.70) were significantly independently associated with higher risk of HCV hospitalization. 3.3. Other hepatitis infections A total of 45 clients had primary diagnoses of hepatitis infections other than HCV (Table 3). The main contributors included acute hepatitis B (n = 19, ICD-10 code = B16.9), acute hepatitis A (n = 15, ICD-10 code = B15.9), and unspecified viral hepatitis (n = 6, ICD-10 code = B19.9). Differences in baseline characteristics are shown in Table 3. A higher proportion of clients diagnosed with other types of hepatitis were men and they were more likely to report hepatitis B positive test at baseline compared to the other clients (P = 0.032 and P = 0.050 respectively). Similarly, they engaged more in injecting, and daily use of their primary drug during past month compared to others (P = 0.003 and P = 0.009 respectively, Table 3). Use of opiates as primary drug was high (44.4%) among clients hospitalized for other hepatitis but this was not statistically different from the other group (P = 0.134). At the end of the follow-up, a total of 50 hospital admissions for other types of hepatitis infections were observed. Relative to the other two BBVIs, clients hospitalized for other hepatitis infections had the

Table 1 Baseline differences and Cox regression analyses of predictors of hospitalization for HIV infection among 4817 drug-using clients. Baseline variables

Clients hospitalized for HIV (n = 81)

Age, mean (SD) Gender, n (%) Females Males Homeless, n (%) Yes No (Missing data) HIV at baseline, n (%)c Yes No Missing data Drug use duration, mean (SD) Primary drugs of abuse, n (%) Alcohol Prescription medicine Opiates Stimulants Others Cannabis Route, primary drug, n (%) Intravenous Non-intravenous Missing data Frequency, primary drugs, n (%) Daily users Non-daily users Missing data

32.0 (7.4)

a

Other clients (n = 4736) 24.4 (8.0)

P-valuea

HR (95% CI)

Adjusted HR (95% CI)

P-valueb

b0.001

1.09 (1.07 – 1.12)*

1.00 (0.92 – 1.09)

0.947

1.09 (0.68 – 1.74) 1.00 (ref)

–

–

26 (32.1) 55 (67.9)

1426 (30.1) 3310 (69.9)

0.699

35 (44.3) 44 (55.7) 2 (-)

985 (21.7) 3552 (78.3) 199 (-)

b0.001

2.89 (1.86 – 4.51)* 1.00 (ref)

1.54 (0.81 – 2.93) 1.00 (ref)

0.184

11 (57.9) 8 (42.1) 62 (-) 18.0 (8.6)

38 (2.6) 1414 (97.4) 3284 (-) 9.3 (7.2)

b0.001

18.20 (9.55 – 34.68)* 0.34 (0.16 – 0.71)** 1.00 (ref) 1.11 (1.08 – 1.14)*

6.58 (2.96 – 14.60) 0.22 (0.09 – 0.52) 1.00 (ref) 1.11 (1.01 – 1.21)

b0.001 0.001 – 0.024

5 (6.2) 1 (1.2) 23 (28.4) 51 (63.0) 0 (0.0) 1 (1.2)

999 (21.1) 95 (2.0) 1409 (29.8) 1283 (27.1) 57 (1.2) 893 (18.9)

b0.001

4.78 (0.56 – 40.95) 10.44 (0.65 – 166.88) 15.37 (2.08 – 113.84)*** 35.74 (4.94 – 258.65)* N.A. 1.00 (ref)

2.31 (0.24 3.36 (0.20 2.00 (0.20 2.68 (0.27 N.A. 1.00 (ref)

66 (84.6) 12 (15.4) 3 (-)

1927 (44.6) 2393 (55.4) 416 (-)

b0.001

6.77 (3.66 – 12.53)* 1.00 (ref)

2.46 (0.76 – 7.95) 1.00 (ref)

0.134

35 (52.2) 32 (47.8) 14 (-)

1824 (43.3) 2385 (56.7) 527 (-)

0.145

1.48 (0.91 – 2.38) 1.00 (ref)

–

–

b0.001

– 22.36) – 57.03) – 19.68) – 26.35)

0.471 0.401 0.553 0.397 N.A.

For Chi-square test or Fischer’s exact test for categorical variable and Mann-Whitney test for continuous variables. For the multivariate model. Missing data used as a reference category due to large amount of missing information. For variables with missing data, % is proportion of complete data. HR – hazard ratio. CI – confidence interval. SD – standard deviation. N.A. - not applicable. *P b 0.001. **P = 0.004. ***P = 0.007. b c

74

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77

Proportions (%) of clients

3.5 3 2.5 2 1.5 1 0.5 0

1997

1998

1999

2000

2001

2002

2003

2004

2005

2006

2007

2008

2009

HIV

0

0.7

1.8

2.3

3

2.5

2.7

2.8

2.4

1.7

0.7

1

1.3

2010 0.7

Hepatitis C

0

1

1.2

0.8

2.7

1.6

1.1

3

2

1.6

0.8

0.5

0.2

0.5

Other hepatitis

0

1.4

1.2

1

0.7

2.4

0.1

0

0

0.1

0

0.1

0

0.1

Fig. 1. Clients with primary discharge diagnoses of HIV, hepatitis C, and other types of hepatitis infections as a proportion of disease-related hospitalizations, 1997 – 2010.

least CHR (1.2/1000 PY, 95% CI = 0.9–1.6). Their total lengths of hospital stay were 308 days (mean = 0.06, SD = 0.82). Trend in other hepatitis hospitalizations decreased from 2.4% in 2002 to 0.1% in 2010 (Fig. 1). In the univariate Cox models, female gender and not having hepatitis B at baseline were protective while using opiates as primary drug, injecting, and past month daily use of primary drug increased the risk of being hospitalized for other hepatitis infections. However, only hepatitis B negative status at baseline (aHR: 0.25, 95% CI = 0.09–0.71) remained statistically significant in the multivariate model. 4. Discussion This retrospective study of treatment-seeking illicit drug users in Finland found that the numbers of persons hospitalized for HCV were higher than the other BBVIs. However, those hospitalized for HIV had

higher total number of admissions, higher CHR, and higher total length of stay compared to HCV and other types of hepatitis infections. HIV positive status at baseline and longer duration of drug use increased the hazard of HIV hospitalization. Female gender and intravenous use of primary drug were significantly associated with HCV hospitalization while having hepatitis B negative status at baseline reduced the hazard of other hepatitis hospitalizations. Although clients with primary discharge diagnoses of HIV were fewer than those with HCV (81 versus 116 persons respectively), the HIV group tended to use more hospital resources as evidenced by higher total number of admissions and LOS, and CHR compared to the HCV group. A likely explanation for this greater utilization of health service is because HIV lowers the body’s immunity. This predisposes HIVinfected persons to opportunistic infections which increase as their CD4 cell counts decrease (Chowdhury, Huq, Roy, Talukder, & Haque,

Table 2 Baseline differences and Cox regression analyses of predictors of hospitalization for hepatitis C among 4817 drug-using clients. P-valuea

HR (95% CI)

Adjusted HR (95% CI)

P-valueb

24.4 (8.0)

b0.001

1.05 (1.03 – 1.07)*

1.02 (0.94 – 1.10)

0.696

61 (52.6) 55 (47.4)

1391 (29.6) 3310 (70.4)

b0.001

2.59 (1.80 – 3.73)* 1.00 (ref)

3.05 (1.71 – 5.46) 1.00 (ref)

b0.001

27 (25.0) 81 (75.0) 8 (-)

993 (22.0) 3515 (78.0) 193 (-)

0.462

1.21 (0.78 – 1.87) 1.00 (ref) –

–

–

16 (64.0) 9 (36.0) 91 (-) 12.3 (9.8)

634 (45.1) 772 (54.9) 3295 (-) 9.3 (7.3)

0.060

1.13 (0.66 – 1.92) 0.52 (0.26 – 1.03) 1.00 (ref) 1.05 (1.02 – 1.09)**

–

–

1.04 (0.96 – 1.12)

0.348

12 (10.3) 2 (1.7) 31 (26.7) 60 (51.7) 0 (0.0) 11 (9.5)

992 (21.1) 94 (2.0) 1401 (29.8) 1274 (27.1) 57 (1.2) 883 (18.8)

b0.001

1.09 (0.48 1.98 (0.44 1.92 (0.96 3.90 (2.05 N.A. 1.00 (ref)

77 (70.6) 32 (29.4) 7 (-)

1916 (44.7) 2373 (55.3) 412 (-)

b0.001

2.97 (1.96 – 4.48)* 1.00 (ref) –

2.78 (1.00 – 7.70) 1.00 (ref)

0.049

47 (45.6) 56 (54.4) 13 (-)

1812 (43.4) 2361 (56.6) 528 (-)

0.655

1.13 (0.77 – 1.67) 1.00 (ref) –

–

–

Baseline variables

Clients hospitalized for hepatitis C (n = 116)

Age, mean (SD) Gender, n (%) Females Males Homeless, n (%) Yes No (Missing data) Hepatitis C at baseline, n (%)c Yes No Missing data Drug use duration, mean (SD) Primary drugs of abuse, n (%) Alcohol Prescription medicine Opiates Stimulants Others Cannabis Route, primary drug, n (%) Intravenous Non-intravenous Missing data Frequency, primary drugs, n (%) Daily users Non-daily users Missing data

27.5 (9.1)

a

Other clients (n = 4701)

0.037

– 2.47) – 8.93) – 3.81) – 7.41)*

0.56 (0.12 1.34 (0.15 0.57 (0.14 0.98 (0.24 N.A. 1.00 (ref)

– 2.51) – 12.15) – 2.42) – 3.97)

0.445 0.796 0.447 0.972 N.A. –

For Chi-square test or Fischer’s exact test for categorical variable and Mann-Whitney test for continuous variables. For the multivariate model. Missing data used as a reference category due to large amount of missing information. For variables with missing data, % is proportion of complete data. HR – hazard ratio. CI – confidence interval. SD – standard deviation. N.A. - not applicable. *P b 0.001. **P = 0.001. b c

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77

75

Table 3 Baseline differences and Cox regression analyses of predictors of hospitalization for other hepatitis among 4817 drug-using clients. P-valuea

HR (95% CI)

Adjusted HR (95% CI)

24.5 (8.1)

0.362

1.02 (0.99 – 1.05)

–

7 (15.6) 38 (84.4)

1445 (30.3) 3327 (69.7)

0.032

0.42 (0.19 – 0.95)* 1.00 (ref)

0.60 (0.26 – 1.38) 1.00 (ref)

0.231

14 (32.6) 29 (67.4) 2 (-)

1006 (22.0) 3567 (78.0) 199 (-)

0.097

1.74 (0.92 – 3.29) 1.00 (ref)

–

–

0 (0.0) 6 (100.0) 39 (-)

71 (5.4) 1244 (94.6) 3457 (-)

1.00

N.A. 0.46 (0.20 – 1.09) 1.00 (ref)

–

–

2 (33.3) 4 (66.7) 39 (-) 7.8 (5.2)

84 (6.2) 1271 (93.8) 3417 (-) 9.4 (7.3)

0.050

2.15 (0.52 – 8.92) 0.30 (0.11 – 0.83)** 1.00 (ref) 0.97 (0.92 – 1.03)

1.59 (0.38 – 6.74) 0.25 (0.09 – 0.71) 1.00 (ref) –

0.527 0.009 – –

7 (15.6) 1 (2.2) 20 (44.4) 14 (31.1) 0 (0.0) 3 (6.7)

997 (20.9) 95 (2.0) 1412 (29.6) 1320 (27.7) 57 (1.2) 891 (18.7)

0.134

2.18 (0.56 3.34 (0.35 4.36 (1.30 3.22 (0.93 N.A. 1.00 (ref)

28 (68.3) 13 (31.7) 4 (-)

1965 (45.1) 2392 (54.9) 415 (-)

0.003

2.63 (1.36 – 5.08)**** 1.00 (ref) –

1.49 (0.62 – 3.63) 1.00 (ref) –

0.375

25 (64.1) 14 (35.9) 6 (-)

1834 (43.3) 2403 (56.7) 535 (-)

0.009

2.39 (1.24 – 4.59)***** 1.00 (ref) –

1.55 (0.74 – 3.27) 1.00 (ref) –

0.246

Baseline variables

Clients hospitalized for other hepatitis (n = 45)

Age, mean (SD) Gender, n (%) Females Males Homeless, n (%) Yes No (Missing data) Hepatitis A at baseline, n (%)c Yes No Missing data Hepatitis B at baseline, n (%)c Yes No Missing data Drug use duration, mean (SD) Primary drugs of abuse, n (%) Alcohol Prescription medicine Opiates Stimulants Others Cannabis Route, primary drug, n (%) Intravenous Non-intravenous Missing data Frequency, primary drugs, n (%) Daily users Non-daily users Missing data

25.6 (8.4)

Other clients (n = 4772)

0.375

– 8.44) – 32.07) – 14.67)*** – 11.20)

0.91 (0.15 3.32 (0.34 3.13 (0.75 2.08 (0.49 N.A. 1.00 (ref)

– 5.50) –32.88) – 13.12) – 8.86)

P-value b

0.919 0.305 0.119 0.324 N.A. –

a

For Chi-square test or Fisher’s exact test for categorical variable and Mann-Whitney test for continuous variables. For the multivariate model. Missing data used as a reference category due to large amount of missing information. HR – hazard ratio. CI – confidence interval. SD – standard deviation. N.A. - not applicable. *P = 0.036. **P = 0.021. ***P = 0.017. ****P = 0.004. *****P = 0.009. b c

2014; Khalsa & Elkashef, 2010), and impacts the rate of hospital admissions. Cox regression analysis showed that positive HIV status at baseline independently increased the risk of hospitalization. This suggests that education and interventions targeted at boosting the immunity and improving the quality of life among HIV-positive drug users, and promoting adherence to highly active antiretroviral therapy (HAART) could help to reduce future morbidity and hospitalizations. Our result also showed that longer duration of drug use increased the risk for HIV hospitalizations. Since clients mainly used drugs intravenously, it means that the longer they use/inject, the greater the chances of contracting BBVIs. Those who are unable or unwilling to discontinue drug use could benefit from interventions that support transition from injecting to non-injecting route of drug administration as a counter measure against BBVIs, in addition to addressing sexual risk factors. A recent cross-sectional study (Des Jarlais et al., 2014) found that this measure significantly lowered HCV prevalence among formerinjectors compared to current-injectors but there were no statistically significant differences in HIV prevalence in both groups. Further research studies are needed to expand on these findings. HCV was a major contributor in terms of the number of clients hospitalized for BBVIs. Our findings were consistent with existing literature whereby the number of HCV exceeded those of HIV and other types of hepatitis (Grogan et al., 2005; Loebstein et al., 2008; Nelson et al., 2011; Partanen et al., 2009). Clients in our study mainly injected their primary drug, and HCV is predominantly transmitted via this route (Friedland, 2010). HCV is transmitted more easily than HIV because HCV titers in the blood on contaminated injecting equipment are higher than those of HIV (Friedland, 2010). The proportion of clients hospitalized for BBVIs declined after 2001/2004 period. We are unsure about the reason for this decreasing trend but it might be related to increased

intervention and prevention activities including education, and access to sterile injecting equipment following an outbreak of BBVI among drug users in this region of Finland around that period (Kivelä et al., 2007). Day-centre for HIV-positive injectors that offers low threshold methadone program, food, and social services free of charge was introduced at the end of 2000 (Kivelä et al., 2007), and this might have contributed to the decline in BBVIs. Improvement in and access to BBVI treatment regimens might be another explanation for the decline. For example, in a study of 604 HIV-infected drug users in New York, researchers found that the rate of hospitalization was lower in HAART users compared to non-HAART users (Floris-Moore et al., 2003). Despite the observed decline in BBVI hospitalizations, these prevention measures need to be sustained to further reduce BBVIs. Female drug users had three times the risk of HCV hospitalizations in males. This reflects gender differences in BBVI transmission. Researchers suggested that HCV positive women tend to be younger, acquire the virus earlier in their drug-using/injecting careers, require help to inject, and have sex partners who inject (Lidman et al., 2009; Miller et al., 2002). Women could contract BBVIs in general through other means. Apart from risks from intimate relationships, women who engage in commercial sex work are further exposed to risks from multiple partners and inability to negotiate safe sex with customers due to economic or drugrelated reasons (Choi, Cheung, & Chen, 2006). In our cohort, female clients were younger (mean: 22.8 versus 25.2 years) and had shorter duration of drug use (mean: 7.7 versus 10.1 years) compared to males (Onyeka et al., unpublished data) but we did not have any information about their sexual behaviors. We, therefore, recommend gender-sensitive approach to BBVI intervention and prevention. Acute hepatitis B accounted for nearly half (n = 19/45) of the subgroup categorized as “other hepatitis”. Having hepatitis B negative

76

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77

status at baseline was protective against hospitalizations. Hence, drug abuse treatment centres should encourage their clients to screen for BBVIs. Depending on the serostatus, those who tested negative should be informed about factors that might predispose them to infections in the future (such as sharing injecting equipment, unprotected sex, and continued drug use), and they should be counselled to repeat the tests periodically to ascertain their current serostatus. Individuals who are sero-positive should be referred for early treatment. It has been suggested that drug users who know their sero-positive status are less likely to further engage in risky behaviors than those unaware (Loebstein et al., 2008), which might reduce their chances of co-infection with other viruses. In this paper, we restricted data analysis only to BBVIs recorded as the main/primary diagnosis in the FHDR. Due to the register-based nature of this data, we are not sure why physicians made these decisions. For example, hospitalization for a HCV is not usually HCV solely, but probably complications secondary to the liver function or other circumstances, or acute infection requiring hospitalization. However, when we explored clients that had both primary diagnoses and secondary diagnoses of BBVIs, the data were as follows: HIV = 104, hepatitis C = 812, and other hepatitis = 120 clients. When we restricted analysis to only primary diagnoses of BBVIs, the clients dropped to: HIV = 81, hepatitis C = 116, and other hepatitis = 45, as reported in this paper. The Finnish Hospital Discharge Register has good accuracy (Sund, 2012); thus, our data may have accurately captured those hospitalized “due to” BBVIs as against those “with” BBVIs.

diagnosis only thereby excluding BBVIs that were secondary diagnoses. This was done to prevent the additional/secondary diagnoses from interacting with the main/primary diagnoses thereby obscuring the most important reasons for the hospital admissions. We only considered the first hospitalization and as such, our analyses did not account for multiple hospitalizations in the same person for the same disease. We could not test for associations with other predictors of BBVI hospitalizations such as sexual behaviors and BBVI treatment status at baseline (e.g. use of HAART) because this information was not available in the database. Due to the limitation of the dataset and the large number of study samples, we could not determine types and levels of coinfection with viral and other types of infections. Information about hepatitis A and B vaccination status was available for few persons in this cohort and therefore was excluded. Furthermore, clients’ characteristics might have changed during the follow-up period and the use of their most current information as covariates would have been more robust. Despite these limitations, findings from this large-scale longitudinal study will assist treatment planners and decision-makers in their ongoing efforts to improve the health of drug users and will guide future research. 4.3. Conclusion In conclusion, illicit drug use coexists with other medical problems like BBVIs that could increase morbidity and the use of hospital services. Appropriate public health response would require combination of efforts from health professionals working in drug abuse treatment centres, and clinicians who treat infectious diseases in the general population.

4.1. Implications for practice Authors’ contributions Our findings have implications for the public healthcare system in Finland. These diagnoses of BBVIs were made in various hospitals including those offering services to the general population. Clinicians treating infectious diseases need to also identify drug use in their patients and provide drug treatment information/referral. Researchers have suggested combined intervention/treatment for substance use and infectious diseases due to the interconnectedness of both disorders (Parry, Rehm, Poznyak, & Room, 2009; Schneider, Chersich, Neuman, & Parry, 2012). Such combination of services might improve access to HCV treatment for intravenous drug users (Wiessing, 2001). More so, ongoing substance use increases the likelihood of non-adherence to infectious disease therapy (Hendershot, Stoner, Pantalone, & Simoni, 2009). Preventive efforts against BBVIs should include risk-reducing measures such as screening, counselling, vaccination for hepatitis, education, substitution therapy for opiate users, and offering sterile injecting equipment via needle exchange programs (NEP) (EMCDDA, 2004). High level of individual syringe coverage percentage could increase the impact of NEP. In a study examining different ranges of syringe coverage among 1577 injectors in 24 NEPs in California, Bluthenthal, Anderson, Flynn, and Kral (2007) found that syringe coverage of 150% or more reduced the odds of risky behaviors such as re-using and sharing syringes, and sharing cookers. The researchers suggested that “achieving 100% syringe coverage is important, but exceeding 100% coverage may be required to maximize the public health impact of syringe access” (Bluthenthal et al., 2007, pg. 219). 4.2. Strengths and limitations This study has some limitations. Our study relied only on data for BBVIs requiring inpatient hospital care and might have underestimated the extent of BBVIs among drug users. This study cohort consisted of treatment-seekers who might have different morbidities and hospital needs from non-treatment seekers and does not represent all drug users in Finland. This limits the generalizability of our findings to nontreatment seekers but because of the large sample size, it is likely that the findings are generalizable to other treatment-seeking drug users in Finland. Our data analyses were restricted to primary discharge

INO, JK, and CMB designed this particular study. INO carried out data analysis with KR. INO conducted literature review with OO. INO prepared the first draft of the manuscript. All authors participated in interpreting the results and critical revision of the manuscript for important intellectual content. All authors read and approved the final manuscript. Conflict of interest INO received an unrelated grant from Lundbeck Ltd. Finland in 2013. All other authors declare that they have no conflicts of interest. Acknowledgement The Academy of Finland (grant number 118584) and the University of Eastern Finland Doctoral Programme provided funding for this study. Funders had no further role in study design; in the collection, analysis and interpretation of data; in the writing of the report; or in the decision to submit the paper for publication. References Aceijas, C., & Rhodes, T. (2007). Global estimates of prevalence of HCV infection among injecting drug users. International Journal of Drug Policy, 18, 352–358. Alter, M. J. (2006). Epidemiology of viral hepatitis and HIV co-infection. Journal of Hepatology, 44, s6–s9. Andıá, J. F., Deren, S., Kang, S. Y., Robles, R. R., Colón, H. M., Oliver-Velez, D., et al. (2001). Residential status and HIV risk behaviors among Puerto Rican drug injectors in New York and Puerto Rico. American Journal of Drug and Alcohol Abuse, 27, 719–735. Barash, E. T., Hanson, D. L., Buskin, S. E., & Teshale, E. (2007). HIV-infected injection drug users: health care utilization and morbidity. Journal of Health Care for the Poor and Underserved, 18, 675–686. Bassetti, S., Hoffmann, M., Bucher, H. C., Fluckiger, U., & Battegay, M. (2002). Infections requiring hospitalization of injection drug users who participated in an injection opiate maintenance program. Clinical Infectious Diseases, 34, 711–713. Bluthenthal, R. N., Anderson, R., Flynn, N. M., & Kral, A. H. (2007). Higher syringe coverage is associated with lower odds of HIV risk and does not increase unsafe syringe disposal among syringe exchange program clients. Drug and Alcohol Dependence, 89, 214–222. Brodish, P., Singh, K., Rinyuri, A., Njeru, C., Kingola, N., Mureithi, P., et al. (2011). Evidence of high-risk sexual behaviors among injection drug users in Kenya PLACE study. Drug and Alcohol Dependence, 119, 138–141.

I.N. Onyeka et al. / Journal of Substance Abuse Treatment 53 (2015) 71–77 Caiaffa, W. T., Zocratto, K. F., Osiman, M. L., Martinez, P. L., Radulich, G., Latorre, L., et al. (2011). Hepatitis C virus among non-injecting cocaine users (NICUs) in South America: Can injectors be a bridge? Addiction, 106, 143–151. Choi, S. Y., Cheung, Y. W., & Chen, K. (2006). Gender and HIV risk behavior among intravenous drug users in Sichuan Province, China. Social Science & Medicine, 62, 1672–1684. Chowdhury, T., Huq, O., Roy, P., Talukder, M. U., & Haque, F. (2014). Opportunistic infections among human immune deficiency virus (HIV) positive injecting drug users of Dhaka City, Bangladesh. Journal of AIDS and HIV Research, 6, 79–84. Des Jarlais, D. C., McKnight, C., Arasteh, K., Feelemyer, J., Perlman, D. C., Hagan, H., et al. (2014). Transitions from injecting to non-injecting drug use: potential protection against HCV infection. Journal of Substance Abuse Treatment, 46, 325–331. European Monitoring Centre for Drugs and Drug Addiction (EMCDDA) (2004). Hepatitis C and injecting drug use: impact, costs and policy options. EMCDDA scientific monograph series no 7 (Available at: http://www.emcdda.europa.eu/html.cfm/index31213EN. html. Retrieved 07.04.2014). Fethers, K., Marks, C., Mindel, A., & Estcourt, C. S. (2000). Sexually transmitted infections and risk behaviours in women who have sex with women. Sexually Transmitted Infections, 76, 345–349. Floris-Moore, M., Lo, Y., Klein, R. S., Budner, N., Gourevitch, M. N., Moskaleva, G., et al. (2003). Gender and hospitalization patterns among HIV-infected drug users before and after the availability of highly active antiretroviral therapy. Journal of Acquired Immune Deficiency Syndromes, 34, 331–337. Friedland, G. (2010). Infectious disease comorbidities adversely affecting substance users with HIV: Hepatitis C and tuberculosis. Journal of Acquired Immune Deficiency Syndromes, 55, s37–s42. Fry, C. L., & Lintzeris, N. (2003). Psychometric properties of the blood-borne virus transmission risk assessment questionnaire (BBV-TRAQ). Addiction, 98, 171–178. Garten, R. J., Zhang, J., Lai, S., Liu, W., Chen, J., & Yu, X. F. (2005). Coinfection with HIV and hepatitis C virus among injection drug users in Southern China. Clinical Infectious Diseases, 41, s18–s24. Gissler, M., & Haukka, J. (2004). Finnish health and social welfare registers in epidemiological research. Norsk Epidemiologia, 14, 113–120. Grogan, L., Tiernan, M., Geogeghan, N., Smyth, B., & Keenan, E. (2005). Bloodborne virus infections among drug users in Ireland: A retrospective cross-sectional survey of screening, prevalence, incidence and hepatitis B immunisation uptake. Irish Journal of Medical Science, 174, 14–20. Haikonen, K., Lunetta, P., Lillsunde, P. M., & Sund, R. (2013). Methodological challenges in using the Finnish Hospital Discharge Register for studying fire-related injuries leading to inpatient care. BMC Medical Informatics and Decision Making, 13, 36. Helsper, C. W., Hellinga, H. L., van Essen, G. A., de Wit, G. A., Bonten, M. J., van Erpecum, K. J., et al. (2012). Real-life costs of hepatitis C treatment. The Netherlands Journal of Medicine, 70, 145–153. Hendershot, C. S., Stoner, S. A., Pantalone, D. W., & Simoni, J. M. (2009). Alcohol use and antiretroviral adherence: review and meta-analysis. Journal of Acquired Immune Deficiency Syndromes, 52, 180–202. Khalsa, J. H., & Elkashef, A. (2010). Interventions for HIV and hepatitis C virus infections in recreational drug users. Clinical Infectious Diseases, 50, 1505–1511. Kivelä, P., Kro, L. A., Simola, S., Vaattovaara, M., Tuomola, P., Brummer-Korvenkontio, H., et al. (2007). HIV outbreak among injecting drug users in the Helsinki region: social and geographical pockets. European Journal of Public Health, 17, 381–386. Li, J., Wang, J., Tian, K., Wang, Y., Zhang, L., & Huang, H. (2006). Epidemiology of hepatitis B, C, D and G viruses and cytokine levels among intravenous drug users. Journal of Huazhong University of Science and Technology. Medical sciences, 26, 221–224. Lidman, C., Norden, L., Kåberg, M., Kåll, K., Franck, J., Aleman, S., et al. (2009). Hepatitis C infection among injection drug users in Stockholm Sweden: Prevalence and gender. Scandinavian Journal of Infectious Diseases, 41, 679–684. Liu, H., Grusky, O., Li, X., & Ma, E. (2006). Drug users: a potentially important bridge population in the transmission of sexually transmitted diseases, including AIDS, in China. Sexually Transmitted Diseases, 33, 111–117.

77

Loebstein, R., Mahagna, R., Maor, Y., Kurnik, D., Elbaz, E., Halkin, H., et al. (2008). Hepatitis C, B, and Human immunodeficiency virus infections in illicit drug users in Israel: Prevalence and risk factors. The Israel Medical Association Journal, 10, 775–778. Lugoboni, F., Quaglio, G., Civitelli, P., & Mezzelani, P. (2009). Bloodborne viral hepatitis infections among drug users: The role of vaccination. International Journal of Environmental Research and Public Health, 6, 400–413. Maher, L., Chant, K., Jalaludin, B., & Sargent, P. (2004). Risk behaviors and antibody hepatitis B and C prevalence among injecting drug users in south-western Sydney, Australia. Journal of Gastroenterology and Hepatology, 19, 1114–1120. Mathers, B. M., Degenhardt, L., Phillips, B., Wiessing, L., Hickman, M., Strathdee, S. A., et al. (2008). Global epidemiology of injecting drug use and HIV among people who inject drugs: a systematic review. Lancet, 372, 1733–1745. Miller, C. L., Johnston, C., Spittal, P. M., Li, K., LaLiberté, N., Montaner, J. S., et al. (2002). Opportunities for prevention: Hepatitis C prevalence and incidence in a cohort of young injection drug users. Hepatology, 36, 737–742. Nelson, P. K., Mathers, B. M., Cowie, B., Hagan, H., Des Jarlais, D., Horyniak, D., et al. (2011). Global epidemiology of hepatitis B and hepatitis C in people who inject drugs: results of systematic reviews. Lancet, 378, 571–583. Niccolai, L. M., Shcherbakova, I. S., Toussova, O. V., Kozlov, A. P., & Heimer, R. (2009). The potential for bridging of HIV transmission in the Russian Federation: sex risk behaviors and HIV prevalence among drug users (DUs) and their non-DU Sex partners. Journal of Urban Health, 86, s131–s143. Onyeka, I. N., Beynon, C. M., Ronkainen, K., Tiihonen, J., Föhr, J., Kuikanmäki, O., et al. (2014). Hospitalization in a cohort seeking treatment for illicit drug use in Finland. Journal of Substance Abuse Treatment, http://dx.doi.org/10.1016/j.jsat.2014.12.009 (pii: S0740-5472(14)00259-1, Epub ahead of print). Onyeka, I. N., Beynon, C., Uosukainen, H., Korhonen, M. J., Ilomäki, J., Bell, J. S., et al. (2013). Coexisting social conditions and health problems among clients seeking treatment for illicit drug use in Finland: The HUUTI study. BMC Public Health, 13, 380. Onyeka, I. N., Uosukainen, H., Korhonen, M. J., Beynon, C., Bell, J. S., Ronkainen, K., et al. (2012). Sociodemographic characteristics and drug abuse patterns of treatmentseeking illicit drug abusers in Finland, 1997–2008: The HUUTI study. Journal of Addictive Diseases, 31, 350–362. Parry, C., Rehm, J., Poznyak, V., & Room, R. (2009). Alcohol and infectious diseases: An overlooked causal linkage? Addiction, 104, 331–332. Partanen, T. A., Vikatmaa, P., Tukiainen, E., Lepäntalo, M., & Vuola, J. (2009). Outcome after injections of crushed tablets in intravenous drug abusers in the Helsinki University Central Hospital. European Journal of Vascular and Endovascular Surgery, 37, 704–711. Pereira, L. M., Martelli, C. M., Moreira, R. C., Merchan-Hamman, E., Stein, A. T., Cardoso, M. R., et al. (2013). Prevalence and risk factors of Hepatitis C virus infection in Brazil, 2005 through 2009: A cross-sectional study. BMC Infectious Diseases, 13, 60. Schneider, M., Chersich, M., Neuman, M., & Parry, C. (2012). Alcohol consumption and HIV/AIDS: the neglected interface. Addiction, 107, 1369–1371. Schoenbaum, E. E., Lo, Y., & Floris-Moore, M. (2002). Predictors of Hospitalization for HIVPositive Women and Men Drug Users, 1996-2000. Public Health Reports, 117, s60–s66. Sund, R. (2012). Quality of the Finnish Hospital Discharge Register: a systematic review. Scandinavian Journal of Public Health, 40, 505–515. Tortu, S., McMahon, J. M., Pouget, E. R., & Hamid, R. (2004). Sharing of noninjection druguse implements as a risk factor for hepatitis C. Substance Use & Misuse, 39, 211–224. Turner, J. M., Rider, A. T., Imrie, J., Copas, A. J., Edwards, S. G., Dodds, J. P., et al. (2006). Behavioural predictors of subsequent hepatitis C diagnosis in a UK clinic sample of HIV positive men who have sex with men. Sexually Transmitted Infections, 82, 298–300. van Haastrecht, H. J., Bindels, P. J., Sluijs, T. A., van den Hoek, A. A., Reijneveld, S. A., & Coutinho, R. A. (1996). The impact of drug users on inpatient hospital care during the human immunodeficiency virus epidemic in Amsterdam. International Journal of Epidemiology, 25, 846–853. Wiessing, L. (2001). The access of injecting drug users to hepatitis C treatment is low and should be improved. Euro Surveillance, 5 (pii = 1709. Available at: http://www. eurosurveillance.org/ViewArticle.aspx?ArticleId=1709 Retrieved 07.04.2014). Xia, X., Luo, J., Bai, J., & Yu, R. (2008). Epidemiology of hepatitis C virus infection among injection drug users in China: Systematic review and meta-analysis. Public Health, 122, 990–1003.