Commercial Tattooing as a Potentially Important Source of Hepatitis C Infection

Introduction Hepatitis C virus (HCV), first identified in 1989 (34), is the most common chronic viral infection in the United States population, affecting at least 1.8%, or 3.9 million, Americans, and placing them at increasing risk of hepatic cirrhosis and carcinoma in coming decades (9,33). This extraordinary population burden of infection appears to have occurred through rapid population-wide spread of mostly asymptomatic HCV infection in the 1980s. Judging from a >80% decline in acute symptomatic cases after 1989, the rate of HCV transmission appears to have slowed substantially in recent years (33). To understand how so large an epidemic could have occurred and how to prevent recurrences, it is important to know what routes of infection were responsible for the most infections during the height of epidemic transmission in the 1980s. Despite numerous epidemiologic studies of risk factors for HCV infection measured by serologic markers, a substantial proportion of infections (perhaps 40% or more) remains unexplained by the accepted risk factors (4,8,31,68,126), suggesting that important risk factors are yet to be identified. Recent suggestions that a higher proportion of cases can be explained were based on studies of risk factor prevalence in HCVinfected patients only, without comparisons with uninfected control groups (5,33). Whereas injection-drug use is clearly an important mechanism of transmission, researchers have generally agreed that nosocomial transmission to either patients or patient-care personnel, perinatal transmission, and household contact do not account for large proportions of infections, but they have differed on the overall importance of promiscuous sexual activity, transfusions, and consumption of alcohol and other recreational drugs such as cocaine (33). While tattooing has been shown repeatedly to be a potent source of HCV infection in other countries, its role in transmission has received little study in the United States (33). To study risk factors thoroughly, an epidemiologic study must measure all important risk factors in the same study sample to assess overlap of association (collinearity) and interactions among risk factors through multivariate analysis. It must include large samples of subjects, enough to identify at least 10 HCV infections per risk factor studied in multivariate models (98). Extraordinary interview techniques must be used to elicit sensitive exposure information from subjects who may be reluctant to divulge prior indiscretions. Ideally both patients and researchers should be unaware of HCV status when assessing risk factors to avoid differential completeness of recollections in case and control groups. Risk factor associations and attributability must be established through demonstration of a higher risk factor prevalence in HCV-positive than in HCV-negative subjects, rather than by ascertainment in HCV-positive subjects only. In early 1991, with the newly released serologic tests for HCV infection, we undertook a prospective clinical epidemiologic study to identify all known risk factors for blood-borne infection in an internal medicine practice serving patients referred to a large orthopedic spinal clinic. This setting provided the advantages of large patient volume selected on factors not particularly related to blood-borne infection and the opportunity to probe deeply for risk-factor information in the context of an internist’s confidential history and physical examination before either the physician or patients were aware of blood test results. We included in the examination and interview all risk factors for blood-borne infection identified in a review of the literature. Anticipating the results of large population surveys started shortly thereafter, we did not report our results until a recent literature review suggested the need for epidemiologic evaluation of a broader range of risk factors (33). Methods Study population The study included 629 consecutive patients from the southwestern United States who visited an orthopedic spinal clinic in Dallas, Texas, between July 1, 1991, and November 15, 1992. Approximately 90% of the clinic’s patients had spinal problems unrelated to their occupation; whereas, the remaining 10% were covered by workers’ compensation. The study protocol was reviewed and approved by the university’s institutional review board. Since the sample was composed almost exclusively of employed working people, the middle age groups, which have higher risks of HCV infection (9), were disproportionately represented compared with the total United States civilian population (101) (Table 1). Blacks, Hispanics, and men were slightly overrepresented, although the gender distribution was close to that of the United States civilian labor force (26). Laborers, craftsmen, and service occupations were overrepresented, and technical, sales, managerial, and professional occupations were somewhat underrepresented. Low and middle income groups were slightly overrepresented, and more highly educated groups were underrepresented.TABLE 1: Sample and occupation-standardized distributions and hepatitis C virus prevalence rates of the 626 patients in a spinal clinic on socioeconomic and demographic characteristics, in comparison with distributions of the 1990 United States civilian populationThe underlying risk of the sample patients for blood-borne infection was similar to that of the adult population of Dallas County, Texas. Only 1 of the 629 subjects was to be for virus compared with the rate of measured in a of a sample of of Dallas in 1989 patients for spinal were first by an a history and physical examination in during which close was the the study and the subjects that the information on risk factors was for not be their or to and not their or in the importance to the of and with a interview a to measure risk factors for HCV infection, measured by the in and or of all and a blood for HCV The risk factors measured in the study included all of identified in a literature on hepatitis virus and The clinical information was and before the results of HCV serologic tests were To ascertainment of injection-drug the the of the the importance of to the of the and for the information on risk The of alcohol in was by the The of was by the in in samples from all 629 patients were for HCV by both the and Dallas, and The tests for the and the tests for the and the that by either the or the tests were for HCV by the and patient was to have been with HCV either of the or the was and at least 1 of the on the test was as or higher we subjects with 1 test as was by that of the had risk factor more the than the In studies of blood with most such test results to be and are whereas, in higher risk they are more to be and should be included (9,33). patients with results were from the their HCV exposure status could not be 626 for analysis. and and were with the and of the and was with tests for prospective studies in the To identify the most of risk we multivariate the risk factor at to avoid results from We the risk for the risk factors that the multivariate to The was for of the for the other by the of The the proportion of all HCV cases that are to risk the that the risk factors are and that all have been for in the The that the of HCV cases that are to the of all risk factors in the was by the of for the risk that the of the not the as in our the distributions of the risk factors are not they of the were by to the distribution of the 1990 United States civilian population and to the distribution of the 1990 United States civilian force the of the sample not to the United States population, for demographic comparison of the to other HCV prevalence the 626 patients were for HCV infection, a sample prevalence rate of The HCV rate was years of age and at the of was higher in men than in than and in and service than in and in middle income groups than in the of and in the groups that of or 1). to the and distributions of the United States civilian population, the overall HCV rate was more to the prevalence rate measured in a sample of the United States population in or more were in of the patients (Table Whereas the risk of HCV infection was with a risk it was most with in but so with at at other or in for more than 1 HCV of HCV infection to risk of HCV infection in a with the the of and of the and the of tattooing but so with the of on the risk was with and but not with the most common The risk was not with the in which the was was a association between the setting in which a was and the in were more than to be of higher to or and to of the use of the patients drugs The risk of HCV infection was with a history of injection-drug use and was higher in injection-drug use for more than 1 than in it for than 1 (Table The risk was particularly for patients who other and but so for who or of HCV infection to and alcohol consumption The risk of HCV infection was higher in patients who and in a with the of per but it was not with or in of the of from all was not with HCV infection after for the of not and related occupations of the patients had in a or related at in the was of an risk of HCV infection in who in contact with patients or or of either or in with contact with patients (Table risk of HCV infection was to be in men who had had This included men who had in and patient and were HCV of HCV infection to other risk factors for viral transmission suggested by prior risk factors had a was with HCV infection but other of a history of and by were not was a for the risk of HCV infection to be related to the of of blood received by for HCV prevalence was slightly in who received in or after than before of blood for viral risk factors was although the was not of the patients in the study had or had who had 10 or more sexual per at their age of sexual had an risk of HCV infection by sexual that men were the at risk of HCV infection although were men and to their risk not a history of or cirrhosis from hepatitis to have a slightly higher risk of HCV infection was association with a of as measured by the for not of the between and identified an risk in men to a men the United States to be at higher risk than the The represented by of patients for and an risk of HCV infection was that of who In a of the to be the most of HCV infection included injection-drug use for a or been a a in a and or more of per (Table that did not the but after for the in the included the measure of the of the covered by after and the for injection-drug use of than 1 after suggesting that the of and injection-drug use are of the other risk of the of the in the was of HCV infection and measure of risk factor in the of the the overlap of risk factors injection-drug use and the the association of in from to and to but highly in the among risk In as the first risk factors the the association of that the of on HCV infection was was highly with HCV infection before other risk factors the but when injection-drug use the association of cocaine to that its association was explained by the of injection-drug The associations of sexual and were explained by of all of the risk factors The association of HCV risk with the of of blood received by was to be to by the risk In the patients with history of injection-drug prior as a and HCV infection was in 1 of patients with a prior blood and in of with prior for the risk factors in a prior was not with HCV In a similar a history of sexual in the of sexual was not the proportion of HCV cases to risk factor or population risk from the to the distribution of the United States civilian population, we that a in a for the of HCV infections, by injection-drug and been a (Table risk their are for of the HCV infections in working of HCV infections to of the risk factors in the study sample and after to the United States The of study in a sample of working from the of the United States a of the of HCV infection than that by recent review results that the of HCV infections may have been from injection-drug and for from blood or promiscuous sexual not or may have an important role in or HCV infection and increasing its population The most important is that a in a a risk suggesting a role in the transmission of The that the or of a was by in the physical examination of all subjects, as in a prior study our from information was by interview or The for tattooing was by our that the risk of HCV infection in a with the measured of covered by the and the of tattooing The association with in than in at is with the of which may transmission, in to the of The association of HCV infection with of and the of such as and in importance was the in a multivariate of risk for more than as HCV infections as injection-drug use and that the risk factors for of HCV It is that our of risk were from of risk factor in both HCV-positive and HCV-negative groups a multivariate to the of HCV cases to risk factor for the of the other risk factors This should the practice of HCV infections to risk factors they have of how common the risk factors be in a control is an the for was first in the by and The range of known to have been by tattooing infections, and blood-borne hepatitis in have for decades been as a of transmission in of have between 10 and with acute symptomatic who were to have in common a from the same on the same or in a and either the or a was to have been of the tattooing techniques of tattooing or of the tattooing or in such as the tattooing or the by the of or In to the of cases and tattooing has been as a risk factor for infection in epidemiologic studies In an epidemiologic of prevalence rates of serologic in of tattooing to be an risk factor for in a which included age in 1 of the and a similar association between tattooing and serologic in men by after all who promiscuous sexual or injection-drug The rate with the of personnel, tattooing to be more with serologic than or years in the service The rate was higher in with or more than in with and in who had been in than in This did not measure injection-drug use or sexual and multivariate analysis. In a of a in in tattooing and to be more with serologic than use of drugs or history of a but multivariate was United States studies of risk factors for acute hepatitis in the did not identify tattooing as a risk factor of hepatitis from HCV were first by and from in and by from and from In patients were to have rates of prior tattooing in clinical studies in and Since epidemiologic studies have the association between tattooing and HCV infection 1). blood in a to be a risk factor for HCV after all with a history of injection-drug blood or sexual who did not in injection-drug use or with a to be a risk factor for HCV the risk increasing with the of for on from in to be with HCV of injection-drug transfusions, and sexual in a association between and HCV in a study of patients a control of hepatitis after all subjects with a history of or injection-drug in a similar association of with HCV in a study of cases and from a after all subjects with a or injection-drug use and for history of alcohol and sexual with an injection-drug and patients to a and that tattooing was not with HCV suggested an association but with subjects, they had for a test of the of patients with a history of viral hepatitis or and problems with of subjects the subjects and in an association with but they did not measure or control for injection-drug 1: risk of hepatitis C with risk factors in the for subjects the most important risk factors injection-drug and multivariate the included that may have substantial the risk studies in which all subjects with the risk factor were Risk factors represented by were not measured or how could tattooing be responsible for so a of HCV cases in our The may in a rapid in the of tattooing in the United States in the and early to a study of tattooing by the of the tattooing as in that was responsible for tattooing from the practice of and of to a of in all of of in 1990 the established tattooing that was to models and of or large in their in that as did the of tattooing In a study of the and of their that most to to and for their from and and to tattooing compared or before a tattooing before or their to be a risk in and a large in that of had and of the were tattooing in the In the of the had tattooing by or tattooing to or setting for and a on tattooing by the of has United States when to have been to and of that of the had for in had at the or with The of infection control was by an of in a the of tattooing of the and early may have provided the same of for the spread of HCV infection as the of promiscuous and injection-drug use of the and did for when it was the use The risk and the association of a HCV test with the of injection-drug use in our study its role in HCV This is been in all studies in which it was 1). is in our study is the of the risk a and the risk a more population than other risk The risk is to the that injection-drug use is common in the population we studied than other such as tattooing and it is a potent it in the population than are by the more common risk ascertainment of injection-drug use could have to we that the to elicit information in the context of an internist’s history and physical examination and the and by the physician This is by our a higher prevalence of injection-drug use and yet a rate in our sample than were in the household of Dallas that a at risk of HCV may to who and as the more risk through exposure to in patient and blood in studies of HCV risk in exclusively on and and as did to be at little or risk of HCV from and from risk in with history of a exposure after or subjects with a history of and injection-drug use but the other studies have yet HCV infection in In serologic surveys of that included that and had rates of as as the risk patient-care and of explained were at that the rates may have been to the the rates in in our study and that of after for the other risk factors the that the rates in men were to the prior studies of in did not the risks in men and that not the that the exposure to HCV may be in that are by exposure to during such as and and is the the recent and of for by the and may have the was the association of HCV with that after for the other important risk factors in the multivariate was with HCV infection, the of in the United States population of the patients in our sample use of or to an association with alcohol of per for to be a risk factor for HCV infection in their study after subjects with and injection-drug use and for other risk factors 1). a association of but not with infection in men in similar to our although that association did not in a multivariate analysis. This to may have been a for that viral transmission it is that consumption of and to viral infection Recent studies have identified of alcohol that may the to In in that chronic alcohol and and in the control by HCV and the to a of to HCV were in from alcohol in the but the to in on alcohol that chronic alcohol the of the between and to in should the to viral and to higher prevalence of HCV infection in alcohol was to similar in to with This mechanism has been by the demonstration that alcohol consumption viral in patients with chronic HCV infection and that alcohol consumption during for HCV infection the of of the virus it appears that the association of with the prevalence of HCV infection in our study may an important to HCV infection in the The for of blood and blood to HCV infection has been by numerous studies in and particularly in such as patients and who large of or blood from that at most a of HCV and little to the population rate of to HCV prevalence was suggested by epidemiologic studies of hepatitis in the early and by more recent studies risk factors in patients without control comparisons The of epidemiologic have either association a association with 1). We that the association was to the of other risk the studies a a in the risk with of blood but did not of blood from and other blood The which the most association from a large that could have a as The studies by from from and from were in the blood may not have been as early as in the United the of of United States blood in the and HCV serologic in the the rate of HCV infections from by or more the of HCV infections from appears to have been in comparison with the from injection-drug and to have had a on the population prevalence for the transmission of HCV by sexual sexual or association between the of sexual at age of sexual was explained by from other risk particularly injection-drug in multivariate In to infection for which sexual is a of transmission, studies based on multivariate have to sexual to be an risk factor for HCV infection and in of sexual transmission either did not control for injection-drug use or had other 1). Only of the studies a but the risks in were to 1). studies have prevalence rates of HCV in the of HCV-positive subjects HCV or between or association of risk factor in multivariate Whereas 1 study in HCV infection between the of HCV-positive and HCV-negative a but injection-drug use and from were not In a study of men in that and infection, for viral infection, were with other but were not with HCV infection and a of between and HCV in men in with a prior of sexual transmission based on the with and overall has not a between sexual and HCV and gender The risk of HCV in from prior studies in an and a clinic was in our study to be almost to by the of been a and are particularly at with of a large study of blood from to by the of and the and by 1). consecutive blood with a or were as or on the of their test all were to the the first who were and they were by who the test results. In a multivariate the and the the they associations with a history of blood cocaine injection-drug sexual and in men 1). was with HCV infection in but was not in the multivariate analysis. We that the at such a rate with prior of the HCV serologic results may have and risk-factor by with risk factors and their infections, risk-factor in the population in which This was by the report of higher of and blood in the than in the and could have associations of the 1). of risk factors of is the of not to such as injection-drug sexual alcohol particularly in with The more are with such the more they are to This is important to the study it has been suggested that of injection-drug use the of the association between HCV infection and other risk such as sexual The to is to in the of risk-factor include placing sensitive in a interview after has been and sensitive with that the of study employed all of the risk factors in a by a physician who first established through a history and physical examination and a is to have higher than that by in the rates have been completeness of could not be We that the of the least of of injection-drug use of it could have the associations between other risk factors and HCV to on the that the who did not their injection-drug use were more to be HCV and to have other risk This was by who that the to in rather than is to the association between risk factors and the in a large population of injection-drug with known they that the as were with the prevalence of such as at and more than a This that were The were not with that and the were similar in the and for the had on the of the associations between the risk factors and This is explained by the of which that to be a a must be with both the risk factor and the Since was not with it could not the association between injection-drug use and In our study is to be with HCV both subjects and the physician were unaware of HCV at the of the risk-factor had of injection-drug use the associations of other risk factors with HCV we have sexual to be with HCV of injection-drug use but it was we it that the associations of and with HCV were or by from of injection-drug The of by must be from the more important of by measured injection-drug use and other risk The is an important that was by the multivariate in We that our in the context of the from prior studies an of to the spread of This should include the and of in in all to for of the to all not in patient from to injection-drug use and to it and the The association of with HCV infection the need for the by which alcohol consumption may HCV and the is to in is known to blood-borne viral infections, C virus (HCV), in other countries, but its to the population prevalence of HCV infection in the United States has been Risk factors for blood-borne infection were by interview of 626 consecutive patients evaluation for spinal problems in and unaware of their HCV all were for HCV infection with and and were with patients were for HCV prevalence prevalence identified risk factors for HCV injection-drug use prevalence by men from and of per risk factors account for of HCV infections, with tattooing injection-drug use and for men promiscuous sexual activity, perinatal or transmission in and other were not with serologic of HCV to be explained by or of other risk tattooing in may have been responsible for more HCV infections than injection-drug The the of the for the study of and of for of serologic and for for on for and for