PREVALENCE OF REACTIVE TUBERCULIN SKIN TEST IN DENTAL HEALTHCARE WORKERS AND STUDENTS

Introduction. The risk for active tuberculosis (TB) infection among dental healthcare workers (DHCWs) is controversial. Specifically, TB incidence is generally low among general dental practitioners, but it is higher than in the general population among hospital/prison dentists, DHCWs who work in endemic areas, students during clinical training with direct patient contact. Aim. To investigate prevalence of reactive tuberculin skin test (TST) among dental students and DHCWs in an Italian dental university hospital. Material and methods. DHCWs (n=76), dental (n=92) and dental hygienist (n=13) students, administrative staff (n=15, reference group), who were not previously vaccinated against TB, provided their consent to undergo the Mantoux test, the standard recommended TST. Unadjusted and adjusted odds ratios (ORs) of reactive TST for working categories and for years of practice in the dental hospital were assessed through logistic regression analysis. Results. Prevalence of reactive TST estimates were 0.0%, 17% (95% confidence interval, 95CI, 11-26%), 20% (95CI, 12-30%) and 7% (95CI, 1-30%) among dental hygienist students, dental students, dental healthcare workers and administrative staff, respectively. The adjusted ORs for working categories were non-significant, but they were significant at 95% level for the years of exposure (OR, 1.09 for each year of practice at the hospital; 95CI, 1.01-1.19). Conclusion. High prevalence values could be due to cross immunization from non-tubercular mycobacteria, frequently found in dental unit waterlines, or boosting from repeated TST occasions. However, these data suggest that the risk for TST conversion is associated with the years of practice in dental hospitals.


Introduction
Mycobacterium tuberculosis is carried in airborne particles called droplet nuclei that are generated when persons with infectious tuber culosis (TB) disease cough, sneeze, shout or talk.During these events, a multitude of drop lets lower than 50 μm in size from the upper aero-digestive tract, known as aerosol, are emit ted in the environment and rapidly desiccate, thus becoming droplet nuclei.They are 1-5 μm in size, are suspended in air for hours and are carried through air currents from a room to an other.The probability of a person exposed to M. tuberculosis to become infected depends on the concentration of infectious droplet nuclei in the air and the duration of the contact with the source of infectious TB disease.Environmental factors, such as long stay in confined spaces, inadequate ventilation or recirculation of air containing infectious droplet nuclei, along with poor general health status of exposed individu als, may increase the likelihood of transmission.The persons at highest risk for M. tuberculosis infection are, therefore, individuals who share air space in households or in other indoor envi ronments with persons with active pulmonary TB disease 1 .After initial infection, immune re sponse generally prevents further multiplication and spread of M. tuberculosis and the subjects do not show symptoms of the disease, although bacteria can remain alive in lungs or other or gans for years.This condition is known as la tent TB infection, which can be diagnosed by reactive tuberculin skin test (TST).Some 5% of infected individuals develop active TB disease soon after infection, while another 5% develop active TB later in life in consequence of condi tions, such as AIDS or malnutrition, associated with immune depression.Symptoms of active TB disease include productive cough, night sweats, fatigue, malaise, fever and weight loss.These are the only contagious TB subjects 2,3 .
The risk for M. tuberculosis infection trans mission in dental healthcare settings could be associated with aerosol production, which is frequent during dental therapy as demonstrated since more than forty years ago 4 .In the year 1970, professional toothcleaning with ultrason ic scaler, which generates abundant aerosol and splatter, was performed on the labial surfaces of the mandibular anterior teeth of patients with suspected active TB disease.The purpose was to determine whether in these circumstances droplet nuclei containing M. tuberculosis were produced in the environment.These microor ganisms were not detected in air samples col lected immediately after the dental prophylaxis.This experiment, which would not be feasible nowadays, suggests that production of M. tu berculosis containing droplet nuclei from pa tients with active TB disease probably occurs only when fluids from the lower respiratory tract are emitted, while during dental therapy fluids from the upper respiratory tract and, most of all, saliva are emitted 5 .However, the risk for M. tuberculosis infection and TB transmission to patients and staff cannot merely be assessed through the investigation of the spread of mi croorganisms in the dental environment, but it is necessary to study the ascertained cases of transmission during dental treatment 6 .Actu ally, an outbreak of TB was reported in the year 1982 at two community dental clinics: primary lesions developed in fifteen patients follow ing dental extractions.Thirteen of the lesions were oral TB, while one case was an active TB disease which involved mouth and lungs.The dental surgeon who did not routinely use protective barriers was the source of infec tion 7 .Such a case could not occur nowadays since almost all dental healthcare wrokers use masks and other protective barriers and, if af fected by active TB disease, are prevented from working.This situation is common with many other infectious diseases 8,9 .A dentist with ac tive pulmonary TB disease who treated at least 450 patients, including 60 children, during the period elapsed between the appearance of the first symptoms and the final active TB diagno sis, did not transmit M. tuberculosis to any of his patients probably because he wore mask routinely 10 .The TB risk for patients is minimal if dental healthcare workers adopt few simple preventive measures 3 , but a different issue is the risk for dental staff.Indeed, TB was consid ered an occupational hazard for this category of healthcare workers few years ago 11 .A survey made in England and Wales in 1988 and 1993 seeking to determine whether healthcare work ers were at increased risk of TB, reported that there were 48 health professionals with diag nosis of active TB and the risk for this disease was almost ten times higher than in the general population.However, only one of the diseased individuals was a dentist 12 .Conversely, accord ing to another survey made in 1995 among the members of the American Association of Hos pital Dentists, 12% of respondents reported that there was at least one case of TST conversion by dental healthcare providers within the past year at their institution 13 .These data suggest that the risk of M. tuberculosis infection would be particularly high among dental healthcare workers who practice in hospitals.

Aim
The aim of this study was to investigate prevalence of reactive TST in a sample of den tal students, and dental healthcare associates and professionals in a large university hospital in Italy.

Material and Methods
This study was made in the year 2008 in a large dental hospital in the centre of Italy, where dental students perform their clinical training.The clinical and administrative staff was invit ed to participate and to provide their informed consent, while the internal review board of the hospital approved the study.In addition, den tal students at their fourth and fifth (i.e., last) course years, as well as dental hygienist stu dents at their second and third (i.e., last) course years were invited to participate.Criteria for el igibility were that subjects must not be affected by diabetes, kidney diseases, Crohn's disease, rheumatoid arthritis, cancer chemotherapy, conditions which may produce unclear TST re sults.Volunteers were also excluded if they had a history of severe reaction to a previous TST, were vaccinated against TB with the Bacille of Calmette-Guérin (BCG) a strain of Mycobacte rium bovis, or had fever the day of the testing occasion.The remaining participants provided their data and underwent the TST.
The TST used was the Mantoux test (Biocine Test PPD Liof 2D 5UI/D -Novartis Vaccines and Diagnostics, Siena Italy-this product is not available anymore in Italy since April 2009).
TST tests are based on the fact that infection with M. tuberculosis produces a delayed-type hypersensitivity skin reaction to certain compo nents of the microorganism.These components are contained in extracts of culture filtrates and are the core elements of the classic tuberculin Purified Protein Derivative (PPD), which is then used for skin testing.The standard rec ommended tuberculin test, known as Mantoux test, is administered by injecting a 0.1 mL of 5 tuberculin units of PPD intradermally in the forearm, choosing an area free of abnormali ties and away from veins.Reaction in the skin is due to sensitized T cells which induce indu ration, that is, a hard, raised area with clearly defined margins around the injection site.The Mantoux test was preferred to the widespread and cheaper Tine test, a multiple puncture test, because the latter provides a higher number of false positive results and tends to overestimate TST positives 14 .
Skin tests were read 48-72 h after the injec tion when the size of the induration was maxi mal.Reactive (positive) TST among students and the administrative staff with normal im mune system must show at least one indura tion greater than 15 mm, while among dental healthcare workers, indurations ≥10 mm were considered reactive TST.Since in this study the response to the TST was crucial, it was neces sary to obtain the most reliable results as pos sible.For this reason, the test was performed by one trained healthcare worker with more than ten years of experience and a specific TST cal liper was used to read the induration size.These two measures, guaranteed the consistency of test results 15 .
Differences in mean age, gender distribu tion and average years of hospital practice (i.e., exposure) among the four categories of dental healthcare workers, dental students, dental hy gienist students and administrative staff mem bers were statistically analysed using the χ2 test (cell χ2 test for pairwise differences) and the analysis of variance -ANOVA (Bonferroni-Dunn post hoc test for pairwise differences).Prevalence of reactive TST with 95% confi dence interval (95CI) and the unadjusted odds ratios (ORs) for reactive TST were estimated within the four categories, with the administra tive staff as reference group.Logistic regression analysis also was used to estimate the ORs for reactive TST test within the four categories, ad justed for gender, age and years of exposure.In order to prevent that excess collinearity may ar tificially inflate the coefficient estimates, Pear son's correlation coefficient between the two continuous variables, that is, age and years of exposure, was estimated.In the event that coef ficient was higher than 0.5, the two variables were used in two different regression models.A level of significance of 95% was chosen.

Results
Four dental healthcare workers were not in cluded because they reported to be vaccinated with BCG.The sample, therefore, included 76 dental healthcare workers, 15 administrative staff members, 92 dental students and 13 den tal hygienist students.Gender distribution was similar in the four categories, with males pre vailing on females, excluding dental hygienist students who were 77% females, a statistically significant difference with the other catego-  1).Mean age was approximately of thirty years in all groups excluding dental stu dents (26 years, statistically significant differ ence with the remaining groups).Years of ex posure to the dental hospital settings was zero in the administrative staff and nine years in the healthcare workers, significantly higher than in the student categories.
Prevalence of reactive TST in the admin istrative staff, used as reference unexposed category, was almost 7% (Table 2) with wide confidence interval due to the small size of this group.Similar situation for dental hygienist students, with 0% reactive TST.Prevalence was considerably high among dental students (17%) and healthcare workers (20%).These values led to high unadjusted ORs for reactive TST of 3 and 3.4, respectively.The Pearson's correlation coefficient be tween age and years of exposure was 0.65 (Zvalue, 10.58; p<0.0001) (data not in Table ).It was, therefore, impossible to include these variables in the same logistic regression model, because this option would produce excess col linearity.Two models were then designed in cluding the four working categories, gender and age (model 1) and the four working categories, gender and years of exposure (model 2).The results, displayed in Table 3, refer to model 1 for all the coefficient estimates excluding for the covariate "years of exposure" which was estimated through model 2. The coefficient es timates for the working categories and gender did not differ appreciably between model 1 and model 2. In addition, the analysis of the whole model fit, assessed through the likelihood ratio test, provided similar values of the statistic test (χ25df=11.63,model 1; p=0.04 -χ25df=12.24model 2; p=0.03), thus suggesting that the dif ference between models was minimal.The only variables significantly associated with reactive TST were age (OR=1.08,95CI, 1.01-1.15for each increasing year) and years of exposure (OR=1.09,95CI, 1.01-1.19for each increasing year).
As for the latter variable, the probability to be TST positive at the Mantoux test increased by 1%-19% for every working year in the den tal hospital.The variable years of exposure was, therefore, more meaningful than the work ing category.
This important result induced to investigate the effect of years of exposure to the hospital setting on reactive TST probability.This vari able was split into four categories of exposure, namely, 0-2 (reference category), 3-5, 6-10, >10 years.The unadjusted and adjusted ORs for reactive TST in these categories were assessed using simple and multiple logistic regression analysis.The covariates included in the mul tivariate model were gender and working cat egory.The results, displayed in Table 4, suggest that the risk for reactive TST was marginally significant (p-values ranging between 0.05 and 0.07 -data not in Table ), within the first ten years, while it was statistically significant and

Discussion
The present study is one of the papers pre sented at the workshop "Advances in Infection Epidemiology and Control in Dental Healthcare Settings", Department of Public Health and In fectious Diseases, Sapienza University, Rome, Italy on February 9th, 2013 1622 .
The main limit of this study lies in the small sized sample, particularly in the control group.Indeed, the administrative staff included only 15 subjects, which resulted in low level of pre cision of the risk estimates.Therefore, high point risk estimates, such as those displayed in Tables 3-5, had wide confidence intervals and did not allow to make robust inference regard ing the risk for TST conversion in dental hos pitals.
Nevertheless, there were two important re sults, which could not be ignored.High preva lence of reactive TST in last course year dental students (17%, Table 2) and the increasing risk for TST conversion with the increasing years of practice in the hospital.
These data are comparable with published data regarding dental healthcare workers.The cited survey made in 1995 by the American Association of Hospital Dentists reported that cases of TST conversion within their institu tion were reported by 12% of respondents.This, of course, does not mean that there was a 12% conversion rate in hospital dentists from US, but that conversion was not unusual 13 .An Italian survey made in the year 2005 reported that 17% dental healthcare workers, including those vaccinated against M. tuberculosis, from a dental hospital were reactive for TST assessed using the Tine test 23 .According to a survey among 474 dentists in the Indian Health Service and the Federal Bureau of Prisons (US), 7.6% resulted reactive for TST and 3.6% reported that conversion occurred after graduation, suggesting that infection was acquired during practice among aborigines or in prisons 24 .An other survey made in the year 1995 compared self-reported reactive TST subsequently treated with antituberculosis medication occurred dur ing the period of practice among Air Force den tists and lawyers.2.4% and 1.5% dentists and lawyers, respectively, reported to have required such medication 25 .Another study made among dentists working in US-Mexico border coun ties of Texas where TB prevalence was higher than in other US counties reported an annual TST conversion rate of 1.7% 26 .These data col lectively suggest that the risk for reactive TST among dental healthcare workers depends on the characteristics of patients, specifically, rel atively high risk in hospitals, as also reported in the present study 13,23 or in areas where TB prevalence is high 26 low risk for general dental practitioners 24,25 .
Paradoxically, dental students are more frequently investigated than dental healthcare workers, probably because in most countries dental students are invited to check their TB sta tus before admission or before beginning their clinical training with direct patient contact.In addition, negative TST subjects are invited to undergo one or more BCG vaccinations, be cause clinical training is often made in hospi tals, where patients are more likely to be affect ed by active TB.A previous study from Italy reported that 14% dental students at their lastcourse years were positive at the Tine test, in cluding some vaccinated subjects as compared with 2% of controls who were not healthcare workers 23 .Another study from Japan reported that 60% medical/dental students were positive It is important to highlight, however, that such a high prevalence of positive TST among dental healthcare workers and dental students could be due to cross immunization between M. tuberculosis and other environmental my cobacteria, defined non-tubercular mycobac teria (NTM) isolated from water, as cases of TST conversion attributable to Mycobacterium scrofulaceum and Mycobacterium intracellu lare are reported among medical students 31 and aerosols coming from dental unit waterlines are often contaminated by NTM 32 .Another possible problem of TST is that in individuals who undergo frequent tests, such as healthcare workers and medical/dental students, "boost ing" may occur.Boosting is a T-cell response due to repeated injections of PPD antigens.Such a response induces TST conversion even in absence of BCG vaccination or M. tubercu losis infection 33 .Therefore, the high prevalence reported in this and other studies among dental healthcare workers and dental students could be partly due to cross immunization from NTM or boosting from repeated TST occasions.

Conclusion
The data from this survey and from the lit erature suggest that prevalence of reactive TST among dental healthcare workers from hospi tals and dental students during clinical training is relatively high.Such a high prevalence in creases with the years of practice in dental hos pitals.However, these data do not necessarily mean that the risk for M. tuberculosis infection is high because the cases of active TB among these working categories are few, and because frequent inhalation of NTM from aerosol gen erated by dental equipment and repeated TST occasions may artificially induce reactive TST conversion in absence of M. tuberculosis infec tion.

Table 1 .
General characteristics of the sample.
Gender: χ23df test = 12.79; p=0.005 -pairwise differences: dental hygienist students with the three other categories, p<0.05 Age: ANOVA F-test = 24.97;p<0.0001 -pairwise differences: dental students with the three other categories, p<0.05 Years of exposure: ANOVA F-test = 52.02;p<0.0001 -pairwise differences: healthcare workers with the three other categories, p<0.05 ries (Table

Table 2 .
Prevalence of reactive tuberculin skin test (TST) among the four categories and unadjusted odds ratio for positive test.

Table 3 .
Multiple logistic regression analysis (adjusted OR) of the association between reactive TST and the fourworking categories, gender, age and years of exposure to the dental healthcare setting.(unadjusted OR, 9.38, 95CI, 2.26-38.90;adjusted OR, 10.10, 95CI, 2.38-42.85)when the years of exposure were more than ten.
*p<0.05 Whole model fit.Likelihood ratio χ25df = 11.63;p=0.04 [model with years of experience and without age]; χ25df = 12.24; p=0.03 [model with age and without years of experience] high

Table 4 .
30mple logistic regression analysis (unadjusted OR) of the association between reactive TST and the four categories of exposure to the healthcare setting.Multiple logistic regression analysis (adjusted OR) including gen-der and the four working categories as covariates.Mantoux test, but 92% of the sampled subjects were vaccinated with BCG between one and four times.Despite such a high vacci nation rate, 1% of them was considered suspi cious of having an episode of TB infection27.A survey performed in 1994 in US, where dental students were invited to read the results of their Mantoux test by themselves, reported that 7.4% of them declared to be reactive to TST28.In the same year, 8.7% US dental students resulted reactive TST before starting their clinical train ing, with conversion rates of 4.2% during the first year and of 2.2% during the second year29.A last survey from US in 1994 reported a con version rate of 10.6% after the first year of in tensive clinical training30.Collectively, these data suggest that dental students are at risk for TST conversion and, perhaps, for M. tuberculo sis infection during the period of clinical train ing, with conversion rates or prevalence values at the end of this period, which may be as high as in dental healthcare workers.