Elevated Risk of Tuberculosis by Occupation with Special Reference to Health Care Workers

We conducted a study to evaluate tuberculosis (TB) risk in Japan by work performed, either paid or unpaid. We collated information on sex, age, employment category, occupation, and family history from 1120 registration cards of new TB cases at two wards in Nagoya City over seven years (1989-1995). We used census data and data from the Survey of Physicians, Dentists and Pharmacists conducted in 1990 to estimate the population at risk by employment category and occupation. Elevated TB incidence rates were observed for female nurses (SIR: 3.81; 95%Cl: 1.97-6.65), clinical laboratory technicians (SIR: 25.00; 6.81-63.99), and males without a paid job (SIR: 1.35; 1.20-1.53). A work environment conducive to transmission may have increased the TB risk in female nurses and clinical laboratory technicians. Male jobless people and institutionalized elderly residents may have enhanced the TB risk for males without a paid job. J Epidemiol, 2000 ; 10 : 1-6

We conducted a study to evaluate tuberculosis (TB) risk in Japan by work performed, either paid or unpaid.We collated information on sex, age, employment category, occupation, and family history from 1120 registration cards of new TB cases at two wards in Nagoya City over seven years (1989)(1990)(1991)(1992)(1993)(1994)(1995).We used census data and data from the Survey of Physicians, Dentists and Pharmacists conducted in 1990 to estimate the population at risk by employment category and occupation.Elevated TB incidence rates were observed for female nurses (SIR: 3.81; 95%Cl: i.e., 257.1 per 100,000.However, since World War II the mortality rate has been decreasing rapidly, reaching 2.6 per 100,000 population in 1995.Outbreaks of TB tend to be concentrated in specific classes and populations 1).Even in communities with relatively low TB incidence rates, certain occupations seem to be associated with a higher risk 2).An increased incidence of TB has been observed in particular types of occupations which are invariably associated with an elevated risk of TB including impoverished and unskilled workers, occupations which increase susceptibility to the infectious organism, such as in mining and pottery work, and occupations which may increase the likelihood of exposure to transmission, including working in acute care hospitals, extended care facilities, and mycobacteriology laboratories 3).
Notwithstanding the interest in work-related TB, there are few recent studies in Japan investigating the risk of TB across a range of occupations.We thus conducted a study to deter-mine the risk of TB by work performed (paid or unpaid), especially in specific health care jobs.

SUBJECTS AND POPULATIONAT RISK
Utilizing the registration cards of TB cases (TB cards), we collated 1120 TB cases (726 males and 394 females) who had been newly registered at the Nakagawa and Meito Health Centers in Nagoya City over seven years (1989)(1990)(1991)(1992)(1993)(1994)(1995).We confirmed information on sex, age, employment category, occupation, and family history.
Data from the Survey of Physicians, Dentists and Pharmacists conducted in 19904) were used to estimate the population at risk in physicians, while national census data in 1990 5) were in other professional employment categories.When the cases changed their employment categories or occupations because of the onset of TB, we classified them into their prior employment or occupation category.

INCIDENCE RATE AND STANDARDIZED INCIDENCE RATIO
The TB incidence rates by sex and age group were calculated according to each employment category and occupation.The number of incident cases between 1989 and 1995 was divided by the population at risk multiplied by seven based on the census data together with data from the Survey of Physicians, Dentists and Pharmacists conducted in 1990.
The standardized incidence ratios (SIRs) were calculated by dividing the observed numbers of TB cases in each employment category or occupation by the expected numbers, computed by multiplying the populations at risk based on census data in 1990 by the TB incidence rates of the total population by sex and age group at two wards over seven years.Ninetyfive percent confidence intervals (95%CI) for the SIRs were estimated by assuming that the observed frequencies were distributed according to a Poisson distribution 6).

NUMBER OF NEWLY REGISTERED TB CASES
There were 726 male TB cases newly registered at the two wards during the study period (

INCIDENCE RATE
The age-specific TB incidence rates for the male PWOP 20 years or over were 1.5-3 times higher than those for PWP (Table 5).Similarly, the age-specific TB incidence rates for female PWOP 40 years or over were higher than those for PWP (Table 6).Elevated TB incidence rates by occupation were observed for male physicians (81.1 per 100,000), male drivers (82.5 per 100,000), female nurses (91.6 per 100,000) and female clinical laboratory technicians (570.0 per 100,000).

TB FAMILY HISTORY
The number of cases who had a family history of TB was 71 (9.8%) out of 756 males, and 70 (17.8%)out of 394 females.Two nurses with a TB family history showed no significant difference from the expected number (2.45) when checked by chi-square test.In each employment category and occupation, there were no statistically significant differences in the rates of cases with a family history of TB.

DISCUSSION
Studies using data of disease registries abroad from the 1960s through the 1980s reported TB risk ratios of 0.6 to 2.0 for health care workers, indicating at most a modest increase in Table 5. Incidence rates of registered male TB cases per 100,000 by age group, employment status and occupation at Nakagawa and Meito Wards in Nagoya City (1989-1995).
Table 6.Incidence rates of registered female TB cases per 100,000 by age group, employment status and occupation at Nakagawa and Meito Wards in Nagoya City (1989-1995).
Table 7. SIRs of registered male TB cases by employment category and occupation at Nakagawa and Meito Wards in Nagoya City (1989)(1990)(1991)(1992)(1993)(1994)(1995).risk [7][8][9][10][11].The TB risk for clinical laboratory technicians was not elevated in U.S. from 1984 to 1985 (SIR: 0.9 (0.5-1.5)) 2).However, our study showed a significantly high TB risk for female nurses and clinical laboratory technicians.Ikari also reported a risk ratio of 1.8 for nurses aged 20 through 29 years in Aichi Prefecture using data from TB cards 121.Similarly, a relative risk of 3.1 was reported in Osaka Prefecture 13).Suzuki et al. estimated a risk ratio of 6.5 for clinical laboratory technicians from their questionnaire survey to assess the incidence of TB among health care workers in hospitals in Chiba Prefecture 14).
Four technicians with no family history of TB in our study engaged in bacteriological or physiological examinations.Two of them took part in bacteriological examinations for TB, which may account for their infection by TB bacteria.As age-specific TB incidence rates for male physicians 20 through 40 years were 2-5 times higher than those for the general population, an additional study may be needed to clarify the associated factors, if any.The SIR for other health care workers (nursing aids or student nurses who work closely with TB cases) was elevated, but not statistically significant.Additional studies of this group are also warranted.Thus, health care workers in Japan seem to have had a higher risk than those in Western countries 2, 7-14).We may have failed in controlling nosocomial transmission of TB.Measures for its control must be established immediately in Japan.
Male PWOP consisting of jobless people, retirees and institutionalized individuals demonstrated an increased risk of TB.Our findings seemed to be compatible with the earlier observations.The odds ratio of TB was significantly elevated for male jobless in Canada 15), and a chronic jobless status remained strongly and independently associated with the risk of TB in U.S. 2).The TB incidence rate for nursing home residents in U.S. was 1.8 times higher than the rate seen in elderly persons who were living in the community 16).
Nine homeless males were registered.However, their incidence rate could not be determined because of uncertainties surrounding the homeless population in these two wards.Yamanaka et al. 17.18) estimated the incidence rate among the homeless in Nagoya City to be 20-30 times higher than that among the non-homeless.No decrease in the incidence of TB cases among the homeless was seen in recent years, while the rate among the non-homeless decreased gradually.Besides TB among health care workers, TB among the homeless is another major problem in large cities.
The SIR for drivers was higher than unity.The relative risk of mortality for TB in workers in transport and communications occupations versus clerical and related workers was 4.6 in Japan 1990 19).As the great majority of the former workers consisted of drivers, the higher SIR for drivers may have contributed to the relative risk of mortality.
1.97-6.65),clinical laboratory technicians (SIR: 25.00; 6.81-63.99),and males without a paid job (SIR: 1.35; 1.20-1.53).A work environment conducive to transmission may have increased the TB risk in female nurses and clinical laboratory technicians.Male jobless people and institutionalized elderly residents may have enhanced the TB risk for males without a paid job.
Tuberculosis (TB) had been prevalent in Japan before World War II .The mortality rate from TB in 1918 was very high:

Table 1 .
Male population at risk by age group, employment status and occupation at Nakagawa and Meito Wards in Nagoya City (1990).

Table 2 .
Female population at risk by age group, employment status and occupation at Nakagawa and Meito Wards in Nagoya City (1990) .

Table 3 )
. Among them, seven health care workers consisted of four physicians (two working in hospitals and two in clinics), one dentist, one nursing aid and one student nurse.Most of the 34 drivers were truck or taxicab drivers.The 357 cases in other occupations consisted of 232 full-time employees, 95 self-employed persons or professionals, 16 temporary employees, and 14 day laborers.PWOP were all PUE including 27 institutionalized elderly residents.The 394 female TB cases newly registered (Table4) included 22 health care workers, consisting of 12 nurses, four clinical laboratory technicians, and six other health care workers.All the nurses worked in hospitals; one in a TB hospital, one in a psychiatric hospital, and 10 in general hospitals.Eight nurses worked on hospital wards, one in the consultation room of a hospital, and three in unclear hospital locations.Two clinical laboratory technicians worked in hospitals specialized in TB, and two worked in general hospitals.Four technicians partici-

Table 3 .
Number of registered male TB cases by age group, employment status and occupation at Nakagawa and Meito Wards in

Table 4 .
Number of registered female TB cases by age group, employment status and occupation at Nakagawa and Meito Wards in