Tuberculosis among aboriginal and nonaboriginal persons in British Columbia

Correspondence: Dr J Mark FitzGerald, British Columbia Centre for Disease Control Society, Division of Tuberculosis Control, 655 West 12th Avenue, Vancouver, British Columbia V5Z 4R4. Telephone 604-660-6127, fax 604-660-1950, e-mail markf@interchange.ubc.ca L Wang, K Noertjojo, RK Elwood, JM FitzGerald. Tuberculosis among aboriginal and nonaboriginal persons in British Columbia. Can Respir J 2000;7(2):151-157.

45.7±19.7),differences in the prevalence of lymphadenopathy (5.9% versus 16.4%, P=0.0008) and pleural disease (21.3% versus 16.4%, P=0.00008), there were no differences in presentation between aboriginal and nonaboriginal people.Aboriginal people were more likely to have a history of contact with a patient with TB (53% versus 17.9%, P<0.05), to have received directly observed therapy (55% versus 33.8%, P=0.00002) and to have contacts who were purified protein derivative (PPD) positive (4±9 versus 2±3, P=0.002).These contacts were more likely to start isoniazid (2±3 versus 1±1, P=0.002).Overall, there was a significant decline in rates of TB among aboriginal persons compared with the general population, but there was a small increase in rates among all subjects in the final year of the study.CONCLUSIONS: In the present study, significant variations in rates of TB among different population groups in British Columbia were found.During the study period, there was a greater decline in the rates of TB among aboriginal persons.A greater use of directly observed therapy and greater use of chemoprophylaxis occurred among aboriginal persons, which may have contributed to this decline, or alternatively, it simply reflects the natural evolution of the TB epidemic.
R ates of tuberculosis (TB) in Canada, after declining for many years (1), plateaued in the late 1980s (2) and have only recently begun to fall slightly again (3).Historically, rates of TB among aboriginal Canadians have been significantly higher than those among the general population (4).Globally, TB is a major cause of morbidity and mortality (5), particularly in developing countries where the rates of the disease have been increasing, especially in areas where there is significant coinfection with HIV (6).
In British Columbia, as in other regions, the rates of TB among aboriginal Canadians have traditionally been much higher than in the general population.Because of this, we have attempted to treat all cases of active TB among aboriginal persons with directly observed therapy (DOT) and to use chemoprophylaxis more extensively among TB-infected aboriginal persons in British Columbia.During the study period, the rates of TB have declined significantly (7) while the rates of TB among the general population have remained static or increased slightly.We decided, therefore, to review the characteristics of aboriginal TB patients, both on and off reserves, and compare them with a random sample of nonaboriginal patients.

PATIENTS AND METHODS
In British Columbia, all cases of TB are reported to the Division of Tuberculosis at the British Columbia Centre for Disease Control in Vancouver.Both the provincial laboratory for the province and the pharmacy for dispensing TB drugs throughout the province are located in this division.In British Columbia, the Aboriginal TB Control Program is run as a separate unit within the larger provincial program.When diagnosed, patients are classified as aboriginal on reserve, aboriginal off reserve, Canadian nonaboriginal and foreign born.For the present study, all cases of active TB that occurred among aboriginal Canadians from January 1992 to December 1996 were identified.For each case of TB diagnosed on and off reserves among aboriginal Canadians, a control patient diagnosed with TB in the same month was picked randomly from the TB registry.This registry contains information on all cases of TB diagnosed in BC.Data are collected in a systematic manner using standardized data sheets.Control subjects were selected randomly from nonaboriginal TB patients diagnosed the same month as the aboriginal patients.Information collected included age, sex, symptoms, whether patients were admitted to hospital, duration of hospitalization and site of disease, whether pulmonary or extrapulmonary, or if both sites were involved.
Extrapulmonary disease was classified as lymphadenopathy, pleural, bone and joint, abdominal or other.In addition, the method of diagnosis was outlined.These included smear and culture positivity, histology consistent with the diagnosis or if the diagnosis was based on clinical and radiological features characteristic of TB in which there was an appropriate response to therapy.Bacteriological information included the percentage of patients with a positive sputum smear, or bronchoalveolar lavage or other body fluids.In addition, sensitivity patterns to first-line drugs including isoniazid (INH), rifampin, pyrazinamide, ethambutol and streptomycin were evaluated.
Historical information included whether there was a prior diagnosis of TB, a history of TB treatment, a contact history or known exposure, a prior tuberculin test or a prior positive tuberculin test.A history of bacille Calmette-Guerin (BCG) vaccination was identified as were risk factors for the reactivation of TB infection.The number of contacts identified per patient was documented.In addition, the percentage of patients considered to have a positive tuberculin test (5 mm or greater in duration to 5TU purified protein derivative) was documented.The number of contacts initiating chemoprophylaxis was also identified.Rates of disease were based upon population data from Statistics Canada for the nonaboriginal population and the aboriginal off-reserve population.Aboriginal on-reserve data were based on population figures supplied by the Medical Services Branch, Health Canada, Ottawa, Ontario, responsible for health care for this population.Statistical analysis: This analysis was performed with the SPSS PC for DOS version 4.0 (SPSS Inc, Chicago, Illinois) and SPSS for Windows version 6.1.3(SPSS Inc) statistical software.c 2 analyses were performed on categorical variables, and a two-sided t-test and ANOVA were performed on continuous variables.A Bonferroni method of multiple comparison adjustment was used when necessary.P<0.05 was adopted as the level of significance (8).The curve estimation procedure on SPSS for Windows was adopted to analyze trends for the rate of TB across time.Regression analysis was performed to identify differences between both groups.

RESULTS
During the study, the rates of disease declined significantly among aboriginal Canadians, both on and off reserves (91/100,000 to 27/100,000, P<0.05).The decline in rates was most notable among on-reserve patients, with the rate falling from 112/100,000 to 17/100,000, while in off-reserve patients the rates declined from 71/100,000 to 36/100,000.The rates for nonaboriginal Canadian-born and foreign-born people remained essentially unchanged during much of this period but rose slightly between 1995 and 1996 (Figure 1).A total of 202 cases of TB were diagnosed among aboriginal persons and 201 control subjects were chosen from cases of TB diagnosed among Canadian nonaboriginal and foreignborn patients.The characteristics and demographics of both aboriginal and nonaboriginal patients are shown in Table 1.Aboriginal patients were more likely to be younger (35.1±20.0versus 45±19.7 years of age, P<0.001).Patients' symptoms as well as number of subjects admitted to hospital did not differ between the groups.In Table 2, the type of disease and bacteriological results are shown.Nonaboriginal patients were more likely to have disease of the lymph nodes (14.4% versus 5.9%, P=0.001) and less likely to have pleural disease (7.5% versus 21.3%, P=0.001).There was no difference in the proportion of patients with culture-proven disease.A history of TB of either diagnosis or treatment and predisposing factors, if any, are shown in Table 3

DISCUSSION
In the present paper, we have outlined the characteristics of TB cases that occur among aboriginal persons, both on and off reserves in British Columbia, and compared them with nonaboriginal Canadian-born and foreign-born subjects.Our primary interest was not to compare separately Canadian-born nonaboriginal and foreign-born patients with aboriginal patients.Although we did complete this analysis, the differences related primarily to the higher rate of lymphadenopathy, the greater likelihood of resistance and the greater use of ethambutol in foreign-born persons.There were 118 foreign-born and 90 Canadian-born people in the comparison group.It has been well documented that primary disease is more likely to occur among aboriginal Canadians (9) and that foreign-born people, especially from Asian countries, are more likely to have TB lymphadenopathy (10).These differences did not change over time and are unlikely to account for the decline in TB rates seen among the aboriginal group.Although there was an excess of TB cases in the early 1990s related to clustering on reserves, when we completed an analysis of trends, the significant differences in rates of decline persisted over the study period.Specifically, we compared rates from 1994 to 1996, and the statistically   significant trend (P<0.024)persisted for both on-reserve and off-reserve patients.We identified no clustering off reserves nor an increase in the absolute number of cases among off-reserve subjects during the study period.Interestingly, there was a much higher proportion of aboriginal persons who had a history of contact and a prior positive tuberculin test.These results indicate that further reductions in the rates of aboriginal TB could be achieved with even more aggressive use of chemoprophylaxis.Not surprisingly, a greater proportion of aboriginal persons had BCG vaccination, and this is consistent with the fact that BCG offers only partial protection against the development of TB and that this protection is time limited (11).Although the efficacy of BCG is controversial, it is unlikely to have affected the decline seen among aboriginal persons because the use of BCG did not change during the study period.In general, risk factors for TB were equivalent in both groups, with a notable exception of alcoholism occurring in over 25% of aboriginal subjects compared with just under 12% among nonaboriginal persons (P<0.001).Interestingly, one would have anticipated that this would have adversely affected the efficacy of the control program, but this did not appear to be the case.Infection with HIV as a risk factor occurred more commonly among nonaboriginal persons, but a significant proportion of both aboriginal and nonaboriginal patients were known to be HIV positive.Also, a significant number of patients were from a group putting them at risk of being HIV positive (especially intravenous drug users) or had confirmed HIV as a risk factor for their TB.In all, 63 of the 403 patients studied (15.6%) fell into these high risk groups.This increase is consistent with a change from previous (12) and more recent reports on HIV-related TB in British Columbia (13), outlining a greater tendency for AIDS-related TB to occur among intravenous drug users and aboriginal persons.This increase in the number of HIV-related TB cases over time highlights the need for targeted surveillance among those at risk of dual HIV-TB infection, especially aboriginal persons.We have also identified clustering of TB cases in these same groups (14).
Both groups were as likely to have bacteriologically proven disease, but there were significant differences in the resistance pattern between the groups studied, with the nonaboriginal persons much more likely to be resistant to isoniazid, rifampin or streptomycin.Multidrug resistant disease was not documented during the study period.Our data with regard to the risk of resistance are consistent with previous reports of resistance patterns in TB cases in western Canada (15), and reflect the impact on immigrants from high prevalence countries where drug resistance is a major problem.The pattern of resistance was unchanged during the study period and therefore is unlikely to account for the differences in rates of TB seen.
Recent attention has focused on the benefits of directly observed therapy versus self-administered therapy (16).A number of reports have been presented recommending the more widespread use of directly observed therapy (17,18).There are limited data, however, to justify its use, although some have argued persuasively with regard to the financial benefits of this type of therapy delivery (18).A recent randomized, controlled trial study from South Africa showed no benefit from the use of directly observed therapy compared with self-administered therapy, but this study has been criticized because of the relatively small percentage of eligible subjects enrolled (19).
In the present study, we found significantly more aboriginal persons received directly observed therapy than did nonaboriginal persons.Given the equivalent characteristics of the patients at the time of presentation, greater use of directly observed therapy may be a factor that accounts for the decline in rates of TB among aboriginal persons both on and off reserves.Contact tracing among aboriginal persons was also more successful.Unfortunately, for the early part of the study, data regarding chemoprophylaxis offered were not coded for nonaboriginal persons.In addition to greater completion rates with directly observed therapy and greater assurance that the therapy has in fact been taken, we have shown recently greater efficacy in terms of completion of medication when directly observed chemoprophylaxis is used (20) among the aboriginal population.This greater surveillance may reflect the fact that aboriginal persons diagnosed with TB, particularly on reserves, live in relatively small communities, and because of the high risk of clustering in these populations (21), the entire communities may be surveyed.We have argued previously that, in high prevalence groups, there should be caution in initiating chemoprophylaxis (22) unless there is a satisfactory program in place to deliver therapy to patients with active cases of TB, especially those that are smear positive.

CONCLUSIONS
We have compared the presentation and management of TB among aboriginal and nonaboriginal persons in British Columbia.In the early part of this study, there was a significant reduction in rates of TB among aboriginal persons in British Columbia.It is unclear if the decline in rates of disease is primarily related to the greater use of directly observed therapy and chemoprophylaxis in the high risk population, or is part of the natural decline in rates expected at this stage of the epidemic among aboriginal persons.More recently, this decline has been reversed in all groups, including aboriginal persons, and indicates a need for continued vigilance and strategies to improve control in all groups.This cautionary note needs to be emphasized particularly for aboriginal persons, given the relatively high prevalence of TB infection in this population, and because recent epidemiological surveillance that we have undertaken indicates a disproportionately high HIV sero-prevalence among aboriginal intravenous drug users in Vancouver (23) compared with the nonaboriginal population.There is a significant risk that these encouraging results may not be sustained.The risk of dual HIV-TB infection is such that ongoing, targeted surveillance of this high risk population will need to persist if the currently reported declining rates of TB among aboriginal persons in British Columbia are to continue, and we are to avoid the documented rising rates of HIV-related TB reported from New York City and other large urban centres in the United States (24).Further studies, especially randomized, controlled trials with appropriate cost effectiveness analysis, are needed urgently to support our recommendations that directly observed therapy should be used on a more widespread basis.Such studies should also account for the significant gaps in knowledge about TB that patients have (25) and the added challenge of controlling TB in the inner city (26).

ACKNOWLEDGEMENTS:
We acknowledge the assistance of Ms Jean Fraser and her staff of the TB registry.We also acknowledge the hard work of the many nurses and outreach workers who participate in the provincial TB program.We thank Ms Fay Hutton for helping with data collection and analysis.Dr Lai Wang was funded by the World Bank.

TB among aboriginal people in British Columbia
. Aboriginal persons were more likely to have a contact history (53% versus 17.9%, P<0.001), a prior tuberculin skin test (75.7%versus 54.7%, P<0.001), a prior positive tuberculin skin test (65.8%versus 46.3%, P<0.001) and BCG vaccination (25.2% versus 10.9%, P<0.001).Aboriginal persons were more likely to report a history of significant alcohol intake Can Respir J Vol 7 No 2 March/April 2000 153

TABLE 3 Tuberculosis history and predisposing factors
*Others include patients prescribed corticosteroids or other immunosuppressive therapy, predisposing condition such as cancer and diabetes mellitus.BCG bacille Calmette-Guerin vaccination; HIV Human immunodeficiency virus

TABLE 4 Treatment administered, drug susceptibility and data on number of contacts and chemoprophylaxis administered to the aboriginal and nonaboriginal groups
INH Isoniazid; PPD Purified protein derivative; PZA pyrazinamide; RIF Rifampin.*Fishersexacttest used because of small numbers.†Thetotal number treated is incomplete because of a small number of subjects in both groups who either did not initiate treatment or left the province