Tuberculosis and HIV: estimates of the overlap in England and Wales

The growing impact of HIV infection on the epidemiology of tuberculosis in England and Wales: 1999 2003

BACKGROUND

Previous studies have estimated the prevalence of tuberculosis and HIV infection in population subgroups in the UK. This study was undertaken to describe recent trends in the proportion of individuals with HIV infection among reported cases of tuberculosis in England and Wales, and to review the implications for clinical and public health care.

METHODS

A population-based matching study using national surveillance databases was used to investigate all persons aged 15 years and over reported with a diagnosis of tuberculosis to the Health Protection Agency in England and Wales in 1999-2003. Record linkage was used to match the national tuberculosis and HIV/AIDS surveillance databases to identify all cases of tuberculosis and determine the proportion of patients with tuberculosis co-infected with HIV. The distribution and characteristics of the cases were determined and the trend examined by year.

RESULTS

Of 30,670 cases of tuberculosis reported in England and Wales between 1999 and 2003, an estimated 1743 (5.7%) were co-infected with HIV. There was a year on year increase in the proportion from 3.1% (169/5388) in 1999 to 8.3% (548/6584) in 2003 (p for trend <0.0001). Co-infected patients contributed to almost a third of the increase in the number of cases of tuberculosis during the 5 year period. Patients co-infected with HIV were predominantly those born abroad. 18.5% (n = 323) of co-infected patients had not been reported as active cases of tuberculosis on the national tuberculosis database.

CONCLUSION

The proportion of patients with tuberculosis co-infected with HIV in England and Wales is increasing, with the greatest impact on those born abroad regardless of their ethnic origin. With HIV infection contributing substantially to the increase in the number of cases of tuberculosis, close cooperation in the clinical management and accurate notification of patients is vital if appropriate care and public health action is to be achieved.

An estimate of the contribution of HIV infection to the recent rise in tuberculosis in England and Wales

BACKGROUND

The number of patients with tuberculosis has been increasing slowly in England and Wales since the late 1980s. HIV infection has been a contributory factor to increases in tuberculosis in a number of comparable industrialised countries. This study investigated the extent of tuberculosis and HIV co-infection in England and Wales in 1993 and 1998, and estimated its contribution to the increase in tuberculosis observed during this period.

METHODS

Patients aged 16-54 years old at diagnosis on the 1993 and 1998 National Tuberculosis Survey databases were matched with those on the HIV/AIDS patient database. A coded process maintained patient confidentiality. Primary outcome measures were the increase between 1993 and 1998 in the numbers with both infections reported and an estimate of the proportion of the increase in tuberculosis during this period attributable to HIV co-infection.

RESULTS

In 1993 61 (2.2%) tuberculosis patients aged 16-54 years matched with patients reported to the HIV database, increasing to 112 (3.3%) in 1998 (p=0.08; OR 1.35; 95% CI 0.97 to 1.87). Patients co-infected with HIV contributed an estimated 8.5% of the increase in number of tuberculosis patients between 1993 and 1998 nationwide (11% in London). In both years prevalence of co-infection was greatest in London and in patients of white and black African ethnic groups.

CONCLUSIONS

In 1998 the number of tuberculosis patients co-infected with HIV in England and Wales, though still small, had nearly doubled since 1993, with most of the increase occurring in London. As HIV infection may be undiagnosed in patients with tuberculosis, and tuberculosis may be unreported in patients with diagnosed HIV infection, the true extent of co-infection will have been underestimated by this study. In addition, constraints in coded matching make it inevitable that some reported co-infections are missed. Routine HIV testing of all patients with tuberculosis should now be considered, particularly in patients of white or black African ethnic origin under 55 years of age.

Tuberculosis at the end of the 20th century in England and Wales: results of a national survey in 1998

BACKGROUND

A national survey of tuberculosis was conducted in England and Wales in 1998 to obtain detailed information on the occurrence of the disease and recent trends. This survey also piloted the methodology for enhanced tuberculosis surveillance in England and Wales and investigated the prevalence of HIV infection in adults with tuberculosis.

METHODS

Clinical and demographic data for all cases diagnosed during 1998 were obtained, together with microbiological data where available. Annual incidence rates in the population were estimated by age, sex, ethnic group, and geographical region using denominators from the 1998 Labour Force Survey. Incidence rates in different subgroups of the population were compared with the rates observed in previous surveys. The tuberculosis survey database for 1998 was matched against the Communicable Disease Surveillance Centre HIV/AIDS database to estimate the prevalence of HIV co-infection in adult patients with tuberculosis.

RESULTS

A total of 5658 patients with tuberculosis were included in the survey in England and Wales (94% of all formally notified cases during the same period), giving an annual rate of 10.93 per 100 000 population (95% CI 10.87 to 10.99). This represented an increase of 11% in the number of cases since the survey in 1993 and 21% since 1988. In many regions case numbers have remained little changed since 1988, but in London an increase of 71% was observed. The number of children with tuberculosis has decreased by 10% since 1993. Annual rates of tuberculosis per 100 000 population have continued to decline among the white population (4.38) and those from the Indian subcontinent, although the rate for the latter has remained high at 121 per 100 000. Annual rates per 100 000 have increased in all other ethnic groups, especially among those of black African (210) and Chinese (77.3) origin. Over 50% of all patients were born outside the UK. Recent entrants to the UK had higher rates of the disease than those who had been in the country for more than 5 years or who had been born in the UK. An estimated 3.3% of all adults with tuberculosis were co-infected with HIV.

CONCLUSIONS

The epidemiology of tuberculosis continues to change in England and Wales and the annual number of cases is rising. More than one third of cases now occur in young adults and rates are particularly high in those recently arrived from high prevalence areas of the world. The geographical distribution is uneven with urban centres having the highest rates. The increase in the number of cases in London is particularly large. Tuberculosis in patients co-infected with HIV makes a small but important contribution to the overall increase, particularly in London. To be most effective and to make the most efficient use of resources, tuberculosis prevention and control measures must be based on accurate and timely information on the occurrence of disease. A new system of continuous enhanced tuberculosis surveillance was introduced in 1999, based on the methodology developed in this national survey.

HIV associated culture proved tuberculosis has increased in north central London from 1990 to 1996

OBJECTIVES

To examine rates of culture proved tuberculosis in HIV infected patients in three specialist centres in north central London.

METHODS

Cases of tuberculosis in patients with previously documented HIV infection from 1990 to 1996 were identified retrospectively from microbiological/clinical records at Chelsea and Westminster, St Mary's, and University College London Hospitals.

RESULTS

Between 1990 and 1996 202 cases of culture proved tuberculosis were identified at the three centres. Of these, 132/202 (65.3%) occurred in homosexual/bisexual men, 41/202 (20.3%) were in patients with heterosexual contact in sub-Saharan Africa, and 29/202 (14.4%) were in "others." Overall 148/202 (73.3%) had pulmonary tuberculosis. The total number of HIV infected individuals seen at the three centres increased from 4298 in 1990 to 5048 in 1996. Rates of tuberculosis in the three centres increased from 0.46% in 1990 to 0.83% in 1996. Part of this increase was due to an increase in tuberculosis among Africans from 1993 to 1996.

CONCLUSIONS

Rates of HIV associated tuberculosis increased in these three centres in north central London between 1990 and 1996. In part this was due to an increase in the number of African patients with HIV infection attending the three centres. In addition, there was circumstantial evidence of recent transmission among homosexual men with HIV infection. Prospective "real time" surveillance of tuberculosis in HIV infected patients is needed in order to detect case clustering and to improve tuberculosis control.

Tuberculosis in the twentieth century: time-series mortality in São Paulo, Brazil, 1900-97

The objective of this study was to characterize tuberculosis mortality trends in the Municipality of São Paulo, Brazil, from 1900 to 1997. Standardized tuberculosis mortality rates and proportional mortality ratios were calculated and stratified by gender and age group based on data provided by government agencies. These measures were submitted to time-series analysis. We verified distinct trends: high mortality and a stationary trend from 1900 to 1945, a heavy reduction in mortality (7.41% per year) from 1945 to 1985, and a resumption of increased mortality (4.08% per year) from 1985 to 1995. In 1996 and 1997 we observed a drop in tuberculosis mortality rates, which may be indicating a new downward trend for the disease. The period from 1945 to 1985 witnessed a real reduction in tuberculosis, brought about by social improvements, the introduction of therapeutic resources, and expansion of health services. Recrudescence of tuberculosis mortality from 1985 to 1995 may reflect the increasing prevalence of Mycobacterium and HIV co-infection, besides loss of quality in specific health programs.

HIV and tuberculosis co-infection in an inner London hospital--a prospective anonymized seroprevalence study

OBJECTIVES

Since 1987 there has been an increase in tuberculosis notifications in the U.K., with this increase disproportionately affecting London. A recent national survey suggests that co-infection with HIV occurs in less than 5% of tuberculosis patients. This study asked if local co-infection rates in Inner London differed from the national results.

METHODS

157 consecutive patients starting antituberculous chemotherapy were venesected 2 weeks into treatment. Anonymized blood samples were screened for antibodies for HIV-1 and HIV-2 by enzyme-linked immunosorbent assay (ELISA). Epidemiological data were collected on each patient which was also coded before HIV test results were known.

RESULTS

Of 157 patients commencing antituberculous therapy, 39 patients (24.8%) were found to be co-infected with HIV-1. HIV-negative and positive patients were similar in terms of age and sex. When 98 patients giving their country of origin as other than Europe were considered there were 22 co-infected with HIV (22.4%). Of the 39 HIV-positive identified in this study, 37 were also identified by our voluntary HIV testing programme.

CONCLUSIONS

This study has shown that there may be very different rates of co-infection at a local level in the U.K. The local variation may be missed by national surveys and diverse local testing procedures. Anonymous testing identified only two patients with tuberculosis and HIV infection who were not identified by our voluntary HIV testing programme and this suggests that offering HIV tests to patients with tuberculosis is largely taken up by those at risk of HIV infection. Surveillance studies of this type are important in identifying marked local variation from the national pattern of HIV and Mycobacterium tuberculosis infection.

HIV-testing study of immigrants with pulmonary tuberculosis

The increasing prevalence of tuberculosis (TB) and human immunodeficiency virus (HIV), particularly in Africa and Asia, led us to investigate the prevalence of HIV infection in immigrants with pulmonary TB at the time of arrival in the UK. We performed anonymous HIV testing of stored sera from 39/65 immigrants referred to our unit between January 1991 and December 1994, who had culture-positive pulmonary TB. None of the 39 patients tested was positive for either HIV-1 or HIV-2, and the characteristics of the 26 patients for whom no serum was available were similar to those of the tested group. Despite the need to consider concomitant HIV infection in any patient with TB, particularly those from an area of HIV endemnicity, the present data do not suggest that recently arrived refugees, asylum seekers, immigrants or long-term visitors to the UK constitute a group in whom dual infection of HIV and Mycobacterium tuberculosis is common.

Prevention and management of tuberculosis in HIV positive patients living in countries with a low prevalence of Mycobacterium tuberculosis

We have re-examined the evidence on which current British Thoracic Society recommendations for primary and secondary prophylaxis and therapy of tuberculosis are based. We suggest that in a country such as the UK with a low prevalence of tuberculosis, primary prophylaxis should be offered primarily to tuberculin positive or anergic patients from high-incidence groups, including immigrants from high-prevalence countries, intravenous drug users and those with previous tuberculosis, that secondary prophylaxis be withheld from all but very high-risk patients and that four drug regimens which include ethambutol should be used for patients originating from, or who have lived in areas of the world with more than 2% primary isoniazid resistance.

Increasing incidence of tuberculosis in the young and the elderly in Scotland

BACKGROUND

Since 1987 there has been an arrest in the previously established decline of tuberculosis notifications in Scotland. A study was undertaken to determine whether age contributed to this phenomenon.

METHODS

Notifications of tuberculosis in Scotland were quantified by year and age group for the years 1981-92 from national statistics supplied by the Information and Statistics Division. Population data were obtained from the 1981 and 1991 national censuses.

RESULTS

Age group analysis of pulmonary tuberculosis notifications showed that, in the 0-14 age group, incidence (per 10(5) population) decreased from 7.4 in 1981 to 2.6 in 1987, rising by an estimated 12.6% per annum to 3.7 in 1992. In the 65+ age group incidence declined from 30.1 in 1981 to 17.3 in 1988, and rose by an estimated 4.1% per annum to 22.2 in 1992. In the age groups 15-44 and 45-64 a continuous decrease in notification rate was seen over the period of the study.

CONCLUSIONS

The plateauing of the incidence of tuberculosis in Scotland is associated with significant increases since 1987 of tuberculosis in the young and elderly. Contributions from ethnic minorities and those infected with HIV are negligible. An ageing population over the decade, with the highest tuberculosis rates seen in the older age group, may explain these findings.

Pulmonary complications of HIV disease: 10 year retrospective evaluation of yields from bronchoalveolar lavage, 1983-93

BACKGROUND

Pulmonary disease is a major contributor to morbidity and mortality in patients with HIV infection and AIDS. The aim of this study was to describe bronchoscopic findings and the spectrum of pulmonary pathogens in HIV seropositive patients undergoing investigation of respiratory disease over a 10 year period in a major UK referral centre.

METHODS

Recruitment was procedure based with data being captured when bronchoscopy was clinically indicated. Data were evaluated from 580 HIV seropositive patients (559 men, age 13-65 years) over a 10 year period from June 1983 to March 1993.

RESULTS

A total of 947 bronchoscopies was performed. The most frequent pulmonary pathogen isolated from bronchoalveolar lavage (BAL) fluid in 44% of all bronchoscopies was Pneumocystis carinii. Of all patients studied, 324 (55%) had at least one cytologically confirmed episode of P carinii pneumonia; this was AIDS defining in 219 (38%) of patients who underwent bronchoscopy. Between 1987 and 1993 the overall diagnostic yield from BAL fluid was 76%; 25% of all bronchoscopies yielded positive microbiological results, the most frequent isolates being Staphylococcus aureus, Streptococcus pneumoniae, Pseudomonas spp, and Haemophilus influenzae. Mycobacteria were identified in 8% of patients; M tuberculosis was the most common being identified in 3% of lavage samples and in 4% of patients. No drug-resistant M tuberculosis was found. Viral isolates (mainly cytomegalovirus) were identified in up to 31% of BAL fluid samples. Endobronchial Kaposi's sarcoma was seen in 15% of patients at bronchoscopy.

CONCLUSIONS

Of the 1956 newly diagnosed HIV seropositive patients receiving clinical care at St Mary's Hospital over this period, approximately 30% underwent bronchoscopy. Diagnostic rates for P carinii pneumonia, endobronchial Kaposi's sarcoma, and bacterial and mycobacterial infection have remained largely constant since 1989. Bronchoalveolar lavage produces high diagnostic yields generally, and P carinii pneumonia remains a common cause of pulmonary disease in these patients.

Mycobacterial infection in HIV seropositive and seronegative populations, 1987-93

BACKGROUND

Although the causes of the worldwide resurgence of tuberculosis are multifactorial, the HIV epidemic is believed to have had a central role. Control is further threatened by the emergence of multidrug-resistant tuberculosis.

METHODS

A retrospective evaluation was undertaken of trends in pulmonary and extrapulmonary culture positive mycobacterial pathology, and the prevalence of drug-resistant tuberculosis in both HIV seropositive and, presumptively, HIV seronegative patients receiving their clinical care at St Mary's Hospital, London. Five hundred and thirty eight patients (188 of whom were known to be HIV seropositive) with positive mycobacterial isolates between January 1987 and March 1993 were identified from laboratory records. These were cross referenced with drug surveillance records.

RESULTS

Overall, between 1987 and 1992 there was a progressive 3.5 fold increase in positive mycobacterial isolates and a 2.5 fold increase in patients with proven mycobacterial infection. This increase was greater within the HIV seropositive population. A total of 663 positive mycobacterial isolates was evaluated; the major pathogen identified was Mycobacterium tuberculosis (379 isolates, 57%). Three hundred and fourteen patients were diagnosed as having M tuberculosis, 49 of whom were HIV seropositive. M tuberculosis was predominantly isolated from the lung. Of 358 positive cultures for M tuberculosis (68 HIV seropositive, 290 presumptively HIV seronegative), only 27 isolates (7.6%), almost exclusively derived from presumed HIV seronegative patients, were resistant to either isoniazid, rifampicin, or both drugs together. No increases in drug-resistant isolates were observed over this period.

CONCLUSIONS

There has been a considerable increase in the incidence of tuberculosis in both HIV seronegative and seropositive populations during the study period. The emergence of drug-resistant tuberculosis was not observed.

Socioeconomic deprivation and notification rates for tuberculosis in London during 1982-91

OBJECTIVES

To investigate the association between four sociodemographic measures (unemployment, overcrowding, low social class, and the proportion of migrants from areas of high prevalence of tuberculosis) and average level and rate of change of notification rates for tuberculosis.

DESIGN

Ecological analysis of both the average and the rate of change of standardised annual notification rates for tuberculosis from 1982-91 and sociodemographic measures from the 1981 and 1991 censuses.

SETTING

32 London boroughs. SUBJECTS AND DATA: Sociodemographic measures from the 1981 and 1991 censuses and tuberculosis notification rates for 1982-91.

MAIN OUTCOME MEASURES

A measure of the association between average levels and rate of change in tuberculosis notification rates and four sociodemographic measures in 1981 and between the rate of change in tuberculosis notification rates between 1981 and 1991 and changes in sociodemographic measures between 1981 and 1991.

RESULTS

The average level of notifications was correlated with overcrowding and the proportion of migrants but not with unemployment or social class. No significant association was found between the rate of change in notification rates and sociodemographic measures in 1981. An association was found between increases in unemployment and the rate of change in notification rates, but the effect was small. Changes in the levels of unemployment explained 23% of the variation between boroughs in the rate of change in their notification rates.

CONCLUSION

The average tuberculosis notification rates were related to overcrowding and the proportion of migrants in 1981. Only increases in unemployment from 1981 to 1991, however, were significantly associated with the rate of change in notifications over the same period.

Isolation of mycobacteria from patients seropositive for the human immunodeficiency virus (HIV) in south east England: 1984-92

BACKGROUND

Tuberculosis and other mycobacterial infections are well recognised complications of HIV infection and surveillance is thus required.

METHODS

All mycobacteria isolated from HIV positive subjects and referred to the Public Health Laboratory Service South East Regional Tuberculosis Centre (SERTC) from the first such case in 1984 until the end of 1992 were reviewed.

RESULTS

A total of 803 mycobacteria isolated from 727 HIV positive subjects were referred to the SERTC during the study period. A single species was isolated from 660 patients: 150 members of the tuberculosis complex (146 M tuberculosis, two M bovis, and two M africanum), 356 M avium-intracellulare (MAI), and 154 other environmental mycobacteria. More than one mycobacterium was isolated from 67 patients. In 12 cases M tuberculosis and MAI were isolated from the same patient, almost always in that sequence, with an interval of 8-41 months between isolations. Most of the 407 isolates of MAI (74%) were considered to be clinically significant and often caused disseminated disease. In other cases single isolates of MAI were obtained from sputum or faeces and occasionally such isolates preceded disseminated disease by several months. Only 33 (14%) of the 229 isolates of environmental mycobacteria other than MAI were considered clinically significant.

CONCLUSIONS

HIV related mycobacterial disease is increasing in incidence in south east England. Further studies are required to determine the significance of single isolates of MAI and other environmental mycobacteria as a guide to the need for preventive chemotherapy or immunotherapy.