Tuberculosis in London Annual Review 2014 - Gov.uk

0 downloads 188 Views 1MB Size Report
information to PHE centres, local health partners, service providers and ... In 2014, 2,572 new cases of tuberculosis (T
Tuberculosis in London: Annual review (2014 data) Data from 2000 to 2014

Tuberculosis in London: Annual review (2014 data)

About Public Health England Public Health England exists to protect and improve the nation's health and wellbeing, and reduce health inequalities. It does this through world-class science, knowledge and intelligence, advocacy, partnerships and the delivery of specialist public health services. PHE is an operationally autonomous executive agency of the Department of Health.

About the Field Epidemiology Service The Field Epidemiology Service (FES) supports PHE centres and partner organisations through the application of epidemiological methods to inform public health action. It does this in two main ways: first by providing a flexible expert resource, available as and when needed, to undertake epidemiological investigations for key health protection work, and second through the expert analysis, interpretation and dissemination of surveillance information to PHE centres, local health partners, service providers and commissioners of services. Within the FES network, excellence and innovation is encouraged. We foster academic collaborations and take active part and lead in research, development and training. You can contact your local FES team at [email protected]. Public Health England Wellington House 133-155 Waterloo Road London SE1 8UG Tel: 020 7654 8000 http://www.gov.uk/phe Twitter: @PHE_uk Facebook: www.facebook.com/PublicHealthEngland Prepared by: Field Epidemiology Service For queries relating to this document, please contact: [email protected] © Crown copyright 2015 You may re-use this information (excluding logos) free of charge in any format or medium, under the terms of the Open Government Licence v3.0. To view this licence, visit OGL or email [email protected]. Where we have identified any third party copyright information you will need to obtain permission from the copyright holders concerned. Published: December 2015 PHE gateway number: 2015535

2

Tuberculosis in London: Annual review (2014 data)

Contents About Public Health England

2

About the Field Epidemiology Service

2

Acknowledgements

4

Executive summary

5

Introduction

8

Objectives

8

Tuberculosis epidemiology

9

Overall numbers, rates and geographical distribution

9

Demographic characteristics

12

Clinical characteristics

19

Microbiological information

21

TB transmission

21

Delay from onset of symptoms to start of treatment of pulmonary TB cases

26

TB outcome in drug sensitive cohort

28

1: Outcomes for TB patients with expected duration of treatment less than 12 months

28

2: Outcomes for drug sensitive cohort of patients with CNS, spinal, miliary or cryptic disseminated TB

30

Drug resistant TB (including outcomes in the drug resistant cohort)

32

Drug resistance

32

TB outcome at 24 months for patients with rifampicin resistant disease

34

TB in those with social risk factors and health inequalities associated with TB

35

Patient care: HIV testing, hospitalisation and directly observed therapy

37

Discussion

38

Conclusion and recommendations

41

References

42

Appendix A: Description of data sources and definitions

43

Appendix B: TB among London residents

45

Appendix C: Local authority TB epidemiological summaries

48

The data presented in this report is correct as at August 2015.

3

Tuberculosis in London: Annual review (2014 data)

Acknowledgements We are grateful to all those who contribute information on tuberculosis cases in London, including nurses, physicians, microbiologists, scientists, outreach and social care and administrative staff. We also acknowledge colleagues at the PHE National Mycobacterium Reference Laboratory for information on culture confirmation, drug susceptibility testing and strain type. Further thanks are due to colleagues in the TB Section at the national PHE Centre for Infectious Disease Surveillance and Control who provided the cleaned matched dataset and supported the analysis for this report; Jacqui Carless, Lamya Kanfoudi and Neil Billingham for their work supporting the London TB Register; and the PHE London Health Protection Team staff.

Authors This report was prepared by Daniele Curtis, Charlotte Anderson, Helen Maguire, Maria Saavedra-Campos and Neil Macdonald of the Field Epidemiology Service (South East and London team), PHE National Infection Service.

Suggested citation Public Health England. (December 2015) Tuberculosis in London: Annual review (2014 data), 2015. Public Health England: London

4

Tuberculosis in London: Annual review (2014 data)

Executive summary In 2014, 2,572 new cases of tuberculosis (TB) were notified among London residents, a rate of 30 per 100,000 population. This was a 15% decrease from the rate observed in 2013, and a 26% decrease from 2012. Compared to the rest of the UK, however, rates remained highest in London.1 As in previous years, the highest numbers and rates of TB were reported among residents of Newham and Brent local authorities. However, these areas saw rates decline by more than 25% compared with 2013. Overall, the largest reductions were in the areas of London with highest incidence. A large decrease in the rate of TB among those aged 20-29 was observed, but rates remained highest in adults 20-39 years of age. Although 82% of individuals with TB in 2014 were born abroad, TB rates decreased in the non-UK born London population relative to 2013. While between 2012 and 2013 numbers decreased among those who entered within the previous two years, in 2014 the greatest decrease was among those who entered 2-4 years previously. This was also reflected in trends of TB rates among different ethnic populations in London. Although residents of Indian ethnicity continued to have the highest rate of TB, rates fell relative to 2013. Rates among Pakistani and black Africans similarly fell, but remained stable among those of white or black Caribbean ethnicity. While the most common country of birth of non-UK born TB cases was still India, the number born here decreased by 25% since 2013, and 37% since 2012. This was largely due to a reduction by almost half in the number of cases in Indian migrants who had arrived in the UK two to four years previously, in particular among those aged 20-29 years. The rate of TB in the UK born London population remained stable, at 8.8 per 100,000. However, the TB notification rate among the UK born population in London was more than double that in England overall.1 Half of all TB patients had pulmonary disease, and just under half of these were sputum smear positive. As in previous years, only a very small proportion of patients had a previous history of TB. Only 59% of cases were culture confirmed, although this was higher among those with pulmonary disease (73%). In 2014, the rate of TB in UK born children under 15 years of age in London, an indirect indicator of recent transmission, was estimated as 5.7 per 100,000. This was more than double the national rate.1 Between 2010 and 2014, half of strain typed London TB 5

Tuberculosis in London: Annual review (2014 data)

cases were identified as being clustered with one or more other resident(s) in London (using 24 loci Mycobacterial Interspersed Repetitive Unit-Variable Number Tandem Repeats). Over this same period, 996 clusters were identified among London residents. The majority of strain typed clusters were small, with just over half comprising two cases. Among those with pulmonary TB, the median time between symptom onset and treatment start was 67 days. The proportion of pulmonary TB cases starting treatment within two months of symptom onset was 43%, increasing to 72% within four months. This was slightly better than that observed in England overall.1 Delays of more than four months were more common among older patients, the UK born and in particular those of white ethnicity. According to the revised outcome categories,2 in 2013 86% of rifampicin sensitive patients without CNS, spinal, miliary or cryptic disseminated disease completed treatment within 12 months, similar to recent years. The most common reason for not completing treatment was being still on treatment (6%). Completion was lower among older, white, and individuals with social risk factors. Among those with central nervous system, spinal, miliary or cryptic disseminated disease, just over half had completed treatment at 12 months (56%). Overall, 3.3% of rifampicin sensitive London cases diagnosed in 2013 died within 12 months of diagnosis: while the median age of those who died was 67, TB caused/contributed to the death of nine individuals under the age of 40. A further 4.3% were lost to follow-up, more common among those born abroad and with social risk factors. Relative to 2013, isoniazid (7.5%) and multi-drug resistance (1.3%) decreased slightly in 2014. As seen previously, resistance was less common among children and older people. White patients had the highest levels of resistance to at least one first-line drug, and those of Indian ethnicity the lowest. The majority of resistant cases occurred in individuals with no previous history of TB. Isoniazid resistance was more common among UK born cases, while all but two of the 20 patients with multi-drug resistant TB were born abroad, most commonly India and Lithuania (four from each). Just over half (58%) of patients with rifampicin resistant disease notified in 2012 had completed treatment at 24 months. Nearly a quarter remained on treatment, 12% were lost to follow up and none had died. One in ten patients were known to have at least one social risk factor (homelessness, drug use, imprisonment or alcohol misuse), a third of whom reported multiple factors. Social risk factors were twice as common among those born in the UK than those born abroad. Approximately a quarter each of those of white or black Caribbean ethnicity had 6

Tuberculosis in London: Annual review (2014 data)

at least one social risk factor. Compared to those without, patients with social risk factors were twice as likely to be infectious. Drug resistance was also more common among those with at least one social risk factor. HIV testing coverage was extremely high, both with respect to the offering (99%) and uptake (98%) of tests. Nearly a third of all TB cases had been a hospital inpatient at some point throughout treatment, increasing to half among those with one or more social risk factor. The sustained decrease in TB numbers and rates in 2014 is promising. To strengthen TB control in the city, locally driven reviews of delays, outcomes and contact tracing should be continued. Although the majority of TB cases still occur in non-UK born individuals, there has been a reduction in cases among new migrants. Changes in migration patterns, pre-entrant screening and reducing TB rates in other parts of the world are likely to have contributed to this. The current roll out of latent TB infection screening will further reduce cases arising in the future among existing, recently arrived residents. Early detection remains essential and rapid molecular tests should be universally available and appropriately targeted to identify MDR. In those with higher risk (such as Eastern Europeans) a low threshold for suspicion should be set. Not all pulmonary cases were culture confirmed and greater efforts are warranted to ensure all pulmonary cases can have drug sensitivity tests performed. More work is required to validate and investigate delays to treatment, to identify areas for improvement. Poor treatment outcome among rifampicin-resistant and MDR cases should also be reviewed through focused pan-London cohort review. As overall numbers decline, mirroring that seen elsewhere in Europe,3 TB also becomes more concentrated in risk groups, emphasising the need to support services that explicitly address the underlying risks in these vulnerable populations. A third of cases were admitted to hospital at some point throughout treatment. Improved understanding of why this proportion is this high would clarify whether it is appropriate and justified on clinical grounds, or indicative of gaps in current community based TB services.

7

Tuberculosis in London: Annual review (2014 data)

Introduction TB remains a serious public health problem in London. Surveillance provides relevant information on the TB cases to local teams, to help plan and evaluate their services. This report is based on surveillance data on patients from TB clinics collected via the London TB Register (LTBR) or national Enhanced TB Surveillance (ETS) system and microbiological information, including drug resistance and strain type, provided by the National Mycobacterium Reference Laboratory (NMRL). This yearly report provides an update on the epidemiology of TB in London residents, including characteristics and distribution of TB cases in London, trends in anti-TB drug resistance, clustering of TB cases, and also the treatment outcome of patients. As part of the Collaborative TB Strategy for England 2015-2020, a suite of TB Strategy Monitoring Indicators have been developed (https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/403231/ Collaborative_TB_Strategy_for_England_2015_2020_.pdf). Data for indicators that are presented at upper tier local authority can be found at: http://fingertips.phe.org.uk/profile/tb-monitoring.

Objectives This report describes the recent epidemiology of TB in London. We aim to update the London TB control board and TB clinical leadership group, as well as public health, clinical and allied colleagues, including clinical commissioning groups and NHS England (London). We aim to inform about strategy indicators – complementing that available on fingertips (see above) – summarise the latest trends, identify areas where there is a high burden of disease, and describe at-risk population groups as well as opportunities for interventions and prevention of future cases.

8

Tuberculosis in London: Annual review (2014 data)

Tuberculosis epidemiology Overall numbers, rates and geographical distribution In 2014, 2,572 cases of TB were notified among London residents, a rate of 30 per 100,000 population. This was a 15% decrease from the rate observed in 2013, and a 26% decrease from 2012. These reflect national trends in TB incidence, with England reporting an overall decrease of 11% since 2013 and 21% since 2012.1 After remaining relatively stable at just over 40 per 100,000 population since 2002, this was the second consecutive year of a marked decrease in the London TB notification rate (Figure 1). However, London cases did still comprise 39% of the 6,520 TB cases in England, making London the PHE Centre with the highest notification rate. Figure 1: TB case notifications and rates, London, 2000-2014 (TB Monitoring Indicator 1)

As in previous years, the highest rate and number of cases were reported among residents of north west London, and the lowest in south west London (Figure 2). Relative to 2013, TB notification rates decreased in all five areas of London, as did overall numbers.

9

Tuberculosis in London: Annual review (2014 data)

Figure 2: TB notification rate by health protection team area of residence, London, 2000-2014

Historically, there has been little change with regards to which local authority residents have the highest rates of TB (Figure 3). In 2014, Newham (255 cases, 79 per 100,000) and Brent (204 cases, 64 per 100,000) had the highest TB notification rate, followed by Ealing (211 cases, 62 per 100,000), Hounslow (152 cases, 57 per 100,000) and Redbridge (133 cases, 45 per 100,000). Rates in Newham and Brent decreased the most however, by more than 25% compared with 2013, while those in Hounslow and Redbridge decreased by 7% and 13% respectively. Ealing showed little change.

10

Tuberculosis in London: Annual review (2014 data)

Figure 3: TB notification rate by local authority of residence, London, 2014

Within some local authorities there existed a high degree of variation in TB incidence, such that overall rates masked a relatively high number of cases in only a small number of middle super output areas (MSOAs). As illustrated in Figure 4, the high overall rates of Newham, Brent, Ealing, Hounslow and Redbridge were predominantly driven by a small number of MSOAs.

11

Tuberculosis in London: Annual review (2014 data)

Figure 4: TB notification rate by Middle Super Output Area of residence, London, 2014

Demographic characteristics Age and sex In 2014, 61% (1,566) of TB cases were male. Rates were also higher among males (37 per 100,000) than females (23 per 100,000), as seen in recent years (with an average of 14 additional cases per 100,000 among males over the past five years). TB notification rates were highest among adults aged 20-39 years, irrespective of gender (Figure 5).

12

Tuberculosis in London: Annual review (2014 data)

Figure 5: TB case notifications and rates by age and sex, London, 2014

Rates fell, or remained stable among all age groups relative to 2013 (Figure 6). The largest decrease was experienced by those aged 20-29 years, for whom the rate fell from 56 to 42 notifications per 100,000; a decrease of 25%. For this age group, the decrease was greater in females than males, and continued further a decrease since 2012. Although the TB notification rate in children under ten years of age increased relative to 2013, this age group accounted for only 3% of total cases, and the rate remained low at 6 per 100,000 population.

13

Tuberculosis in London: Annual review (2014 data)

Figure 6: TB notification rates by age group, London, 2000-2014

In 2014, 133 cases of TB were reported in children under the age of 16 (8 per 100,000 population), 71% (95) of whom were born in the UK. Similar to recent years, 38 cases occurred in children aged less than five years old (6 per 100,000). Among those under the age of five, 87% (33) were UK born, of whom most were of mixed/other (34%, 11) or black African ethnicity (31%, 10).

Place of birth and time since entry In 2014, 82% (2,089/2,563) of cases were among individuals born outside of the UK, similar to previous years. Although the TB notification rate for those born abroad was nearly eight times higher than for those born in the UK (67.8 per 100,000 vs. 8.8 per 100,000), this was the narrowest difference in over a decade (Figure 8). This was attributable to decreasing rates among the non-UK born population, which reduced by 17% from 2013, and 28% since 2012. The TB notification rate among the UK born population, however, showed little change from recent years, as did the number of cases (474 in 2014). The rate in London was more than double that of the UK born in England overall (3.9 per 100,000).1

14

Tuberculosis in London: Annual review (2014 data)

Figure 8: TB case notifications and rates by place of birth, London, 2000-2014 (TB Monitoring Indicator 2)

In 2014, information on the time between entry to the UK and TB notification was available for 91% (1,898/2,089) of those born outside the UK. Whereas only 228 (12%) of these cases were among recent entrants to the UK (diagnosed less than two years after entry), 827 (44%) occurred in individuals ten or more years after their arrival. Relative to 2013, cases among those born abroad reduced in all groups other than those who had been in the UK for ten or more years (Figure 9). While in 2013 the greatest decrease was among those who entered within the previous two years, in 2014 the greatest decrease was among those who entered 2-4 years previously.

15

Tuberculosis in London: Annual review (2014 data)

Figure 9: Time between entry to the UK and TB notification for non-UK born cases by year, London, 2000-2014

In 2014, country of birth was known for 2,029 (97%) of those born abroad. As in previous years, the highest proportion of cases occurred among those born in India (28%, 567 cases). This represented a 25% decrease in the number born in India relative to 2013, and a 37% reduction compared with 2012 (Table 1). This was largely due to a reduction by almost half in the number of cases in Indian migrants who had arrived in the UK two to four years earlier (142 in 2014 vs. 257 in 2013), in particular among those aged 20-29 years (137 in 2014 vs. 215 in 2013). After India, the most common countries of birth were Pakistan, Somalia, Bangladesh, Nepal and Nigeria. Together, cases in individuals from these countries represented 59% (1,197) of non-UK born TB cases, and 47% of all TB cases. This compares with the most common countries of birth in the non-UK born general population of London, which in 2014 were India, Poland, Pakistan, Bangladesh and the Republic of Ireland.4

16

Tuberculosis in London: Annual review (2014 data)

Table 1: Ten most common countries of birth of non-UK born TB cases, London, 2014

Ethnicity In 2014, ethnicity data was available for 99% (2,555) of all cases. TB cases most commonly occurred among those of Indian ethnicity in London (28%, 705). Although London residents of Indian ethnicity also had the highest rate of TB (123 notifications per 100,000 population), this was a 24% decrease from 2013, and a 38% decrease since 2012. After consistent increases between 2001 and 2012, this was the second consecutive year of a decrease among those of Indian ethnicity, and was the lowest observed rate since 2002 (Figure 10). The second highest rate was among those of Pakistani ethnicity, although these accounted for only 9% of cases (237). Rates in this group also reduced from 145 per 100,000 population in 2013, to 96 per 100,000 in 2014, a decrease of a third and the largest decrease seen in any ethnic group. The next highest rates were among those of black African ethnicity, with 85 per 100,000 population. Rates in this population have been steadily decreasing since 2003, and there was a reduction of 12% since 2013. Black African represented the second most common ethnic group (21%, 537). With the exception of those of white or black Caribbean ethnicity, rates fell for all ethnic groups in 2014 compared with 2013. Over half of those of white and black Caribbean ethnicity were born in the UK (53%, 167/313 and 59%, 53/90 respectively). In 2014 the rate (26 per 100,000) among those of black Caribbean ethnicity was more than four times higher that of the white population (6.4 per 100,000), but these accounted for only a small proportion (3.6%, 91) of London TB cases.

17

Tuberculosis in London: Annual review (2014 data)

Figure 10: TB case rate by ethnic group, London, 2001-2014

Ethnicity among UK born TB cases The ethnicity distribution in the UK born patient population is notably different from that of the patient population overall. As in recent years, in 2014, the most common ethnic group was white (36%, 167/469) (Figure 11). One in five was black African, 14% were mixed/other ethnicity, and around ten percent each were black Caribbean and Indian. Figure 11: Proportion of UK born TB cases by ethnic group, London, 2000-2014

18

Tuberculosis in London: Annual review (2014 data)

Occupation In 2014, occupation information was complete for 90% (2,160) of the 2,395 cases aged 18 years or older (Table 2). Of these, 211 (10%) cases occurred in those working or engaged in education, almost all full time students (96%, 202). Most of these students were born outside of the UK (80%, 162/202). Of the 835 cases in people not currently working, 30% (252) were retired and another 46% (383) unemployed. Table 2: Occupational category of persons with TB aged 18 years and older, London, 2014

Clinical characteristics Site of disease In 2014, half of all TB patients had pulmonary disease (Table 3). The second most common site was extra-thoracic lymph node TB, accounting for nearly a quarter of cases.

19

Tuberculosis in London: Annual review (2014 data)

Table 3: Site of disease of TB patients, London, 2014

*patients may have disease at more than one site, so the total % will not equal 100%

Pulmonary TB was more common among UK-born patients (68%, 324/474) than those born abroad (46%, 955/2,089). Among non-UK born pulmonary TB patients, 13% (117/871) had entered the UK less than two years previously. Although consistent with 2013 (13%, 125/969) this was a reduction of more than 40% relative to 2012 (22%, 240/1095). As in recent years, it was also more common among those of white (78%, 248/316) and Chinese ethnicity (68%, 17/25), and least common among those of Bangladeshi (33%, 49/148) and Indian (39%, 275/705) ethnicity.

Previous diagnosis of tuberculosis In 2014, data on previous diagnosis was available for 98% (2,520) of cases. As in previous years, 6.3% (160) were previously diagnosed with TB.

BCG vaccination In 2014, information on BCG vaccination was available for 1957 cases (76%), of whom 76% (1,492/1,957) were vaccinated (Table 4). Overall, comparable proportions of UK born and non-UK born cases had been vaccinated.

20

Tuberculosis in London: Annual review (2014 data)

Table 4: Number and proportion of TB patients with BCG vaccination, London, 2014

Microbiological information Culture confirmation and speciation As in previous years, 59% (1,530) of cases were culture confirmed. This was higher among those with pulmonary TB (73%, 940/1,284 vs. 46%, 590/1,288, of patients with exclusively extra-pulmonary TB). Of those cases that were culture confirmed, the vast majority were Mycobacterium tuberculosis (98%, 1,506), with only seven cases of M. bovis, and 16 cases of M. africanum. One case could not be categorised beyond belonging to the M. tuberculosis complex.

Sputum smear In 2014, sputum-smear results were known for 76% (974/1,284) of patients with pulmonary TB. Of these, 46% (446) were smear-positive, consistent with recent years.

TB transmission It is not currently possible to directly measure TB transmission at a population level, so proxy measures are required. The rate of TB in children is widely accepted to be a good indicator of TB transmission in a community. Molecular genotyping of the organisms causing TB in a population can also provide insight into putative transmission chains.

Rate of TB in UK born children In 2014, the rate of TB in UK born children under 15 years of age in London, an indirect indicator of recent transmission, was estimated at 5.7 per 100,000. This represents a reduction in this rate since the peak of 9.2 per 100,000 in 2008 (Figure 12). Small numbers of cases among children, however, mean year-on-year changes should be interpreted with caution.

21

Tuberculosis in London: Annual review (2014 data)

Figure 12: Rate of TB in UK born children aged under 15 years, London, 2000-2014 (TB Monitoring Indicator 5)

Strain typing and clustering The National TB Typing service in England was established in 2010 and since that time all TB isolates have been typed using 24 loci Mycobacterial Interspersed Repetitive Unit-Variable Number Tandem Repeats (MIRU-VNTR). Such strain typing identifies clusters of cases with indistinguishable strains that may be due to recent transmission. 5 While these clustered cases may reflect cases that are part of the same chain of recent transmission, this could also reflect common endemic strains circulating either within England or abroad. Thus the detection of a common strain type among cases does not confirm recent transmission. Additional epidemiological information is required to assess whether a common strain type is likely to reflect recent transmission. MIRUVNTR strain trying can be used to refute transmission between individuals who have different strain types. It is hoped that a higher level of resolution provided by whole genome sequencing (WGS) will improve our understanding of TB transmission in England.

22

Tuberculosis in London: Annual review (2014 data)

Proportion of cases clustered In 2014, 99.8% (1,527/1,530) of culture confirmed cases in London residents had an isolate that was strain typed and 93% (1,427/9,429) had at least 23 loci typed (Table 5). Overall, during 2010 to end 2014, 86% (8,148/9,429) of isolates were typed with at least 23 loci; 58.3% (4,753/8,148) were identified as belonging to 1,556 molecular clusters in England, with the remaining 41.7% (3,395) having a unique strain. Table 5: Number and proportion of culture confirmed cases typed, with 23 or 24 loci typed, London, 2010-2014

a

culture confirmed cases which have at least one loci typed culture confirmed cases which have had at least 23 loci typed c culture confirmed cases which have had at least 24 loci typed

b

The proportion of London residents that clustered with at least one other case within London from 2010 to 2014 was relatively stable by year (range 47% to 54%, Table 6) and these cases were distributed across 996 clusters within London. The number of new clusters identified each year1 has been around 200 a year since 2010. Table 6: Number and proportion of unique cases, clustered cases and new clusters by year, London, 2010-2014

a

b

London TB cases clustered with at least one other case in a London resident between 2010 and 2014 New clusters identified that included at least one London resident. A new cluster is identified at the point when a second case is notified with the same MIRU-VNTR strain type as an existing case within London

1 A new cluster is identified at the point when a second case in a London resident is notified with the same MIRU-VNTR strain type as an existing case in London 23

Tuberculosis in London: Annual review (2014 data)

Size of clusters Over the five year period of 2010 to 2014, there were a total of 969 clusters that included London residents, with a median cluster size of two (range 2-102). The majority of clusters (52%) composed two cases, with 20% five or more and 6% ten or more cases (Figure 13). Figure 13: Proportion of clusters by size, London, 2010-2014

Proportion of clusters (%)

60 50

40 30 20 10 0 2 cases

3 cases

4 cases

5 to 9 cases

10 or more cases

Size of clusters

Cluster lineage The majority of London clusters were of Euro American lineage, which accounted for 42% (422/996) of clusters observed between 2010 and 2014 (Table 7). Clusters of Central Asian lineage comprised 26% (260/996), followed by East African Indian 15% (149) and Beijing 5% (52). The distribution of cluster size in London tended to be similar across lineage (median cluster size 2-2.5), however 15% (8/52) of Beijing lineage clusters contained ten or more cases compared with less than 7% for the others. Table 7: Cluster lineage and size, London 2010 – 2014

*includes cases with M.bovis, M.africanum, multiple lineages and cases where no lineage has been identified 24

Tuberculosis in London: Annual review (2014 data)

Characteristics of cases in clusters Of the 4,129 clustered TB cases notified between 2010 and 2014, 62% were male, 71% were aged 15-44 and 79% were born outside the UK (Table 8). For those born outside the UK, 20% had been in the country for less than two years and 36% for ten or more. The most common ethnicities for those in clusters were Indian and black African (both 26%). Fourteen percent of clustered cases had one or more social risk factor: 6% reported drug use, 6% alcohol misuse, 5% had a history of homelessness and 4% of imprisonment. In terms of clinical characteristics, 65% of clustered cases had pulmonary disease of which 50% were sputum smear positive. Only 6% had a previous history of TB, 9% had isoniazid drug resistance and overall 2% were multi drug resistant (MDR). Between 2010 and 2014 the proportions of these characteristics remained broadly similar for clustered cases in London. Among non UK born clustered cases there was a decline in the proportion who entered the UK less than two years before TB notification, from 26% (127/486) in 2010 to 11% (53/467) in 2014. The proportion who had entered the UK ten or more years before notification increased from 30% (146/486) in 2010 to 46% (215/467) in 2014.

25

Tuberculosis in London: Annual review (2014 data)

Table 8: Characteristics of clustered cases*, London, 2010-2014

* denominator may vary depending on completeness of variable a

of pulmonary cases

Delay from onset of symptoms to start of treatment of pulmonary TB cases Time symptomatic Information on time from symptom onset to treatment start was available for 77% (999/1,284) of pulmonary TB cases in 2014. Sixteen patients were asymptomatic at diagnosis. The median time between symptom onset and treatment start was 67 days, with an interquartile range (IQR) of 35-128 days. This was a week shorter than the median delay to treatment start for England overall (74 days, IQR 39-139).1

26

Tuberculosis in London: Annual review (2014 data)

In 2014, 43% (427/999) of London residents with pulmonary TB started treatment within two months of symptom onset, and 72% (723/999) within four months. This was slightly better than what was observed in England overall (where 40% of pulmonary cases started treatment within two months, and 70% within four months of symptom onset).1 Table 9: Time between symptom onset and treatment start in pulmonary TB cases*, London, 2014 (TB Monitoring Indicators 6 and 7)

*excluding those with missing onset or treatment start dates

Characteristics of pulmonary TB cases with a delay from onset of symptoms to treatment of more than four months Comparable proportions of males and females had a greater than four month delay between symptom onset and treatment (27%, 176/645 vs. 28%, 100/354). The proportion of pulmonary TB cases with a delay of more than four months increased with age, from only 7% (3/45) in children under the age of 15, to 45% (53/118) in cases 65 years or older. It was slightly more common for UK born cases to wait in excess of four months compared with non-UK born cases (31%, 75/242 vs. 27%, 200/754). The highest proportion of cases with a delay of more than four months to treatment occurred in those of white ethnicity (41%, 76/185). This was followed by Chinese (36%, 5/14), black Caribbean (32%, 15/47) and mixed/other ethnicity (26%, 57/216).

27

Tuberculosis in London: Annual review (2014 data)

TB outcome in drug sensitive cohort Drug sensitive cohort For the purposes of TB outcome reporting the drug sensitive cohort excludes all TB cases with rifampicin resistant TB (initial or amplified) including multi-drug resistant (MDR) TB (initial or amplified), and non-culture confirmed cases treated as MDR-TB.2 Treatment outcomes for the drug sensitive cohort are reported separately for the following groups: 1. For cases with an expected duration of treatment less than 12 months, TB outcomes at 12 months are reported. This group excludes cases with CNS disease, who have an expected duration of treatment of 12 months. In addition, those with spinal, cryptic disseminated or miliary disease are excluded from this group, as CNS involvement cannot be reliably ruled out for the purposes of reporting 2. For cases with CNS, spinal, cryptic disseminated or miliary disease, the last recorded treatment outcome is reported. For cases reported in 2013, however, information on final outcome was collected in 2014 so may be only one year after start for many patients. In 2013, 2,965 TB cases were notified, of which 2,922 (99%) were sensitive to rifampicin, and were thus included in the drug sensitive cohort.

1: Outcomes for TB patients with expected duration of treatment less than 12 months Of those with rifampicin-sensitive TB in 2013, 89% (2,592) had non-CNS, spinal, miliary or cryptic disseminated disease. Of the 2591 patients for which information regarding treatment outcome was available, 86% (2,232) completed treatment within 12 months, similar to recent years (Table 10).

28

Tuberculosis in London: Annual review (2014 data)

Table 10: Number and proportion completing treatment at 12 months, London, 2002-2013*

*excludes rifampicin resistant TB, and patients with CNS, spinal, miliary or cryptic disseminated disease

At 12 months, 5.6% (146) of patients were still on treatment, making this the most common reason for failure to complete treatment within 12 months (Table 11). Of these, 54% (79) were on a planned treatment regimen that exceeded 12 months (29% of these due to initial isoniazid resistance), 33% (48) had their treatment regimen changed (equally attributable to poor clinical response, intolerance, and initial drug resistance), and 12% (18) had their treatment interrupted. Table 11: TB outcome at 12 months, London, cases diagnosed in 2013*

*excludes initial and amplified rifampicin resistant TB cases, and patients with CNS, spinal, miliary or cryptic disseminated disease

Treatment completion was slightly lower among males (85%, 1,262/1,493) as compared with females (88%, 970/1,098). Overall, treatment completion decreased with age, with 95% (94/99) of patients less than 15 years of age completing treatment within 12 months, 29

Tuberculosis in London: Annual review (2014 data)

compared with 77% (190/246) of patients aged 65 years or older. Although equivalent proportions of non-UK born (86%) and UK born (87%) patients completed treatment, the former were ten times more likely to be lost to follow-up (4.8% vs. 0.5%). As illustrated in Figure 14, treatment completion was lowest among black Caribbean individuals born outside of the UK (75%, 24/32), followed by white UK born individuals (80%, 122/152). Treatment completion was worse among those with at least one social risk factor (80%, 179/224) as compared with those without any social risk factors (87%, 1936/2,215). Figure 14: Treatment completion within 12 months by place of birth and ethnic group, London, cases diagnosed in 2013

2: Outcomes for drug sensitive cohort of patients with CNS, spinal, miliary or cryptic disseminated TB Of the 330 cases of CNS, spinal, miliary or cryptic disseminated disease in 2013, only 56% (185) had completed treatment within 12 months, and 65% (215) had completed at the last recorded outcome (Table 12). Among the 207 patients for whom duration of treatment was known, the median treatment time was 337 days (IQR 269-365); approximately 11 months. A considerable number of cases (19%, 62) were still on treatment at the last recorded outcome, and three (0.9%) had had their treatment stopped. 30

Tuberculosis in London: Annual review (2014 data)

Table 12: Last recorded TB outcome for cases of rifampicin sensitive, CNS, spinal, miliary or cryptic disseminated disease, London, cases diagnosed in 2013*

*excludes initial and amplified rifampicin resistant TB cases

Equivalent proportions of females (57%, 70/123) and males (56%, 115/207) had completed treatment at 12 months since diagnosis. There was no obvious trend between treatment completion and age (50-59 year olds reported the highest proportion at 69%, 27/39). A lower proportion of cases among UK born individuals (48%, 26/54) had completed treatment at 12 months than among non-UK born individuals (58%, 158/274). Among the 301 patients for whom social risk factor data was complete, only 8.6% (26) had one or more risk factor. Of these, 46% (12) completed treatment, compared with 59% (163/275) of those without any risk factors.

Deaths and lost to follow-up in the drug sensitive cohort (at 12 months since diagnosis) Overall, 3.3% (96/2,922) of rifampicin sensitive cases diagnosed in 2013 died before completing treatment. TB caused or contributed to 42 of these deaths (44%), was incidental to 26 (27%), and had an unknown relationship to the remaining 28 (29%). Four cases were diagnosed at post-mortem. The median age of those who died was 66.5 years (IQR 53-78), but TB caused/contributed to the death of nine individuals under the age of 40, including five under the age of 30 (one of which was 19 years old and UK born). Deaths were more common among UK born (4.4%, 21/481) than those born abroad (3%, 72/2,417). As in previous years, the proportion of rifampicin sensitive cases lost to follow-up at 12 months was 4.3% (125/2,922). Just over half of these (53%, 66) had left the UK: the proportion lost among non-UK born cases was 12 times higher than among those born in the UK (5.1%, 122/2,417 vs. 0.4%, 2/481), with all but three of the 125 cases lost to follow-up known to have been born abroad. A higher proportion of males were lost to follow-up (5.0% vs. 3.3% of females), as were those with at least one social risk factor (6.8%, 17/251 vs.3.8%, 95/2,490 of those without).

31

Tuberculosis in London: Annual review (2014 data)

Drug resistant TB (including outcomes in the drug resistant cohort) Drug resistance Overall drug resistance and geographical distribution In 2014, resistance profiles were available for all but five of the 1,530 culture confirmed cases. The proportion of culture-confirmed TB cases resistant to one or more first line drugs decreased slightly to 8.3% (126/1,525) relative to 2013 (9.0%, 157/1,750). Over the past 15 years, the proportion with first-line resistance has remained between 8 and 11% (Figure 15). The decrease since 2013 was seen in both the proportion resistant to isoniazid (from 8.3%, 145/1,750 to 7.5%, 115/1,525) and MDR (from 2.1% to 1.3%). Figure 15: Proportion of TB cases with first line drug resistance, London, 2000-2014

32

Tuberculosis in London: Annual review (2014 data)

Characteristics of patients with drug resistant TB Any first line drug resistance In 2014 equal proportions of males (8.3%, 78/941) and females (8.2%, 48/584) with TB were resistant to at least one first-line drug. As in previous years, drug resistance was less common among younger (only one person aged less than 15 years had drug resistant disease) and older persons (3.9%, 6/154 of those aged over 64 years), as compared with other age groups. A slightly higher proportion of UK born TB cases had drug resistant disease (10.7%, 28/262 vs.7.8%, 98/1,258 among those born abroad). This was consistent with recent years, other than 2013, and was due to more cases of isoniazid resistant TB among UK born individuals. Overall, in 2014, white TB patients had the highest levels of resistance (12.9%, 26/201), and those of Indian ethnicity the lowest (6.2%, 26/420). Among those born in the UK, the proportion of cases with resistance to at least one first line drug was similar for black Caribbean (14.7%, 5/34), white (14.4%, 14/97), Indian (13.3%, 4/30) and Pakistani (13.3%, 2/15) ethnic groups. Among those born abroad, resistance was most common in white patients (11.7%, 12/103), five of whom came from Lithuania. A third (33%, 7/21) of cases from Lithuania was resistant to at least one first line drug. Drug resistance was slightly more common among those with pulmonary TB (9.1%, 85/938) than those with exclusively extra-pulmonary disease (7.0%, 41/587). Similar to 2013, 10.6% (45/426) of sputum smear positive cases had drug resistant TB. Drug resistance was nearly twice as common among cases with a previous history of TB (15.5%, 11/71 vs. 8.0%, 113/1,422 among those with no previous history). However, these accounted for only 8.9% (11/124) of resistant cases. MDR In 2014, 20 (1.3%) cases were MDR, all but two of whom were born abroad. The most common countries of birth were India and Lithuania (four from each). Among non-UK born cases, those of white ethnicity had the highest proportion of MDR TB (3.9%, 4/103), while only 1.0% of those of Indian ethnicity had MDR disease. Only two patients with MDR disease had a previous history of TB. One case in an individual born in Lithuania notified in 2014 had extensively drug resistant (XDR) disease. Of the nine cases of culture confirmed XDR TB since 2009, five were in individuals born in Lithuania.

33

Tuberculosis in London: Annual review (2014 data)

TB outcome at 24 months for patients with rifampicin resistant disease In 2012, 41 TB cases were rifampicin resistant at start of treatment. Two additional cases became resistant while on treatment, bringing the total to 43. Nine were in individuals previously treated for TB (21%), and 38 (88%) were born abroad. Although none were XDR, 37 had MDR TB initially, and two more became MDR while on treatment (overall 91% of patients with rifampicin resistant TB were also MDR). At 12 months, three had completed treatment (two MDR, and one rifampicin resistant), five were lost to follow-up (three abroad), and the remaining 35 were still on treatment. Of those that remained on treatment, 22 had completed treatment within 24 months, ten were still on treatment and two had stopped treatment (Table 13). Table 13: TB outcome at 24 months for patients with rifampicin resistant disease, London, cases diagnosed in 2012

34

Tuberculosis in London: Annual review (2014 data)

TB in those with social risk factors and health inequalities associated with TB Social risk factors In 2014, information on social risk factors was available for 2,377 cases, of whom 10% (232) had one or more social risk factor (Table 14). A social risk factor was defined as current/history of homelessness, current/history of drug use, current/history of imprisonment or current alcohol misuse. Information on risk factors was completed for more than 93% of cases in 2014. Drug use was the most common risk factor (4.1%, 103/2,482). The next most common risk factors were homelessness (3.6%, 89/2,498) and alcohol misuse (3.6%, 86/2,404), followed by imprisonment (2.7%, 67/2,487). These proportions were all consistent with previous years. Among individuals with at least one social risk factor, many in fact had multiple, with a third reporting more than one (33%, 77/232). Table 14: Social risk factors among TB cases, London, 2009-2014

Among the 2,371 cases for whom risk factor and place of birth data was complete, social risk factors were twice as common among those born in the UK (16.6%, 73/439 vs. 8.2%, 158/1,932). Among the 158 non-UK born cases with at least one social risk factor, 38% (60) were black African, 20% (32) were of mixed/other ethnicity and 18% were white. Conversely, among UK born cases, 53% (39/73) of those with at least one social risk factor were white, and 23% (17) were black Caribbean. Overall, approximately a quarter of cases of white or black Caribbean ethnicity had at least one social risk factor, while less than 5% of those of Indian, Pakistani, Bangladeshi or Chinese ethnicity reported any such risk factors. Individuals with social risk factors were twice as likely to be infectious than those without (43% had sputum smear positive pulmonary TB vs. 20% among those without social

35

Tuberculosis in London: Annual review (2014 data)

risk factors), and although representing just 10% of cases, they accounted for 22% (89/411) of infectious cases. Drug resistance was more common among those with at least one social risk factor (13.9%, 24/173 vs. 7.8%, 96/1,232 among those without any risk factors). Patients with social risk factors accounted for 20% (24/120) of resistant cases in 2014, the majority of whom were isoniazid resistant.

Deprivation Since 2010, when the Index of Multiple Deprivation was first introduced, the distribution of TB case notifications by local deprivation quintile has remained stable. In 2014, 30% (767) of TB patients were resident in those areas that constituted the most deprived quintile in London, compared with 6.5% (168) in the least deprived areas. Similarly, TB rates were highest in the most deprived areas of London, and lowest in the least deprived areas (Figure 16). Figure 16: TB case rate by deprivation, London, 2014

36

Tuberculosis in London: Annual review (2014 data)

Patient care: HIV testing, hospitalisation and directly observed therapy HIV testing In 2014, information on HIV testing was available for 99% of cases. Of these, 98.5% (2,512/2,550) were either offered an HIV test, or were already aware of their HIV status. Testing uptake was very high: 98.4% of those to whom a test was offered were tested for HIV (2,387/2,426).

Hospital inpatient stay and directly observed therapy In 2014, information on whether or not a case had been a hospital inpatient at some point throughout their treatment was available for 2,541 (99%) cases, of whom nearly a third (31%, 783) had been an inpatient. A higher proportion of those 65 years or older (42%, 108/260), and of those with at least one social risk factor (51%, 118/231) had been inpatients. In particular 61% of cases in individuals with a history of homelessness were hospitalised (54/89). Almost half of initially MDR cases were hospitalised at some point (45%, 9/20). Overall, 15% (391/2,570) of cases notified in 2014 reportedly received directly observed therapy (DOT). Among those with one or more social risk factor, 63% (145/232) received DOT, compared with only 10% (218/2,144) of those without any social risk factors.

37

Tuberculosis in London: Annual review (2014 data)

Discussion As seen across England, rates of TB in London decreased in 2014, continuing further a reduction observed since 2012. These reductions were largest in the areas of London with highest incidence. However rates remain high and London still accounted for 39% of the 6,520 TB cases in England, and was the PHE Centre with the highest notification rate.1 The decrease was seen in those born abroad, particularly young adults from India who arrived 2-4 years previously. These reflect migration trends, which saw the number of Indians entering the UK on study visas reduce by more than 60% between 2010 and 2012.6 Decreasing rates of TB in India may also have had an impact.7 India, however, remained the most common country of birth among non-UK born patients, and despite two consecutive years of marked reductions, London residents of Indian ethnicity had the highest rates of TB. Pre-entry screening8 may be contributing to reductions in cases diagnosed within a few years of arrival to the UK and the implementation of LTBI screening as part of the Collaborative Tuberculosis Strategy for England will further reduce future cases (that might otherwise have arisen later).9 Although cases among London’s large foreign-born population accounted for the majority (82%) there is a sizeable contribution from UK born individuals too. Unlike those born abroad, rates in the UK born population have remained remarkably stable, suggesting that TB control measures introduced over the last decade have had minimal impact in this population. Also, rates in the London UK born population, although low, are more than double that in England overall, reflecting greater risk of TB in London. This may be due to persistent risk in second and third generation migrants, as well as the increased risk associated with social risk factors, more common among the UK born. As a major international urban centre, London has a greater prevalence of both these demographics as compared with other PHE centres. During the five years since universal strain typing of TB isolates has been carried out, 51% of culture confirmed London residents were molecularly clustered with one or more other case(s) in London. While this will reflect some recent transmission it is hard to determine precisely what the proportion directly attributable to this is. The implementation of WGS across PHE after 2016 will permit greater discrimination and better identify where recent transmission is occurring. This will provide evidence for targeted interventions in those groups. The identification of 996 MIRU-VNTR clusters already, reveals both the extent and complexity of TB molecular epidemiology in London. The time between the onset of symptoms and starting treatment was shorter among London residents with pulmonary TB compared to nationally. In recent years, however, 38

Tuberculosis in London: Annual review (2014 data)

the proportion with a delay greater than four months has increased, and in 2014 this was most common among older people, UK born and white patients. It is unclear whether delays are attributable to the patient presenting late to healthcare services, late diagnosis by clinicians, or procedural problems in the post diagnosis but pre-treatment patient pathway. The detailed information collected through cohort reviews across London should be used to better describe these delays and identify where improvements could be made. However, given the possibility that many – particularly older people, and those with social risk factors – present with comorbidities (making it difficult to reliably infer symptom onset date), caution is urged when interpreting these findings. Among those with sensitive and non-CNS disease, a high proportion continued to complete treatment within 12 months. The proportion completing was considerably lower among those with sensitive but CNS, spinal, miliary or cryptic disseminated disease, for which the current recommendation is to treat for 12 months. Only patients with rifampicin resistant disease (including MDR) would be expected to be on planned treatment regimens exceeding 12 months. Among these, 58% had completed treatment at 24 months. This was slightly higher than that in England overall (56%), as well as in the European region (49%).10 However, it was markedly lower than that in many highburden developing countries, such as Kenya (83%) and Myanmar (79%).10 While there was a reduction in the overall levels of drug resistance, the fact that isoniazid resistance remains more common among UK born patients demonstrates the continued transmission of drug resistant TB within London. In contrast, nearly all MDR cases occurred in those born abroad, particularly those from India or Lithuania. The only patient to be diagnosed with XDR disease in 2014 was also born in Lithuania, as were half of all XDR patients in recent years. A previous diagnosis of TB was reported by less than 10% of resistant cases, indicating that most drug resistance was a primary acquisition, rather than a result of treatment failure. The rate of culture confirmation was just 73% for pulmonary cases and greater efforts are needed to increase that proportion, so that tailored treatment is possible and resistant cases do not go undiagnosed. Although just one in ten patients experienced at least one social risk factor, these individuals were more likely to be infectious, have drug resistant TB, be hospitalised and were less likely to complete their treatment. They were also more likely to be in clusters (suggesting they transmit more often). Social risk factors were twice as common among UK born patients, and with higher proportions among those of white and black Caribbean ethnicity, these marginalised groups had an ethnicity profile distinct from that of the general London TB population. This underlines the need for good management of such patients, as described in the NICE guidance for vulnerable patients.11 As overall TB cases decline, it tends to become more concentrated in risk groups and there is a

39

Tuberculosis in London: Annual review (2014 data)

need to support services that explicitly address homelessness, alcohol misuse, drug use and crime in vulnerable populations. Despite recommendations that those with social risk factors be offered DOT as part of standard care,12 only 63% reportedly received DOT in 2014. Efforts to provide HIV testing to TB patients should be commended, as HIV testing coverage remains high, in keeping with the Collaborative Tuberculosis Strategy for England. 9 Overall, 1.4 % of rifampicin sensitive cases died with TB as a cause or contributing factor, and there were nine such deaths in under 40 year olds. Local audit and review of deaths in TB cases should be undertaken so that preventable factors can be avoided.

40

Tuberculosis in London: Annual review (2014 data)

Conclusion and recommendations The sustained decrease in TB numbers and rates in 2014 is promising. Rates in London, however, remain high compared with the rest of England and western Europe. Although the majority of TB cases still occur in non-UK born individuals, there has been a reduction in cases among new migrants. Changes in migration patterns, pre-entrant screening for active TB and reducing TB rates in other parts of the world are likely to have contributed to this. The current roll out of latent TB infection screening and treatment should further reduce future TB cases among existing recently arrived migrants. The roll out of this programme will be monitored by PHE nationally, and future reports from the national centre will include coverage of the target population. The apparently long delays from symptom onset to treatment should be validated and investigated using cohort review data. Not all pulmonary cases were culture confirmed and greater efforts are warranted to ensure all pulmonary cases can have drug sensitivity tests performed. While drug resistance remained relatively stable, early detection is important and rapid molecular tests should be universally available and appropriately targeted to identify MDR. In those with higher risk (such as Eastern Europeans) a low threshold for suspicion should be set. Although treatment completion among rifampicin-sensitive cases remains high, it was considerably lower among resistant cases. Given the complexity of these cases, it is likely that many factors contribute to these poor outcomes. These should be reviewed through focussed pan-London cohort reviews, to support the best possible care for these individuals. This also underscores the importance of reducing the drug resistant TB disease burden, as outlined in the Collaborative TB Strategy for England.9 Patients with social risk factors pose particular challenges to the NHS and the wider social care system. Guidance exists and should be followed.11 This will become increasingly important as overall TB cases decline (including those born abroad) and the disease becomes concentrated in such vulnerable populations. To strengthen TB control in the city, locally driven reviews of delays, outcomes and contact tracing as well as deaths should be carried out, with monitoring by the London TB Control Board. Key performance indicators at trust level should be agreed and regularly reviewed.

41

Tuberculosis in London: Annual review (2014 data)

References 1.

Public Health England. Tuberculosis in England: 2015 report. London: Public Health England; 2014 October. https://www.gov.uk/government/publications/tuberculosis-tb-in-the-uk

2.

World Health Organization. Definitions and reporting framework for tuberculosis – 2013 revision. WHO 2013. http://apps.who.int/iris/bitstream/10665/79199/1/9789241505345_eng.pdf

3.

European Centre for Disease Prevention and Control, WHO Regional Officer for Europe. Tuberculosis surveillance and monitoring in Europe 2015. Stockholm: ECDC; 2015. http://ecdc.europa.eu/en/publications/Publications/tuberculosis-surveillance-monitoring-Europe-2015.pdf

4.

Office for National Statistics. Population of the United Kingdom by country of birth and nationality tables, January 2014 to December 2014. London: Office for National Statistics; 2015. Table 1.4, Overseas-born population in the United Kingdom.

5.

Public Health England. TB strain typing cluster investigation handbook, 3 Edition, February 2014. http://webarchive.nationalarchives.gov.uk/20140714084352/http:/www.hpa.org.uk/webc/HPAwebFile/HPAw eb_C/1317140774833

6.

Office for National Statistics. Immigration statistics, January to March 2014. Office for National Statistics; 2014.Table ad_03, Passengers given leave to enter the United Kingdom by purpose of journey by main category and country of nationality. https://www.gov.uk/government/publications/immigration-statisticsjanuary-to-march-2014/immigration-statistics-january-to-march-2014#admissions-3

7.

World Bank. Incidence of tuberculosis. http://data.worldbank.org/indicator/SH.TBS.INCD

8.

Public Health England. UK pre-entry tuberculosis screening report 2014. London: Public Health England; 2015 June. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/442201/UK_preentry_tuberculosis_screening_report_2014_gateway_approved.pdf

9.

Public Health England. Collaborative tuberculosis strategy for England. Public Health England: London, 2015 January. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/403231/Collaborative_TB_Str ategy_for_England_2015_2020_.pdf

10.

World Health Organization. Global tuberculosis report 2015. World Health Organization: 2015. http://www.who.int/tb/publications/global_report/gtbr15_annex04.pdf?ua=1

11.

National Institute of Health and Clinical Excellence. Tuberculosis in vulnerable groups [internet]. NICE: 2013 September. https://www.nice.org.uk/advice/lgb11/resources/tuberculosis-in-vulnerable-groups60521141388229

12.

National Institute of Health and Clinical Excellence. Tuberculosis: identification and management in underserved groups [internet]. NICE: 2012 March. https://www.nice.org.uk/guidance/ph37

rd

42

Tuberculosis in London: Annual review (2014 data)

Appendix A: Description of data sources and definitions Data sources Data on TB cases in London comes from the PHE London TB Register (LTBR). The data contributes to the national Enhanced TB surveillance (ETS) system. Data collected includes notification details, and demographic, clinical and microbiological information, including drug resistance and strain type, provided by the National Mycobacterium Reference Laboratory (NMRL).

Definitions Social risk factors and directly observed therapy (DOT) have been defined in the RCN TB case management guidance.

Treatment outcome Information on outcomes was reported for all cases reported in the previous year, excluding those with known rifampicin resistant disease: outcomes for these cases were reported at 24 months. Definitions for outcome are based on World Health Organization (WHO) and European definitions, but adapted to the UK context. In this report, all data was obtained from the ETS matched dataset provided in August 2014.

Proportions All proportions in this report are calculated among cases with known information or a known result, except where otherwise stated.

Confidence intervals A 95% confidence interval for incidence was obtained using the relevant procedure in Stata, assuming a Poisson distribution.

Population denominator Tuberculosis rates by geographical area (centre, local authority, MSOA and LSOA), age, sex and place of birth were calculated using ONS mid-year population estimates for the most recently available year. Tuberculosis rates by ethnicity were calculated 43

Tuberculosis in London: Annual review (2014 data)

using the population estimates provided by the Greater London Authority via the London Data Store https://files.datapress.com/london/dataset/2013-round-ethnic-grouppopulation-projections/gla-egpp-2013rnd-trend-central-5yr-borough.zip

Cluster definitions Strain typing was performed at the TB reference laboratories using 24 MIRU-VNTR profiling. Analysis was undertaken on strain type clusters defined as two or more people with TB caused by indistinguishable strains, with at least 23 complete VNTR loci. Analysis of clustering in London was carried out on cases that clustered in London and notified between 2010 and 2014.

44

Tuberculosis in London: Annual review (2014 data)

Appendix B: TB among London residents Table Bi: TB cases numbers by local authority of residence, London, 2003-2014

45

Tuberculosis in London: Annual review (2014 data)

Table Bii: TB rate* per 100,000 by local authority of residence, London, 2003-2014

*rates calculated using ONS mid-year population estimates

46

Tuberculosis in London: Annual review (2014 data)

Table Biii: TB case numbers and rate by age and sex, London, 2014

Table Biv: Drug resistance among TB patients with culture confirmed disease*, London, 2000-2014

*culture confirmed cases with drug susceptibility testing results for at least isoniazid and rifampicin

47

Tuberculosis in London: Annual review (2014 data)

Appendix C: Local authority TB epidemiological summaries Local authority TB epidemiological summaries will provide further information about TB cases among residents of London upper tier local authorities with an average of at least 50 TB cases per year over the previous three years. These will be published online shortly by your local FES team.

48