OUP user menu

From contact investigation to tuberculosis screening of drug addicts and homeless persons in Rotterdam

Gerard de Vries, Rob A. van Hest
DOI: http://dx.doi.org/10.1093/eurpub/cki203 133-136 First published online: 17 October 2005

Abstract

Background: In early 2001 there were indications that tuberculosis (TB) was increasingly becoming a problem among drug addicts and homeless persons in Rotterdam, after a periodical screening was discontinued in 1997. A contact investigation around a homeless drug addicted man in Rotterdam with infectious pulmonary TB is described. Contact investigation: A total of 507 drug addicts, homeless persons, and staff of facilities for these risk groups were examined with tuberculin skin testing (TST) and chest radiography. DNA fingerprinting of mycobacteriological cultures through Restricted Fragment Length Polymorphism methodology and molecular epidemiology investigation through cluster analysis were performed. Outcome: TST showed an infection prevalence of 29%, especially among staff of services for drug addicts and homeless persons. Six persons with active intrathoracic TB were identified. Cluster analysis demonstrated no relation with the initial case but showed intense transmission of TB among drug addicts and homeless persons in Rotterdam by multiple sources. As a consequence of the findings, a proposal to the Council of the City of Rotterdam resulted in the re-introduction of a comprehensive TB screening programme among these risk groups with mobile digital X-ray units (MXUs). Conclusion: This contact investigation gradually obtained the characteristics of a screening of drug addicts and homeless persons. Novel technologies, such as MXUs, facilitate appropriate and efficient outreach approaches to TB control among difficult-to-reach groups. This method and knowledge of individual fingerprints and clusters of TB patients are indispensable for underpinning proposals for change of local TB control strategies and convincing local authorities of the rationale.

  • DNA fingerprinting
  • policy implications
  • risk groups
  • screening
  • tuberculosis

Introduction

In the mid-1990s the number of hard drug addicts in Rotterdam (population 600 000) was estimated at ∼3700 through the capture–recapture method.1 The hard core of homeless persons was estimated ∼1100 persons, many of them also hard drug or alcohol addicted.2 On average 10 drug addicts or homeless persons were diagnosed with tuberculosis (TB) annually. Although a fraction of drug addicts and homeless persons were periodically screened for TB until 1996, still the majority of TB cases were found passively with symptoms. In 1997 screening was stopped because of limited financial and human resources and the Department of TB Control of the Rotterdam Municipal Health Service adopted a reactive strategy, i.e. an institution for drug addicts or homeless persons was examined only after an infectious case of TB was detected (contact investigation).

In early 2001 there were indications that TB was increasingly becoming a problem among drug addicts and homeless persons in Rotterdam. In 1999, TB was diagnosed in 13 drug addicts or homeless persons (8.8% of the total number of TB cases in Rotterdam), and in 2000 there were 17 cases (12.6%). DNA fingerprinting of the Mycobacterium tuberculosis isolates showed that a specific cluster had developed in Rotterdam with a substantial number of patients being drug addicted or homeless. According to the quarterly reports of the National Institute of Public Health and the Environment (RIVM), this was the fastest growing and largest TB cluster in the Netherlands.

We describe a contact investigation among drug addicts and homeless persons in Rotterdam turning into a screening exercise, and how DNA fingerprint cluster analysis contributed to the re-introduction of a comprehensive programme of periodic active case finding among these risk groups.

Initial case

A homeless, schizophrenic, drug addicted 50-year-old male came to the Department of TB Control in 2001 with a persistent cough that had worsened since 2 months. The chest X-ray (CXR) showed extensive infiltrative lesions with small cavities, compatible with infectious pulmonary TB (figure 1). It appeared that the man had been referred by his general practitioner twice before but had never reported himself. This time he was accompanied by a social worker. After the CXR was made the man left immediately but the social worker managed to bring a sputum sample from the patient the next day. Microscopical examination of the sputum after Ziehl–Neelsen staining showed many alcohol and acid fast bacteria. The patient could be traced the same day and was admitted to a general hospital for isolation and treatment.

Figure 1

Chest X-ray of the initial patient

Considering the duration of his complaints and the extent of the abnormalities on the CXR, it was assumed that the patient had been infectious for a long time. Besides, the infectiousness was regarded high, considering the quantitative count of alcohol and acid fast bacteria in the sputum. Furthermore, the patient had a very social lifestyle, chatting with many people every day, while his cough hygiene was minimal. In the months preceeding the diagnosis, the man had stayed in three different shelters for homeless persons. He visited the regional institution for ambulatory mental health care, daily, for supervised intake of antipsychotic drugs.

Outline of the contact investigation

The Department of TB Control planned an extensive contact investigation that was initially aimed at the staff and visitors of the three homeless shelters, as well as the staff of the regional institution for ambulatory mental health care. According to the national guidelines for contact investigation, contacts with a previous history of TB, latent tuberculosis infection (LTBI) or BCG vaccination, and persons born before January 1, 1945 are excluded from tuberculin skin testing (TST) and are examined by CXR. Among persons examined with TST in the first round 20% had a LTBI. These results lead to an extension of the second round of the contact investigation to a larger group of residents of various shelters for homeless persons, as they often stay at different locations. The outlines of the contact investigation were, therefore, deliberately obscured and it obtained more the characteristics of a screening.

In the second round, contacts were checked at the Department of TB Control with TST or CXR. In addition, 312 residents of shelters for homeless persons were examined for TB on location. A mobile digital X-ray unit (MXU) was used, allowing one of the TB physicians to read the CXR immediately in the van (figure 2). In this way evident abnormalities could be detected instantly and follow-up examination could be initiated at once. All CXRs were later reviewed at the Department of TB Control. In this group TST was not performed because of expected poor compliance with preventive treatment.

Figure 2

TB screening at a hostel for homeless persons in Rotterdam with a mobile digital X-ray unit

Outcome of the contact investigation

In the contact investigation a total of 507 persons were examined. For 127 persons the examination consisted of one or more TSTs (table 1). In 28 persons—18 staff members and 10 residents—LTBI was found, resulting in an infection prevalence of 29% (only persons in whom the TST examination could be completed were counted in the denominator). The CXR of one shelter resident with a positive TST in the first round of the contact investigation was suspicious for pulmonary TB. The sputum culture for M. tuberculosis became positive after several months. In two larger shelters for homeless persons the infection prevalence among staff members was high with 26% (6 out of 23 persons) and 27% (4 out of 15 persons). Preventive treatment with isoniazid during 6 months was started for 17 persons while for the remaining 10 individuals it was decided to follow them up with CXRs every 6 months during 2 years.

View this table:
Table 1

Results of tuberculin skin testing

ConclusionNumber of persons testedRemarks
No LTBIa67
Recent LTBIb2823 persons had a positive TST in the first round (including one resident in whom active TB was diagnosed later)
1 person with a positive TST in the second round
4 persons with a TST conversion
Infection with mycobacterium other than TB2
Examination not completedc30
Total number of persons examined with TST127
  • LTBI: latent tuberculosis infection; TB: tuberculosis; TST: tuberculin skin test

  • a: TST negative in first and second round, or TST negative when only the second round was performed

  • b: TST positive in the first round, or TST positive when only the second round was performed, or TST negative in the first round but positive in the second round (conversion)

  • c: No TST performed in the second round, or TST in the second round not read

The other 380 persons were examined for active TB by means of a CXR (table 2). Individuals examined with a CXR as a result of a positive TST are excluded from this number. As a result of the TB examination on location with the MXU, active intrathoracic TB was diagnosed in five persons: two persons with smear-positive pulmonary TB, two persons with culture-positive pulmonary TB, and one person with a tuberculous pleurisy. The results of this contact investigation translate into a prevalence rate of 1183 TB cases per 100 000 persons examined. All six patients with active TB completed 6 months of therapy under Directly Observed Therapy. Remarkably, two of these patients, one of them smear-positive, stayed in a shelter of which the initial case had been a resident for 3 months, which until then was unknown. During follow-up, LTBI was found in several staff members of this shelter and in one staff member culture-positive TB was diagnosed. These results are not presented in the tables since it is considered a new contact investigation.

View this table:
Table 2

Results of chest radiography examinations

ConclusionNumber of persons examinedRemarks
No indication of active intrathoracic TB375
Active intrathoracic TB5Second round:
2 persons with smear-positive pulmonary TB
2 persons with culture-positive pulmonary TB
1 person with tuberculous pleurisy
Total number of persons examined with a CXR380
  • TB: tuberculosis; CXR: chest X-ray

DNA fingerprint results

When M. tuberculosis is cultured from sputum or other patients' material in the Netherlands, the microbiology laboratory sends the isolate to the national mycobacteriological reference laboratory at the RIVM for DNA fingerprinting through Restricted Fragment Length Polymorphism (RFLP) methodology3. When a RFLP pattern matches that of the mycobacterial DNA from at least one other patient, they belong to a so-called cluster. This often indicates transmission of TB between these persons or infection by a common source or sources.

The mycobacterial RFLP pattern of the initial case was unique, i.e. not previously found in the Netherlands. None of the six other TB cases found had the same DNA fingerprint as the initial patient. The DNA fingerprint of four out of these six TB cases—as well as the fingerprint of the staff member mentioned above—belonged to the largest TB cluster in the Netherlands, almost entirely consisting of patients from Rotterdam. One fingerprint belonged to another major cluster related to homelessness in Rotterdam and one fingerprint was unique.

Discussion

In this contact investigation, which gradually obtained the characteristics of a screening, many cases of LTBI and six patients with active TB were found. RFLP pattern analysis showed that none of the secondary TB cases were related to the initial patient but reflected intense transmission of TB in the group of drug addicts and homeless persons in Rotterdam by multiple sources.

Under Dutch public health laws the Local Authorities are responsible for infectious disease control, implemented by Municipal Health Services. Metropolitan areas harbour a disproportionate number of inhabitants belonging to various risk groups for TB, such as recent immigrants from TB high-endemic countries, hard drug users, or homeless persons. High rates of TB among homeless persons in large cities are well documented almost universally.49 Outbreaks of TB in homeless shelters have been reported.10,11 Molecular epidemiology investigation through RFLP cluster analysis has contributed to the understanding of the significance and spread of TB in a community and demonstrated the importance of (recent) transmission of TB, e.g. among homeless persons.1215 Transmission of TB among homeless persons can also affect shelter personnel and other staff of medical and social services contacted by homeless persons.16,17 The many cases of LTBI we found among staff members of institutions that dealt with homeless persons were probably the result of intense exposure to TB during several years. Active case finding has been reported to decrease TB transmission and overall case rate among urban homeless persons.17,18 However, the yield of some TB screening programmes of homeless persons may be low due to poor health care system uptake and poor treatment completion rates.5,1923 Among risk groups such as homeless persons, conventional methods of contact investigation are not feasible for various reasons.8,9 Outreach interventions, bringing active case finding and case holding services closer to certain risk groups, with incentives to maximize compliance, and in co-ordinated co-operation with hospitals, social services and the community, have been described.17,22,24 Among some groups of careless care-avoiders strong persuasion or even mandatory screening may be necessary and has resulted in a drastic reduction of TB incidence.17,18 Novel technologies, such as MXUs with digital X-ray equipment facilitate appropriate and efficient approaches of targeted TB control among difficult-to-reach groups. A TB physician can read the CXR immediately and follow-up examination can be initiated instantly, such as a short medical history, spot sputum collection, or referral to a clinic accompanied by a TB nurse or social worker. In this way, the failure of many individuals in attending the medical services as a result of their care-avoiding behaviour can be prevented. Also, the bond created in the first encounter with health care providers will positively influence the trust of the patient and contribute to better compliance.

Many homeless persons are also alcohol or hard drug addicted. High rates of TB among hard drug users are also reported.2527 Digital MXUs can also be employed in co-operation with needle exchange posts, methadone substitution programmes, or day-care facilities for hard drugs users. In the Netherlands MXUs have been used effectively for TB control among prisoners and in centres for asylum seekers.28 This has facilitated its use in case of an outbreak of TB or screening of other risk groups on location.

Considering the outcome of the DNA fingerprint analysis, this ‘contact investigation’ should be regarded as a screening, revealing several undetected TB cases among hard drug users and homeless persons in Rotterdam. As a consequence of these findings, a proposal to re-introduce TB screening for hard drug users, methadone users, and homeless persons in Rotterdam gained momentum from the Municipal Health Service and the institutions providing social services to these risk groups. It contributed to making TB screening among drug users, methadone users, and homeless persons one of the priority issues in the Council Programme 2003–6 of the Council of the City of Rotterdam. This resulted in the re-introduction of a comprehensive TB screening programme among these risk groups, aimed at semi-annual radiological examination, with a digital MXU already used for screening prisoners and asylum seekers. Also, arrangements were made with the institutions for social services for drug users, methadone users, and homeless persons for periodic examination of their staff members.

Conclusions

Contact investigation has not been effective among hard drug users and homeless persons in Rotterdam. The use of MXUs with digital X-ray equipment facilitates TB control among certain difficult-to-reach TB risk groups. Molecular technologies in TB control such as mycobacterial strain identification with DNA fingerprinting can recognize local epidemiological patterns previously undetectable. The knowledge of individual fingerprints and clusters of TB patients not only contributes to better risk group identification but is also indispensable for underpinning proposals for change of local TB control strategies and convincing local authorities of the rationale behind appeals for support, in this case for a screening programme.

Disclaimer

No financial support received. No conflict of interest to report. Part of this article was previously published in Dutch in Tegen de Tuberculose 2003;99:33-35 (non-indexed).

Key points

  • Discontinuing screening in certain risk groups can lead to unrecognized on-going transmission for several years before resurgence becomes visible.

  • The use of mobile X-ray units with digital equipment facilitates TB control among certain difficult-to-reach risk groups.

  • Molecular technologies in TB control can recognize local epidemiological patterns previously undetected.

  • Knowledge of transmission within communities is indispensable for underpinning proposals for targeted TB control strategies.

References

  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
  6. 6.
  7. 7.
  8. 8.
  9. 9.
  10. 10.
  11. 11.
  12. 12.
  13. 13.
  14. 14.
  15. 15.
  16. 16.
  17. 17.
  18. 18.
  19. 19.
  20. 20.
  21. 21.
  22. 22.
  23. 23.
  24. 24.
  25. 25.
  26. 26.
  27. 27.
  28. 28.
View Abstract