Get access

Directly observed therapy for treating tuberculosis

  • Review
  • Intervention


  • J Volmink,

  • P Garner

Prof Jimmy Volmink, Deputy Dean: Research and SACC Director, Faculty of Health Sciences, Stellenbosch University, PO Box 19063, Tygerberg, 7505, SOUTH AFRICA.



People with tuberculosis require treatment for at least six months. As many patients do not complete their treatment, policies have been introduced to encourage adherence to treatment regimens. One such policy is directly observed therapy, which involves people directly observing patients taking their antituberculous drugs.


To compare directly observed therapy (DOT) with self administration of treatment in people requiring treatment for clinically active tuberculosis or prevention of active disease.

Search strategy

We searched the Cochrane Infectious Diseases Group Specialized Register (November 2005), CENTRAL (The Cochrane Library 2005, Issue 4), MEDLINE (1966 to November 2005), EMBASE (1974 to November 2005), LILACS (1982 to November 2005), and reference lists of articles. We also contacted researchers and organizations working in the field.

Selection criteria

Randomized and quasi-randomized controlled trials comparing a health worker, family member, or community volunteer routinely observing people taking antituberculous drugs compared with routine self administration of treatment at home. We include patients requiring treatment for clinically active tuberculosis or medication for preventing active disease.

Data collection and analysis

Both authors independently assessed trial methodological quality and extracted data. Data were analysed using relative risks (RR) with 95% confidence intervals (CI) and the fixed-effect model when there was no statistically significant heterogeneity (chi square P > 0.1). Trials of drug users were analysed separately.

Main results

Ten trials with 3985 participants met the inclusion criteria. There was no statistically significant difference between DOT and self administration of treatment for the number of people cured (RR 1.02, 95% CI 0.86 to 1.21, random-effects model; 1603 participants, 4 trials) or who were cured or completed treatment (RR 1.06, 95% CI 1.00 to 1.13; 1603 participants, 4 trials). Stratifying the location of the DOT by home or at a clinic suggests a possible small effect with home-based DOT (RR 1.10, 95% CI 1.02 to 1.18; 1365 participants, 3 trials). Two small trials of tuberculosis prophylaxis in intravenous drugs users found no statistically significant difference between DOT and self administration (199 participants, 1 trial), or a choice of location for DOT for completion of treatment (108 participants, 1 trial).

Authors' conclusions

The results of randomized controlled trials conducted in low-, middle-, and high-income countries provide no assurance that directly observed therapy compared with self-administered treatment has any quantitatively important effect on cure or treatment completion in people receiving treatment for tuberculosis.

Plain language summary

Plain language summary

Directly observing people taking their tuberculosis drugs did not improve the cure rate compared with people without direct monitoring of treatment

Tuberculosis (TB) is a very serious health problem with two million people dying each year, mostly in low-income countries. Effective drugs for TB have been available since the 1940s, but the problem still abounds. People with TB need to take the drugs for at least six months, but many do not complete their course of treatment. For this reason, services for patients with TB often use different approaches to encourage people to complete their course of treatment. This review found no evidence that direct observation by health workers, family members, or community members, of people taking their medication showed better cure rates that people having self-administered treatment. The intervention is expensive to implement, and there appears to be no sound reason to advocate its routine use until we better understand the situations in which it may be beneficial.

Get access to the full text of this article