and drug-resistant tuberculosis cases,-5 the. Centers .... (Range). HIV-negative with drug-sensitive tuberculosis .... Frieden TR, Sterling T, Pablos-Mendez A, Kil-.
The Influence of Program Acceptability on the Effectiveness of Public Health Policy: A Study of Directly Observed Therapy for Tuberculosis S. Jody Heymann, MD, PhD, Randall Sell, ScD, and Timothy F Brewer; MD, MPH
Introduction In response to the epidemic of overall and drug-resistant tuberculosis cases,-5 the Centers for Disease Control and Prevention and the World Health Organization recommend the widespread or universal use of directly observed therapy.6'7 In directly observed therapy, patients are required to take all medication in front of health care workers or other service providers. Some directly observed therapy programs have demonstrated increased completion rates relative to self-administered therapy8; however, some patients delay seeking treatment to avoid mandatory therapy or detainment.9 Changes in patient behavior in response to public health interventions may offset potential gains or even produce paradoxical increases in disease spread.'0"'1 Effective tuberculosis control depends on patients' both promptly seeking therapy and successfully completing it. In this study, we examined how program acceptability influences the effectiveness of self-administered therapy, directly observed therapy, and confinement in getting patients with active tuberculosis to seek, initiate, and successfully complete therapy. While we examined the influence of program acceptability on effectiveness in the United States, the same methods are applicable elsewhere. These methods are also applicable to examining the impact of program acceptability on the effectiveness of other mandatory public health interventions.
ment. The likelihood that a patient seeking treatment completes it depends on the treatment program used. In the baseline analysis, 60% of patients beginning self-administered therapy successfully completed treatment within 1 year.'2 We conducted a survey of 10 US directly observed therapy programs and found that 67% to 95% (weighted mean: 77%) of patients in such therapy complete treatment. The treatment program offered also affects the number of people seeking care. When programs are less acceptable to patients, fewer patients seek care. The present model provides quantitative answers to the question of how the overall effectiveness of directly observed therapy and confinement is influenced by their acceptability to patients. Values used in the model and sensitivity analyses are described in Table 1.
Results Initial Treatmentfor Patients with Tuberculosis
Methods
If 6% fewer people with active tuberculosis seek care within a year to avoid mandatory directly observed therapy, then requiring directly observed therapy would be less effective than providing self-administered therapy. Sensitivity analyses around directly observed therapy completion rates were conducted. For directly observed therapy programs with a completion rate of 67%, directly observed therapy is less effective than self-administered treatment in
According to the decision tree model used here, a health program decides to offer self-administered therapy or directly observed therapy as first-line therapy (Figure 1). Programs decide whether persons failing self-administered therapy are restarted on self-administered therapy or are required to begin directly observed therapy and whether individuals failing to complete directly observed therapy are continued on that therapy or required to go into confine-
S. Jody Heymann is with the Harvard School of Public Health and Harvard Medical School, Boston, Mass. Randall Sell is with the Columbia School of Medicine, New York City. Timothy F. Brewer is with Harvard Medical School and Brigham and Women's Hospital, Boston. Requests for reprints should be sent to S. Jody Heymann, MD, PhD, Department of Health and Social Behavior, Harvard School of Public Health, 677 Huntington Ave., Boston, MA 02115. This paper was accepted June 13, 1997.
March 1998, Vol. 88, No. 3
Public Health Briefs
Repeat Self-administered Therapo
Initial Therapy Decision
vs
DOT care completed
care completed
Seeks=care Ther
Selt-adminstered
as{ care
RepeatSett-adminsteredThenapy
not completed
seeks
seeks care
s
rotrnmpteted ccnpetedcare cmlte
a
Reeat DOT vs Confinement
cafe oomplet gsk/
careks care
completed
~~~~~~~acane ovods avoId Set-admiostened Therapy care nolcampleted ~ ~ ~ ~ ~ ~ caeseeks carea DO I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~Rpa avoids cane Repeat DOT Begin DOT
sek
T
I~~~~~~~~niWal
DOT
completed
caee
car
DOT vs Confinemert Reoeat seeks car~~~~~~~~~~~~~~~a
c~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a car
caepcmplted
(i
ekscn
care completed
seeks care
DOT Repegi ae
os cane
c arnotcomspleted e
completedned
DOT--_ \ Begin Confinernent {
