Are we making a health impact

1
Acquired TB Multidrug Resistance in California
Travis C. Porco Jennifer M. Flood Peter Oh
California Department of Health
Services Tuberculosis Control Branch 12 May 2006
2
Acquired MDR TB

• Multidrug-resistant tuberculosis (MDR TB) is
  defined as TB resistant to the most important
  first line drugs, isoniazid (INH) and rifampin
• MDR TB requires longer treatment durations and is
  associated with poorer patient outcomes
• How much multi-drug resistance is acquired during
  therapy in California?

3
Monitoring of MDR-TB

• Use the RVCT, Report of Verified Case of
  Tuberculosis, 1995-2003 to ensure data
  completeness.
• Out of 33,452 TB cases, 1,330 had both initial
  and follow-up susceptibility testing results
  reported.
• Twenty-eight cases of MDR have been acquired
  during therapy during 1995-2003 (i.e., not MDR on
  initial sensitivity testing, but MDR on follow-up
  testing).

4
Data limitations

• Incomplete reporting of drug susceptibilities
  (esp. of second line drugs).
• MDR arising from initially culture negative cases
  may be underreported.
• AIDS registry match may possibly fail to identify
  some individuals.
• Individuals whose continuing culture-positivity
  enables follow-up susceptibility testing may not
  be representative.

5
Description

• Mean age, 45.0 years
• Pulmonary major site in all but 1
• 23/28 sputum smear positive (82)
• 10/27 cavitary (37)
• 9/28 AIDS (32)
• 7/28 died (25) (4 with AIDS)
• Origin Mexico, 9 US, 7 Philippines, 4
  Vietnam, 3 Laos, 2 Peru, China, Cambodia, 1.
• Sputum smear-negative, cavitary disease,
  non-AIDS 1/28

6
When
7
Where
North/Sierra 0 Bays 5 Valley
3 Southland 20
Note Southern California contains most of the
people statistically the rate is not
disproportionate.
8
MDR is a bivariate outcome
A44
1084
141
For this analysis, we are interested only in C,
D, and E, because only these correspond to
the new acquisition of MDR TB (CDE28). Numbers
inside circles refer to initial
follow-up susceptibility results (patients with
initial MDR are not considered).
B39
12
9
Crude risk
Initially sensitive 7/21716 3 x 10-4 (1 x 10-4
, 7 x 10-4) Initially INH resistant 17/2184
0.008 (0.005, 0.012) Initially RIF
resistant 4/89 0.045 (0.012,0.11) (95 exact
CI by Clopper-Pearson method)
Denominator consists of all individuals with
initial susceptibility testing and RVCT follow-up
2 reported.
10
Directly observed therapy

• Poor adherence well-known risk factor for drug
  resistance (e.g. Bangsberg, Porco et al. J Infect
  Dis. 2004 Jul 1190(1)162-5 for HIV).
• DOT may be important proxy for adherence, but
• DOT may be both a cause of better adherence, but
  a consequence of (anticipated) poor adherence,
  i.e. DOT may be targeted to those who may need it.

11
Multivariate analysis (1)
Initially INH resistant, RIF sensitive
This table only includes individuals of known
follow-up susceptibility.
12
Multivariate analysis (2)
Multivariate logistic regression All
subjects with known follow-up susceptibility resul
ts
Similar findings are obtained when using all
individuals with initial susceptibility findings
for whom the RVCT follow-up 2 exists.
13
Summary

• Surveillance data suggest that AIDS, prior
  resistance to INH or rifampin, and cavitary
  disease without DOT are predictors of acquisition
  of MDR.
• Other surveillance variables, such as smear
  status, prior TB, alcohol and drug use, were not
  statistically significant predictors of
  acquisition of MDR (but the number of cases is
  small.)

14
Acknowledgments

• Sarah Royce
• Joan Sprinson
• Bill Elms
• Janice Westenhouse
• Jennifer Allen