A Quantitative Study of Tuberculosis Outcomes in Two Districts

1
A Quantitative Study of Tuberculosis Outcomes in
Two Districts
Mabuza Cesphina, Frank Bonsu, Moalasi Gasekgale,
Kayla Laserson
2
Background

• Data from the Electronic TB Register from a
  Southern African country from 1996-2000
• Data from two districts

3
Objectives

• To examine risk factors associated with overall
  poor treatment outcome, treatment default, and
  transfers

4
Methods Three Studies

• First study
• Poor outcome (default death fail)
  vs.Treatment success (treatment cure treatment
  completion)
• Transfers not included
• Second study
• Default vs. All other outcomes except transfers

5
Methods

• Third study
• Transfer vs. All other outcomes
• Analysis performed by Epi-INFO

6
Results
7
Risk Factors Associated with Overall Poor Outcome
8
Age Distribution Poor Outcome
Chi square 24.20 Plt.05 transfers not included
9
Risk Factors Associated with Default
transfers not included
10
Age Distribution Default
Chi square 45.80 Plt .05 transfers not included
11
Age Distribution Default
transfers not included
12
Risk Factors Associated with Transfer
13
Age Distribution Transfer
Chi square 32.3 Plt 0.05
14
Findings Poor Outcome

• Male patients are more likely to have a poor
  outcome compared to females
• Being in either district is no risk factor for
  having a poor treatment outcome

15
Findings Poor Outcome

• You are less likely to have a poor treatment
  outcome if your pretreatment sputum was positive
• Possible explanations
• Commitment of health workers to the follow-up of
  smear-positive patients
• Smear negative first sputum patients are likely
  to be HIV positive and hence the poor outcome
  (due to death)

16
Findings Poor Outcome

• Patients who have lower sputum conversion rates
  at the end of the initial phase are likely to
  have poor treatment outcomes

17
Findings Default

• Males are more likely to default
• Possible explanations
• Men are employed less likely to go for
  treatment
• Men are more stigmatized about TB-HIV
• We observed a high defaulter rate in the age
  group 0-14, even higher than the overall rate of
  9

18
Findings Transfer

• Gender was not a risk factor for transfer
• You are more likely to transfer out when you are
  in district 3 than in district 15
• We observed a high rate of transfer in the age
  group 0-14, even higher than the overall rate of
  33

19
Limitations

• Missing data (Registered Cases Vs Outcome
  assessment)
• Data not validated
• Data may not represent all transfers defaulters
  in the districts

20
Conclusions

• Epi-Info is a useful tool to analyze TB data
• Statistical methods should be used for
  interpreting TB data to conclude or interpret TB
  epidemiology in your area/ country
• Data can be used for planning purposes and
  improving programs

21
Recommendations

• Data for the two districts need to be validated
• Some of the findings need to be investigated
  further
• For example. The default and transfer rate among
  the 0-14 year olds

22
Acknowledgements

• Data from a Southern African Country
• CENQAM
• Facilitators